emerging signaling pathways in tumor biology 2010

Emerging Signaling Pathways in Tumor Biology

2010

Editor

Pedro A. Lazo

IBMCC-Centro de Investigación del Cáncer, CSIC-Universidad de Salamanca, Campus Miguel de Unamuno, E-37007 Salamanca, Spain

Transworld Research Network, T.C. 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India

Published by Transworld Research Network 2010; Rights Reserved Transworld Research Network T.C. 37/661(2), Fort P.O., Trivandrum-695 023, Kerala, India Editor Pedro A. Lazo Managing Editor S.G. Pandalai Publication Manager A. Gayathri Transworld Research Network and the Editor assume no responsibility for the opinions and statements advanced by contributors ISBN: 978-81-7895-477-6

Preface The specificity of biological effects in different cells types is determined by the composition of their proteome, the physical and functional interaction between different signaling pathways and theirs spatial and temporal organization in signaling complexes. Although the implication of some specific pathways with major phenotypes, such as proliferation, apoptosis or differentiation is well known, the final effect varies depending on cell type, which might even be the opposite. These phenotypes play an essential role, and are frequently deregulated in tumor cells, which make some of them potential targets for development of new drugs. In this context, protein kinases represent the main regulatory family, which can be activated and deactivated by reversible phosphorylation, or by integration in spatially organized signaling complexes with scaffold proteins. The human kinome is composed of 518 kinases and for many of them their characterization from the point of view of regulation and substrate specificity, as well as their integration in novel signaling pathways, or their effect on other better know pathways such as MAPK, are not yet understood. In this book we have focused the attention on some of the less well known kinases such as Cot1/Tpl2, LKB1, WNK, GSK3, VRK, and PLK3. Also there is a chapter on development of specific inhibitors for mitotic kinases. There are two additional chapters. One focused on the KSR scaffold protein, an organizer of signaling complexes and intracellular signal compartmentalization, as well as another on DUSP phosphatases, the key deactivator of regulated kinases, which comparatively receive little attention, but are an essential component in regulatory mechanisms.

Pedro A. Lazo

Contents

Chapter 1 Modulation of Ras signaling by the KSR family of molecular scaffolds 1 Luís G. Pérez-Rivas, Stephanie Prieto and José Lozano Chapter 2 The biological function of the proto-oncogene Cot/tpl-2 (MAP kinase kinase kinase 8) 25 Margarita Fernández, Irene Soria, Marta López-Peláez Antonio Martín-Duce and Susana Alemany Chapter 3 WNK kinase signalling in cancer biology 43 Sónia Moniz and Peter Jordan Chapter 4 The tumour suppressor role of LKB1 in human cancer 71 Paola A. Marignani and Montse Sanchez-Cespedes Chapter 5 The role of glycogen synthase kinase-3 in the decision between cell survival and cell death 95 Roberta Venè and Francesca Tosetti Chapter 6 Polo-like kinase 3 in normal and tumor biology 117 Dazhong Xu and Wei Dai

Chapter 7 Vaccinia-related kinase (VRK) signaling in cell and tumor biology 135 Marta Sanz-García, Alberto Valbuena, Inmaculada López-Sánchez Sandra Blanco, Isabel F. Fernández, Marta Vázquez-Cedeira and Pedro A. Lazo

Chapter 8 Targeting mitotic kinases for anti-cancer treatment 157 Hae-ock Lee, Yoo-Kyung Lee, and Hyunsook Lee Chapter 9 Atypical DUSPs: 19 phosphatases in search of a role 185 Yolanda Bayón and Andrés Alonso

Transworld Research Network 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India

Emerging Signaling Pathways in Tumor Biology, 2010: 1-23 ISBN: 978-81-7895-477-6 Editor: Pedro A. Lazo

1. Modulation of Ras signaling by the KSR family of molecular scaffolds

Luís G. Pérez-Rivas, Stephanie Prieto and José Lozano

Dpto. Biología Molecular y Bioquímica, Universidad de Málaga, 29071 Málaga, Spain

Abstract. KSR proteins were initially identified in C. elegans and D. melanogaster as positive regulators of Ras signaling. Murine and human homologs were subsequently isolated, allowing for their characterization at the molecular level. The mammalian homologs are functionally classified as pseudokinases due to the lack of a catalytic lysine involved in the phosphotransfer reaction and found in most active kinases. Over the past 10 years, work done in several groups has allowed a detailed molecular characterization of KSR proteins, mainly at the functional level. In the near future, a more complete picture of the mode of action of KSR proteins will require additional studies at the structural level. However, the wealth of data currently available clearly indicates that KSR is a critical player in Ras signal transduction, functioning as a molecular scaffold that complexes the individual components of the Raf/MEK/ERK cascade into multiprotein signaling platforms (signalosomes). Here, we will review some of the well-established concepts regarding the regulation of Ras signaling by KSR proteins, in both normal and malignant cells. We will also include some recent and provocative data that sheds new light on the mechanism of KSR function.

Correspondence/Reprint request: Dr. José Lozano, Dpto. Biología Molecular y Bioquímica, Universidad de Málaga. Campus de Teatinos s/n. 29071 Málaga, Spain. Phone: (+34) 95 213 6661, E-mail: [email protected]

Luís G. Pérez-Rivas et al. 2

1. Introduction The kinase suppressor of Ras (KSR) family of proteins comprise an evolutionary conserved group of molecular scaffolds that function as modulators of Ras signaling by bringing together the different components of the Raf/MEK/ERK cascade [1, 2]. KSR1, the founding member of the KSR family of proteins, was identified in 1995 by means of genetic screens in D. melanogaster and C. elegans [3-5]. At that time, a murine ksr1 gene was also cloned by Therrien et al. [5]. The corresponding mKSR1 protein has been, since then, the most studied member of the KSR family and currently serves as a model to explain the biological function of these proteins. To date, the family includes one gene in Drosophila (ksr), two genes in mammalian cells (KSR1 and KSR2) and two genes in C. elegans (ksr-1 and ksr-2) [4-8]. All KSR proteins identified so far have a C-terminal sequence which includes the 11 highly conserved subdomains found in most eukaryotic kinases and, therefore, they were initially named as if they were true kinases. However, the failure to find a KSR-specific direct substrate and, more importantly, the lack of an invariant catalytic residue (K) in subdomain II of mammalian KSR isoforms has led to the classification of these proteins in the pseudokinase subgroup of the human kinome, together with other proteins also containing degenerated kinase domains [9]. Currently, KSR is considered to function mainly as a molecular scaffold that facilitates the assembling of the individual components of the Raf/MEK/ERK signaling cascade [10-13]. Mammalian KSR proteins are expressed in most adult tissues, although their levels are usually very low and hard to detect by immunoblotting [10, 14, 15]. The highest expression level has been reported for a brain-specific isoform (B-KSR1) [16]. Titration experiments, showed that, in fact, the in vivo KSR levels need to be carefully regulated and kept below a certain limit in order for KSR to properly function as a molecular scaffold [17]. This also applies to other known scaffolds, such as JNK-interacting protein 1 (JIP1) [18] and the molecular explanation for this requirement is summarized in Figure 1. Essentially, the concentration of a scaffold protein cannot be increased above its optimal level (defined by the stoichiometry of the reaction) to avoid chelation of its ligands into non-competent complexes [1]. For this reason, some of the early works aimed to define the role of KSR1 in Ras signaling resulted in contradictory outcomes, with some investigators reporting a negative role in ERK activation and Ras transformation and others reporting just the opposite [17, 19-24]. KSR1 function was mainly investigated by overexpressing mKSR1 in cultured cells and, therefore, the amount of ectopic mKSR1 was usually way above the endogenous (suboptimal) level, resulting in inhibition of Ras signaling.

KSR proteins in Ras signaling 3

Under circumstances of low transfection efficiency, the ectopically expressed mKSR1 helped to reach the optimal scaffold concentration, enhancing signal transduction. To avoid this problem, which is inherent to the plasmid transfection methods, Morrison and coworkers developed an assay to measure the Ras-dependent ERK scaffolding activity of KSR in Xenopus laevis oocytes. The system allows for a careful titration of the amount of KSR expressed in each oocyte an scores their meiotic maturation (germinal vesicle breakdown, or GVBD) as the biological readout [1, 17]. More recently, the availability of immortalized KSR1-deficient cells has greatly facilitated the study of KSR function since variable amounts of mKSR1 proteins (wild-type or mutant) can be reintroduced in these cells by retroviral infection and subsequent sorting of cell populations with different KSR1 expression levels [25-27]. The following sections are a summary of the current molecular knowledge regarding KSR structure and function.

Figure 1. The effective concentration of KSR1 is critical for an optimal scaffolding activity. This is achieved when the scaffold (KSR1, in red) and its ligands (C-Raf, MEK and ERK, in yellow, green and orange, respectively) are in a concentration near the stoichiometry of the reaction (B). Below such levels, a local increase in the effective concentration of KSR1 results in an increased signaling (A). However, an excess of the scaffold (i.e., a concentration higher than the optimal) results in the formation of non-productive complexes that block signaling (C).


2. Domain organization and function KSR proteins are similar to members of the Raf family of S/T kinases [4-7], and as such are included in the tyrosine-like kinase (TLK) group of kinases [9]. All KSR proteins identified so far have all or some of five distinct domains named Conserved Area (CA) 1 to CA5 (Figure 2) [5]. The CA1 domain consists of about 40 amino acids and is unique to the KSR family of proteins. Although the functional role of this conserved region has not been fully characterized, two reported evidences suggest an important role in mediating Raf activation by means of a direct interaction. Thus, an intact CA1 domain is required for the interaction of Drosophila KSR (D-KSR) with D-Raf [28] while a D-KSR with two specific single amino acid substitutions in CA1 (L50G and R51S) behaves as a weak loss-of-function in Ras signaling [5]. Similarly, mutation of the equivalent sites in the CA1 region of mouse KSR1 (L56G and R57S), impaired binding to B-Raf [13] which was also found to be mediated by the N-terminal half of mKSR1 [29]. Intriguingly,

Figure 2. Domain organization of KSR proteins. With the exception of ceKSR-2A (which lacks CA4), all members of the KSR family contain the CA2-CA5 domains. Two sequences for hKSR1 and hKSR2 have been published lacking the CA1 domain (blue boxes) but analysis of the human genome indicates that those sequences are incomplete. The domain organization of the related S/T kinase C-Raf is included for comparison. The length (number of residues) of each protein is indicated in parenthesis. GenBank accession numbers are also indicated.


all C. elegans KSR isoforms lack the CA1 domain, suggesting that a different region and/or mechanism participates in Raf activation [4, 6]. The CA2 region is a short (13 Aas) proline-rich domain whose function has yet to be determined. Fifty residues comprise the CA3 region, a cysteine-rich domain (CRD), defined by the consensus sequence HX10-12CX2CX11-19-CX2CX4HX2-4CX5-9C and similar to those found in Raf and the ζ isoform of protein kinase C (ζPKC) [5, 30, 31]. The CRDs of KSR, Raf and ζPKC are of the so-called atypical C1 class and, as opossed to the typical C1 domains –which bind diacylglycerol (DAG)- they are mainly involved in protein-protein interactions and membrane localization [32]. The primary sequences of both types of C1 domains are similar, however structural differences seem to account for their different functions. Thus, the CA3 atypical C1 domain of KSR1 is composed of two anti-parallel β-sheets an a C-terminal α helix (Figure 3), similar to typical C1 domains, but lacks four conserved residues in the loop between the β3 and β4 strands (β3-β4 loop) [31]. This deletion, which also occurs in C-Raf, makes the loop more rigid and impairs DAG binding [33]. The issue of lipid binding to the CA3 domain has long been a matter of debate. Work done in the Kolesnick´s group identified KSR1 as their previously reported ceramide-activated protein kinase (CAPK), a 97 kDa membrane-bound, proline-directed and ceramide-enhanced activity [34, 35]. This work raised the possibility that KSR might be a direct target of ceramide [30, 35]. Very recently, the same group have examined the ceramide-binding capabilities of an isolated GST-CA3 domain in an ELISA-based assay and found a marked preference for binding to different ceramides species over other lipids such as diacylglycerol, the ganglioside GM1, sphingomyelin and phosphatidylcholine [36]. Disrupting the tertiary structure of the CA3 domain by mutation of two conserved cysteines (C359S/C362S) reduced, although did not prevent, binding to C16-ceramide [36]. Interestingly, these authors also showed a rapid and transient EGF-dependent colocalization of an overexpressed KSR1 protein with different lipid rafts markers, including ceramide, caveolin-1 and ganglioside GM1 [36]. Although EGF did not induced a ceramide increase, the distribution of KSR1 into the ceramide-rich platforms was reduced by preincubation of the cells with Fumonisin B1 (FB1) a competitive inhibitor of ceramide synthase [36]. Collectively, these results indicate that the ceramide content of the lipid rafts plays a role in the CA3-mediated recruitment of KSR1 to signaling platforms at the plasma membrane.


Figure 3. The CA3 region of KSR is an atypical cysteine-rich C1 domain. (A) Structures of the C1 domain of KSR1 (left) and C-Raf (right) in cartoon format. The β-sheets are colored in yellow while α-helices are in blue. In KSR1, residues C346, C349, C370 (all in green) and H367 (cyan) coordinate one Zn2+ atom while C359, C362, C77 (all in red) and H334 (magenta) coordinate the second Zn2+ atom. (B) The same structures shown in (A) rendered in surface view to highlight the charge distribution (acid = red, basic = blue, hydrophobic = green). Note that, compared with C-Raf, KSR1 has a unique patch of basic residues while lacking a Ras binding site. This basic region is located between chains β4 and β5 and mediates binding to CK2. The hydrophobic pocket is potentially involved in interactions with membrane lipids. Direct ceramide binding to ζPKC or C-Raf has been reported by several groups [37-39] however, to date only one group has reported direct in vitro [14C16]-ceramide binding to an isolated, bacterially-expressed CA3 domain [40] and most other investigators have failed to observe neither binding nor a direct activation of KSR1 by ceramide [20, 41, 42]. It has been suggested that an additional positively charged residue in the β1-β2 loop might contribute to an overall structure competent for ceramide binding since the atypical C1 domains of the ceramide-binding proteins C-Raf and ζPKC contains such a residue and KSR1 does not [31]. In summary, a direct evidence of lipid binding


to the CA3 region of KSR still remains to be unambiguously demonstrated. On the contrary its role as a multifunctional domain involved in protein-protein interactions and subcelullar localization has been firmly stablished. The CA3 region is neccessary for the interaction of KSR1 with 14-3-3 proteins [17, 43-45], the βγ subunits of heterotrimeric G-protein [46], and the Cdc25C-associated kinase 1 (C-TAK1) [44, 47] and, most importantly, an intact CA3 domain is required for KSR1 to modulate Ras signaling at the plasma membrane [20, 31]. Thus, a KSR1 mutant with a disrupted CA3 domain (CRM mutant) is unable to accelerate Ras-induced maturation in Xenopus oocytes and to localize KSR1 to the membrane in response to Ras activation by oncogenic mutations or growth factor treatment [17, 20, 31]. CA3 could be mediating the membrane targeting of KSR1 through association to an unidentified membrane lipid or by means of interaction with a membrane-bound protein. In this regard, there is no evidence of KSR binding to Ras but an interaction of the membrane-associated βγ subunits of G proteins with the CA3 domain has been reported [2, 46]. The CA3 domain of KSR1 can serve specific functions that cannot be replaced by other atypical C1 domains. Thus, a chimeric KSR1 protein in which its own C1 domain has been replaced with that of C-Raf (KRK mutant) did not change its subcelullar localization in NIH 3T3 cells after PDGF treatment [31]. However, we an others have demonstrated that the KRK mutant binds 14-3-3 as efficiently as the wild-type protein [31, 45], suggesting that a general C1 domain structure can mediate interaction with the cytosolic anchors that keep KSR1 in the cytosol while Ras-dependent translocation to the plasma membrane is strictly dependent on specific features of its CA3 domain. In marked contrast, the C1 domain of C-Raf contributes to its membrane association but it is the Ras-binding domain (RBD) that is required for the translocation [48, 49]. CA4 is a serine/threonine-rich domain of variable length in D-KSR, mKSR1 and ceKSR1 (61, 53 and 47 residues, respectively) while it is absent in ceKSR2 [5, 6]. This region contains a motif known as docking site for ERK, FxFP (DEF) which mediates the interaction of ERK with KSR to get phosphorylated on proline-directed S/T sites [13, 50]. As expected for such a critical function, DEF motifs are conserved in KSR proteins from different species, including ceKSR (FLFP), D-KSR (FNFP) and mKSR1/2 (FSFP) [50]. In mKSR1 the stretch of S/T residues is located in the N-terminal region of CA4 and includes one ERK phosphorylation site (S443) while the DEF motif is positioned in the C-terminal side of CA4 [5, 17]. The co-existence of S/T-rich stretches and a DEF motif in CA4 makes it resemble the C boxes of ETS proteins in the Elk subfamily [50]. Mutant KSR proteins lacking a functional DEF motif are poorly phosphorylated by ERK both in vitro and in vivo and this defect is due to a reduced affinity of the KSR1 mutants for


activated ERK [13, 50]. Thus, ERK phosphorylation assays using isolated CA4 domains as substrates showed that inactivating mutations in the conserved FxFP docking site of KSR proteins dramatically increased the Km (7-50-fold) of the reactions [50], indicating that a functional DEF motif is necessary for a productive KSR/ERK interaction. While the CA1-4 regions fill the N-terminal half of KSR, the CA5 region expands over the C-terminal half and contains some features of a kinase domain [2-5]. The CA5 domain is closely related to the kinase domain of Raf proteins [5] and contains the eleven conserved subdomains found in all kinases [9, 51]. However, the issue of the kinase activity of KSR1 has been a matter of dispute. While some investigators reported that KSR1 can function as a Raf kinase [35, 52-54] others have been unable to reproduce neither the KSR1-dependent Raf phosphorylation nor any kinase activity at all [19-21, 23, 24, 55]. In support of a non-catalytic function of mammalian KSR is the lack of a highly conserved lysine in kinase subdomain II of CA5 [5]. This invariant residue is found in the vast majority of kinases identified so far and participates in the positioning of the ATP molecule and in the phosphotransfer reaction [51]. Curiously, this conserved residue is present in D-KSR (K705) and ceKSR1 (K503) while in murine and human KSR1 is substituted by an arginine (R589 in mKSR1 and R571 in hKSR2) [5, 7]. A second ksr gene was identified in C. elegans (ksr-2) coding for two proteíns isoforms generated by alternative splicing (ceKSR-2A and ceKSR-2B) [6]. Similar to the mammalian isoforms, these two proteins have an arginine instead of the catalytic lysine [6], suggesting that in C. elegans KSR may serve both catalytic and non-catalytic functions. Even if KSR was proven to have kinase activity, it is not clear whether KSR is a tyrosine or a S/T kinase as the sequence YI(L)APE in subdomain VIII, which is conserved among S/T kinases, is present in all KSR genes cloned so far, while C. elegans and D. melanogaster KSR also contains the sequence HKDLR indicative of tyrosine kinases [5]. More recently, the full sequencing of the human kinome (the complete set of 518 human kinases) has revealed that KSR belongs to a small subgroup of kinases that differ significantly with respect to the key residues that are necessary for catalysis [9]. Members of this subgroup, which represents 10% of the kinome, are called pseudokinases and besides KSR1-2, also include vaccinia-related kinase 3 (VRK3), Janus kinases 1-3 (JAK1-3), integrin-linked kinase (ILK) and ErbB3, among others [9, 56, 57]. The pseudokinases are predicted to be enzimatically inactive due to sequence changes in essential motifs. However they are conserved throughout metazoans, suggesting that their unusual kinase domain still serve important functions [58]. The recently solved first 3D structures of pseudokinases VRK3 [57] and ROP2 [59] have provided evidence for a regulatory role of the their inactive


kinase domain. Thus, despite having the invariant lysine in the appropiate position, VRK3 has a highly degraded catalytic site with bulky residues that hinder ATP binding and prevent any kinase activity [57]. ROP2 has a more favorable open conformation and could accommodate nucleotide binding, however several unusual side chains lining the cavity prevent ATP binding [59]. Similar to KSR1, the S/T kinase WNK1 lacks the catalytic lysine and was initially considered to be a pseudokinase. Despite that, WNK1 can display kinase activity phosphorylating myelin basic protein and itself, due to a different lysine in kinase subdomain I that provides the catalytic function [60, 61]. Therefore, in addition to its enzymatic funtion, kinases have clearly evolved to provide docking sites for specific protein-protein interactions. 3. KSR-binding proteins As true scaffolds, KSR proteins do not function alone and are tipically bound to other signaling proteins. Early work from different groups showed that KSR1 associates with all three components of the signaling module Raf/MEK/ERK. Thus, MEK1/2 binds constitutively to the CA5 region of KSR1 while the interaction with Raf and ERK1/2 is dependent on Ras activation and KSR1 translocation to the plasma membrane [17, 21, 24, 43]. The functional relevance of such interactions was later investigated by Therrien and co-workers, demonstrating that KSR functions as a scaffold linking MEK to its activator Raf [28, 62]. The first evidence that KSR proteins participate in signaling complexes came from work done by Stewart et al. [55] showing that KSR1 co-immunoprecipitates with a group of 10 different proteins, including MEK1/2, 14-3-3 and the heat-shock proteins HSP90, HSP70, HSP68 and p50CDC37, among others (Figure 4). The role of the heat-shock family of proteins in KSR regulation is still uncertain but they may stabilize KSR molecules or the whole complex, since treatment of cells with geldanamycin, a known inhibitor of HSP90, reduces the KSR half-life [55].

Figure 4. KSR-interacting proteins. The figure shows the name of some of the most relevant and well-studied KSR partners. They are positioned close to their mapped interaction sites.


By using gel-filtration chromatography, Stewart et al. demonstrated that KSR1 exists in mammalian cells as part of multiprotein signaling complexes with apparent molecular masses ranging from 250 to 1000 kDa [10, 55]. Interestingly, MEK1/2 was incorporated into such complexes in a manner strictly dependent on its interaction with KSR1, since overexpression of a KSR1-C809Y mutant, defective in MEK1/2 binding displaced MEK1/2 from the high-molecular fractions to the 40-50 kDa fractions, indicative of a monomeric status [55]. The KSR/MEK interaction is constitutive and does not seem to change after growth factor stimulation or Ras activation. KSR1 has also been reported to bind the βγ subunits of heterotrimeric G proteins in response to lysophosphatidic acid stimulation [46]. Interestingly, the cysteine-rich domain of KSR1 is both required and sufficient for the KSR1/βγ interaction, since an isolated CA3 domain is able to bind both subunits [46]. These results provide an alternative mechanism for the CA3-mediated translocation of KSR proteins to the plasma membrane in the absence of lipid binding: CA3 would localize KSR by interacting with other membrane-bound proteins, such as the G proteins. The KSR/14-3-3 interaction is critical for the regulation of KSR function since it helps to sequester KSR proteins in the cytosolic compartment in the absence of Ras activation [17, 44]. 14-3-3 proteins bind to their targets at specific phosphoserine (pS) residues contained in two types of consensus motifs: RSXpSXP (mode 1) and RXXXpSXP (mode 2) [63, 64]. In particular, mKSR1 associates with 14-3-3 through mode 1 sequences RSKpS297HE and RTEpS392VP, which flank the CA3 domain [43, 44]. Our group have recently demonstrated the existence of functional specificity among 14-3-3 isoforms (β, γ, ε, η, σ, τ and ζ in mammals) by showing that KSR1 interacts preferentially with 14-3-3γ, regulating its ability to translocate to the plasma membrane and facilitating Ras-induced ERK2 activation [45]. Further, we have provided data demonstrating that the flexible C-terminal tail of 14-3-3γ is required for a full and specific interaction with KSR1 and that 14-3-3γ heterodimerizes preferentially with selected isoforms when bound to KSR1 [45]. KSR1 also interacts with regulatory enzymes such as the S/T kinases C-TAK1 [44] and CK2 [65] and the S/T phosphatase PP2A [66, 67]. The coordinated action of these three enzymes facilitate the activation of mKSR1 as a scaffold for the RAF/MEK/ERK cascade (see below). The KSR1 residues that are relevant for the specific interaction with these regulatory proteins have been identified by the Morrison group following a mutagenesis approach. Thus, the stable interaction of KSR1 with C-TAK1 is mediated by residues 377-424, which include the phosphoacceptor site S392 and the characteristic motifs PxxR/K and ΦPK (Φ being a hydrophobic residue),


located C-terminal to the S392 [44, 47]. In particular I397, a residue located at the +5 position relative to the S392 phosphorylation site is absolutely required for the KSR1/C-TAK1 interaction [47]. The interaction of KSR1 with the S/T phosphatase PP2A involves residues 249-320, which include the CA2 domain [66]. PP2A is a heterotrimeric holoenzyme composed of a structural A subunit, a regulatory B subunit and a catalytic C subunit. While subunits A and C associate constitutively with KSR1, binding to the B subunit is induced by growth factor stimulation, resulting in dephosphorylation of S392 and KSR1 translocation to the plasma membrane [66]. Interestingly, in addition to PP2A, KSR2 binds specifically to the Ca2+/calmodulin-regulated phosphatase calcineurin [8]. The interaction is constitutive and mediated by a region located between CA2 and CA3 that is unique to KSR2 and contains the calcineurin binding motif LSVP (residues 390-393) [8]. Casein kinase 2 (CK2) is a heterotetrameric kinase composed by two regulatory (β) and two catalytic (α/α´) subunits [65]. The interaction of CK2 with KSR1 is constitutive and the region involved in the binding has been mapped to KSR1 residues 320-424 which includes the CA3 domain [65]. Swapping the CA3 domain with the C1 region of C-Raf abolishes binding to CK2, indicating that residues specific to the cysteine-rich domain of KSR participate in the interaction. In particular, a basic surface region in the β1 sheet of the C1 domain is unique to KSR1 and mutation of K360 and R363 to A (KR/AA mutant) impairs binding to CK2 [65]. The scaffolding activity of KSR1 requires its interaction with CK2 since the KR/AA mutant is also impaired in its ability to facilitate activation of the associated MEK and ERK proteins [65]. This deficiency is due to the fact that the KSR1-associated CK2 activity functions as a Raf N-region kinase, phosphorylating residues S338 in C-Raf and S446 and S430 in B-Raf [65]. Thus, for Raf proteins to be activated, KSR1 has to be competent in CK2 binding. Activated KSR1 translocates to the plasma membrane where it associates with a Raf homodimer which then becomes phosphorylated by the pool of CK2 molecules associated to KSR1 [65]. KSR1 also interacts with the ubiquitin E3 ligase IMP (Impedes Mitogenic Signal Propagation), a novel Ras effector that functions by limiting the formation of RAF/MEK complexes and inhibiting ERK activation [68]. KSR1 and IMP bind to each other via their N-terminal halves and this interaction provokes the hyperphosphorylation of KSR1 and its partitioning to a detergent (NP-40)-insoluble subcellular fraction [68]. It is unclear how IMP induces such changes in KSR1 but it has been demonstrated that IMP inhibits KSR1 homo-oligomerization without affecting its ability to interact with MEK [69].


Mouse KSR1 interacts with the F-actin binding protein leukocyte-specific protein 1 (LSP1) and, as a result, the KSR/MEK/ERK complex gets targeted to peripheral actin filaments. The Rap guanine nucleotide exchange factor C3G has also been reported to bind KSR1 in complex with PP2A, MEK and ERK at the subcortical actin cytoskeleton, where it could contribute to ERK signaling by increasing the phosphatase activity of PP2A [70]. A comprehensive set of KSR-associated proteins (the KSR interactome) has recently been reported by Morrison and coworkers [8]. Using mass spectrometry, they found >70 different proteins associated with recombinant mouse KSR1 and KSR2 isoforms immunopurified from overexpressing 293T cells. Importantly, they could discriminate between common and isoform-specific KSR partners. Thus, Raf, MEK, ERK, 14-3-3, PP2A, C-TAK1 and CK2 bound to both KSR1 and KSR2 while Regulatory-associated protein of mTOR (Raptor) and the S/T phosphatase calcineurin (PP2B) bound specifically to KSR1 and KSR2, respectively [8]. Following the same functional proteomics strategy, Liu et al. have identified a set of 100 proteins specifically bound to overexpressed human KSR2, many of which are different from those identified by the Morrison group [12]. They also identified a group of 43 proteins, including A-Raf, p38 MAPK, PI3K and the IκB kinase complex-associated protein (IKAP), that were recruited to a KSR2 signaling complex in response to TNF-α treatment [12]. This, together with the fact that the authors did not detect binding to either B-Raf nor C-Raf would suggest specific roles for KSR2 in A-Raf and p38 MAPK signaling. However, Morrison and coworkers did detect B-Raf in association to both KSR1 and KSR2 in their study [8]. In addition to using KSR2 proteins from different species (human and mouse) the construct used by Liu et al. lacked the first two coding exons of human KSR2 [8], which might account for the discrepancies observed between the two groups. It is not likely that all KSR-associated proteins are interacting directly with KSR in vivo. As part of multiprotein complexes, some of those proteins could associate with KSR through their specific interaction with other direct KSR-binding proteins. Alternatively, they could interact directly with different pools of KSR molecules defined by its phosphorylation status or subcellular localization. At least MEK1, ERK2, 14-3-3 and the γ subunit of the G-protein appear to interact directly based on studies using the yeast two-hybrid system or peptide competition [24, 43, 45, 46]. In any case, both studies indicate that, similar to KSR1, the KSR2 isoform also functions as a molecular scaffold in MAPK pathways.


4. KSR kinases and phosphatases In resting cells, KSR1 is phosphorylated on several S and T residues and growth factor stimulation induces dephosphorylation of specific sites as well as phosphorylation of new sites. Figure 5 shows the twelve KSR1 phosphorylation sites reported so far and their known kinases. It is worth noting, however, that residues S190, S256, S320, S429 and S434 were identified initially by Lewis and coworkers in KSR mutants lacking at least 5 major sites [71]. Therefore, they could be artifactual or represent in vivo phosphorylation events that function concomitantly with dephosphorylation of other sites. In favor of the later option, S320 has recently been shown to get phosphorylated by ERK after EGF treatment [13] and S434 has been reported to be phosphorylated by the Nm23-H1 metastasis suppressor only when phosphorylation was prevented on S392 [72] or S429 [71]. A mutant KSR1 protein, lacking de pseudokinase domain produced the same tryptic phosphopeptides than the wild-type version, indicating that none of the reported phosphosites were generated by autophosphorylation [71]. In resting cells, KSR2 is also constitutively phosphorylated on seven residues: S198, T287, S310, S469, S485, T492 and S623 [8].

Figure 5. KSR1 kinases and their phosphorylation sites. Residues S392 and S518 are phosphorylated constitutively by C-TAK1 and CK2, respectively. S297 is also constitutively phosphorylated by an unknown kinase. Ras activation induces phosphorylation of T260, T274, S320 and S443 by ERK and a concomitant dephosphorylation of S392 by PP2A. The Nm23-H1 metastasis suppressor phosphorylates S434 and also S392, although less efficiently than C-TAK1. The reported phosphorylation of KSR1 on residues S190, T256, T411 and S429 still awaits a detailed functional characterization.


The identification and functional characterization of the major KSR1 and KSR2 phosphorylation sites have been reported by the Morrison group in a series of landmark papers [8, 17, 44, 65, 66]. Thus, phosphorylation on S297 by an unknown kinase and on S392 by C-TAK1 creates two mode 1 motifs that allow binding to 14-3-3 and cytoplasmic retention of KSR1 [44, 45]. The major phosphorylation site S392 can also get phosphorylated by Nm23-H1 but it is unclear how this event impacts KSR1 function [72]. Breast cancer cells expressing high levels of Nm23-H1 show an enhanced degradation of KSR1, likely due to its increased binding to the Hsp90 chaperone [73]. However, this effect seems to be independent of the kinase activity of Nm23-H1 and relay exclusively on its expression level [73]. In murine KSR2, residues S310 and S469 mediate binding to 14-3-3 proteins and are constitutively phosphorylated [65]. Although the possibility exists that C-TAK1 phosphorylates one of these sites (KSR2 also binds C-TAK1), it remains to be determined. Casein kinase 2 (CK2) accounts for the constitutive phosphorylation of KSR1 on S518 [65, 71]. As mentioned above, binding to CK2 is essential for the scaffolding activity however, phosphorylation at S518 is not required for that function since a KSR1-S518A mutant is as efficient as the wild-type in two different assays measuring ERK activation [65]. The functional significance of S518 phosphorylation remains to be determined. The scaffolding activity of KSR contributes to the Ras-dependent activation of ERK which, in turn, phosphorylates KSR1 on residues T260, T274, S320 and S443 [13, 17]. Thus, phosphorylation of these sites is minimal in resting cells and increases after mitogenic stimulation. Until recently, however, there was no known biological consequence of such event. Work done by McKay et al. indicates that, once activated, ERK phosphorylates both KSR1 and B-Raf to promote their dissociation and release of KSR1 from the plasma membrane [74]. These authors observed that a KSR1 mutant with a defective docking site for ERK had an increased and prolonged interaction with B-Raf after mitogenic stimulation, an effect that was mimicked by preincubation of the cells with a MEK inhibitor [74]. Since that mutant had reduced levels of phosphorylated proline-directed S/T (pS/TP) sites, the most plausible explanation to their data is that feedback phosphorylation of KSR1 by ERK impairs binding to B-Raf. This was further supported by experiments showing that KSR1 and B-Raf proteins with mutated ERK phosphorylation sites interacted strongly even in the absence of growth factor stimulation [74]. Several KSR phosphatases have been identified that counteract the effect of KSR kinases. In particular, PP2A plays a critical role in the activation of KSR1 as a molecular scaffold for ERK. Thus, the phosphorylation status of


S392 is regulated by the coordinated action of the kinase C-TAK1 and the phosphatase PP2A, which dephosphorylates this major 14-3-3 binding site in response to growth factor stimulation and Ras activation [66, 67]. KSR2 also associates with PP2A and, similar to KSR1, gets dephosphorylated on a 14-3-3 binding site (S469) after EGF stimulation [8]. This event is blocked by okadaic acid (OA) but not by cyclosporin A (CsA), suggesting that PP2A is the phosphatase responsible for pS469 dephosphorylation [8]. On the contrary, the phosphatase calcineurin binds selectively to KSR2 and dephosphorylates residues S198, T287 and S310 in response to Ca2+ signaling [8]. As a control, KSR1 phosphorylation was not altered after elevation of Ca2+ intracellular levels by PMA/ionomycin treatment, further indicating a specific role of KSR2 in Ca2+ signaling. 5. Role of KSR in normal Ras signaling The phosphorylation status of KSR proteins greatly affects their subcellular localization and, more particularly, their transport to the plasma membrane and the nucleus [8, 13, 44, 66, 68, 72, 75, 76]. In fact, KSR1 translocation to the plasma membrane is a crucial step during Ras-dependent ERK activation. The current model to explain the scaffolding activity of KSR1 is depicted in Figure 6. A KSR1 multiprotein complex exists in resting cells which includes, among others, MEK, 14-3-3 and subunits A and C of the heterotrimeric PP2A phosphatase (Figure 6A). Ras activation by growth factors allows reconstitution of a PP2A activity by recruitment of the regulatory subunit B to the complex [66]. As a result, the 14-3-3 binding site pS392 gets dephosphorylated by PP2A, KSR1 partially dissociates from 14-3-3 and the CA3 domain gets exposed (Figure 6B). Exposure of the cysteine-rich domain allows translocation of KSR1 to the plasma membrane, presumably by interaction with lipids [36]. Simultaneously, Raf kinases also get dephosphorylated by PP2A on 14-3-3 binding sites (S259 in C-Raf) and recruited to the membrane by active, GTP-loaded Ras (Figure 6B) [66]. The KSR1-associated CK2 activity then function as a Raf N-region kinase phosphorylating B-Raf on residue S446 and C-Raf on S338 (Figure 6C) [65]. A negative charge at position +3 from the phosphorylation site is required for the Raf N-region kinase activity of CK2 [65]. Therefore, to get phosphorylated by CK2, C-Raf needs to be previously phosphorylated on Y341 by c-Src, whereas the presence of a negatively-charged aspartic acid residue at position 449 (D449) overcomes that requirement in B-Raf [65]. Raf proteins are able to form homo- and hetero-oligomers, with B-Raf/C-Raf dimers having a greater specific activity than the corresponding homodimers or monomers [77]. The molecular


Figure 6. Mechanism of KSR1-mediated activation of ERK. (A) In resting cells, KSR1 is kept in an “inactive” state by 14-3-3 interaction and IMP-mediated partitioning to a detergent-insoluble fraction. Raf proteins are also inactivated by forming a complex with 14-3-3. (B) Ras activation disrupts the KSR1/IMP complex by recruiting IMP and promoting its autoubiquitination. In addition, the regulatory B subunit of PP2A associates with the catalytic core (subunits A and C), previously bound to KSR1 and Raf, dephosphorylating a 14-3-3 binding site (pS392 in KSR1). KSR1 then exposes its membrane-targeting C1 domain and a docking site for ERK. Dephosphorylation of Raf by PP2A, among other events, contributes to the formation of B-Raf/C-Raf heterodimers. Side-to-side dimerization and 14-3-3 bridging are likely involved in the formation of B-Raf/C-Raf complexes. (C) The KSR1-associated CK2 enzyme contributes to Raf activation by functioning as a Raf N-region kinase. (D) Activated Raf phosphorylate KSR1-bound MEK, which in turn phosphorylate and activate ERK. Recent data indicate that activated ERK phosphorylate KSR1 and B-Raf proteins on several feedback S/TP sites which results in KSR1 dissociation from B-Raf and attenuation of ERK signaling. determinants involved in the formation of Raf oligomers are beginning to be uncovered. Binding of 14-3-3 proteins to a C-terminal site in both B-Raf (pS729) and C-Raf (pS621) bridges the two monomers in B-Raf/C-Raf heterodimers [77, 78]. Surprisingly, C-Raf can be fully activated by heterodimerization with a kinase-inactive B-Raf mutant, indicating that B-Raf acts allosterically on C-Raf [78]. A very recent study has provided an explanation for this puzzling evidence. Thus, Therrien and coworkers have described the activation of Raf by a new and specific mode of dimerization of its kinase domain, which they call side-to-side dimerization [79]. Interestingly, the dimerization surface is conserved in the Raf-related pseudokinase domain (CA5) of KSR, which could allow side-to-side heterodimerization of both proteins and explain the intrinsic Raf


activating potential of mamalian KSR in the absence of catalytic activity [79]. Whether or not the associated CK2 activity contributes to the KSR scaffolding activity in different cellular contexts needs to be clarified since depletion of CK2 levels by RNA interference in Drosophila S2 cells had no impact on the side-to-side activation of Raf by D-KSR [79]. KSR1 can also form homodimers through interactions in the CA5 region [69] which could be mediated by side-to-side dimerization. The ubiquitin E3 ligase IMP blocks KSR1 homodimerization and B-Raf/C-Raf heterodimerization, resulting in inhibition of MEK activation and signaling through the ERK cascade [29, 68, 69]. This effect seems to be a consequence of the KSR1/IMP complex partitioning to a detergent-insoluble cellular compartment, where KSR1 is kept “inactive” by hyperphosphorylation on S392 and 14-3-3 binding [68]. Active, GTP-bound Ras prompts dissociation of the KSR1/IMP complex by a direct interaction with IMP and stimulation of its auto-polyubiquitination activity (Figure 6B) [68]. Therefore, active Ras contributes to the scaffolding activity of KSR by a dual mechanism involving the potentiation of positive signals (dephosphorylation of KSR and Raf proteins) and blockade of the negative ones (IMP interaction) [29, 68, 69]. A functional CA1 domain and MEK binding are required for KSR1 to associate with B-Raf and form a ternary complex competent for downstream signal transduction [13]. Once activated, Raf kinases phosphorylate and activate KSR-associated MEK which, in turn phosphorylate and activate ERK (Figure 6D). As mentioned above, activated ERK phosphorylates KSR1 and B-Raf on several feedback S/TP residues to foster dissociation of the KSR1/B-Raf complex, release of KSR1 from the plasma membrane and attenuation of signaling through the pathway [13]. In addition to modulating the scaffolding activity of KSR1 at the plasma membrane, phosphorylation has also been reported to regulate the nucleocytoplasmatic transport of KSR1 [75]. In contrast to wild-type KSR1, which shuttles continuously between the nucleus and the cytoplam, a T274V/S392A double mutant is constitutively located in the nucleus [75]. Thus, the evidence suggests that 14-3-3 release, in the absence of ERK phosphorylation mediates the nuclear translocation of KSR1. The molecular mechanism involved in this process awaits its characterization, although proteomic analysis of immunopurified KSR has revealed several members of the importin family as part of the multiprotein complex [8]. 6. Role of KSR in oncogenic Ras signaling Since KSR was initially identified in C. elegans and D. Melanogaster by the isolation of mutants able to revert the phenotype of transgenic animals


expressing an hyperactive (“oncogenic”) form of Ras, it has always been considered an attractive target to investigate therapeutic strategies against Ras-dependent tumors [3-5, 80]. Initial efforts were aimed at the molecular characterization of KSR proteins and elucidation of the mechanism used to transmit Ras-dependent signals [17, 19, 21-24, 43, 55, 81]. However, it was not until the development of ksr1 knockout mice that an in vivo evidence for its role in Ras-mediated tumorigenesis was obtained. Two KSR1-deficient mice were generated by work done independently in the groups of Shaw [10] and Kolesnick [15]. In both cases, ksr1-/- mice were viable and developed normally, indicating that either KSR1 is dispensable for development or, more likely, KSR2 compensates for KSR1 deficiency. However, cultured primary embrionic fibroblasts (MEFs) and T cells obtained from KSR1-deficient mice showed some important limitations, such as the exclusion of MEK and ERK from high molecular weight signaling complexes and, as a consequence, impaired MEK/ERK activation and reduced proliferation [10, 15]. Further, ksr1-/- MEFs transduced with retroviral versions of the c-myc and Ras oncogenes were not transformed and did not form colonies in soft agar [15]. To address the in vivo relevance of KSR1 in Ras tumorigenesis, Lozano et al. bred ksr1-/- mice to hemizygous Tg.AC mice containing an oncogenic form of the ν-HA-ras transgene in the skin [82]. This animal model has been widely used to study two-stage skin carcinogenesis. The ν-HA-ras oncogene is transcriptionally silent in latent neoplastic cells until induced by topical treatment with the tumor promoter 2-O-tetradecanoylphorbol 13-acetate (TPA), wich induces papilloma formation [82]. Notably, while most Tg.AC mice (70%) in a ksr1+/+ background (Tg.AC/ksr1+/+) developed papillomas after repeated TPA treatment, only 10% of the Tg.AC/ksr1-/- mice displayed papillomas [15], strongly indicating a critical role for KSR1 in Ras-mediated tumor formation. In the work by Nguyen et al. their ksr1-/- mice were crossed with mice transgenic for the polyomavirus middle-T antigen (MT+), which develop breast tumors soon after birth [83]. MT functions like a constitutively active receptor at the membrane by recruiting several signaling molecules, including c-Src, PI3K and the adaptor protein ShcA. Tyrosine phosphorylation of ShcA creates binding sites for Grb2/Sos1 which initiates Ras/ERK signaling [84]. Therefore, breast tumors produced in MT+ mice can be considered Ras-dependent [10]. Although they also developed tumors, the onset of tumor formation in MT+ ksr1-/- mice was almost twice slower than in MT+ ksr1+/- indicating that KSR1 deficency also impairs Ras-dependent tumorigenesis in this knockout model [10]. The different response observed in the two KSR1-deficient mice after Ras activation (blockage in Tg.AC/ksr1-/- mice vs delayed tumor progression in MT+ ksr1-/- mice) might be related to the fact that MT


mammary tumors signals through wild-type Ras and PI3K [83, 85] while tumors arising in the Tg.AC/ksr1-/- are dependent on oncogenic Ras and ERK signaling [86]. In spite of this small difference, it is worth noting that both knockdown models are refractory to Ras-mediated tumorigenesis. Xing et al. have provided proof-of-concept data for the consideration of KSR1 as a bona fide therapeutic target in the treatment of Ras-dependent tumors [87, 88]. Thus, A431 epidermoid carcinoma cells stably transfected with an antisense KSR1 construct to reduce KSR1 levels by 60%, did not form tumors when implanted subcutaneously in nude mice (the control, vector-transfected A431 cells were highly tumorogenic) [87]. In addition, delivery of a KSR1-specific synthetic phosphorothioate antisense oligonucleotide (AS-ODN-1) to human PANC-1 pancreatic cancer cells, reduced KSR1 levels by 90% and inhibited proliferation, invasion and transformation. More importantly, continuous infusion of AS-ODN-1 via implanted osmotic pumps in mice xenografted with PANC-1 tumors attenuated or completely abolished (depending on the dose) tumor growth without affecting signaling upstream of Ras [87]. When human A549 non-small-cell lung tumors were xenografted instead of PANC-1, infusion of AS-ODN-1 not only inhibited tumor growth but also reduced the number of metastatic foci in the lungs by 60-80% [87]. The tumorigenic activity of A431 cells is driven by EGFR overexpression and Ras hyperactivation while both PANC-1 and A549 express an oncogenic K-Ras protein (mutations G12D and G12S, respectively) [89, 90]. Therefore, the above experiments highlight the importance of KSR proteins in Ras-dependent tumor formation and further support its evaluation as a therapeutic target in neoplasias with a high incidence of oncogenic Ras mutations, such as pancreatic, gastrointestinal (GI) tract and lung carcinomas. Interestingly, KSR1 has also been suggested to be a potential prognostic marker in oncology since evaluation of a panel of cancer cell lines expressing different levels of KSR1 indicates that KSR1 expression inversely correlates with anticancer drug sensitivity [91]. Thus, cancer cells with higher levels of KSR1 were more resistant to cytotoxic drugs while those with lower KSR1 levels showed enhanced sensitivity and increased ERK activation [91]. References 1. Ritt, D.A., I.O. Daar, and D.K. Morrison. 2006, Methods Enzymol. 407, 224. 2. Claperon, A. and M. Therrien. 2007, Oncogene. 26, 3143. 3. Kornfeld, K., D.B. Hom, and H.R. Horvitz. 1995, Cell. 83, 903. 4. Sundaram, M. and M. Han. 1995, Cell. 83, 889. 5. Therrien, M., H.C. Chang, N.M. Solomon, F.D. Karim, D.A. Wassarman, and

G.M. Rubin. 1995, Cell. 83, 879.


6. Ohmachi, M., C.E. Rocheleau, D. Church, E. Lambie, T. Schedl, and M.V. Sundaram. 2002, Curr Biol. 12, 427.

7. Channavajhala, P.L., L. Wu, J.W. Cuozzo, J.P. Hall, W. Liu, L.L. Lin, and Y. Zhang. 2003, J Biol Chem. 278, 47089.

8. Dougherty, M.K., D.A. Ritt, M. Zhou, S.I. Specht, D.M. Monson, T.D. Veenstra, and D.K. Morrison. 2009, Mol Cell. 34, 652.

9. Manning, G., D.B. Whyte, R. Martinez, T. Hunter, and S. Sudarsanam. 2002, Science. 298, 1912.

10. Nguyen, A., W.R. Burack, J.L. Stock, R. Kortum, O.V. Chaika, M. Afkarian, W.J. Muller, K.M. Murphy, D.K. Morrison, R.E. Lewis, J. McNeish, and A.S. Shaw. 2002, Mol Cell Biol. 22, 3035.

11. Brown, M.D. and D.B. Sacks. 2009, Cell Signal. 21, 462. 12. Liu, L., P.L. Channavajhala, V.R. Rao, I. Moutsatsos, L. Wu, Y. Zhang, L.L. Lin,

and Y. Qiu. 2009, Biochim Biophys Acta. 13. McKay, M.M., D.A. Ritt, and D.K. Morrison. 2009, Proc Natl Acad Sci U S A.

106, 11022. 14. Giblett, S.M., D.J. Lloyd, Y. Light, R. Marais, and C.A. Pritchard. 2002, Cell

Growth Differ. 13, 307. 15. Lozano, J., R. Xing, Z. Cai, H.L. Jensen, C. Trempus, W. Mark, R. Cannon, and

R. Kolesnick. 2003, Cancer Res. 63, 4232. 16. Muller, J., A.M. Cacace, W.E. Lyons, C.B. McGill, and D.K. Morrison. 2000,

Mol Cell Biol. 20, 5529. 17. Cacace, A.M., N.R. Michaud, M. Therrien, K. Mathes, T. Copeland, G.M. Rubin,

and D.K. Morrison. 1999, Mol Cell Biol. 19, 229. 18. Yasuda, J., A.J. Whitmarsh, J. Cavanagh, M. Sharma, and R.J. Davis. 1999, Mol

Cell Biol. 19, 7245. 19. Therrien, M., N.R. Michaud, G.M. Rubin, and D.K. Morrison. 1996, Genes Dev.

10, 2684. 20. Michaud, N.R., M. Therrien, A. Cacace, L.C. Edsall, S. Spiegel, G.M. Rubin, and

D.K. Morrison. 1997, Proc Natl Acad Sci U S A. 94, 12792. 21. Denouel-Galy, A., E.M. Douville, P.H. Warne, C. Papin, D. Laugier, G. Calothy,

J. Downward, and A. Eychene. 1998, Curr Biol. 8, 46. 22. Joneson, T., J.A. Fulton, D.J. Volle, O.V. Chaika, D. Bar-Sagi, and R.E. Lewis.

1998, J Biol Chem. 273, 7743. 23. Sugimoto, T., S. Stewart, M. Han, and K.L. Guan. 1998, Embo J. 17, 1717. 24. Yu, W., W.J. Fantl, G. Harrowe, and L.T. Williams. 1998, Curr Biol. 8, 56. 25. Kortum, R.L., H.J. Johnson, D.L. Costanzo, D.J. Volle, G.L. Razidlo, A.M.

Fusello, A.S. Shaw, and R.E. Lewis. 2006, Mol Cell Biol. 26, 2202. 26. Razidlo, G.L., R.L. Kortum, J.L. Haferbier, and R.E. Lewis. 2004, J Biol Chem.

279, 47808. 27. Kortum, R.L. and R.E. Lewis. 2004, Mol Cell Biol. 24, 4407. 28. Roy, F., G. Laberge, M. Douziech, D. Ferland-McCollough, and M. Therrien.

2002, Genes Dev. 16, 427. 29. Matheny, S.A. and M.A. White. 2009, J Biol Chem. 284, 11007. 30. van Blitterswijk, W.J. 1998, Biochem J. 331 ( Pt 2), 679.


31. Zhou, M., D.A. Horita, D.S. Waugh, R.A. Byrd, and D.K. Morrison. 2002, J Mol Biol. 315, 435.

32. Colon-Gonzalez, F. and M.G. Kazanietz. 2006, Biochim Biophys Acta. 1761, 827. 33. Hurley, J.H., A.C. Newton, P.J. Parker, P.M. Blumberg, and Y. Nishizuka. 1997,

Protein Sci. 6, 477. 34. Yao, B., Y. Zhang, S. Delikat, S. Mathias, S. Basu, and R. Kolesnick. 1995,

Nature. 378, 307. 35. Zhang, Y., B. Yao, S. Delikat, S. Bayoumy, X.H. Lin, S. Basu, M. McGinley,

P.Y. Chan-Hui, H. Lichenstein, and R. Kolesnick. 1997, Cell. 89, 63. 36. Yin, X., M. Zafrullah, H. Lee, A. Haimovitz-Friedman, Z. Fuks, and R.

Kolesnick. 2009, Cell Physiol Biochem. 24, 219. 37. Muller, G., P. Storz, S. Bourteele, H. Doppler, K. Pfizenmaier, H. Mischak, A.

Philipp, C. Kaiser, and W. Kolch. 1998, EMBO J. 17, 732. 38. Huwiler, A., J. Brunner, R. Hummel, M. Vervoordeldonk, S. Stabel, H. van den

Bosch, and J. Pfeilschifter. 1996, Proc Natl Acad Sci U S A. 93, 6959. 39. Muller, G., M. Ayoub, P. Storz, J. Rennecke, D. Fabbro, and K. Pfizenmaier.

1995, EMBO J. 14, 1961. 40. Grassme, H., H. Schwarz, and E. Gulbins. 2001, Biochem Biophys Res Commun.

284, 1016. 41. Galve-Roperh, I., C. Sanchez, M.L. Cortes, T. Gomez del Pulgar, M. Izquierdo,

and M. Guzman. 2000, Nat Med. 6, 313. 42. Zhou, Y.J., M. Chen, N.A. Cusack, L.H. Kimmel, K.S. Magnuson, J.G. Boyd, W.

Lin, J.L. Roberts, A. Lengi, R.H. Buckley, R.L. Geahlen, F. Candotti, M. Gadina, P.S. Changelian, and J.J. O'Shea. 2001, Mol Cell. 8, 959.

43. Xing, H., K. Kornfeld, and A.J. Muslin. 1997, Curr Biol. 7, 294. 44. Muller, J., S. Ory, T. Copeland, H. Piwnica-Worms, and D.K. Morrison. 2001,

Mol Cell. 8, 983. 45. Jagemann, L.R., L.G. Perez-Rivas, E.J. Ruiz, J.A. Ranea, F. Sanchez-Jimenez,

A.R. Nebreda, E. Alba, and J. Lozano. 2008, J Biol Chem. 283, 17450. 46. Bell, B., H. Xing, K. Yan, N. Gautam, and A.J. Muslin. 1999, J Biol Chem.

274, 7982. 47. Muller, J., D.A. Ritt, T.D. Copeland, and D.K. Morrison. 2003, Embo J. 22, 4431. 48. Luo, Z., B. Diaz, M.S. Marshall, and J. Avruch. 1997, Mol Cell Biol. 17, 46. 49. Roy, S., A. Lane, J. Yan, R. McPherson, and J.F. Hancock. 1997, J Biol Chem.

272, 20139. 50. Jacobs, D., D. Glossip, H. Xing, A.J. Muslin, and K. Kornfeld. 1999, Genes Dev.

13, 163. 51. Hanks, S.K., A.M. Quinn, and T. Hunter. 1988, Science. 241, 42. 52. Xing, H.R. and R. Kolesnick. 2001, J Biol Chem. 276, 9733. 53. Xing, H.R., J. Lozano, and R. Kolesnick. 2000, J Biol Chem. 275, 17276. 54. Yan, F. and D.B. Polk. 2001, Cancer Res. 61, 963. 55. Stewart, S., M. Sundaram, Y. Zhang, J. Lee, M. Han, and K.L. Guan. 1999, Mol

Cell Biol. 19, 5523. 56. Kannan, N. and S.S. Taylor. 2008, Cell. 133, 204.


57. Scheeff, E.D., J. Eswaran, G. Bunkoczi, S. Knapp, and G. Manning. 2009, Structure. 17, 128.

58. Manning, G., G.D. Plowman, T. Hunter, and S. Sudarsanam. 2002, Trends Biochem Sci. 27, 514.

59. Labesse, G., M. Gelin, Y. Bessin, M. Lebrun, J. Papoin, R. Cerdan, S.T. Arold, and J.F. Dubremetz. 2009, Structure. 17, 139.

60. Xu, B., J.M. English, J.L. Wilsbacher, S. Stippec, E.J. Goldsmith, and M.H. Cobb. 2000, J Biol Chem. 275, 16795.

61. Xu, B.E., X. Min, S. Stippec, B.H. Lee, E.J. Goldsmith, and M.H. Cobb. 2002, J Biol Chem. 277, 48456.

62. Roy, F. and M. Therrien. 2002, Curr Biol. 12, R325. 63. Yaffe, M.B., K. Rittinger, S. Volinia, P.R. Caron, A. Aitken, H. Leffers, S.J.

Gamblin, S.J. Smerdon, and L.C. Cantley. 1997, Cell. 91, 961. 64. Yang, X., W.H. Lee, F. Sobott, E. Papagrigoriou, C.V. Robinson, J.G.

Grossmann, M. Sundstrom, D.A. Doyle, and J.M. Elkins. 2006, Proc Natl Acad Sci U S A. 103, 17237.

65. Ritt, D.A., M. Zhou, T.P. Conrads, T.D. Veenstra, T.D. Copeland, and D.K. Morrison. 2007, Curr Biol. 17, 179.

66. Ory, S., M. Zhou, T.P. Conrads, T.D. Veenstra, and D.K. Morrison. 2003, Curr Biol. 13, 1356.

67. Raabe, T. and U.R. Rapp. 2003, Curr Biol. 13, R635. 68. Matheny, S.A., C. Chen, R.L. Kortum, G.L. Razidlo, R.E. Lewis, and M.A.

White. 2004, Nature. 427, 256. 69. Chen, C., R.E. Lewis, and M.A. White. 2008, J Biol Chem. 283, 12789. 70. Martin-Encabo, S., E. Santos, and C. Guerrero. 2007, Exp Cell Res. 313, 3881. 71. Volle, D.J., J.A. Fulton, O.V. Chaika, K. McDermott, H. Huang, L.A. Steinke,

and R.E. Lewis. 1999, Biochemistry. 38, 5130. 72. Hartsough, M.T., D.K. Morrison, M. Salerno, D. Palmieri, T. Ouatas, M. Mair, J.

Patrick, and P.S. Steeg. 2002, J Biol Chem. 277, 32389. 73. Salerno, M., D. Palmieri, A. Bouadis, D. Halverson, and P.S. Steeg. 2005, Mol

Cell Biol. 25, 1379. 74. McKay, M.M. and D.K. Morrison. 2007, J Biol Chem. 282, 26225. 75. Brennan, J.A., D.J. Volle, O.V. Chaika, and R.E. Lewis. 2002, J Biol Chem.

277, 5369. 76. Robertson, S.E., S.R. Setty, A. Sitaram, M.S. Marks, R.E. Lewis, and M.M.

Chou. 2006, Mol Biol Cell. 17, 645. 77. Rushworth, L.K., A.D. Hindley, E. O'Neill, and W. Kolch. 2006, Mol Cell Biol.

26, 2262. 78. Garnett, M.J., S. Rana, H. Paterson, D. Barford, and R. Marais. 2005, Mol Cell.

20, 963. 79. Rajakulendran, T., M. Sahmi, M. Lefrancois, F. Sicheri, and M. Therrien. 2009,

Nature. 80. Downward, J. 1995, Cell. 83, 831. 81. Basu, S., S. Bayoumy, Y. Zhang, J. Lozano, and R. Kolesnick. 1998, J Biol

Chem. 273, 30419.


82. Leder, A., A. Kuo, R.D. Cardiff, E. Sinn, and P. Leder. 1990, Proc Natl Acad Sci U S A. 87, 9178.

83. Guy, C.T., R.D. Cardiff, and W.J. Muller. 1992, Mol Cell Biol. 12, 954. 84. Ichaso, N. and S.M. Dilworth. 2001, Oncogene. 20, 7908. 85. Dilworth, S.M., C.E. Brewster, M.D. Jones, L. Lanfrancone, G. Pelicci, and P.G.

Pelicci. 1994, Nature. 367, 87. 86. Hamilton, M. and A. Wolfman. 1998, J Biol Chem. 273, 28155. 87. Xing, H.R., C. Cordon-Cardo, X. Deng, W. Tong, L. Campodonico, Z. Fuks, and

R. Kolesnick. 2003, Nat Med. 9, 1266. 88. Zhang, J., M. Zafrullah, X. Yang, X. Yin, Z. Zhang, Z. Fuks, and R. Kolesnick.

2008, Cancer Biol Ther. 7, 1490. 89. Almoguera, C., D. Shibata, K. Forrester, J. Martin, N. Arnheim, and M. Perucho.

1988, Cell. 53, 549. 90. Valenzuela, D.M. and J. Groffen. 1986, Nucleic Acids Res. 14, 843. 91. Stoeger, S.M. and K.H. Cowan. 2009, Cancer Chemother Pharmacol. 63, 807.



2. The biological function of the proto-oncogene Cot/tpl-2

(MAP kinase kinase kinase 8)

Margarita Fernández, Irene Soria, Marta López-Peláez Antonio Martín-Duce and Susana Alemany

Instituto de Investigaciones Biomédicas, CSIC, Fac. Medicina UAM. Arturo Duperíer 4, 28029 Madrid, Spain

Abstract. Cot, as well as its murine homologue tpl-2, was discovered in a COOH–terminal truncated form that unmasks the transformation capacity of the protein. The COOH-terminal domain of wt Cot contains an amino acid sequence that is a recognition signal for degradation via proteasome, besides, this domain of wt Cot is also an autoinhibitory domain of the specific activity of the wild type form. These data explain the transformation capacity of trunc-Cot/tpl-2, that when overexpressed is capable of activating several MAP kinases pathways as well as AP-1, NFAT, and NF-κB2 transcriptional activities. Earlier sobreexpression experiments lead to the proposal that Cot/tpl-2 could be involved in proliferative signalling, but the use of new technologies such as genetically modifies mice and interference RNA end up with the already accepted hypothesis that Cot/tpl-2 is involved in immune innate and adaptive processes. Cot/tpl-2 is activated in response to the activation of the TLR/IL-1 receptor superfamily as well as in response to the activation of some receptors of the TNF family. Independently of the cell system

Correspondence/Reprint request: Dr. Susana Alemany, Instituto de Investigaciones Biomédicas, CSIC, Arturo Duperier 4, 28029, Madrid, Spain. E-mail: [email protected]

Margarita Fernández et al. 26

it is accepted that in resting cells Cot/tpl-2 forms a stable and inactive complex with p105 NF-κB among other proteins to protect it from degradation, adequate TLR/IL-1R stimulation induces the activation of the IKK complex that targets p105 NF-κB to be rapidly degraded by the proteasome pathway to p50 NF-κB, a subunit of the NF-κB transcription factor. Consequently Cot/tpl-2 is released from the complex and susceptible to transduce the activatory signal, leading to the activation of the MEK1-Erk1/Erk2 pathway. However, actually it is not completely understood all the requests that Cot/tpl-2 needs to be fully active and to this end it is also accepted that Cot/tpl2 requires to be phosphorylated. In addition the possibility that the requirements vary from cell system to cell system cannot be excluded. Physiologically, Cot/tpl-2 is involved in provoking innate immunity to establish adaptive immunity. In fact it is the unique MAP3K that activates Erk1/Erk2 when the TLRs/IL-1 receptors are activated and mediates the production of pro-inflammatory cytokines, such as TNFα, IL-1, or IL-6. More recently it has been shown that Cot/tpl-2 has the capacity to regulate the balance between Th1 and Th2 cytokines. All these data indicate that, although mutations in Cot gene result in the expression of a protein linked with cell malignancies, physiologically wt Cot/tpl-2 is involved in innate and adaptive immunity.

Characterization of Cot/tpl-2 protein Cot/tpl-2 is a MAP kinase kinase kinase (MAP3K8) capable of activating the MEK1-Erk1/Erk2 pathway. The human Cot gene was identified in a 3´ rearrangement form that leads to the expression of a truncated/modified protein. The normal homologue has an open reading frame encoding 467 amino acids. The first 397 are identical to the truncated form (trunc-Cot), while the 69 amino acids from the COOH-terminal of wt Cot are replaced by 18 amino acids without sequence homology in the truncated form (1). The murine homologue of the human Cot gene is the tpl-2 gene. The tpl-2 gene was also identified in a similar truncated/modified form (2). The COOH-terminal domains of both wt Cot and wt tpl-2 are almost identical. The only difference between the last 77 COOH-terminal amino acids of the human wt Cot and the murine wt tpl-2 are the two conservative substitutions of the residues M437 for V and K439 for R. However, there is no similarity in the amino acids inserted in either trunc-Cot or trunc-tpl-2. These truncations/modifications unmask the transforming capacity of this MAP kinase kinase kinase, a property of the trunc-Cot/tpl-2 protein (Fig. 1). Nevertheless, it has been shown that overexpression of the proto-oncogenic form is also capable of conferring a transformed phenotype to established cell lines (3-5).

