disruption of protein kinase a regulation causes ...disruption of protein kinase a regulation causes...

Disruption of Protein Kinase A Regulation Causes Immortalization

and Dysregulation of D-Type Cyclins

Kiran S. Nadella1and Lawrence S. Kirschner

1,2

1Human Cancer Genetics Program and 2Division of Endocrinology, Diabetes and Metabolism, Department of Internal Medicine,The Ohio State University, Columbus, Ohio

Abstract

Phosphorylation is a key event in cell cycle control, anddysregulation of this process is observed in many tumors,including those associated with specific inherited neoplasiasyndromes. We have shown previously that patients with theautosomal dominant tumor predisposition Carney complexcarry inactivating mutations in the PRKAR1A gene, whichencodes the type 1A regulatory subunit of protein kinase A(PKA), the cyclic AMP–dependent protein kinase. This defectwas associated with dysregulation of PKA signaling, andgenetic analysis has suggested that complete loss of the genemay be required for tumorigenesis. To determine the mecha-nism by which dysregulation of PKA causes tumor formation,we generated in vitro primary mouse cells lacking the Prkar1aprotein. We report that this genetic disruption of PKA regu-lation causes constitutive PKA activation and immortalizationof primary mouse embryonic fibroblasts (MEFs). At themolecular level, knockout of Prkar1a leads to up-regulationof D-type cyclins, and this increase occurs independently ofother pathways known to increase cyclin D levels. Despitethe immortalized phenotype, known mediators of cellularsenescence (e.g., p53 and p19ARF) seem to remain intact inPrkar1a�/� MEFs. Mechanistically, cyclin D1 mRNA levels arenot altered in the knockout cells, but protein half-life ismarkedly increased. Using this model, we provide the firstdirect genetic evidence that dysregulation of PKA promotesimportant steps in tumorigenesis, and that cyclin D1 is anessential target of PKA. (Cancer Res 2005; 65(22): 10307-15)

Introduction

Carney complex is an autosomal dominant inherited neoplasiasyndrome characterized by spotty skin pigmentation, myxomas,endocrine tumors, and schwannomas (1). We and others have shownthat this disease is caused by inactivating mutations in the PRKAR1Agene, encoding the type 1A regulatory subunit of the cyclic AMP(cAMP)–dependent protein kinase, protein kinase A (PKA; refs. 2–4).Furthermore, we have shown that mutation of this ubiquitouslyexpressed regulatory subunit causes increased PKA activity intumors from patients with Carney complex (2). PRKAR1A (R1A)mutations in Carney complex tumors induce nonsense-mediatedmRNA decay resulting from premature termination codons (3).Partial or complete loss of heterozygosity of PRKAR1A is seen inCarney complex–associated tumors (2) as well as in sporadic tumors

of the thyroid and adrenal glands (5, 6), indicating that it is a tissue-specific tumor suppressor gene.Because patient samples are rare, we have generated a mouse

model to study this condition. Mice homozygous for a null allele ofPrkar1a exhibit embryonic lethality, whereas heterozygotes, liketheir human counterparts, are tumor prone (7–9). To determine themechanism by which loss of Prkar1a promotes tumors, we havechosen to use primary cultures of mouse embryonic fibroblasts(MEFs), which lack secondary alterations commonly found inimmortalized cell lines. Using Cre-lox technology, we generatedPrkar1a�/� MEFs in vitro and compared them with otherwisegenetically identical cells. We report that removal of Prkar1a fromcells leads to immortalization in the absence of transformation. Inprobing the molecular basis for this observation, we find that thereis strong up-regulation of cyclin D1, which is likely linked to theimmortalized phenotype. This increase in cyclin D1 protein levelsoccurs independently of other known pathways, suggesting thatPKA itself functions as a regulator of cyclin D1 protein levels.