The biological function of the proto-oncogene Cot/tpl-2 (MAP kinase kinase kinase 8 27

The COOH-terminal domain of wt Cot contains the EMLKRQRSLYIDLGALAGYFNL amino acid sequence that is a recognition signal for degradation via proteasome. This degron confers instability to other unrelated proteins and accounts for the 2.6-fold lower t1/2 of wt Cot compared (35 min), to trunc-Cot (90 min). On the other hand, the deletion of the 44 amino acids of the COOH-terminal of wt Cot increases its specific activity in 3.8 fold, measured as MEK1-Erk1/Erk2 activation, indicating that the COOH-terminal domain of wt Cot is also an autoinhibitory domain of the specific activity of the wild type form (5) (Fig. 1). In fact, the COOH-terminal of tpl-2 is capable of inhibiting the kinase activity of trunc-tpl-2 “in vitro” (2). Moreover, phosphorylation of the residue S400, which is contained in the carboxi-terminal of the wt form but not in the truncated/modified form, by extracellular stimuli, increases the specific activity of the protein (6). In conclusion, all these data indicate that deletion of the carboxi-terminal region of the murine and human wt forms increases their kinase activity through two independent mechanisms that leads to an increase in the t1/2 and an increase in the specific activity of the kinase, both potentiate each other in such a manner that equal amounts of trunc- and wt Cot mRNA results in a 10-fold increase in the activity of trunc-Cot.

P P P

DSSCTGSTEESEMLKRQRSLYIDLGALAGYFN VRGPPTLEY

Kinase activity inhibitory domain

4543

4642

inhibitory domain destabilizing region

kinase domain

COOH COOH M29

NH2NH2 M1

21

MSYGSDNKEEIDLLIKHLNVSDVIDdestabilizing region

B

Cell transformation

A Innate

inmunity

Figure 1. Schematic representation of Cot protein kinase.PartA compares the structures of the wt Cot (above) and trunc-Cot (below) proteins. The last 69 amino-acids of the carboxi-terminal region from the wt protein have been substituted by 18 unrelated amino-acids. The alternative translation start methionine (M29) is also indicated. Part B shows the different domains of the wt Cot protein. The amino-acid sequence of the destabilizing region located at the N terminus is shown. The C-terminus sequence include both a destabilizing region and a kinase activity inhibitory domain. The main phosphorylation sites (residues S62, T290, and S400) are indicated.


Two alternative translation initiation sites M(1) and M(29) has been described for wt Cot/tpl-2 (7). M(29) wt Cot does not contain the first 29 amino acids of M(1) wt Cot and exhibits a three -fold higher t1/2 than M(1) wt Cot (7). This data suggests that the N-terminal of wt Cot contains another signal that triggers the protein to degradation. In fact antibodies against the C-terminal of Cot wt recognized two bands in Western blot analysis, being the one with higher mass more rapidly degraded (8-10), although the sequence of amino acids in the N-terminal of the protein involved in this process remains to be established. Mutations in Cot/or tpl-2 gene are associated with tumorigenesis Increased Cot/tpl-2 activity has been observed in different transformed cell types. In fact, different modifications in Cot/tpl-2 have been proposed to be responsible for cellular transformation (Table1). Cot was identified in a truncated/modified form that occurred as a consequence of a rearrangement in the 3´ terminal region of human Cot gene during transfection of the genomic DNA of a human carcinoma cell line into hamster SHOK cells; this rearrangement induced the transformation of the murine cells (1). The insertion of the Moloney Leukaemia Provirus into tpl-2 gene also induced a rearrangement in the 3´terminal region of tpl-2 gene. This tpl-2 gene reorganization plays an essential role in the progression of T lymphomas (11, 12). Similarly, the insertion of the Mouse Mammary Tumour Virus (MMTV) in the last intron of tpl-2 gene is associated with transformation

Table 1

Type of mutation Cell line/ tissue Reference 3´ rearrangement of Cot/tpl-2 gene

Hamster SHOK cells. In vitro transfection of human genomic DNA

Miyoshi (1991)

3´ rearrangement of Cot/tpl-2 gene

Murine T cell. Insertion of the Moloney Leukemia provirus

Patriotis (1993)

3´ rearrangement of tpl-2 gene

Mammary gland. Insertion of Murine Mammary provirus

Erny (1996)

3´ rearrangement of Cot gene

Primary human adenocarcinoma Clark (2004)

Genomic locus amplification

Human breast cancer cells Sourvinos (1999)

Postraductional modifications

Infection of the human T cell leukemia virus type-I

Babu (2006)

Unknown Human large granular lymphocyte proliferative disorder

Christoforidou (2004)

Unknown Epstein-Barr virus infection Eliopoulos (2002)


of mammary gland cells (13). More recently, in some human adenocarcinoma primary tumors a modification in the 3´ region of Cot gene has been confirmed (14). The deletion of the 3´ region of the Cot and tpl-2 genes provoke not only the expression of a truncated/modified protein, as explained above, but also the deletion of the 3´ untranslated region of both genes that harbor three copies of the RNA destabilizing sequence AUUUA. This, results in the higher expression levels of their corresponding mRNAs (12, 13). On the other hand, the human Cot genomic locus, without modifications in the coding sequence of Cot gene, is amplified in some human breast cancers (15). Based on these data it has been proposed that Cot could be a prognostic factor in breast cancer development (16). Moreover, a correlation between high Cot mRNA levels and human large granular lymphocyte proliferative disorder has been also proposed (17). Besides the infection of the human T cell leukemia virus type-I (HTLV-I) induces high steady state expression levels of the M(29) wt Cot, that as previously mentioned, exhibit a high stability, resulting in high levels of Cot activity (18). Cot is also commonly highly expressed in Epstein Barr virus -associated malignancies, such as nasopharyngeal carcinoma and gastric cancer (19). All these data, indicate that increased Cot/tpl-2 activity plays a role in cell transformation, besides more recently it has been also proposed that Cot could also participate in cell invasion processes (20, 21). However in contrast to the established role of Cot/tpl-2 as an oncogene, it has been also reported that wt Cot/tpl-2 can function as a tumour suppressor gene, since CD8+ T cell lymphomas are developed in old Cot/tpl-2 KO mice (22). In these cells the Cot/tpl-2 downregulation decreases the expression of the T cell co-receptor CTL4, which negatively regulate de signals transducer through the T cell receptor (TCR). Regulation of Cot/tpl-2 activity by extracellular signals Cot and tpl-2 genes were identified in an oncogenic form; thereby for many years, it was expected, that the proto-onco Cot/tpl-2 protein could be involved in the intracellular proliferative signals pathways. However, the fact that the tpl-2 knock-out mice was resistant to LPS/D-Galactosamine-induced pathology, due to the low production of TNFα among other cytokines, indicated that, physiologically, endogenous Cot/tpl-2 would be involve in innate and adaptive immunity rather in proliferative signals (23). Now it is accepted that Cot/tpl-2 is activated in response to the activation of the TLR/IL-1 receptor superfamily as well as in response to the activation of some receptors of the TNF family (24-26) and that Cot/tpl-2 plays a unique role in inflammatory processes. The TLRs, included in the TLR/IL-1R


superfamily, contain an extracellular leucine-rich repeat region and are sensors of the infection by pathogens since they recognized the different pathogen-associated molecular patterns (PAMPS): bacterial lipopeptides (TLRs1, 2, and 6) lipopolysaccharide (TLR4), flagellin (TLR5) and nucleic acids (TLRs3,7,8, and 9). The IL-1 receptor subfamily, with an Igg like extracellular region, includes receptors for IL-1 (IL-1R), IL-18 (IL-18R, and IL-33 (T1/ST2 R). However, despite their different extracellular structure, both TLRs and IL-1Rs contain in their intracellular region the TIR domain and activate similar intracellular pathways including the recruitment to the receptor of some adaptors like MyD88, the activation of the kinases IRAK, and the ubiquitination of the TRAF proteins, resulting in the activation of the IKK complex (reviewed in 27-29). Independently of the cell system, it is accepted that in resting cells, Cot/tpl-2 forms a stable and inactive complex with p105 NF-κB among other proteins to protect it from degradation. The interaction between p105 NF-κB and wt Cot/tpl-2 was first described as a result of a yeast two-hybrid screen (30) and was further corroborated, by the fact that p105 NF-κB knock-out macrophages, show normal tpl-2 mRNA levels but tpl-2 protein, could be only detected when cells are treated with proteasome inhibitors (8,9). Studies performed with overexpressed p105 NF-κB and wt Cot/tpl-2, have indicated that Cot/tpl-2 protein interacts with p105 NF-κB trough two different domains (31). One interaction takes place through the COOH-domain of wt Cot/tpl-2, where a proteasome degradation signal is located. This interaction explains why the interaction between p105 NF-κB and Cot/tpl-2 protects Cot/tpl-2 from degradation. Another interaction between the kinase domain of Cot/tpl-2 and p105 NF-κB has been also proposed (31). This interaction does not allow the substrates of Cot/tpl-2, like MEK1, to interact with the kinase and to be subsequently phosphorylated (32). Besides, cotransfection experiments have also demonstrated that ABIN2 (A20-binding inhibitor of NF-κB2) also interacts with wt Cot/tpl-2 in addition to p105 NF-κB (33) (Fig. 2). This interaction has been also further confirmed in knock-out ABIN2 macrophages where, as occurred in p105 NF-κB knockout macrophages, endogenous wt Cot/tpl-2 protein is neither detected due to its rapid degradation. But, in contrast to p105 NF-κB, the interaction of ABIN2 with Cot/tpl-2 does not block its kinase activity (34, 35). All these data indicate that both p105 NF-κB and ABIN2 are necessary to form a complex with wt Cot/tpl-2 to protect the kinase from the proteasome degradation. Besides ABIN2 and p105 NF-κB it has been also reported that a member of the kinase suppressor family of RAS (KSR2) interacts with Cot/tpl-2 when both


Figure 2. Mechanisms of Cot/tpl-2 activation. The figure shows a schematic representation of the mechanisms involved in the activation of Cot/tpl-2 by extracellular stimuli (TLRs/IL-1Rs). A) Cot protein status in basal conditions. In unstimulated cells Cot/tpl-2 protein is forming a complex with at least two proteins (p105 - NF−κB and ABIN-2), this association maintains Cot/tpl-2 inactive and stable. B) Upon ligand binding, activated IKKβ phosphorylates p105 tagging it for proteosoma degradation. Other signals, not clearly understood yet, allow Cot phosphorylation in different residues C) Cot is then released from its interaction with p105 and phosphorylated Cot/tpl-2 activates Erk1/2 pathway which is subsequently rapidly degraded by the proteasome. proteins are overexpressed. The co-expression of KSR2 with wt Cot/tpl-2 significantly and specifically reduces Erk1/Erk2 and NF-κB activation by Cot/tpl-2 (36); however the physiological role of this interaction remains to be established. Further studies have to be performed to determine if other proteins are part of the p105 NF-κB-ABIN2-Cot/tpl-2 complex (37). Adequate TLR/IL-1R stimulation induces the activation of the IKK complex; active IKKβ kinase phosphorylates p105 NF-κB in different residues (38-40). These phosphorylations target p105 NF-κB to be rapidly degraded by the proteasome pathway to p50 NF-κB, a subunit of the NF-κB transcription factor (38-41). Consequently Cot/tpl-2 is released from the complex and susceptible to transduce the activatory signal (8, 9, 31, 42). Once Cot/tpl-2 is dissociated from the complex is rapidly degraded just upon activation, by phosphorylation, of MEK1 (8,9). However, actually it is not completely understood all the requisites that Cot/tpl-2 needs to be fully active and the possibility that the requirements vary from cell system to cell system can not be excluded. In one hand, overexpression of TRAF6, an adaptor that mediates the activation of the IKK complex, is sufficient to induce Cot/tpl-2


activation in MEF cells (43), but not in HeLa cells (10). Overexpression of TRAF6 in HeLa cells induces IκBα degradation, indicating that TRAF6 overexpression induces the activation of the IKK complex. However, the overexpression of this adaptor is not sufficient to induce Cot/tpl-2 activation, although TRAF6 is required to activate Cot in response to IL-1 stimulation (10). Similarly, in macrophages other TRAF protein, TRAF2, participates in the intracellular signalling to activate Cot/tpl-2 in response to TNFα but as in the case of HeLa it is not sufficient to induce the activation of the kinase (44). All together these data provide evidence that at least two intracellular signals are required to induce Cot activation (Fig. 2). One of which, is the dissociation of Cot from the inactive complex, being this intracellular pathway TRAF dependent, but others signal(s) is (are) required to fully activate the kinase. In this context it has been reported that upon cellular stimulation Cot/tpl-2 becomes phosphorylated in different residues, but the exact role of each of these phosphorylations remains to be elucidated (Fig. 1). It has been reported by several groups that in response to extracellular stimuli like LPS, TNFα or IL-1, Cot/tpl-2 is phosphorylated in the activation loop of the kinase, on residue T290 (45-47). This T290 phosphorylation, seems also to play a role in the dissociation of Cot/tpl-2 protein from the NF−κB p105 complex and it is essential for the activation of Erk1/Erk2 by Cot/tpl-2 (45, 48). Once Cot/tpl-2 is phosphorylated on residue T290, it phosphorylates itself in S62, this autophosphorylation enhances the capacity of Cot/tpl-2 to activate the Erk1/Erk2 pathway (45). On the other hand, a phosphorylation on the S400 of Cot/tpl-2 by LPS stimulation, is also required for a full Cot/tpl-2 activation (6). AKT is capable of phosphorylating this residue “in vitro” (49), but does not plays any role in the phosphorylation of this residue “in vivo” and the identity of this kinase is still unknown (6). All these data indicate that in general to activate physiologically Cot/tpl-2 two requirements are necessary, dissociation of the inactive complex and an adequate phosphorylation state of the protein. The specific kinases that phosphorylate both residues T290 and S400 on Cot/tpl-2 protein remain to be identified. However, it has been reported, that in response to IL-1 and TNFα stimulation Cot/tpl-2 activation requires the activity of a tyrosine kinase of the SRC family (10, 44), thereby it can not be excluded the possibility that a SRC tyrosine kinase activity, mediates the phosphorylation on residues T290 or S400 or both. By using immobilized metal affinity chromatography (IMAC) and a linear trap mass spectrometer analysis, other phosphorylation sites on Cot/tpl-2 proteins have been also identified, but the role of these phosphorylations on Cot/tpl-2 protein are still unknown (50).


More recently, it has been proposed that Cot/tpl-2 also mediates the activation of Erk1/Erk2 in response to the activation of the proteinase activating factor 1 (PAR-1) (21) and the adiponectin receptor (51). These two receptors contain in their structure the classical seven transmembrane domains, but not the TIR domain. Thereby, it is unlikely that these receptors could activate Cot/tpl-2 through the Myd88 adaptor. All these data open the possibility that Cot/tpl-2 take also part in other type of intracellular signaling pathways. Signal pathways activated by Cot/tpl-2 Both overexpressed wt Cot/tpl-2 and trunc Cot/tpl-2 are capable of increasing the activity of several MAP kinases pathways. Overexpression of wt or trunc-Cot/tpl-2 induce an increase in the MAP kinases pathways Erk1/Erk2, JNK (52), p38γ, and Erk5 (4). Cot/tpl-2, partially as a consequence of the up-regulation of the different MAP kinases pathways, induces the activation of several transcription factors, such as AP-1, NFAT and NF-κB. Overexpression of Cot/tpl-2 induces AP-1 activity in different cell systems (53, 54). This activation is probably mediated by the capacity of Cot to stimulate c-Jun and c-Fos activities (4, 55). In this context, it has been proposed that the ability of Cot/tpl-2 to induce cellular transformation is dependent on its ability to activate these two transcription factors (4, 55). Overexpressed Cot/tpl-2 also up-regulates the NFAT transcriptional activity (54). In fact, a physical interaction among both overexpressed Cot/tpl-2 and NFATc2 has been also demonstrated (56). In one hand, Cot/tpl-2 potentiates NFAT-induced transactivation (57), on the other one it has been also reported that Cot/tpl-2 induces an NFAT accumulation in the nucleus (58, 59). Since the NFAT transcription factors are regulated at two different levels, firstly at the level of subcellular localization and secondly

at the level of the intrinsic DNA binding activity, it remains to be established, whether the reported accumulation of NFAT by Cot/tpl-2 in the nucleus, could be a consequence of the increased transactivation of NFAT triggered by the kinase, resulting in higher nuclear retention, or Cot/tpl-2 could also have an active role in the modulation of the NFAT in/out shuttling of the nucleus. Different groups have also shown that overexpressed Cot/tpl-2 induces the activation of the NF-κB, although the mechanism it still unclear. It has been claim that Cot/tpl-2 induces an increase in p65 NF-κB transcription factor activity by phosphorylating the residues S276 (60), S311(61), or S536


(62), or both S468 and 536 (63). Several groups have stated that the NF-κB activation by Cot/tpl-2 is through the activation of the IKK complex (64-70), due to the capacity of Cot/tpl-2 to activate NIK (63, 65), a kinase that mediates the activation of the IKK complex (71). Besides, it has been also proposed that Cot/tpl-2 is required to induce p105 NF-κB degradation (19, 30) by phosphorylating p105 NF-κB in a different site than the IKK complex (32). The cis-elements of these transcription factors modulated by Cot/tpl-2 overexpression (AP-1, NFAT, and NF-κB) are conserved in many promoter regions, thereby, it should be expected that Cot/tpl-2 overexpression, would induce the transcription of many different genes (54, 57, 58, 72). However, more recent studies performed in cells; where the Cot/tpl-2 protein is knock-out suggest that in a physiological context Cot/tpl-2 may be more specific than when overexpressed. This is in part due to the fact that Cot/tpl-2 share substrates with other kinases. Indeed, overexpression of the Cot/tpl-2 dead kinase mutant sequesters MEK-1, that physiologically is also phosphorylated by other kinases, such as c-RAF. Involvement of Cot innate and adaptive immunity The physiological activation of Cot/tpl-2 has been extensively studied in macrophages after LPS stimulation. Cot/tpl-2 is the unique MAP3K that activates Erk1/Erk2 in response to the activation of TLR4 by LPS in macrophages (9, 23, 73, 74) (Fig. 2) without modulating the activation of p38, JNK, Erk5, and NFκB (23). In TNFα- or LPS-stimulated macrophages, in CD-40- stimulated-macrophages or B cells, as well as in IL-1-stimulated HeLa cells, Cot/tpl-2 does not modulated JNK and p38 activation (10, 23, 43). However in tpl-2-/- MEF cells not only Erk1/Erk2, but also JNK, and NF-κB activation are defective when stimulated with TNF whereas only Erk1/Erk2 activation is impaired when these cells are stimulated with IL-1. Thereby, it has been proposed that the signalling role of endogenous Cot/tpl-2 is stimulus and cell type specific (60). The ablation of Cot/tpl-2 expression in LPS-stimulated macrophages partially blocks TNFα (23) and IL-10 secretion (24), and reduces the expression of COX-2 (74). Cot/tpl-2 activation is also required to induce Erk1/Erk2 activation when bone marrow derived macrophages are activated with peptidoglycan, double-stranded RNA or loxoribine, that are ligands for TLR2, TLR3, and TLR7 respectively (24, 26). Cot/tpl-2 is also the sole MAP3K that activates Erk1/Erk2 pathway in B cells in response to anti-CD40 and participates in the secretion IgE (43). In carcinoma Hela cells


Cot/tpl-2- MEK1-Erk1/Erk2 pathway is required to produce IL-8 and MIP-1β in response to IL-1 stimulation (10). On the other hand, and although Cot/tpl-2 is involved in the intracellular signalling generated by the activation of TLR9 by CpG-DNA, it is not clear whether other MAP3K, besides Cot/tpl-2, upregulates the Erk1/Erk2 pathway in response to CpG-DNA stimulation (26, 75, 76). The role of Cot in inflammation is not only confirmed by experiments performed in intact cells due to its capacity to modulate the production of cytokines/chemokines involved in this process, but it also has been assessed “in vivo”. In fact Cot/tpl-2 KO mice have a reduction in caerulein-induced pancreatic inflammation upon tpl-2 ablation (77) and it have been also shown that Cot/tpl-2 plays a role in the onset of Crohn’s disease (78)(Fig.3) Besides, experiments performed with the Cot/tpl-2 KO mice, as well as with intact cells were Cot/tpl-2 is down modulated, have shown that this kinase modulates the strict balance between Th1 and Th2 cytokines. It has been reported that Cot/tpl-2 is an important negative regulator of Th1-type adaptive response due to its capacity to downregulate IL-12 production (75) (Fig. 3). Antigen presenting cells from Cot/tpl-2 KO mice stimulated with CpG, an activator of TLR9, showed an increase in the IL-12 production (75, 76), and this has been confirmed “in vivo” by performing studies with the whole animal (75). More recently, it has been also shown, that antigen presenting cells from Cot/tpl-2 KO mice stimulated with LPS or CpG, showed not only an increase in IL-12 but also in the production of IFNβ, being concomitant with a reduction in the levels of IL-10, a typical Th2

Figure 3. Cot/tpl-2 modulates Th1/Th2 response and inflammation. In the figure are summarized some data from the bibliography using mice null for the Cot/tpl-2 gene. The Cot/tpl-2 KO mice shows a decreased production of inflammatory cytokines such as TNFα or IL-1β and an increased production of cytokines characteristic of the Th1 response like IL-12, IFNβ and decreased production of IL-10.


cytokine (76). However, further studies have to be carried on to determine the exact role of Cot/tpl-2 in polarizing the immune response to type Th1, since IL-12-stimulation of CD4+ T lymphocytes of Cot/tpl-2 KO mice do not to produce IFNγ, another key cytokine in the Th1 response (79). Cot/tpl-2 a target of anti-inflammatory drugs Cot/tpl-2 and RAF isoforms are capable of activating the MEK1-Erk1/Erk2 pathway, being Erk1/Erk2 a MAP kinase that phosphorylates multiple substrates. However RAF-MEK1-Erk1/Erk2 pathway is activated in response to proliferative signals, whereas Cot/tpl-2 mediates the intracellular signalling of PAMPs and pro inflammatory cytokines, indicating that Cot/tpl-2 is involved in innate and adaptive immunity and thereby in inflammation. Besides, the Cot/tpl-2 physiological role can not be replace by any other protein, indicating that this protein is a very good target to develop new and improved anti-inflammatory drugs (80, 81). In fact, several groups and companies, have focused in the development of small organic compounds to specifically block the kinase activity by occupying the active center of Cot/tpl-2. To date, 1,7-napthyridine-3-carbonitriles, quinoline-3-carbonitriles and thieno[2,3-c]pyridines, have been reported as Cot/tpl-2 kinase inhibitors (82, 85). It has been shown, that the 7-amino substituted thieno[2,3-c]pyridines are capable to inhibit activation of Erk1/Erk2 in bone marrow derived macrophages in response to LPS stimulation as well as TNFα secretion in human whole blood (85). Among the quinoline-3-carbonitriles, it has been proposed that the 8-bromo-4-(3-chloro-4-fluorophenylamino)-6-[(1-methyl-1H-imidazol-4-yl)met hylamino]quinoline-3-carbonitrile compound is capable to inhibit, at an oral dose of 50 mg/kg, the production TNF-α in a LPS-induced rat model (83). More recently, it has appeared the possibility to obtain a specific ribozyme to degrade the Cot mRNA. This catalytic RNA is an adenine-dependent hairpin ribozyme that specific cleaves “in vitro” the mRNA of Cot/tpl-2 between bp +225 and +266 bp (86, 87). Further studies are required to determine whether this type of compounds or other improved ones will be useful to inhibit specifically the expression of Cot/ tpl-2 “in vivo”. Conclusion Cot gene was identified in 1991, and since then we are beginning to understand its physiological role in biological processes. Today, there is also a very clear link between inflammation and cancer and Cot/tpl-2 is a good example of how a protein that is required to control immune responses


properly, playing a role in cell transformation when des-regulated. Now, it is also clear that this gene has a unique role in the development of the immune responses and in inflammation and thus Cot/tpl-2 is emerging as a very interesting target to arise new anti-inflammatory drugs. However, and as it has been discussed here, there are still many questions unresolved before to complete the picture of Cot/tpl-2 activation and function. Acknowledgements Our research is supported by the Plan Nacional (SAF 2008-00819) and Mutua Madrileña. References 1. Miyoshi, J., Higashi, T., Mukai, H., Ohuchi, T., and Kakunaga, T. (1991) Mol

Cell Biol 11, 4088-4096 2. Ceci, J. D., Patriotis, C. P., Tsatsanis, C., Makris, A. M., Kovatch, R., Swing, D.

A., Jenkins, N. A., Tsichlis, P. N., and Copeland, N. G. (1997) Genes Dev 11, 688-700

3. Chan, A. M., Chedid, M., McGovern, E. S., Popescu, N. C., Miki, T., and Aaronson, S. A. (1993) Oncogene 8, 1329-1333

4. Chiariello, M., Marinissen, M. J., and Gutkind, J. S. (2000) Mol Cell Biol 20, 1747-1758

5. Gandara, M. L., Lopez, P., Hernando, R., Castano, J. G., and Alemany, S. (2003) Mol Cell Biol 23, 7377-7390

6. Robinson, M. J., Beinke, S., Kouroumalis, A., Tsichlis, P. N., and Ley, S. C. (2007) Mol Cell Biol 27, 7355-7364

7. Aoki, M., Hamada, F., Sugimoto, T., Sumida, S., Akiyama, T., and Toyoshima, K. (1993) J Biol Chem 268, 22723-22732

8. Waterfield, M., Jin, W., Reiley, W., Zhang, M., and Sun, S. C. (2004) Mol Cell Biol 24, 6040-6048

9. Waterfield, M. R., Zhang, M., Norman, L. P., and Sun, S. C. (2003) Mol Cell 11, 685-694

10. Rodriguez, C., Pozo, M., Nieto, E., Fernandez, M., and Alemany, S. (2006) Cell Signal 18, 1376-1385

11. Makris, A., Patriotis, C., Bear, S. E., and Tsichlis, P. N. (1993) J Virol 67, 4283-4289 12. Patriotis, C., Makris, A., Bear, S. E., and Tsichlis, P. N. (1993) Proc Natl Acad

Sci U S A 90, 2251-2255 13. Erny, K. M., Peli, J., Lambert, J. F., Muller, V., and Diggelmann, H. (1996)

Oncogene 13, 2015-2020 14. Clark, A. M., Reynolds, S. H., Anderson, M., and Wiest, J. S. (2004) Genes

Chromosomes Cancer 41, 99-108 15. Sourvinos, G., Tsatsanis, C., and Spandidos, D. A. (1999) Oncogene 18,

4968-4973


16. Krcova, Z., Ehrmann, J., Krejci, V., Eliopoulos, A., and Kolar, Z. (2008) Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub 152, 21-25

17. Christoforidou, A. V., Papadaki, H. A., Margioris, A. N., Eliopoulos, G. D., and Tsatsanis, C. (2004) Mol Cancer 3, 34

18. Babu, G., Waterfield, M., Chang, M., Wu, X., and Sun, S. C. (2006) J Biol Chem 281, 14041-14047

19. Eliopoulos, A. G., Davies, C., Blake, S. S., Murray, P., Najafipour, S., Tsichlis, P. N., and Young, L. S. (2002) J Virol 76, 4567-4579

20. Rodriguez, C., Lopez, P., Pozo, M., Duce, A. M., Lopez-Pelaez, M., Fernandez, M., and Alemany, S. (2008) Cell Signal 20, 1625-1631

21. Hatziapostolou, M., Polytarchou, C., Panutsopulos, D., Covic, L., and Tsichlis, P. N. (2008) Cancer Res 68, 1851-1861

22. Tsatsanis, C., Vaporidi, K., Zacharioudaki, V., Androulidaki, A., Sykulev, Y., Margioris, A. N., and Tsichlis, P. N. (2008) Proc Natl Acad Sci U S A 105, 2987-2992

23. Dumitru, C. D., Ceci, J. D., Tsatsanis, C., Kontoyiannis, D., Stamatakis, K., Lin, J. H., Patriotis, C., Jenkins, N. A., Copeland, N. G., Kollias, G., and Tsichlis, P. N. (2000) Cell 103, 1071-1083

24. Banerjee, A., Gugasyan, R., McMahon, M., and Gerondakis, S. (2006) Proc Natl Acad Sci U S A 103, 3274-3279

25. Banerjee, A., and Gerondakis, S. (2007) Immunol Cell Biol 85, 420-424 26. Loniewski, K. J., Patial, S., and Parameswaran, N. (2007) Mol Immunol 44,

3715-3723 27. O'Neill, L. A., and Bowie, A. G. (2007) Nat Rev Immunol 7, 353-364 28. Akira, S., and Takeda, K. (2004) Nat Rev Immunol 4, 499-511 29. O'Neill, L. A., and Dinarello, C. A. (2000) Immunol Today 21, 206-209 30. Belich, M. P., Salmeron, A., Johnston, L. H., and Ley, S. C. (1999) Nature 397,

363-368 31. Beinke, S., Deka, J., Lang, V., Belich, M. P., Walker, P. A., Howell, S., Smerdon,

S. J., Gamblin, S. J., and Ley, S. C. (2003) Mol Cell Biol 23, 4739-4752 32. Babu, G. R., Jin, W., Norman, L., Waterfield, M., Chang, M., Wu, X., Zhang, M.,

and Sun, S. C. (2006) Biochim Biophys Acta 1763, 174-181 33. Lang, V., Symons, A., Watton, S. J., Janzen, J., Soneji, Y., Beinke, S., Howell,

S., and Ley, S. C. (2004) Mol Cell Biol 24, 5235-5248

34. Papoutsopoulou, S., Symons, A., Tharmalingham, T., Belich, M. P., Kaiser, F., Kioussis, D., O'Garra, A., Tybulewicz, V., and Ley, S. C. (2006) Nat Immunol 7, 606-615

35. Verstrepen, L., Carpentier, I., Verhelst, K., and Beyaert, R. (2009) Biochem Pharmacol 78, 105-114

36. Channavajhala, P. L., Wu, L., Cuozzo, J. W., Hall, J. P., Liu, W., Lin, L. L., and Zhang, Y. (2003) J Biol Chem 278, 47089-47097

37. Savinova, O. V., Hoffmann, A., and Ghosh, G. (2009) Mol Cell 34, 591-602

38. Heissmeyer, V., Krappmann, D., Wulczyn, F. G., and Scheidereit, C. (1999) EMBO J 18, 4766-4778


39. Salmeron, A., Janzen, J., Soneji, Y., Bump, N., Kamens, J., Allen, H., and Ley, S. C. (2001) J Biol Chem 276, 22215-22222

40. Orian, A., Gonen, H., Bercovich, B., Fajerman, I., Eytan, E., Israel, A., Mercurio, F., Iwai, K., Schwartz, A. L., and Ciechanover, A. (2000) EMBO J 19, 2580-2591

41. Vallabhapurapu, S., and Karin, M. (2009) Annu Rev Immunol 27, 693-733 42. Beinke, S., Robinson, M. J., Hugunin, M., and Ley, S. C. (2004) Mol Cell Biol

24, 9658-9667 43. Eliopoulos, A. G., Wang, C. C., Dumitru, C. D., and Tsichlis, P. N. (2003)

EMBO J 22, 3855-3864 44. Eliopoulos, A. G., Das, S., and Tsichlis, P. N. (2006) J Biol Chem 281,

1371-1380 45. Stafford, M. J., Morrice, N. A., Peggie, M. W., and Cohen, P. (2006) FEBS Lett

580, 4010-4014 46. Cho, J., and Tsichlis, P. N. (2005) Proc Natl Acad Sci U S A 102, 2350-2355 47. Cho, J., Melnick, M., Solidakis, G. P., and Tsichlis, P. N. (2005) J Biol Chem

280, 20442-20448 48. Luciano, B. S., Hsu, S., Channavajhala, P. L., Lin, L. L., and Cuozzo, J. W.

(2004) J Biol Chem 279, 52117-52123 49. Kane, L. P., Mollenauer, M. N., Xu, Z., Turck, C. W., and Weiss, A. (2002) Mol

Cell Biol 22, 5962-5974 50. Black, T. M., Andrews, C. L., Kilili, G., Ivan, M., Tsichlis, P. N., and Vouros, P.

(2007) J Proteome Res 6, 2269-2276 51. Zacharioudaki, V., Androulidaki, A., Arranz, A., Vrentzos, G., Margioris, A. N.,

and Tsatsanis, C. (2009) J Immunol 182, 6444-6451 52. Salmeron, A., Ahmad, T. B., Carlile, G. W., Pappin, D., Narsimhan, R. P., and

Ley, S. C. (1996) EMBO J 15, 817-826 53. Hagemann, D., Troppmair, J., and Rapp, U. R. (1999) Oncogene 18, 1391-1400 54. Ballester, A., Velasco, A., Tobena, R., and Alemany, S. (1998) J Biol Chem 273,

14099-14106 55. Choi, H. S., Kang, B. S., Shim, J. H., Cho, Y. Y., Choi, B. Y., Bode, A. M., and

Dong, Z. (2008) FASEB J 22, 113-126 56. Gomez-Casero, E., San-Antonio, B., Iniguez, M. A., and Fresno, M. (2007) Cell

Signal 19, 1652-1661 57. de Gregorio, R., Iniguez, M. A., Fresno, M., and Alemany, S. (2001) J Biol Chem

276, 27003-27009 58. Tsatsanis, C., Patriotis, C., Bear, S. E., and Tsichlis, P. N. (1998) Proc Natl Acad

Sci U S A 95, 3827-3832 59. Tsatsanis, C., Patriotis, C., and Tsichlis, P. N. (1998) Oncogene 17, 2609-2618 60. Das, S., Cho, J., Lambertz, I., Kelliher, M. A., Eliopoulos, A. G., Du, K., and

Tsichlis, P. N. (2005) J Biol Chem 280, 23748-23757 61. Sanchez-Valdepenas, C., Punzon, C., San-Antonio, B., Martin, A. G., and Fresno,

M. (2007) Cell Signal 19, 528-537 62. Mattioli, I., Sebald, A., Bucher, C., Charles, R. P., Nakano, H., Doi, T., Kracht,

M., and Schmitz, M. L. (2004) J Immunol 172, 6336-6344 63. Wittwer, T., and Schmitz, M. L. (2008) Biochem Biophys Res Commun 371, 294-297


64. Chikamatsu, S., Furuno, T., Kinoshita, Y., Inoh, Y., Hirashima, N., Teshima, R., and Nakanishi, M. (2007) Mol Immunol 44, 1490-1497

65. Lin, X., Cunningham, E. T., Jr., Mu, Y., Geleziunas, R., and Greene, W. C. (1999) Immunity 10, 271-280

66. O'Mahony, A., Lin, X., Geleziunas, R., and Greene, W. C. (2000) Mol Cell Biol 20, 1170-1178

67. Ouellet, M., Roy, J., Barbeau, B., Geleziunas, R., and Tremblay, M. J. (2003) Biochemistry 42, 8260-8271

68. Sebald, A., Mattioli, I., and Schmitz, M. L. (2005) Eur J Immunol 35, 318-325 69. Cismasiu, V. B., Duque, J., Paskaleva, E., Califano, D., Ghanta, S., Young, H.

A., and Avram, D. (2009) Biochem J 417, 457-466 70. Hammerman, P. S., Fox, C. J., Cinalli, R. M., Xu, A., Wagner, J. D., Lindsten,

T., and Thompson, C. B. (2004) Cancer Res 64, 8341-8348 71. Claudio, E., Brown, K., Park, S., Wang, H., and Siebenlist, U. (2002) Nat

Immunol 3, 958-965 72. Ballester, A., Tobena, R., Lisbona, C., Calvo, V., and Alemany, S. (1997) J

Immunol 159, 1613-1618 73. Caivano, M., Rodriguez, C., Cohen, P., and Alemany, S. (2003) J Biol Chem 278,

52124-52130 74. Eliopoulos, A. G., Dumitru, C. D., Wang, C. C., Cho, J., and Tsichlis, P. N.

(2002) EMBO J 21, 4831-4840 75. Sugimoto, K., Ohata, M., Miyoshi, J., Ishizaki, H., Tsuboi, N., Masuda, A.,

Yoshikai, Y., Takamoto, M., Sugane, K., Matsuo, S., Shimada, Y., and Matsuguchi, T. (2004) J Clin Invest 114, 857-866

76. Kaiser, F., Cook, D., Papoutsopoulou, S., Rajsbaum, R., Wu, X., Yang, H. T., Grant, S., Ricciardi-Castagnoli, P., Tsichlis, P. N., Ley, S. C., and O'Garra, A. (2009) J Exp Med

77. Van Acker, G. J., Perides, G., Weiss, E. R., Das, S., Tsichlis, P. N., and Steer, M. L. (2007) J Biol Chem 282, 22140-22149

78. Kontoyiannis, D., Boulougouris, G., Manoloukos, M., Armaka, M., Apostolaki, M., Pizarro, T., Kotlyarov, A., Forster, I., Flavell, R., Gaestel, M., Tsichlis, P., Cominelli, F., and Kollias, G. (2002) J Exp Med 196, 1563-1574

79. Watford, W. T., Hissong, B. D., Durant, L. R., Yamane, H., Muul, L. M., Kanno, Y., Tato, C. M., Ramos, H. L., Berger, A. E., Mielke, L., Pesu, M., Solomon, B., Frucht, D. M., Paul, W. E., Sher, A., Jankovic, D., Tsichlis, P. N., and O'Shea, J. J. (2008) J Exp Med 205, 2803-2812

80. Cohen, P. (2009) Curr Opin Cell Biol 21, 317-324

81. Gaestel, M., Mengel, A., Bothe, U., and Asadullah, K. (2007) Curr Med Chem 14, 2214-2234

82. Kaila, N., Green, N., Li, H. Q., Hu, Y., Janz, K., Gavrin, L. K., Thomason, J., Tam, S., Powell, D., Cuozzo, J., Hall, J. P., Telliez, J. B., Hsu, S., Nickerson-Nutter, C., Wang, Q., and Lin, L. L. (2007) Bioorg Med Chem 15, 6425-6442

83. Green, N., Hu, Y., Janz, K., Li, H. Q., Kaila, N., Guler, S., Thomason, J., Joseph-McCarthy, D., Tam, S. Y., Hotchandani, R., Wu, J., Huang, A., Wang, Q.,


Leung, L., Pelker, J., Marusic, S., Hsu, S., Telliez, J. B., Hall, J. P., Cuozzo, J. W., and Lin, L. L. (2007) J Med Chem 50, 4728-4745

84. George, D., Friedman, M., Allen, H., Argiriadi, M., Barberis, C., Bischoff, A., Clabbers, A., Cusack, K., Dixon, R., Fix-Stenzel, S., Gordon, T., Janssen, B., Jia, Y., Moskey, M., Quinn, C., Salmeron, J. A., Wishart, N., Woller, K., and Yu, Z. (2008) Bioorg Med Chem Lett 18, 4952-4955

85. Cusack, K., Allen, H., Bischoff, A., Clabbers, A., Dixon, R., Fix-Stenzel, S., Friedman, M., Gaumont, Y., George, D., Gordon, T., Grongsaard, P., Janssen, B., Jia, Y., Moskey, M., Quinn, C., Salmeron, A., Thomas, C., Wallace, G., Wishart, N., and Yu, Z. (2009) Bioorg Med Chem Lett 19, 1722-1725

86. Li, Y. L., Torchet, C., Vergne, J., and Maurel, M. C. (2007) Biochimie 89, 1257-1263

87. Li, Y. L., Vergne, J., Torchet, C., and Maurel, M. C. (2009) FEBS J 276, 303-314



3. WNK kinase signalling in cancer biology

Sónia Moniz and Peter Jordan Departamento de Genética Humana, Instituto Nacional de Saúde 'Dr. Ricardo Jorge'

Lisboa, Portugal

Abstract. The subfamily of WNK (With No K [lysine]) protein kinases is characterised by a unique sequence variation in the catalytic domain: a conserved lysine residue that is essential for catalytic activity in most eucaryotic protein kinases is located in an alternative position within the catalytic domain and this variation may result in unique substrate binding properties. The human genome contains four WNK genes, with different tissue-specific expression patterns. Mutations in WNK1 or WNK4 cause a hereditary hypertension syndrome due to increased renal salt retention. At the molecular level, WNK1, WNK3 or WNK4 have been shown to regulate different ion transporters in both the kidney and extrarenal tissues. Growing evidence has also revealed additional roles for WNK kinases in multiple signalling cascades related to tumour biology. There is strong evidence for a role as upstream regulators of MAPK cascades involved in cell proliferation control. In addition, a requirement of some WNK members for cell survival has been demonstrated. Here, we review the experimental evidence linking WNK kinases to tumorigenesis and discuss their role in major aspects of tumour biology: G1/S cell cycle progression, metabolic tumour cell adaptation, evasion of apoptosis, and metastasis.

Correspondence/Reprint request: Dr. Peter Jordan, Departamento de Genética Humana, Instituto Nacional de Saúde 'Dr. Ricardo Jorge', Avenida Padre Cruz, 1649-016 Lisboa, Portugal E-mail: [email protected]

Sónia Moniz & Peter Jordan 44

Introduction The phosphorylation of cellular proteins alters their function and is catalysed by protein kinases in response to extra- or intracellular stimuli. A superfamily of protein kinases has been recognized based on conserved sequence elements in their catalytic domains and human genome-wide analyses have concluded that there are 518 different protein kinase genes [1-3]. Of these, 478 genes present the classical eukaryotic protein kinase catalytic domain [4], while 40 are atypical kinases. Of the 478 classical kinases, 428 possess known or likely kinase activity, while the remaining 50 proteins lack conserved key sequence elements. Among the 428 classical kinases, 365 fall into seven major families (TK, CAMK, AGC, CMGC, STE, TKL, and CK1), whereas 63 kinases present sequence variations in their catalytic domains and have been classified separately as 'Other' [3]. This group contains unique kinase genes or small subfamilies, including the WNK (With No [K] = lysine) subfamily. Phylogenetically, WNKs define a separate protein kinase branch, most closely related to the STE (mammalian homologs of the Saccharomyces cerevisiae STE families of serine/threonine kinases) and TKL (tyrosine kinase-like) family branches [1, 3]. The unique sequence variation that characterises WNK protein kinases [5, 6] is the lack of a highly conserved catalytic lysine in subdomain II, important for the correct positioning of adenosine triphosphate (ATP) within the classical catalytic domain [4]. The crystal structure for the recombinant WNK1 kinase domain has been determined [7] and revealed that an alternative lysine from subdomain I reaches into the position normally occupied by the conserved lysine residue from subdomain II and therefore confers catalytic activity [5] (see Figure 1). The sequence variation in the catalytic domain can be expected to induce conformational changes responsible for unique ATP and substrate binding properties of WNK kinases. Otherwise, the overall structure and folding of the WNK1 catalytic domain resembles that of other protein kinases with typical bilobal domain architecture [8]. WNK proteins are serine/threonine kinases that undergo autophosphorylation and require phosphorylation of at least one serine residue within the WNK activation loop, S382 in WNK1, to become active [9, 10]. This serine is part of an activation site motif (S378FAKS382) that is conserved in MEK family kinases. Sequence alignment identifies homologue serine residues in positions 356, 308 and 332 for WNK2, 3 and 4, respectively.

WNK signalling in cancer 45

N-terminal lobe(ATP-binding)

C-terminal lobe(Substrate binding)

GxGxxGxVxxxxxxxxxxxxxKxxxGxGxxKxVxxxxxxxxxxxxxCxxx

glycine flap

Distinct catalytic lysine in WNKs(Subdomain I)

Distinct catalytic lysine in WNKs(Subdomain I)

typical catalytic lysine(Subdomain II)

MAPK motif:WNK motif:

Subdomain I Subdomain II

II III IV V VIA VIB VII VIII IX X XII II III IV V VIA VIB VII VIII IX X XII

With no K (lysine) =WNK

Figure 1. Sequence variation in the catalytic domain that characterises the WNK subfamily of protein kinases. The typical protein kinase catalytic domain is subdivided into 12 subdomains. A conserved lysine in subdomain II, which binds ATP, is absent in WNK kinases and functionally substituted by another lysine located in subdomain I [5, 7], as indicated. Another primary level in the regulation of WNK kinase activity can be provided by the autoinhibitory domain encompassing about 70 residues just C-terminal to the catalytic domain and these are conserved in WNK protein kinases across species [9, 11]. The characteristic catalytic lysine in subdomain I together with 7 other adjacent amino acid residues form an invariant WNK signature sequence [6] (see Figure 2). A database search using the WNK signature sequence revealed the existence of a variable number of WNK genes in animals and plants that are absent in uni-cellular organisms. The number of WNK genes increases from one WNK gene in C. elegans or Drosophila melanogaster to four different WNK genes in mouse and man, while the higher plant Arabidopsis thaliana has 9 WNK genes [6, 12, 13]. The human genome contains four WNK genes with chromosomal locations at 12p13.3 (WNK1), 9q22.31 (WNK2), Xp11.23-p11.21 (WNK3) and 17q21-q22 (WNK4). The human WNK1 gene contains 28 exons and extends over more than 150 kb, WNK2 is encoded by 31 exons that span 136 kb, WNK3 contains 24 exons and spans 165 kb, whereas WNK4 is encoded by 19 exons contained within 16 kb of genomic DNA [6, 14, 15]. These genes differ in their tissue specific expression patterns. WNK1 is expressed in most human foetal and adult tissues; WNK3 is mainly expressed in brain,


liver and small intestine, WNK2 is mainly present in brain, heart and colon crypt, and WNK4 predominantly in the kidney, colon, skin, liver, and lung [5, 6, 14-19]. Expression analysis of WNK kinases has also revealed the existence of tissue-specific alternative promoter usage, alternative splicing and polyadenylation variants [6, 14, 15, 19-21].

I II III IV V VIA VIB VII VIII IX X XI

N-terminal lobe(ATP-binding)

C-terminal lobe(Substrate binding)

WNK signature sequence

GENE CATALYTIC SUBDOMAINS I+II GENE ID hs WNK1 LKFDIEIGRGSFKTVYKGLDTETTVEVAWCELQ 65125 hs WNK2 LKFDIELGRGSFKTVYKGLDTETWVEVAWCELQ 65268 hs WNK3 LKFDIELGRGAFKTVYKGLDTETWVEVAWCELQ 65267 hs WNK4 LKFDIEIGRGSFKTVYRGLDTDTTVEVAWCELQ 65266 mm WNK1 LKFDIEIGRGSFKTVYKGLDTETTVEVAWCELQ 232341 mm WNK2 LKFDIELGRGSFKTVYKGLDTETWVEVAWCELQ 75607 mm WNK3 LKFDIELGRGAFKTVYKGLDTETWVEVAWCELQ 279561 mm WNK4 LKFDIEIGRGSFKTVYRGLDTDTTVEVAWCELQ 69847 at WNK1 GRYNEVLGKGASKTVYRAFDEYEGIEVAWNQVK 819651 at WNK2 GRYDEILGKGASKTVYRAFDEYEGIEVAWNQVK 821810 at WNK3 GRYKEVLGKGAFKEVYRAFDQLEGIEVAWNQVK 823984 at WNK4 GRFAEILGRGAMKTVYKAIDEKLGIEVAWSQVK 835947 at WNK5 GRFREVLGKGAMKTVYKAFDQVLGMEVAWNQVK 824326 at WNK6 IRYKEVIGKGAFKTVYKAFDEVDGIEVAWNQVR 821406 at WNK7 IRYKEVIGKGASKTVFKGFDEVDGIEVAWNQVR 841339 at WNK8 IRYDDVLGRGAFKTVYKAFDEVDGIEVAWNLVS 834204 at WNK9 GRYNEVLGKGSSKTVYRGFDEYQGIEVAWNQVK 832881 rn WNK1 LKFDIEIGRGSFKTVYKGLDTETTVEVAWCELQ 116477 pt WNK1 LKFDIEIGRGSFKTVYKGLDTETTVEVAWCELQ 451739 clf WNK1 LKFDIEIGRGSFKTVYRGLDTDTTVEVAWCELQ 477728 gg WNK1 LKFDIEIGRGSFKTVYKGLDTDTTVEVAWCELQ 427925 dm CG7177 FKYDKEVGRGSFKTVYRGLDTLTGVPVAWCELL 40391 ce C46C2.1 LKFDEELGRGSFKTVFRGLDTETGVAVAWCELQ 177743

Invariant G G K V D VAW Figure 2. Alignment of subdomains I and II of the catalytic domains of some WNK kinases from various species. The invariant WNK signature sequence is indicated below. Species abbreviations are: hs, Homo sapiens; mm, Mus musculus; at, Arabidopsis thaliana; rn, Rattus norvegicus; pt, Pan troglodytes; clf, Canis lupus familiaris; gg, Gallus gallus; dm, Drosophila melanogaster; ce, Caenorhabditis elegans (updated from Veríssimo and Jordan, 2001); NCBI sequence IDs are also given. The four human WNK kinases are large proteins containing 2382, 2297, 1743 or 1243 amino acids with predicted molecular weights of 251, 243, 192 and 135 kDa, for WNK1, 2, 3 and 4 respectively [6, 14, 15, 21]. They share high homology within their kinase domains and the adjacent auto-inhibitory domain [9], however, homology outside their catalytic domains is low except for the presence of three short WNK homology regions (Figure 3).


WNK Homology Region I

WNK Homology Region III

Coiledcoil

CatalyticDomain

WNK Homology Region II

Conservedin WNK 1-3

AutoinhibitoryDomain

WNK Homology Region I

WNK Homology Region III

Coiledcoil

CatalyticDomain

WNK Homology Region II

Conservedin WNK 1-3

AutoinhibitoryDomain

1 147-405 1743

1 195-453 2297

1 221-479 2382I II III c4c3c2c1

1 174-432 1243

WNK4

WNK3

WNK2

WNK1cc kinase cc cc

kinase cc

kinase cc

kinase cc cc

91%

93%

83%

Figure 3. Schematic representation of the four human WNK proteins. Catalytic, autoinhibitory and coiled-coil domains are shown as well as conserved regions of homology between WNK kinases. The white numbers indicate sequence identity of the respective catalytic domain with regard to the WNK1 sequence. The most C-terminal homology region III contains a coiled-coil domain and WNK1 and WNK4 have at least one further coiled-coil domain outside the conserved regions. In addition, four short regions of unknown function are conserved between WNK1, WNK2 and WNK3. Multiple PXXP motifs, the typical binding sites for Src-homology 3 (SH3) domains [22], are also present in WNK protein kinases, suggesting a potential scaffolding function for these proteins [6, 23]. In this chapter we will address the experimental evidence linking WNK kinases to tumorigenesis and discuss four main aspects: their roles in G1/S cell cycle progression, metabolic tumour cell adaptation, evasion of apoptosis and, finally, invasion and metastasis. 1. Role of WNK kinases in cell cycle progression Normal cells require extracellular growth stimulatory signals in order to leave the quiescent and enter a proliferative state. In epithelial cells the stimulation of cell proliferation is intimately linked to the receptor-mediated activation of the mitogen-activated protein kinase (MAPK) cascades. 1.1. WNKs regulate MAP kinase cascades The best characterized MAPK pathways in mammals involve the extracellular signal-regulated kinases 1 and 2 (ERK1/2), the p38 isoforms,


the c-jun N-terminal kinases (JNK1-3), and ERK5. In general, MAPK cascades consist of three protein kinase layers: MAPKs are phosphorylated and activated by MAPK kinases (MAP2Ks) which are in turn activated by phosphorylation by MAPK kinase kinases (MAP3Ks). Other kinases, including p21-activated protein kinases (PAKs), can act as upstream regulators (MAP4Ks) for these MAPK cascades and connect them with additional signalling pathways. MAPK signalling cascades regulate fundamental biological functions such as cell growth, proliferation, differentiation, migration and apoptosis and are found to be deregulated in approximately one third of all human cancers [24]. Sequence alignments using the catalytic domain of WNK1 showed that the closest human homologous kinases are the MAP3Ks MEKKs or Rafs and the MAP4K PAKs, which all display around 30 % sequence identity and 50 % sequence homology [6]. It is therefore not surprising that effects of WNK kinases on various MAPK cascades have been observed. WNK1 was reported to be required for EGF-dependent stimulation of ERK5 but the activation of ERK1/2, JNK or p38 MAP kinases was not significantly affected. In vitro, WNK1 interacts with and phosphorylates MEKK2 and MEKK3, however, this phosphorylation does not seem to stimulate MEKK2/3 activity. In contrast, transfection of cells with either WNK1 or its kinase-dead mutant WNK1K233M stimulates MEKK3 autophosphorylation and activity towards its substrate MEK6. Apparently, WNK1 acts by protein-protein interaction to assemble an ERK5 activation complex and thus acts as an upstream regulator of the ERK5 pathway (Figure 4). WNK1 was required for activation of ERK5 by EGF, in HeLa cells, but in the presence of high concentrations of EGF, this effect became less pronounced [25]. In agreement with these findings, the down-regulation of WNK1 in C17.2 mouse neural progenitor cells also suppressed activation of ERK5 and greatly reduced cell growth [26]. In contrast to WNK1, human WNK2 had no effect on ERK5 but modulated activation of ERK1/2. Experimental depletion of WNK2 or overexpression of a kinase-dead WNK2K207M mutant led to increased phospho-ERK1/2 levels but only when a basal ERK stimulation was present and not in serum-free culture conditions [19]. This increase in ERK1/2 activation promoted cell cycle progression through G1/S and sensitized cells to respond to lower concentrations of EGF. From these data one might predict that loss of WNK2 expression can promote cell cycle progression in tumour cells. Interestingly, WNK2 expression is silenced in a significant percentage of human gliomas [27, 28, see section 5 below] suggesting that this pathway may be used in some tumour types to promote cell proliferation. The mechanism by which a reduction in WNK2 expression increased


ERK1/2 activation involves phosphorylation of MEK1 at serine 298, a modification that increases MEK1 affinity towards ERK1/2. Apparently, WNK2 affects Rac1 activation with subsequent stimulation of the Rac-effector PAK1, the kinase responsible for MEK1 S298 phosphorylation [29] (Figure 4). Overexpression of WNK4 was also claimed to increase the phosphorylation of ERK1⁄2 and p38 MAPKs following EGF stimulation or hyperosmotic stress [30]. However, the mechanism of WNK4 involvement remains unclear and the corresponding effect of endogenous WNK4 was not addressed. Together, these data demonstrate that WNK kinases can act upon different MAPK cascades and modulate cell proliferation. It may also be anticipated that this modulation is concerted because WNK kinases appear to regulate each others activities. For example, the recombinant kinase domain of WNK1 was shown to phosphorylate recombinant WNK2 and WNK4 catalytic domains in vitro, in particular WNK4 on serine 332 in its activation

loop [31]. WNK4 can also phosphorylate the catalytic domain of WNK1 and expression of the autoinhibitory domains of WNK1 and WNK4 are able to inhibit each others catalytic activities [11, 31]. Gel filtration profiles further indicate that endogenous WNK1 exists as a tetramer [31]. Although the functional role of oligomerization is still unclear, there is some evidence that the autoinhibitory domain of one WNK1 molecule may inhibit the catalytic activity of another WNK1 molecule within the tetramer [9, 31]. Additionally, given the presence of conserved coiled-coil protein interaction domains in all

WNK2MEKK2/3

MKK5

ERK5

Ras

Cell proliferation

S218S222

S298

activePAK1

activeRac1

WNK1

plasma membrane

EGFEGF EGFEGFEGFEGF EGFEGF

B-Raf

MEK1

ERK1/2

Ras

Figure 4. Schematic representation of the role of human WNK1 and WNK2 in the cellular response to EGF. (Left side) Physiological activation of the ERK5 pathway is enhanced in the presence of WNK1. (Right side) Expression of WNK2 controls a Rho GTPase/PAK-mediated signalling cross talk that can increase ERK1/2 activation.


four human WNK proteins, it is likely that hetero-oligomers form under certain conditions. In this way, the activity of different WNK proteins on parallel MAPK cascades might be coordinated within a single complex [23]. WNKs also regulate upstream activators of MAPK pathways (MAP4Ks), such as those belonging to the STE20-related protein kinases. These consist of 10 subfamilies that include PAK and the germinal centre kinases (GCKs). The proline-alanine-rich kinase (SPAK) and oxidative-stress responsive 1 (OSR1, also refered to as OXSR1) are members of the GCK VI subfamily and primarily known for their role in cell volume sensing and osmotic homeostasis. Given that cell size and volume are often linked to cell cycle progression, it is not surprising that these kinases have roles as upstream activators of MAPK pathways (MAP4Ks) [32, 33]. For example, SPAK is a known activator of the p38 and JNK cascades [34-36] and OSR1 acts upstream of CaMKII and p38 in C. elegans [37]. OSR1 and SPAK interact with WNK1, WNK3 and WNK4 [38-44] and WNK1 and WNK4 phosphorylate these proteins on two residues: T233 and S373 on SPAK and T185 and S325 on OSR1. Residues T233 or T185 are located in the activation loops of these kinases' catalytic domains and mediate their activation [38]. The molecular details of how the phosphorylation of SPAK and OSR1 by WNK kinases relates to cell proliferation have not yet been investigated. As mentioned above, WNK2 also acts upon another STE20 member, PAK1, but direct phosphorylation does not seem to be involved. Rather, WNK2 affects activation of Rac1, a known upstream stimulator of PAK1 [29] and active PAK1 phosphorylates MEK1 on S298, leading to increased activity of ERK1/2 MAPK [45-49]. 1.2. Effect of WNKs on receptor traffic MAP kinases are activated in response to growth factor binding to their cell surface receptors and can either stimulate or inhibit cell proliferation. After ligand binding, most receptors become internalized into signalling-competent endosomes and then either recycle back to the cell surface or become degraded in lysosomes as a mechanism to downregulate receptor signalling in cells [50, 51]. Vesiclular traffic can therefore modulate cellular signalling intensity and WNK4 has been proposed to promote cargo delivery to lysosomes. Expression of WNK4 inhibited the activity of the sodium chloride cotransporter NCC by diverting the channel to lysomal degradation [52, 53] and because this effect involved a general adaptor protein (AP) 3-dependent endosomal sorting mechanism, WNK4 may also modulate the ratio of degradation of endocytosed growth factor receptors. This mechanism


is known to downregulate receptor signalling in cells and can have consequences for cell proliferation [50]. Similarly, WNK1 and WNK4 have been shown to stimulate clathrin-dependent endocytosis of the renal outer medullar potassium 1 channel (ROMK1), thus inhibiting potassium secretion [54-56]. This process involves the interaction of WNK1 and WNK4 with the scaffold protein intersectin 1 (ITSN1), which binds to specific proline-rich WNK motifs via its SH3 domains and does not require WNK kinase activity [56]. Additional support of a role for WNKs in endocytosis comes from a genome-wide RNAi screen to determine the effect of human protein kinases on endocytosis. In this screen virus entry was used as a measure of clathrin-mediated (vesicular stomatitis virus- VSV) or caveolin-dependent (simian virus- SV40) endocytosis and it was found that WNK4 interfered with VSV entrance whereas WNK2 inhibited caveolin-mediated SV40 uptake [57, Table S1]. 1.3. WNKs as targets of insulin signalling and the PI3K/Akt pathway There is substantial evidence that many tissues and derived cancer cells express insulin and insulin-like growth factor 1 (IGF1) receptors and that elevated circulating levels of IGF1 and insulin reflect a nutritional energy imbalance with increased cancer risk [58]. Insulin and IGF1 receptors are important activators of the phosphatidylinositol 3-kinase (PI3K)/Akt signalling network and increase the proliferation of neoplastic cell lines. Following IGF1 treatment of cells, Akt is activated downstream of PI3K and phosphorylates WNK1 in its threonine residue 60, although this phosphorylation does not appear to affect WNK1 kinase activity, or its subcellular localization [59-63]. This WNK1 phosphorylation is dependent on PDK1, which phosphorylates Akt within its activation loop [64], and is blocked by PI3K inhibitors. IGF1 also activates the Akt-homologous serum- and glucocorticoid-induced protein kinase SGK1. While SGK1 is known to regulate sodium transport in response to aldosterone in the kidney, it also promotes degradation of the cyclin-dependent kinase inhibitor protein p27kip in other cell types [65]. Xu and collaborators proposed that WNK1 phosphorylation by PKB/Akt contributes to SGK1 activation and that WNK1 is required for SGK1 activation by IGF1 [61]. There is evidence that SGK1 may in turn phosphorylate WNK1 on Thr60 and/or other sites and create a positive feedback [62].


When 3T3-L1 adipocytes were treated with insulin, an increased WNK1 phosphorylation was observed, and RNAi-induced depletion of WNK1 enhanced insulin-stimulated thymidine incorporation by about 2-fold without significantly affecting insulin-stimulated glucose transport in these cells [60]. This observation suggests a negative regulatory role for WNK1 in insulin-mediated proliferation control. SGK1 apparently also phosphorylates WNK4 on serine residue 1169 in response to aldosterone [66] but a role related to cell proliferation has not been investigated yet. 1.4. WNKs in transforming growth factor β signalling

The transforming growth factor (TGF) β transduces signals by stimulating the formation of heteromeric complexes of type I and type II serine/threonine kinase receptors which then phosphorylate and activate cytoplasmic transcription activators of the Smad family. Signalling from the TGFβ-Smad pathway often plays an anti-proliferative role in many cell types, acting as a tumour suppressor [67]. WNK1 and WNK4 kinase domains both interacted with and phosphorylated Smad2 in vitro. The siRNA-mediated knockdown of WNK1 reduced overall Smad2 levels but increased Smad2/3-dependent transcriptional responses [68]. This suggests that WNK1 imposes an inhibitory constraint on Smad2 and TGFβ signalling. Interestingly, a systematic mapping of the TGFβ receptor interactome revealed a link to another WNK1 substrate, the OSR1 protein kinase [69]. 1.5. WNKs in calcium signalling An increase in cytoplasmic calcium serves as a second messenger for calmodulin- and calcineurin-activated signalling pathways that promote cancer cell proliferation [70, 71]. One mechanism of increasing cytoplasmic calcium is uptake from the extracellular medium by calcium-channels, such as the transient receptor potential vanilloid (TRPV) channels and TRPV6 overexpression has been reported in carcinomas of the colon, breast, thyroid, and ovary [72] Expression of WNK4 specifically enhanced TRPV5-mediated calcium uptake, which correlates with increased membrane expression of the channel [73], whereas WNK3 stimulated TRPV5 and TRPV6-mediated transport [74]. In contrast, WNK4 and WNK1 decreased cell surface expression of TRPV4, which plays a role in osmoregulation following hypotonic cell swelling [75].


Additionally, WNK1 was also implicated in intracellular calcium sensing, because it selectively binds synaptotagmin 2 and phosphorylates the C2 domains, involved in calcium binding. WNK1 and synaptotagmin 2 co-localize in secretory vesicles in an insulin secreting pancreatic β cell line and phosphorylation by WNK1 increased the amount of calcium necessary for synaptotagmin 2 binding to phospholipid vesicles [76]. Together, these results show that WNK kinases can affect multiple signalling pathways related to cell proliferation; however, at present their best defined roles are as upstream regulators of MAPK cascades. 2. WNK kinases and metabolic adaptation of tumour cells Cell adaptation to the extracellular environment is required during the initiation of tumour formation when cells start to proliferate and in consequence experience suboptimal supply of oxygen and nutrients due to diffusion limits. In response, cancer cells upregulate the production of glycolytic energy leading to the generation of lactate and acidosis of the extracellular medium. Tumour cells have to compensate for this acid-induced toxicity in order to remain viable and continue to grow [77, 78]. 2.1. WNKs and glucose uptake Malignant cells have accelerated metabolism and the increased requirements for ATP production are satisfied by aerobic glycolysis [79]. One adaptation is to increase glucose transport into malignant cells by overexpression of glucose transporters (GLUTs) [80] and elevated expression of GLUT1 has been described in many cancers. In 3T3L1 adipocytes, the syntaxin 4-regulatory protein Munc18c functions in insulin-stimulated GLUT4 vesicle translocation and fusion [81]. A role for WNK1 has been reported in vesicle/granule exocytosis, because it forms a complex with Munc18c. WNK1 binds Munc18c via its kinase domain but does not phosphorylate it, suggesting a scaffolding function for WNK1 in recruiting Munc18c as a substrate for another kinase or phosphatase [82]. Interestingly, increased glucose supply decreased the expression of WNK4 in mouse kidneys, indicating the existence of a regulatory mechanism that links WNK expression to extracellular glucose concentrations [83]. Because synaptotagmins and Munc18c are implicated in a variety of membrane trafficking and vesicle fusion events, their interaction with WNK1 provides a putative mechanism for how WNKs may regulate retention or insertion of plasma membrane proteins [76].


2.2. Regulation of extracellular pH and the plasma membrane conductances Mutations in the WNK1 and WNK4 genes were initially discovered to cause pseudo-hypoaldosteronism type II (or Gordon’s syndrome), a rare familial form of hypertension [14] characterised by increased renal salt reabsorption accompanied by hyperkalemia and metabolic acidosis due to impaired K+ and H+ excretion. Subsequently, a huge body of evidence has shown that WNK1, WNK3 and WNK4 are involved in the regulation of a variety of renal but also extra-renal ion channels (for WNK2 no such studies have been reported yet). Summarizing these studies, evidence links WNKs to the regulation of cell surface expression or channel activity of the cotransporters NCC, NKCC and KCC, the ROMK and Kir1.1 potassium channels, the CFTR chloride channel, the Cl-/HCO3

- exchangers SLC26A6 and A9, the epithelial sodium channel (ENaC), and the TRPV4 and TRPV5 calcium channels [reviewed in 18, 84-86]. These data indicate that WNK kinases are important to maintain fundamental cell functions related to electrolyte homeostasis. The regulation of electrolyte homeostasis is important for tumour cells and one relevant aspect is pH regulation. The extracellular environment in the tumour is more acidic than the intracellular pH (6.2–6.9 compared with 7.3–7.4). The increased production and export of glycolytic lactate together with overexpression of the Na+/H+ exchanger NHE1, create a reversed pH gradient across the tumour cell membrane. Through NHE1 action, the inwardly directed sodium gradient can drive the uphill extrusion of protons, alkalinize intracellular pH and further acidify the extracellular pH. A supporting activity in pH regulation is provided by the chloride/bicarbonate (Cl–/HCO3–) anion exchangers of the SLC26A family [87]. The combined transport activities of NHE and SLC26A channels lead to electroneutral NaCl absorption with net H+ secretion. Recent findings suggest that the expression of the SLC26A family is regulated by WNKs. When expressed in Xenopus oocytes, WNK4 inhibits the expression at the plasma membrane of SLC26A6 [17] and WNK1, WNK3 and WNK4 inhibit the expression of SLC26A9 [88], consequently decreasing channel activities. Another aspect is the electrochemical gradient across the plasma membrane which most cells require to sustain nutrient uptake or release of metabolic products. For example, the electrochemical gradient for Na+ drives the uptake of nutrients and is in part maintained by the negative membrane potential. In case of enhanced Na+/H+ exchanger activity, as observed in tumour cells, the intracellular Na+ concentration increases so that the ATP-consuming Na+/K+ ATPase becomes activated to exchange cytosolic Na+ for


K+. In consequence, K+ ions need to recycle through membrane potasium channels in order to keep the Na+/K+ ATPase going. A solid body of evidence documents that various types of tumours overexpress voltage-gated K+ channels, in particular the ether-a-go-go (EAG) and BK (Ca2+-activated K+ channels) subfamilies. These channels open in response to a depolarization of the cell membrane, thus allowing an efflux of K+ ions. Experimental overexpression of these channels promotes tumour cell proliferation, probably because it helps to maintain the membrane potential [89]. Although WNK kinases have not yet been linked to the direct regulation of voltage-gated K+ channels, their known effects on channels involved in Na+, K+, Cl- and HCO3– homeostasis already indicate that these proteins play at least an overall regulatory role on electrochemical cell homeostasis and lead us to suggest that these channels may represent a further link between WNKs and tumour biology that should be explored. 3. Role of WNK kinases in the evasion of apoptosis The apoptotic program is a major barrier to tumour development that must be inactivated to achieve net tumour cell proliferation. A complex interplay between different pro- or anti-apoptotic factors determines whether cells survive or activate the cell death programme [90]. 3.1. WNK3 modulates caspase 3 Experimental data and mathematical models have suggested that apoptosis activation depends on molecular threshold values that trigger either the cell death pathway or maintain cell survival [91, 92]. Thus, either over-production of anti-apoptotic factors, or loss of expression of pro-apoptotic factors, or changes in their activation status can shift the balance towards cell survival. WNK3 has been shown to act on this balance by promoting cell survival in a caspase-3-dependent pathway [21]. Suppression of endogenous WNK3 by RNA interference accelerated the apoptotic response of HeLa cells and promoted the activation of caspase-3. The mechanism of WNK3 action involves interaction with procaspase-3 and heat-shock protein 70. The prosurvival role was only in part dependent on the catalytic activity of WNK3 because a kinase-dead WNK3K159M mutant also interacted with procaspase-3 and increased cell survival to some extent. This indicates an adaptor or scaffold function for WNK3 within a protein complex that controls procaspase-3 activation. Accordingly, the level of WNK3 expression or activity may determine sensitivity or tolerance of cells towards apoptotic stimuli.


3.2. Genome-wide screening results The role of protein phosphorylation in apoptosis induction was recently studied by systematic depletion of each human protein kinase or phosphatase in HeLa cells using RNA interference [93]. This approach identified 73 kinases whose suppression increased the level of apoptosis by at least twofold over control, defining them as survival kinases. In this screen WNK1 and WNK3 kinases also scored positive, albeit below the threshold value (WNK1- 1.98 fold; WNK3- 1.58 fold), whereas WNK2 and WNK4 had no effect on cell survival [J. Blenis, Boston, personal communication]. These results are further supported by a genome wide screen for cell survival factors in Drosophila melanogaster. The single Drosophila WNK gene (designated CG7177; NM_141072) is most homologous to WNK3 and WNK1, and its depletion by RNAi affected cell survival in fly S2R+ cells [94]. Although WNK1 and WNK3 were detected in these screens, they did not stand out as essential genes for sustaining cell survival. However, these screens scored for spontaneous induction of apoptosis and a different physiological response can be expected when cells are exposed to conditions that challenge cell survival. For example, the single Caenorhabditis elegans WNK kinase (designated C46C2) was found essential for worm survival but only following hyperosmotic stress conditions [13]. It can be expected that the roles of WNK1 and WNK3 in cell survival will become more evident when cells face metabolic stress situations, including the above mentioned glycolytic acidosis and cell volume regulation. 3.3. Putative role of WNK substrates OSR1 and SAPK Which candidate signalling pathways could be involved in the anti-apoptotic function of WNK kinases? The survival under osmotic stress depends on OSR1 [37]. As mentioned previously, OSR1 and SPAK interact with WNK1, WNK3 and WNK4 (see section 1.2) and are key regulators of the NCC, NKCC and KCC ion channels involved in the regulation of ion homoeostasis and volume control in mammalian cells [95, 96]. Cells under hyperosmotic stress conditions are challenged to induce an apoptotic response and such conditions, including high concentrations of NaCl, KCl, glucose, sucrose, manitol or sorbitol, provoke a marked and reproducible increase in WNK1 activity [5, 9, 31] including towards its substrate OSR1 [10]. Thus, the expression or activity levels of WNK1 and WNK3 may determine the cellular capacity to cope with osmotic or acid-related stress. In a yeast two hybrid screen with the SPAK C-terminal substrate binding domain, the apoptosis-associated tyrosine kinase (AATYK) was identified as


an interacting protein [97], and in myeloid precursor [98] or cerebellar granule cells [99], AATYK expression promotes apoptosis. In addition, SPAK and OSR1 participate in other signalling pathways related to cell survival. Overexpression of the TNF receptor RELT led to activation of both the p38 and JNK survival pathways in 293 cells [35] and this was dependent on SPAK kinase activity, which is regulated by WNKs. Also, following extrinsic pro-apoptotic stimuli such as TRAIL receptor binding, SPAK is cleaved at two distinct sites by a caspase 3-like protease, which removes its substrate-binding domain. Accordingly, experimental depletion of SPAK expression by siRNA increased the sensitivity of HeLa cells to TRAIL-induced apoptosis [100]. Thus, down-regulation of SPAK activity is an important target to enhance the apoptotic effect of TRAIL, where WNK kinases may influence cell survival. 4. WNK kinases in invasion and metastasis Cancer cells can escape from the constraints of their tissue of origin and enter into the circulation to reach distant organs and eventually form secondary tumours, called metastases. 4.1. Epithelial–mesenchymal transition In many solid tumour types, the epithelial tumour cells switch to a highly motile fibroblastoid or mesenchymal phenotype, a process called epithelial–mesenchymal transition (EMT). A well characterised inducer of EMT is TGFβ signalling [101, 102] that uses Smad-mediated gene expression to induce the transcription factors Snail and Slug. These repress expression of the E-cadherin gene required for epithelial cell adhesion, a hallmark phenotype of EMT [103]. Since WNK1 was shown to phosphorylate Smad2 in vitro and apparently negatively controls Smad2/3-dependent transcriptional responses and TGFβ Signalling [68], this suggests that loss of expression or inactivating WNK1 mutations could promote EMT of epithelial tumour cells. 4.2. Neuronal cell invasion The ganglioside GD3 is a sialic acid-containing glycosphingolipid normally expressed during development but also in pathological conditions such as cancer. In neural tumour cells, WNK1 expression was found to correlate with that of GD3 [104]. In particular, experimental suppression of the GD3-synthase gene in F-11 tumour cells led to a reduced rate of cell migration and invasiveness [105], conditions under which a dramatic


decrease in WNK1 expression was observed. This suggests that WNK1 may contribute to the invasive phenotype of F-11 cells but further mechanistic details are not yet available. 4.3. Role of WNK-regulated Rho GTPases Rho GTPases control the dynamics of the actin cytoskeleton and are important for cell migration and invasiveness [106]. WNK2 controls, through a yet unknown mechanism, the activation of the small GTPase RhoA, which in a reciprocal way regulates activation of Rac1 [29]. The suppression of WNK2 in cell lines leads to reduced RhoA, but increased Rac1 activation. It remains to be determined whether changes in WNK2 expression or activity affect RhoA/Rac1 activation in tumours, but in this sense, the reported epigenetic silencing of WNK2 in infiltrative gliomas [27] is highly suggestive. A recent report also links WNK1 to Rho GTPases. In neuronal cells WNK1 can be isolated in a complex with Rho-GDI and was proposed to mediate the regulation of Nogo66-induced RhoA activation and neurite outgrowth [107]. Interestingly, a reciprocal relation between RhoA and Rac1 signalling has also been reported to mediate NHE1-dependent changes in motility and invasion of breast cancer cells [107]. For example, NHE1 has a crucial role in driving invasion, possibly by acting as an integrator protein that links the cytoskeleton and various metastasis-specific signalling complexes that mediate invasive activity. NHE1 can directly regulate cytoskeletal dynamics independently of its ion-transporting capabilities because it binds the ezrin, radixin and moesin (ERM) family of cytoskeleton regulators [109]. This feature of actin cytoskeleton remodeling has also been reported downstream of other ion channels [110], including some that are regulated by WNKs. 5. Mutational mechanisms affecting WNK genes in tumours If WNK signalling were important for tumour development, one can expect that genetic alterations affect either WNK gene expression or their coding sequence with consequences for protein activity. 5.1. Changes in gene expression The strongest evidence for such alterations was reported for WNK2. Expression of the human WNK2 gene is silenced in a large percentage of human gliomas due to extensive methylation in the CpG island encompassing the 5’ end of this gene [27]. Likewise, the WNK2 promoter was hyper-methylated in 83% and 71% of grade II and III meningiomas, respectively and this was associated with decreased WNK2 expression in primary


tumours. In contrast, promoter methylation was rare in a total of 209 tumours from thirteen other tumour types [28]. This finding makes WNK2 a candidate tumour suppressor gene in gliomas. WNK2 indirectly inhibits MEK1 and restrains growth-promoting signals through the EGF receptor. Thus, it is possible that the epigenetic silencing of WNK2 interacts on a functional level with genetic alteration of EGFR signalling, a common abnormality in glioblastomas especially due to EGFR gene amplifications [111]. It is interesting to note that colon tumours also show frequent hyper-stimulation of the EGFR signalling pathway and that a partial clone of WNK2 has previously been isolated as the colon cancer antigen SDCC43 [112]. An incomplete WNK2 fragment was also isolated as a T-cell recognized pancreatic cancer cell antigen P/OKcl13 [113]. The other three human WNK genes share with WNK2 the presence of large CpG islands surrounding their promoter regions and extending into the first exons [6, 20, 15, 27]. This observation suggests some common mechanism of regulation under physiological conditions but also indicates that hyper-methylation of the WNK1, WNK3 or WNK4 promoters may exist in other tumour types. In future studies, the presence of genomic deletions in WNK genes should also be explored because one mutation type that predisposes patients to Gordon syndrome is a deletion within intron 1 of WNK1. This is a very large intron of 58 kb which contains a kidney-specific repressor element [114] and its deletion results in WNK1 overexpression [14]. The structure of the WNK2 and WNK3 genes is conserved with respect to a large intron 1 (44 kb and 22 kb, respectively), raising the possibility that a similar mutational mechanism may lead to their overexpression. However, a homology search using the kidney-specific WNK1 repressor element C1 [114] did not identify any homologous sequences in introns 1 from the WNK2 and WNK3 genes [unpublished observation]. WNK expression studies will further need to test for the presence of alternative promoters. In WNK1 intron 4 an intragenic promoter generates a kidney-specific isoform (KS-WNK1) lacking kinase activity, which is thus functionally different from the long ubiquitous isoform (L-WNK1). Epigenetic changes could shift promoter usage in tumours and thus generate dominant-negative WNK isoforms, a possibility that remains to be explored as a potential mutagenic mechanism. An additional avenue that requires more attention is the role of WNK alternative splicing variants. For example, one WNK1 variant lacks exons 11 and 12 [6], two WNK2 transcripts exist that differ in their C-terminal exon with different C-terminal sequences [19], and for WNK3 two variants differ in usage of mostly brain-specific exons [15, 21]. Alternative splicing variants


can affect transcript turnover or encode protein isoforms. At present no simple structure/function relation exists to predict functional differences between two splicing variants from the same gene, however, many examples of variants have been documented that significantly differ regarding regulation or signalling activities [115-118]. Indeed, a recent report revealed that the two WNK3 variants have opposite effects on expression of the ion channel NCC [119]. 5.2. WNK point mutations in tumours In the last years, large-scale cancer genome sequencing efforts were conducted to determine the full spectrum of mutations in a given tumour. Among the provided lists of mutated genes a variety of point mutations in several WNK genes were identified in breast, colon, lung or brain tumours (Figure 5) [120-124]. No experimental evidence exists, however, showing whether the resulting missense or frameshift mutations have a functional impact on the corresponding WNK protein. It is possible that these mutations represent just 'passenger' mutations in a genetically unstable cancer cell genome that are not causally implicated in oncogenesis [120].

Tissue Histology/Type Gene Zygosity cDNA Protein Mutation ReferenceBreast IDC WNK1 Het c.5395C>G p.Q1799E Missense 2,3,4Breast pleomorphic lobular ca WNK1 Het c.1255G>C p.E419Q Missense 2,4Breast pleomorphic lobular ca WNK1 Het c.6569C>G p.S2190C Missense 2,4Lung adenocarcinoma WNK1 Hom c.7086C>A p.F2362L Missense 1,4Ovary serous carcinoma WNK1 Het c.2829C>T p.Y943Y Silent 4Colon colorectal WNK1 c.3596A>G p.E1199G Missense 3Brain glioblastoma WNK1 Het c.5293G>A p.G1765S Missense 5Brain glioblastoma WNK2 Het c.3799G>A p.A1267T Missense 5Colorectal adenocarcinoma WNK2 Het c.1922delC p.P641fs*2 Frameshift deletion 4Stomach adenocarcinoma WNK2 Het c.4116delC p.S1373fs*5 Frameshift deletion 4Lung adenocarcinoma WNK2 Het c.5933G>T p.S1978I Missense 1,4Lung neuroendocrine carcinoma WNK2 Het c.4856G>A p.G1619E Missense 1,4Ovary serous carcinoma WNK2 Het c.1486G>T p.V496L Missense 4Ovary Mucinous carcinoma WNK2 Het c.6642delC p.T2215fs*31 Frameshift deletion 4Glioma glioblastoma WNK3 Het c.2784C>T p.H928H Silent 4Lung squamous cell carcinoma WNK3 Het c.2561C>G p.S854C Missense 4Lung large cell carcinoma WNK3 Het c.4599G>T p.L1533F Missense 4Kidney clear cell carcinoma WNK3 Het c.3809C>A p.T1270N Missense 4Kidney clear cell carcinoma WNK3 Het c.4900T>C p.S1634P Missense 4Stomach adenocarcinoma WNK4 Het c.1786delG p.V596fs*53 Frameshift deletion 4Melanoma metastatic WNK4 Het c.1402C>T p.L468L Silent 4Melanoma WNK4 Het c.2974C>T p.P992S Missense 4Melanoma WNK4 Het c.3154C>T p.P1052S Missense 4Ovary Mucinous carcinoma WNK4 Het c.1302C>G p.D434E Missense 4Breast IDC WNK4 Het c.441C>G p.F147L Missense 4

References: 1= Davies et al., 2005; 2= Stephens et al., 2005; 3=Sjöblom et al., 2006; 4= Greenman et al., 2007; 5=Parsons et al., 2008 Figure 5. List of WNK mutations identified in different unbiased cancer genome sequencing efforts (corresponding references are indicated).


6. Summary and perspective Considering the reviewed data, the evidence for a role of WNK kinases in cancer cell signalling is just beginning to emerge. At present, experimental evidence is strongest concerning their role as upstream regulators of MAPK cascades, as modulators of Rho-GTPases and inhibitors of apoptosis (see Figure 6). The next few years will certainly shed light on further aspects of cancer biology that are affected by WNKs. It will be of particular interest to elucidate how their abundantly documented role on the regulation of activity or cell surface expression of ion channels can be related to tumour development. In order to understand the underlying signalling network, it will be important to systematically identify WNK interacting proteins and physiological substrate proteins, cellular phenotypes following WNK suppression, and the genetic changes affecting WNK genes during tumorigenesis.

Figure 6. Schematic summary of major WNK-regulated cellular processes with a role in cancer cell biology. Arrows symbolise stimulation and t-shaped lines inhibition. Acknowledgements The authors wish to thank Drs. Fátima Veríssimo (Heidelberg), Karl Kunzelmann (Regensburg) and Jonathan Morris (London) for their helpful suggestions on the manuscript. Work in the authors' laboratory was supported by the Fundação para a Ciência e a Tecnologia, Portugal (Programa de Financiamento Plurianual do CIGMH, grants POCTI/33221/99, POCTI/56294/04 and fellowship BD 11180/02 to S.M.).


References 1. Manning, G., Whyte, D.B., Martinez, R., Hunter, T., Sudarsanam, S. 2002, The

protein kinase complement of the human genome. Science., 298, 1912-34. 2. Kostich, M., English, J., Madison, V., Gheyas, F., Wang, L., Qiu, P., Greene, J.,

Laz, T.M. 2002, Human members of the eukaryotic protein kinase family. Genome Biol., 3, RESEARCH0043.

3. Hanks, S,K. 2003, Genomic analysis of the eukaryotic protein kinase superfamily: a perspective. Genome Biol., 4, 111.1-111.7.

4. Hanks, S,K., Hunter, T. 1995, The eucaryotic protein kinase superfamily: kinase (catalytic) domain structure and classification. FASEB J., 9, 576-96.

5. Xu, B.E., English, J.M., Wilsbacher, J.L., Stippec, S., Goldsmith, E.J., Cobb, M.H. 2000, WNK1, a novel mammalian serine/threonine protein kinase lacking the catalytic lysine in subdomain II. J Biol Chem., 275, 16795-801.

6. Veríssimo, F., Jordan, P. 2001, WNK kinases, a novel protein kinase subfamily in multi-cellular organisms. Oncogene., 20, 5562-9.

7. Min, X., Lee, B.H., Cobb, M.H., Goldsmith, E.J. 2004, Crystal structure of the kinase domain of WNK1, a kinase that causes a hereditary form of hypertension. Structure., 12, 1303-11.

8. Knighton, D.R., Zheng, J.H., Ten Eyck, L.F., Ashford, V.A., Xuong, N.H., Taylor, S.S., Sowadski, J.M. 1991, Crystal structure of the catalytic subunit of cyclic adenosine monophosphate-dependent protein kinase, Science., 253, 407-14.

9. Xu, B.E., Min, X., Stippec, S., Lee, B.H., Goldsmith, E.J., Cobb, M.H. 2002, Regulation of WNK1 by an autoinhibitory domain and autophosphorylation. J Biol Chem., 277, 48456-62.

10. Zagórska, A., Pozo-Guisado, E., Boudeau, J., Vitari, A.C., Rafiqi, F.H., Thastrup, J., Deak, M., Campbell, D.G., Morrice, N.A., Prescott, A.R., Alessi, D.R. 2007, Regulation of activity and localization of the WNK1 protein kinase by hyperosmotic stress. J Cell Biol., 176, 89-100.

11. Wang, Z., Yang, C.L., Ellison, D.H. 2004, Comparison of WNK4 and WNK1 kinase and inhibiting activities, Biochem Biophys Res Commun., 317, 939-44.

12. Nakamichi, N., Murakami-Kojima, M., Sato, E., Kishi, Y., Yamashino, T., Mizuno, T. 2002, Compilation and characterization of a novel WNK family of protein kinases in Arabiodpsis thaliana with reference to circadian rhythms. Biosci Biotechnol Biochem., 66, 2429-36.

13. Choe, K., Strange, K. 2007, Evolutionarily conserved WNK and Ste20 kinases are essential for acute volume recovery and survival following hypertonic shrinkage in Caenorhabditis elegans. Am J Physiol Cell Physiol., 293, C915-27.

14. Wilson, F.H., Disse-Nicodeme, S., Choate, K.A., Ishikawa, K., Nelson-Williams, C., Desitter, I., Gunel, M., Milford, D.V., Lipkin, G.W., Achard, J.M. 2001, Human hypertension caused by mutations in WNK kinases. Science., 293, 1107-12.

15. Holden, S., Cox, J., Raymond, F.L. 2004, Cloning, genomic organization, alternative splicing and expression analysis of the human gene WNK3 (PRKWNK3). Gene., 335, 109-19.


16. Choate, K.A., Kahle, K.T., Wilson, F.H., Nelson-Williams, C., Lifton, R.P. 2003, WNK1, a kinase mutated in inherited hypertension with hyperkalemia, localizes to diverse Cl- -transporting epithelia. Proc Natl Acad Sci U S A., 100, 663-8.

17. Kahle, K.T., Gimenez, I., Hassan, H., Wilson, F.H., Wong, R.D., Forbush, B., Aronson, P.S., Lifton, R.P. 2004, WNK4 regulates apical and basolateral Cl- flux in extrarenal epithelia. Proc Natl Acad Sci U S A., 101, 2064-69.

18. Kahle, K.T., Wilson, F.H., Lalioti, M., Toka, H., Qin, H., Lifton, R.P. 2004, WNK kinases: molecular regulators of integrated epithelial ion transport. Curr Opin Nephrol Hypertens., 13, 557-62.

19. Moniz, S., Veríssimo, F., Matos, P., Brazão, R., Silva, E., Kotelevets, L., Chastre, E., Gespach, C., Jordan, P. 2007, Protein kinase WNK2 inhibits cell proliferation by negatively modulating the activation of MEK1/ERK1/2. Oncogene., 26, 6071-81.

20. Delaloy, C., Lu, J., Houot, A.M., Disse-Nicodeme, S., Gasc, J.M., Corvol, P., Jeunemaitre, X. 2003, Multiple promoters in the WNK1 gene: one controls expression of a kidney-specific kinase-defective isoform. Mol Cell Biol., 23, 9208-21.

21. Veríssimo, F., Silva, E., Morris, J.D., Pepperkok, R., Jordan, P. 2006, Protein kinase WNK3 increases cell survival in a caspase 3-dependent pathway. Oncogene., 25, 4172-82.

22. Kay, B.K., Williamson, M.P., Sudol, M. 2000, The importance of being proline: the interaction of proline-rich motifs in signaling proteins with their cognate domains. FASEB J., 14, 231-41.

23. Xu, B.E., Lee, B.H., Min, X., Lenertz, L., Heise, C.J., Stippec, S., Goldsmith, E.J., Cobb, M.H. 2005, WNK1: analysis of protein kinase structure, downstream targets, and potential roles in hypertension. Cell Res., 15, 6-10.

24. Dhillon, A.S., Hagan, S., Rath, O., Kolch, W. 2007, MAP kinase signalling pathways in cancer, Oncogene., 26, 3279-90.

25. Xu, B.E., Stippec, S., Lenertz, L., Lee, B-H., Zhang, W., Lee, Y.K., Cobb, M.H. 2004, WNK1 activates ERK5 by an MEKK2/3-dependent mechanism. J Biol Chem., 279, 7826-31.

26. Sun, X., Gao, L., Yu, R.K., Zeng, G. 2006, Down-regulation of WNK1 protein kinase in neural progenitor cells suppresses cell proliferation and migration. J Neurochem., 99, 1114-21.

27. Hong, C., Moorefield, K.S., Jun, P., Aldape, K.D., Kharbanda, S., Phillips, H.S., Costello, J.F. 2007, Epigenome scans and cancer genome sequencing converge on WNK2, a kinase-independent suppressor of cell growth. Proc Natl Acad Sci USA., 104, 10974-79.

28. Jun, P., Hong, C., Lal, A., Wong, J.M., McDermott, M.W., Bollen, A.W., Plass, C, Held, W.A., Smiraglia, D.J., Costello, J.F. 2008, Epigenetic silencing of the kinase tumor suppressor WNK2 is tumor-type and tumor-grade specific. Neuro Oncol., in press.

29. Moniz, S., Matos, P., Jordan, P. 2008, WNK2 modulates MEK1 activity through the Rho GTPase pathway. Cell Signal., 20, 1762-8.


30. Shaharabany, M., Holtzman, E.J., Mayan, H., Hirschberg, K., Seger, R., Farfel, Z. 2008, Distinct pathways for the involvement of WNK4 in the signaling of hypertonicity and EGF. FEBS J., 275, 1631-42.

31. Lenertz, L.Y., Lee, B.H., Min, X., Xu, B.E., Wedin, K., Earnest, S., Goldsmith, E.J., Cobb, M.H. 2005, Properties of WNK1 and implications for other family members. J Biol Chem., 280, 26653-8.

32. Dan, I., Watanabe, N.M., Kusumi, A. 2001, The Ste20 group kinases as regulators of MAP kinase cascades. Trends Cell Biol., 11, 220-30.

33. Strange, K., Denton, J., Nehrke, K. 2006, Ste20-type kinases: evolutionarily conserved regulators of ion transport and cell volume. Physiology., 21, 61-8.

34. Johnston, A.M., Naselli, G., Gonez, L.J., Martin, R.M., Harrison, L.C., Deaizpurua, H.J. 2000, SPAK, a Ste20/SPS1-related kinase that activates the p38 pathway. Oncogene., 19, 4290-97.

35. Polek, T.C., Talpaz, M., Spivak-Kroizman, T. 2006, The TNF receptor, RELT, binds SPAK and uses it to mediate p38 and JNK activation. Biochem Biophys Res Commun., 343, 125-34.

36. Yan, Y., Nguyen, H., Dalmasso, G., Sitaraman, S.V., Merlin, D. 2007, Cloning and characterization of a new intestinal inflammation-associated colonic epithelial Ste20-related protein kinase isoform. Biochim Biophys Acta., 1769, 106-16.

37. Solomon, A., Bandhakavi, S., Jabbar, S., Shah, R., Beitel, G..J., Morimoto, R.I. 2004, Caenorhabditis elegans OSR-1 regulates behavioral and physiological responses to hyperosmotic environments. Genetics., 167, 161-70.

38. Vitari, A.C., Deak, M., Morrice, N.A., Alessi, D.R. 2005, The WNK1 and WNK4 protein kinases that are mutated in Gordon's hypertension syndrome phosphorylate and activate SPAK and OSR1 protein kinases. Biochem. J., 391, 17-24.

39. Moriguchi, T., Urushiyama, S., Hisamoto, N., Iemura, S., Uchida, S., Natsume, T., Matsumoto, K., Shibuya, H. 2005, WNK1 regulates phosphorylation of cation-chloride-coupled cotransporters via the STE20-related kinases, SPAK and OSR1. J Biol Chem., 280, 42685-93.

40. Anselmo, A.N., Earnest, S., Chen, W., Juang, Y.C., Kim, S.C., Zhao, Y., Cobb, M.H. 2006, WNK1 and OSR1 regulate the Na+, K+, 2Cl- cotransporter in HeLa cells. Proc Natl Acad Sci USA., 103, 10883-8.

41. Gagnon KB, England R, and Delpire E. 2006, Volume sensitivity of cation-chloride cotransporters is modulated by the interaction of two kinases: SPAK and WNK4. Am J Physiol Cell Physiol 290: C134-C142.

42. Vitari, A.C., Thastrup, J., Rafiqi, F.H., Deak, M., Morrice, N.A., Karlsson, H.K., Alessi, D.R, 2006, Functional interactions of the SPAK/OSR1 kinases with their upstream activator WNK1 and downstream substrate NKCC1. Biochem J., 397, 223-31.

43. Richardson, C., Rafiqi, F.H., Karlsson, H.K., Moleleki, N., Vandewalle, A., Campbell, D.G., Morrice, N.A., Alessi, D.R. 2008, Activation of the thiazide-sensitive Na+-Cl- cotransporter by the WNK-regulated kinases SPAK and OSR1. J Cell Sci., 121, 675-84.


44. Ponce-Coria, J., San-Cristobal, P., Kahle, K.T., Vazquez, N., Pacheco-Alvarez, D., de Los Heros, P., Juárez, P., Muñoz, E., Michel, G., Bobadilla, N.A., Gimenez, I., Lifton, R.P., Hebert, S.C., Gamba, G. 2008, Regulation of NKCC2 by a chloride-sensing mechanism involving the WNK3 and SPAK kinases. Proc Natl Acad Sci USA., 105, 8458-63.

45. Frost, J.A., Steen, H., Shapiro, P., Lewis, T., Ahn, N., Shaw, P.E., Cobb, M.H. 1997, Cross-cascade activation of ERKs and ternary complex factors by Rho family proteins. EMBO J., 16, 6426-38.

46. Coles, L.C., Shaw, P.E. 2002, PAK1 primes MEK1 for phosphorylation by Raf-1 kinase during cross-cascade activation of the ERK pathway. Oncogene., 21, 2236-44.

47. Eblen, S.T., Slack, J.K., Weber, M.J., Catling, A.D. 2002, Rac-PAK signaling stimulates extracellular signal-regulated kinase (ERK) activation by regulating formation of MEK1-ERK complexes. Mol Cell Biol., 22, 6023-33.

48. Slack-Davis, J.K., Eblen, S.T., Zecevic, M., Boerner, S.A., Tarcsafalvi, A., Diaz, H.B., Marshall, M.S., Weber, M.J., Parsons, J.T., Catling, A.D. 2003, PAK1 phosphorylation of MEK1 regulates fibronectin-stimulated MAPK activation. J Cell Biol., 162, 281-91.

49. Park, E.R., Eblen, S.T., Catling, A.D. 2007, MEK1 activation by PAK: A novel mechanism. Cell Signal., 19, 1488-496.

50. Sorkin, A., von Zastrow, M. 2002, Signal transduction and endocytosis: close encounters of many kinds. Nat Rev Mol Cell Biol., 3, 600-14.

51. González-Gaitán, M. 2003, Signal dispersal and transduction through the endocytic pathway. Nat Rev Mol Cell Biol., 4, 213-24.

52. Cai, H., Cebotaru, V., Wang, Y.H., Zhang, X.M., Cebotaru, L., Guggino, S.E., Guggino, W.B. 2006, WNK4 kinase regulates surface expression of the human sodium chloride cotransporter in mammalian cells. Kidney Int., 69, 2162-70.

53. Subramanya, A.R., Liu, J., Ellison, D.H., Wade, J.B., Welling, P.A. 2009, WNK4 diverts the thiazide-sensitive NaCl cotransporter to the lysosome and stimulates AP-3 interaction. J Biol Chem., in press.

54. Kahle, K.T., Wilson, F.H., Leng, Q., Lalioti, M.D., O'Connell, A.D., Dong, K., Rapson, A.K., MacGregor, G.G., Giebisch, G., Hebert, S.C., Lifton, R.P. 2003, WNK4 regulates the balance between renal NaCl reabsorption and K+ secretion. Nat Genet., 35, 372-6.

55. Cope, G., Murthy, M., Golbang, A.P., Hamad, A., Liu, C.H., Cuthbert, A.W., O'Shaughnessy, K.M. 2006, WNK1 affects surface expression of the ROMK potassium channel independent of WNK4. J Am Soc Nephrol., 17, 1867-74.

56. He, G., Wang, H.R., Huang , S.K., Huang, C.L. 2007, Intersectin links WNK kinases to endocytosis of ROMK1. J Clin Invest., 117, 1078-87.

57. Pelkmans, L., Fava, E., Grabner, H., Hannus, M., Habermann, B., Krausz, E., Zerial, M. 2005, Genome–wide analysis of human kinases in clathrin- and caveolae/raft-mediated endocytosis. Nature., 436, 78–86.

58. Pollak, M. 2008, Insulin and insulin-like growth factor signalling in neoplasia. Nat Rev Cancer., 8, 915-28.

59. Vitari, A.C., Deak, M., Collins, B.J., Morrice, N., Prescott, A.R., Phelan, A.,Humphreys, S., Alessi, D.R. 2004, WNK1, the kinase mutated in an inherited


high-blood-pressure syndrome, is a novel PKB (protein kinase B)/Akt substrate. Biochem. J., 378, 257-68.

60. Jiang, Z.Y., Zhou, Q.L., Holik, J., Patel, S., Leszyk, J., Coleman, K., Chouinard, M., Czech, M.P. 2005, Identification of WNK1 as a substrate of Akt/protein kinase B and a negative regulator of insulin-stimulated mitogenesis in 3T3-L1 cells. J Biol Chem., 280, 21622-28.

61. Xu, B.E., Stippec, S., Chu, P.Y., Lazrak, A., Li, X.J., Lee, B.H., English, J.M., Ortega, B., Huang, C.L., Cobb, M.H. 2005, WNK1 activates SGK1 to regulate the epithelial sodium channel. Proc Natl Acad Sci USA., 102, 10315-20.

62. Xu, B.E., Stippec, S., Lazrak, A., Huang, C.L., Cobb, M.H. 2005, WNK1 activates SGK1 by a phosphatidylinositol 3-kinase-dependent and non-catalytic mechanism. J Biol Chem., 280, 34218-23.

63. Sale, E.M., Hodgkinson, C.P., Jones, N.P., Sale, G.J. 2006, A new strategy for studying protein kinase B and its three isoforms. Role of protein kinase B in phosphorylating glycogen synthase kinase-3, tuberin, WNK1, and ATP citrate lyase. Biochemistry., 45, 213-23.

64. Alessi, D.R., James, S.R., Downes, C.P., Holmes, A.B., Gaffney, P.R., Reese, C.B., Cohen, P. 1997, Characterization of a 3-phosphoinositide-dependent protein kinase which phosphorylates and activates protein kinase Balpha. Curr Biol., 7, 261-9.

65. Hong, F., Larrea, M.D., Doughty, C., Kwiatkowski, D.J., Squillace, R., Slingerland, J.M. 2008, mTOR-raptor binds and activates SGK1 to regulate p27 phosphorylation. Mol Cell., 30, 701-11.

66. Ring, A.M., Leng, Q., Rinehart, .J, Wilson, F.H., Kahle, K.T., Hebert, S.C., Lifton, R.P. 2007, An SGK1 site in WNK4 regulates Na+ channel and K+ channel activity and has implications for aldosterone signaling and K+ homeostasis. Proc Natl Acad Sci USA., 104, 4025-9.

67. Massagué, J., Blain, S.W., Lo, R.S. 2000, TGFbeta signaling in growth control, cancer, and heritable disorders. Cell., 103, 295-309.

68. Lee, B.H., Chen, W., Stippec, S., Cobb, M.H. 2007, Biological cross-talk between WNK1 and the transforming growth factor beta-Smad signaling pathway. J Biol Chem., 282, 17985-96.

69. Barrios-Rodiles, M., Brown, K. R., Ozdamar, B., Bose, R., Liu, Z., Donovan, R. S., Shinjo, F., Liu, Y., Dembowy, J., Taylor, I. W., Luga, V., Przulj, N., Robinson, M., Suzuki, H., Hayashizaki, Y., Jurisica, I., and Wrana, J. L. 2005, High-throughput mapping of a dynamic signaling network in mammalian cells. Science., 307, 1621-25.

70. Monteith, G.R., McAndrew, D., Faddy, H.M., Roberts-Thomson, S.J. 2007, Calcium and cancer: targeting Ca2+ transport. Nat Rev Cancer., 7, 519-30.

71. Roderick, H.L., Cook, S.J. 2008, Ca2+ signalling checkpoints in cancer: remodelling Ca2+ for cancer cell proliferation and survival. Nat Rev Cancer., 8, 361-75.

72. Prevarskaya, N., Zhang, L., Barritt, G. 2007, TRP channels in cancer. Biochim Biophys Acta., 1772, 937-46.

73. Jiang, Y., Ferguson, W.B., Peng, J.B. 2007, WNK4 enhances TRPV5-mediated calcium transport: potential role in hypercalciuria of familial hyperkalemic


hypertension caused by gene mutation of WNK4. Am J Physiol Renal Physiol., 292, F545–F554.

74. Zhang, W., Na, T., Peng, J.B. 2008, WNK3 positively regulates epithelial calcium channels TRPV5 and TRPV6 via a kinase-dependent pathway. Am J Physiol Renal Physiol., 295, F1472-84.

75. Fu, Y., Subramanya, A., Rozansky, D., Cohen, D.M. 2006, WNK kinases influence TRPV4 channel function and localization. Am J Physiol Renal Physiol 290: F1305–F1314,

76. Lee, B.H., Min, X., Heise, C.J., Xu, B.E., Chen, S., Shu, H., Luby-Phelps, K., Goldsmith, E.J., Cobb, M.H. 2004, WNK1 phosphorylates synaptotagmin 2 and modulates its membrane binding. Mol. Cell., 15, 741-51.

77. Gatenby, R.A., Gillies, R.J. 2008, A microenvironmental model of carcinogenesis. Nat Rev Cancer., 8, 56-61.

78. Denko, N.C. 2008, Hypoxia, HIF1 and glucose metabolism in the solid tumour. Nat Rev Cancer., 8, 705-13.

79. Gatenby RA and Gillies RJ. 2004, Why do cancers have high aerobic glycolysis? Nat Rev Cancer. 4(11):891-9.

80. Macheda, M.L., Rogers, S., Best, J.D. 2005, Molecular and cellular regulation of glucose transporter (GLUT) proteins in cancer. J Cell Physiol., 202, 654-62.

81. Thurmond, D.C., Pessin, J.E. 2001, Molecular machinery involved in the insulin-regulated fusion of GLUT4-containing vesicles with the plasma membrane. Mol Membr Biol., 18, 237-45.

82. Oh, E., Heise, C.J., English, J.M., Cobb, M.H., Thurmond, D.C. 2007, WNK1 is a novel regulator of Munc18c-syntaxin 4 complex formation in SNARE-mediated vesicle exocytosis. J Biol Chem., 282, 32613-22.

83. Song, J., Hu, X., Riazi, S., Tiwari, S., Wade, J.B., Ecelbarger, C.A. 2006, Regulation of blood pressure, the epithelial sodium channel (ENaC), and other key renal sodium transporters by chronic insulin infusion in rats. Am J Physiol Renal Physiol., 290, F1055-64.

84. Gamba G. (2005). Role of WNK kinases in regulating tubular salt and potassium transport and in the development of hypertension. Am J Physiol Renal Physiol., 288, F245-F252.

85. Kahle, K.T., Rinehart, J., Ring, A., Gimenez, .I, Gamba, G., Hebert, S.C., Lifton, R.P. 2006, WNK protein kinases modulate cellular Cl- flux by altering the phosphorylation state of the Na-K-Cl and K-Cl cotransporters. Physiology (Bethesda)., 21, 326-35.

86. Kahle, K.T., Ring, A.M., Lifton, R.P. 2008, Molecular physiology of the WNK kinases. Annu Rev Physiol., 70, 329-55.

87. Cardone, R.A., Casavola, V., Reshkin, S.J. 2005, The role of disturbed pH dynamics and the Na+/H+ exchanger in metastasis. Nat Rev Cancer., 5, 786-95.

88. Dorwart, M.R., Shcheynikov, N., Wang, Y., Stippec, S., Muallem, S. 2007, SLC26A9 is a Cl(-) channel regulated by the WNK kinases. J Physiol., 584, 333-45.

89. Kunzelmann, K. 2005, Ion channels and cancer. J Membr Biol., 205, 159-73.


90. Igney, F.H., Krammer, P. 2002, Death and anti-death: tumour resistance to apoptosis. Nat Rev Cancer., 2, 277-88.

91. Bagci, E.Z., Vodovotz, Y., Billiar, T.R., Ermentrout, G.B., Bahar, I. 2006, Bistability in apoptosis: roles of bax, bcl-2, and mitochondrial permeability transition pores. Biophys J., 90, 1546-59.

92. Dogu,Y., Díaz, J. 2009, Mathematical model of a network of interaction between p53 and Bcl-2 during genotoxic-induced apoptosis. Biophys Chem., 143, 44-54.

93. MacKeigan, J.P., Murphy, L.O., Blenis, J. 2005, Sensitized RNAi screen of human kinases and phosphatases identifies new regulators of apoptosis and chemoresistance. Nat Cell Biol., 7, 591-600.

94. Boutros, M., Kiger, A.A., Armknecht, S., Kerr, K., Hild, M., Koch, B., Haas, S.A., Paro, R., Perrimon, N., Heidelberg Fly Array Consortium. 2004, Genome-wide RNAi analysis of growth and viability in Drosophila cells. Science., 303, 832-835.

95. Delpire, E., Gagnon, K.B. 2008, SPAK and OSR1: STE20 kinases involved in the regulation of ion homoeostasis and volume control in mammalian cells. Biochem J., 409, 321-31.

96. Richardson, C., Alessi, D.R. 2008, The regulation of salt transport and blood pressure by the WNK-SPAK/OSR1 signalling pathway. J Cell Sci., 121, 3293-304.

97. Piechotta, K., Garbarini, N., England, R., Delpire, E. 2003, Characterization of the interaction of the stress kinase SPAK with the Na+-K+-2Cl- cotransporter in the nervous system: evidence for a scaffolding role of the kinase. J Biol Chem., 278, 52848-56.

98. Gaozza, E., Baker, S.J., Vora, R.K., Reddy, E.P. 1997, AATYK: a novel tyrosine kinase induced during growth arrest and apoptosis of myeloid cells. Oncogene., 15, 3127-35.

99. Tomomura, M., Furuichi, T. 2005, Apoptosis-associated tyrosine kinase (AATYK) has differential Ca2+-dependent phosphorylation states in response to survival and apoptotic conditions in cerebellar granule cells. J Biol Chem., 280, 35157-63.

100. Polek, T.C., Talpaz, M., Spivak-Kroizman, T.R. 2006, TRAIL-induced cleavage and inactivation of SPAK sensitizes cells to apoptosis. Biochem Biophys Res Commun., 349, 1016-24.

101. Thiery, J.P. 2003, Epithelial–mesenchymal transitions in development and pathologies. Curr Opin Cell Biol., 15, 740-6.

102. Padua, D., Massagué, J. 2009, Roles of TGFβ in metastasis. Cell Research., 19, 89-102.

103. Thuault, S., Tan, E.J., Peinado, H., Cano, A., Heldin, C.H., Moustakas, A. 2008, HMGA2 and Smads co-regulate SNAIL1 expression during induction of epithelial-to-mesenchymal transition. J Biol Chem., 283, 33437-46.

104. Zeng, G., Gao, L., Xia, T., Gu, Y., Yu, R.K. 2005, Expression of the mouse WNK1 gene in correlation with ganglioside GD3 and functional analysis of the mouse WNK1 promoter. Gene., 344, 233-9.

105. Zeng, G., Gao, L., Yu, R.K. 2000, Reduced cell migration, tumor growth and experimental metastasis of rat F-11 cells whose expression of ganglioside GD3 was suppressed. Int J Cancer., 88, 53–57.


106. Ridley, A.J., Schwartz, M.A., Burridge, K., Firtel, R.A., Ginsberg, M.H., Borisy, G., Parsons, J.T., Horwitz, A.R. 2003, Cell migration: integrating signals from front to back. Science., 302, 1704-9.

107. Zhang, Z., Xu, X., Zhang, Y., Zhou, J., Yu, Z., He, C. 2009, LINGO-1 interacts with WNK1 to regulate Nogo-induced inhibition of neurite extension. J Biol Chem., 284, 15717–28.

108. Paradiso, A., Cardone, R.A., Bellizzi, A., Bagorda, A., Guerra, L., Tommasino, M., Casavola, V., Reshkin, S.J. 2004, The Na+-H+ exchanger-1 induces cytoskeletal changes involving reciprocal RhoA and Rac1 signaling, resulting in motility and invasion in MDA-MB-435 cells. Breast Cancer Res., 6, R616-28.

109. Denker, S.P., Huang, D.C., Orlowski, J., Furthmayr, H., Barber, D.L. 2000, Direct binding of the Na-H exchanger NHE1 to ERM proteins regulates the cortical cytoskeleton and cell shape independently of H(+) translocation. Mol. Cell., 6, 1425-36.

110. Denker, S.P., Barber, D.L. 2002, Ion transport proteins anchor and regulate the cytoskeleton. Curr Opin Cell Biol., 14, 214-20.

111. Ekstrand, A.J., James, C.D., Cavenee, W.K., Seliger, B., Pettersson, R.F., Collins,V.P. 1991, Genes for epidermal growth factor receptor, transforming growthf actor a, andepidermal growthf actor and their expression in human gliomas in vivo. Cancer Res., 51, 2164-72.

112. Scanlan, M.J., Chen, Y.T., Williamson, B., Gure, A.O., Stockert, E., Gordan, J.D., Türeci, O., Sahin, U., Pfreundschuh, M., Old, L.J. 1998, Characterization of human colon cancer antigens recognized by autologous antibodies. Int J Cancer., 76, 652-8.

113. Ito, M., Shichijo, S., Tsuda, N., Ochi, M., Harashima, N., Saito, N., Itoh, K. 2001, Molecular basis of T cell-mediated recognition of pancreatic cancer cells. Cancer Res., 61, 2038-46.

114. Delaloy, C., Elvira-Matelot, E., Clemessy, M., Zhou, X.O., Imbert-Teboul, M., Houot, A.M., Jeunemaitre, X., Hadchouel, J. 2008, Deletion of WNK1 first intron results in misregulation of both isoforms in renal and extrarenal tissues. Hypertension., 52, 1149-54.

115. Schwerk, C., Schulze-Osthoff, K. 2005, Regulation of apoptosis by alternative pre-mRNA splicing. Mol Cell., 19, 1-13.

116. Venables, J.P. 2006, Unbalanced alternative splicing and its significance in cancer. BioEssays., 28, 378-86.

116. Venables, J.P. (Ed). 2006, Alternative Splicing and Cancer, Transworld Research Network, Kerala (India), (ISBN: 81-7895-235-1).

118. Srebrow, A., Kornblihtt, A.R. 2006, The connection between splicing and cancer. J Cell Sci., 119, 2635-41.

119. Glover, M., Zuber, A.M., O'Shaughnessy, K.M. 2009, Renal and brain isoforms of WNK3 have opposite effects on NCCT expression. J Am Soc Nephrol. 20, 1314-22.

120. Davies, H., Hunter, C., Smith, R., Stephens, P., Greenman, C., Bignell, G., Teague, J., Woffendin, H., Garnett, M.J., Bottomley, W. et al. 2005, Somatic mutations of the protein kinase gene family in human lung cancer. Cancer Res., 65, 7591-5.


121. Stephens, P., Edkins, S., Davies, H., Greenman, C., Cox, .C, Hunter, C. Bignell G, Teague, J., Smith, R., Stevens, C. et al. 2005, A screen of the complete protein kinase gene family identifies diverse patterns of somatic mutations in human breast cancer. Nat Genet. 37, 590-2.

122. Sjöblom, T., Jones, S., Wood, L.D., Parsons, D.W., Lin, J., Barber, T.D., Mandelker, D., Leary, R.J., Ptak, J., Silliman, N., et al. 2006, The consensus coding sequences of human breast and colorectal cancers. Science., 314, 268-74.

123. Greenman, C., Stephens, P., Smith, R., Dalgliesh, G.L., Hunter, C., Bignell, G., Davies, H., Teague, J., Butler, A., Stevens, C., et al. 2007, Patterns of somatic mutation in human cancer genomes. Nature., 446, 153-8.

124. Parsons, D.W., Jones, S., Zhang, X., Lin, J.C., Leary, R.J., Angenendt, P., Mankoo, P., Carter, H., Siu, I.M., Gallia, G.L., et al., 2008, An integrated genomic analysis of human glioblastoma multiforme. Science., 321, 1807-12.



4. The tumour suppressor role of LKB1 in human cancer

Paola A. Marignani1 and Montse Sanchez-Cespedes2

1Faculty of Medicine, Department of Biochemistry and Molecular Biology, Dalhousie University Halifax, Nova Scotia, Canada; 2Programa Epigenetica i Biologia del Cancer-PEBC –IDIBELL

Hospital Duran i Reynals. Av. Gran Via de l'Hospitalet, 199-203 08907-Hospitalet de Llobregat-Barcelona, Spain

Abstract. Germ-line LKB1 mutations are responsible for Peutz-Jeghers syndrome (PJS) and somatic mutations are involved in the development of some sporadic tumours, in particular lung cancer. The gene encodes the LKB1 protein with serine-threonine kinase activity, which is involved in the regulation of the cell energetic checkpoint through the phosphorylation and activation of AMP-dependent kinase (AMPK) and other substrates. LKB1 is also involved in other important cell processes such as the control of cell polarity in various types of tissues. Here we provide an overview of the alterations of LKB1 in human cancer, Lkb1-mice models and what is currently known about the biological function, substrates and binding partners of LKB1 protein, with a special focus on cancer development.

Correspondence/Reprint request: Dr. M. Sanchez-Cespedes, PEBC –IDIBELL. Hospital Duran i Reynals. Av. Gran Via de l'Hospitalet, 199-203. 08907-Hospitalet de Llobregat-Barcelona, Spain E-mail: [email protected]

Paola A. Marignani & Montse Sanchez-Cespedes 72

Introduction The human LKB1 gene, also known as STK11, spans 23 kb and is made up of nine coding exons and a final non-coding exon. LKB1 codes for an mRNA of ~2.4 kb transcribed in a telomere-to-centromere direction and for a protein of 433 amino acids and approximately 48 kDa. LKB1 is widely expressed in embryonic and adult tissues, albeit with some differences. In mouse, early embryos have high levels of Lkb1 expression in embryonic and extra-embryonic tissues and, along with progressive embryonic development, Lkb1 protein concentrates in the heart, esophagus, pancreas, kidney, colon, lung, small intestine and stomach (1, 2). In adult tissues, high levels of LKB1 protein can be found in most epithelia, in the follicles and corpus luteum of the ovary, the seminiferous tubules of the testis, in myocytes from skeletal muscle, and in glial cells (2-3). The LKB1 gene gives rise to at least two transcripts (LKB1L and LKB1S) from an alternative splice site at exon 9 (4). Both isoforms are widely expressed in rodent and human tissues, although LKB1S is particularly abundant in haploid spermatids in the testis. LKB1 maps to 19p13.3, a chromosomal region to which the autosomal dominant Peutz-Jeghers syndrome (PJS) (OMIM 175200) was first linked by genetic analysis. Individuals with PJS typically exhibit mucocutaneous melanin pigmentation and suffer from hamartomatous polyps in the gastrointestinal tract and an increased risk of developing cancer (5-6). Following the linkage of PJS to chromosome 19p13, the identification of germ-line-inactivating mutations at LKB1 in these patients was definitive in the conclusion that LKB1 is the gene causing the syndrome (7-8). The LKB1 protein kinase LKB1 is a multi-tasking tumour suppressor serine-threonine kinase (9-10) and is a member of the CAMK (Calcium/ Calmodulin Regulated Kinase-Like) family of kinases. LKB1 orthologues include Xenopus laevis egg and embryonic kinase 1 (XEEK1)(11), mouse Lkb1 (12), Caenorhabditis elegans partitioning-defective gene 4 (Par-4)(13) and Drosophila lkb1 (dLKB1) (14). Par-4 and dLKB1 share 26% and 44% overall identity with human LKB1, respectively, and share 42% and 66% identity with the LKB1 kinase domain, respectively. Post-translational modifications Phosphorylation is a post-translational modification that alters the chemical properties of a protein allowing it to recognize, bind, activate, or

LKB1 kinase in cancer 73

deactivate substrate. Mouse Lkb1 is autophosphorylated on the residues Thr185, Thr189, Thr336 and, Ser404, and phosphorylated by upstream kinases on Ser31, Ser325, Thr366, and Ser431 (15-18). The residues Thr366, Ser404, and Ser431 in the mouse sequence correspond to residues Thr363, Ser402 and Ser428 respectively, of human LKB1 (19) (Figure 1).

Figure 1. Post-translational modifications of murine LKB1 tumour suppressor kinase. The kinase domain is depicted in yellow, noncatalytic domains are depicted in orange and the nuclear localization sequence (NLS) is depicted in white. Autophosphorylation sites are depicted in black and sites phosphorylated by upstream kinases are depicted in red (modified from Alessi et al. 2006). Currently there is no evidence that phosphorylation of LKB1 is necessary for its activity (15-16), thus the function of LKB1 may be indirectly regulated. LKB1 is phosphorylated at Ser431 by cAMP-dependent protein kinase A (PKA) and p90 ribosomal S6 kinase (RSK) (15-16). An earlier study suggested that phosphorylation at Ser431 is required for LKB1 mediated growth arrest (16). However, a recent study reported that the LKB1 mutants S431A, S431E, and the splice variant LKB1S, that lacks Ser431, retained the ability to cause cell cycle arrest (20), indicating that phosphorylation on Ser431 is not required for this function. A role for phosphorylation of Ser431 has been implicated in axon differentiation (21), and neuronal polarization (22). In Xenopus, the LKB1 homologue XEEK1, is also phosphorylated by protein kinase A (PKA), and increased PKA activity is important for maintaining the oocytes at the G2 phase of cell cycle (15). However, phosphorylation of Ser431 is not essential for LKB1 catalytic activity, since mutagenesis of this residue does not alter the ability of LKB1 to autophosphorylate, or activate TP53 (16), AMP-activated protein kinase (AMPK) or brain-specific kinases 1 and 2 (BRSK1 and BRSK2 respectively) (20). Neither the cellular localization nor prenylation of LKB1 is affected by phosphorylation of Ser431. Phosphorylation at Ser31, Ser325 and Thr366 are not necessary for the catalytic activity of LKB1 since mutations to alanine to prevent phosphorylation, or to glutamic acid to mimic phosphorylation, or

farnesylation


addition of phosphatases did not affect LKB1 autophosphorylation, and its ability to activate TP53 (17). On the other hand, LKB1 autophosphorylation at Thr336, is important for LKB1-mediated growth arrest in G361 cells. Activation of ataxia telangiectasia (ATM) by ionizing radiation, leads to double strand breaks, and phosphorylation of LKB1 on Thr366 (18) (Figure 1). Phosphorylation at this site also plays a role in LKB1 mediated cell cycle arrest. The significance of LKB1 autophosphorylation at Thr185, Thr189, Ser404 is currently not known. Moreover, the upstream kinase responsible for phosphorylating LKB1 at S31 and S325 is not known, but has been proposed to be AMPK, a proline-directed kinase (23), where in LKB1, a proline residue follows Ser325 (17). A recent study indicates that LKB1 is acetylated on a minimum of nine lysine residues (24), with acetylation at Lys48 being the most significant. Sirtuin 1 (SIRT1), a NAD-dependent protein deacetylase, deacetylates LKB1 at Lys48, resulting in an increased LKB1 interaction with STRAD. This in turn shifts the LKB1 cytoplasmic to nuclear distribution ratio in favour of the cytoplasm, ultimately increasing LKB1 catalytic activity. Mutations of Lys48 to Arg, to mimic effects of deacetylation, resulted in similar effects as overexpression of SIRT1. Overexpression of SIRT1, and consequently deacetylation of LKB1 Lys48, resulted in increased levels of phospho-Ser428- and phospho-Thr336- LKB1. Mutations of Lys48Arg displayed similar effects. As discussed above, phosphorylation of these residues (human Ser428 and Thr336) is required for LKB1 growth suppression function. LKB1 catalytically deficient mutants: Oncogenic mutants A previous study from the Marignani laboratory demonstrated a novel function for LKB1 catalytically deficient mutants (25). Since these LKB1 mutants are present in epithelial-derived cancers and PJS, it is important to understand the biological activity of these mutants within the context of disease. Marignani and colleagues discovered that LKB1 catalytically deficient mutants, such as SL26 and D194A, display oncogenic properties, whereby LKB1 mutants act as oncogenes or LKB1 mutants enhance expression of an oncogene (25). In a series of experiments, LKB1 catalytic deficient mutants (referred to as oncogenic mutants) enhanced the expression of the oncogene cyclin D1, which is required for cells to progress from G0 phase to G1 phase of the cell cycle. This study shows that the introduction of LKB1 oncogenic mutants into colorectal DLD1TP53-/-,P21WAF-/- cells resulted in elevated gene and protein levels of cyclin E, RB and cyclin D1, three proteins required for the progression of cells through S-phase of cell cycle. The cell cycle profile of these cells were unaltered following expression of ectopic TP53 or P21WAF,


suggesting LKB1 is capable of arresting cells in G1 independent of TP53 and P21WAF expression and LKB1 catalytic activity is required for this arrest, since LKB1 oncogenic mutants did not induce the G1 arrest. In this same study (25), chromatin immunoprecipitation experiments were conducted to demonstrate that endogenous LKB1, but more importantly LKB1 oncogenic mutants, were recruited to response elements within the endogenous cyclin D1 promoter, namely the AP1, Sp1 and CRE binding protein/Ets sites, but not other sites, such as the E2F site. The authors postulate that in cancers where LKB1 is mutated, LKB1 oncogenic mutants are recruited to the promoter of cyclin D1, resulting in increased cyclin D1 expression, cell cycle progression and, cell division; perpetuating the uncontrolled cell growth characteristic of hamartomatous polyps. These findings represent the first characterized function for LKB1 oncogenic mutants. Interestingly, mutants of another tumour suppressor, TP53, have also been shown to exert oncogenic properties, and are therefore often referred to as ‘gain-of-function’ mutants (26). Future studies of mutant tumour suppressor proteins may lead to the discovery of other gain-of-function mutants, thereby providing insight into how mutants of tumour suppressor proteins may function in the context of disease. LKB1 interacting partners After the discovery of LKB1 as the gene responsible for PJS, the discovery of LKB1 signalling partners was pursued. LKB1 interacts with numerous proteins, highlighting the multifunctionality of this protein (Figure 2), as discussed below. Brahma-related gene 1 (BRG1) In an effort to identify a binding partner for LKB1, Marignani and colleagues (27) developed an in vitro expression cloning strategy. A HeLa cell cDNA library was in vitro transcribed and translated (TnT) in the presence of radiolabelled methionine. The TnT products were incubated with recombinant GST-LKB1 and GST fusion proteins, and subjected to a pull-down using GSH beads. A single TnT pool containing a 35-kDa protein that bound specifically to LKB1 was further identified by sib-selection as Brahma-related gene 1 (BRG1). BRG1 is a member of the mammalian SWI-SNF chromatin remodelling complex and contains ATP-dependent helicase activity required for chromatin remodelling. To confirm the LKB1-BRG1 interaction in vivo, immunoprecipitation (IP) of LKB1 from Saos-2 cells was conducted, and BRG1 was identified in the IP complex, indicating endogenous LKB1 and endogenous BRG1 interact


in vivo. To identify the interface between LKB1 and BRG1, a series of LKB1 and BRG1 truncations were generated. LKB1 was found to bind exclusively to the helicase domain of BRG1 while the N-terminus of LKB1 was found to be sufficient for LKB1 to bind the BRG1 helicase domain. Next, it was confirmed that the oncogenic mutant SL26 interacted with BRG1, indicating kinase activity is not required for this interaction. Functional studies determined that the association of LKB1 with BRG1 in the presence of DNA increases the ATPase activity of BRG1 approximately 6-fold compared to BRG1 alone. SL26 increased BRG1 ATPase function, indicating kinase activity of LKB1 is not required for BRG1-ATPase activity. Since LKB1 was previously shown to induce G1 growth arrest (28) and since BRG1 is involved in the retinoblastoma (RB)-dependent growth arrest pathway (29-30), Marignani and colleagues determined that LKB1 co-expressed with BRG1, induced growth arrest at the same level as BRG1 alone. However, co-expression of SL26 and BRG1 lead to cell growth and proliferation (27), suggesting that LKB1 is required for BRG1-mediated growth arrest, and LKB1 kinase activity is required for this function.