Materials and Methods

Generation of Prkar1a�/� mouse embryonic fibroblasts. Micecarrying a conditional null allele of Prkar1a (8) were crossed, and primary

MEFs were prepared from embryonic day 13.5 embryos. Each MEF line was

derived from an individual embryo and PCR genotyped as described (8).Cells were cultured in DMEM with 1% penicillin/streptomycin (Invitrogen,

Carlsbad, CA) containing 14% fetal bovine serum (Hyclone, South Logan,

UT). All the experiments presented here were done in two to three

independent MEF lines prepared from different litters. Mice used forpreparation of MEFs were maintained in accordance with the highest

standards of animal care in accordance with Institutional Laboratory

Animal Care and Use Committee guidelines.

Retroviral infections. High titer retroviruses were produced bytransient transfection of retroviral constructs using Phoenix-Eco packaging

cells (10) by calcium phosphate precipitation (ProFection calcium

phosphate kit, Promega, Madison, WI). Passage 2 MEFs were infected withvirus and selected for 4 days in the presence of either puromycin (2.5 Ag/mL) or hygromycin (400 Ag/mL). The following previously described

plasmids were used: pBABE-puro, pBABE-puro-Cre, pBABE-hygro (11), and

pBABE-puro-cyclin D1 (12). pBABE-hygro-hemagglutinin (HA)-R1A wasprepared by subcloning the HA-tagged human PRKAR1A cDNA from

pREP4-HA-R1A (13) into the pBABE-hygro vector.

Cell proliferation and synchronization. The 3T3 assay was done

essentially as described (14). To assess serum-free growth, 300,000 cellswere plated in 100-mm dishes containing complete medium. After 24

hours, cells were trypsinized and counted. The medium in the remaining

plates was replaced by serum-free DMEM, and cells were counted every24 hours. For cell cycle studies, subconfluent MEFs were synchronized

by incubation in serum-free DMEM for 72 hours and then stimulated

to proliferate by addition of DMEM containing 14% serum. Cells were

harvested at the indicated times for Western blotting.Western analyses.MEFs were lysed in M-PER protein extraction reagent

(Pierce, Rockford, IL) unless otherwise stated. Proteins were resolved in

SDS-PAGE gels and transferred to nitrocellulose (Pall, East Hills, NY), and

Requests for reprints: Lawrence S. Kirschner, Human Cancer Genetics Program,Ohio State University, 544 TMRF, 420 West 12th Avenue, Columbus, OH 43210. Phone:614-292-1190; Fax: 614-688-4006; E-mail: [email protected].

I2005 American Association for Cancer Research.doi:10.1158/0008-5472.CAN-05-3183

www.aacrjournals.org 10307 Cancer Res 2005; 65: (22). November 15, 2005

Research Article

Research. on August 28, 2021. © 2005 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

the blots were developed with Western lightning reagents (Perkin-Elmer,Boston, MA). The antibodies used in this study were as follows: PKA-R1A,

PKA-R2A, PKA-R2B, PKA-CA, and RB (554136) were from BD Biosciences

(San Jose, CA). Cyclin D1 (SC-718), cyclin A (SC-596), cyclin E (SC-481), cyclin

B1 (SC-245), proliferating cell nuclear antigen (SC-56), p16 (SC-1207), p19(SC-1063), p21Cip1 (SC-471), cyclin-dependent kinase (CDK) 4 (SC-601), CDK2

(SC-163), and E2F1 (SC-251) were from Santa Cruz Biotechnology (Santa

Cruz, CA). p27Kip1 (2552), phosphorylated extracellular signal-regulated

kinase (ERK) p44/p42 (9101), ERK p44/p42 (9102), phosphorylated Creb(9196), phosphorylated glycogen synthase kinase 3h (GSK3h; 9336),

phosphorylated p38 (9216), and p38 (9212) were from Cell Signaling

Technologies (Beverly, MA). Monoclonal anti-HA (MMS-101P) was from

Covance (Princeton, NJ). Actin (A5060) was from Sigma (St. Louis, MO).PTEN monoclonal antibody was a kind gift from Dr. C. Eng (The Ohio State

University, Columbus, OH), and cyclin D3 monoclonal antibody was a gift

from Dr. D. Guttridge (The Ohio State University, Columbus, OH).Quantitative real-time PCR analyses. mRNA was isolated from