Figure 2. LKB1 interacting partner. A schematic representation of LKB1 interacting partners and signalling pathways. Dashed lines (---) indicate interacting partners, solid arrows ( ) indicate phosphorylation, dotted arrows (-->) indicate downstream signalling effects. LKB1 interacting protein 1 (LIP1) Smith and colleagues (31) used a yeast-two-hybrid screen to identify potential interacting partners for LKB1 and found the novel protein LKB1 interacting partner-1 (LIP1). LIP1 is approximately 1,100 amino acids, with a molecular weight of about 121 kDa and was found to be ubiquitously

phosphrorylate mTOR


expressed in a wide variety of tissues. Co-immunoprecipitation experiments demonstrated that LKB1 and LIP1 interact in COS cells and LKB1 kinase activity was not required for this interaction since D194A catalytic deficient mutant co-immunoprecipitated along with LIP1. Kinase assays using the immunoprecipitated complexes showed that levels of LIP1 phosphorylation did not increase in the presence of LKB1, although low levels of LIP1 phosphorylation were seen. The authors suggest this is likely due to co-immunoprecipitation of an unidentified kinase, since similar levels of phosphorylated LIP1 were observed in the absence of LKB1. Thus LIP1 was determined not to be a substrate for LKB1. Immunofluorescent microscopy demonstrated that the LIP1 localized to the cytoplasm in punctuate structures, suggesting LIP1 may be localizing to a particular organelle. However, ectopic expression of LKB1 and LIP1 led to the re-localization of LKB1 to the cytoplasm, co-localizing with LIP1. This study also showed that LIP1 co-immunoprecipitates with SMAD4, a TGFβ-regulated transcription factor. LKB1 does not directly interact with SMAD4; suggesting LIP1 acts as an adaptor between LKB1 and SMAD4. Having shown that LIP1 facilitates in the re-localization of LKB1, the authors suggest that LIP1 may play a role in regulating LKB1 function by contributing to LKB1 subcellular localization, allowing LKB1 to act upon both nuclear and cytoplasmic substrates. In summary, Smith and colleagues identified another LKB1 interacting partner, which regulates LKB1 function through partial control of LKB1 subcellular location. STRAD and MO25

Another series of studies, using a combination of yeast two-hybrid analysis and affinity purification experiments, have shown that LKB1 exists in a ternary complex with STE20-related adaptor (STRAD) (32) and mouse protein 25 (MO25) (33) in mammalian cells. STRAD is homologous to the STE20 family of kinases but lacks several key catalytic amino acid residues within its kinase domain (residues 58-401) (34), namely the glycine in the glycine-rich loop (subdomain I), the lysine residue of VAIK (subdomain II), the catalytic Asp residue of HRD (subdomain VIb), a conserved Asn residue (subdomain VIb) as well as the DFG motif in subdomain VII (35), thus is referred to as a pseudokinase (35, 36). Van Aalten and colleagues (35) reported the structure of STRADα in complex with MO25α, forming a heterodimer. STRADα was previously determined to bind to MO25α through its C-terminal Trp-Glu-Phe residues (WEF motif) (33), however, recent structural analysis confirms extensive interface with MO25α through seven structurally similar α-helical repeats that form a horseshoe shaped concave surface (36). Although STRADα lacks the prerequisite residue for catalytic


activity, STRADα adopts an active conformation when bound to ATP. While the affinity of STRADα for ATP is enhanced by MO25α, the binding of both MO25α and ATP to STRADα, are essential for the activation of LKB1 (36). LKB1-STRADα-MO25α heterotrimeric complexes (Figure 3) can be purified from mammalian cells, with each component in approximately equal stoichiometry (33). When in complex with STRADα and MO25α, LKB1 becomes catalytically active, thus unlike most serine-threonine kinases LKB1 does not require the phosphorylation of its conserved threonine in the T-loop for activity (32-35). Individually, STRADα and MO25α are expressed in both the cytoplasm and nucleus; however when co-expressed, both are located exclusively in the cytoplasm. Expression of LKB1 with STRADα and MO25α results in extensive re-localization of LKB1 to the cytoplasm (32-33). Most oncogenic mutants of LKB1 that have been identified in human cancers are unable to interact with STRADα and MO25α (37), thus the majority of these mutants remain localized in the nucleus. The interaction of STRADα and MO25α with LKB1 increases the catalytic activity of LKB1 approximately 10-fold (32). As previously mentioned, LKB1 induces a G1 cell cycle arrest, while LKB1 oncogenic mutants cannot (28,38); this difference may be attributed to the fact that LKB1 oncogenic mutants do not interact with STRADα-MO25α, since the majority of mutations are localized to the kinase domain of LKB1. STRADα interaction is required for LKB1-induced G1 arrest, since siRNA-mediated knockdown of STRADα expression prevented LKB1 from inducing G1 arrest (32). LKB1 can also form a complex with heat shock protein 90 (Hsp90) and the Cdc37 kinase-specific targeting subunit for Hsp90 (39). It is likely that these are two separate complexes, since STRADα-MO25α immunoprecipitate with LKB1, but not with Cdc37 or Hsp90 (19). The authors of this study suggest that Hsp90 and Cdc37 may facilitate in the formation of the LKB1-STRADα-MO25α complex (39). A more complex role for STRADα has recently been identified. A series of experiments by Dorfman and colleagues (34) demonstrated that STRADα plays an important role in the nucleo-cytoplasmic shuttling of LKB1. The authors showed that STRADα prevents LKB1 from binding to importinα/β, preventing LKB1 from entering the nucleus. Furthermore, STRADα exits the nucleus in a CRM1- (exportin1) mediated fashion. LKB1 is exported from the nucleus in complex with STRADα. LKB1 alone is not exported by CRM1. STRADβ can prevent the binding of LKB1 to importinα/β but cannot facilitate nuclear export of LKB1 through CRM1, indicating functional differences for STRADα and STRADβ. STRADα when in complex with LKB1 can also be exported from the nucleus by exportin7; this is the first protein identified as cargo for both CRM1 and exportin7. Both STRADα and


MO25α can passively diffuse in and out of the nucleus through nuclear pores. Due to the presence of various STRAD isoforms and splice variants, it is possible that when LKB1 is in complex with different STRAD isoforms, LKB1 localization will vary, providing another means of control of LKB1 function. This was recently demonstrated by Marignani and colleagues (40) that identified eleven different STRADα isoforms isolated from human colorectal cancer cell lines, each of which differently interact with LKB1 and MO25α. Levels of LKB1 activity and localization also varied with the different STRADα isoforms. The authors of the study postulate that in PJS where LKB1 expression is normal, the syndrome manifests due to the presence of a STRADα isoforms that render LKB1 catalytically deficient (40).

Figure 3. Schematic representation of how STRADα/MO25α may interact to activate LKB1. The model is based on known mutagenesis and structural data detailed in Zeqiraj et al. 2009. Binding of either ATP and/or MO25a to STRADα induces STRADα to adopt a closed conformation, leading to the assembly of a fully active LKB1 complex. From Zeqiraj et al, PloS Biology, 2009. Other interacting partners of LKB1 In addition to the proteins described above, LKB1 has been reported to regulate the expression of transcriptions factors. LKB1 associates with LIM domain only 4 (LMO4), an adaptor that assembles transcription factors such as GATA-6 (41). LKB1 kinase activity enhances GATA-6 mediated transactivation in HeLa cells. Furthermore this study demonstrated that, in contrast to previous reports, LKB1 enhances the promoter activity of P21 WAF1 in the presence of LIMO4, GATA-6 and LDB1, in a TP53 independent manner (41). LKB1 along with its downstream kinase, the salt inducible kinase (SIK), regulates another transcription factor, cyclin AMP response element binding protein (CREB) by phosphorylating and repressing its coactivator, transducer of regulated CREB activity (TORC) (42). It has been reported that Lkb1 null fibroblasts from mice exhibit high expression levels of signal transducer and activator of transcription 3 (Stat3)-regulated genes such as,


matrix metalloproteinase 2 (Mmp2), Mmp9, vascular endothelial growth factor (Vegf), insulin-like growth factor binding protein 5 (IGFBP5) and Cyclooxygenase-2 (COX-2) (43-44). This has been attributed to the suppression of the oncogenic transcription factor Stat3 by Lkb1 and its kinase deficient mutants, which interact with Stat3 and reduce its association with target promoters (46). An alternative mechanism for the enhanced expression of COX-2 in PJS hamartomas (47), has been ascribed to the regulation of the transcription factor polyomavirus enhancer activator 3 (PEA3) by LKB1. LKB1 associates with PEA3, induces its phosphorylation, and targets it to the ubiquitination and proteosome-dependent degradation pathway (48). Hence LKB1 plays a role in gene transcription through the regulating of transcription factors. LKB1 gene inactivation in human cancer There is no doubt that LKB1 is an important tumour suppressor gene involved in cancer development. Its genetic inactivation is not only the cause of PJS but also a critical step in the development of some forms of sporadic tumours. Since PJS patients have an increased risk of developing several types of malignancies, mutations at LKB1 have been sought in a wide variety of sporadic tumours. Tumour-specific LKB1 alterations have been identified in many tumour types, although their frequency is apparently low (49). Intriguingly, in lung tumours of the non-small-cell lung cancer (NSCLC) type, especially adenocarcinomas, LKB1-inactivating mutations are commonly detected, occurring in from one-third to a half of cases (50-51). The type and pattern of mutations in LKB1 in the PJS and in lung cancer have been extensively reviewed elsewhere (49,52). These observations are fully consistent with the high frequency of loss of heterozygosity (LOH) at chromosome 19p13, reported in lung primary tumours and cancer cell lines (53). In tumours, frequent LOH at a particular chromosomal region constitutes a classic hallmark for pinpointing the location of a tumour suppressor gene. Thus, it can be concluded that LKB1 is one of the tumour suppressor genes targeted by the high frequency of LOH at chromosome 19p in lung cancer. More recently, another tumour suppressor on chromosome 19p13.2, BRG1 (also called SMARCA4), has been found to be frequently inactivated by somatic mutations in lung cancer cell lines (54), indicating that LOH at chromosome 19p in lung cancer targets two distinct tumour suppressors (55). The reasons for the differences in the frequency of LKB1 mutations between lung tumours and other tumour types are not yet understood and are puzzling because lung cancer is not among the commonest tumours arising in PJS. There are several possible explanations for these differences, and these have been discussed elsewhere (49).


The pattern of mutations in LKB1 in tumours of sporadic origin and in those arising in PJS is that of a classic tumour-suppressor gene. First of all, mutations are homozygous, as predicted by Knudson’s two hit hypothesis (56). Usually, mutations or small deletions/insertions arise in one of the alleles while the remaining allele is lost by LOH. PJS patients are born with an LKB1 alteration (first hit) and the remaining allele is altered (second hit) during tumour development. In contrast, in tumours of sporadic origin both alleles arise during tumour development. LKB1 mutations are scattered throughout exons 1 to 8, but no mutations in exon 9 have so far been identified. There is a large proportion of nonsense, indel, frameshift and intronic mutations in splicing-conserved sites, as well as large deletions. These changes predict the generation of truncated and, therefore, completely inactivated LKB1. Some missense mutations, which lead to aminoacid changes, have also been found, especially in PJS patients. Most of these are substitutions of highly conserved aminoacids in the kinase domain and abrogate the activity of LKB1 (52). In lung cancer, LKB1 mutations occur preferentially in smokers and are found concomitantly with alterations at other cancer genes such as TP53, PIK3CA, MYC, CDKN2A and KRAS, but not with EGFR mutations (57-58). Moreover, LKB1 inactivation is less frequent in lung adenocarcinomas from patients of Asian origin (57-59), which exhibit a significantly higher frequency of EGFR alterations than found in lung adenocarcinomas arising in patients from western countries. LKB1 mouse models and cancer Genetic mouse models have provided insight into molecular mechanisms by which LKB1 elicits its tumour suppressor function. The first Lkb1 mouse model was developed by Mäkelä and colleagues in 2001, targeted disruption of Lkb1 resulted in embryonic lethality at midgestation characterized by neural tube defects, and abnormal vasculature due to aberrant regulation of vascular endothelial growth factor (Vegf) expression (60). A second Lkb1 gene knockout mouse model generated by Taketo and colleagues (61) characterized hamartomatous tumours in heterozygous Lkb1+/- mice that were histologically similar to PJS hamartomas. Analysis of the gastrointestinal hamartomas confirmed expression of wild-type Lkb1 and targeted Lkb1 allele, thus suggesting that the initiation and development of polyps was due to Lkb1 haploinsufficiency and not due to Lkb1 loss of heterozygosity (61). In addition to gastrointestinal hamartomas, greater than seventy per cent of the male Lkb1+/- mice developed hepatocellular carcinomas (HCC) compared to approximately twenty per cent of the female Lkb1+/- mice (62). Furthermore, histopathology confirmed the presence of trabecular, pseudoglandular and


sarcomatous forms of HCC, that shared histological characteristics with human HCC and displayed biallelic inactivation of Lkb1(62). Since knockout mouse models of the Lkb1 gene were found to be embryonic lethal, DePinho and colleagues developed a mouse model that carried a conditional Lkb1 allele (Figure 4) (43). Similar to previous models, conditional Lkb1-/- mice were embryonic lethal, while conditional Lkb1+/- mice developed polyps that were histologically similar to PJS polyps. Like the polyps from people with PJS (63-64), polyps from conditional Lkb1+/- mice were not positive for activating ras mutations however, Lkb1-/- mouse embryonic fibroblasts (MEFs) were sensitive to Ras-induced senescence and activated ras-mediated transformation (43). Inconsistent with previous Lkb1 knockout models (60-61), 25% of the conditional Lkb1+/- mice displayed LOH at the Lkb1 locus. The reason for the discrepancy in Lkb1 LOH between Lkb1 knockout and conditional Lkb1 mouse models is not known.

Figure 4. Targeting strategy and analysis of Lkb1. a) Lkb1 genomic structure and recombinant alleles. The 3' untranslated region (black bars), and noncanonical splice sequences (asterisks) are indicated. The transcript from the Lkb1 null allele eliminates exons 2–6, resulting in a translational frameshift (crosshatched bars). b) Southern analysis of DNA from ES cell clones showing targeting of the Lkb1 locus. c) Southern analysis of the lox, null (-), and wild-type (+) alleles. d) Southern analysis of DNA from Lkb1lox/- MEFs, untreated (-) or infected with a Cre retrovirus (+). e) Western analysis of Lkb1lox/- (-/-) and wild-type (+/+) MEFs after Cre expression, using antibodies to the amino (left) and carboxy (right) termini of Lkb1. From Bardeesy et al, Nature 2002.


In 2005, Sakamoto and colleagues (65) generated mice conditional for expression of Lkb1 by replacing exons 5-7 of the Lkb1 gene with a cDNA cassette encoding exons V-VIII and IXb, flanked by loxP Cre excision sequence (Figure 5) (65). Surprisingly, male Lkb1fl/fl mice were sterile however a plausible explanation was made available several years later with the discovery of an alternative splice variant of Lkb1 located at exon IXa, referred to as Lkb1 short form (Lkb1S). Lkb1S was found to be highly expressed in testis whereby male mice lacking Lkb1S were found to be sterile (66) and presented with defects in spermatogenesis (67).

Figure 5. Generation of LKB1-deficient mice. a) Diagram illustrating the positions of exons 3–8 (█) in the wild-type Lkb+Cre- allele. In the Lkb1flCre- allele, exon 4 of the Lkb gene is flanked with loxP Cre recombinase excision sites (▲), and exons 5–7 encoding the catalytic domain of Lkb1 are replaced with a cDNA construct encoding the remainder of the Lkb1 sequence, as well as a neomycin (Neo) selection gene. The expression of the neomycin gene is driven by the Lkb1 promoter and it is made as fusion mRNA with Lkb1. Its translation is directed by an internal ribosome entry site. In the Lkb1flCre+ allele, exons 4–7 of the Lkb1 gene are deleted through action of the Cre recombinase, there by ablating functional expression of LKB1. The positions of the PCR primers used to genotype the mice are indicated with arrows. b) Breeding strategy employed to generate Lkb1fl/fl and Lkb1-muscle-deficient (Lkb1fl/flCre+/-) mice, where MCKCre denotes transgenic mice expressing the Cre recombinase under the muscle creatine kinase promoter. The number and percentage of each genotype obtained are indicated. From Sakamoto et al, EMBO J, 2005.


The strategy used by the Alessi’s laboratory (65) to generate a floxed Lkb allele, resulted in no expression of testicular Lkb1S, thus explaining the sterility of these male mice. In addition to male sterility, protein expression in Lkb1fl/fl mice was found to be five- to ten-fold reduced compared to Lkb1+/+,

indicative of a hypomorphic phenotype for homozygous Lkb1 mice, whereas Lkb1+/fl mice expressed two-fold less Lkb1 than Lkb1+/+ mice (65). When Lkb1fl/fl mice were crossed with mice transgenic for Cre-recombinase under the muscle creatine kinase promoter (MCKCre), resulting progeny had Lkb1 excised from skeletal and heart muscles prior to birth (68). AMPKα2 activation was evaluated in Lkb1 lacking skeletal muscle and found to be significantly lower than AMPKα2 activity in skeletal muscle from wild-type or Lkb1+/fl mice (65), despite comparable AMPKα2 expression levels. Interestingly, AMPKα1 activity was modestly increased by 2-fold in Lkb1-deficient muscle. These observations were explained by the possible contamination of nonmuscle cells in the sample preparations (65). Finally, since AMPK activation during muscle contraction regulates the balance of energy, ADP:ATP and AMP:ATP ratios were evaluated in contracting tibialis anterior and extensor digitorum longus muscles. In muscles lacking expression of Lkb1, ADP:ATP and AMP:ATP ratios were significantly elevated compared to wild-type, Lkb1+/fl, or hypomorphic Lkb1fl/fl mice. Overall, the work from Alessi and colleagues (65) confirms a critical role for Lkb1 in AMPK-mediated activation of skeletal muscle and energy balance. The loss of LKB1 expression in pancreatic cancer is well described in the literature with neoplasms ranging from ductal adenocarcinomas, to intraductal papillary mucinous neoplasia and serous cytadenomas (69-70). Conditional ablation of Lkb1 expression in mice was conducted to determine the functional role of Lkb1 in the development of the pancreas (71). Genetic cross between conditional null allele of Lkb1 and transgenic Pdx1-Cre (expressed in pancreatic progenitors between E8-E12) resulted in initially normal development, however shortly after birth, these mice exhibited increased weight gain, elevated serum amylase, steatorrhea and by 10 weeks, all mice became increasingly cachetic and therefore euthanized (71). Analysis of the abdominal cavity revealed complete replacement of pancreatic tissues with cystic masses that closely resembled serous cystadenoma observed in PJS. The penetrance of pancreatic disease was complete in these mice compared to heterozygous Lkb1-Pdx1-Cre or wild-type mice (71). These findings confirm a functional role for Lkb1 in pancreas development. Since polarized organization of acina is necessary for directional secretion of zymogen from the pancreatic ductal network and given that LKB1 is reported to be involved in mediating cell polarity (13-14, 72), as it is discussed later, acinar development was evaluated in the Lkb1-Pdx1-Cre mice (71). There


was a loss of acinar polarity in pancreatic tissues from Lkb1-Pdx1-Cre at E18.5, as determined by loss of basal positioning of nuclei, lateralization of microvilli and a reduction/loss of tight and/or adherents junctions, was observed. Overall, the authors of this study establish a physiological role for LKB1 in pancreatic cellular polarity and the maintenance of ductal intregrity. LKB1 and AMPK: Cellular energetic metabolism and the cancer connection

As a tumour suppressor, the restitution of wild type LKB1 has the ability to suppress the growth of tumour cells that lack the LKB1 protein (28, 73). The molecular mechanisms underlying the tumour-suppressor function of LKB1 are still not completely understood, although it seems plausible that it depends on its serine-threonine kinase activity. The best characterized LKB1 substrate is the AMP-activated protein kinase (AMPK), a so-called 'metabolic master switch'. An increase in the intracellular AMP:ATP ratio triggers the phosphorylation and activation of AMPK, which is allosterically activated by AMP and subsequently phosphorylated at Thr172 by upstream kinases such as LKB1 (74). Activation of AMPK modulates the activity of multiple downstream targets in normalizing ATP levels, e.g. stimulating fatty acid oxidation to generate more ATP and simultaneously inhibiting ATP-consuming processes including fatty acid and protein synthesis (75). Enzymes essential for free fatty acid synthesis, including fatty acid synthase and acetyl CoA carboxylase (ACC), can be inhibited by AMPK by virtue of direct phosphorylation and/or regulation of transcription. The upstream kinase that phosphorylates and activates AMPK (AMPKK) was a long-sought protein. The first evidence suggesting that LKB1 is an AMPKK came from the work by Hong et al. (76), who searched for kinases that triggered the phosphorylation/activation of Snf1, the homolog of AMPK in yeast. They identified three yeast kinases, Pak1p, Tos3p and Elm1p, that activate Snf1 kinase in vivo. Moreover, Tos3p was found to be able to phosphorylate and activate mammalian AMPK, suggesting functional conservation of the upstream kinases between yeast and mammals. Finally, it was shown that LKB1, which is related to Tosp3, is able to phosphorylate and activate AMPK in vitro and could be the long-sought mammalian AMPKK. Almost simultaneously it was described that LKB1, in a complex with STRAD alpha/beta and MO25 alpha/beta, activates AMPK via phosphorylation of Thr172. The heterotrimeric complex was essential for the proficient activation of AMPK in vitro and in vivo (74, 77). In further support of the AMPKK activity of LKB1, the AMPK-activating drugs 5-aminoimidazole-4-carboxamide ribonucleoside (AICAR) and phenformin


does not activate AMPK in the cervical cancer-derived and LKB1-mutated HeLa cells or in lung cancer-derived LKB1-mutant cells, but activation can be restored by stably expressing wild type, but not catalytically inactive, LKB1 (33, 78). Finally, the comparison of the expression profiles among lung tumours according to LKB1 mutation status revealed that LKB1-mutant tumours has differential levels of many transcripts, including components of the PTEN/PI3K pathway as well as enzymes that participate in AMP metabolism, such as AMPD3 (adenosine monophosphate deaminase, isoform E) and APRT (adeninephosphoribosyl transferase) (79), further linking LKB1 function with that of AMPK. How LKB1 and AMPK inactivation contribute to cancer development is still not completely clear. Abrogation of energetic checkpoints may be required for a cancer cell to ensure the maintenance of cell survival and growth in an environment with limited access to nutrients or oxygen. In a normal cell, these circumstances will trigger cessation of cell growth or apoptosis due to the compromised availability of ATP. Another link between the LKB1-AMPK pathway and cancer development is the tuberin protein, which is the product of the tuberous sclerosis complex 2 gene (TSC2), which represses of mTOR when activated by AMPK (80-81). Germ-line mutations of TSC1 or TSC2 genes cause the tuberous sclerosis syndrome, which has similarities with PJS, including the presence of hamartomas (82). Wild type LKB1 has been shown to be required for modulating AMPK activity and that of AMPK-downstream targets, including mTOR, in a variety of energetically depleted cancer cells (78, 80).

In addition to cancer, LKB1 may be important in type II diabetes mellitus, a metabolic disorder characterized by resistance to the actions of insulin to stimulate skeletal muscle glucose disposal. The involvement of LKB1 in type II diabetes may arise from its function in the activation of AMPK. Deletion of LKB1 in the liver of adult mice abolished almost all AMPK activity and resulted in hyperglycemia with increased gluconeogenic and lipogenic gene expression (83). Furthermore, metformin, a drug widely prescribed to type II diabetes patients, lowers blood glucose concentrations by indirectly promoting AMPK activity (84). The effectiveness of metformin depends on the presence of LKB1 (83). Among the drugs that activate AMPK are AICAR and the biguanides, metformin and phenformin. All of these require LKB1 activity, and so therapies based on AMPK activators would be effective only in LKB1 wild type tissues. A recent population-based case-control study showed that type II diabetic patients treated with the AMPK activator metformin were less likely to get cancer (85), suggesting that AMPK activators could also be used in cancer prevention. Although the role of these compounds in cancer prevention looks


promising, their effectiveness in patients who have already developed cancer may not be so clear. More recently, LKB1 and AMPK activity have also been related to SIRT1, a NAD-dependent protein deacetylase that regulates energetic homeostasis in response to nutrient availability. SIRT1 overexpression diminished lysine acetylation in specific lysine residues of LKB1 and concurrently increased its activity, cytoplasmic/nuclear ratio, association with the LKB1 activator STRAD and, as expected, increased AMPK and ACC phosphorylation (86). It has also been reported that resveratrol, a polyphenol that protects against metabolic disease by activating SIRT1, increased LKB1 phosphorylation at Ser(428) and, thus, AMPK activity. These effects are abolished by pharmacological and genetic inhibition of SIRT1 (24). Taken together, these findings suggest that SIRT1 functions as a novel upstream regulator for LKB1/AMPK signalling. Finally, the LKB1-AMPK signalling pathway has also been functionally associated with the BRAF oncogene. In melanoma cells, AMPK activation is suppressed by activated BRAF, V600E mutation. Conversely, downregulation of BRAF signalling triggers activation of AMPK. In these cells LKB1 was found to be phosphorylated by ERK and Rsk, two kinases downstream of BRAF, which compromised the ability of LKB1 to bind and activate AMPK (87). The role of LKB1 in the regulation of cell polarity The identification of AMPK as a key substrate of LKB1 kinase activity provides definitive evidence that the abrogation of energetic checkpoints, probably with the purpose of maintaining energetically costly processes such as DNA replication and cell division, is an obligatory event for cancer development. However, LKB1 functions as a master upstream protein kinase of many AMPK-related kinases (88), indicating that LKB1 modulates other important cell processes. Some of these AMPK-related kinases are the MAP/microtubule affinity-regulating kinases (MARK1, MARK2, MARK3 and MARK4) involved in microtubule stabilization, which may be related to the proposed role of LKB1 in cell polarity. The substrates of LKB1, their downstream targets and associated biological roles are summarized in Figure 2. Watts et al. (13) provided the first hint about the role of LKB1 in the control of cell polarity. They observed that the maternally expressed par genes were asymmetrically distributed in the first cell cycle of Caenorhabditis elegans embryogenesis. Establishing asymmetries at this early stage is essential for determining the subsequent developmental fates of


the daughter cells. They found that the par4 gene of C. elegans encodes a putative serine-threonine kinase that was similar to LKB1 and to the Xenopus egg and embryo kinase, XEEK1. In support of this, the protein encoded by the homolog of par4 in Drosophila, lkb1, has also been implicated in embryonic polarity, since it is required for the early anterior-posterior (A-P) polarity of the oocyte and for the repolarization of the oocyte cytoskeleton that defines the embryonic A-P axis (14). Finally, in mammalian development, it has been reported that Lkb1 is asymmetrically localized to the animal pole of the mouse oocyte and that it associates with the microtubules at metaphase during oocyte maturation, indicating that Lkb1 protein participates in the polarization of the mouse oocyte and in the regulation of the asymmetry of meiotic divisions during oogenesis (89). LKB1 has also been associated with the polarization of adult somatic cells of a number of cell types. In human epithelial intestinal cells it has been observed that the inducible expression of STRAD activates LKB1, which leads the individual cells to the rapid remodeling of the actin cytoskeleton whereby, in the absence of cell-cell contacts, they form an apical brush border and redistribute junctional proteins in a dotted circle at the periphery of the brush border. This implies that LKB1 can induce complete polarity in intestinal epithelial cells (72). In addition, LKB1 is required for axon differentiation and specification during neuronal polarization in the mammalian cerebral cortex through the phosphorylation and activation of brain-specific kinase, BRSK (21-22). Finally, immunostaining of LKB1 in a variety of normal human tissues has revealed apical accumulation of this protein in several types of epithelial cells (3). LKB1 has also been shown to be essential for NSCLC polarity and to colocalize at the leading edge of the cell with the small rho GTPase cdc42 and its downstream binding partner, p21-activated kinase (PAK), two key components of the polarity pathway (90). The exact mechanism and downstream targets that mediate the role of LKB1 in the control of cell polarity are still under debate. Some observations suggest AMPK and its target non-muscle myosin regulatory light chain (MRLC; also known as MLC2), which is directly phosphorylated by AMPK at an important regulatory site (91). Since the microtubule affinity-regulating kinases (MARK1, MARK2, MARK3 and MARK4) are among the AMPK-related kinases phosphorylated by LKB1 they may also mediate the role of LKB1 in cell polarity. The MARK family phosphorylates microtubule-associated proteins (tau/MAP2/MAP4), causing detachment from microtubules, and their disassembly (92-94). Interestingly, aberrant phosphorylation of the microtubule-associated protein tau is one of the pathological features of neuronal degeneration in Alzheimer’s disease, which points towards LKB1 being a candidate target for novel therapies against this neurodegenerative disease.


LKB1 in regulation of hormone receptor signalling Clinical evidence suggests a role for the tumour suppressor kinase LKB1 in the development of breast carcinoma (95-97). Early evidence in the literature suggests that a loss of LKB1 expression leads to papillary breast carcinoma (98). More recently, a study that evaluated the expression of LKB1 in 85 cases of breast cancers, the authors found that in a subset of high-grade in situ and invasive mammary carcinomas, the expression of LKB1 was completely lost as determined by immunohistochemistry (96). In a separate study, 116 patients with confirmed breast carcinoma were evaluated for LKB1 protein expression, of which expression was low in approximately one third of the patients compared to control population of women. The authors of this study concluded that low expression of LKB1 correlated with higher histological grade, tumour size, and presence of lymph node status (95). Both studies conclude that LKB1 expression profile may serve as a prognostic marker for breast carcinoma, however further investigation is warranted. These recent findings suggest a role for LKB1 in breast cancer however the molecular mechanism by which this occurs is not fully understood. Recent work from the Marignani lab (98) highlight the discovery of a novel function for LKB1, coactivator of estrogen receptor alpha (ERα) signalling. In a series of experiments, LKB1 was discovered to bind to ERα, thereby enhancing ERα-mediated transcription. Furthermore, it was determined that LKB1 catalytic activity contributes to enhanced ERα-mediated transcription and that LKB1 is recruited to the promoters of ERα-responsive genes. Clinical evidence suggests a role for LKB1 in breast cancer since low expression of LKB1 is correlated with poor clinical outcomes. This said the molecular mechanisms by which LKB1 is involved in mammary gland development and/or malignancies are not known. The significance of this work is the demonstration, for the first time, of a functional link between LKB1 and ERα, placing LKB1 in a coactivator role for ERα signalling, similar to the other tumour suppressors such as RB (99-102), P21WAF/CIP (103-104), TSC2 (105) and BRG1 (106), that are reported to have coactivator function for hormone receptors. This discovery broadens the scientific scope of LKB1 and lays the ground work for further investigation into the role of LKB1 in mammary gland development and tumourigenesis. References 1. Luukko, K., Ylikorkala, A., Tiainen, M., and Makela, T.P. 1999, Mech. Dev.,

83,187-190. 2. Rowan, A., Churchman, M., Jefferey, R., Hanby, A., Poulsom, R., and

Tomlinson, I. 2000, J. Pathol., 192, 203-6.


3. Conde, E., Suarez-Gauthier, A., Garcia-Garcia, E., Lopez-Rios, F., Lopez-Encuentra, A., Garcia-Lujan, R, Morente, M., Sanchez-Verde, L., and Sanchez-Cespedes, M. 2007, Hum. Pathol., 38,1351-60.

4. Towler, M.C., Fogarty, S., Hawley, S.A., Pan, D.A., Martin, D.M., Morrice, N.A., McCarthy, A., Galardo, M.N., Meroni, S.B., Cigorraga, S.B., Ashworth, A., Sakamoto, K., and Hardie, D.G. 2008, Biochem. J., 416, 1-14.

5. Jeghers, H., McKusic, V.A., Katz, K.H. 1949, N. Engl. J. Med., 241, 1031-1036. 6. Giardiello, F.M., Brensinger, J.D., Tersmette, A.C., Goodman, S.N., Petersen,

G.M., Booker, S.V., Cruz-Correa, M., and Offerhaus, J.A. 2000, Gastroenterology, 119,1447-53.

7. Hemminki, A., Markie, D., Tomlinson, I., Avizienyte, E., Roth, S., Loukola, A. et al. 1998, Nature, 18,184-187.

8. Jenne, D.E., Reimann, H., Nezu, J., Friedel, W., Loff, S., Jeschke, R. et al. 1998, Nat. Genet., 18,38-44.

9. Marignani, P. A. 2005, J. Clin. Pathol, 58, 15-19. 10. Hemminki, A., Tomlinson, I., Markie, D., et al. 1997, Nat. Genet., 15, 87-90. 11. Su, J.Y., Erikson, E., and Maller, J.L. 1996, J. Biol. Chem., 271, 14430-14437. 12. Smith, D.P., Spicer, J., Smith, A., Swift, S., and Ashworth, A. 1999, Hum. Mol.

Genet., 8, 1479-1485. 13. Watts, J.L., Morton, D.G., Bestman, J., and Kemphues, K.J. 2000, 127, 1467-75 14. Martin, S.G., and St Johnston, D. 2003, Nature, 421, 379-84. 15. Collins, S. P., Reoma, J. L., Gamm, D. M. and Uhler, M.D. 2000,. Biochem. J.,

345 Pt 3, 673-680. 16. Sapkota, G.P., Kieloch, A., Lizcano, J.M., Lain, S., Arthur, J.S., Williams, M R.,

Morrice, N., Deak, M., and Alessi, D.R. 2001, J. Biol. Chem., 276, 19469-19482. 17. Sapkota, G.P., Deak, M., Kieloch, A., Morrice, N., Goodarzi, A.A., Smythe, C.,

Shiloh, Y., Lees-Miller, S.P., and Alessi, D.R. 2002, Biochem. J., 368, 507-516. 18. Sapkota, G.P., Boudeau, J., Deak, M., Kieloch, A., Morrice, N., and Alessi, D.R.

2002, Biochem. J., 362, 481-490. 19. Alessi, D.R., Sakamoto, K., and Bayascas, J.R. LKB1-dependent signaling

pathways. 2006, Annu. Rev. Biochem., 75,137-163. 20. Fogarty, S., and Hardie, D.G. 2009, J. Biol. Chem., 284, 77-84. 21. Barnes, A.P., Lilley, B.N., Pan, Y.A., Plummer, L.J., Powell, A.W., Raines,

A.N., Sanes, J.R., and Polleux, F. 2007, Cell, 129, 549-563. 22. Shelly, M., Cancedda, L., Heilshorn, S., Sumbre, G. and Poo, M. 2007, Cell, 129,

565-577. 23. Songyang, Z., Blechner, S., Hoagland, N., Hoekstra, M. F., Piwnica-Worms, H.,

and Cantley, L.C. 1994, Curr. Biol., 4, 973-982. 24. Lan, F., Cacicedo, J.M., Ruderman, N., Ido, Y. 2008, J. Biol. Chem., 283,

27628-35. 25. Scott, K.D., Nath-Sain, S., Agnew, M.D. and Marignani, P.A. 2007. Cancer Res.,

67, 5622-5627. 26. Sigal, A., and Rotter, V. 2000, Cancer Res, 60, 6788-6793. 27. Marignani, P.A., Kanai, F., and Carpenter, C. L. 2001, J. Biol. Chem., 276,

32415-32418.


28. Tiainen, M., Ylikorkala, A., and Makela, T.P. 1999, Proc. Natl. Acad. Sci., USA. 96, 9248-51. 29. Dunaief, J.L., Strober, B.E., Guha, S., Khavari, P.A., Alin, K., Luban, J.,

Begemann, M., Crabtree, G.R., and Goff, S.P. 1994, Cell, 79, 119-130. 30. Strober, B.E., Dunaief, J.L., Guha and Goff, S.P. 1996, Mol. Cell Biol., 16,

1576-1583. 31. Smith, D.P., Rayter, S.I., Niederlander, C., Spicer, J., Jones, C.M., and Ashworth,

A. 2001, Hum. Mol. Genet., 10, 2869-2877. 32. Baas, A.F., Boudeau, J., Sapkota, G.P., Smit, L., Medema, R., Morrice, N.A.,

Alessi, D.R., and Clevers, H C. 2003, EMBO J., 22, 3062-3072. 33. Boudeau, J., Baas, A.F., Deak, M., Morrice, N.A., Kieloch, A., Schutkowski, M.,

Prescott, A.R., Clevers, H.C. and Alessi, D.R. 2003, EMBO J., 22, 5102-5114. 34. Dorfman, J., and Macara, I.G. 2008, Mol. Biol. Cell, 19, 1614-1626. 35. Boudeau, J., Miranda-Saavedra, D., Barton, G. J., and Alessi, D.R. 2006, Trends

Cell Biol., 16, 443-452. 36. Zeqiraj, E., Filippi, B.M., Goldie, S., Navratilova, I., Boudeau, J., Deak, M.,

Alessi, D.R., and van Aalten, D.M.F. 2009, PLoS Biol., 7, e1000126. 37. Boudeau, J., Scott, J.W., Resta, N. et al. 2004, J. Cell Sci., 117, 6365-6375. 38. Tiainen, M., Vaahtomeri, K., Ylikorkala, A., and Makela, T.P. 2002, Hum. Mol.

Genet., 11, 1497 1504. 39. Boudeau, J., Deak, M., Lawlor, M.A., Morrice, N.A., and Alessi, D.R. 2003,

Biochem. J., 370, 849-857. 40. Marignani, P.A., Scott, K.D., Bagnulo, R., Cannone, D., Ferrari, E., Stella, A.,

Guanti, G., Simone, C., and Resta, N. 2007, Cancer Biol. Ther., 6, 1627-1631. 41. Setogawa, T., Shinozaki-Yabana, S., Masuda, T., Matsuura, K., and Akiyama, T.

2006, Biochem. Bioph. Res. Comm., 19, 1186-1190. 42. Katoh, Y., Takemori, H., Lin, X. et al. 2006, FEBS Journal, 273, 2730-2748. 43 Bardeesy, N., Sinha, M., Hezel, A.F., Signoretti, S., Hathaway, N.A., Sharpless,

N. E., Loda, M., Carrasco, D.R., and DePinho, R. A. 2002, Nature, 419, 162-167. 44. Rossi, D.J., Ylikorkala, A., Korsisaari, N. et al. 2002, Proc. Natl. Acad. Sci. U.S

A, 99, 12327-12332. 45. Kim, D.W., Chung, H.K., Park, K.C., Hwang, J.H., Jo, Y.S., Chung, J.,

Kalvakolanu, D.V., Resta, N., and Shong, M. 2007, Vol. 21, pp. 3039-3049. 47. McGarrity, T., Peiffer, L., Amos, C., Frazier, M., Ward, M., and Howett, M.

2003, Am. J. Gastroenterol., 98, 671-678. 48. Upadhyay, S., Liu, C., Chatterjee, A. et al. 2006, Cancer Res., 66, 7870-7879.

49. Sanchez-Cespedes, M. 2007, Oncogene, 26, 7825-32. 50. Sanchez-Cespedes, M., Parrella, P., Esteller, M., Nomoto, S., Trink, B., Engles, J.M.,

Westra, W.H., Herman, J.G., and Sidransky, D. 2002, Cancer Res., 62, 3659-3662. 51. Carretero, J., Medina, P.P., Pio, R., Montuenga, L.M., and Sanchez-Cespedes, M.

2004, Oncogene, 23,4037–4040. 52. Launonen, V. 2005, Hum. Mutat. 26:291-297. 53. Sanchez-Cespedes, M., Ahrendt, S.A., Piantadosi, S., Rosell, R., Monzo, M., Wu,

L., Westra, W.H., Yang, S.C., Jen, J., and Sidransky, D. 2001, Cancer Res., 61, 1309-1313.


54 Medina, P., Romero, O.A., Kohno, T., Montuenga, L.M., Pio, R., Yokota, J., and Sanchez-Cespedes, M. 2009, Hum. Mutat., 29, 617-22.

55. Rodriguez-Nieto, S., and Sanchez-Cespedes, M, 2009, Carcinogenesis, 30:547-54 56. Knudson, A.G. 1971, Proc. Natl. Acad. Sci. U.S.A., 68:820–823. 57. Matsumoto, S., Iwakawa, R., Takahashi, K., Kohno, T., Nakanishi, Y., Matsuno,

Y., Suzuki, K., Nakamoto, M., Shimizu, E., Minna, J.D., and Yokota, J. 2007, Oncogene, 26, 5911-5918.

58. Blanco, R. Iwakawa, R., Tang, M., Kohno, T., Angulo, B., Pio, R., Montuenga, L.M., Minna, J.D., Yokota, J., and Sanchez-Cespedes, M. 2009, Hum. Mut., May 20 [Epub ahead of print].

59. Koivunen, J.P., Kim, J., Lee, J., Rogers, A.M., Park, J.O., Zhao, X., Naoki, K., Okamoto, I., Nakagawa, K., Yeap, B.Y., Meyerson, M., Wong, K.K., Richards, W.G., Sugarbaker, D.J., Johnson, B.E., and Janne, P.A. 2008, Br. J. Cancer, 99, 245-252.

60. Ylikorkala, A., Rossi, D. J., Korsisaari, N., Luukko, K., Alitalo, K., Henkemeyer, M., and Makela, T.P. 2001, Science, 293, 1323-1326.

61. Miyoshi, H., Nakau, M., Ishikawa, T.O., Seldin, M. F., Oshima, M., and Taketo, M. M. 2002, Cancer Res., 62, 2261-2266.

62. Nakau, M., Miyoshi, H., Seldin, M. F., Imamura, M., Oshima, M., and Taketo, M.M. 2002, Cancer Res., 62, 4549-4553.

63. Gruber, S. B., Entius, M.M., Petersen, G.M., Laken, S.J., Longo, P.A., Boyer, R., Levin, A.M., Mujumdar, U.J., Trent, J.M., Kinzler, K.W., Vogelstein, B., Hamilton, S.R., Polymeropoulos, M.H., Offerhaus, G.J., and Giardiello, F.M. 1998, Cancer Res., 58, 5267-5270.

64. Entius, M.M., Keller, J.J., Westerman, A.M., van Rees, B.P., van Velthuysen, M. L., de Goeij, A.F., Wilson, J.H., Giardiello, F.M., and Offerhaus, G. J. 2001, J. Clin. Pathol., 54, 126-131.

65. Sakamoto, K., McCarthy, A., Smith, D., Green, K.A., G. Hardie, D., Ashworth, A., and Alessi, D.R. 2005, EMBO J., 24, 1810-1820.

66. Denison, F.C., Hiscock, N. J., Carling, D., and Woods, A. 2009, J. Biol. Chem., 284, 67-76

67. Xie, Z., Dong, Y., Scholz, R., Neumann, D., and Zou, M.H. 2008, Circulation, 117, 952-962.

68. Brüning, J.C., Michael, M.D., Winnay, J.N., Hayashi, T., Hörsch, D., Accili, D., Goodyear, L.J., and Kahn, C.R. 1998, Mol. Cell, 2, 559-569.

69. Sato, N., Rosty, C., Jansen, M. et al. 2001, Am. J. Pathol., 159, 2017-2022. 70. Yee, N.S., Furth, E.E., and Pack, M. 2003, Cancer Biol. Ther., 2, 38-47. 71. Hezel, A.F., Gurumurthy, S., Granot, Z., Swisa, A., Chu, G.C., Bailey, G., Dor,

Y., Bardeesy, N., and DePinho, R.A. 2008, Mol. Cell. Biol., 28, 2414-2425. 72. Baas, A.F., Kuipers, J., van der Wel, N.N., Batlle, E., Koerten, H.K., Peters, P.J.,

and Clevers, H.C. 2004, Cell, 116,457-466. 73. Jimenez, A.I., Fernandez, P., Dominguez, O. Dopazo, A., and Sanchez-Cespedes,

M. 2003, Cancer Res, 63, 1382-8. 74. Hawley, S.A., Boudeau, J., Reid, J.L., Mustard, K.J., Udd, L., Makela, T.P. et al.

2003, J. Biol, 2,28.


75. Hardie, D.G. 2003, Endocrinology, 144, 5179-83. 76. Hong, S.P,, Leiper, F.C,, Woods, A., Carling, D., and Carlson, M. 2003, Proc.

Natl. Acad. Sci. USA., 100,8839-43. 77. Woods, A., Johnstone, S.R., Dickerson, K., Leiper, F.C., Fryer, L.G., Neumann,

D., Schlattner, U., Wallimann, T., Carlson, M., and Carling, D. 2003, Curr. Biol., 13,2004-8

78. Carretero, J., Medina, P.P., Blanco, R., Smit, L., Tang, M., Roncador, G. et al. 2007, Oncogene 26,1616-1625.

79. Fernandez P, Carretero J, Medina PP, Jimenez AI, Rodriguez-Perales, S. Paz, M.F., Cigudosa, J.C., Esteller, M., Lombardia, L., Morente, M., Sanchez-Verde, L., Sotelo, T., and Sanchez-Cespedes, M. 2004, Oncogene 23:5084-5091.

80. Shaw, R.J., Kosmatka, M., Bardeesy, N., et al. 2004, Proc. Natl. Acad, Sci. USA., 101,3329-35.

81. Corradetti, M.N., Inoki, K., Bardeesy, N., DePinho, R.A., and Guan, K.L. 2004, Genes Dev., 18,1533-1538.

82. Consortium TECS. 1993, Cell, 75,1305-1315. 83. Shaw, R.J,, Lamia, K.A., Vasquez, D., Koo, S.H., Bardeesy, N., Depinho, R.A.,

Montminy, M., and Cantley, L,C. 2005, Science, 2005; 310,1642-6. 84. Musi, N., Hirshman, M.F., Nygren, J., Svanfeldt, M., Bavenholm, P.,

Rooyackers, O., Zhou, G., Williamson, J.M. Ljunqvist, O., Efendic, S., Moller, D.E., Thorell, A., and Goodyear, L.J. 2002, Diabetes, 51:2074-81.

85. Evans, J.M., Donnelly, L.A., Emslie-Smith, A.M., Alessi, D.R., and Morris, A.D. 2005, BMJ., 330,1304-5.

86. Hou, X., Xu, S., Maitland-Toolan, K.A., Sato, K., Jiang, B., Ido, Y., Lan, F., Walsh, K., Wierzbicki, M., Verbeuren, T.J., Cohen, R.A., and Zang, M. 2008, J. Biol. Chem., 283, 20015-26.

87. Zheng, B., Jeong, J.H., Asara, J.M., Yuan, Y.Y., Granter, S.R., Chin, L., and Cantley, L.C.2009, Mol. Cell, 33, 237-47

88. Lizcano, J.M., Goransson, O., Toth, R., Deak, M., Morrice, N.A., Boudeau, J., Hawley, S.A., Udd, L., Mäkelä, T.P., Hardie, D.G., and Alessi, D.R. 2004, EMBO J., 23:833-43.

89. Szczepańska, K., and Maleszewski, M. 2005, Gene Expr. Patterns, 6, 86-93. 90. Zhang, S., Schafer-Hales, K., Khuri, F.R., Zhou, W., Vertino, P.M., and Marcus,

A.I. 2008, Cancer Res, 68, 740-8. 91. Lee, J.H., Koh, H., Kim, M., Kim, Y., Lee, S.Y., Karess, R.E., Lee, S.H., Shong,

M., Kim, J.M., Kim, J., and Chung, J. 2007, Nature, 447, 1017-20. 92. Drewes, G., Ebneth, A., Preuss, U. ,Mandelkow, E.M., and Mandelkow, E.. 1997,

Cell, 89:297-308. 93. Trinczek, B., Brajenovic, M., Ebneth, A., and Drewes, G. 2004, J. Biol. Chem.,

279,5915-23. 94. Mandelkow, E.M., Thies, E., Trinczek, B., Biernat, J., and Mandelkow, E. 2004,

J. Cell Biol., 167,99-110. 95. Shen, Z., Wen, X.F., Lan, F., Shen, Z.Z., and Shao, Z.M. 2002, Clin. Cancer

Res., 8, 2085-2090.


96. Fenton, H., Carlile, B., Montgomery, E., Carraway, H., Herman, J., Sahin, F., Su, G., and Argani, P. 2006, Mol. Morphol., 14, 146-153.

97. Esteller, M., Avizienyte, E., Corn, P.G., Lothe, R.A., Baylin, S.B., Aaltonen, L.A., and Herman, J.G. 2000, Oncogene, 19, 164-168.

98. Nath-Sain, S., and Marignani, P.A. 2009, Mol. Biol. Cell, 20, 2785-2795. 99. Singh, P., Coe, J., and Hong, W. 1995, Nature, 374, 562-565. 100. Lu, J., and Danielsen, M. 1998, J. Biol. Chem., 273, 31528-31533. 101. Yeh, S., Miyamoto, H., Nishimura, K., Kang, H., Ludlow, J., Hsiao, P., Wang,

C., Su, C., and Chang, C. 1998, Biochem. Bioph. Res. Comm., 248, 361-367. 102. Batsche, E., Desroches, J., Bilodeau, S., Gauthier, Y., and Drouin, J. 2005, J.

Biol. Chem., 280, 19746-19756. 103. Wakasugi, E., Kobayashi T, Tamaki Y, Ito Y, Miyashiro I, Komoike Y, Takeda

T, Shin E, Takatsuka Y, Kikkawa N, Monden T, and Monden M. 1997, Am. J. Clin. Pathol., 107, 684-691.

104. Fritah, A., Saucier, C., Mester, J., Redeuilh, G., and Sabbah, M. 2005, Mol. Cell Biol., 25, 2419-2430.

106. Henry, K.W., Yuan, X., Koszewski, N.J., Onda, H., Kwiatkowski, D.J., and Noonan, D.J. 1998, J. Biol. Chem., 273, 20535-20539.

107. Trotter, K., and Archer TK. 2008, Nucl. Recept. Signal, 6, 1-12.



5. The role of glycogen synthase kinase-3 in the decision between cell survival and

cell death

Roberta Venè and Francesca Tosetti Department of Translational Oncology, Cell Biology Laboratory, National Cancer Research

Institute (IST), Genova, 16132, Italy

Abstract. Glycogen synthase kinase 3 (GSK3), originally recognized for its role in glycogen synthesis, is a multifunctional kinase involved in important biological processes such as insulin sensitivity, hematopoietic stem cell repopulation, survival of tumor cells, cytoprotection in normal and pathological neurons and in cardiomyocytes. Tissue-specific effects and segregated functions within intracellular compartments (mitochondria, nucleus, cytoplasm) further contribute to branch out the activity of GSK3. Due to the extensive properties of this kinase, novel drugs targeting GSK3 are under development for potential therapeutic use in neurological disorders, diabetes and cancer. Emerging evidence indicates that GSK3 is a central regulator of a stress-activated network acting in concert with metabolic sensors of the bioenergetic requirements of the cells. The crosstalk between the PI3K/AKT/mTOR, the p90 RSK/MAP kinase (ERK1/2, JNK, p38) pathways and other key regulators of intracellular homeostasis including the transcription factors hypoxia inducible transcription

Correspondence/Reprint request: Dr. Francesca Tosetti, Cell Biology Laboratory, Istituto Nazionale per la Ricerca sul Cancro, (IST), Largo R. Benzi 10, 16132 Genova, Italy. E-mail: [email protected]

Roberta Venè & Francesca Tosetti 96

factor-1 (HIF-1), NFkB, Nrf2 and the LKB1/AMPK/malonyl-CoA fuel sensing pathway set the threshold of tolerance to stress through modification of glucose metabolism and redox balance. Modulation of GSK3 activity thus operates at the crossroad between cell survival and cell death and can arbitrate cell fate by promoting adaptive mechanisms that lead to cytoprotection against mild and transient stress, or to cell death under conditions of severe threat such as acute oxidative damage. This work provides an overview of recent findings on the role of GSK3 in tumor establishment and progression, with a special focus on the regulation of metabolic parameters related with energy and redox homeostasis in response to GSK activation or inhibition, which could be exploited as potential targets for cancer treatment. Introduction GSK3 is a ubiquitous serine/threonine kinase that controls a wide range of cellular processes, including insulin sensitivity, glucose metabolism and cell survival through regulation of energy homeostasis. GSK3 is composed of two isoforms, GSK3α and GSK3β. GSK3 is a unique kinase normally active (unphosphorylated) in unstimulated cells and is negatively regulated by phosphorylation of GSK3α at Ser21 and GSK3β at Ser9 by AKT and other kinases of the AGC (protein kinase A, protein kinase G, protein kinase C) family including p90 ribosomal S6 kinase (p90RSK) [1] upon stimulation by growth factors. Thus, GSK3 activity is kept under control by survival signals propagated by growth factors mainly through the PI3K/AKT pathway upon AKT phosphorylation at Ser473. GSK3 is also regulated by activating phosphorylation of GSK3α and β at Tyr276 and Tyr216, respectively. Phosphorylation of GSK3α at Ser21 and GSK3β at Ser9 by AKT can show non-overlapping and even opposite effects, for example in cardiac hypertrophy [2]. As a downstream target of the PI3K/AKT survival pathway, the inhibition of GSK3 triggered by tyrosine kinase (RTK) and G-protein-coupled receptors (GPCR) can block mitochondrial apoptosis [3,4]. This property is beneficial to protect from damage untransformed cells including neurons [5], cardiomyocytes [6,7] and pancreatic β cells [8]. Similarly, pharmacologic inhibitors of GSK3 can exert cytoprotective effects [9]. The cytoprotective activity of lithium chloride and numerous, more specific GSK3 inhibitors has been extensively studied in pathological neuronal cells [10,11] and in cardiomyocytes under stress [12]. Among other functions, GSK3 inhibition or, more recently, genetic deficiency has been shown to normalize glucose metabolism in animal models of insulin resistance [13] and to support pancreatic β cells survival [8], thus fostering the development of GSK3 inhibitors as antidiabetic agents [14].

GSK3 in regulation of cell fate 97

In cancer, an important function of GSK3, in particular GSK3β, is to regulate the central step in the canonical Wnt/β-catenin signaling pathway. The overactivation of this pathway has been implicated in the pathogenesis of several types of tumors, such as colon and prostate adenocarcinomas [15]. GSK3 phosphorylation stimulated by hepatocyte growth factor (HGF) and insulin-like growth factor-1 (IGF-1) through the oncogenic AKT pathway can enhance tumor cell proliferation, motility, invasion and metastatic spread [16]. Since the formulation of the Warburg hypothesis on the generation of excessive lactate in the presence of oxygen by tumor cells, numerous studies have progressively unraveled a specific metabolic signature of epithelial cells in tumors characterized by glucose usage for energy production (aerobic glycolysis) [17]. GSK3 intervenes in key molecular events that regulate glucose metabolism such as binding of overexpressed hexokinase II (HK II) to mitochondria [18] and stabilization of HIF-1 [19]. On the other hand, GSK3 has also been implicated in the inhibition of caspase 8-mediated apoptosis triggered by death receptors, thus GSK3 inhibitors can potentiate cell death propagated through the extrinsic pathway [4,20]. More in general, new data add to mounting evidence that, at least in some cell types, GSK3 inhibition can be detrimental to tumor cell survival [21-23]. This work will focus on seminal and recent studies that have shed light on GSK3 as an important node in a regulatory network involved in the adaptive response to critical extracellular conditions affecting intracellular homeostasis (Fig. 1). In particular we will discuss in which way the metabolic consequences of GSK3 activation or inhibition support cell survival or dictate cell death in cells under stress and, in particular, in transformed cells. 1. GSK3 in cell survival and cell death: A matter of timing GSK3 inhibition has been reported to induce cytoprotection in neuronal cells and cardiomyocytes. The cytoprotective properties of GSK3 inhibition appear particularly relevant in the context of the biological phenomenon defined as preconditioning [7] or hormesis [24] and in postconditioning, consisting of a series of adaptive responses to mild stress that protect cells against a subsequent, severe injury. These concepts have informed the improvement of therapeutic strategies in neuroprotection (neurohormesis) and cardioprotection to avoid the adverse effects of oxidative stress and necrotic cell death during reperfusion after cerebral or myocardial ischemia [7].


One of the central events in reperfusion is the opening of the mitochondrial permeability transition pore (mPTP). This protein complex is formed, among other less well-identified subunits, by cyclophilin D and the voltage dependent anion channel (VDAC). The involvement of adenine nucleotide translocase (ANT), hexokinase II and creatine kinase has been recently questioned [25]. mPTP controls selective exchanges of ions and metabolites between the cytoplasm and mitochondria. Opening of the mPTP located in the inner mitochondrial membrane is an on-off process that leads to mitochondrial depolarization, ATP depletion and eventually mitochondrial swelling. Mitochondrial damage and mPTP opening at reperfusion is caused by a burst of ROS mostly originating as superoxide anion during mitochondrial respiration (probably at complex 1 and 3 of the mitochondrial respiratory chain) and elevated calcium levels. Hydrogen peroxide derived by the activity of superoxide dismutase (SOD) can in turn be transformed in highly reactive hydroxyl radicals that damage mitochondrial proteins and lipids. Mild mitochondrial depolarization with limited production of ROS is one of the mechanisms of preconditioning that allows cells to better tolerate oxidative stress during ischemia/reperfusion injury. GSK3 plays a prominent role in preconditioning by elevating the threshold of tolerance to ROS damage, thus protecting mitochondrial membrane integrity against mPTP opening. Proteins forming the mPTP complex and antiapoptotic proteins regulating the susceptibility to the mitochondrial permeability transition such as Bcl2 may be the GSK3 targets involved in the cytoprotective effects [7]. The temporal dynamic of preconditioning is relevant to obtain the cytoprotective effect. Cells must have time to respond to the noxious stimulus by transiently perturbing mitochondrial respiration to produce a limited amount of ROS with prevailing signaling function. Then, following a biphasic kinetics, the adaptive response ensues. Several lines of evidence indicate that GSK3 inhibition is a key event integrating survival signals emerging from the stimulation of different molecular pathways including those mediated by insulin and IGF-1, PKC, PKA. 1.1. The role of GSK3 in cardioprotection A well-characterized experimental model of preconditioning governed by GSK3 inhibition has been described by Sollott and coworkers in cardiomyocytes [12]. The consequences of ATP depletion caused by impaired mitochondrial oxidative phosphorylation are particularly dangerous in cardiomyocytes that, differently from neurons, cannot rely on glucose as an energy substrate for ATP production and do not tolerate low pH due to lactic acid production by anaerobic glycolysis. The role of GSK3 has been


studied by analyzing the effects of different pharmacological inhibitors of the mPTP. Two pathways leading to inhibition of the mitochondrial pool of GSK3, which differ in their ability to confer transient or stable, memory-associated protection against oxidative stress have been identified by using different pharmacological agents, including lithium chloride, insulin and mitochondrial ATP-dependent K+ channel openers [7,12,25]. Phosphorylation of GSK3β at Ser9 has been proposed to elevate the threshold for mPTP opening by preserving the binding of hexokinase II to the mPTP complex [18]. At present, the exact molecular mechanism through which GSK3 inhibition elevates the threshold for mPTP activation remains undefined. However, based on the encouraging results that indicate GSK inhibition in preconditioning as a valuable tool to limit ischemia/reperfusion injury, this therapeutic approach deserves in depth exploration in further preclinical and clinical studies. 1.2. The role of GSK3 in neuronal cell survival and death The role of active GSK3 in neuronal cell death has been extensively studied in several models of neurotoxicity and neurodegeneration. Although the adaptations to limit neuronal cell loss and to maintain tissue integrity in the brain differ in acute damage such as ischemic/reperfusion and neurotoxic injury, and in chronic pathological conditions, some common mechanisms include oxidative stress and disruption of calcium homeostasis. These conditions eventually lead to impairment of mitochondrial function followed by apoptotic or necrotic cell death [26]. GSK3 has been found to regulate different molecular activators of the mitochondrial (intrinsic) apoptotic pathway [4]. In neuronal cells, GSK3 has been reported to phosphorylate Bax, favoring its mitochondrial translocation, and the mixed lineage kinase 3 (MLK3), a mitogen-activated protein kinase kinase kinase that activates the proapoptotic c-Jun N-terminal kinase (JNK) pathway [27,28]. The recent finding that Wnt-regulated GSK3β is implicated in cell death of cerebellar granule neurons induced by trophic factor deprivation assigns a new role to the Wnt GSK3 pool. GSK3 overexpression or overactivation in the brain has been found in animal models of Alzheimer’s and Parkinson’s disease [29]. Most neurodegenerative diseases are characterized by impaired mitochondrial function and glucose hypometabolism. Of note, neuroprotection can apparently be achieved by metabolic changes enhancing glucose metabolism in neurons. Two main molecular mechanisms underlying neuroprotection by enhanced glycolysis have been identified: HIF-1 induction boosting glucose flux through glycolysis and the pentose phosphate pathway [30], and


induction of HKII binding to mitochondria induced by GSK3 inhibition [31]. Enhanced glucose metabolism can result in the production of reducing equivalents in the form of NADPH. This HIF-mediated mechanism has been identified as an adaptation to tolerate oxidative stress due to β-amyloid toxicity and to contribute to the selection of a β-amyloid-resistant phenotype [30]. Paradigmatic models of neuronal cell death upon trophic factor withdrawal show that GSK3 inhibition by growth factors is an essential event to preserve neuronal cell survival [32]. In accordance with the activity of other growth factors, NGF neuroprotection is linked to GSK3 inhibition through activation of PI3K/AKT signaling and the effects of NGF deprivation have provided an informative set of data on the mechanisms implicated in GSK-mediated cell death [28]. Lithium chloride has found application in psychiatry as a mood stabilizer in manic-depression illness (bipolar disorders) for over fifty years. In 1996, Woodgett and coworkers discovered that lithium chloride is a pharmacological inhibitor of GSK3 [33]. Since then, numerous GSK3 inhibitors other than lithium, including alsterpaullone, kenpaullone, SB216763 and AR-A014418 have been synthesized and reported to protect neurons against a variety of stressful conditions. The effects of lithium and other GSK3 inhibitors at molecular level have been studied in in vitro and in vivo experimental models recapitulating oxidative neuronal damage found in Parkinson’s and Alzheimer’s disease and other injuries including hypoxia-ischemia in the immature brain, oxygen-glucose deprivation, potassium deprivation, glutamate excitotoxicity. Some of the biological effects of lithium cannot be ascribed to GSK3 inhibition in that lithium can also inhibit inositol monophosphatase (IMP) and induce macroautophagy due to free inositol and myo-inositol-1,4,5-triphosphate (IP3) depletion [34]. This biological effect is therapeutically relevant in neurodegenerative disorders induced by intracellular self-aggregation and accumulation of peptides such as β-amyloid, α-synuclein, tau or mutant huntingtin, in neurofibrillary tangles, which are autophagy targets. Nevertheless, the neuroprotective properties of selective GSK3 inhibitors confirm that GSK3 plays a central role in neuronal survival and plasticity. Interestingly, other neuroprotective drugs such as valproic acid indirectly induce GSK3 phosphorylation [11]. GSK3 inhibitors belong to two categories: ATP competitive inhibitors localized in the ATP-binding pocket (CHIR-99021, AR-A014418, SB216763, SB415286) and non ATP competitive compounds (TDZD8, TWS119). Lithium chloride competes with magnesium and causes Ser9 and Ser21 autophopshorylation in GSK3β and GSK3α, respectively. Different GSK3


inhibitors including AR-A014418, SB216763, SB415286, do not interfere with GSK3α/β phosphorylation at Ser21 and Ser9. Thus, the effects of the pharmacological inhibition of GSK3 on downstream targets can be independent of phosphorylation induced by PI3K/AKT and other AGC kinases (Fig. 1). The question of why dysregulated, hyperactive GSK3 appears particularly damaging in neurons remains to be answered. The identification of oxidative stress as one of the most important event in the pathogenesis of classical models of neuronal injury by neurotoxic agents such as amyloid-β peptide and glutamate excitotoxicity and the use of pharmacological inhibitors of mitochondrial damage to maintain neuronal cell survival has provided important information about the requirements of this highly specialized cell type with unique biological functions. Several recent reports suggest that, among other factors, neuronal GSK3 activation appears particularly harmful due to the unique set of stress-related GSK substrates in this cell type. As an example, active GSK3 plays a pivotal role in Alzheimer’s disease by increasing tau phosphorylation and β-amyloid production [35]. Most of the GSK3 inhibitors, except for CHIR-99021, show low affinity and weak inhibition for other kinases [1]. Nevertheless, GSK3 inhibitors represent a promising category of therapeutic agents in neurodegenerative disorders and, consequently, in recent years the search for more specific GSK3 inhibitors to avoid side effects due to off-site targeting has been intensified [14]. 1.3. The opposing roles of GSK3 in tumors Over-activation of survival signaling pathways mediated by AKT and NFkB is a common feature of tumor cells. In recent years, great efforts have been made to advance the search for synthetic or naturally occurring antitumor drugs able to selectively inhibit protein kinases regulating cell survival pathways. The clinical utility of highly selective protein kinase inhibitors, however, has been questioned. To overcome the emergence of molecular resistance to specific inhibitors, it has been proposed the alternative strategy of pointing to multi-targeted protein kinase inhibitors or cocktails of selective antagonists [36]. Another possibility is the targeting of multifunctional kinases that act as master regulators of different and cross-talking signaling pathways. This appears to be the case for GSK3. In some tumor cell types active GSK3 mediates mitochondrial apoptosis, while in others it negatively regulates cell death induced by death receptors and


caspase 8, thus GSK3 has been shown to enhance cell death induced by TRAIL [4]. Thus, the effects of GSK3 inhibition on tumor cell survival depend on cell type-specific contexts. The increasing interest in histone deacetylase (HDAC) inhibitors as neuroprotective and cardioprotective drugs [11,37] and, at the same time, as promising anticancer agents underscores the evidence that GSK3 inhibition by lithium chloride, valproate and other HDAC inhibitors can lead to opposite outcomes in different cellular contexts. Evidence from in vitro and in vivo studies with specific GSK3 inhibitors like AR-A014418 confirm this observation: AR-A014418 has been indicated as a potential therapeutic tool in neurodegenerative disorders, however it potentiates the activity of anticancer drugs in glioblastoma and melanoma cells [38] and other tumor cell types [4]. Similarly, the regulation of NFkB by GSK3 exerts antiapoptotic effects in pancreatic cancer cells [39], while it induces cell death in astrocytes [40]. We have shown that GSK3 inhibition modulates the threshold of sensitivity to anticancer synthetic triterpenoids in prostate adenocarcinoma cell lines. We observed that GSK3 inhibition or genetic depletion changes the shape of cell death from apoptotic to necrotic by enhancing caspase 8 activity and mitochondrial and nuclear apoptosis mediated by downstream caspase 9 and caspase 3 [22]. The activation of the Wnt/βcatenin pathway can promote tumor progression. Active GSK3β phosphorylates and promotes β-catenin proteosomal degradation by the destruction complex formed by the adenomatous polyposis coli (APC) and Axin proteins. This process inhibits Wnt-stimulated TCF/LEF-dependent transcription and accumulation of an oncogenic, transcriptionally active form of β-catenin [41]. In those tumor cell types in which the role of active GSK3β in the Wnt/β-catenin pathway prevails over other functions, the inhibitory phosphorylation of GSK3β can thus be functional to tumor progression through the stabilization and activation of β-catenin. Recently, multiple proteins have been identified as potential candidates for GSK3-regulated phosphorylation and Wnt-regulated proteasomal degradation, thus broadening the range of cellular activities controlled by the Wnt/GSK3 destruction complex. These targets, besides β-catenin-mediated transcription, are involved in cellular functions including RNA processing, cytoskeletal dynamics, and cell metabolism [42]. Several recent excellent reviews have exhaustively treated the role of GSK3 and the Wnt-β-catenin pathway in the control of cell fate [43], therefore this topic will not be discussed further in this review. The paradoxical effects produced by the same compound in different pathological contexts occur despite some neurodegenerative disorders like


Alzheimer’s disease, diabetes and cancer show common pathophysiological traits including mitochondrial dysfunction, excessive generation of ROS, a precarious redox equilibrium due to altered expression of antioxidant enzymes and insulin resistance. The first consideration is that pathological neurons, cardiomyocytes and pancreatic β cells are not transformed cells; secondly, they are non-proliferating, post-mitotic cells. These observations could imply that untransformed and rapidly proliferating, transformed cells adopt different metabolic rearrangements to cope with similar pathological manifestations. Here we will illustrate some of the metabolic changes resulting from GSK3 modulation and will discuss several hypothesis on why the attempt to restore intracellular homeostasis can produce diverging effects in untransformed and in transformed cells. 2. Linking the metabolic effects of GSK3 with cytoprotective and anticancer properties It is now established that the metabolic control of intracellular homeostasis is intimately linked to cell signaling networks integrating information from environmental changes or intracellular pathological events such as peptide or protein aggregation and tangle formation in neurons, endoplasmic reticulum (ER) stress due to glucose deprivation, or oncogene activation in cells living in chronic stressful conditions. In particular AKT, mTOR, GSK3, FoxO and sirtuins (Sirt 1-3) belong to a signaling network acting in concert with energy sensors such as AMP-activated protein kinase (AMPK) and transcription factors including the hypoxia inducible transcription factor 1 (HIF-1) and NF-E2-related factor 2 (Nrf2) capable of producing metabolic and redox adjustments in response to the bioenergetic requirements of the cells. While short-term damage elicits adaptive responses usually functional to cell defense and repair, long-term activation of homeostatic molecules, for example HIF-1, AKT/GSK3 and Nrf2, can help cells to substantially reprogram metabolism in pathological directions (Fig. 1). 2.1 GSK3 supports the glycolytic phenotype The glycolytic phenotype of most tumor cells is supported by the overactivation of the PI3K/AKT/GSK3 pathway that can confer resistance to apoptosis. Other specific cell types such as immune cells with intense biosynthetic activity show an hypermetabolic state and preferentially use glucose as a primary energy source [44]. However, it is not clear whether the glycolytic switch may have a causative role or be a consequence of transformation. The observation that metabolic diseases characterized by


Figure 1. Redox imbalance is a common trait of different pathological conditions including cardiovascular diseases, neurodegeneration, metabolic disorders, diabetes, and cancer. One of the main line of antioxidant defense against oxidative stress involves the activation of cell signaling pathways regulating energy expenditure and adjustment of glucose metabolism. The enhanced transcriptional activity of Nrf2 following GSK3 inhibition provides a battery of antioxidant enzymes, while HIF induction of glycolytic enzymes can compensate the impairment of mitochondrial ATP production. This mechanism is effectively involved in neuroprotection. In cardiomyocytes, a central event apparently necessary to protection against ischemia/reperfusion injury is the limitation of mPTP, which can be obtained by GSK3 inhibition. In tumor cells under constitutive oxidative stress, the perturbation of a precarious redox equilibrium by ROS induction or GSH depletion can lead to cell death and represents a promising anticancer strategy. Different redox-active drugs have been shown to enhance GSK3 phosphorylation. Due to the up-regulation of GSH and antioxidant enzymes consequent to GSK3 inhibition, this event could be an attempt to counteract oxidative stress. altered glucose balance and defective mitochondrial energy production including insulin resistance and obesity are risk factors for some types of cancer suggest that impairment of glucose homeostasis is important in the early phases of cancer development [45]. On the other hand, the hypermetabolic state of frank tumor cells showing activated oncogenes like ras and myc indicate that glycolysis supported by AKT could be an adaptation to cope with the high bioenergetic requirements of tumor cells due to oncogene activation and increased ROS production.


Emerging evidence is elucidating the link between glucose metabolism and inhibition of tumor cell death. GSK3 can regulate glucose metabolism through multiple mechanisms, with varied effects in normal and tumor cells [14,46]. During growth factor deprivation, but in conditions of enhanced glucose metabolism, GSK3 inhibition by protein kinase C can stabilize the antiapoptotic Bcl-2 family protein Mcl-1, which is degraded by a GSK3-mediated mechanism [47]. Overexpression of HKII and enhanced binding to mitochondria promotes glucose consumption and exerts an antiapoptotic effect. It has been reported that active GSK3 can phosphorylate HK II, thus inhibiting its binding to mitochondria, and trigger cell death in tumor cells [48]. As a result of AKT activation by growth factors, the inhibitory phosphorylation of GSK3 promotes binding of HK II to mitochondrial outer membrane, an event that enhances glucose metabolism. [18,49]. This same adaptive mechanism involving HK II overexpression and binding to mitochondria, with the consequent enhancement of glucose metabolism, can be also due to chronic inhibition of GSK3 and plays an opposing, neuroprotective role in neuronal cells treated with the mitochondrial complex I inhibitor rotenone [23,31,48]. GSK has been shown to mediate HIF-1α stabilization upon ROS-induced RhoB activation in hypoxic conditions [50], or during early hypoxia in other experimental models [51], thus reinforcing glucose utilization through HIF-mediated transcription. Conversely, in prolonged hypoxia GSK3 dephosphorylation and activation contributes to HIF-1α degradation [19] and probably hypoxic cell death. 2.2 GSK3 in regulation of energy balance Cross-talk and reciprocal regulation between GSK3 and AMPK, a major molecular mediator of energy balance within the cell [52], plays a key role in cellular and systemic metabolic regulation. AMPK is a heterotrimeric complex that is composed of a catalytic α-subunit, a β-subunit, and a γ-subunit that binds AMP. Low energy levels in the cell due to oxygen or glucose deprivation or enhanced ATP consumption are sensed as a decreased ATP/AMP ratio. Allosteric AMPK activation by AMP binding is accompanied by phosphorylation of the active site by upstream kinases including the tumor suppressor LKB1 and calmodulin-dependent kinase kinase-β (CaMKKβ). AMPK then phosphorylates various substrates including metabolic enzymes, transcription factors and co-activators to inhibit energy consuming biosynthetic pathways, first of all protein synthesis promoted by mTOR activation, along with lipid and carbohydrate biosynthesis, while it activates catabolic reactions to preserve ATP. The


metabolic effects of AMPK inhibition by AICAR or metformin, include improved glucose uptake and glycolysis and lipid metabolism [53]. AMPK phosphorylates AcetylCoA carboxylase 1 and 2 (ACC1, ACC2), blocking fatty acid synthesis and improving fatty acid oxidation. It has been recently demonstrated that AMPK is inhibited upon binding of glycogen to its β-subunit [53]. In this way, through the modulation of glycogen binding, AMPK can also sense and control the availability energy stores. AMPK inhibitors such as metformin are in clinical use for the treatment of the metabolic syndrome, dyslipidemia and diabetes. Activation of AMPK triggers a phosphorylation cascade involving the tuberous sclerosis complex proteins (TSC1 or hamartin, TSC2 or tuberin), that inhibit the mTOR activator Rheb, and leading to mTOR inhibition. In conditions of critical energy levels due to low ATP, glucose or oxygen, AMPK and GSK3 cooperate to activate a pro-survival pathway by blocking mTOR activity. In different non-tumorigenic cell lines including HEK293, MEF, 293T, AMPK directly phosphorylates TSC2 on serine residues 29–32 and primes serine residues for subsequent phosphorylation by GSK3 [54]. AMPK has been shown to induce AKT-mediated Ser9 phosphorylation of GSK3 in HepG2 cells that is required for inhibition of cAMP-response element (CRE)-dependent genes, such as phosphoenolpyruvate carboxykinase C (PEPCK-C) [55]. Conversely, in both hyppocampal neurons and the neuroblastoma cell line SH-SY5Y both the AMPK inhibitors AICAR and phenformin induced AKT and GSK3 dephosphorylation [56]. AMPK activation not only limits ATP-consuming biosynthetic processes, but also inhibits cell growth and proliferation by inducing the accumulation of cell-cycle inhibitors such as p53, p27 and p21. Based on the evidence of the proapoptotic effects on energy demanding cancer cells, AMPK inhibitors have received attention as potential antineoplastic drugs [53]. In conclusion, GSK3 signaling essentially contributes to the network involving modulation of mTOR and AMPK activation, all integrating at the level of the TSC2 complex, in order to arrange growth control with energy availability within the cell (Fig. 2). Other major metabolic changes affecting energy balance in tumor cells involve the diversion of lipogenesis by fatty acid synthase (FAS) to enhance cell survival at the expenses of energy storage [57] and myc-stimulated high rate glutamine metabolism to support anabolic growth [58]. Several reports show that GSK3 plays a role in these key tumor-specific metabolic adaptations. Active GSK3 has been reported to down-regulate the transcriptional activity of adipocyte determination and differentiation-dependent factor 1 (ADD1)/SREBP1c that controls FAS expression [59] and to phosphorylate and promote degradation of a Drosophila homologue of c-myc upon priming by casein kinase I (CKI) [60].


Figure 2. The main intracellular stress-sensitive signaling pathways converge on GSK3 regulation. Dissection of the different branches of the network by the use of pharmacological inhibitors including the MEK inhibitors PD98059 and UO126, the PI3K inhibitors wortmannin and LY294002, the mTOR inhibitor rapamycin and the p90 RSK inhibitor BI-D1870 served to set the hierarchical order of the single molecular components of the network. Although targeted genetic depletion is a straight forward approach to assign a role to the different molecular components, the therapeutic perspectives to modulate the activity of the network foresee the use of protein kinase inhibitors. The reciprocal regulation and feed-back inhibition or activation in the network are not represented for simplification, but contribute to finely tune the duration and the effectiveness of the adaptive signals aimed at preserving cell survival as a first goal. If the intra- and extracellular conditions do not allow the completion of the survival programs, the attempts to balance adverse conditions result in a futile cycle and cells will eventually undergo cell death. 2.3. GSK3 in intracellular redox regulation One of the most remarkable properties of active GSK3 is to modulate the activity of the transcription factor Nrf2, a major regulator of redox homeostasis [61]. This property directly links the neuroprotective effects of GSK3 inhibition with the antioxidant defenses of the cells. A pool of nuclear Nrf2 is constitutively present in unstressed cells to maintain basal levels of transcription of antioxidant phase II genes regulated by the cis-acting element antioxidant response element (ARE). To maintain homeostasis, part of Nrf2 is negatively regulated by Keap1 which promotes its ubiquitylation and degradation in the cytoplasm. In conditions of redox imbalance due to enhanced production of hydrogen peroxide and ROS, or decreased antioxidant defenses including GSH, cytoplasmic Nrf2 is stabilized and,


following translocation, contributes to the nuclear active pool. ARE activation by a transcriptional complex formed by Nrf2 and small Maf proteins induces the expression of antioxidant proteins including heme oxygenase (HO-1), NADPH:quinone oxidoreductase (NQO1), thioredoxin (Trx) and enzymes involved in GSH metabolism such as glutathione S-transferase A1/A2 (GST A1/A2), glutathione peroxidase (GPx), γ-glutamyl cysteine ligase (γGCL). The induction of this last enzyme causes an adaptive elevation of intracellular GSH. Once normal levels of GSH are reached, a negative feed-back operated by GSH itself inhibits further GSH synthesis by γGCL. Active GSK3 has been reported to phosphorylate Nrf2 [62]. This post-translational modification excludes Nrf2 from the nucleus, thus limiting the expression of Nrf2-dependent genes. Thus, upon GSK3 inhibition, Nrf2 can activate the transcription of ARE-regulated phase II antioxidant enzymes. As a consequence, GSK3 inhibition can enhance GSH utilization in cells with overall antioxidant effects. Importantly, persistent oxidative stress has been shown to inhibit cytoprotective Nrf2 activation by GSK3 in neuronal cells [63], confirming the neuroprotective effects of GSK3 inhibition. Nrf2 plays a relevant role in tumorigenesis. Nrf2 promotes detoxification and cytoprotection against oxidative, electrophilic and xenobiotic stress and its activation has been implicated in prevention of tumor development and progression in numerous animal models. Nrf2 is apparently a major molecular target of cancer chemopreventive compounds, in particular of chemopreventive phytochemicals or their synthetic derivatives [64]. In established tumors, Nrf2 targeting by different anticancer agents induce caspase-dependent and independent cell death essentially by glutathione depletion and redox imbalance [65]. In some tumor cell types, Nrf2 activation, as judged by induction of HO-1, occurs in cells undergoing cell death [66]. Therefore, while the cytoprotective and antioxidant response in normal or preneoplastic cells in the early phases of transformation elicits an antitumor effect by preserving cell survival and tissue integrity, the same response may apparently contribute to cell death in established tumor cells. Several lines of evidence indicate that GSK3 itself is sensitive to redox regulation [62,63,67,68]. Of note, key molecular targets of survival pathways with proinflammatory and proangiogenic activity including the nuclear transcription factors NF-κB and HIF, AKT and mTOR are redox-sensitive molecules (Fig. 1). Different redox-active pharmacological agents and intracellular conditions that perturb intracellular redox homeostasis modulate GSK3 activity [67]. The modification of intracellular redox conditions toward a more oxidized state, as already discussed in the case of GSH metabolism regulated by Nrf2, can affect, in particular, GSK3 intracellular localization and activity.


The molecular events following GSK3 inhibition, including induction of phase II genes such as heme oxygenase, elevation of intracellular GSH and of enzymes related with GSH metabolism, indicate that GSK3 inhibition exerts a cytoprotective effect against oxidative stress (Fig.3). This cytoprotective response is apparently evoked by transient, acute stress in untransformed as well as in tumor cells. A different condition seems to be established in chronic degenerative disorders, where oxidative stress becomes a permanent condition that necessitates metabolic adaptations, and active GSK3 accumulates in neuronal cells with detrimental effects on cell survival. Time-dependent effects

Figure 3. The expression of antioxidant enzymes such as heme oxygenase-1 (HO-1) and a set of stress-inducible Nrf2-regulated genes stimulated by GSK3 inhibition requires the coordinated activation of ancillary pathways providing energy substrates and cofactors necessary to guarantee the proper functioning of the antioxidant machinery. The expression of antioxidant enzymes requires transient up-regulation of G6PD providing reducing equivalents in the form of NADPH, concomitantly with enhanced glucose uptake. The attempt to mount an antioxidant adaptive response against oxidative stress can be elicited in normal and stressed untransformed cells, as well as in cells under selective pressure imposed by tumorigenic conditions such as chronic inflammation, and in frank tumor cells. The final outcome changes depending on the intracellular bioenergetic state and microenvironmental conditions that must be permissive to effectively obtain cell repair and survival.


have been described for ERK1/2. Transient ERK activation can induce an adaptive elevation of intracellular GSH [69]. However, sustained ERK activation can both lead to oncogenic adaptations or to cell death [70]. Several lines of evidence suggest that the consequences of transient versus sustained GSK3 inhibition can have similar effects in some cellular contexts. 2.4. GSK3 as a molecular target of redox-active phytochemicals Emerging evidence indicates that natural or synthetic derivatives of phytochemicals interfere with key redox-regulated events in tumorigenesis and tumor-predisposing conditions such as metabolic disorders including hyperinsulinemia, obesity and diabetes, and in aging [71]. Remarkably, several redox-active dietary phytochemicals have shown neuroprotective [72] and cardioprotective effects [73] in preclinical studies. It is now established that metabolic disorders, besides defining risk factors for cardiovascular diseases and diabetes, also increase the risk for cancer. Starting from the natural compounds, novel synthetic derivatives are under development for the production of more effective drugs with increased therapeutic efficacy. This class of compounds are presently under intense pre-clinical and clinical scrutiny for potential use in diabetes, neurodegenerative and cardiovascular disorders and in cancer treatment [74]. Nrf2 is a major target of cancer chemopreventive and cytoprotective phytochemicals and their synthetic derivatives [64]. Few data are presently available on the involvement of GSK3-mediated Nrf2 activation in the mechanism of action of these compounds. However, owing to the fact that Nrf2 regulation by GSK3 could explain several of the cytoprotective effects of GSK3 inhibition in different pathophysiological contexts, here we will briefly mention recent reports on the involvement of Nrf2 in the activity of model phytochemicals which appear promising therapeutic agents: resveratrol, sulforaphane and curcumin. Resveratrol, a polyphenol found in red grapes, has shown a range of remarkable biological properties in preclinical models of aging that resemble the beneficial metabolic effects of dietary restriction [75]. Similarly to other phytochemicals with potential application in long-term cancer chemoprevention, resveratrol shows both cytoprotective and anticancer properties. GSK3 inhibitory phosphorylation induced by resveratrol has been shown to protect against mitochondrial damage induced by arachidonic acid [76], and to be implicated in cardioprotection [77]. Induction of Nrf2-regulated antioxidant genes depending on GSK3 inhibition induced by sulforaphane has been shown to protect hyppocampal neurons against lethal kainate-induced oxidative stress [78]. Curcumin and sulforaphane activate an antioxidant defensive response mediated by Nrf2-induced GPx expression [79].


3. Effects of GSK3 on stem cells An unexpected function of GSK3 recently discovered is to regulate embryonic stem cell (ESC) pluripotency and self-renewal. GSK3 inhibition has been shown to sustain the expression of three molecular targets involved in stem cell repopulation: the pluripotent state-specific transcription factors Oct-3/4, Rex-1 and Nanog [80]. GSK3 regulates different signaling pathways implicated in HSC function, including Wnt, Hedgehog and Notch pathways. Interestingly, GSK3 inhibitors were found to modulate genes regulated by Wnt, Hedgehog and Notch pathways selectively in hematopoietic stem cells (HSC) of the primitive hematopoietic compartment, while mature cells were unaffected [81]; most importantly, GSK inhibitors were toxic to MLL leukemia cells [82]. Again, GSK3 can act in opposite ways in cells of the same origin depending on transformation. Similarly to the role of GSK3 in different intracellular compartments and molecular complexes, the partition of GSK3 in subcellular pools appears a regulatory mechanism per se in stem cell renewal [83]. The PI3K/AKT pathway and c-myc sustain ESC self-renewal. GSK3 phopshorylates and induce degradation of c-myc, thus antagonizing ESC self-renewal. The mechanism operating in mouse ESC depends on GSK3 phosphorylation by AKT that excludes shuttling of GSK3 into the nucleus and causes phospho-GSK3 accumulation in the cytoplasm. In this way AKT neutralizes the antagonizing effect of GSK3 on ESC self-renewal. Summary The regulation of cell survival occurs through the activity of interrelated sensing systems that monitor, among others, nutrient and glucose availability, the level of tissue oxygenation and redox homeostasis. NF-κB, HIF, FOXO, Sirt1, PI3K/AKT/mTOR/ GSK3 and the MAP kinases ERK1/2, JNK and p38 belong to a signaling network deciphering the overall metabolic state of the cell. The cross-talk between the different branches of the network results in the activation of effector mechanisms that engage both pro-survival and pro-death pathways in adverse environmental conditions. The modulation of these pathways propagates information to effector systems involved in glucose metabolism, energy expenditure and redox homeostasis that allow cells to adapt to transient environmental changes, with an overall positive effect on cell survival. Under conditions of chronic stress, however, the persistent activation of these salvage pathways to which GSK3 belongs can contribute to the transcriptional reprogramming involved in tumor insurgence, stabilization and progression. Active GSK3 phosphorylates and promotes the degradation of substrates involved in molecular pathways


including Wnt/βcatenin and Hedgehog/MYCN in normal cells, thus blunting important oncogenic signaling. However, GSK3 overactivation has apparently detrimental effects on cell survival. Under physiological conditions, GSK3 is kept under negative control by cell survival signaling triggered by trophic factors and supported by adequate energy availability. In different pathological conditions, such as neurodegenerative and cardiovascular disorders, GSK3 shows excessive accumulation and over-activation, thus its inhibition exerts cytoprotective effects. In cancer, GSK3 probably plays a different role during the early phases of tumorigenesis and in established tumor cells. Several lines of evidence indicate that dysregulated GSK3 inhibition, depending on constitutive activation of AKT and other oncogenes, can support tumor-associated metabolic adaptations such as the glycolytic switch and can play an antiapoptotic role by preserving mitochondrial integrity in stressful conditions. On the other hand, Nrf2, which can be activated upon GSK3 inhibition, has been identified as a major target of several cancer chemopreventive drugs, providing a cellular defense against tumorigenic oxidative damage. The up-regulation of Nrf2-regulated antioxidant enzymes following GSK3 inhibition, while accounting for the cytoprotective effects against cell death in brain cells, can confer resistance to anticancer cytotoxic agents. Paradoxically, different drugs inducing GSK3 inhibition, including several phytochemicals or their synthetic derivatives, specific pharmacologic inhibitors and HDAC inhibitors, have shown promising anticancer effects in preclinical models. One hypothesis to explain the opposing roles of GSK3 inhibition is that different effects depend on the bioenergetic requirements of the cells. In cells that rely on oxidative phosphorylation for energy production and in cells that can tolerate the use of glycolysis as an alternate source of ATP, such as neuronal cells, GSK3 inhibition exerts cytoprotective effects. GSK3-mediated cytoprotection can thus be beneficial in cells that are not subjected to oncogenic stimulation, while can promote cell survival in tumorigenic cells. Once the metabolic reprogramming that leads to stabilization of tumor cells, first of all the glycolytic switch, is fully realized, GSK3 inhibition in sensitive cells might perturb a precarious equilibrium characterized by a high metabolic rate fueled by enhanced glucose metabolism and adaptive adjustments of redox balance to tolerate increased ROS production. An emerging therapeutic approach highlights the possibility to exploit specific metabolic adaptations such as enhanced glucose usage, elevated production of reactive oxygen species and antioxidant enzymes to selectively target tumor cells [84]. GSK3 pharmacologic targeting could offer the opportunity to investigate the application in oncology of drugs already in use for the treatment of neurological disorders, which have been proved to be safe upon long-term administration.


Acknowledgments This work was supported by grants from the Associazione Italiana per la Ricerca sul Cancro (AIRC), the Ministero della Salute, ISS Italy-USA Progetto Malattie Rare and the Compagnia di San Paolo. References 1. Bain, J., Plater, L., Elliott, M., Shpiro, N., Hastie, J., McLauchlan, H., et al. 2007,

Biochem J, 13, 13. 2. Matsuda, T., Zhai, P., Maejima, Y., Hong, C., Gao, S., Tian, B., et al. 2008, Proc

Natl Acad Sci U S A, 105, 20900. 3. Watcharasit, P., Bijur, G. N., Song, L., Zhu, J., Chen, X., and Jope, R. S. 2003, J

Biol Chem., 278, 48872. 4. Beurel, E., and Jope, R. S. 2006, Prog Neurobiol., 79, 173. 5. Soriano, F. X., Papadia, S., Hofmann, F., Hardingham, N. R., Bading, H., and

Hardingham, G. E. 2006, J Neurosci, 26, 4509. 6. Das, A., Xi, L., and Kukreja, R. C. 2008, J Biol Chem, 283, 29572. 7. Juhaszova, M., Zorov, D. B., Yaniv, Y., Nuss, H. B., Wang, S., and Sollott, S. J.

2009, Circ Res, 104, 1240. 8. Mussmann, R., Geese, M., Harder, F., Kegel, S., Andag, U., Lomow, A., et al.

2007, J Biol Chem., 282, 12030. 9. Cohen, P., and Goedert, M. 2004, Nat Rev Drug Discov, 3, 479. 10. Liang, M. H., and Chuang, D. M. 2007, J Biol Chem, 282, 3904. 11. Leng, Y., Liang, M. H., Ren, M., Marinova, Z., Leeds, P., and Chuang, D. M.

2008, J Neurosci, 28, 2576. 12. Juhaszova, M., Zorov, D. B., Kim, S. H., Pepe, S., Fu, Q., Fishbein, K. W., et al.

2004, J Clin Invest., 113, 1535. 13. Tanabe, K., Liu, Z., Patel, S., Doble, B. W., Li, L., Cras-Meneur, C., et al. 2008,

PLoS Biol, 6, e37. 14. Medina, M., and Castro, A. 2008, Curr Opin Drug Discov Devel, 11, 533. 15. Segditsas, S., and Tomlinson, I. 2006, Oncogene, 25, 7531, Pearson, H. B.,

Phesse, T. J., and Clarke, A. R. 2009, Cancer Res, 69, 94. 16. Zhu, L., and Pollard, J. W. 2007, Proc Natl Acad Sci U S A, 104, 15847. 17. DeBerardinis, R. J., Lum, J. J., Hatzivassiliou, G., and Thompson, C. B. 2008,

Cell Metab, 7, 11, Robey, R. B., and Hay, N. 2009, Semin Cancer Biol, 19, 25. 18. Chiara, F., Castellaro, D., Marin, O., Petronilli, V., Brusilow, W. S., Juhaszova,

M., et al. 2008, PLoS ONE., 3, e1852. 19. Flugel, D., Gorlach, A., Michiels, C., and Kietzmann, T. 2007, Mol Cell Biol,

27, 3253. 20. Sun, M., Song, L., Li, Y., Zhou, T., and Jope, R. S. 2008, Cell Death Differ,

15, 1887. 21. Matsui, Y., Watanabe, J., Ikegawa, M., Kamoto, T., Ogawa, O., and Nishiyama,

H. 2008, Oncogene, 7, 7.


22. Vene, R., Larghero, P., Arena, G., Sporn, M. B., Albini, A., and Tosetti, F. 2008, Cancer Res., 68, 6987.

23. Kotliarova, S., Pastorino, S., Kovell, L. C., Kotliarov, Y., Song, H., Zhang, W., et al. 2008, Cancer Res, 68, 6643.

24. Calabrese, E. J., Bachmann, K. A., Bailer, A. J., Bolger, P. M., Borak, J., Cai, L., et al. 2007, Toxicol Appl Pharmacol, 222, 122.

25. Juhaszova, M., Wang, S., Zorov, D. B., Nuss, H. B., Gleichmann, M., Mattson, M. P., et al. 2008, Ann N Y Acad Sci, 1123, 197.

26. Mattson, M. P., Gleichmann, M., and Cheng, A. 2008, Neuron, 60, 748. 27. Linseman, D. A., Butts, B. D., Precht, T. A., Phelps, R. A., Le, S. S., Laessig, T.

A., et al. 2004, J Neurosci, 24, 9993. 28. Mishra, R., Barthwal, M. K., Sondarva, G., Rana, B., Wong, L., Chatterjee, M., et

al. 2007, J Biol Chem, 282, 30393. 29. Chen, G., Bower, K. A., Ma, C., Fang, S., Thiele, C. J., and Luo, J. 2004, Faseb J,

18, 1162, MacAulay, K., Doble, B. W., Patel, S., Hansotia, T., Sinclair, E. M., Drucker, D. J., et al. 2007, Cell Metab, 6, 329.

30. Soucek, T., Cumming, R., Dargusch, R., Maher, P., and Schubert, D. 2003, Neuron, 39, 43.

31. Gimenez-Cassina, A., Lim, F., Cerrato, T., Palomo, G. M., and Diaz-Nido, J. 2009, J Biol Chem, 284, 3001.

32. Hetman, M., Cavanaugh, J. E., Kimelman, D., and Xia, Z. 2000, J Neurosci, 20, 2567.

33. Stambolic, V., Ruel, L., and Woodgett, J. R. 1996, Curr Biol, 6, 1664. 34. Sarkar, S., Floto, R. A., Berger, Z., Imarisio, S., Cordenier, A., Pasco, M., et al.

2005, J Cell Biol., 170, 1101. 35. Hooper, C., Killick, R., and Lovestone, S. 2008, J Neurochem, 104, 1433. 36. Daub, H., Specht, K., and Ullrich, A. 2004, Nat Rev Drug Discov, 3, 1001. 37. Trivedi, C. M., Luo, Y., Yin, Z., Zhang, M., Zhu, W., Wang, T., et al. 2007, Nat

Med, 13, 324. 38. Panka, D. J., Cho, D. C., Atkins, M. B., and Mier, J. W. 2008, J Biol Chem, 283,

726, Miyashita, K., Kawakami, K., Nakada, M., Mai, W., Shakoori, A., Fujisawa, H., et al. 2009, Clin Cancer Res, 15, 887.

39. Ougolkov, A. V., Fernandez-Zapico, M. E., Savoy, D. N., Urrutia, R. A., and Billadeau, D. D. 2005, Cancer Res., 65, 2076.

40. Sanchez, J. F., Sniderhan, L. F., Williamson, A. L., Fan, S., Chakraborty-Sett, S., and Maggirwar, S. B. 2003, Mol Cell Biol, 23, 4649.

41. Rasola, A., Fassetta, M., De Bacco, F., D'Alessandro, L., Gramaglia, D., Di Renzo, M. F., et al. 2007, Oncogene., 26, 1078.

42. Kim, N. G., Xu, C., and Gumbiner, B. M. 2009, Proc Natl Acad Sci U S A, 106, 5165.

43. Shitashige, M., Hirohashi, S., and Yamada, T. 2008, Cancer Sci, 99, 631, Angers, S., and Moon, R. T. 2009, Nat Rev Mol Cell Biol, 10, 468.

44. Woodland, R. T., Fox, C. J., Schmidt, M. R., Hammerman, P. S., Opferman, J. T., Korsmeyer, S. J., et al. 2008, Blood, 111, 750.

45. Giovannucci, E., and Michaud, D. 2007, Gastroenterology, 132, 2208.


46. Summers, S. A., Kao, A. W., Kohn, A. D., Backus, G. S., Roth, R. A., Pessin, J. E., et al. 1999, J Biol Chem., 274, 17934, Clodfelder-Miller, B., De Sarno, P., Zmijewska, A. A., Song, L., and Jope, R. S. 2005, J Biol Chem., 280, 39723.

47. Zhao, Y., Altman, B. J., Coloff, J. L., Herman, C. E., Jacobs, S. R., Wieman, H. L., et al. 2007, Mol Cell Biol, 27, 4328.

48. Pastorino, J. G., Hoek, J. B., and Shulga, N. 2005, Cancer Res, 65, 10545. 49. Rathmell, J. C., Fox, C. J., Plas, D. R., Hammerman, P. S., Cinalli, R. M., and

Thompson, C. B. 2003, Mol Cell Biol., 23, 7315. 50. Skuli, N., Monferran, S., Delmas, C., Lajoie-Mazenc, I., Favre, G., Toulas, C., et

al. 2006, Cancer Res, 66, 482. 51. Mottet, D., Dumont, V., Deccache, Y., Demazy, C., Ninane, N., Raes, M., et al.

2003, J Biol Chem, 278, 31277. 52. King, T. D., and Jope, R. S. 2005, Neuroreport., 16, 597. 53. McBride, A., Ghilagaber, S., Nikolaev, A., and Hardie, D. G. 2009, Cell Metab,

9, 23. 54. Inoki, K., Ouyang, H., Zhu, T., Lindvall, C., Wang, Y., Zhang, X., et al. 2006,

Cell., 126, 955. 55. Horike, N., Sakoda, H., Kushiyama, A., Ono, H., Fujishiro, M., Kamata, H., et al.

2008, J Biol Chem, 283, 33902. 56. King, T. D., Song, L., and Jope, R. S. 2006, Biochem Pharmacol., 71, 1637. 57. Menendez, J. A., and Lupu, R. 2007, Nat Rev Cancer., 7, 763. 58. Wise, D. R., DeBerardinis, R. J., Mancuso, A., Sayed, N., Zhang, X. Y., Pfeiffer,

H. K., et al. 2008, Proc Natl Acad Sci U S A, 105, 18782. 59. Kim, K. H., Song, M. J., Yoo, E. J., Choe, S. S., Park, S. D., and Kim, J. B. 2004,

J Biol Chem., 279, 51999. 60. Galletti, M., Riccardo, S., Parisi, F., Lora, C., Saqcena, M. K., Rivas, L., et al.

2009, Mol Cell Biol, 29, 3424. 61. Nguyen, T., Nioi, P., and Pickett, C. B. 2009, J Biol Chem, 284, 13291. 62. Salazar, M., Rojo, A. I., Velasco, D., de Sagarra, R. M., and Cuadrado, A. 2006, J

Biol Chem., 281, 14841. 63. Rojo, A. I., Sagarra, M. R., and Cuadrado, A. 2008, J Neurochem, 105, 192. 64. Sporn, M. B., and Liby, K. T. 2005, Nat Clin Pract Oncol., 2, 518. 65. Gopalakrishnan, A., and Tony Kong, A. N. 2008, Food Chem Toxicol, 46, 1257,

Hyer, M. L., Shi, R., Krajewska, M., Meyer, C., Lebedeva, I. V., Fisher, P. B., et al. 2008, Cancer Res., 68, 2927.

66. Konopleva, M., Tsao, T., Estrov, Z., Lee, R. M., Wang, R. Y., Jackson, C. E., et al. 2004, Cancer Res., 64, 7927.

67. Nemoto, S., Takeda, K., Yu, Z. X., Ferrans, V. J., and Finkel, T. 2000, Mol Cell Biol., 20, 7311.

68. Nair, V. D., and Olanow, C. W. 2008, J Biol Chem, 283, 15469. 69. Burdo, J., Schubert, D., and Maher, P. 2008, Brain Res, 1189, 12. 70. Sasagawa, S., Ozaki, Y., Fujita, K., and Kuroda, S. 2005, Nat Cell Biol, 7, 365. 71. Surh, Y. J. 2003, Nat Rev Cancer, 3, 768. 72. Dasgupta, B., and Milbrandt, J. 2007, Proc Natl Acad Sci U S A, 104, 7217. 73. Baur, J. A., and Sinclair, D. A. 2006, Nat Rev Drug Discov, 5, 493.


74. Tosetti, F., Noonan, D. M., and Albini, A. 2009, Int J Cancer, 75. Pearson, K. J., Baur, J. A., Lewis, K. N., Peshkin, L., Price, N. L., Labinskyy, N.,

et al. 2008, Cell Metab, 8, 157. 76. Shin, S. M., Cho, I. J., and Kim, S. G. 2009, Mol Pharmacol, 77. Xi, J., Wang, H., Mueller, R. A., Norfleet, E. A., and Xu, Z. 2009, Eur J

Pharmacol, 604, 111. 78. Rojo, A. I., Rada, P., Egea, J., Rosa, A. O., Lopez, M. G., and Cuadrado, A.

2008, Mol Cell Neurosci, 39, 125. 79. Banning, A., Deubel, S., Kluth, D., Zhou, Z., and Brigelius-Flohe, R. 2005, Mol

Cell Biol, 25, 4914. 80. Sato, N., Meijer, L., Skaltsounis, L., Greengard, P., and Brivanlou, A. H. 2004,

Nat Med, 10, 55. 81. Trowbridge, J. J., Xenocostas, A., Moon, R. T., and Bhatia, M. 2006, Nat Med,

12, 89. 82. Wang, Z., Smith, K. S., Murphy, M., Piloto, O., Somervaille, T. C., and Cleary, M. L. 2008, Nature, 455, 1205. 83. Bechard, M., and Dalton, S. 2009, Mol Cell Biol, 29, 2092. 84. Pelicano, H., Martin, D. S., Xu, R. H., and Huang, P. 2006, Oncogene, 25, 4633.



6. Polo-like kinase 3 in normal and tumor biology

Dazhong Xu and Wei Dai

Department of Environmental Medicine, New York University Langone Medical Center Tuxedo, NY 10987, USA

Abstract. Plk3 is a member of the Polo-like kinase family. It has multiple biological functions including the regulation of cell cycle progression, cell cycle checkpoint control, cellular response to hypoxia, and apoptosis. Deregulated Plk3 is also strongly implicated in tumorigenesis. Aberrant expression of Plk3 is detected in tumors of various origins. Plk3 knockout mice develop tumors in multiple organs at an advanced age. In the past, much of the attention regarding the role of Plks in tumor biology has been primarily directed toward Plk1. The importance of Plk3 in tumor biology and its value as a target for therapeutic intervention has not yet been fully appreciated. Recent progress on Plk3 research has revealed new roles that may lead to better understanding of this elusive protein kinase in regulating stress responses and suppressing tumorigenesis.

Correspondence/Reprint request: Dr. Wei Dai, Department of Environmental Medicine & Pharmacology New York University Langone Medical Center, 57 Old Forge Road, Tuxedo, NY 10987, USA E-mail: [email protected] or [email protected]

Dazhong Xu & Wei Dai 118

Introduction Serine/threonine (Ser/Thr) protein kinases play a central role in tumor biology. Deregulated kinase activities and/or their expression frequently lead to perturbation in cell growth control and differentiation, resulting in tumor formation. A prominent example is B-Raf kinase, whose oncogenic mutations contribute to the development of numerous human malignancies [1]. As deregulated kinase activities are often the main reason for their oncogenicity, small chemical compounds have been developed as potential anticancer drugs that target the kinase activities of various Ser/Thr protein kinases [2]. On the other hand, regulatory domains of protein kinases have also been explored as anti-proliferative or anticancer drugs [2]. Given the sheer number of genes encoding unique Ser/Thr protein kinases in the mammalian genome and the necessity of developing target-specific chemical compounds for cancer therapies, it is imperative for us to understand the biochemical and biological properties of various protein kinases in growth control and their deregulation in tumorigenesis. Polo-like kinases (Plks) are a family of Ser/Thr protein kinases that, among other functions, regulate cell cycle progression, apoptosis, and stress responses [3-6]. These kinases, when deregulated, have been strongly implicated in tumor initiation and development [7, 8]. Therefore there is considerable interest in exploring these kinases as targets for cancer drug development [9] although the major efforts have been devoted primary to Plk1 [10, 11]. Compared with that of Plk1, the biological role of Plk3 is much less clear. In this chapter, we attempt to summarize the work on characterization of Plk3 and discuss its significance in suppressing genomic instability and tumor angiogenesis. Polo-like kinases Polo-like kinases were named after Polo, a fruit fly (Drosophila melanogaster) Ser/Thr protein kinase that is involved in mitosis and meiosis [12]. Polo was originally discovered from a polo mutation that is associated with mitotic and meiosis defects in fruit fly [13]. It was subsequently found to be a highly conserved kinase [14]. Polo thus serves as the founding member of a family of evolutionarily conserved Ser/Thr protein kinases. Polo is an essential component of the cell cycle machinery and its kinase activity and expression level are tightly regulated during the cell cycle [3, 15]. Mutations of the polo gene lead to abnormal mitosis and meiosis [3]. The counterparts of Polo kinase in the budding yeast and the fission yeast are CDC5 and Plo1, respectively [16]. As with the fruit fly, these proteins have a pivotal role in the mitotic stage of the cell cycle [16]. In mammals, the

Polo-like kinase 3 in normal and tumor biology 119

Plk kinase family consists of four members including Plk1, Plk2, Plk3, and Plk4. These proteins all share significant sequence homology to Polo [17]. Structurally, all Plks are comprised of a highly conserved kinase domain at the amino–terminus and a Polo box domain (PBD) at the carboxyl-terminus [5, 17] (Fig. 1). PBD is critical for the kinase subcellualr localization as well as the substrate recognition by Plks [5]. The biological functions of mammalian Plks, however, are more diverse than that of their counterparts in lower eukaryotes. Plk1 is the most studied and its functions are more in line with those of Polo in the regulation of mitosis [6]. Plk1 expression peaks at G2/M phases, consistent with its role in mitosis [18]. The functions of Plk2, Plk3, and Plk4, however, are much less understood. Available evidence indicates that these proteins may have more diverse functions than that of Plk1 [6]. This notion is reinforced by findings revealing that these proteins have expression patterns quite different from those of Plk1 during the cell cycle [18]. In addition to the potential role in mitosis, Plk2, Plk3, and Plk4 appear to have functions in regulating the G1/S transition, S phase progression [19, 20], centrosome dynamics, stress responses [20-24], and/or synaptic activities [25]. Plks, or their deregulated activities, have also been strongly implicated in tumor development, especially Plk1, Plk3 and Plk4 [7, 8, 26, 27].

Figure 1. Basic Structures of Plk Family Members. Polo-like kinase 3 (Plk3) Mouse Plk3 was initially identified as a fibroblast growth factor-inducible kinase (Fnk) from NIH 3T3 cells in a differential display analysis [28]. The human counterpart was independently isolated and named Prk as a proliferation related kinase [29]. The amino acid sequence of human Plk3


shares 36% and 33% overall identity to Plk1 and Polo, respectively [29]. Plk3 has the structure of a typical Plk family member, with an N-terminal kinase domain and a C-terminal PBD with two Polo boxes [17] (Figure 1). Plk3 is considered an immediate early response gene, whose expression does not depend on de novo protein synthesis [18, 28, 29]. Its mRNA expression level reaches the maximal level one hour after stimulation with fibroblast growth factor or serum addition to the serum-starved NIH 3T3 cells and returns to the basal level 8 hours post-stimulation [18, 28, 30]. Plk3 protein level is fairly constant in mitogen-stimulated cells throughout the cell cycle [18, 21, 30]. A recent report, however, showing that Plk3 protein level is also cell cycle-regulated with a pattern similar to that of its mRNA, peaking at the G1 phase [19]. The kinase activity of Plk3 also oscillates during the cell cycle [31]. Immunocomplex kinase assays show that whereas the Plk3 kinase activity is low in G1 and G1/S phases, it dramatically increases during late S/G2 and remains at a moderate level at M phase [31]. Plk3 activity is also subjected to regulation by a variety of stress conditions [21, 23, 24, 32, 33]. Plk3 and tumorigenesis Tumorigenesis is a cellular transformation process that causes uncontrolled cell division and proliferation, resulting in a disease phenotype in multi-cellular organisms. Perturbations of many biological processes that control cell proliferation, apoptosis, stress responses, and angiogenesis have the potential to trigger malignant transformation [34]. The association of Plk3 with tumor development was first revealed when the human Plk3 (Prk) was examined in human primary tumor samples [29]. In contrast to Plk1, whose activity and expression are generally positively correlated with tumor formation, Plk3 activity or expression is often negatively correlated with tumor development. Thus, Plk3 is commonly regarded as a tumor suppressor [7, 8]. This notion has been confirmed by a mouse genetic study [22]. Compared with wild-type littermates, Plk3 null mice are prone to the development of malignancies in several organs [22]. Through analysis of more than a dozen of lung cancer patient samples, Plk3 mRNA was found to be significantly downregulated in a majority of lung carcinoma samples [29]. This abnormal mRNA expression appears to be a result of reduced PLK3 mRNA transcription [29], suggesting that reduced Plk3 expression may be associated with tumor development. This notion is supported by the observation that although three polymorphisms were identified from 40 lung tumor cell lines, no missense or nonsense mutations were found in Plk3 [35]. Thus, mutations of Plk3 coding sequences appear to be rare in lung cancer.


Reduced expression of PLK3 also occurs in human head and neck cancer [36]. In a survey of primary head and neck squamous-cell carcinomas (HNSCC) from 35 patients, 26 samples have reduced PLK3 mRNA levels compared with corresponding normal tissues. Two samples have an undetectable level of PLK3 expression [36]. This connection of PLK3 expression with head and neck cancer is reinforced by the finding that the PLK3 gene is located at chromosome 8p21 where high frequency of loss of heterozygosity happens in many human cancers [36-38]. Plk3 is also implicated in the development of colon cancer [39]. Expression of PLK3 mRNA is significantly lower in rat colon tumors that have been induced by azoxymethane than it is in the corresponding normal tissues [39]. Moreover, PLK3 mRNA levels in rat colon tumors appear to be influenced by diet since tumors isolated from rats fed high fat diet supplemented with fish oil exhibit less reduction in PLK3 expression than those isolated from rats fed the same diet supplemented with corn oil [39]. A positive correlation between Plk3 expression/activity and cancer formation has also been reported. An immunohistochemical study shows that Plk3 is expressed at a moderately high level in cystadenomas [40]. More than 5% of ovarian carcinomas are Plk3-positive whereas Plk3 expression was low in normal ovarian surface epithelium and borderline tumors [40]. In addition, LFM-A13, an inhibitor of Plks with a significant inhibitory activity toward Plk3, is able to delay progression of MMTV/neu transgenic mouse model of HER2 positive breast cancer [41]. Plk3 and cell cycle progression Plk3 in regulating G2/M phases Plk3 is involved in regulating multiple phases of the cell cycle. In most cells, Plk3 is predominantly located around the nuclear membrane in interphase of the cell cycle [42]. It concentrates at the centrosomes in a microtubule-dependent manner [42]. During mitosis, Plk3 is closely associated with spindle poles and mitotic spindles [42]. A significant amount of Plk3 can also be found in midbody in telophase [42]. Plk3 overexpression induces rapid cell cycle arrest at the M phase followed by apoptosis, likely as a result deregulation of microtubule dynamics and centrosomal function [42]. The subcellular localization of Plk3 to the centrosomes, spindle poles and the midbody is driven by the PBD domain of the kinase [43]. Both Polo boxes of Plk3 are required for its localization [43]. Interestingly, overexpression of the PBD domain alone can cause significant cell cycle arrest and abnormal cytokinesis [43]. This is consistent with another report that overexpression of


Plk3 inhibits cell proliferation and colony formation as a result of incomplete cytokinesis [44]. The functional conservation of Plk3 is clearly illustrated by two lines of evidence. Human PLK3 transcripts greatly enhance progesterone-induced meiotic maturation of Xenopus oocytes, whereas the corresponding antisense PLK3 transcripts inhibit the maturation process [31]. In S. cerevisiae, ectopic expression of human Plk3 was able to rescue the temperature-sensitive phenotype of a CDC5 (the yeast Plk homolog) deficient strain of the budding yeast [31]. These studies thus strongly suggest that Plk3 may also regulate the onset and/or progression of mitosis and meiosis. Plk3 phosphorylates and regulates the subcellular localization of Cdc25C phosphatase [45, 46], a key regulator of the G2/M transition by dephosphorylating inhibitory phosphorylation sites of CDK1 thereby activating the master regulator of mitosis [47]. Plk3 directly interacts with Cdc25C and phosphorylates it in vitro at Ser216 [45], a residue critical for binding to scaffold protein 14-3-3 when phosphorylated [48, 49]. Binding of Cdc25C to 14-3-3 sequestrates Cdc25C to the cytosol, thereby effectively inhibiting its function in the nuclei [50]. Plk3 is also found to phosphorylate Ser191 and Ser198 of the Cdc25C [46]. Similar to Ser216, phosphorylation of these two residues also promotes accumulation of Cdc25C in the nuclei [46]. Thus, serine to alanine mutation of Ser191 prevents the nuclear accumulation [46]. Consistent with this observation, ectopically expressed Plk3 promotes the accumulation of Cdc25C protein in the nuclei whereas Plk3 knockdown with RNA interference inhibits its nuclear accumulation [46]. Together, these findings indicate that Plk3 is an important regulator of the cell cycle at the G2/M transition. Plk3 in regulating G1/S phases Plk3 also mediates the G1/S transition and is required for the S phase entry [19, 51]. Plk3 protein levels are regulated during the cell cycle that peak at the G1 phase [19, 51], consistent with the expression pattern of Plk3 mRNA [18, 28, 30]. Knockdown of Plk3 by transfecting siRNA prevents cells from entering the S phase of the cell cycle [19]. Mechanistically, Plk3 appears to be required for cyclin E expression since Plk3 depletion significantly reduces cyclin E levels [19]. Regulation of cyclin E by Plk3 appears to be mediated through a posttranscriptional mechanism [19]. It has been proposed that Plk3 may regulate cyclin E levels by modulating the activity of phosphatase Cdc25A [19], which is a potential target of Plk3 [52]. Cdc25A is known to regulate the activity of the cyclin E/Cdk2 complex [53, 54].


Plk3 also phosphorylates proteins that are important for regulating DNA replication [20, 55]. Plk3 but not Plk1 can phosphorylate topoisomerase IIα at Thr1342 [55]. Topoisomerase IIα has an essential role in DNA replication/transcription, cell cycle regulation, cell proliferation, and cancer development [56]. Although the functional significance of Thr1342 phosphorylation remains poorly understood, this finding may reveal a potential new mechanism by which cell proliferation and tumor formation is differentially regulated by Plk3 and Plk1 [55]. Moreover, Plk3 interacts with and phosphorylates p125, the major subunit of DNA polymerase δ (Pol δ), and the phosphorylation occurs on Ser60 [20]. The significance of the physical interaction and phosphorylation of Pol δ by Plk3 remains to be elucidated. It is speculated that the phosphorylation may regulate the subcellular localization of Pol δ because the region of p125 contains a nuclear localization signal. Plk3 in Golgi fragmentation Fragmentation of the Golgi apparatus in mammalian cells is an essential step in mitosis entry. It is a process in which Golgi stacks are reversibly disassembled into small vesicles and tubules during mitosis to ensure correct partitioning of the Golgi content between daughter cells [57]. Several lines of evidence show that Plk3 is involved in the Golgi fragmentation process [58-60]. Plk3 is found to localize at the Golgi apparatus in interphase of the cell cycle; it disintegrates and redistributes in a manner similar to that of Golgi stacks [59]. Nocodazole, a drug that disrupts microtubules and induces mitotic arrest and Golgi fragmentation, activates Plk3 [59, 60]. Plk3 interacts with Golgi-specific protein giantin; a Golgi-specific poison brefeldin A disperses Plk3 signals in a manner similar to that of the Golgi apparatus [59]. Ectopic expression of wild-type but not the kinase-defective mutant of Plk3 leads to marked Golgi breakdown [59, 60]. Subsequent studies show that MEK1, a MAP kinase kinase known to mediate Golgi fragmentation, interacts and activates Plk3 [60]. Plk3 also colocalizes with phosphorylated MEK1 at the spindle poles [60]. Activation of Plk3 by nocodazole and Golgi fragmentation induced by Plk3 can be blocked by specific MEK1 inhibitors [60]. These results indicate that Plk3 is an essential regulator of Golgi fragmentation during mitosis and that MEK1 is an upstream activator of Plk3 in mediating this process. A recent study identifies vaccinia-related kinase1 (VRK1) to be the direct downstream target of Plk3 in regulating Golgi fragmentation [58]. VRK1 interacts and colocalizes with the Golgi complex and Plk3 throughout mitosis [58]. Plk3 phosphorylates VRK1 at Ser342 of the C-terminal region and this phosphorylation is necessary for VRK1 to mediate Golgi fragmentation.


Thus, a serine-to-alanine point mutation abolishes the capability of VRK1 to promote Golgi fragmentation induced by Plk3 or MEK1 [58]. Furthermore, both VRK1 kinase activity and phosphorylation by Plk3 at Ser342 are necessary for its function in Golgi fragmentation. Moreover, knockdown of VRK1 by siRNA or overexpression of kinase-dead VRK1 also inhibits Plk3 or MEK1-induced Golgi fragmentation [58]. Combined, available evidence has revealed the MEK1→Plk3→VRK1 signaling axis in the regulation of Golgi fragmentation in mitosis. Plk3 in regulating apoptosis Apoptosis or programmed cell death is a controlled suicidal process that is vitally important in maintaining homeostasis of tissues in multicellular organisms [61]. Apoptosis serves as a tumor suppressing mechanism by which old, abnormal, and/or excessive cells in tissues are removed [34, 62, 63]. Dysregulated apoptosis often contributes to tumor development [34, 63]. The role of Plk3 in apoptosis was initially revealed in the study of ectopically expressed Plk3 [32, 44]. Ectopic expression of Plk3 causes cell cycle arrest followed by chromatin condensation and apoptosis of cultured cells [32, 42, 44]. Plk3 appears to mediate apoptosis through a kinase-dependent and a kinase-independent mechanism [42, 64]. It is also shown to induce apoptosis in a biphasic manner: a rapid apoptosis that requires Plk3 kinase activity and p53, and a delayed onset apoptosis that is independent of Plk3 kinase activity and p53 [64]. Consistently, kinase defective Plk3 is less capable of inducing apoptosis compared to wild type Plk3 [42, 64]. Plk3 promotes the kinase-independent apoptotic function mainly through its C-terminal Polo box domains because ectopic expression of the Polo box domains alone is sufficient to induce apoptosis [32, 43, 44]. However, another study has identified the first 26 amino acids at the N-terminus of Plk3 to be essential for induction of apoptosis when it is overexpressed [64]. DNA damage- or superoxide-induced apoptosis requires the kinase activity of Plk3 [32, 64]. Plk3 phosphorylates tumor suppressor protein p53 at Ser20 in response to these stimuli and induces apoptosis through the p53-mediated proapoptotic pathway [32, 64]. Moreover, Plk3 can induce apoptosis through disrupting microtubule integrity [42]. Plk3 has been identified as a key component in NF-κB-dependent apoptosis [64]. A κB enhancer element is identified at the promoter region of PLK3 gene [64]. This element is responsible for NF-κB mediated Plk3 expression in response to a panel of known NF-κB activators, including chemotherapeutic drug Doxycycline [64]. Plk3 appears to play an essential role in Doxycycline-induced apoptosis, since knockdown of Plk3 expression


with siRNA dramatically inhibits Doxycycline-induced apoptosis [64]. Increase of Plk3 levels alone does not seem sufficient for apoptosis because TNF-α strongly activates Plk3 expression but fails to induce apoptosis [64]. Thus, Plk3 is necessary but not sufficient for NF-κB-mediated apoptosis. Plk3 in stress responses Plk3 in DNA damage response Genotoxic stresses such as ultraviolet light and ionizing radiation generate reactive oxygen species (ROS). ROS induces DNA damages and mutations that lead to genomic instability, thereby promoting tumor formation and progression [65, 66]. ROS are known to regulate DNA damage responses through the activation of p53, which leads to cell cycle arrest, apoptosis or replicative senescence [67]. ATM (mutated in ataxia telangiectasia) functions as an upstream regulator of p53 that is activated by oxidative stresses [68]. ATM regulates phosphorylation of p53 directly or indirectly and is thought to be a sensor for oxidative stresses [68]. Plk3 is activated by and mediates cellular responses to DNA damage that produces ROS [21, 23, 32, 69, 70]. Genotoxic agents such as hydrogen peroxide (H2O2), ionizing radiation, methylmethane sulfonate, ultraviolet light, and adriamycin strongly activate Plk3 [21, 32, 69, 70]. This is different from Plk1 whose activity is inhibited by these agents [32]. Plk3 mRNA is also inducible by ionizing radiation [33]. Plk3 activation induced by genotoxic agents is ATM-dependent [21, 32, 69], Thus, H2O2 and ionizing radiation mimetics are unable to activate Plk3 in ATM-deficient cells [21, 69]. Furthermore, inhibition of ATM using caffeine blocks adriamycin- or ionizing radiation-induced Plk3 activation [21, 32]. Plk3 is phosphorylated when treated with ionizing radiation and this phosphorylation is also ATM-dependent [21]. Further experimentation, however, fails to detect direct phosphorylation of Plk3 by ATM, indicating that ATM may promote Plk3 phosphorylation and activation through another kinase(s) [21]. One of the candidate kinases that phosphorylates Plk3 appears to be Chk 2, which is known be activated by ATM in response to genotoxic stresses [21]. The downstream effecter of Plk3 in mediating DNA damage responses is p53 [21, 32, 69]. Plk3 is able to phosphorylate p53 in response to DNA damage [21, 32, 69]. Activation of Plk3 by H2O2 is accompanied by phosphorylation of p53 at Ser9, Ser15, and Ser20 residues [69]. Phosphorylation of p53 at these residues in response to H2O2 correlates with its transcriptional activity since the level of p21, whose expression is positively regulated by p53, is also elevated [69]. Further analysis reveals that


phosphorylation of Ser20, but not other residues, after H2O2 treatment is dependent on Plk3 [69]. Furthermore, H2O2 fails to induce Ser20 phosphorylation of p53 in a Plk3-deficient cell line; the kinase-defective Plk3 suppresses H2O2 –induced Ser20 phosphorylation [69]. Phosphorylation of Ser20 of p53 is confirmed in an in vitro assay using recombinant Plk3 and p53 proteins [32]. Ser20 of p53 is also a target of protein kinases Chk1 and Chk2 that are downstream components of ATM and are essential for DNA damage checkpoint control [71, 72]. Plk3 is thus proposed to work in parallel with Chk1 and Chk2 to reinforce DNA damage checkpoint responses or to preferentially mediate certain types of DNA damage insults [32]. Plk3 is also found to directly regulate Chk2 and vise versa [21, 70]. Chk2 interacts directly with and activates Plk3 both in vitro and in vivo; this interaction is enhanced after DNA damage [70]. Intriguingly, Plk3 also phosphorylates Chk2 and contributes to its full activation [21]. Collectively, these results underscore the dynamic role of Plk3 in the DNA damage checkpoint control. Plk3 in hypoxic stress responses Tumor progression and metastasis require blood supply to provide oxygen and nutrients to the newly formed tumor masses. The formation of new blood vesicles or angiogenesis is therefore essential for tumor growth [73, 74]. Cellular responses to hypoxia play an important role in tumor growth since hypoxia triggers tumor angiogenesis [75]. Hypoxia-inducible factors (HIFs) are the key sensors and regulators of hypoxic responses, which, upon induction by hypoxia, promote expression of multiple genes essential for mediating angiogenesis. Among the HIF-activated gene products are: fibroblast growth factor (FGF), vascular endothelial growth factor (VEGF), and glucose transporters [75]. Plk3 can be activated by hypoxia [24]. Hypoxia-activated Plk3 increases the activity of AP-1 transcription factors and promotes apoptosis [24]. A recent study has demosntrated that Plk3 is involved in controlling HIF-1α levels in response to hypoxic conditions [22]. Specifically, compared with wild-type mouse embryonic fibroblasts (MEFs), PLK3 null MEFs exhibit a hyper-sensitive response to the treatment with hypoxia or with nickel ions and contain an elevated level of HIF-1α but not of HIF-1β [22]. Consistently, expression of VEGF-A is also higher in PLK3-/- MEFs than that in wild type MEFs [22]. More importantly, PLK3 null mice display an increased tumor incidence; tumors developed in these mice display an increased vasculature [22]. Transfection analyses also confirm the role of Plk3 in regulating the stability of HIF-1α [22]. Ectopically expressed Plk3 is able to suppress the expression and nuclear accumulation of HIF-1α in HeLa cells [22]. The


inhibition of HIF-1α nuclear translocation appears to be dependent on its kinase activity since overexpression of Plk3 kinase domain alone is sufficient to suppress HIF-1α accumulation in the nucleus under hypoxic conditions [22]. The exact mechanism by which Plk3 regulates HIF-1α levels and its activities remains to be elucidated. HIF-1α is a very labile protein that is rapidly degraded under normoxia via the proteosome [76]. After hydroxylation, which is an O2-sensitive process, hydroxylated HIF-1α is recognized by von Hippel-Lindau factor (VHL, a ubiquitin E3 ligase) for polyubiquitination and subsequent proteasomal degradation [77]. HIF-1α degradation can also be positively controlled by direct phosphorylation. GSK3β phosphorylates HIF-1α in the oxygen-dependent degradation domain, resulting in its destabilization via a VHL-independent but proreasome-dependent process [78]. Given that HIF-1α is super-induced by hypoxia or hypoxic mimetics in PLK3 null MEFs [22], it is tempting to speculate that Plk3 may regulate HIF-1α protein stability via a phosphorylation-dependent process as well.

Figure 2. Plk3 Regulatory Network. Conclusion remarks The relationship between Plk3 and normal or abnormal cell proliferation has been established since its initial identification. To date, interest in the role of Plk3 in tumor biology has lagged behind that of Plk1, a structurally close, and perhaps functionally-related component. Earlier work has focused primarily on the role of Plk3 in cell cycle regulation. Recent studies suggest


that Plk3 may have broader functions. The potential role of Plk3 in mediating hypoxic responses is particularly interesting since it can lead to the discovery of a new player important to the regulation of tumor angiogenesis. Despite extensive research efforts in the past decade, Plk3 remains largely an elusive protein kinase in terms of its physiological substrates and biological functions. Much needs to be done in order to obtain a clear picture on Plk3’s roles in normal biology, as well as to explore it as a potential anti-tumor target. Summarized in Figure 2 is the current knowledge about molecular interactions between Plk3 and cellular gene products in the regulation of normal and abnormal biological processes. Acknowledgements We thank Jeffery Winkles for his assistance in the study of PLK3 knockout mice. We are grateful for co-workers in the laboratory for various technical support and intellectual input. We also thank Mrs. Nedda Tichi for administrative support. This work was supported in part by grants from the National Institutes of Health to WD (CA074229, CA090658). References 1. Wan, P. T., Garnett, M. J., Roe, S. M., Lee, S., Niculescu-Duvaz, D., Good, V.

M., Jones, C. M., Marshall, C. J., Springer, C. J., Barford, D., and Marais, R. (2004). Mechanism of activation of the RAF-ERK signaling pathway by oncogenic mutations of B-RAF. Cell 116, 855-67.

2. Noble, M. E., Endicott, J. A., and Johnson, L. N. (2004). Protein kinase inhibitors: insights into drug design from structure. Science 303, 1800-5.

3. Glover, D. M., Hagan, I. M., and Tavares, A. A. (1998). Polo-like kinases: a team that plays throughout mitosis. Genes Dev 12, 3777-87.

4. Nigg, E. A. (1998). Polo-like kinases: positive regulators of cell division from start to finish. Curr Opin Cell Biol 10, 776-83.

5. Lowery, D. M., Lim, D., and Yaffe, M. B. (2005). Structure and function of Polo-like kinases. Oncogene 24, 248-59.

6. Archambault, V., and Glover, D. M. (2009). Polo-like kinases: conservation and divergence in their functions and regulation. Nat Rev Mol Cell Biol 10, 265-75.

7. Eckerdt, F., Yuan, J., and Strebhardt, K. (2005). Polo-like kinases and oncogenesis. Oncogene 24, 267-76.

8. Takai, N., Hamanaka, R., Yoshimatsu, J., and Miyakawa, I. (2005). Polo-like kinases (Plks) and cancer. Oncogene 24, 287-91.

9. Dai, W., Yang, Y., and Jiang, N. (2008). Plks as Novel Targets for Cancer Drug Design. In "Cancer Drug Discovery and Development" (W. Dai, Ed.), Humana Press, Totowa, NJ.


10. Spankuch, B., Kurunci-Csacsko, E., Kaufmann, M., and Strebhardt, K. (2007). Rational combinations of siRNAs targeting Plk1 with breast cancer drugs. Oncogene 26, 5793-807.

11. Gumireddy, K., Reddy, M. V., Cosenza, S. C., Boominathan, R., Baker, S. J., Papathi, N., Jiang, J., Holland, J., and Reddy, E. P. (2005). ON01910, a non-ATP-competitive small molecule inhibitor of Plk1, is a potent anticancer agent. Cancer Cell 7, 275-86.

12. Fenton, B., and Glover, D. M. (1993). A conserved mitotic kinase active at late anaphase-telophase in syncytial Drosophila embryos. Nature 363, 637-40.

13. Sunkel, C. E., and Glover, D. M. (1988). polo, a mitotic mutant of Drosophila displaying abnormal spindle poles. J Cell Sci 89 ( Pt 1), 25-38.

14. Llamazares, S., Moreira, A., Tavares, A., Girdham, C., Spruce, B. A., Gonzalez, C., Karess, R. E., Glover, D. M., and Sunkel, C. E. (1991). polo encodes a protein kinase homolog required for mitosis in Drosophila. Genes Dev 5, 2153-65.

15. Glover, D. M., Ohkura, H., and Tavares, A. (1996). Polo kinase: the choreographer of the mitotic stage? J Cell Biol 135, 1681-4.

16. Lee, K. S., Park, J. E., Asano, S., and Park, C. J. (2005). Yeast polo-like kinases: functionally conserved multitask mitotic regulators. Oncogene 24, 217-29.

17. Dai, W. (2005). Polo-like kinases, an introduction. Oncogene 24, 214-6. 18. Winkles, J. A., and Alberts, G. F. (2005). Differential regulation of polo-like kinase 1,

2, 3, and 4 gene expression in mammalian cells and tissues. Oncogene 24, 260-6. 19. Zimmerman, W. C., and Erikson, R. L. (2007). Polo-like kinase 3 is required for

entry into S phase. Proc Natl Acad Sci U S A 104, 1847-52. 20. Xie, S., Xie, B., Lee, M. Y., and Dai, W. (2005). Regulation of cell cycle

checkpoints by polo-like kinases. Oncogene 24, 277-86. 21. Bahassi el, M., Conn, C. W., Myer, D. L., Hennigan, R. F., McGowan, C. H.,

Sanchez, Y., and Stambrook, P. J. (2002). Mammalian Polo-like kinase 3 (Plk3) is a multifunctional protein involved in stress response pathways. Oncogene 21, 6633-40.

22. Yang, Y., Bai, J., Shen, R., Brown, S. A., Komissarova, E., Huang, Y., Jiang, N., Alberts, G. F., Costa, M., Lu, L., Winkles, J. A., and Dai, W. (2008). Polo-like kinase 3 functions as a tumor suppressor and is a negative regulator of hypoxia-inducible factor-1 alpha under hypoxic conditions. Cancer Res 68, 4077-85.

23. Wang, L., Dai, W., and Lu, L. (2007). Stress-induced c-Jun activation mediated by Polo-like kinase 3 in corneal epithelial cells. J Biol Chem 282, 32121-7.

24. Wang, L., Gao, J., Dai, W., and Lu, L. (2008). Activation of Polo-like kinase 3 by hypoxic stresses. J Biol Chem 283, 25928-35.

25. Seeburg, D. P., Pak, D., and Sheng, M. (2005). Polo-like kinases in the nervous system. Oncogene 24, 292-8.

26. Hudson, J. W., Kozarova, A., Cheung, P., Macmillan, J. C., Swallow, C. J., Cross, J. C., and Dennis, J. W. (2001). Late mitotic failure in mice lacking Sak, a polo-like kinase. Curr Biol 11, 441-6.

27. Lu, L. Y., Wood, J. L., Minter-Dykhouse, K., Ye, L., Saunders, T. L., Yu, X., and Chen, J. (2008). Polo-like kinase 1 is essential for early embryonic development and tumor suppression. Mol Cell Biol 28, 6870-6.


28. Donohue, P. J., Alberts, G. F., Guo, Y., and Winkles, J. A. (1995). Identification by targeted differential display of an immediate early gene encoding a putative serine/threonine kinase. J Biol Chem 270, 10351-7.

29. Li, B., Ouyang, B., Pan, H., Reissmann, P. T., Slamon, D. J., Arceci, R., Lu, L., and Dai, W. (1996). Prk, a cytokine-inducible human protein serine/threonine kinase whose expression appears to be down-regulated in lung carcinomas. J Biol Chem 271, 19402-8.

30. Chase, D., Feng, Y., Hanshew, B., Winkles, J. A., Longo, D. L., and Ferris, D. K. (1998). Expression and phosphorylation of fibroblast-growth-factor-inducible kinase (Fnk) during cell-cycle progression. Biochem J 333 ( Pt 3), 655-60.

31. Ouyang, B., Pan, H., Lu, L., Li, J., Stambrook, P., Li, B., and Dai, W. (1997). Human Prk is a conserved protein serine/threonine kinase involved in regulating M phase functions. J Biol Chem 272, 28646-51.

32. Xie, S., Wu, H., Wang, Q., Cogswell, J. P., Husain, I., Conn, C., Stambrook, P., Jhanwar-Uniyal, M., and Dai, W. (2001). Plk3 functionally links DNA damage to cell cycle arrest and apoptosis at least in part via the p53 pathway. J Biol Chem 276, 43305-12.

33. Kis, E., Szatmari, T., Keszei, M., Farkas, R., Esik, O., Lumniczky, K., Falus, A., and Safrany, G. (2006). Microarray analysis of radiation response genes in primary human fibroblasts. Int J Radiat Oncol Biol Phys 66, 1506-14.

34. Evan, G. I., and Vousden, K. H. (2001). Proliferation, cell cycle and apoptosis in cancer. Nature 411, 342-8.

35. Wiest, J., Clark, A. M., and Dai, W. (2001). Intron/exon organization and polymorphisms of the PLK3/PRK gene in human lung carcinoma cell lines. Genes Chromosomes Cancer 32, 384-9.

36. Dai, W., Li, Y., Ouyang, B., Pan, H., Reissmann, P., Li, J., Wiest, J., Stambrook, P., Gluckman, J. L., Noffsinger, A., and Bejarano, P. (2000). PRK, a cell cycle gene localized to 8p21, is downregulated in head and neck cancer. Genes Chromosomes Cancer 27, 332-6.

37. Spurr, N. K., Blanton, S., Bookstein, R., Clarke, R., Cottingham, R., Daiger, S., Drayna, D., Faber, P., Horrigan, S., Kas, K., and et al. (1995). Report and abstracts of the second international workshop on human chromosome 8 mapping 1994. Oxford, United Kingdom, September 16-18, 1994. Cytogenet Cell Genet 68, 147-64.

38. Bockmuhl, U., Wolf, G., Schmidt, S., Schwendel, A., Jahnke, V., Dietel, M., and Petersen, I. (1998). Genomic alterations associated with malignancy in head and neck cancer. Head Neck 20, 145-51.

39. Dai, W., Liu, T., Wang, Q., Rao, C. V., and Reddy, B. S. (2002). Down-regulation of PLK3 gene expression by types and amount of dietary fat in rat colon tumors. Int J Oncol 20, 121-6.

40. Weichert, W., Denkert, C., Schmidt, M., Gekeler, V., Wolf, G., Kobel, M., Dietel, M., and Hauptmann, S. (2004). Polo-like kinase isoform expression is a prognostic factor in ovarian carcinoma. Br J Cancer 90, 815-21.

41. Uckun, F. M., Dibirdik, I., Qazi, S., Vassilev, A., Ma, H., Mao, C., Benyumov, A., and Emami, K. H. (2007). Anti-breast cancer activity of LFM-A13, a potent inhibitor of Polo-like kinase (PLK). Bioorg Med Chem 15, 800-14.


42. Wang, Q., Xie, S., Chen, J., Fukasawa, K., Naik, U., Traganos, F., Darzynkiewicz, Z., Jhanwar-Uniyal, M., and Dai, W. (2002). Cell cycle arrest and apoptosis induced by human Polo-like kinase 3 is mediated through perturbation of microtubule integrity. Mol Cell Biol 22, 3450-9.

43. Jiang, N., Wang, X., Jhanwar-Uniyal, M., Darzynkiewicz, Z., and Dai, W. (2006). Polo box domain of Plk3 functions as a centrosome localization signal, overexpression of which causes mitotic arrest, cytokinesis defects, and apoptosis. J Biol Chem 281, 10577-82.

44. Conn, C. W., Hennigan, R. F., Dai, W., Sanchez, Y., and Stambrook, P. J. (2000). Incomplete cytokinesis and induction of apoptosis by overexpression of the mammalian polo-like kinase, Plk3. Cancer Res 60, 6826-31.

45. Ouyang, B., Li, W., Pan, H., Meadows, J., Hoffmann, I., and Dai, W. (1999). The physical association and phosphorylation of Cdc25C protein phosphatase by Prk. Oncogene 18, 6029-36.

46. Bahassi el, M., Hennigan, R. F., Myer, D. L., and Stambrook, P. J. (2004). Cdc25C phosphorylation on serine 191 by Plk3 promotes its nuclear translocation. Oncogene 23, 2658-63.

47. Nilsson, I., and Hoffmann, I. (2000). Cell cycle regulation by the Cdc25 phosphatase family. Prog Cell Cycle Res 4, 107-14.

48. Peng, C. Y., Graves, P. R., Thoma, R. S., Wu, Z., Shaw, A. S., and Piwnica-Worms, H. (1997). Mitotic and G2 checkpoint control: regulation of 14-3-3 protein binding by phosphorylation of Cdc25C on serine-216. Science 277, 1501-5.

49. Sanchez, Y., Wong, C., Thoma, R. S., Richman, R., Wu, Z., Piwnica-Worms, H., and Elledge, S. J. (1997). Conservation of the Chk1 checkpoint pathway in mammals: linkage of DNA damage to Cdk regulation through Cdc25. Science 277, 1497-501.

50. Lopez-Girona, A., Furnari, B., Mondesert, O., and Russell, P. (1999). Nuclear localization of Cdc25 is regulated by DNA damage and a 14-3-3 protein. Nature 397, 172-5.

51. Zimmerman, W. C., and Erikson, R. L. (2007). Finding Plk3. Cell Cycle 6, 1314-8. 52. Myer, D. L., Bahassi el, M., and Stambrook, P. J. (2005). The Plk3-Cdc25

circuit. Oncogene 24, 299-305. 53. Hoffmann, I., Draetta, G., and Karsenti, E. (1994). Activation of the phosphatase

activity of human cdc25A by a cdk2-cyclin E dependent phosphorylation at the G1/S transition. Embo J 13, 4302-10.

54. Jinno, S., Suto, K., Nagata, A., Igarashi, M., Kanaoka, Y., Nojima, H., and Okayama, H. (1994). Cdc25A is a novel phosphatase functioning early in the cell cycle. Embo J 13, 1549-56.

55. Iida, M., Matsuda, M., and Komatani, H. (2008). Plk3 phosphorylates topoisomerase IIalpha at Thr(1342), a site that is not recognized by Plk1. Biochem J 411, 27-32.

56. Wang, J. C. (2002). Cellular roles of DNA topoisomerases: a molecular perspective. Nat Rev Mol Cell Biol 3, 430-40.


57. Colanzi, A., and Corda, D. (2007). Mitosis controls the Golgi and the Golgi controls mitosis. Curr Opin Cell Biol 19, 386-93.

58. Lopez-Sanchez, I., Sanz-Garcia, M., and Lazo, P. A. (2009). Plk3 interacts with and specifically phosphorylates VRK1 in Ser342, a downstream target in a pathway that induces Golgi fragmentation. Mol Cell Biol 29, 1189-201.

59. Ruan, Q., Wang, Q., Xie, S., Fang, Y., Darzynkiewicz, Z., Guan, K., Jhanwar-Uniyal, M., and Dai, W. (2004). Polo-like kinase 3 is Golgi localized and involved in regulating Golgi fragmentation during the cell cycle. Exp Cell Res 294, 51-9.

60. Xie, S., Wang, Q., Ruan, Q., Liu, T., Jhanwar-Uniyal, M., Guan, K., and Dai, W. (2004). MEK1-induced Golgi dynamics during cell cycle progression is partly mediated by Polo-like kinase-3. Oncogene 23, 3822-9.

61. Kerr, J. F., Wyllie, A. H., and Currie, A. R. (1972). Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer 26, 239-57.

62. Kaufmann, S. H., and Gores, G. J. (2000). Apoptosis in cancer: cause and cure. Bioessays 22, 1007-17.

63. Williams, G. T. (1991). Programmed cell death: apoptosis and oncogenesis. Cell 65, 1097-8.

64. Li, Z., Niu, J., Uwagawa, T., Peng, B., and Chiao, P. J. (2005). Function of polo-like kinase 3 in NF-kappaB-mediated proapoptotic response. J Biol Chem 280, 16843-50.

65. Storz, P. (2005). Reactive oxygen species in tumor progression. Front Biosci 10, 1881-96.

66. Akman, S. A., O'Connor, T. R., and Rodriguez, H. (2000). Mapping oxidative DNA damage and mechanisms of repair. Ann N Y Acad Sci 899, 88-102.

67. Liu, Y., and Kulesz-Martin, M. (2001). p53 protein at the hub of cellular DNA damage response pathways through sequence-specific and non-sequence-specific DNA binding. Carcinogenesis 22, 851-60.

68. Rotman, G., and Shiloh, Y. (1997). Ataxia-telangiectasia: is ATM a sensor of oxidative damage and stress? Bioessays 19, 911-7.

69. Xie, S., Wang, Q., Wu, H., Cogswell, J., Lu, L., Jhanwar-Uniyal, M., and Dai, W. (2001). Reactive oxygen species-induced phosphorylation of p53 on serine 20 is mediated in part by polo-like kinase-3. J Biol Chem 276, 36194-9.

70. Xie, S., Wu, H., Wang, Q., Kunicki, J., Thomas, R. O., Hollingsworth, R. E., Cogswell, J., and Dai, W. (2002). Genotoxic stress-induced activation of Plk3 is partly mediated by Chk2. Cell Cycle 1, 424-9.

71. Matsuoka, S., Huang, M., and Elledge, S. J. (1998). Linkage of ATM to cell cycle regulation by the Chk2 protein kinase. Science 282, 1893-7.

72. Sanchez, Y., Bachant, J., Wang, H., Hu, F., Liu, D., Tetzlaff, M., and Elledge, S. J. (1999). Control of the DNA damage checkpoint by chk1 and rad53 protein kinases through distinct mechanisms. Science 286, 1166-71.

73. Folkman, J. (2006). Angiogenesis. Annu Rev Med 57, 1-18. 74. Zetter, B. R. (1998). Angiogenesis and tumor metastasis. Annu Rev Med 49,

407-24.


75. Pugh, C. W., and Ratcliffe, P. J. (2003). Regulation of angiogenesis by hypoxia: role of the HIF system. Nat Med 9, 677-84.

76. Yee Koh, M., Spivak-Kroizman, T. R., and Powis, G. (2008). HIF-1 regulation: not so easy come, easy go. Trends Biochem Sci 33, 526-34.

77. Kaelin, W. G., Jr., and Ratcliffe, P. J. (2008). Oxygen sensing by metazoans: the central role of the HIF hydroxylase pathway. Mol Cell 30, 393-402.

78. Flugel, D., Gorlach, A., Michiels, C., and Kietzmann, T. (2007). Glycogen synthase kinase 3 phosphorylates hypoxia-inducible factor 1alpha and mediates its destabilization in a VHL-independent manner. Mol Cell Biol 27, 3253-65.



7. Vaccinia-related kinase (VRK) signaling in cell and tumor biology

Marta Sanz-García, Alberto Valbuena, Inmaculada López-Sánchez

Sandra Blanco, Isabel F. Fernández, Marta Vázquez-Cedeira and Pedro A. Lazo

Experimental Therapeutics and Translational Oncology Program, Instituto de Biología Molecular y Celular del Cáncer, Consejo Superior de Investigaciones Científicas (CSIC) –

Universidad de Salamanca, Campus Miguel de Unamuno, 37007 Salamanca, Spain

Abstract. VRK (vaccinia-related kinase) is a group of three proteins in the human kinome. These proteins, mainly VRK1 and VRK2, have been studied in the context of their substrates and interacting proteins in order to identify and characterize their signaling pathway, as well as their effect on other signaling pathways. VRK1 is mostly a nuclear kinase that specifically phosphorylates p53, c-Jun, ATF2, CREB, BAF and histone H3. VRK1 is an early response gene and is implicated in regulation of cell cycle progression. VRK1 is activated in response to DNA damage phosphorylating p53, which is stabilized and activated; this active p53 induces a downregulatory mechanism of VRK1 that permits the reversal of p53 induced effects. The activity of nuclear VRK1 is regulated by its interaction with the Ran small GTPase. Also, VRK1 is a downstream component of the signaling pathway of MEK-Plk3 that induces Golgi fragmentation in mitosis. VRK2

Correspondence/Reprint request: Dr. P.A. Lazo, IBMCC-Centro de Investigación del Cáncer, CSIC-Universidad de Salamanca, Campus Miguel de Unamuno, E-37007 Salamanca, Spain E-mail: [email protected];

Marta Sanz-García et al. 136

has two isoforms; VRK2A is cytosolic and bound to endoplasmic reticulum and mitochondrial membranes. VRK2B is a shorter isoform free in cytosol and nucleus. VRK2A affects cellular signaling by interaction with scaffold proteins, as JIP1. The JIP1-VRK2A signalosome blocks the incorporation of JNK, preventing its activation, and thus reducing the stress response to inflammatory cytokines as interleukin-1β and to hypoxia. 1. The vaccinia-related kinase (VRK) family The completion of the human genome project has led to the identification of 518 proteins that constitute the human kinome [1]. Approximately half of them are not well characterized regarding the signaling pathways in which they are integrated, and their biological functions. But in most likelihood they can probably account for the specificity and regulation of many biological functions. VRK (vaccinia-related kinases) were classified as a new group of Ser-Thr kinases in the human kinome [1], and originally were identified as two human EST, VRK1 and VRK2, which have homology to catalytic region of the B1R kinase of vaccinia virus [2] that is expressed early in viral infection since it is required for viral DNA replication [3-5]. The VRK gene family appeared late in evolution. In lower eukaryotes, D. melanogaster (NHK-1) and C. elegans (Vrk1), there is only one member, and no ortholog has been identified in unicellular eukaryotes, like yeasts S. cerevisiae or S. pombe. In mammals there are three members, two of which (VRK1 and VRK2) are catalytically active [6-8], and a third one, VRK3, is not active and might function as a scaffold protein [9]. The VRK catalytic domain is distantly related to the casein kinase group, forming an early divergent branch in the human kinome [1], but the conservation of the kinase domain is weak and has substitutions in several key residues in this domain [10]. These proteins maintain the structure of the kinase domain, but the rest of the protein is very divergent among them, suggesting that the CK1-like catalytic domain recombined early in evolution with other proteins to form a divergent protein family with new regulatory properties, affecting interactions, regulation, and subcellular localization. In one of these recombination events the kinase domain was picked up by pox viruses (vaccinia, variola, smallpox) and generated the B1R protein, which is the only kinase in poxviruses genome; B1R is required for poxvirus replication and its inactivating mutations can be partially recovered by overexpression of mammalian VRK1, or VRK2 proteins, indicating that in poxvirus there was an additional functional divergence due to viral life requirement [11]. Structurally, the three VRK proteins differ in their regulatory region located C-terminal for VRK1 and VRK2 and N-terminal for VRK3. These regulatory regions are unrelated

Vaccinia-related kinase (VRK) signaling in cell and tumor biology 137

among them and have no homology to any known feature identified in other proteins. The length of the regulatory region is variable. It is approximately one hundred amino acids in VRK1 and two hundred aminoacids in VRK2, thus their subcellular localization and regulation is likely to be different among mammalian VRK proteins and those of lower eukaryotes. The relative conservation of catalytic domain of VRK proteins indicates they can have an overlap of potential substrates. The catalytic domain of these kinases have some key substitutions which make them differ from other closely related kinases, nevertheless two of them VRK1 and VRK2 are catalytically active and have a 56 % sequence identity [6, 8, 12]. These differences might be exploited for the design of VRK specific inhibitors [13], which may be achieved by determination of their crystal structure. The crystal structure of the catalytic domain of VRK2 and VRK3, with 38 % sequence identity and without any significant difference, has been determined providing the framework for future inhibitor development [10]. Human VRK1 gene is located on the chromosomal region 14q32.2. This gene has a polymorphism, marker rs722869, that in combination with other nine polymorphisms in other genes has proven very useful and is included in the most informative set to study the identification of human continental population structure and genetic ancestry [14]. VRK1 gene expression has been studied in hematopoietic murine development, and its highest expression coincides with the time of maximum cellular expansion in days E11.5 to E13.5 [15]. Endogenous VRK1 protein is detected in different

Figure 1. Structure of the three human VRK proteins. ATP: ATP binding site; KD: kinase domain; Plk: consensus target for Plk3. NLS: nuclear localization signal. BAB: Basic-acid-basic region. TM: transmembrane region.


intracellular compartments, suggesting that localization is regulated and that it might play different functional roles depending on its localization. VRK1 has a canonical nuclear localization signal (KKRKK) in residues 356 – 360, which accounts for its nuclear localization [6, 8]. In interphase the endogenous VRK1 has a granular aspect, probably forming aggregates, dispersed within the nucleus, and in mitosis it is dispersed throughout the cell and does not bind to chromatin. In some cell types the endogenous VRK1 protein is mostly cytosolic presenting a particulate appearance, and a minor fraction is nuclear [16], but the underlying cause is unknown. Also there is a subpopulation in the Golgi apparatus where VRK1 colocalizes with giantin where it participates in the control of Golgi fragmentation, and perhaps in VRK1 recycling [17]. This differential distribution is tissue specific [16]. These different locations of human VRK1 are also detected in human normal tissue biopsies. For example in testes and esophagus VRK1 is detected as nuclear with all antibodies. However in small intestine or kidney, a nuclear and a cytosolic subpopulation are detected depending on the antibody used [16]. The factors determining, and regulating, these different subcellular localizations are unknown. The human VRK2 gene has 14 exons and can generate two different transcripts by alternative splicing, one lacking exon 13 that contains a termination codon, encoding two proteins of 508 (VRK2A) or 397 (VRK2B) aminoacids respectively. The largest protein is called VRK2A, and the other is VRK2B [12], both are identical up to residue 393 and have a common kinase domain. The VRK2A protein has, at the end of its carboxy terminus, a hydrophobic region of 17 aminoacids that anchors the protein to membranes in endoplasmic reticulum and mitochondria. VRK2A colocalizes with calreticulin and calnexin indicating its presence in the endoplasmic reticulum [12]; it also colocalizes with mitotracker, a lipidic marker of mitochondria [12]. VRK2B, lacking its hydrophobic tail, is free in the cytosol and the nucleus of the cell. The shorter isoform B is more similar in size to VRK1, lacking the membrane anchor sequence, and detected in those cells where VRK1 is mostly cytosolic, suggesting they might have some functional redundancy, but how their expression is coordinated is not known. Therefore, although both proteins might have similar or identical kinase properties, their natural substrates may be different due to their different subcellular localization. Additional variants of VRK2 have been identified by RT-PCR, but the expected corresponding proteins have not been detected and may represent splicing intermediates. Many cell types have been studied for the expression of both isoforms, such as normal B-cells, several lymphoma cell lines and several frequently used carcinoma cell lines such as HeLa, A549, H1299, MCF7, among others. All of them expressed both messages [12]. All


of them express VRK2A protein, but the level of the VRK2B protein was variable, and in some cell lines was not detectable [12]. The human VRK3 gene is located on 19q13.33, and codes for a protein with a degenerate kinase domain located in the carboxy terminal region that has no kinase activity due to critical aminoacid substitutions [8]. Its crystal structure has revealed important structural differences that make it unable to bind ATP [10]. VRK3 non-functional kinase-dead domain maintains its overall structure so that it can function as scaffold for other proteins [10]. The VRK3 protein is detected in the nucleus and has a bipartite nuclear localization signal [8, 9]. In humans the expression of VRK proteins has been detected in all tissues studied; but not all cells in a tissue express these proteins. The reason for this heterogeneous expression is not yet known, but it may be associated to the proliferation and differentiation state of individual cells. In normal epithelium higher levels are detected near the basal layer, and the intensity of the signal decreases as the cell matures [18]. 2. Biological roles of VRK proteins The biological information available on the three VRK proteins is very limited, and most of it was obtained by studying the human proteins in two main directions. One of them implies the characterization of the signaling pathway in which VRK proteins are implicated, either by substrate identification or by interaction with other proteins. The other is directed at their effect on some biological response, such as hypoxia, interleukin1β or DNA damage. In both situations it is expected that information will expand as these proteins acquire more relevance in the context of tumor biology, and start to attract more attention. In the case of VRK1 some of its targets have been identified, including several transcription factors (Fig. 2), which can serve as starting points to characterize the VRK pathway by itself, or their interaction with other pathways. In this review we will summarized the different biological processes where VRK proteins have been shown to participate. 3. VRK proteins and cell signaling

3.1. VRK2 downregulates the response to hypoxia or interleukin-1β mediated by JNK Many biological processes induced by growth factors, stress responses send signals that activate gene expression. These signals are channeled by a


protein complex of mitogen-associated protein kinases (MAPK) which are activated by two additional kinases, MAPKK and MAPKKK, ordered as consecutive steps. MAP kinases represent the core of multiple signaling pathways in response to a variety of stimuli, ranging from growth factors to stress responses. The response implies multiple biological effects such as proliferation, apoptosis, growth arrests and senescence among others [19, 20]. For each of these steps there are multiple kinases, making up multiple combinations that are very likely to determine the specificity of the responses. The three kinases are often anchored on a scaffold protein; the best known are JIP1, of which there are four members, and KSR1 [21]. But there are also other less well known scaffolds. The levels and subcellular localization of scaffold proteins are critical to determine the distribution of a signal among different signaling pathways, and the asymmetry of this signal distribution can determine the final biological effect. These complexes might be further regulated by additional proteins that might interact, activate or inhibit the signal transmission; among these proteins are the two VRK2 isoforms. Therefore signal specificity is likely to be determined by interactions between components of the core signaling pathway with other proteins, most of which have not yet been identified, but is an emerging field. Among these proteins, VRK2A and VRK3 have been shown to interact and regulate MAP kinase signaling in response to different types of stimulation. 3.2. VRK2 downregulates MAP kinase signaling and JNK activation Many biological processes are controlled by regulation of the activity of mitogen-activated protein kinases which function like a central hub where signals are received and distributed to exert specific roles. MAPK (mitogen-activated protein kinases) are implicated in a large number of cellular responses to growth factors and stress signals of different types. MAPK signals are transmitted by a complex of three consecutive kinases and for each step there are several possibilities. The combination of these kinases contributes to signal specificity. The three consecutive MAP kinases are assembled on scaffold proteins, of which there are several types [21]. These modules permit the organization of specific signal transmission and can be further modulated by specific interactions with other proteins. One of these new regulators is VRK2, which by its interaction with scaffold–MAPK complexes can alter the balance between different pathways responding to a common signal. Signals regulating proliferation are mainly transmitted by the RAF-MEK-ERK proteins assembled on the KSR1 scaffold, while signals implicated in stress responses are channeled by the TAK1-MKK4/7-JNK complex assembled on JIP proteins [21]. The effect of VRK proteins on


signaling by scaffold proteins has been studied using two different experimental systems; one based on the response to stress mediated by the JNK (c-Jun N-terminal kinase) pathway and the other in response to growth factors such as EGF and mediated by the ERK pathway. The best characterized one is the complex assembled on JIP1. JIP1 interacts with JNK and MKK4, or MKK7, as the third and second kinase respectively. However, responses to a common stimulation using this complex may vary depending on cell type, and the reasons for the difference are not known, but are likely to be related to additional proteins which might modulate the signalosome or signaling complex. The JIP1 scaffold protein is usually thought of as a monomer to which different kinase can bind. However, it is known that JIP1 can also be associated to membranes and form large complexes [22]. In our laboratory we have identified that JIP1 form complexes with a size of 1200 kDa, which includes the VRK2A protein [23, 24]. These complexes can incorporate VRK2, and the composition of the complex can be altered if cells are stimulated [23].

Figure 2. Model of the complexes assembled on the JIP1 scaffold protein which can incorporate VRK2 protein and is modulated by interleukin-1β or hypoxia.


The VRK2A protein subpopulation that is anchored on the endoplasmic reticulum (ER) [12], colocalizes with JIP1 on the ER which suggest that they can interact and modulate MAPK signaling [24]. JIP1 by interacting with ER-anchored VRK2 can be incorporated and form larger complexes; JIP1 up to now has been thought of as a monomer, and by this incorporation in membrane complexes JIP1 can be compartmentalized instead of being free in the cytosol. The formation of these complexes provides the cell with two different signaling alternatives, depending on location and aggregation state, and probably with different substrates. The formation of the signalosome is likely to be assembled on VRK2A protein anchored to membranes, because it is not formed by the VRK2B isoform that is free in the cytosol and nuclei (Fig. 2). The presence of VRK2A is a major requirement to induce the oligomerization state and assembly of the signalosome. In the presence of VRK2A complexes are very large (1000 kDa), but when VRK2A was knocked down the complexes were disassembled, and many of its components were detected as free or forming smaller and incomplete complexes [23, 24]. The effect of VRK2-JIP1 interaction has been studied under two types of stimulation, one in response to the inflammatory interleukin-1β (IL-1 β)[24] and the other in response to hypoxia [23], IL-1 β and hypoxia induce the formation of a large complexes of more than 1000 kDa, known as signalosomes, and containing MAPK and JIP1 protein, and among its components is present the VRK2A [23, 24]. In the complex VRK2A directly interacts with JIP1 and TAK1. The stable interaction of VRK2A with JIP1 can alter the composition of MAPK bound to JIP1. Binding of VRK2A to JIP1 does not affect JIP1 interaction with TAK1 or MKK7, but reduces its binding to JNK, and thus limits the potential activation of c-Jun dependent transcription, which can not be activated because JNK is inactive (Fig. 2). 3.3. VRK3 downregulates ERK activity by interaction with the VHR phosphatase The nuclear VRK3 protein is able to downregulate ERK signaling by a mechanism different from the one involving VRK2 and MAPK. Because VRK2 is not catalytically active it can only function by mediating protein-protein interactions, where it plays a scaffold role. The effect of VRK3 on MAPK signaling has been characterized by its role in modulating nuclear signals mediated by ERK. ERK (extracellular regulated kinase) is a MAPK that has been implicated in the control of proliferation and cell growth [25, 26]. ERK is the last kinase in the MAPK complex responding to EGF, ERBB2, Ras and Raf, which is most commonly assembled on the KSR1


scaffold. Activation of the pathway by growth factors (EGF) or oncogenes (ERBB2, RasG12V, or BRAFV600E) induces a phosphorylation of ERK (p-ERK) that has substrates in the cytosol and in the nucleus. Among its cytosolic targets is p90RSK (ribosomal S6 kinase) that activates pleiotropic response associated to cell growth and increase in cell mass [27]. The activated p-ERK is translocated to the nucleus and mainly phosphorylates a large number of transcription factors including ETS, ELK1, Pax6 or c-FOS among others, and their relative levels and activation might condition activation of one type of response or another. One of these pathways is represent by ERK (extracellular-signal regulated kinase). These kinases are regulated by phosphorylation and the level of phosphorylation is control by proteins known as mitogen-activated protein kinase phosphatases (MKPs), the expression or stabilization of MKP activity thus negatively controls ERK activation forming a feed-back regulatory mechanism [28]. In this context, the VRK3 protein, the least known member of the VRK family, appears to control the level of ERK activation [9, 29]. This effect is indirect and mediated by an interaction with the VHR (vaccinia H1-related). In the nucleus VHR dephosphorylates ERK resulting in its inactivation. The nuclear binding of VRK3 to VHR increases its phosphatase activity, and for this process the kinase activity is not necessary, thus this effect is postactivation in the cytosol and therefore its role is to mediate a downregulation, or silencing, the nuclear action of phospho-ERK [9, 29]. In this context, VRK3 functions as a scaffold protein that binds to the VHR phosphatase, which removes the phosphate from an activated MAPK, as is the case of ERK, and thus inhibits or downregulates its nuclear signal. 4. Modulation of transcription factors by VRK proteins 4.1. VRK1 activation of transcription factors can cooperate or be an alternative to MAPK signals Among the phosphorylation targets of hVRK1 there are transcription factors that are generally regulated by MAPK signaling. Some of these transcription factors can be directly phosphorylated by VRK1 and VRK2B isoform resulting in enhancement of their transcriptional activity, including c-Jun [30], ATF2 [31], CREB [32] and p53 [6]. The nature of these transcription factors reveals potential functional implications for VRK proteins. ATF2, CREB and c-Jun are members of the bZIP family of proteins that can bind to the AP-1 sites to regulate gene expression [33]; this activation is mediated by dimers of these proteins, and depending on the two partners the specificity of the gene selected can be determined.


The c-Jun protein, a component of the AP1 transcription sites, is activated by phosphorylation by JNK (c-Jun N-terminal kinase) in response to stress signals that regulate apoptosis, proliferation and development [34, 35]. The phosphorylation of c-Jun by VRK1 results in the transcriptional activation of Jun independently of the JNK signal. The oncogenic c-Jun can be directly phosphorylated by VRK1 and VRK2B in residues Thr63 and Ser73, which are the same as those phosphorylated by JNK (N-terminal kinase of c-Jun), activating c-Jun dependent transcription [30]. Thus, in situation of suboptimal stimulation, the two kinases might have an additive effect (Fig. 3) and reach maximum c-Jun transcriptional activation [30]. This generates an interesting situation in which c-Jun phosphorylation might be inhibited by the action of cytosolic VRK2A that blocks its activation as a consequence of VRK2A interaction with the JIP1 scaffold protein [24], thus the stress signal response is blocked, as is the case in hypoxia response [23], but at the same time permits its activation by a different route mediated by VRK1, which is regulated in a different way that remains to be identified, but which is associated with entry into cell cycle or responses to DNA damage.

Figure 3. Diagram illustrating the different substrates identified for the human VRK1 kinases and the overlap of its signals with other signaling pathways [6, 30-32].


The ATF2 and CREB transcriptions factors are implicated in pleiotropic responses. VRK1 phosphorylates ATF2 in Ser62 and Thr73 and activates transcription [31]. These two residues are very proximal to those targeted by JNK, Thr69 and Thr71; all within the same region (Fig. 3). These slightly different positions might regulate ATF2 by differential interaction properties with other transcriptional cofactors, and thus affect the specificity of gene transcription. Furthermore, these differences in phosphorylation target can generate a potential cooperative activation by suboptimal stimulation of each kinase, VRK1 and JNK, or respond alternatively to one signaling route or the other. But also the ATF2 residues targeted by VRK1 are the same as those targeted by calmodulin-dependent kinase IV or protein kinase A [31], permitting cooperation between more than one signaling pathways [31]. Thus at least four signaling routes can converge on ATF2 (Fig. 3). Another target of VRK1 is CREB, a factor identified in response to cAMP. The CREB transcription factor can also be activated by phosphorylation in Ser133 by VRK1, and is required for transcription of cyclinD1 (CCND1), VRK1 forms part of the transcriptional complex of this gene in the G1 phase of the cell cycle. This CREB activation was previously known as an effect mediated by a cAMP-response element in CCND1 promoter [32]. Thus two transcription factors responding to cAMP, ATF2 and CREB, are regulated by VRK1, but the functional interaction between VRK1-mediated pathway and cAMP responses are not known. 4.2. Regulation of p53 by VRK1, a novel regulatory loop between VRK1 and p53 in response to DNA damage mediated by DRAM The tumor suppressor p53 was the first target identified for VRK proteins. VRK proteins can stabilize and regulate p53 by a unique phosphorylation of Thr18. VRK1 [6, 36] and VRK2 [12] phosphorylate p53 specifically is Thr18, a residue that is critical to maintain the loop structure in its N-terminal tansactivation domain (TAD). The stabilization is a consequence of the disruption of the hydrogen bond between Thr18 and Asp21 in p53 necessary to maintain the structure of the loop that binds to Hdm2/Mdm2 [37, 38]. This Thr18 is the critical residue controlling the selection of binding partner by p53 TAD, its phosphorylation increases seven fold its binding to transcriptional coactivators, such as p300 [39], and dephosphorylated p53 interacts preferentially with Hdm2. The effect on p53 interactions is very consistent with the consequences of phosphorylation; when p53 is phosphorylated in Thr18 there is a reduction in ubiquitination and an increase in acetylation, as well as in transcriptional activity [36]. This Thr18 specific phosphorylation changes by three orders of magnitude the


differential binding from Hdm2 to p300 and TAZ1 [39]. Other well-known residues such as Ser15 or Ser20 have a much weaker contribution to selection of binding partner and play a secondary role in this context, although they are the most characterized in response to DNA damage by the ATM/CHK2 [40, 41]or ATR/CHK1 pathways [42]. The stabilization of p53 by VRKs renders a p53 molecule that cannot be degraded and thus accumulates in the cell; a persistent accumulation of p53 will not permit life because it induces either a blockade of cell cycle progression or apoptosis. In addition, the VRK1 protein that functions as a p53 activator is very stable with a half life of four days thus if activated it will maintain p53 in a non-degradable and stabilized form. These characteristics suggested that regulation of VRK proteins level probably requires an active mechanism of protein degradation, which will permit p53 downregulation at the same time. Therefore it is very likely that an autoregulation between p53 and VRK1 must exist. There was an initial observation which suggested that such potential mechanism was functioning because in cell lines that have a high level of p53 there was a lower level of VRK1 compared with cell lines that have a smaller amount of p53, and a higher level of VRK1 [43]. In experimental systems, overexpression of p53 was always followed by a reduction in VRK1 level [43]. The drop in VRK1 protein was not accompanied by a reduction in VRK1 gene expression suggesting it was an indirect effect of p53. Nevertheless, this downregulation was dependent on p53 transcriptional activity, probably by regulating the expression of a not yet identified protein that is the one that targets VRK1 for degradation [43]. The structural requirements of the p53 molecule required to induce VRK downregulation have been identified using different p53 variants containing mutations, deletions or conformational mutants [43]. The N-terminal TAD (trans activation domain) was studied with the p53L22Q/23WS conformational mutant, phosphorylation mutants in all its Ser or Thr residues; none of them affected VRK1 downregulation [43]. The p53 DNA binding domain was studied using the three most common mutations detected in sporadic or hereditary (Li-Fraumeni syndrome) human cancer. The conformational mutant p53R175H, or the p53R248W and p53R273H DNA-contact mutants. All these mutants were unable to induce downregulation of VRK1 levels [43]. Thus a functional p53 DNA binding domain is necessary and supports the requirement of p53 transcriptional dependence of this VRK1 downregulation. The VRK1 protein is not ubiquitinated by Hdm2, or other ubiquitin ligases, as demonstrated by its insensitivity to either overexpression of mdm2 or to proteasome inhibitors [43], suggesting that the proteasome pathway is not implicated in VRK1 downregulation. Alternatively, downregulation of


VRK1 is sensitive to inhibitors of late-endosome to lysosome transport such as chloroquine [43], and to inhibitors of lysosomal protease activity. These results suggested that p53 induces the expression of a protein that regulates VRK1 and targets it for lysosomal degradation. This protein is DRAM, a lysosomal membrane protein of 238 aminoacids and has six transmembrane domains equally spaced [44]. DRAM expression is dependent on activation of transcription by p53, and expression is lost in case of common p53 mutations in human cancer, such as p53R175H, p53R248W and p53R273H. These requirements are identical as those for VRK1 downregulation induced by p53, thus DRAM is a candidate to be its mediator, and its overexpression of DRAM downregulates VRK1. The induction of DNA damage by ultraviolet light results in accumulation of p53, followed by expression of DRAM and downregulation of VRK1. Thus p53 has a double autoregulatory loop to control its intracellular levels, activated p53 is phosphorylated and induces HDM2 and DRAM gene expression. But hdm2 ca not degrade p53 if it is phosphorylated, thus the removal of VRK1 by its lysosomal degradation mediated by DRAM permits the accumulation of unphosphorylated p53 that is now susceptible to ubiquitination by hdm2 and can be degraded in the proteasome (Fig. 4).

Figure 4. Autoregulatory loop of VRK1 dependent on p53. Phosphorylated p53 activates transcription of Mdm2 that downregulates p53 and of DRAM that downregulates VRK1.


5. VRK proteins in processes required for mitosis 5.1. VRK1 expression and cell cycle Expression of VRK1 and VRK2 was originally associated to proliferation, because of their expression in proliferating tissues such as testis, thymus, and fetal liver [2] and in the rapid proliferation stage of B-cell development in murine embryos [15]. The expression of human VRK1 gene is regulated in cell cycle progression. Serum withdrawal results in endogenous VRK1 gene silencing in human fibroblast, parallel to the loss of phopho-Rb, and also in loss of expression dependent on the human VRK1 gene promoter coupled to a luciferase reporter [45]. Addition of serum induces expression of VRK1 RNA at the same time as MYC, FOS and CCND1 (cyclinD1), indicating VRK1 is an early response gene. Knock-down of endogenous VRK1 results in lack of induction of cyclin D1 and a cell cycle block, early in G1 [36, 45]. In human head and neck squamous cell carcinomas (HNSCC), VRK1 protein is expressed at high levels and positively correlates with several proliferation markers such as Ki67, cyclins B1 and A, cdk2, cdk6 and cdc2. In addition in human lung carcinomas, there were higher levels of VRK1 in those tumors that have a p53 mutation, consistent with an inactivation of the p53 dependent autoregulatory loop [46]. 5.2. VRK1, in the control of nuclear membrane formation The VRK1 and VRK2 proteins, as well as the vaccinia virus B1R kinase, phosphorylate the BAF1 protein (barrier to autointegration factor 1), a 10 kDa protein that forms dimers. The BAF1 protein binds to DNA in a non-sequence specific manner [47], and to proteins containing LEM domains, which are present in the inner nuclear membrane [48]. VRKs and B1R phosphorylate residues Ser4 or Thr2/Thr3 in the N-terminal domain of BAF1. The phosphorylated BAF1 presents a reduction in its binding to DNA and LEM-containing proteins [49]. Phosphorylation of BAF1 in Ser4 delocalizes emerin and interferes with emerin binding to lamin A, both in mitosis and interphase [50]. The overexpression of VRKs reduces BAF1 interaction with nuclear chromatin and results in its dispersal [49]. Nuclear disassembly is a process necessary for cell cycle progression, thus BAF1 phosphorylation by VRK1 might be dependent on functional changes in the activity of VRK1, a kinase that is regulated in cell cycle, and required for G0 exit and G1 progression [45], and its knock down causes a block of cell cycle progression [36, 45]. VRK1 is expressed as the cells enter the cycle from G0 to G1 and


the highest levels are achieved at G1/S transition [18]. The VRK1 protein is very stable and therefore is ready to act when the nuclear envelope has to disintegrate in order to permit chromosome segregation. The C. elegans Vrk-1 protein, a unique ortholog of human VRK proteins, also phosphorylates the BAF1 protein and affects nuclear membrane structure [51]. But C. elegans Vrk-1 protein is much larger, 610 aminoacids, completely unrelated to that of human VRK1 protein.. The conservation is restricted to the kinase domain, but there is no relation between their carboxyterminal regions. Thus C. elegans and human VRK1 have a common substrate, but probably a different regulation because of their different C-terminal region. Vaccinia virus B1R kinase also has among its substrates the BAF protein that is required in vaccinia life cycle at a time when disintegration of the nuclear envelope takes place, and this function can be partially rescued by overexpressed human VRK1 in virus with B1R mutations [49]. 5.3. Regulation of VRK1 activity by interaction with Ran and nuclear dynamics The enzymatic activity of VRK1 can be regulated by different mechanisms that can affect either its kinase activity or its substrate specificity. These regulatory mechanisms include either protein interactions or covalent modifications. Using a proteomics based approach several interacting proteins have been identified and some of them have already been validated and are now components of emerging VRK signaling pathways. The first allosteric regulator of VRK proteins is the small GTPase Ran, the only nuclear member of the large Ras GTPase family. Ran stably interacts with the three VRK proteins, but only the Ran interaction with VRK1 has been characterized [52]. Small GTPases have two forms depending on the bound nucleotide; inactive or bound to GDP, and active or bound to GTP. Ran-GDP is mostly located in the cytosol and participates in the nuclear transport mechanism; Ran-GTP, is located in the nucleus, and the nucleotide exchange from Ran-GDP to Ran-GTP is mediated by the action of RCC1 its GEF (Guanine exchange factor) in the nucleus. During mitosis Ran-GTP can form a concentration gradient, with the highest concentration near condensed chromosomes [53]. VRK1 contributes to chromatin condensation by phosphorylation of histone H3 [52, 54]. Ran-GTP binds to VRK1 but does not alter its activity, but VRK1 binding to Ran-GDP inhibits VRK1 kinase activity [52], permitting the formation of a nuclear gradient of VRK1 activity in the nucleus depending on its partner interactions. VRK1 will be activated near the chromosomes, where it can participate in chromatin condensation by


phosphorylation of histone H3 [54], a phosphorylation that is lost by knocked-down of VRK1 [52]. A diagram of the regulation of VRK1 by the Ran small GTPase is shown in Fig. 5. Also some histones are substrates of VRK proteins. NHK (nucleosomal histone kinase), is the unique VRK ortholog in Drosophila melanogaster, and phosphorylates mitotic histones H2A which is required for acetylation of H3 [55-57], a phosphorylation also required for mitotic progression [58]. Human VRK1 is able to phosphorylate histone H3 in Ser10 resulting in chromatin condensation, and cooperating with Aurora B, a process also required for progression of mitosis [54]; this phosphorylation of H3 can be inhibited if human VRK1 interacts with RanGDP [52].

Figure 5. Interaction between VRK1 and Ran. Effect on VRK1 kinase activity depending on the loading state of the small GTPase Ran. [52].


5.5. The MEK-Plk3-VRK1 pathway: Specific phosphorylation of VRK1 in Golgi fragmentation Golgi fragmentation is a necessary process during mitosis in order to redistribute this organelle into daughter cells. The regulation of this fragmentation is not well known, but mitogenic signals mediated by MAP kinases, particularly MEK1, that induce Golgi fragmentation [59-62], representing an upstream component. A step downstream of MEK1 is mediated by plk3 (polo-like kinase 3) [63], a member of the plk protein family [64]. There is a subpopulation of VRK1, detected with a specific antibody, which is partially located in Golgi colocalizing with markers such as giantin of GM130 [17], which suggests VRK1 might participate in processes regulating Golgi functions. VRK1 participates in this process as a downstream target of Plk3, since VRK1 has a consensus sequence for phosphorylation by Plk proteins. Plk3 phosphorylates VRK1 in Ser340 [17]. Knocking-down VRK1, or the use of a catalytically inactive VRK1K179E, blocks Golgi fragmentation induced by either MEK1 or Plk3 [17]. Thus, one of the cytosolic VRK1 roles is to contribute to Golgi fragmentation in G2/M in mitosis, as a new downstream step of this pathway in mitosis (Fig. 6).

Figure 6. VRK1 is a downstream step in the signaling pathway required for Golgi fragmentation in mitosis [17]. Conclusions VRK proteins are a group of new human serine-threonine kinases forming a distinctive branch in the human kinome. Among its targets there are a variety of transcription factors implicated in the response to DNA damage, such as p53, or stress responses such as p53 or c-Jun. VRK1


participate in several biological processes including regulation of cell cycle entry and progression, response to DNA damage, nuclear envelope assembly and Golgi fragmentation. VRK2 also modulates signaling by MAPK pathways playing and inhibitory role after its incorporation in a signalosome with scaffold proteins. The VRK signaling pathway is still partially characterized and is expected that its contribution to cell and tumor biology will increase significantly in a near future. Acknowledgements This work was supported by grants from Ministerio de Educación y Ciencia e Innovación (SAF2007-60242 and CSD2007-0017), Junta de Castilla y León (Consejería de Educación, CSI-14A08 and GR15; Consejería de Sanidad) and Fundación Sandra Ibarra. References 1. Manning G, Whyte DB, Martinez R, Hunter T & Sudarsanam S (2002) The

protein kinase complement of the human genome. Science 298, 1912-1934. 2. Nezu J, Oku A, Jones MH & Shimane M (1997) Identification of two novel

human putative serine/threonine kinases, VRK1 and VRK2, with structural similarity to vaccinia virus B1R kinase. Genomics 45, 327-331.

3. Banham AH & Smith GL (1992) Vaccinia virus gene B1R encodes a 34-kDa serine/threonine protein kinase that localizes in cytoplasmic factories and is packaged into virions. Virology 191, 803-812.

4. Lin S, Chen W & Broyles SS (1992) The vaccinia virus B1R gene product is a serine/threonine protein kinase. J Virol 66, 2717-2723.

5. Rempel RE & Traktman P (1992) Vaccinia virus B1 kinase: phenotypic analysis of temperature-sensitive mutants and enzymatic characterization of recombinant proteins. J Virol 66, 4413-4426.

6. Lopez-Borges S & Lazo PA (2000) The human vaccinia-related kinase 1 (VRK1) phosphorylates threonine-18 within the mdm-2 binding site of the p53 tumour suppressor protein. Oncogene 19, 3656-3664.

7. Barcia R, Lopez-Borges S, Vega FM & Lazo PA (2002) Kinetic Properties of p53 Phosphorylation by the Human Vaccinia-Related Kinase 1. Arch Biochem Biophys 399, 1-5.

8. Nichols RJ & Traktman P (2004) Characterization of three paralogous members of the Mammalian vaccinia related kinase family. J Biol Chem 279, 7934-7946.

9. Kang TH & Kim KT (2006) Negative regulation of ERK activity by VRK3-mediated activation of VHR phosphatase. Nat Cell Biol 8, 863-869.

10. Scheeff ED, Eswaran J, Bunkoczi G, Knapp S & Manning G (2009) Structure of the Pseudokinase VRK3 Reveals a Degraded Catalytic Site, a Highly Conserved Kinase Fold, and a Putative Regulatory Binding Site. Structure 17, 128-138.


11. Boyle KA & Traktman P (2004) Members of a Novel Family of Mammalian Protein Kinases Complement the DNA-Negative Phenotype of a Vaccinia Virus ts Mutant Defective in the B1 Kinase. J. Virol. 78, 1992-2005.

12. Blanco S, Klimcakova L, Vega FM & Lazo PA (2006) The subcellular localization of vaccinia-related kinase-2 (VRK2) isoforms determines their different effect on p53 stability in tumour cell lines. FEBS J 273, 2487-2504.

13. Fedorov O, Marsden B, Pogacic V, Rellos P, Muller S, Bullock AN, Schwaller J, Sundstrom M & Knapp S (2007) A systematic interaction map of validated kinase inhibitors with Ser/Thr kinases. Proc Natl Acad Sci U S A 104, 20523-20528.

14. Lao O, van Duijn K, Kersbergen P, de Knijff P & Kayser M (2006) Proportioning whole-genome single-nucleotide-polymorphism diversity for the identification of geographic population structure and genetic ancestry. Am. J. Hum, Genet. 78, 680-690.

15. Vega FM, Gonzalo P, Gaspar ML & Lazo PA (2003) Expression of the VRK (vaccinia-related kinase) gene family of p53 regulators in murine hematopoietic development. FEBS Lett 544, 176-180.

16. Valbuena A, Lopez-Sanchez I, Vega FM, Sevilla A, Sanz-Garcia M, Blanco S & Lazo PA (2007) Identification of a dominant epitope in human vaccinia-related kinase 1 (VRK1) and detection of different intracellular subpopulations. Arch Biochem Biophys 465, 219-226.

17. Lopez-Sanchez I, Sanz-Garcia M & Lazo PA (2009) Plk3 interacts with and specifically phosphorylates VRK1 in Ser342, a downstream target in a pathway that induces Golgi fragmentation. Mol Cell Biol 29, 1189-1201.

18. Santos CR, Rodriguez-Pinilla M, Vega FM, Rodriguez-Peralto JL, Blanco S, Sevilla A, Valbuena A, Hernandez T, van Wijnen AJ, Li F, de Alava E, Sanchez-Cespedes M & Lazo PA (2006) VRK1 Signaling Pathway in the Context of the Proliferation Phenotype in Head and Neck Squamous Cell Carcinoma. Mol Cancer Res 4, 177-185.

19. Murphy LO & Blenis J (2006) MAPK signal specificity: the right place at the right time. Trends Biochem. Sci. 31, 268-275.

20. Turjanski AG, Vaque JP & Gutkind JS (2007) MAP kinases and the control of nuclear events. Oncogene 26, 3240-3253.

21. Morrison DK & Davis RJ (2003) Regulation of MAP kinase signaling modules by scaffold proteins in mammals. Ann. Rev. Cell Develop. Biol. 19, 91-118.

22. Yasuda J, Whitmarsh AJ, Cavanagh J, Sharma M & Davis RJ (1999) The JIP Group of Mitogen-Activated Protein Kinase Scaffold Proteins. Mol Cell Biol 19, 7245-7254.

23. Blanco S, Santos C & Lazo PA (2007) Vaccinia-Related Kinase 2 Modulates the Stress Response to Hypoxia Mediated by TAK1. Mol Cell Biol 27, 7273-7283.

24. Blanco S, Sanz-Garcia M, Santos CR & Lazo PA (2008) Modulation of Interleukin-1 Transcriptional Response by the Interaction between VRK2 and the JIP1 Scaffold Protein. PLoS ONE 3, e1660.

25. Meloche S & Pouyssegur J (2007) The ERK1/2 mitogen-activated protein kinase pathway as a master regulator of the G1- to S-phase transition. Oncogene 26, 3227-3239.


26. Raman M, Chen W & Cobb MH (2007) Differential regulation and properties of MAPKs. Oncogene 26, 3100-3112.

27. Anjum R & Blenis J (2008) The RSK family of kinases: emerging roles in cellular signalling. Nat Rev Mol Cell Biol 9, 747-758.

28. Owens DM & Keyse SM (2007) Differential regulation of MAP kinase signalling by dual-specificity protein phosphatases. Oncogene 26, 3203-3213.

29. Kang TH & Kim KT (2008) VRK3-mediated inactivation of ERK signaling in adult and embryonic rodent tissues. Biochem Biophys Acta 1783, 49-58.

30. Sevilla A, Santos CR, Barcia R, Vega FM & Lazo PA (2004) c-Jun phosphorylation by the human vaccinia-related kinase 1 (VRK1) and its cooperation with the N-terminal kinase of c-Jun (JNK). Oncogene 23, 8950-8958.

31. Sevilla A, Santos CR, Vega FM & Lazo PA (2004) Human Vaccinia-related Kinase 1 (VRK1) Activates the ATF2 Transcriptional Activity by Novel Phosphorylation on Thr-73 and Ser-62 and Cooperates with JNK. J Biol Chem 279, 27458-27465.

32. Kang TH, Park DY, Kim W & Kim KT (2008) VRK1 phosphorylates CREB and mediates CCND1 expression. J Cell Sci 121, 3035-3041.

33. Kerppola TK & Curran T (1993) Selective DNA bending by a variety of bZIP proteins. Mol. Cell. Biol. 13, 5479-5489.

34. Dunn C, Wiltshire C, MacLaren A & Gillespie DA (2002) Molecular mechanism and biological functions of c-Jun N-terminal kinase signalling via the c-Jun transcription factor. Cell. Signal. 14, 585-593.

35. Shaulian E & Karin M (2002) AP-1 as a regulator of cell life and death. Nat. Cell Biol. 4, E131-136.

36. Vega FM, Sevilla A & Lazo PA (2004) p53 Stabilization and Accumulation Induced by Human Vaccinia-Related Kinase 1. Mol Cell Biol 24, 10366-10380.

37. Schon O, Friedler A, Bycroft M, Freund S & Fersht A (2002) Molecular Mechanism of the Interaction between MDM2 and p53. J. Mol. Biol. 323, 491-501.

38. Jabbur JR, Tabor AD, Cheng X, Wang H, Uesugi M, Lozano G & Zhang W (2002) Mdm-2 binding and TAF(II)31 recruitment is regulated by hydrogen bond disruption between the p53 residues Thr18 and Asp21. Oncogene 21, 7100-7113.

39. Teufel DP, Bycroft M & Fersht AR (2009) Regulation by phosphorylation of the relative affinities of the N-terminal transactivation domains of p53 for p300 domains and Mdm2. Oncogene 28, 2112-2118.

40. Banin S, Moyal L, Shieh S, Taya Y, Anderson CW, Chessa L, Smorodinsky NI, Prives C, Reiss Y, Shiloh Y & Ziv Y (1998) Enhanced phosphorylation of p53 by ATM in response to DNA damage. Science 281, 1674-1677.

41. Buscemi G, Perego P, Carenini N, Nakanishi M, Chessa L, Chen J, Khanna K & Delia D (2004) Activation of ATM and Chk2 kinases in relation to the amount of DNA strand breaks. Oncogene 23, 7691-7700.

42. Tibbetts RS, Brumbaugh KM, Williams JM, Sarkaria JN, Cliby WA, Shieh SY, Taya Y, Prives C & Abraham RT (1999) A role for ATR in the DNA damage-induced phosphorylation of p53. Genes Develop, 13, 152-157.


43. Valbuena A, Vega FM, Blanco S & Lazo PA (2006) p53 Downregulates Its Activating Vaccinia-Related Kinase 1, Forming a New Autoregulatory Loop. Mol Cell Biol 26, 4782-4793.

44. Crighton D, Wilkinson S, O'Prey J, Syed N, Smith P, Harrison PR, Gasco M, Garrone O, Crook T & Ryan KM (2006) DRAM, a p53-induced modulator of autophagy, is critical for apoptosis. Cell 126, 121-134.

45. Valbuena A, Lopez-Sanchez I & Lazo PA (2008) Human VRK1 Is an Early Response Gene and Its Loss Causes a Block in Cell Cycle Progression. PLoS ONE 3, e1642.

46. Valbuena A, Suarez-Gauthier A, Lopez-Rios F, Lopez-Encuentra A, Blanco S, Fernandez PL, Sanchez-Cespedes M & Lazo PA (2007) Alteration of the VRK1-p53 autoregulatory loop in human lung carcinomas. Lung Cancer 58, 303-309.

47. Bradley CM, Ronning DR, Ghirlando R, Craigie R & Dyda F (2005) Structural basis for DNA bridging by barrier-to-autointegration factor. 12, 935-936.

48. Segura-Totten M & Wilson KL (2004) BAF: roles in chromatin, nuclear structure and retrovirus integration. Trends Cell Biol. 14, 261-266.

49. Nichols RJ, Wiebe MS & Traktman P (2006) The vaccinia-related kinases phosphorylate the N' terminus of BAF, regulating its interaction with DNA and its retention in the nucleus. Mol Biol Cell 17, 2451-2464.

50. Bengtsson L & Wilson KL (2006) Barrier-to-Autointegration Factor Phosphorylation on Ser-4 Regulates Emerin Binding to Lamin A In Vitro and Emerin Localization In Vivo. Mol Biol Cell 17, 1154-1163.

51. Gorjanacz M, Klerkx EP, Galy V, Santarella R, Lopez-Iglesias C, Askjaer P & Mattaj IW (2008) Caenorhabditis elegans BAF-1 and its kinase VRK-1 participate directly in post-mitotic nuclear envelope assembly. EMBO J 26, 132-143.

52. Sanz-Garcia M, Lopez-Sanchez I & Lazo PA (2008) Proteomics identification of nuclear Ran GTPase as an inhibitor of human VRK1 and VRK2 (vaccinia-related kinase) activities. Mol Cell Proteomics 7, 2199-2214.

53. Kalab P, Pralle A, Isacoff EY, Heald R & Weis K (2006) Analysis of a RanGTP-regulated gradient in mitotic somatic cells. Nature 440, 697-701.

54. Kang TH, Park DY, Choi YH, Kim KJ, Yoon HS & Kim KT (2007) Mitotic histone H3 phosphorylation by vaccinia-related kinase 1 in mammalian cells. Mol Cell Biol 27, 8533-8546.

55. Ivanovska I & Orr-Weaver TL (2006) Histone modifications and the chromatin scaffold for meiotic chromosome architecture. Cell cycle 5, 2064-2071.

56. Aihara H, Nakagawa T, Yasui K, Ohta T, Hirose S, Dhomae N, Takio K, Kaneko M, Takeshima Y, Muramatsu M & Ito T (2004) Nucleosomal histone kinase-1 phosphorylates H2A Thr 119 during mitosis in the early Drosophila embryo. Genes Develop. 18, 877-888.

57. Brittle AL, Nanba Y, Ito T & Ohkura H (2007) Concerted action of Aurora B, Polo and NHK-1 kinases in centromere-specific histone 2A phosphorylation. Exp. Cell Res. 313, 2780-2785.

58. Cullen CF, Brittle AL, Ito T & Ohkura H (2005) The conserved kinase NHK-1 is essential for mitotic progression and unifying acentrosomal meiotic spindles in Drosophila melanogaster. J. Cell Biol. 171, 593-602.


59. Acharya U, Mallabiabarrena A, Acharya JK & Malhotra V (1998) Signaling via mitogen-activated protein kinase kinase (MEK1) is required for Golgi fragmentation during mitosis. Cell 92, 183-192.

60. Colanzi A, Sutterlin C & Malhotra V (2003) RAF1-activated MEK1 is found on the Golgi apparatus in late prophase and is required for Golgi complex fragmentation in mitosis. J. Cell Biol. 161, 27-32.

61. Shaul YD & Seger R (2006) ERK1c regulates Golgi fragmentation during mitosis. J. Cell Biol. 172, 885-897.

62. Kano F, Takenaka K, Yamamoto A, Nagayama K, Nishida E & Murata M (2000) MEK and Cdc2 kinase are sequentially required for Golgi disassembly in MDCK cells by the mitotic Xenopus extracts. J. Cell Biol. 149, 357-368.

63. Xie S, Wang Q, Ruan Q, Liu T, Jhanwar-Uniyal M, Guan K & Dai W (2004) MEK1-induced Golgi dynamics during cell cycle progression is partly mediated by Polo-like kinase-3. Oncogene 23, 3822-3829.

64. Archambault V & Glover DM (2009) Polo-like kinases: conservation and divergence in their functions and regulation. Nat Rev Mol Cell Biol 10, 265-275.



8. Targeting mitotic kinases for anti-cancer treatment

Hae-ock Lee1, Yoo-Kyung Lee1, and Hyunsook Lee

Department of Biological Sciences, College of Natural Sciences, Seoul National University San 56-1, Shillim-Dong, Gwanak-Gu, Seoul 151-742, Republic of Korea

1. Background Cancer cells proliferate unlimitedly. Therefore, targeting the proliferative capacity of cancer is one of the smartest ways to treat cancer. In this vein, understanding the basis of mitosis is essential for the advanced treatment of cancer, let alone its importance in basic biology. Indeed, recent progress in anti-cancer strategy has come from the development of a group of inhibitors targeting mitosis. Opposed to meiosis, mitosis aims to preserve its genome during proliferation. In mitosis, genetic information of the parental cell is passed to two daughter cells through equal segregation of the replicated genome. Thus, the genetic integrity in proliferating cells is guaranteed through accurate chromosome separation in mitosis. Mitosis is an ordered process orchestrated by a series of mitotic kinases and associated proteins. Sets of checkpoint mechanisms ensure the accurate segregation of the replicated genomes. Prior to mitosis, G1 restriction point, Correspondence/Reprint request: Dr. Hyunsook Lee, Department of Biological Sciences, College of Natural Sciences, Seoul National University, San 56-1, Shillim-Dong, Gwanak-Gu, Seoul 151-742, Republic of Korea E-mail: [email protected] 1H-O Lee and Y-K Lee contributed equally.

Hae-ock Lee et al. 158

G1/S checkpoint, G2 checkpoint ensures that any type of DNA damage is repaired before entry into mitosis. p53 plays the pivotal role in interphase checkpoints; hence numerous anti-cancer strategies have focused to activate the p53 pathway. In mitosis, failure in the bipolar attachment of microtubule spindles to even one chromosome would result in loss or gain of chromosomes in daughter cells, resulting in massive alteration of genetic information. Therefore, spindle assembly checkpoint has evolved in eukaryotes to ensure that not a single chromosome is left behind in mitosis. Errors in this process can be detrimental; deformity in development, abortion, Down’s syndrome, neurodegenerative disorders, and cancer. Principal strategy of anticancer agents was the stabilization or destabilization of microtubule polymerization. This effectively controlled mitotic cells, the cancer cells, and not the quiescent normal cells. These include taxanes and vinca alkaloids. Recently, mitotic kinases emerged as the target for anti-cancer therapy.

Figure 1. Dynamic localization of mitotic kinases. At the prophase, Plk1 recruits Aurora A and B kinases to the centrosomes and chromosome arms respectively. BubR1 is recruited to the outer kinetochores after nuclear envelope break-down (NEBD). During prometaphase, Plk1 and Aurora B converge to the kinetochores; Aurora B to the inner centromeres and Plk1 to the outer kinetochores. At the metaphase when all the chromosomes are attached to the bipolar spindles and under proper tension, BubR1 is degraded and anaphase begins. At the anaphase, Plk1 and Aurora B are relocalized to the midzone and regulate cytokinesis. With completion of cytokinesis, Plk1 and both Aurora kinases banish.

Targeting mitotic kinases for anti-cancer treatment 159

Table 1. Inhibitors for Aurora kinases and Plk1 in clinical trials. Aurora kinase inhibitors Company Name Remarks Astex Therapeutics AT9283 Phase I-II trials, Aurora A/B inhibitor AstraZeneca AZD1152 Phase I-II trials, Aurora B/C inhibitor ZM447439 Preclinical development, Aurora B inhibitorCyclacel CYC116 Phase I trials, pan Aurora inhibitor GlaxoSmithKline GSK1070916 Phase I trials, selective Aurora B/C inhibitorMillenium MLN8054 Phase I trials, Aurora A inhibitors MLN8237 Phase I-II trials, Aurora A inhibitors Pfizer PHA-739358 Phase II trials, pan Aurora inhibitor PHA-680632 Aurora A inhibitor PF-03814735 Phase I trials, Aurora A/B inhibitor Sunesis Pharmaceuticals SNS-314 Phase I trials, pan Aurora inhibitor Vertex / Merck VX-680 (MK-0457) Phase I-II trials, Aurora A/B inhibitor Plk1 inhibitors Company Name Remarks Boehringer Ingelheim BI 2536 Phase I-II trials BI 6727 Phase I-II trials GlaxoSmithKline GSK461364 Phase I trials Nippon Shinyaku HMN-214 Phase I trials Onconova ON01910Na Phase I-II trials As illustrated in Figure 1, mitosis is orchestrated by several mitotic kinases at distinct cellular locations. Thus, malfunction in any of these key kinases can activate the spindle assembly checkpoint and ultimately end up in apoptosis in an as-yet-undefined mechanism. Among these mitotic kinases, Aurora kinases and Polo-like kinases have proven to be effective targets for cancer therapy; specific inhibitors are on clinical trials (Table 1). In this chapter, we will review on how mitosis is regulated by key kinases, such as Aurora kinases, Plk, and BubR1, and discuss on the potential of the recently developed anti-cancer drugs. The future of anti-cancer strategy based on the understanding of mitosis and its novel mechanisms will also be discussed. 2. Aurora kinases

Aurora kinases are Serine/Threonine kinases that are associated with microtubule apparatus and chromosome in mitosis. While cyclin-dependent kinse-1 (CDK1) and Polo-like kinase-1 (Plk1) trigger cell division and work as global controllers of mitosis, Aurora kinases directly regulate mitotic events largely at three critical points; at the mitotic entry building up centrosomes for the bipolar spindle formation, at the prometa- to metaphase orchestrating amphitellic attachment of spindles to kinetochores, and during cytokinesis delaying abscission until chromosome segregation is complete. In yeasts, a single Aurora kinase exists, Ipl1 [increased in ploidy] in the budding


yeast and Ark1 [Aurora related kinase] in fission yeast. In metazoans Aurora A and B were discovered; Aurora A is prominently localized at the centrosomes whereas Aurora B is localized at the inner kinetochores early in the mitosis and relocalized to the midbody during cytokinesis (Figure 1). Yeast Ipl1 is localized at the chromosome to midbody and forms a complex with INCENP-Survivin (Sli15-Bir1) similar to mammalian Aurora B. Reminiscent of their localization, Aurora A plays a crucial function in the centrosome maturation and bipolar spindle formation whereas Aurora B is involved in the chromosome condensation, microtubule-kinetochore attachment, and cytokinesis. Vertebrates have a third Aurora kinase C; the expression of Aurora C is restricted to the testis and shows similar localization and function as Aurora B. 2.1. Structure Aurora kinases have a variable N-terminal domain, a highly conserved protein kinase domain and a regulatory C-terminal domain. Aurora A and Aurora B kinases share over 60% amino acid identity and structural similarities. The substrate binding surface of Aurora A and Aurora B catalytic domain is almost identical except for a few amino acid residues. The difference is responsible for the binding of a unique partner for each kinase; Aurora A to Tpx2 and Aurora B to INCENP. A single amino acid substitution, Gly-198 of Aurora A with the Aurora B equivalent Asn residue disrupts Tpx2 association and confers INCENP binding for the mutant Aurora A [1]. The exchange of binding partner leads to the localization of Aurora A at the inner centromeres and midzone. More importantly, Aurora A G198N compensates the loss of Aurora B. These indicate that the basis for the divergent function of Aurora A and B is their interaction partners. Aurora A-Tpx2 complex and Aurora B-INCENP complex have distinct structure and different substrate specificities [2,3] Autophosphorylation of Thr-288 in Aurora A stimulated by Tpx2 and Thr-232 in Aurora B by INCENP are required for the full activation of kinase activities as well. The variable N-terminus mediates many protein-protein interactions for Aurora A and Aurora B kinases. C-terminal D box and N-terminal A box are responsible for APC/Cdh1 dependent degradation during mitotic exit [4] for both kinases. 2.2. Function of Aurora A Aurora A is prominently localized at the centrosomes and along the microtubules during mitosis. At the centrosome, Aurora A is involved in the centrosome separation and centrosome maturation. Aurora A also regulates


centrosome-dependent and independent spindle assembly. In the absence of Aurora A, mitotic entry is significantly delayed, implicating Aurora A in cell cycle progression as well. Also important is that Aurora A locus is frequently amplified and the expression elevated in cancer cells including breast and colon. We will discuss the role of Aurora A in the regulation of centrosome and bipolar spindle formation, mitotic entry, and its oncogenic potential. 2.2.1. Centrosomal function The first discovery of Aurora A kinase was made in Drosophila, in search for mutations that affect centrosome cycle [5]. In the aur mutant embryos, two centrosomes were paired and acted as a single pole to nucleate monopolar spindles. Those embryos had circular chromosome arrangements around a large centrosomal structure, resembling aurora in the polar region. Centrosomes and the pericentriolar materials (PCM) function as a main microtubule nucleation center in animal cell division. A single centrosome comprises of two centrioles perpendicularly arranged. In G1-S phase, centrosomes duplicate in concert with DNA replication, giving rise to two centriole pairs surrounded by PCM. In late G2-prophase, the two paired centrosomes undergo maturation, expanding PCM to nucleate sufficient mitotic microtubules. Centrosome separation occurs before or after nuclear envelope break down (NEBD) depending on the cell types. After NEBD, the centrosomes nucleate microtubule asters to form bipolar spindles around metaphase chromosomes. In all organisms examined, Aurora A inhibition resulted in defects in centrosome maturation, failing to build up enough PCM components such as γ-tubulin. In C. elegans, Aurora A depleted air-1(RNAi) embryos and somatic cells failed to accumulate γ-tubulin and additional PCM components at the centrosome [6]. In flies, aur mutant sensory organ precursor cells failed to recruit γ-tubulin, Centrosomin [7], and Minispindles (XMAP215 homolog) . In HeLa cells, RNAi for aurora A impaired gamma-tubulin recruitment and induced split centrioles [8]. The failure to recruit enough PCM components may perturb in centrosome separation and bipolar spindle formation. Apart from the role in centrosome maturation, Aurora A plays a direct function in mitotic spindle formation. In animal cells, centrosome-dependent and chromosome-dependent spindle assembly pathways exist and Aurora A is an important player in both. In centrosome-dependent spindle assembly, a link between Aurora A and the spindle formation is found in TACC, a highly transforming acidic-coiled-coil-containing protein. TACC is a substrate of Aurora A whose centrosomal targeting relies on Aurora A. TACC forms a complex with TOG/XMAP215 protein which directly binds to microtubules.


Aurora A mediated recruitment of TACC-TOG complex to centrosomes stabilizes microtubule minus ends by counteracting microtubule-destabilizing kinesins. Involving Aurora A’s function in centrosome maturation and bipolar spindle formation several upstream and downstream factors were identified in various model systems. Although the contribution and exact function of those factors are not fully defined, Barr and Gergely [9] put together those factors in a model where, the kinase CDK11 is responsible for the localization of Plk1, which in turn recruits Aurora A to the centrosome. Centrosomal Aurora A phosphorylates factors like PAK1, Tpx2 and Ajuba, all of which facilitate Aurora A autophosphorylation. The active Aurora A kinase recruits downstream factors like TACC and Ndel1. Ndel1 is known to restore many defects caused by Aurora A depletion. In the chromosome-dependent spindle assembly, microtubules assemble around chromosomes which can be readily shown in oocytes of many animal species. This pathway is known to require the activity of Ran-GTP. In fact, microtubule nucleation takes place even in the absence of centrosome or chromatin when the activity of Ran-GTP is present in xenopus egg extracts. The high concentration of Ran-GTP around the chromosomes promotes release of spindle assembly proteins from Importins at the nuclear envelope. The released factors form a complex called EXTAH, which contains Eg5, XMAP215, Tpx2, Aurora A, and HURP as well as γ-TURC. Aurora A coated beads in the xenopus egg extracts could form EXTAH complex and promote microtubule assembly in the absence of centrosome or chromatin. Aurora A recruits and phosphorylates Tpx2, which leads to the autophosphorylation of Aurora A. The active Aurora kinase recruits other components of EXTAH complex where, γ-TURC nucleates; XMAP215 stabilizes; HURP bundles; and Eg5 and Tpx2 crosslink microtubules. Thus Aurora A is at the center of mitotic microtubule organization. 2.2.2. Mitotic entry Aurora A depletion causes delay in mitotic entry. Mitotic entry and exit are controlled by the key mitotic kinase Cdk1, and Aurora A is known to regulate it in two ways. First, the earliest activation of Cdk1-cyclin B occurs at the centrosome and Aurora A is essential for the centrosomal recruitment and activation. In human cell lines, Aurora A depletion impaired the recruitment and initial activation of Cdk1-cyclin B at the centrosome. As Cdk1 inhibition impairs Aurora A activation, there seems to be a positive feedback loop between Aurora A and Cdk1. When Aurora A is depleted, delay in mitotic entry was also observed in C. elegans. In the first mitotic division of C. elegans embryos, male pronuclei-associated centrosomes are


crucial for the timely mitotic entry and Aurora A is the major player in this process. Secondly, Aurora A regulates mitotic entry via regulating CPEB (Cytoplasmic Polyadenylation Element Binding Protein). The polyadenylation-induced translation of cyclin B was initially shown to be critical to the meiotic progression in xenopus oocytes. Similar translational control was found in xenopus embryos for mitotic progression where phosphorylation of CPEB by Aurora A regulates polyadenylation and cell cycle [10]. Embryonic CPEB is localized at the centrosome with other factors involved in the polyadenylation, indicating that CPEB phosphorylation by Aurora A is associated with centrosome. As oocytes do not have centrosome and centrosome disruption does not cause as significant mitotic delay as Aurora A depletion, the role of Aurora A in mitotic entry is exerted in both centrosome dependent and independent ways. 2.2.3. Aurora A and cancer In 1997, human Aurora A was identified as a ‘breast tumor activated kinase (BTAK)’ mapped to the locus 20q11-13 that is frequently amplified in breast cancer [11]. The gene amplification as well as over expression of Aurora A is found in many other epithelial cancers [12]. The over-expression of Aurora A kinase transformed Rat1a cells and NIH3T3 cells to form colonies in vitro and tumor mass in nude mice [13,14]. These studies assigned Aurora A as an oncogene. However, Aurora A failed to transform primary mouse embryonic fibroblasts [15] and failed to induce tumorigenesis when expressed as a transgene in mice [16]. These results indicate that Aurora A is not a bona fide oncogene, and other genetic alterations are necessary for the transformation. Nevertheless, Aurora A surely contributes to the tumorigenesis. First, over-expression of Aurora A kinase induces polyploidy and abnormal centrosome number. This is likely to cause genetic instability and may contribute to tumorigenesis, although not sufficient for transformation. Secondly, Aurora A phosphorylates p53 and inhibits its transactivation and also induces its degradation. The inactivation of p53 would abrogate the G1 checkpoint, and abnormal cells may enter cell cycle. Thirdly, reports suggest that Aurora A might regulate Ras signal and influence cell growth and enhance Ras-mediated transformation [17,18]. The association between Aurora A and cancer prompted the development of Aurora A targeted small molecules, which showed promising drug response in cancer cells and animal models. When aurora A is depleted or inhibited, abnormal spindle formation takes place. The cells -both normal and transformed- undergo abnormal cell division after delay in mitotic entry and


anaphase onset. The resulting cells end up with polyploidy. The long mitotic delay and the abnormal chromosome trigger apoptosis in most cells. Recently, it was found that the abnormal chromosome segregation leads to cell death depending on the time spent in mitosis trying to fix the problems. If cells have normal checkpoint function, apoptotic pathway is the preferred pathway after Aurora A inhibition because the absence of aurora A does not override checkpoints. This is why the aurora A specific inhibitor is a promising but potentially dangerous cancer therapeutic target. Primary mechanism how Aurora A inhibition induces cell death is through generation of abnormal cell division. If cells escape the death pathway, Aurora A inhibition could generate more genetically unstable, potentially cancer-prone cells in the treatment. Development of anti-cancer drugs targeting Aurora A specifically will be discussed later. 2.3. Function of Aurora B Aurora B forms the chromosomal passenger complex (CPC) and regulates chromosome cohesion, chromosome-microtubule attachment, and cytokinesis. To coordinate these processes, the CPC moves dynamically from chromosomes to midbody during mitosis. 2.3.1. Chromosome passenger complex Aurora B is the enzymatic subunit of CPC; whose nonenzymatic subunits INCENP, Survivin, and Borealin are conserved from yeast to mammals. The nonenzymatic components are phosphorylated by Aurora B and regulate the localization and kinase activity of Aurora B. INCENP binds Aurora B through the C terminal IN box, which is required for the full activation of the Aurora B kinase. INCENP also functions as a scaffold to accommodate Survivin and Borealin at the N-terminus. Survivin is primarily responsible for the CPC targeting to centromeres and central spindles. Borealin is also required for targeting of CPC by bringing INCENP and Survivin together. In fact, Survivin-INCENP fusion proteins can target functional CPC to centromeres and central spindle in the absence of Borealin. Depletion of any CPC component disturbs proper localization and function of the complex and manifests similar mitotic defects, indicating that the CPC functions as a single structural unit. 2.3.2. Chromosome condensation Aurora B is localized at chromosome arms during prophase and concentrated on the centromere at prometaphase. The chromosome


association continues to the onset of anaphase, then Aurora B moves to the central spindle. While associated with chromosome, Aurora B is known to regulate the mitotic chromosome structure. First, Aurora B phosphorylates Histone H3 at Ser10, which is a hallmark of mitosis and linked to the chromosome condensation. In tetrahymena, nonphosphorylated S10A H3 causes elongation of mitotic micronuclei. Histone H3 phosphorylation is thought to cause dissociation of HP1 from heterochromatin, which allows chromosome to access to the condensin complexes for chromosome condensation and cohesion during mitosis. Second, Aurora B phosphorylates condensin I subunits to allow its chromosome association. Condensin I is required for the chromosome compaction and also for the displacement of cohesins from chromosome arms for chromatid separation. Third, the maximal chromosome compaction is achieved during anaphase and Aurora B inhibition causes decompaction in an unknown mechanism. Therefore Aurora B regulates the chromosome structure throughout mitosis. 2.3.3. Microtubule-kinetochore attachment

Accurate chromosome segregation is the climax of mitosis. The segregation is mediated by bi-oriented microtubule spindles attaching to the sister kinetochores and pulling them to the opposite ends. This amphitellic microtubule-kinetochore (MT-KT) attachment is critical to ensure accurate chromosome segregation. Molecular machineries in monitoring the false attachments have been evolved to sense unattached kinetochores or the lack of tension. How these physical states are biochemically translated has been a major issue in cell biology. Now Aurora B is thought to be the key translator, which destabilizes the MT-KT attachment by phosphorylating kinetochore substrates in the absence of proper tension. Microtubule attachment is a dynamic process that many weak interactions, both enforcing and destabilizing, determine stable attachment. A protein called Ndc80 destabilizes the MT-KT interaction, which requires phosphorylation by Aurora B. Thus phosphorylation by Aurora B could re-enforce MT-KT dynamics to destabilization and generate unattached kinetochores, which will resume microtubule attachment until proper tension is generated. The phosphorylation by Aurora B could be limited by regulation of kinase activity and/or spatial separation between Aurora B at the inner centromeres and its substrates at the outer kinetochores. Liu had shown that the spatial segregation indeed regulated Aurora B mediated phosphorylation [19], and proper tension led to dephosphorylation of kinetochore substrates. Alternatively, and not exclusively, regulation of kinase activity per se plays a role. Aurora B kinase activity is known to require microtubules and


microtubule-associated factor TD-60 [20]. Microtubule attached to the inner centromere in the absence of tension activates Aurora B kinase activity. Nucleosomal structure also affects Aurora B kinase activity by allowing close association of CPCs and activation of Aurora B kinase activity by other CPC components in trans [21]. Contribution from the different pathway is not yet clear, but clearly Aurora B is required for the bipolar MT-KT attachment. Aurora B, the crucial kinase for amphitellic MT-KT attachment, also plays a role for spindle assembly checkpoint (SAC) that prevents the activation of APC/Cdc20 until all the chromosomes are properly aligned at the metaphase plate and bipolar spindle attachment is established. After the inhibition of Aurora B with small molecule inhibitors, ZM447439 or Hesperadin, abnormal cell division takes place indicative of SAC inactivation and premature anaphase onset [22,23]. Consistent with these findings, microtubule stabilizing agent Taxol or Eg5 inhibitor Monastrol do not arrest cells when Aurora B is inhibited or other CPC component is depleted [24,25]. Interestingly, however, in the presence of ZM447439 or Hesperadin, cells arrest after microtubule destruction with Nocodazole, suggesting that SAC activation could take place regardless of Aurora B. In this case, concomitant depletion of Bub1 overrides SAC and cells do not arrest with Nocodazole [26]. Thus Morrow and Taylor suggested that SAC activation occur via Bub1 monitoring unattached kinetochores and Aurora B monitoring bipolar attachment. Although how SAC activation is achieved by Aurora B is not fully defined, involvement of Ndc80 [27] has been suggested. 2.3.4. Cytokinesis

When Aurora B function is disrupted, multinucleated cells are formed, suggesting for cytokinesis failure [28,29]. Cytokinesis is the last stage of cell division where actin, myosin, and other cytoskeletal components are redistributed to the contractile ring to create cleavage furrow and induce division of plasma membrane in animal cells. At the anaphase onset, Aurora B is relocated from chromosomes to central spindles and regulates cleavage furrow formation. The signals for cleavage furrow formation come from both microtubule asters and central spindles at the midzone [30,31] and Aurora B is implicated in both [31]. The stabilization of midzone requires microtubule bundling activity of microtubule associated proteins (MAPs) and centralspindlin complex composed of Mklp1 and RacGAP. The localization of centralspindlin is dependent on Aurora B and either the depletion of Aurora B or Mklp1 show similar cytokinesis defects ranging from the complete lack of furrow formation [32] to asymmetric furrow formation depending on cell types. As RacGAP/CYK-4 is a putative GTPase activating


protein for RhoA GTPase that directs contractile ring assembly, Aurora B might indirectly affect the contractile ring formation as well. While Aurora B plays a role in the initiation of furrow formation, Aurora B also delays completion of cytokinesis until all the chromosome segregation is complete [33-35]. This “NoCut” checkpoint was dissected in the budding yeast that midspindle defects or chromosome bridges delay abscission, which is dependent on Ipl1. Similar “Abscission checkpoint” was also identified in mammalian cells, as chromosome bridges located at the cleavage furrow delayed abscission in Aurora B dependent manner. The kinase activity of Aurora B is required for the delay, and Aurora B-mediated phosphorylation of Mklp1 is implicated. Thus Aurora B regulates the initiation and completion of cytokinesis. 2.4. Inhibitors of Aurora kinases in clinical trials for anti-cancer treatment The Aurora kinase inhibitors such as ZM447439, MK-0457 (VX-680), and PHA-739358 are pan-Aurora inhibitors, which inhibit Aurora A, B, and sometimes C activities in vitro [22]. Although ZM447439 is still on the preclinical development, MK-0457 (VX-680) and PHA-739358 are actively being evaluated in the clinical fields. MK-0457 (VX-680) is a pyrimidine derivative with affinity for aurora A, B, and C at a value of nanomolar concentrations. It inhibits the growth of tumor xenograft, and is effective in chronic myeloid leukemia cells and other solid cancer cells.[36,37]. In the phase I clinical trial, MK-0457 was observed to stabilize the refractory solid tumors in 3 patients [38]. Three patients with T315I phenotype-refractory CML or Philadelphia-positive ALL have shown clinical responses to doses of MK-04547 that are not related with adverse events [39]. Phase II clinical trials are currently underway in the patients with advanced NSCLC (NCT00290550, www.clinicaltrial.gov), and in the patients with CML and Philadelphia-positive ALL (NCT00405054). PHA-739358 is a pan-Aurora kinase inhibitor with anti-cancer activity on variable tumor xenograft models [40,41]. The results of phase I trial were presented in 2006, and PHA-739358 is currently being investigated in a phase II clinical trial in CML patients relapsed after imatinib mesylate or c-ABL therapy, including patients with T315I mutation (NCT00335868). Other phase II trials are currently being investigated in adult patients with Multiple Myeloma who have a history of at least two previous lines of treatment for the disease (NCT00872300), and in patients with metastatic hormone refractory prostate cancer (NCT00766324). AT9283 is another multi-targeted kinase inhibitor recently developed with potent activity against Aurora A and B kinases in the clinical


development [42,43]. Side effects and maximum tolerated dose of AT9283 are being examined in phase I trials when treating patients with advanced or metastatic solid tumors or non-Hodgkin's lymphoma (NCT00443976), ALL, AML, CML, high-risk myelodysplastic syndromes, or myelofibrosis with myeloid metaplasia (NCT00522990). MLN8054 is the first orally administered Aurora kinase inhibitor, and the first one to inhibit Aurora A specifically, and much less so to Aurora B. Treatment of MLN8054 leads to spindle defects, resulting in the activation of spindle assembly checkpoint and inhibition of proliferation in various human cancer cells [44]. Growth of human tumor xenografts in nude mice was inhibited after oral administration at well-tolerated doses, and the tumor growth inhibition was sustained even after the discontinuation of the treatment. In xenografts, MLN8054 induced mitotic arrest and apoptosis. The preliminary results of a phase I study showed that MLN8054 was absorbed rapidly with a reasonably long half-life. It inhibits Aurora B at higher doses. However, dose-limiting toxicities limited dose escalation before mechanism-based toxicity was seen [45]. MLN8237 is a second-generation, selective Aurora A kinase inhibitor designed for greater potency and fewer benzodiazepine-like effects observed in first-generation agent, MLN8054. In the phase I trial, MLN8237 was tolerable with adequate doses and exhibited favorable clinical antitumor activity when treated for the advanced solid tumors [46]. The results supported the continuation for phase II development, and MLN8237 is currently in phase II clinical trials. MLN8237 is administered for the treatment of patients with platinum-refractory or platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal carcinomas (NCT00853307). MLN8237 is also effective on hematological malignancies, as well as on solid tumors. The phase I study in patients with advanced hematological malignancies who have limited standard treatment options is active (NCT00697346). In addition, phase I/II trial is currently recruiting subjects for treating pediatric patients with relapsed, refractory solid tumors or acute lymphoblastic leukemia (NCT00739427). Phase II studies for patients with relapsed or refractory non-Hodgkin’s lymphoma (NCT00807495), AML or myelodysplastic syndrome (MDS) (NCT00830518) are underway. AZD1152 is a dihydrogen phosphate pro-drug of a pyrazoloquinazoline Aurora kinase inhibitor. It showed a potent and selective inhibition of Aurora B in tumor xenografts [47,48]. AZD1152 was reported to be effective in inhibiting growth of acute leukemia cells when used in combination with vincristine and daunorubicin both in vitro and in vivo [49]. Phase I trial on patients with colon cancer, melanoma, or various other solid tumors revealed that AZD1152 made significant disease stabilization with tolerable toxicity


[50]. Phase I studies to assess the effect of AZD1152 on the rate of complete remission in patients with relapsed AML (NCT00530699) and in patients with advanced solid tumors (NCT00338182, NCT00497679, NCT00497731) are currently on going. Various Aurora kinase inhibitors, apart from the inhibitors described above, are currently being evaluated in clinical setting. Phase I trial is studying escalated dose and the side effects of CYC116, a pan-Aurora kinase inhibitor, in treating patients with advanced solid tumors (NCT00530465, NCT00560716). PF-03814735, a pan-Aurora kinase inhibitor, which is administered orally as single agent in patients with advanced solid tumors with preliminary results (NCT00424632) [51] and SNS-314, a selective Aurora A kinase inhibitor, in advanced solid tumors with preliminary results (NCT00519662) [52] are on phase I clinical trials. Moreover, there are many more Aurora inhibitors in preclinical development: CHR-3520, CTK-110, ENMD-981693, JNJ-7706621, PHA-680632, MP-529, MP-235, GSK1070916 and so on [53,54]. 3. Polo-like kinase 3.1. Entry and exit from mitosis with Plk1 The master regulator of mitosis is Cdk1, bound and activated by cyclin B. Mitotic entry and exit is controlled by Cdk1. Following Cdk1, Polo-like kinases (Plks) play critical roles in the progression of mitosis; mitotic entry, centrosome maturation, spindle assembly, chromosome alignment, APC/C regulation, and cytokinesis. Four Plk family members, Plk1-4 are found in vertebrates. Of the four Plks, Plk1 is believed to carry out most of the functions attributed to Cdc5, Plo1, and Polo of budding yeast, fission yeast, and Drosophila, respectively. Plk1 exerts its multifaceted functions in mitosis through localizing to all of the important places like chromosomes, centrosomes, central spindle at the right time. As Erich Nigg describes, Cdk1 is the conductor in mitosis and Plk1 is the first violin [55]. Plks have serine/threonine kinase domain at the N-terminus, D-box degradation signals for APC/C in the middle, and the unique PBD (Polo box domain) in the C-terminus. Two polo box domains, PB1 and PB2, are thought to confer substrate selectivity and regulate subcellular localization of the kinase. The PBD domain serves as an inhibitor of the kinase when substrate binding is absent. When the docking proteins are phosphorylated by priming kinases, often Cdk1, they bind to the PBD domain in Plk1, relieving the PBD’s autoinhibition of the kinase, and being further phosphorylated by Plk1. This way, substrate specificity is elegantly coordinated in a cell division


cycle-dependent manner, following CDK1-cyclin B [55,56]. It is also shown that Plk1 can activate Cdk1 at G2/M transition indirectly [55], and also can phosphorylate cyclin B and promote accumulation of this mitotic cyclin B in the nucleus [57,58]. Conceivably, this synergistic and sequential interplay between Cdk1 and Plk1 is thought to regulate timely mitotic entry. Without Plk1, mitosis cannot be continued; depletion of Plk1 from Xenopus oocyte extracts impaired mitotic entry, whereas depletion of Plk1 resulted in delay in anaphase progression in human cells, flies [59], and in zebrafish embryogenesis (our unpublished data). The mitotic arrest by Plk1 abrogation is thought to result from defective spindle assembly, impaired centrosome maturation, and impaired MT-KT attachments. All of these defects provoke the spindle assembly checkpoint (SAC) activation, since SAC ensures the bipolar spindle attachments and tension at kinetochores. The outcome is the delay in mitosis followed by frequent apoptosis, through the mechanism not yet understood satisfactorily. These results indicate that Plk1 is essential for progression in mitosis, and the primary role of Plk1 may be in microtubule organization, microtubule spindle formation, and MT-KT attachments. It was believed that Plk1 on its own is not involved in SAC, evidenced by the fact that abrogating Plk1 resulted in the activation of SAC. Nonetheless, Plk1 can regulate SAC through phosphorylating BubR1 in a tension-sensitive manner [60]. With these notions, the interaction and interplay between SAC, Plk1, microtubule spindles are not as simple as it first seemed. A good reason why revealing the mechanism orchestrating mitosis is crucial, even for the sake of developing anti-cancer drugs targeting mitosis. Multifaceted roles of Plk1 in mitotic entry and progression have hampered the finding of its role in cytokinesis; when Plk1 function is compromised, the cells first arrest in mitosis. With the development of chemicals inhibiting Plk1 function, it became clear that Plk1 localizes to central spindle in telophase and even in the place where abscission takes place and plays essential roles in cytokinesis [59,61,62]. The advantage of the chemical biology is that treating the cells with the drugs can be controlled in a timely- controlled manner. Cells were treated with specific inhibitors to Plks after they committed chromosome segregation. With chemical biology, unveiling the function of Plk1 in the last step of cell division, cytokinesis, has just begun. 3.2. Is Plk1 an oncogene? Like Aurora A, Plk1 over expression is observed in many tumors of diverse origin, including breast, ovary, colon, stomach, pancreas, lung, head and neck, skin, esophagus, and brain [63-66]. Its over expression correlates


with poor outcome [64,67,68]. In immortalized rodent cells, over expression of Plk1 induced transformation, growth in soft agar assays, and formed tumor in nude mice, suggesting for an oncogenic potential [69]. However, depletion of Plk1 with siRNA or small molecular inhibitors caused genomic instability accompanied by sudden death as well [70-74], suggesting that both depletion and over expression of Plk1 can attack genome integrity. Live-imaging in zebrafish embryogenesis revealed that both depletion of Plk1 and over expression of it cause spindle defects and centrosome abnormality in mitosis, resulting in arrest in mitosis (our unpublished results). These results altogether point to the hypothesis that the adequate level of Plk1 is essential for progression in mitosis and genome integrity. In this regard, Plk1 is not an oncogene. Rather, it behaves as a balance for regulated mitosis and abnormal tumorigenesis. Taken together, it is conceivable to think that targeting Plk1 with chemically designed inhibitors can be cancer cell-selective. Indeed, varieties of small molecular compounds have been developed so far. 3.3. Plk1 inhibitors for anti-cancer treatment First generation Plk1 inhibitors such as LY294002 and other related compounds are in clinical trials. Most recently, BI 2536 has been developed, which showed increased sensitivity towards Plks. In vitro, BI 2536 inhibits Plk1 and induces mitotic arrest with low concentration; IC50 value below 1 nM, and inhibits the growth of cancer cell lines in the IC50 value ranged between 2 to 30 nM [61,62]. BI 2536 binds to the catalytic domain of Plk1 with high potency. Results of phase I trials have reported that BI 2536 was well tolerated and showed a favorable pharmacokinetic profile in advanced solid tumors [75]. Phase II trial has been conducted to evaluate the efficacy, safety and pharmacokinetics of BI 2536 in the treatment of unresectable advanced pancreatic cancer as first line or second line therapy (NCT00710710), in the treatment of advanced or metastatic NSCLC of stage IIIB or IV in patients who relapsed after or failed first-line therapy (NCT00376623), in the treatment of prostate cancer (NCT00706498), in second line treatment in sensitive-relapse SCLC patients (NCT00412880). ON01910Na is a small-molecule Plk1 inhibitor that is not an ATP-mimetic but competes for the substrate-binding site of Plk1 [76]. Recent phase I study with an accelerated titration dose-escalation design of ON01910Na showed objective response with moderate toxicities, especially in refractory ovarian cancer patients [77]. Another phase I trial was opened for evaluation of ON01910Na when given together with gemcitabine for treating patients with advanced or metastatic solid tumors (NCT00822939).


Plk1 inhibitors in preclinical or clinical development include GSK461364, BI 6727, and HMN-214. GSK461364 is a benzimidazolyl-thiophene which inhibits Plk1 in an ATP-competitive way [78]. GSK461364 leads to G2 phase arrest at high concentrations and M phase arrest at low concentrations. BI 6727, the second generation dihydropteridinone derivative, shows highly potent and selective anti-tumor activity in cancer models including taxane-resistant colorectal cancer [79]. Both GSK461364 and BI 6727 are in early clinical trials. Phase I clinical trial is being conducted to determine the maximum dose and adverse effects of GSK461364 in adult patients with solid tumors and Non-Hodgkins lymphoma (NCT00536835). Preliminary results for escalating dose in the subjects with solid tumors are reported [80]. The preliminary results of phase I trial with BI 6727 also showed anti-cancer effects with tolerable dose [81]. Phase II study is underway to evaluate whether BI 6727 monotherapy or in combination with pemetrexed is effective in the treatment of advanced or metastatic NSCLC in patients with recurrent or refractory first-line platinum based therapy (NCT00824408). Phase I/IIa is also on the way to investigate for monotherapy and in combination with low dose cytarabine in patients with relapsed or refractory AML that are not eligible for intensive treatment. In the phase IIa part, the combination of BI 6727 at MTD with cytarabine and cytarabine monotherapy is under investigation to explore the efficacy of the combination schedule in previously untreated AML patients who are not eligible for intensive treatment (NCT00804856). HMN-214, an oral stilbene derivative, affects indirect disruption of Plk1 and leads to G2/M phase arrest causing anti-tumor activity in vitro and in vivo mouse model [82]. The phase I trial demonstrated the tolerable dose for the patients with advanced solid tumors [83]. Besides these Plk1 inhibitors in clinical development, Plk1 inhibitors, such as ZK-thiazolidinone and DAP-81, are under the preclinical development [84,85]. 4. BubR1 kinase 4.1. The key spindle assembly checkpoint component, BubR1 The climax of mitosis is at the metaphase to anaphase transition where the chromosome segregates in a pole-ward direction. This event is critical in ensuring the genetic instability. At every kinetochores of chromosomes, amphitellic attachments of microtubule spindles are checked by the spindle assembly checkpoint (SAC). SAC inhibits APC/C E3 ligase, the multisubunit E3 ligase that is responsible for the destruction of Securin and cyclin B.


Securin is the inhibitor for Separase that cleaves Cohesins. Cyclin B is the mitotic cyclin that binds to Cdk1 and conducts mitosis. Therefore, chromosome segregation and anaphase onset, are governed by APC/C, and SAC inhibits the APC/C until the chromosomes are ready to segregate [86]. Until now, the modifications and control on the 13 subunits of APC/C, the key feature in understanding the regulation of APC/C, have not been fully elucidated yet. However, the fact that APC/C activation requires WD40 domain-containing coactivators Cdc20 in mitosis or Cdh1 in mitotic exit is firmly established [87]. Therefore, the inhibition or the control of APC/C activity in mitosis can be controlled in this layer; titrating out the coactivator Cdc20 away from APC/C. Indeed, Mad2 (first identified from the mad2 mitotic arrest deficient mutant in yeast) binds to Cdc20 in vitro and in vivo, sequestrating it from APC/C thereby inhibiting Securin and cyclin B destruction [87]. Identified later, but not less important inhibitor of APC/C is the kinase BubR1. Unlike Mad2, BubR1 is a kinase not found in yeast. The N-terminus of BubR1 resembles Mad3 of yeast, but it has the kinase domain like Bub1 at the C-terminus. Therefore, it would be fair to say that the vertebrates have evolved another kinase BubR1 to modulate the crucial APC/C activity in mitosis. In vitro experiments suggest that BubR1 is a more potent inhibitor for APC/C-Cdc20 than Mad2 [88]. At first, it seemed that the way BubR1 inhibits APC/C is similar to Mad2 by sequestrating Cdc20 away from APC/C. However, accumulating evidences indicate otherwise. Studies revealed that the way SAC functions is through forming a complex of composed of Mad2, Bub3, BubR1 and Cdc20, named MCC (Mitotic Checkpoint Complex), to inhibit APC/C [89]. This way, the coactivator Cdc20 is sequestered away from APC/C and the anaphase is delayed (Figure 2a). However, this model harbors some problems. BubR1 is capable of binding to both Cdc20 and APC/C. Furthermore, BubR1 was bound to APC/C before and after anaphase onset [90]. More detailed biochemical studies showed that Mad2 only loads Cdc20 to BubR1 to kinetohcores then leaves the complex. Cdc20 is ubiquitinated and degraded until SAC is satisfied. From this model, the MCC model in SAC activation has been refined; BubR1 is bound to APC/C at kinetochores, Mad2 loads Cdc20 to BubR1 and APC/C then leaves the complex. Until bipolar spindle attachment is achieved, Cdc20 is degraded. When the kinetochores are all attached with bipolar spindles, BubR1-APC/C-Cdc20 complex changes their conformation and APC/C gets activated with Cdc20 now bound [91]. In this model, as well as the MCC model, BubR1 is the key player in the inhibition of APC/C until SAC is satisfied (Figure 2b).


Figure 2. Proposed models of spindle assembly checkpoint. (a) In the MCC model, BubR1, Mad2, Bub3 and Cdc20 form a complex and this complex sequesters Cdc20 coactivator from APC/C, therefore APC/C E3 ligase activity is inhibited. (b) Alternatively, recent biochemical and cell biological studies suggested that Mad2 only loads Cdc20 onto BubR1, which is complexed with APC/C, and leaves the complex. Cdc20 is ubiquitinated and degraded when SAC is active and in SAC off condition, Cdc20 ubiquitination is prohibited and it serves as a coactivator for APC/C. Activated APC/C E3 ligase activity results in the anaphase onset. In addition, BubR1 acetylation/deacetylation provides another layer of modulating APC/C activity. Then how is bipolar spindle attachments and tension sensed to SAC? The key molecule to this question seems to be BubR1. BubR1 kinase activity is dispensable for inhibiting APC/C E3 ligase activity in vitro [88]. However, a microtubule motor protein CENP-E is a binding partner of BubR1 and


functions like cyclins for Cdks. This way, microtubule spindle binding to kinetochores controls the activity of BubR1 kinase [92-96]. In this delicate signaling of bipolar spindle attachments to kinetochores, activation of BubR1, and inhibition of APC/C, and Plk1 [60,97] play essential roles by phosphrylating BubR1 kinase. It is noteworthy that BubR1 in prometaphase or upon spindle disruption shows a characteristic slower migrating form of phosphorylation in SDS-PAGE. 4.2. Signals that activate BubR1 All of APC/C substrates have short stretches of destruction motifs called D box or KEN boxes that are recognized by APC/C and coactivators [90]. APC/C inhibitor BubR1 also has D box and two KEN boxes. However, how BubR1 makes use of these motifs to bind to APC/C and Cdc20 and not being degraded by APC/C has not been understood. Recently, it was found that BubR1 is acetylated by PCAF in prometaphase when it is active. Acetylated BubR1 is prohibited from being a substrate of APC/C. When the SAC is satisfied, BubR1 is deacetylated and becomes a substrate of APC/C. Degradation of BubR1 further activates APC/C, initiating the anaphase onset, thus BubR1 acetylation/deacetylation is a molecular switch for BubR1 from being a inhibitor to a substrate of APC/C. [98] (Figure 2b). This study revealed that in addition to BubR1 phosphorylation, acetylation provides another layer of signaling in SAC control. 4.3. Potentials of BubR1 as a molecular target for cancer therapy Notably, BubR1 function is crucial for cells in mitosis. In interphase, BubR1 is mainly at cytoplasm. As cells enter mitosis, Bub3 brings BubR1 to the kinetochores [99]. Localization of BubR1 at outer kinetochores is crucial for SAC function for it bound with CENP-E monitors the bipolar spindle attachment at kinetochores. Only in proliferating cells, the localization and signaling of BubR1 becomes crucial. Conventional anti-cancer therapy has been the DNA damaging drugs or alkylating drugs to kill proliferating cancer cells. It had been noticed that upon the treatment of DNA damaging drugs, cancer cells suddenly die in mitosis. This so called mitotic catastrophe is specific to cancer cells because normal cells with intact G2 checkpoint would arrest the cells in G2 after genotoxic insult. Thus, cancer cells with abrogated interphase checkpoints, e.g. by mutation of p53, would be the ones targeted for mitotic catastrophe after treatment with DNA damaging drugs. Then what is the key molecule receiving and sending the death signals after DNA damage in mitosis?


Studies suggest that BubR1 is phosphorylated and activated in response to adriamycin [100]. Indeed, BubR1 has been suggested to induce mitotic catastrophe in HeLa cells [101]. With these notions, it is conceivable that BubR1 can be an efficient target for cancer therapy. That acetylation of BubR1 regulates SAC activity and modulation of APC/C renders another possibility that specific inhibition of BubR1 deacetylation can be adopted for targeting cancer, since HDAC inhibitors are actively being investigated for anti-cancer therapy in clinical trials [102]. So far, our understanding of BubR1 kinase substrates are poor, and targeting the kinase activity of this important protein in mitosis is yet to be explored. 5. What’s next? We are now living in the interesting times to face the most rapid development of bench to bedside research on mitotic kinase inhibitors for anti-cancer strategy. Hundreds of preclinical results indicate that mitotic kinases are effective targets for the treatment of variety of tumors. However, as we have seen from the cases of Aurora A and Plk1 where both depletion and over expression of them can be dangerous for genomic integrity, inhibition of these kinases do not straightforwardly kill cancer. In fact, we learn more on the mechanisms of how these kinases work in cell division using the specific inhibitors. Thus, clinical trials to validate the efficacy and safety of the mitotic kinase inhibitors should be accompanied with the basic side of science. Right now, we are still waiting for the further phase III clinical trials. Acknowledgements We thank the members of the HL lab (Lab of Tumor suppressors and Cell Division, TSCD) for helpful discussions. We are grateful to Dr. Stephen Taylor (Univ. of Manchester, UK) for informal comments on MLN 8054. Work in HL laboratory are supported by National Research Laboratory Program (NRL, ROA-2008-000-20023-0) and SRC program (R11-2005-009-04003-0). References 1. Fu J, Bian M, Liu J, Jiang Q, Zhang C: A single amino acid change converts

Aurora-A into Aurora-B-like kinase in terms of partner specificity and cellular function. Proc Natl Acad Sci U S A 2009, 106:6939-6944.

2. Bayliss R, Sardon T, Vernos I, Conti E: Structural basis of Aurora-A activation by TPX2 at the mitotic spindle. Mol Cell 2003, 12:851-862.


3. Sessa F, Mapelli M, Ciferri C, Tarricone C, Areces LB, Schneider TR, Stukenberg PT, Musacchio A: Mechanism of Aurora B activation by INCENP and inhibition by hesperadin. Mol Cell 2005, 18:379-391.

4. Littlepage LE, Ruderman JV: Identification of a new APC/C recognition domain, the A box, which is required for the Cdh1-dependent destruction of the kinase Aurora-A during mitotic exit. Genes Dev 2002, 16:2274-2285.

5. Glover DM, Leibowitz MH, McLean DA, Parry H: Mutations in aurora prevent centrosome separation leading to the formation of monopolar spindles. Cell 1995, 81:95-105.

6. Hannak E, Kirkham M, Hyman AA, Oegema K: Aurora-A kinase is required for centrosome maturation in Caenorhabditis elegans. J Cell Biol 2001, 155:1109-1116.

7. Berdnik D, Knoblich JA: Drosophila Aurora-A is required for centrosome maturation and actin-dependent asymmetric protein localization during mitosis. Curr Biol 2002, 12:640-647.

8. De Luca M, Brunetto L, Asteriti IA, Giubettini M, Lavia P, Guarguaglini G: Aurora-A and ch-TOG act in a common pathway in control of spindle pole integrity. Oncogene 2008, 27:6539-6549.

9. Barr AR, Gergely F: Aurora-A: the maker and breaker of spindle poles. J Cell Sci 2007, 120:2987-2996.

10. Groisman I, Jung MY, Sarkissian M, Cao Q, Richter JD: Translational control of the embryonic cell cycle. Cell 2002, 109:473-483.

11. Sen S, Zhou H, White RA: A putative serine/threonine kinase encoding gene BTAK on chromosome 20q13 is amplified and overexpressed in human breast cancer cell lines. Oncogene 1997, 14:2195-2200.

12. Katayama H, Brinkley WR, Sen S: The Aurora kinases: role in cell transformation and tumorigenesis. Cancer Metastasis Rev 2003, 22:451-464.

13. Bischoff JR, Anderson L, Zhu Y, Mossie K, Ng L, Souza B, Schryver B, Flanagan P, Clairvoyant F, Ginther C, et al.: A homologue of Drosophila aurora kinase is oncogenic and amplified in human colorectal cancers. Embo J 1998, 17:3052-3065.

14. Zhou H, Kuang J, Zhong L, Kuo WL, Gray JW, Sahin A, Brinkley BR, Sen S: Tumour amplified kinase STK15/BTAK induces centrosome amplification, aneuploidy and transformation. Nat Genet 1998, 20:189-193.

15. Anand S, Penrhyn-Lowe S, Venkitaraman AR: AURORA-A amplification overrides the mitotic spindle assembly checkpoint, inducing resistance to Taxol. Cancer Cell 2003, 3:51-62.

16. Zhang D, Shimizu T, Araki N, Hirota T, Yoshie M, Ogawa K, Nakagata N, Takeya M, Saya H: Aurora A overexpression induces cellular senescence in mammary gland hyperplastic tumors developed in p53-deficient mice. Oncogene 2008, 27:4305-4314.

17. Tatsuka M, Sato S, Kitajima S, Suto S, Kawai H, Miyauchi M, Ogawa I, Maeda M, Ota T, Takata T: Overexpression of Aurora-A potentiates HRAS-mediated oncogenic transformation and is implicated in oral carcinogenesis. Oncogene 2005, 24:1122-1127.


18. Tatsuka M, Sato S, Kanda A, Miki T, Kamata N, Kitajima S, Kudo Y, Takata T: Oncogenic role of nuclear accumulated Aurora-A. Mol Carcinog 2009.

19. Liu D, Vader G, Vromans MJ, Lampson MA, Lens SM: Sensing chromosome bi-orientation by spatial separation of aurora B kinase from kinetochore substrates. Science 2009, 323:1350-1353.

20. Rosasco-Nitcher SE, Lan W, Khorasanizadeh S, Stukenberg PT: Centromeric Aurora-B activation requires TD-60, microtubules, and substrate priming phosphorylation. Science 2008, 319:469-472.

21. Coelho PA, Queiroz-Machado J, Carmo AM, Moutinho-Pereira S, Maiato H, Sunkel CE: Dual role of topoisomerase II in centromere resolution and aurora B activity. PLoS Biol 2008, 6:e207.

22. Ditchfield C, Johnson VL, Tighe A, Ellston R, Haworth C, Johnson T, Mortlock A, Keen N, Taylor SS: Aurora B couples chromosome alignment with anaphase by targeting BubR1, Mad2, and Cenp-E to kinetochores. J Cell Biol 2003, 161:267-280.

23. Hauf S, Cole RW, LaTerra S, Zimmer C, Schnapp G, Walter R, Heckel A, van Meel J, Rieder CL, Peters JM: The small molecule Hesperadin reveals a role for Aurora B in correcting kinetochore-microtubule attachment and in maintaining the spindle assembly checkpoint. J Cell Biol 2003, 161:281-294.

24. Carvalho A, Carmena M, Sambade C, Earnshaw WC, Wheatley SP: Survivin is required for stable checkpoint activation in taxol-treated HeLa cells. J Cell Sci 2003, 116:2987-2998.

25. Lens SM, Wolthuis RM, Klompmaker R, Kauw J, Agami R, Brummelkamp T, Kops G, Medema RH: Survivin is required for a sustained spindle checkpoint arrest in response to lack of tension. Embo J 2003, 22:2934-2947.

26. Morrow CJ, Tighe A, Johnson VL, Scott MI, Ditchfield C, Taylor SS: Bub1 and aurora B cooperate to maintain BubR1-mediated inhibition of APC/CCdc20. J Cell Sci 2005, 118:3639-3652.

27. McCleland ML, Gardner RD, Kallio MJ, Daum JR, Gorbsky GJ, Burke DJ, Stukenberg PT: The highly conserved Ndc80 complex is required for kinetochore assembly, chromosome congression, and spindle checkpoint activity. Genes Dev 2003, 17:101-114.

28. Schumacher JM, Golden A, Donovan PJ: AIR-2: An Aurora/Ipl1-related protein kinase associated with chromosomes and midbody microtubules is required for polar body extrusion and cytokinesis in Caenorhabditis elegans embryos. J Cell Biol 1998, 143:1635-1646.

29. Terada Y, Tatsuka M, Suzuki F, Yasuda Y, Fujita S, Otsu M: AIM-1: a mammalian midbody-associated protein required for cytokinesis. Embo J 1998, 17:667-676.

30. Bringmann H, Hyman AA: A cytokinesis furrow is positioned by two consecutive signals. Nature 2005, 436:731-734.

31. Yabe T, Ge X, Lindeman R, Nair S, Runke G, Mullins MC, Pelegri F: The maternal-effect gene cellular island encodes aurora B kinase and is essential for furrow formation in the early zebrafish embryo. PLoS Genet 2009, 5:e1000518.


32. Adams RR, Tavares AA, Salzberg A, Bellen HJ, Glover DM: pavarotti encodes a kinesin-like protein required to organize the central spindle and contractile ring for cytokinesis. Genes Dev 1998, 12:1483-1494.

33. Mendoza M, Norden C, Durrer K, Rauter H, Uhlmann F, Barral Y: A mechanism for chromosome segregation sensing by the NoCut checkpoint. Nat Cell Biol 2009, 11:477-483.

34. Norden C, Mendoza M, Dobbelaere J, Kotwaliwale CV, Biggins S, Barral Y: The NoCut pathway links completion of cytokinesis to spindle midzone function to prevent chromosome breakage. Cell 2006, 125:85-98.

35. Steigemann P, Wurzenberger C, Schmitz MH, Held M, Guizetti J, Maar S, Gerlich DW: Aurora B-mediated abscission checkpoint protects against tetraploidization. Cell 2009, 136:473-484.

36. Harrington EA, Bebbington D, Moore J, Rasmussen RK, Ajose-Adeogun AO, Nakayama T, Graham JA, Demur C, Hercend T, Diu-Hercend A, et al.: VX-680, a potent and selective small-molecule inhibitor of the Aurora kinases, suppresses tumor growth in vivo. Nat Med 2004, 10:262-267.

37. Okabe S, Tauchi T, Ohyashiki JH, Ohyashiki K: Mechanism of MK-0457 efficacy against BCR-ABL positive leukemia cells. Biochem Biophys Res Commun 2009, 380:775-779.

38. Rubin EH, Shapiro GI, Stein MN, Watson P, Bergstrom D, Xiao A, Clark JB, Freedman SJ, Eder JP: A phase I clinical and pharmacokinetic (PK) trial of the aurora kinase (AK) inhibitor MK-0457 in cancer patients. J Clin Oncol (Meeting Abstracts) 2006, 24:3009-.

39. Giles FJ, Cortes J, Jones D, Bergstrom D, Kantarjian H, Freedman SJ: MK-0457, a novel kinase inhibitor, is active in patients with chronic myeloid leukemia or acute lymphocytic leukemia with the T315I BCR-ABL mutation. Blood 2007, 109:500-502.

40. Carpinelli P, Ceruti R, Giorgini ML, Cappella P, Gianellini L, Croci V, Degrassi A, Texido G, Rocchetti M, Vianello P, et al.: PHA-739358, a potent inhibitor of Aurora kinases with a selective target inhibition profile relevant to cancer. Mol Cancer Ther 2007, 6:3158-3168.

41. Gontarewicz A, Balabanov S, Keller G, Colombo R, Graziano A, Pesenti E, Benten D, Bokemeyer C, Fiedler W, Moll J, et al.: Simultaneous targeting of Aurora kinases and Bcr-Abl kinase by the small molecule inhibitor PHA-739358 is effective against imatinib-resistant BCR-ABL mutations including T315I. Blood 2008, 111:4355-4364.

42. Curry J, Angove H, Fazal L, Lyons J, Reule M, Thompson N, Wallis N: Aurora B kinase inhibition in mitosis: strategies for optimising the use of aurora kinase inhibitors such as AT9283. Cell Cycle 2009, 8:1921-1929.

43. Howard S, Berdini V, Boulstridge JA, Carr MG, Cross DM, Curry J, Devine LA, Early TR, Fazal L, Gill AL, et al.: Fragment-based discovery of the pyrazol-4-yl urea (AT9283), a multitargeted kinase inhibitor with potent aurora kinase activity. J Med Chem 2009, 52:379-388.

44. Manfredi MG, Ecsedy JA, Meetze KA, Balani SK, Burenkova O, Chen W, Galvin KM, Hoar KM, Huck JJ, LeRoy PJ, et al.: Antitumor activity of


MLN8054, an orally active small-molecule inhibitor of Aurora A kinase. Proc Natl Acad Sci U S A 2007, 104:4106-4111.

45. Macarulla T, Rodriguez-Braun E, Tabernero J, Rosello S, Baselga J, Lee Y, Manfredi M, Liu H, Fingert H, Cervantes A: Phase I pharmacokinetic (PK) and pharmacodynamic (PD) study of the selective aurora A kinase (AAK) inhibitor MLN8054 in patients (pts) with advanced solid tumors. J Clin Oncol (Meeting Abstracts) 2009, 27:2578-.

46. Cervantes-Ruiperez A, Elez ME, Rosello S, Macarulla T, Rodriguez-Braun E, Lee Y, Ecsedy J, Liu H, Fingert H, Tabernero J: Phase I pharmacokinetic (PK) and pharmacodynamic (PD) study of MLN8237, a novel selective aurora A kinase (AAK) inhibitor, in patients (pts) with advanced solid tumors. J Clin Oncol (Meeting Abstracts) 2009, 27:2565-.

47. Wilkinson RW, Odedra R, Heaton SP, Wedge SR, Keen NJ, Crafter C, Foster JR, Brady MC, Bigley A, Brown E, et al.: AZD1152, a selective inhibitor of Aurora B kinase, inhibits human tumor xenograft growth by inducing apoptosis. Clin Cancer Res 2007, 13:3682-3688.

48. Mortlock AA, Foote KM, Heron NM, Jung FH, Pasquet G, Lohmann JJ, Warin N, Renaud F, De Savi C, Roberts NJ, et al.: Discovery, synthesis, and in vivo activity of a new class of pyrazoloquinazolines as selective inhibitors of aurora B kinase. J Med Chem 2007, 50:2213-2224.

49. Yang J, Ikezoe T, Nishioka C, Tasaka T, Taniguchi A, Kuwayama Y, Komatsu N, Bandobashi K, Togitani K, Koeffler HP, et al.: AZD1152, a novel and selective aurora B kinase inhibitor, induces growth arrest, apoptosis, and sensitization for tubulin depolymerizing agent or topoisomerase II inhibitor in human acute leukemia cells in vitro and in vivo. Blood 2007, 110:2034-2040.

50. Schellens JH, Boss D, Witteveen PO, Zandvliet A, Beijnen JH, Voogel-Fuchs M, Morris C, Wilson D, Voest EE: Phase I and pharmacological study of the novel aurora kinase inhibitor AZD1152. J Clin Oncol (Meeting Abstracts) 2006, 24:3008-.

51. Jones SF, Burris HA, III, Dumez H, Infante JR, Fowst C, Gerletti P, Xu H, Jakubczak J, Mellaerts N, Schoffski P: Phase I accelerated dose-escalation, pharmacokinetic (PK) and pharmacodynamic study of PF-03814735, an oral aurora kinase inhibitor, in patients with advanced solid tumors: Preliminary results. J Clin Oncol (Meeting Abstracts) 2008, 26:2517-.

52. Robert F, Hurwitz H, Verschraegen CF, Advani R, Berman C, Taverna P, Evanchik M: Phase 1 trial of SNS-314, a novel selective inhibitor of aurora kinases A, B, and C, in advanced solid tumor patients. J Clin Oncol (Meeting Abstracts) 2008, 26:14642-.

53. Anderson K, Lai Z, McDonald OB, Stuart JD, Nartey EN, Hardwicke MA, Newlander K, Dhanak D, Adams J, Patrick D, et al.: Biochemical characterization of GSK1070916, a potent and selective inhibitor of Aurora B and Aurora C kinases with an extremely long residence time1. Biochem J 2009, 420:259-265.

54. Agnese V, Bazan V, Fiorentino FP, Fanale D, Badalamenti G, Colucci G, Adamo V, Santini D, Russo A: The role of Aurora-A inhibitors in cancer therapy. Ann Oncol 2007, 18 Suppl 6:vi47-52.


55. Barr FA, Sillje HH, Nigg EA: Polo-like kinases and the orchestration of cell division. Nat Rev Mol Cell Biol 2004, 5:429-440.

56. Petronczki M, Lenart P, Peters JM: Polo on the Rise-from Mitotic Entry to Cytokinesis with Plk1. Dev Cell 2008, 14:646-659.

57. Toyoshima-Morimoto F, Taniguchi E, Nishida E: Plk1 promotes nuclear translocation of human Cdc25C during prophase. EMBO Rep 2002, 3:341-348.

58. Toyoshima-Morimoto F, Taniguchi E, Shinya N, Iwamatsu A, Nishida E: Polo-like kinase 1 phosphorylates cyclin B1 and targets it to the nucleus during prophase. Nature 2001, 410:215-220.

59. Petronczki M, Glotzer M, Kraut N, Peters JM: Polo-like kinase 1 triggers the initiation of cytokinesis in human cells by promoting recruitment of the RhoGEF Ect2 to the central spindle. Dev Cell 2007, 12:713-725.

60. Elowe S, Hummer S, Uldschmid A, Li X, Nigg EA: Tension-sensitive Plk1 phosphorylation on BubR1 regulates the stability of kinetochore microtubule interactions. Genes Dev 2007, 21:2205-2219.

61. Lenart P, Petronczki M, Steegmaier M, Di Fiore B, Lipp JJ, Hoffmann M, Rettig WJ, Kraut N, Peters JM: The small-molecule inhibitor BI 2536 reveals novel insights into mitotic roles of polo-like kinase 1. Curr Biol 2007, 17:304-315.

62. Steegmaier M, Hoffmann M, Baum A, Lenart P, Petronczki M, Krssak M, Gurtler U, Garin-Chesa P, Lieb S, Quant J, et al.: BI 2536, a potent and selective inhibitor of polo-like kinase 1, inhibits tumor growth in vivo. Curr Biol 2007, 17:316-322.

63. Takai N, Hamanaka R, Yoshimatsu J, Miyakawa I: Polo-like kinases (Plks) and cancer. Oncogene 2005, 24:287-291.

64. Weichert W, Kristiansen G, Schmidt M, Gekeler V, Noske A, Niesporek S, Dietel M, Denkert C: Polo-like kinase 1 expression is a prognostic factor in human colon cancer. World J Gastroenterol 2005, 11:5644-5650.

65. Weichert W, Kristiansen G, Winzer KJ, Schmidt M, Gekeler V, Noske A, Muller BM, Niesporek S, Dietel M, Denkert C: Polo-like kinase isoforms in breast cancer: expression patterns and prognostic implications. Virchows Arch 2005, 446:442-450.

66. Weichert W, Schmidt M, Jacob J, Gekeler V, Langrehr J, Neuhaus P, Bahra M, Denkert C, Dietel M, Kristiansen G: Overexpression of Polo-like kinase 1 is a common and early event in pancreatic cancer. Pancreatology 2005, 5:259-265.

67. Kanaji S, Saito H, Tsujitani S, Matsumoto S, Tatebe S, Kondo A, Ozaki M, Ito H, Ikeguchi M: Expression of polo-like kinase 1 (PLK1) protein predicts the survival of patients with gastric carcinoma. Oncology 2006, 70:126-133.

68. Yamada S, Ohira M, Horie H, Ando K, Takayasu H, Suzuki Y, Sugano S, Hirata T, Goto T, Matsunaga T, et al.: Expression profiling and differential screening between hepatoblastomas and the corresponding normal livers: identification of high expression of the PLK1 oncogene as a poor-prognostic indicator of hepatoblastomas. Oncogene 2004, 23:5901-5911.

69. Smith MR, Wilson ML, Hamanaka R, Chase D, Kung H, Longo DL, Ferris DK: Malignant transformation of mammalian cells initiated by constitutive


expression of the polo-like kinase. Biochem Biophys Res Commun 1997, 234:397-405.

70. Guan R, Tapang P, Leverson JD, Albert D, Giranda VL, Luo Y: Small interfering RNA-mediated Polo-like kinase 1 depletion preferentially reduces the survival of p53-defective, oncogenic transformed cells and inhibits tumor growth in animals. Cancer Res 2005, 65:2698-2704.

71. Liu X, Erikson RL: Polo-like kinase (Plk)1 depletion induces apoptosis in cancer cells. Proc Natl Acad Sci U S A 2003, 100:5789-5794.

72. Spankuch-Schmitt B, Bereiter-Hahn J, Kaufmann M, Strebhardt K: Effect of RNA silencing of polo-like kinase-1 (PLK1) on apoptosis and spindle formation in human cancer cells. J Natl Cancer Inst 2002, 94:1863-1877.

73. Sumara I, Gimenez-Abian JF, Gerlich D, Hirota T, Kraft C, de la Torre C, Ellenberg J, Peters JM: Roles of polo-like kinase 1 in the assembly of functional mitotic spindles. Curr Biol 2004, 14:1712-1722.

74. Nappi TC, Salerno P, Zitzelsberger H, Carlomagno F, Salvatore G, Santoro M: Identification of Polo-like kinase 1 as a potential therapeutic target in anaplastic thyroid carcinoma. Cancer Res 2009, 69:1916-1923.

75. Mross K, Frost A, Steinbild S, Hedbom S, Rentschler J, Kaiser R, Rouyrre N, Trommeshauser D, Hoesl CE, Munzert G: Phase I dose escalation and pharmacokinetic study of BI 2536, a novel Polo-like kinase 1 inhibitor, in patients with advanced solid tumors. J Clin Oncol 2008, 26:5511-5517.

76. Gumireddy K, Reddy MV, Cosenza SC, Boominathan R, Baker SJ, Papathi N, Jiang J, Holland J, Reddy EP: ON01910, a non-ATP-competitive small molecule inhibitor of Plk1, is a potent anticancer agent. Cancer Cell 2005, 7:275-286.

77. Jimeno A, Li J, Messersmith WA, Laheru D, Rudek MA, Maniar M, Hidalgo M, Baker SD, Donehower RC: Phase I study of ON 01910.Na, a novel modulator of the Polo-like kinase 1 pathway, in adult patients with solid tumors. J Clin Oncol 2008, 26:5504-5510.

78. Lansing TJ, McConnell RT, Duckett DR, Spehar GM, Knick VB, Hassler DF, Noro N, Furuta M, Emmitte KA, Gilmer TM, et al.: In vitro biological activity of a novel small-molecule inhibitor of polo-like kinase 1. Mol Cancer Ther 2007, 6:450-459.

79. Rudolph D, Steegmaier M, Hoffmann M, Grauert M, Baum A, Quant J, Haslinger C, Garin-Chesa P, Adolf GR: BI 6727, a Polo-like kinase inhibitor with improved pharmacokinetic profile and broad antitumor activity. Clin Cancer Res 2009, 15:3094-3102.

80. Olmos D, Allred A, Sharma R, Brunetto A, Smith D, Murray S, Barker D, Taegtmeyer A, de Bono J, Blagden S: Phase I first-in-human study of the polo-like kinase-1 selective inhibitor, GSK461364, in patients with advanced solid tumors. J Clin Oncol (Meeting Abstracts) 2009, 27:3536-.

81. Schoffski P: Polo-Like Kinase (PLK) Inhibitors in Preclinical and Early Clinical Development in Oncology. Oncologist 2009, 14:559-570.

82. Takagi M, Honmura T, Watanabe S, Yamaguchi R, Nogawa M, Nishimura I, Katoh F, Matsuda M, Hidaka H: In vivo antitumor activity of a novel sulfonamide, HMN-


214, against human tumor xenografts in mice and the spectrum of cytotoxicity of its active metabolite, HMN-176. Invest New Drugs 2003, 21:387-399.

83. Garland LL, Taylor C, Pilkington DL, Cohen JL, Von Hoff DD: A phase I pharmacokinetic study of HMN-214, a novel oral stilbene derivative with polo-like kinase-1-interacting properties, in patients with advanced solid tumors. Clin Cancer Res 2006, 12:5182-5189.

84. Santamaria A, Neef R, Eberspacher U, Eis K, Husemann M, Mumberg D, Prechtl S, Schulze V, Siemeister G, Wortmann L, et al.: Use of the novel Plk1 inhibitor ZK-thiazolidinone to elucidate functions of Plk1 in early and late stages of mitosis. Mol Biol Cell 2007, 18:4024-4036.

85. Peters U, Cherian J, Kim JH, Kwok BH, Kapoor TM: Probing cell-division phenotype space and Polo-like kinase function using small molecules. Nat Chem Biol 2006, 2:618-626.

86. Nasmyth K: How do so few control so many? Cell 2005, 120:739-746. 87. Yu H: Cdc20: a WD40 activator for a cell cycle degradation machine. Mol

Cell 2007, 27:3-16. 88. Tang Z, Bharadwaj R, Li B, Yu H: Mad2-Independent inhibition of APCCdc20

by the mitotic checkpoint protein BubR1. Dev Cell 2001, 1:227-237. 89. Sudakin V, Chan GK, Yen TJ: Checkpoint inhibition of the APC/C in HeLa

cells is mediated by a complex of BUBR1, BUB3, CDC20, and MAD2. J Cell Biol 2001, 154:925-936.

90. Peters JM: The anaphase promoting complex/cyclosome: a machine designed to destroy. Nat Rev Mol Cell Biol 2006, 7:644-656.

91. Nilsson J, Yekezare M, Minshull J, Pines J: The APC/C maintains the spindle assembly checkpoint by targeting Cdc20 for destruction. Nat Cell Biol 2008.

92. Abrieu A, Kahana JA, Wood KW, Cleveland DW: CENP-E as an essential component of the mitotic checkpoint in vitro. Cell 2000, 102:817-826.

93. Mao Y, Abrieu A, Cleveland DW: Activating and silencing the mitotic checkpoint through CENP-E-dependent activation/inactivation of BubR1. Cell 2003, 114:87-98.

94. Mao Y, Desai A, Cleveland DW: Microtubule capture by CENP-E silences BubR1-dependent mitotic checkpoint signaling. J Cell Biol 2005, 170:873-880.

95. Weaver BA, Bonday ZQ, Putkey FR, Kops GJ, Silk AD, Cleveland DW: Centromere-associated protein-E is essential for the mammalian mitotic checkpoint to prevent aneuploidy due to single chromosome loss. J Cell Biol 2003, 162:551-563.

96. Yao X, Abrieu A, Zheng Y, Sullivan KF, Cleveland DW: CENP-E forms a link between attachment of spindle microtubules to kinetochores and the mitotic checkpoint. Nat Cell Biol 2000, 2:484-491.

97. Huang H, Hittle J, Zappacosta F, Annan RS, Hershko A, Yen TJ: Phosphorylation sites in BubR1 that regulate kinetochore attachment, tension, and mitotic exit. J Cell Biol 2008, 183:667-680.

98. Choi E, Choe H, Min J, Choi JY, Kim J, Lee H: BubR1 acetylation at prometaphase is required for modulating APC/C activity and timing of mitosis. EMBO J 2009, 28:2077-2089.


99. Taylor SS, Ha E, McKeon F: The human homologue of Bub3 is required for kinetochore localization of Bub1 and a Mad3/Bub1-related protein kinase. J Cell Biol 1998, 142:1-11.

100. Choi E, Lee H: Chromosome damage in mitosis induces BubR1 activation and prometaphase arrest. FEBS Lett 2008, 582:1700-1706.

101. Nitta M, Kobayashi O, Honda S, Hirota T, Kuninaka S, Marumoto T, Ushio Y, Saya H: Spindle checkpoint function is required for mitotic catastrophe induced by DNA-damaging agents. Oncogene 2004, 23:6548-6558.

102. Minucci S, Pelicci PG: Histone deacetylase inhibitors and the promise of epigenetic (and more) treatments for cancer. Nat Rev Cancer 2006, 6:38-51.



9. Atypical DUSPs: 19 phosphatases in search of a role

Yolanda Bayón and Andrés Alonso

Instituto de Biología y Genética Molecular, CSIC-Universidad de Valladolid c/ Sanz y Forés s/n, 47003 Valladolid, Spain

Abstract. Atypical Dual Specificity Phosphatases (A-DUSPs) are a group of 19 phosphatases poorly characterized. They are included among the Class I Cys-based PTPs and contain the active site motif HCXXGXXR conserved in the Class I PTPs. These enzymes present a phosphatase domain similar to MKPs, but lack any substrate targeting domain similar to the CH2 present in this group. Although most of these phosphatases have no more than 250 amino acids, their size ranges from the 150 residues of the smallest A-DUSP, VHZ/DUSP23, to the 1158 residues of the putative PTP DUSP27. The substrates of this family include MAPK, but, in general terms, it does not look that MAPK are the general substrates for the whole group. In fact, other substrates have been described for some of these phosphatases, like the 5’CAP structure of mRNA, glycogen, or STATs and still the substrates of many A-DUSPs have not been identified. In addition to the PTP domain, most of these enzymes present no additional recognizable domains in their sequence, with the exception of CBM-20 in laforin, GTase in HCE1 and a Zn binding domain in DUSP12. Although some of these enzymes have been now studied for years there is a great lack of data about the true physiological role of this group of phosphatases.

Correspondence/Reprint request: Dr. Andrés Alonso, Instituto de Biología y Genética Molecular, CSIC-Universidad de Valladolid, c/ Sanz y Forés s/n, 47003 Valladolid, Spain. E-mail: andré[email protected]

Yolanda Bayón & Andrés Alonso 186

Introduction Protein tyrosine phosphorylation is a post-translational modification that, in a reversible manner, regulates many cellular functions such as proliferation, differentiation or citokinesis to cite a few. This process is controlled by two groups of enzymes, protein tyrosine kinases (PTKs) and protein tyrosine phosphatases (PTPs). PTKs constitute a family of proteins with the same evolutionary origin [1], whereas PTPs share a similar enzymatic activity but show a diverse evolutionary origin [2]. Thus, PTPs have been classified, according to the key aminoacid involved in the catalysis, on Cys-based PTPs or Asp-based PTPs [2]. The Cys-based PTPs share a canonical motif (P-loop) in the phosphatase domain, CX5R, that contains the Cys required to form the thiol-phosphate intermediate and can be further subdivided based on the characteristics of the catalytic domain in 3 classes. Class I is the more populated with 99 proteins and contains 2 groups, the well-known classical PTPs with 38 sequences [3] and the VH1-like PTPs with 61 members, a quite diverse group with an ample range of substrates, from the 5’CAP structure of mRNA to inositol phospholipids, including phospho-Ser, phospho-Thr and phospho-Tyr. Class-I consensus sequence for the P-loop can be extended to HCX2GX2R(S/T). In addition, these phosphatases present an Asp, approximately 30 amino acids upstream of the catalytic Cys, which acts in the catalysis as a general acid, protonating the leaving group. The phosphatase domain of Class I PTPs presents a characteristic structure, different from the topology that shows the phosphatase domain of the other two classes of Cys-PTPs. LMW-PTP, the only member of Class II, presents a distinctive fold and CDC25 phosphatases, which constitute the class III, contain a phosphatase domain that displays a rhodanese fold. The Asp, N-terminal to the catalytic Cys in Class I PTPs, is located in the C-terminus of LMW-PTP and is absent in Class II CDC25 enzymes. Although CDC25 phosphatases have been included frequently in a group of dual-specificity phosphatases along with VH1 phophatases, they belong to a different group that shows an exquisite specificity for their substrates, cyclin-dependent kinases. In addition to the different phosphatase domain topology, CDC25 proteins lack any sequence similarity to other DSPs, apart from the CX5R motif common to all Cys-based hydrolases. There is a fourth class of PTPs that belong to the family of Haloacid Dehalogenases (HAD) and use Asp as the nucleophile in the catalysis. The activity of these phosphatases depends on the presence of divalent cations. The first proteins of this group characterized as phosphatases were the transcription factors Eyes absent (EYA) [4]. While EYA proteins seem to

Atypical DUSPs: 19 phosphatases in search of a role 187

dephosphorylate phosphor-Tyr [5], other members of this group dephosphorylate substrates in Ser/Thr: Chronophin, which regulates the cytoskeleton by targeting Ser-3 in cofilin [6]; FCP1, which dephosphorylates the C-terminal domain of RNA polymerase II, regulating transcription [7]; and dullard, which is implicated in nuclear membrane biogenesis through dephosphorylation of the phosphatidic acid phosphatase lipin, functioning in a phosphatase cascade [8]. Table 1. A-DUSPs present in the human genome. The table includes their names and their synomins as well as the uniprot accession numbers to identify these phosphatases. Under the column of substrates, proteins that have been shown to be dephosphorylated have been included, along with those that have been tested but are not targets of these phosphatases.

PROTEIN SYNONIMS AA UNIPROT SUBSTRATES

VHR DUSP3 185 P51452 ERK, JNK, p38, STAT5

MDSP BEDP, DUSP13a 188 Q6B8I1 Unknown

TMDP SKRP4, TS-DSP6, DUSP13b

198 Q9UII6 Unknown and Negative on ERK, JNK, p38 and DYRkβ

DUSP14 MKP-6, MKP-L 198 O95147 ERK, JNK

DUSP15 VHY 295 Q9H1R2 Unknown

DUSP18 LMW-DSP20 188 Q8NEJ0 JNK

DUSP19 DUSP17, SKRP1, TS-DSP1, LMW-DSP3, LDP-2

217 Q8WTR2 JNK

DUSP21 LMW-DUSP21 190 Q9H596 Unknown and negative on ERK, JNK, p38

DUSP22 JSP-1, MKP-X, JKAP, LMW-DSP2, VHX

184 Q9NRW4 ERK, JNK (upstream protein), p38, ERα, STAT3

DUSP23 LDP3, VHZ 150 Q9BVJ7 ERK, JNK (upstream protein), p38

DUSP26 DUSP24, LDP4, MKP8, NATA1, SKRP3, NEAP

211 Q9BV47 ERK, JNK, p38, Akt, Kap3

DUPD1 DUSP27; FMDSP 220 Q68J44 Unknown

STYX 223 Q8WUJ0 Inactive

PTPM1 MOSP, PLIP, PNAS-129

201 Q8WUK0 PI(5)P

Laforin 331 O95278 Glycogen, GSK3β

DUSP11 PIR1 330 O75319 5’CAP

HCE1 CAP1A, HCAP1A 597 O60942 5’CAP

DUSP12 HYVH1, T-DSP4, GKAP

340 Q9UNI6 GKAP Negative on ERK, JNK, p38

DUSP27 STYXL2 1158 Q5VZP5 Inactive


In this chapter, we will discuss the group of VH1-like phosphatases known as A-DUSPs. These enzymes are included in the VH1-like group of Class I PTPs and contain 19 proteins (see table 1) that will be discussed in detail, focused on the physiological role played by these phosphatases. The viral protein VH1 (Vaccinia virus H1 open reading frame) was the first DUSP identified by Dixon and colleges in 1991 [9]. The next year, VHR (VH1-related PTP) was isolated and cloned [10]. Efforts to identify the PTPs present in the human genome and other genomes have advanced in parallel to the sequencing of the human genome. Thus, some sequences were isolated at the end of the 90s, but most the A-DUSPs sequences were identified in the new milleniun. DUPD1 is the most recently characterized phosphatase [11] and it still remains in the databases, DUSP27, an inactive A-DUSP whose mRNA has been detected but no study has been published yet on this protein. Substrates of these enzymes are diverse, as they target not only proteins phosphorylated on Ser, Thr, and Tyr but also the mRNA capping structure or glycogen. A great effort has been dedicated to show that these enzymes dephosphorylate MAPKs (mitogen activated protein kinases) with negative results in some cases and controversial data in others. In general terms, data available on many of these phosphatases is still limited, given the recent identification of many of them. VHR /DUSP3 VHR is a phosphatase of 185 amino acids that was cloned in 1992 with an expression cloning strategy [10]. In all these years, VHR has been widely used as a model to research general aspects of Class I PTPs, as the kinetic mechanism of PTPs [12-14] or to design inhibitors for these enzymes [15]. However, research about the physiological role has not been very productive, although VHR has started to reveal its role in cell cycle regulation [16] and in the activation of several signalling pathways related with the immune response [17, 18]. VHR was the first A-DUSP to be crystallized [19] and its three dimensional structure showed that VHR presents a catalytic cleft 6 Å deep, able to accommodate phospho-Ser and phospho-Tyr, in contrast to the deeper pocket (9 Å) of classic PTPs, like PTP1B [20]. This difference would explain the substrate preference of classic PTPs toward phospho-Tyr. VHR can be found in the cytosol as well as in the nucleus, and the localization might have important implications for its pathological function associated with proliferation [16]. Expression of VHR is ubiquitous [21] and is regulated during the cell cycle [16]. Similarity between VHR and MKPs catalytic domain suggested that VHR could dephosphorylate MAPKs in a similar way to MKPs. However, it


took a few years to demonstrate this point. In fact, it was not until 1999 when VHR was shown to dephosphorylate ERK (extracellular regulated kinase) [22], and a few years later when VHR was proven to dephoshorylate JNK (c-jun N-terminal kinase) [21, 23]. In addition, some studies have claimed dephosphorylation of p38 by VHR [24, 25]. In experiments in vitro using synthetic diphosphorylated peptides, corresponding to the activation loop of ERK and JNK, it was determined that VHR preferencially hydrolyze the phospho-Tyr residue in that sequence, thereby VHR would dephosphorylate phospho-Tyr followed by slow dephosphorylation of phospho-Thr [26]. However, a later study using phosphorylated recombinant ERK as substrate demonstrated that, in vitro, ERK is a poor substrate for VHR [27], indicating that, in vivo, it must exist some mechanism that allows the dephosphorylation of ERK and JNK by VHR. One possibility is the existence of a scaffold that would bring together this phosphatase and its substrate, function that could be developed by VRK3 [28] (see below). Reduction in VHR expression using siRNA interference has also shown an increase in the phosphorylation of ERK and JNK [16]. Although the capacity of A-DUSPs, including VHR, to dephosphorylate MAPK remains controversial, at least in this case, the available data suggest that VHR is a true MAPK phosphatase. Regulation of VHR function is accomplished by Tyr phosphorylation. In T cells, VHR is phosphorylated by the tyrosine kinase Zap70 [17], which is critical for antigen signalling through the TCR (T cell receptor), in Tyr-138, phosphorylation that is required for VHR to inhibit ERK2; conversely, the mutation of Tyr-138 by Phe in VHR increases TCR-induced ERK2 kinase activity as well as activation of interleukin 2 (IL-2) promoter [17]. In a different situation, it has been described that VHR dephosphorylates STAT5 in cells stimulated by IFN-α and IFN-β [18]. Again, phosphorylation of VHR at Tyr-138 was required for its phosphatase activity toward STAT5. The tyrosine kinase Tyk2, which phosphorylates STAT5, is also responsible for the phosphorylation of VHR at Tyr-138. Thereby, these two reports suggest that phosphorylation of VHR in Tyr-138 links kinases that participate in the activation of signalling pathways to its ulterior down-regulation of the pathway by returning the components to the basal state. The biological function of VHR has started to be uncovered in recent years. In this sense, it has been shown, using siRNA interference that cells with a reduced expression of VHR arrest at the G1-S and G2-M transitions of the cell cycle and exhibit senescence [16]. In agreement with this notion, cells lacking VHR were found to upregulate the cyclin dependent kinase inhibitor p21 (Cip-Waf1), and to reduce the expression of genes for cell-cycle regulators, DNA replication, transcription and mRNA processing. Loss of VHR also caused, after serum-induced activation, a several-fold increase of


its substrates, the MAPKs ERK and JNK. VHR-induced cell-cycle arrest was dependent on this hyperactivation of JNK and ERK, and was reversed by ERK and JNK inhibition or knockdown by RNA interference. These data supports the involvement of VHR in cell-cycle progression, role that is linked to its regulation of MAPK activation in the cell-cycle. A few reports have implicated VHR in different types of cancer [29-31]. This role of VHR seems to be related to its role as a MAPK phosphatase. In the prostate cancer cell line LNCaP androgens protect from TPA (12-O-tetradecanoylphorbol-13-acetate) or thapsigargin-induced apoptosis via down-regulation of JNK activation. This effect could be explained by an increased in VHR expression induced by androgens [29]. Transfection of wild-type VHR, but not a catalytically inactive mutant, interfered with TPA and TG-induced apoptosis. Consistently, siRNA-mediated knockdown of endogenous VHR increased apoptosis in response to TPA or TG in the presence of androgens. In this work, it is also found that VHR expression is augmented in prostate cancer cells compared with normal prostate cells. Overexpression of BRCA1-IRIS, a product of the BRCA1 gene, in breast cancer, augments Cyclin D1 expression and increases cell proliferation. Increase in Cyclin D1 is explained by a reduction in VHR expression regulated by BRCA1-IRIS by an unknown mechanism [30]. Furthermore, overexpression of Cyclin D1 could be blocked by transfection of VHR in those cells. An additional study in cervix cancer has reported an increase in the expression of VHR and a change in its location [31]. Thus, in invasive cervix cancer, in addition to be highly expressed, VHR changes its localization to the nucleus, while in normal tissue VHR is always in the cytoplasm. In addition, cervix cancer cell lines such as HeLa, SiHa, CaSki, C33 and HT3 also present a higher expression of VHR, compared to primary keratinocytes. Altogether, the role that plays VHR in cell cycle progression along with VHR implications in several cancer types, suggest that this phosphatase could be a therapeutical target in this disease. DUSP26 Several independent groups initiated the characterization of this phosphatase and published their studies between 2005 and 2007. The results obtained by these researches were as diverse as to show one effect, for example inhibition of p38 [32, 33] and JNK [33, 34], and the opposite, stimulation of p38 and JNK [35]. It has also been reported inhibition of ERK [34], in this case mediated by the interaction of DUSP26 with the protein Hsf4 (heat shock transcription factors 4), a substrate of MAPKs. Furthermore, an independent study in PC12 cells found no effect in any of


Figure 1. Domains present in the A-DUSPs. Apart of the characteristic PTP domain (150 aminoacids) presented by these proteins, only three of these phosphatases present an additional domain: CBM-20 (carbohydrate binding module-family 20) in Laforin, a GTase (guanylyltransferase) domain in HCE1, the mRNA capping enzyme and a Zn binding domain (Zn) in DUSP12. All the phosphatases have drawn to scale except DUSP27 that is the longest (1158 aminoacids). the main MAPKs, ERK, JNK or p38 [36]. Apart of MAPKs, an additional substrate has been described for DUSP26 in PC12 cells, Akt, since overexpression of DUSP26 reduced Akt phosphorylation on Ser-473 [37]. DUSP26 is localized mainly in the nucleus [32, 35], although in PC12 cells is found in the cytosol [36]; and its expression is enriched in the brain [34, 35], particularly in neurons [34]. In addition, this phosphatase has been detected in several cancer cell lines, retinoblastoma, neuroepithelioma, and neuroblastoma [38]. In this sense, 8p12, the chromosomal region where DUSP26 gene is present, has been found to be amplified in anaplastic thyroid cancer (ATC) and, consequently, DUSP26 expression was increased in primary ATC tumors and ATC cell lines. In three ATC cell lines, cell growth was inhibited and apoptosis was increased by DUSP26 siRNA. In this study,


DUSP26 was detected in a complex with p38, which was dephosphorylated by this phosphatase. Therefore, in that report the authors propose that this phosphatase acts as an oncogen by promoting cell survival of ATC cells through inhibition of p38 induced apoptosis. In PC12 cells, DUSP26 expression is up-regulated by nerve growth factor (NGF) and suppression of DUSP26 expression by siRNA enhanced NGF-induced neurite outgrowth and Akt activation. Moreover, DUSP26 impaired cell growth in response to EGF (epidermal growth factor) by reducing EGFR (EGF receptor) mRNA and protein, through down-regulation of the Akt pathway and Wilms' tumor gene product (WT1). Taken together, these studies suggest that DUSP26 could be involved in PC12 differenciation in response to NGF and EGF via regulation of the PI3K/Akt signalling pathway. Recently, DUSP26 has also been involved in intracellular transport and cell adhesion [38], since it was found that DUSP26 binds KIF3, a microtubule-directed protein motor involved in transport of β-catenin/N-Cadherin to the plasma membrane. DUSP26 would be recruited to the KIF3 complex through interaction with one subunit of this complex, Kif3a, to dephosphorylate Kap3, which should be critical for transport of β-catenin/N-Cadherin. In this sense, over-expression of DUSP26 increases cell-cell contacts and adhesion. These authors also described a down-regulation of DUSP26 in gliomas, suggesting a tumor-suppressive role of this phosphatase. DUSP12 DUSP12 was identified based on similarity with the yeast VH1 (YVH1) phosphatase, and then it was named human orthologue of YVH1 (hYVH1) phosphatase [39]. Several studies support a role for this phosphatase in cell growth [40], and meiosis and sporulation [41] in yeast. Independently, the rat orthologue was cloned from a liver cDNA library in a yeast-two hybrid screen with glucokinase and, thus, named glucokinase-associated phosphatase (GKAP) [42]. This protein contains an N-terminal DUSP domain and a cysteine-rich C-terminal domain capable of binding 2 mol of zinc/mol of protein, defining it as a novel zinc finger domain, which is essential for in vivo function [39]. DUSP12 in human cells is expressed in all tissues analyzed [39] and similar results have been obtained for GKAP. Expression of GKAP was also verified by RT-PCR in β-pancreatic cells [42]. DUSP12, in vitro, dephosphorylated phospho-Tyr, but not the artificial substrate phosphoseryl kemptide, although it could hydrolyze phosphate from Ser phosphorylated glucokinase [42]. GKAP fused with GFP (green fluorescent protein) was found in the cytosol, where glucokinase


phosphorylates glucose. Furthermore, addition of DUSP12 enhanced glucokinase activity in vitro in a dose-dependent manner, suggesting that this association is of functional significance. Implications of DUSP12 in glucose metabolism have received further support from genetic studies that have shown that the chromosomal region 1q21-23, where the DUSP12 gene resides, is linked to type 2 diabetes (T2D) [43]. Furthermore, it has been found that several polymorphisms in DUSP12 are associated with T2D [44]. In addition, this chromosomal region is amplified in human liposarcomas and DUSP12 gene showed the highest level of amplification within that region [45]. A recent work has reported the interaction of Hsp70 (heat-shock protein 70) with DUSP12 [46]. The Zn-binding domain is required for this interaction, along with the ATPase domain of Hsp70. DUSP12 phosphatase activity towards an exogenous substrate under non-reducing conditions was increased by Hsp70. This study also supports a role of this phosphatase in cell survival, since overexpression of DUSP12 reduced the number of apoptotic cells in response to various apoptotic inducers, such as heat shock, H2O2 and Fas receptor activation. Activity of DUSP12 on ERK, JNK or p38 was tested under the stress conditions analysed, finding that this phosphatase does not dephosphorylate any of these kinases [46]. Furthermore, purified DUSP12 was unable to dephosphorylate purified activated MAPKs in vitro in the absence or presence of Hsp70. Therefore, substrates for DUSP12 remained to be identified. The role of this phosphatase on cell survival is further supported by data from a large-scale siRNA screening conducted by Blenis and co-workers that showed that knockdown of DUSP12 with siRNA caused significant apoptosis [47]. However, overexpression of DUSP12 could not rescue the cells exposed to cisplatin, which initiates apoptosis via the DNA damage response pathway, and suggest that hYVH1’s effects are selective to specific cytotoxic stimuli, just mentioned above, which involve redox-sensitive signalling pathways. Both phosphatase and Zn-binding domains are required for the anti-apoptotic effects caused by this phosphatase. In an independent study, looking for genes that increase resistance to Fas-induced apoptosis, the expression of DUSP12 was found to be down regulated [48], confirming the role of this phosphatase in resistance to apoptosis and cell survival. LAFORIN Mutations in this DUSP cause the Lafora disease (LD) (OMIM254780) [49, 50], an autosomal recessive neurodegenerative disorder that results in progressive myoclonus epilepsy with onset in adolescence and death within 10 years of appearance. This disease was described at the beginning of last


century, in 1911, by the neurologist Gonzalo Rodríguez Lafora [51], a disciple of Santiago Ramón y Cajal. The hallmark of the disease is the formation of Lafora bodies, which contain insoluble, poorly branched glycogen-like polysaccharide or “polyglucosan”. These bodies develop in different tissues: liver, muscle, heart, skin, and neurons that normally accumulate little glycogen. Approximately 60% of the cases of LD result from mutations in the gene that encodes laforin, EPM2A. This phosphatase contains a glycogen binding domain in the N-terminus (GBM-20), followed by the phosphatase domain. Another gene has been identified, EPM2B (NHLRC1), whose mutation causes 20–30% of the cases of Lafora disease [52, 53]. EPM2B codes for a protein called malin that is an E3 ubiquitin ligase, which promotes ubiquitination and proteasomal degradation of laforin [54, 55], among other proteins related to glycogen metabolism. The first protein identified to interact with laforin was PTG (protein targeting to glycogen) or R5 [56], a protein that targets the Ser/Thr phosphatase PP1 to glycogen [57]. PP1 regulates glycogen synthesis by dephosphorylating key enzymes involved in glycogen metabolism. Thus, PP1 activates glycogen synthetase (GS) and inactivates the glycogen degradation enzymes phosphorylase (Ph) and phosphorylase kinase (PhK). Overexpression of PTG/R5 markedly increases glycogen accumulation, and reduction of PTG expression diminishes glycogen stores. In contrast to PP1 phosphatase, several kinases such as AMPK (AMP-activated protein kinase), PKA, CKI and GSK3 could inhibit glycogen synthesis by acting on the same proteins than PP1. Later, it was shown that laforin and malin form a complex that promotes the ubiquitination and proteasome dependent degradation of PTG/R5, and, accordingly, inhibit glycogen accumulation in neurons [58]. In these cells, which do not express the enzymes involved in glycogen hydrolysis Ph and PhK, it was demonstrated that the malin-laforin complex caused proteasome-dependent degradation of MGS (muscle glycogen synthase) [58], thus mutant laforin and malin proteins would impair the capacity to degrade MGS with the consequence of increasing the amount of glycogen in neurons. Furthermore, Solaz-Fuster et al. [59] demonstrated, in FTO2B hepatoma cells, that malin, in concert with laforin, decreases PTG/R5 levels and glycogen stores without affecting glycogen synthase. Degradation of PTG/R5 by this complex could explain how the mutations that impair the physiological function of laforin or malin increase PTG levels and glycogen synthesis in the cells. In laforin-malin complex, laforin seems to behave as a substrate adaptor, recruiting proteins present in glycogen particles for ubiquitination by malin and subsequent degradation by the proteasome. Additional glycogen metabolism-related proteins have been reported to be degraded by this complex, as the glycogen debranching enzyme


(GDE)/amylo-1, 6-glucosidase,4-alpha-glucanotransferase (AGL)[60]. It has been also shown that malin promotes ubiquitination of laforin, which would mean that the same E3 ubiquitin ligase complex regulates its own quantity by controlling the amount of one of its components, the substrate adaptor laforin. Interaction between laforin and malin is regulated by phosphorylation of laforin by AMPK, which binds directly to laforin [59]. Moreover, AMPK also phosphorylates PTG/R5 on Ser-8 and Ser-268, which accelerates its ubiquitination by the laforin-malin complex [61], following the classical mechanism of tagging proteins for ubiquitination by phosphorylation. Interaction of laforin with other proteins has been described, although the functional significance of these interactions is unknown. Among these proteins are: EPM2AIP1 [62], a protein of unknown function; HIRIP5, a cytosolic protein that contains a NifU-like domain that could be involved in iron homeostasis [63]; GSK3β, which was shown to be dephosphorylated by laforin [55, 64], although this remains controversial. In addition, it has been described that laforin forms homodimers [56, 64], which are critical for its phosphatase activity. The complex malin-laforin could develop additional functions related to the clearing of misfolded proteins through the ubiquitin–proteasome system [65]. In this sense, it has been observed that malin and laforin co-localize in the endoplasmic reticulum (ER) and form aggregates when neurons are treated with proteasomal inhibitors [66]. In the same direction, knockdown of laforin by siRNA treatment in HEK293 and SH-SY5Y cells increased the sensitivity to agents that triggers ER-stress, impairing the ubiquitin-proteasomal pathway and augmenting apoptosis [67]. Therefore, the regulation of glycogen synthesis by proteasomal degradation of key proteins required for this process could be a particular case in the general function that laforin and malin develop in the cell, which could be the control of the accumulation of misfolded proteins under stress conditions. It is well known that protein aggregates in neurons lead to several fatal neurodegenerative disorders. In addition to its function in protein degradation it has been proposed that laforin presents phosphatase activity against glycogen, dephosphorylation that is required for the maintenance of normal cellular glycogen [68]. Glycogen, a branched polymer of glucose, contains a small amount of covalently linked phosphate that can be release by Laforin in vitro. Moreover, glycogen from laforin-deficient mice is increased in several tissues, and this glycogen presents an increase in the covalent phosphate content [69]. These authors proposed that an increase in glycogen phophorylation leads to aberrant branching and Lafora body formation. Thus, the fisiological role of laforin would be to prevent excessive glycogen phosphorylation.


DUSP14 DUSP14 was initially identified as a CD28 interacting protein, called MKP6, in a yeast two-hybrid assay [70]. The potential substrates of DUSP14 can be ERK, JNK and p38, as GST-DUSP14 dephosphorylates them in vitro. Nevertheless, the expression of a catalytically inactive DUSP14 mutant, Cys-111 to Ser, in T-cells produces only an increase in ERK and JNK phosphorylation, whereas p38 phosphorylation remains invariable. This augmented phosphorylation in ERK and JNK is accompanied by an enhanced interleukin-2 production. DUSP14 seems to negatively regulate co-stimulatory signalling in T-cells, possibly by interacting directly with CD28. This would localize DUSP14 close to the cell membrane where the phosphatase could interact with the MAPKs activated early after stimulation. In fact, DUSP14 may function as an early response gene, since DUSP14 mRNA and protein expression are rapidly induced after co-stimulation of peripheral blood T lymphocytes through the TCR and CD28 receptors [48]. Apart of T cells, DUSP14 has been associated with different functions in other cell types. A recent study has suggested that DUSP14 could be the non-specific suppressor factor that regulates some types of hypersensitivity, such as delayed-type hypersensitivity or contact hypersensitivity [71]. There is also recent evidence that involves DUSP14 in the proliferation of pancreatic β-cells, possibly acting through modulation of ERK activity [72]. In this cell type, DUSP14 expression is augmented in response to Glucagon-like peptide-1 (GLP-1), which is a growth and differentiation factor for β-cells. However, GLP-1 has very limited proliferative skills. Thus, finding mechanisms that increase this proliferative ability could be of great interest in the treatment of diabetes. In this sense, blocking DUSP14 expression increases the proliferative effect of GLP-1 on β-cells. Another study, in a different context, has shown that DUSP14 expression is induced by the transcription factor Klf4 and by corticosteroids in the development of the epidermis in mice [73], suggesting a possible role of this phosphatase in the development of this tissue. This gene is also up-regulated in mice dorsal striatum by acute administration of MDMA (3-4-methylenedioxymethamphetamine), active compound of the widely abused drug ecstasy [74], along with two other DUSPs, DUSP1 and DUSP5. Induction of DUSP14 is partially due to activation of dopamine D1 receptors and independent of dopamine D2 receptors. Overall, it appears that DUSP14 is highly regulated at the transcriptional level in the cells studied. An interesting fact reported by Nakano [71] is the possible existence of DUSP14 as a dimer in non-reducing conditions, suggesting that this


phosphatase might be functional in cells as a dimer. This is in agreement with a recent work that reports the crystal structure of VH1, which adopts a dimeric quaternary structure mediated by an N-terminal α-helix that protrudes from the PTP domain [75]. STYX: An inactive A-DUSP STYX (phospho-Ser or Thr or Tyr interaction protein) is an inactive phosphatase that presents a Gly instead of the catalytic Cys in the P-loop of its phosphatase domain. The rest of the features of A-DUSPs are conserved in this protein. In fact, mutation of Gly to Cys recovers the phosphatase activity of this protein to dephosphorylate p-NPP and phospho-Tyr and phospho-Thr residues from peptide sequences of MAPK [76]. The PTP domain of this protein has been crystallized showing the same overall topology of the PTP fold [77]. Expression of STYX is found in all tissues studied with higher abundance in skeletal muscle, testis, and heart [76]. The generation of a knockout mouse that does not express this protein has shown that STYX is essential for normal spermiogenesis [78]. These mice present a great reduction in spermatozoa production with the result of male infertility. In addition, this study has described the interaction of STYX with an RNA-binding protein, CRHSP-24 (Ca2

+-regulated heat-stable protein of 24 kDa), which is phosphorylated in Ser and also targeted by the Ser/Thr phosphatase calcineurin (PP2B). It has been speculated that inactive phosphatases like STYX used the inactive domain to bind phosphorylated proteins in a similar way to SH2 domains and this way may regulate their function, for example, by protecting them from dephosphorylation by active phosphatases. In any case, as judging from the results obtain with the STYX knockout mouse, this inactive phosphatase does develop a key function in the cell that is required for spermiogenesis, although the mechanism is unknown. PIR1 and HCE (RNGTT): RNA phosphatases Two RNA phosphatases have been found in the human genome, PIR1 (Phosphatase that interacts with RNA/RNP complex 1)/DUSP11 [79, 80] and HCE1 [81]. Both possess an N-terminus phosphatase domain that contains the P-loop consensus motif for Cys-PTPs HCX2GX2R(S/T), although they lack the Asp required for the catalysis and present in most of the Cys-based PTPs. These two RNA phosphatases belong to different subgroups: HCE1 is included among the bifunctional metazoan mRNA capping enzymes, where the triphosphatase domain exclusively hydrolyzes the γ-phosphate of RNA; and PIR1 is included in monofunctional enzymes that hydrolize RNA 5’


triphosphate to 5’-diphosphate and 5’-monophosphate [82]. In this last group, it is also included the baculovirus phosphatase BVP, the first enzyme shown to have RNA phosphatase activity, isolated from the Autographa californica nuclear polyhedrosis virus [83, 84]. Bifuctional metazoan capping enzymes as HCE possess two domains, the previously mentioned N-terminal RNA 5'-triphosphatase domain and a C-terminal guanylyltransferase domain [82]. It is this second domain that is absent in monofuctional enzymes such as PIR1 and BVP. HCE1 RNA triphosphatase domain removes the γ-phosphate from the 5' end of nascent mRNA to leave a diphosphate terminus and the guanylyltransferase domain catalyzes the subsequent transfer of a guanylyl group from GTP to produce the unmethylated 5' cap structure [82]. This modification facilitates transcript translation and later processing of the mRNA. Then, PIR1/DUSP11 lacks the guanilyltransferase domain, and although shows greater activity for RNA than for proteins, its function is unclear. To our knowledge a function for PIR1 has not been established in eukaryotic cells and since PIR1 dephosphorylates RNA 5’ triphosphate to 5’-monophosphate, its function as a capping enzyme seems unlikely because it needs a 5’diphosphate in order to form the CAP structure [80]. Consistent with this, BVP, the other monofunctional domain RNA phosphatase, is not required for the formation of a 5’CAP structure on late viral mRNAs [85] and is not essential for viral replication [86]. Although its function is not fully understood, a recent report has shown that could be related to a change in the insect host behaviour that favours dispersion of the viruses by increasing the locomotory pattern during the larval stage [87]. In addition, A. californica nuclear polyhedrosis virus contains a proper bifunctional capping enzyme with RNA triphosphatase and guanylyltransferase activities, LEF-4. BVP and the mouse orthologue of HCE1 phosphatase (PDB 2C46) domains have been crystallized and both show the typical fold of the PTP domain of Class I phosphatases [88]. DUSP13: One gene, two phosphatases The gene DUSP13 is located in human chromosome 10q22.2, 50kb downstream of the gene of another A-DUSP, DUPD1, which suggests that the two genes might have originated by gene duplication. DUSP13 and DUPD1 have been also found in mammals and in chicken, then the gene duplication that gave rise to the DUPD1/DUSP13 pair occurred prior to the divergence of mammals from birds [11]. The mRNA transcribed from DUSP13 gene is translated in two different phosphatases using alternative open reading frames [89]. These two proteins are encoded from different exons, but their transcripts share the same polyadenylation signal in exon 9.


These phosphatases are TMDP (testis and skeletal muscle-specific dual specificity phosphatase) also called DUSP13b, and MDSP (muscle-restricted dual specificity phosphatase) or DUSP13a. The finding that a gene is transcribed into an mRNA that produces two proteins from alternative open reading frames that belong to the same family is unique among eukaryotes. Expression of these two proteins is quite restricted; while TMDP is mainly expressed in testis, MDSP is found in skeletal muscle. Expression of both proteins is increased with tissue maturation [89]. TMDP has been found by in situ hybridization in spermatocytes and round spermatids [90]; and this has been taken as an indication of the role of this phosphatase in spermatogenesis. TMDP dephosphorylates phospho-Tyr and phospho-Ser in artificial substrates as MBP (myelin basic protein) [90], but no physiological substrates have been identified yet. MDSP has been tested in co-transfection experiments for its ability to dephosphorylate several MAPkinases (ERK2, JNK1, p38α and β) producing a negative result [89]. Given the restricted expression of this phosphatase to the skeletal muscle, other kinases were tested, in particular dual specificity tyrosine-regulated kinases (DYRKs) because of the similarity of the activation loop with MAPK, in addition to the role of DYRK1B in skeletal muscle development [89]. However, MDSP was also unable to dephosphorylate this kinase in cotransfection experiments. The structure of the complete TMDP has been solved at a resolution of 2.4 Å [91]. Although the crystal showed an asymmetric unit with four molecules in two potential dimers, however, the dimeric contacts were not significant enough to hold the two molecules in solution. Furthermore, additional experiments such as gel filtration chromatography and dynamic light scattering showed that TMDP exists as a monomer in solution. TMDP structure is most similar to VHR. The main differences between the structures of these PTPs are found in the loop α1–β1, involved in substrate recognition, and in the loop β3–α4, part of so-called variable insert, which is shorter in TMDP. Unlike VHR, TMDP contains a wide and shallow pocket that seems to be adequate for accommodation of both phospho-Tyr and phospho-Thr and could explain its preference for dually phosphorylated substrates. DUSP19 and DUSP23: Putative scaffold A-DUSPs DUSP19 DUSP19 was initially described as SAPK (stress activated protein kinase) pathway-regulating phosphatase 1 (SKRP1) [92], LDP-2 (low-


molecular-mass DSP-2) [25] and LMW-DSP3 (low molecular weight dual specificity phosphatase 3) [93]. The mRNA of this phosphatase has been found in all tissues examined [25, 92]. DUSP19 sequence presents an Ala instead of a Ser after the Arg155 in the P-loop, HCXXGXXRS, which is conserved in most of the DUSPs. Endogenous DUSP19 is mainly detected in the cytoplasm, where it co-localizes with JNK. Recombinant DUSP19 protein showed extremely low phosphatase activity towards p-nitrophenyl phosphate (pNPP) and mutation of Ala-156 to a Ser increased this activity. However, DUSP19 and the Ala-156 Ser mutant exhibited similar activity to both phospho-Ser/Thr and phospho-Tyr residues of myelin basic protein. DUSP19 was found to be specific for JNK [25, 92]. This phosphatase showed a greater inhibition when JNK was activated by TNF-α (tumor necrosis factor alpha) or thapsigargin, in contrast to UV light or anisomycin [92]. DUSP19 formed a complex with the JNK upstream activating kinase MKK7, but not with MKK4, and did not bind directly to JNK. Thereby, DUSP19 targeted its substrate, JNK, through MKK7 interaction. In addition, it has been reported that DUSP19 co-precipitated MKK7-activating MAPKK kinase (MAPKKKs) ASK1 (apoptosis signal-regulating kinase 1), but not MEKK1 (MAP kinase kinase kinase 1) [94]. In agreement with this, DUSP19 overexpression increased ASK1-MKK7 complexes and enhanced the activation of MKK7 by ASK1. The authors of this work proposed that this phosphatase acts as a scaffold in the JNK activation pathway, although the mechanism remains unknown and deserves further study. DUSP23: The smallest A-DUSP At the end of 2004 several studies reported the initial characterization of this phosphatase as VHZ [95], LDP-3 (low-molecular-mass dual-specificity phosphatase-3) [24] and DUSP23 [96]. This A-DUSP is the smallest active phosphatase with 150 aminoacids (16 kDa), whereas the catalytic domain of classical PTPs presents 280 aa [3]. Expression of this phosphatase has been detected as mRNA in every tissue tested, in mouse [24] and human [95]. As a protein, DUSP23 has been detected in peripheral blood lymphocytes (PBL), and in different cell lines of hematopoietic origin: Jurkat T cells, RAMOS B cells, HL-60 myeloid cells as well as the tumor cell lines M45 and HT29 [95]. Endogenous DUSP23 is detected in the cytosol [95] and overexpression of this protein produced the same results [24, 95, 96], although in one study some cells presented nuclear staining, in some cases in the nucleolus [95]. The structure of VHZ/DUSP23 reported recently [97] is particularly interesting since two aminoacids from another molecule in the crystal, Thr-135–Tyr-136, bind in the catalytic pocket with a malate ion, mimicking the


TXY motif found in the activation loop of MAPK, in a similar way to an enzyme-substrate complex. The fold of VHZ/DUSP23 is closer to VHR, but binding of the substrate peptide to the active site resembles more closely to PTP1B due to the presence of similar residues surrounding the active pocket, which provide a similar orientation to the substrate. DUSP23 is an active phoshatase against pNPP and also dephosphorylates Tyr and Ser/Thr phosphorylated peptides [24, 96]. Its ability to dephosphorylate MAPK has been tested by different laboratories obtaining different results. In one study, recombinant GST-DUSP23 dephosphorylated ERK2 in vitro without affecting JNK or p38 [96]. In a different report, overexpression of DUSP23 produced an increase in JNK and p38 activity by sorbitol [24]. However, DUSP23 did not affect ERK activation by EGF. Osmotic stress stimuli other than sorbitol, such as mannitol, glucose or glycine, also increased the activation of the stress kinases JNK and p38, but other kinds of stresses, oxidative stress, heat shock, or UV irradiation, did not induce that effect. Interestingly, it was also observed a sorbitol-induced activation of upstream kinases of JNK and p38, MKK4 and MKK6. Surprisingly, the effect observed on JNK and p38 is independent of its phosphatase activity since Cys to Ser or Asp to Ala inactive mutants produced the same increase in sorbitol-induced activation of these kinases [24]. One explanation for these results would be that DUSP23 works as a scaffold in these pathways, although no interaction with proteins in this pathway has been reported. An alternative hypothesis would be that this PTP dephosphorylates a negative regulatory site on a protein acting on the initial steps of this signalling pathways. PTPMT1, DUSP18 and DUSP21: Mitochondrial atypical-DSPs PTPMT1 PTPMT1 (Protein-tyrosine phosphatase mitochondrial 1) is the first A-DUSP that has been demonstrated to be localized in the mitochondria [98]. Afterwards, another two DUSPs were also found in the mitochondria, DUSP18 and DUSP21 [99]. PTPMT1 was referred previously as PLIP (PTEN-like phosphatase) [100] and MOSP [2]. Initially this PTP was cloned from Dictiostelium after a database search for phophatases similar to PTEN [101]. In fact, it presents two basic aminoacids in the P-loop, Arg-133 and Lys-136, in equivalent positions to PTEN Lys-125 and Lys-128, aminoacids that give PTEN specificity for PI(3,4,5)P3. Initial characterization of the substrate specificity of this phosphatase using several phosphoinositides


showed as the preferred substrate PI(5)P [100]. PTPMT1 is targeted to the mitochondria by an N-terminal signal sequence and it is found anchored to the matrix face of the inner membrane (IM) in this organelle, where it colocalizes with members of the respiratory chain. Expression of this phosphatase, although is highly enriched in testis tissue, has been shown in every tissue tested by PCR. This PTP presents orthologues in all phylogenetic kingdoms, including eubacteria. Disruption of PTPMT1 expression by siRNA in the pancreatic β cell line INS-1 832/13 caused an increase in ATP production and markedly enhanced insulin secretion under both basal- and glucose-stimulated conditions [98]. At the same time, knockdown of this phosphatase produced an increase in Ser and Thr phosphorylation of some proteins, whereas no change was observed in protein tyrosine phosphorylation. However, a reduced expression of PTPMT1 did not affect the PI(5)P levels in the mitochondria , despite the in vitro observed activity of PTPMT1 [100]. These data support the idea of PTPMT1 playing a critical role in mitochondrial function by regulating the status of phosphorylation in Ser or Thr of mitochondrial proteins. DUSP18 This PTP was characterized by two different groups as LMW-DSP20 [102] and DUSP18 [103]. This phosphatase is conserved in the different groups of vertebrates, is expressed in all tissues studied, and is localized in the mitochondria [99]. Unlike, PTPMT1 that presents an N-terminal mitochondrial localization signal, DUSP18 contains an internal mitochondrial localization signal that allows its internalization in the mitochondria. Inside the mitochondria, DUSP18 is localized in the IM, looking to the IMS (intermembrane space) compartment. Although it has been shown to be present in the cytosol and the nucleus, these results could be likely due to overexpression of these PTPs, in addition to the masking effect of N-terminal tags over the mitochondrial localization signals [102]. Kinetic analysis of DUSP18 with pNPP as a substrate showed that DUSP18 was an active PTP and was more efficient using pNPP than the other mitochondrial DUSPs, PTPMT1 and DUSP21, but not as efficiently as VHR [99]. DUSP18 dephosphorylates in vitro diphosphorylated synthetic MAPK peptides, with preference for the phospho-Tyr and diphosphorylated forms over phospho-Thr. However, in vivo, overexpression of DUSP18 was unable to dephosphorylate MAPKs in this study [99]. In contrast, there is other report showing that DUSP18 targets JNK in vivo [104]. However, in a different report has not been possible to detect the presence of JNK in mitochondria [99], suggesting that JNK is not a substrate for DUSP18.


Induction of apoptosis, a process in which mitochondria plays a critical role, resulted in release of DUSP18 from the mitochondria to the cytosol, in a similar way to cytochrome c, a key mediator of apoptosis. Treatment of cells with known inducers of apoptosis, such as starurosporine, an inespecific kinase inhibitor, or etoposide, a topoisomerase II inhibitor, showed cytosolic localization of DUSP18 after the mitochondrial outer membrane permeabilization produced during the apoptotic process [99]. DUSP21 Expression of DUSP21 mRNA has been detected exclusively in testes [102]. DUSP21 presents 69% sequence identity with DUSP18, and, as DUSP18, it also contains an internal mitochondrial signal sequence that, in this case, localizes the phosphatase to the matrix compartment of this organelle. Interestingly, despite this great similarity, these two DUSPs localize to different sides of the inner mitochondrial membrane. DUSP21 dephosphorylates synthetic MAPK peptides in vitro with preference for phospho-Tyr or dual-phosphorylated peptides over phospho-Thr. Co-transfection of this PTP with MAPK resulted in no activity against this family of kinases [102]. DUSP15 and DUSP22: Myristoilated A-DUSPs DUSP15 DUSP15 was first described as VHY (VH1-like member Y) because its closest relative was VHX (VH1-like member X), currently known as DUSP22 [105]. DUSP15 has a very narrow tissue expression pattern. Thus, its mRNA was detected only in testes and very weakly in brain, spinal cord and thyroid. Protein was also detected in testes by Western Blot and histochemistry showed the presence of this phosphatase in spermatocytes [105]. The N-terminus of DUSP15 presents high similarity to the consensus myristoylation sequence of SFK (Src family kinases) and, as well as them, DUSP15 is post-transcriptionally modified by the addition of myristic acid to Gly in position 2, as well as DUSP22 [105]. This myristoylation is essential for DUSP15 recruitment to the plasma membrane. DUSP15 is an active phosphatase as it can dephosphorylate pNPP, but there is no information about its physiological substrates. The crystal structure of DUSP15 has been reported and shows several particular surface properties, which indicate that DUSP15 may have a unique substrate specificity [106].


DUSP22 DUSP22 presents a broad expression and was characterized under four names by different groups: LMW-DSP2, VHX, JSP-1 (JNK Stimulatory Phosphatase-1) and JKAP (JNK pathway-associated phosphatase). Results obtained in these initial reports were also diverse in relation with MAPK activation. Two reports showed that DUSP22 activated JNK [107, 108], other showed inhibition of ERK2 and NF-AT by this phosphatase [109]; and the fourth showed that DUSP22 bound to and inhibited p38 [110]. Chen et al. showed that targeted gene disruption of DUSP22 in embryonic stem cells abolished JNK activation by TNF-α and TGF-β, but not by ultraviolet-C irradiation [107]. DUSP22, in this report, was also shown to interact in vivo with JNK and MKK7, but not with SEK1. However, in vitro, JNK did no bind to DUSP22, suggesting that targeting of JNK is mediated by binding to other protein. Other group showed that DUSP22 acts as a positive regulator of JNK pathway, and its overexpression led to the activation of MKK4 [108], another upstream MAPKK. Altogether, the data presented by these two papers suggests that DUSP22 would exert its effect on JNK in an indirect manner, acting on a protein upstream of this kinase. The fact that DUSP22 is myristilated in the N-terminus to target it to the plasma membrane [105] should be taken into consideration when interpreting the previous data, which has been obtained mainly with overexpressed N-terminal tagged phosphatase. In addition to its role in MAPK signalling pathways, DUSP22 has also been implicated in the regulation of the ERα (estrogen receptor-alpha) and STAT3. DUSP22 regulates ERα-mediated transcriptional activation by dephosphorylation of this receptor on Ser-118 [111]. The function of DUSP22 on STAT3 dephosphorylation has been investigated in the context of IL-6 stimulation [112]. Interaction between these two proteins has been observed in vivo. Despite what this data might suggest, disruption of DUSP22 in mice to generate knockout animals does not produce any observable phenotype in these animals. The structure of this phosphatase has been solved for the peptide that contains the phosphatase domain, aminoacids 1-163 [113]. The topology resembles that of other DUSPs, but with a shallower catalytic pocket, 4.5 Å of depth, smaller than that of VHR, 6 Å, and half the pocket of PTP1b, 10 Å. DUPD1 and DUSP27: A little of confusion DUPD1 (DUSP and pro-isomerase domain-containing 1) was initially called DUSP27 [2], but in the HUGO (Human Genome Organisation) database appears as DUPD1 with DUSP27 as alias. The name DUSP27 is


used to refer to another open reading frame that codes a phosphatase with 1158 amino acids that corresponds to Swiss-Prot accession number Q5VZP5, which is an inactive phosphatase that has been detected as mRNA but no report has been published yet about this PTP. First attempts to characterized DUPD1 (Swiss-Prot accession number Q68J44) predicted an open reading frame that contained a dual specific phosphatase and a pro-isomerase domains [2]. Subsequent studies showed that the pro-isomerase domain was not included in the sequence [11]. DUPD1 is able to dephosphorylate synthetic phopho-Ser and phopho-Thr peptides but it acts more efficiently on phospho-Tyr peptides. The catalytic groove of DUSP27 can accommodate a dual-phosphorylated substrate where the two phosphorylated residues are separated by two amino acids instead of one as in the case of MAPKs. This finding indicates that DUPD1 substrates are not MAPKs. Although further studies are needed to identify DUPD1 substrates, its specific tissue distribution, restricted to adult fat, skeletal muscle and liver [11], suggest that DUSP27 may participate in the regulation of energy metabolism. Conclusions and futures perspectives A-DUSPs are a heterogeneous group of enzymes according to the range of substrates targeted. The controversial results obtained regarding substrates and the rich synonimia used to name these proteins have produced some confusion over this group of proteins. Nomenclature has been fixed, adopting the name DUSP followed by a number, and new substrates identified lately for these phosphatases have widen the repertoire of this family outside the MAPK world. The search for substrates has been done under the influence of sequence similarity between these proteins and MKPs. Therefore, most of the studies tried to show the effect of these phosphatases on MAPK. As we have shown along this review, results were negative in many cases and controversial in most of the rest. Nevertheless, some A-DUSPs have been shown to be true MKPs, for example VHR. Even in these cases, the absence of a targeting domain similar to the rhodanese in MKPs raises the question about the mechanism by which these phosphatases target MAPKs. In this sense, in at least two cases, VHR and DUSP26, two proteins have been found that may work as adaptors that facilitate phosphatase access to their substrates, VRK3 [28] and Hsf4 [34], respectively. At the same time, it has become evident that other substrates could exist for these phosphatases, as, for example, STAT transcription factors. Although the information about these phosphatases is still scarce and the physiological role for most of them is unknown, some studies have suggested the implication of these phosphatases in some cellular functions, such as


apoptosis, DUSP12, or cell-cycle regulation, VHR; and in some pathologies, such as cancer, VHR and DUSP26, or diabetes, DUSP12. In this sense, the lack of knockout mice for A-DUSPs has been an obstacle in the determination of the physiological role of these phsophatases. So far, out of 19 A-DUSPs, only 3 have been disrupted in mice, STYX [78], DUSP22 [107] and laforin [114]. STYX deficient mice showed that this phosphatase is involved in spermatogenesis. It seems likely that the next few years will see a critical advance in our understanding of A-DSPs. This will be thanks to approaches that manipulate gene expression, such as siRNA interference or generation of knockout mice, as well as proteomics and mass spectrometry technologies that will provide new data to dissect the substrate repertoire of A-DUSPs and the physiological function of these phosphatases. References 1. Manning, G., et al., 2002, Science,298, 2. Alonso, A., et al., 2004, Cell,117, 3. Andersen, J.N., et al., 2001, Mol Cell Biol,21, 4. Jemc, J. and I. Rebay, 2007, Annu Rev Biochem,76, 5. Krishnan, N., et al., 2009, J Biol Chem, 6. Gohla, A., J. Birkenfeld, and G.M. Bokoch, 2005, Nat Cell Biol,7, 7. Suh, M.H., et al., 2005, Proc Natl Acad Sci U S A,102, 8. Kim, Y., et al., 2007, Proc Natl Acad Sci U S A,104, 9. Guan, K.L., S.S. Broyles, and J.E. Dixon, 1991, Nature,350, 10. Ishibashi, T., et al., 1992, Proc Natl Acad Sci U S A,89, 11. Friedberg, I., et al., 2007, FEBS Lett,581, 12. Zhou, G., et al., 1994, J Biol Chem,269, 13. Denu, J.M., et al., 1996, Proc Natl Acad Sci U S A,93, 14. Hengge, A.C., J.M. Denu, and J.E. Dixon, 1996, Biochemistry,35, 15. Lyon, M.A., et al., 2002, Nat Rev Drug Discov,1, 16. Rahmouni, S., et al., 2006, Nat Cell Biol,8, 17. Alonso, A., et al., 2003, Nat Immunol,4, 18. Hoyt, R., et al., 2007, J Immunol,179, 19. Yuvaniyama, J., et al., 1996, Science,272, 20. Barford, D., A.J. Flint, and N.K. Tonks, 1994, Science,263, 21. Alonso, A., et al., 2001, J Biol Chem,276, 22. Todd, J.L., K.G. Tanner, and J.M. Denu, 1999, J Biol Chem,274, 23. Todd, J.L., et al., 2002, Oncogene,21, 24. Takagaki, K., et al., 2004, Biochem J,383, 25. Nakamura, K., et al., 2002, J Biochem,132, 26. Denu, J.M., et al., 1995, J Biol Chem,270, 27. Zhou, B., et al., 2002, J Biol Chem,277,


28. Kang, T.H. and K.T. Kim, 2006, Nat Cell Biol,8, 29. Arnoldussen, Y.J., et al., 2008, Cancer Res,68, 30. Hao, L. and W.M. ElShamy, 2007, Int J Cancer,121, 31. Henkens, R., et al., 2008, BMC Cancer,8, 32. Vasudevan, S.A., et al., 2005, Biochem Biophys Res Commun,330, 33. Yu, W., et al., 2007, Oncogene,26, 34. Hu, Y. and N.F. Mivechi, 2006, Mol Cell Biol,26, 35. Takagaki, K., et al., 2007, Mol Cell Biochem,296, 36. Wang, J.Y., et al., 2006, J Neurochem,98, 37. Wang, J.Y., et al., 2008, J Neurochem,107, 38. Tanuma, N., et al., 2009, Oncogene,28, 39. Muda, M., et al., 1999, J Biol Chem,274, 40. Guan, K., et al., 1992, Proc Natl Acad Sci U S A,89, 41. Park, H.D., et al., 1996, Yeast,12, 42. Munoz-Alonso, M.J., et al., 2000, J Biol Chem,275, 43. Elbein, S.C., et al., 1999, Diabetes,48, 44. Das, S.K., et al., 2006, Diabetes,55, 45. Kresse, S.H., et al., 2005, Mol Cancer,4, 46. Sharda, P.R., et al., 2009, Biochem J,418, 47. MacKeigan, J.P., L.O. Murphy, and J. Blenis, 2005, Nat Cell Biol,7, 48. Abraham, R., et al., 2005, J Biol Chem,280, 49. Serratosa, J.M., et al., 1999, Hum Mol Genet,8, 50. Minassian, B.A., et al., 1998, Nat Genet,20, 51. Lafora, G.R., 1911, Virchows Arch. f. Path. Anat.,205, 52. Gomez-Abad, C., et al., 2005, Neurology,64, 53. Chan, E.M., et al., 2003, Nat Genet,35, 54. Gentry, M.S., C.A. Worby, and J.E. Dixon, 2005, Proc Natl Acad Sci U S A,102, 55. Lohi, H., et al., 2005, Hum Mol Genet,14, 56. Fernandez-Sanchez, M.E., et al., 2003, Hum Mol Genet,12, 57. Printen, J.A., M.J. Brady, and A.R. Saltiel, 1997, Science,275, 58. Vilchez, D., et al., 2007, Nat Neurosci,10, 59. Solaz-Fuster, M.C., et al., 2008, Hum Mol Genet,17, 60. Cheng, A., et al., 2007, Genes Dev,21, 61. Vernia, S., et al., 2009, J Biol Chem,284, 62. Ianzano, L., et al., 2003, Genomics,81, 63. Ganesh, S., et al., 2003, Hum Mol Genet,12, 64. Liu, Y., et al., 2006, J Biol Chem,281, 65. Garyali, P., et al., 2009, Hum Mol Genet,18, 66. Mittal, S., et al., 2007, Hum Mol Genet,16, 67. Vernia, S., et al., 2009, PLoS One,4, 68. Worby, C.A., M.S. Gentry, and J.E. Dixon, 2006, J Biol Chem,281, 69. Tagliabracci, V.S., et al., 2007, Proc Natl Acad Sci U S A,104, 70. Marti, F., et al., 2001, J Immunol,166, 71. Nakano, Y., 2007, Br J Dermatol,156, 72. Klinger, S., et al., 2008, Diabetes,57,


73. Patel, S., et al., 2006, Proc Natl Acad Sci U S A,103, 74. Marie-Claire, C., et al., 2008, Brain Res, 75. Koksal, A.C., J.D. Nardozzi, and G. Cingolani, 2009, J Biol Chem,284, 76. Wishart, M.J., et al., 1995, J Biol Chem,270, 77. Almo, S.C., et al., 2007, J Struct Funct Genomics,8, 78. Wishart, M.J. and J.E. Dixon, 2002, Proc Natl Acad Sci U S A,99, 79. Yuan, Y., D.M. Li, and H. Sun, 1998, J Biol Chem,273, 80. Deshpande, T., et al., 1999, J Biol Chem,274, 81. Yamada-Okabe, T., et al., 1998, Nucleic Acids Res,26, 82. Shuman, S., 2001, Cold Spring Harb Symp Quant Biol,66, 83. Takagi, T., et al., 1998, Proc Natl Acad Sci U S A,95, 84. Gross, C.H. and S. Shuman, 1998, J Virol,72, 85. Li, Y. and L.A. Guarino, 2008, J Virol,82, 86. Li, Y. and L.K. Miller, 1995, J Virol,69, 87. Kamita, S.G., et al., 2005, Proc Natl Acad Sci U S A,102, 88. Changela, A., et al., 2005, J Biol Chem,280, 89. Chen, H.H., et al., 2004, J Biol Chem,279, 90. Nakamura, K., et al., 1999, Biochem J,344 Pt 3, 91. Kim, S.J., et al., 2007, Proteins,66, 92. Zama, T., et al., 2002, J Biol Chem,277, 93. Cheng, H., et al., 2003, Int J Biochem Cell Biol,35, 94. Zama, T., et al., 2002, J Biol Chem,277, 95. Alonso, A., et al., 2004, J Biol Chem,279, 96. Wu, Q., et al., 2004, Int J Biochem Cell Biol,36, 97. Agarwal, R., S.K. Burley, and S. Swaminathan, 2008, J Biol Chem,283, 98. Pagliarini, D.J., et al., 2005, Mol Cell,19, 99. Rardin, M.J., et al., 2008, J Biol Chem,283, 100. Pagliarini, D.J., C.A. Worby, and J.E. Dixon, 2004, J Biol Chem,279, 101. Merlot, S., et al., 2003, J Biol Chem,278, 102. Hood, K.L., J.F. Tobin, and C. Yoon, 2002, Biochem Biophys Res Commun,298, 103. Wu, Q., et al., 2003, Biochim Biophys Acta,1625, 104. Wu, Q., et al., 2006, Front Biosci,11, 105. Alonso, A., et al., 2004, J Biol Chem,279, 106. Yoon, T.S., et al., 2005, Proteins,61, 107. Chen, A.J., et al., 2002, J Biol Chem,277, 108. Shen, Y., et al., 2001, Proc Natl Acad Sci U S A,98, 109. Alonso, A., et al., 2002, J Biol Chem,277, 110. Aoyama, K., et al., 2001, J Biol Chem,276, 111. Sekine, Y., et al., 2007, Oncogene,26, 112. Sekine, Y., et al., 2006, Oncogene,25, 113. Yokota, T., et al., 2007, Proteins,66, 114. Ganesh, S., et al., 2002, Hum Mol Genet,11,

emerging signaling pathways in tumor biology 2010

Documents

map kinase kinase kinase

lazo contents chapter

protein kinases

specific pathways

known kinases

susana alemany chapter

peter jordan chapter

wei dai chapter