Prkar1a+/+ and Prkar1a�/� MEFs by Trizol reagent (Invitrogen) and

repurified on a RNeasy column (Qiagen, Valencia, CA). RNA (1 Ag) was

reverse transcribed using iScript cDNA synthesis kit (Bio-Rad, Hercules,

CA). Quantitative real-time PCR (qRT-PCR) was done using an iCycler and

iCycler SYBR Green reagents (Bio-Rad). Cyclin D1 primers were forward:

ATGTGAAGTTCATTTCCAACCC and reverse: TTGACTCCAGAAGGGC-

TTCA. mRNA fold changes were calculated by the DDCt method using

Gapdh as a standard. All PCR reactions were done in triplicate, and each

analysis was representative of two Prkar1a�/� and Prkar1a+/+ cell lines.

Immunofluorescence studies. MEFs were grown on glass coverslips,

fixed in 4% paraformaldehyde, and permeabilized using 0.2% Triton X-100

for 6 minutes. After blocking (2% bovine serum albumin/PBS), slides were

incubated at 4jC overnight with the primary antibody, washed thrice in

PBS, and incubated with FITC-conjugated secondary antibodies (1 hour

at room temperature). Cells were mounted using Vectashield containing

4V,6-diamidino-2-phenylindole (DAPI; Vector Laboratories, Burlingame,

CA) and imaged using a LSM510 confocal microscope (Carl Zeiss, Inc.,

Thornwood, NY).Kinase assays. CDK2 kinase assays were done as described (15) using

histone H1 as a substrate. CDK4 kinase assays were done as described (16)

using glutathione S-transferase (GST)-Rb (791-928) as a substrate. For PKAassays, MEFs were washed in ice-cold PBS and lysed in buffer containing 10

mmol/L Tris-HCl (pH 7.1), 1 mmol/L EDTA, 1 mmol/L DTT, 0.5% NP40, 0.5

mmol/L phenylmethylsulfonyl fluoride, 0.1 mmol/L vanadate, 1 mmol/L

NaF, and HALT protease inhibitors (Pierce). Cell lysates were centrifuged10,000 � g for 20 minutes at 4jC and the supernatants were collected for

measurement of free (�cAMP) and total (+cAMP) PKA activity. Reactions

were done in 30 AL containing 30 Ag protein lysate, 5 Ag of 6� His-Creb,

50 mmol/L ATP, 50 mmol/L Tris, 10 mmol/L MgCl2, and 1 mmol/L DTTwith and without 5 Amol/L PKA inhibitor and cAMP. The reaction was

incubated 30 minutes at 30jC and terminated by boiling in sample buffer.

Proteins were subject to Western blot with a phosphorylated Creb antibodyand bands were quantitated with J image software.

Phosphatase treatment. Whole-cell lysates from knockout MEFs were

made in radioimmunoprecipitation assay buffer containing 50 mmol/L

HEPES, 150 mmol/L NaCl, 0.5 mmol/L EDTA, 2.5 mmol/L EGTA, 1 mmol/LDTT, 0.1% Tween, and 10% glycerol with protease inhibitors with and

without phosphatase inhibitors (0.1 mmol/L sodium vanadate and 1 mmol/L

NaF). Lysates (50 Ag) without phosphatase inhibitors were treated with Ephosphatase according to the manufacturer’s instructions (New EnglandBiolabs, Beverly, MA). Protein samples were subjected to SDS-PAGE and

immunoblotted for cyclin D1 as above.

Reporter assays. MEFs were plated in triplicate in six-well dishes andcotransfected using Superfect (Invitrogen) with a TK-Renilla luciferase

plasmid and a plasmid containing firefly luciferase driven by a promoter

carrying wild-type (WT) or a mutant (inactive) p53 response elements.

After 36 hours, firefly luciferase activity was measured using the dual-luciferase reporter assay system (Promega) and normalized to Renilla

luciferase activity. Ras activity was measured using the Ras activation assay

kit (Upstate, Waltham, MA) according to the manufacturer’s instructions.

Drug treatments. Asynchronous cells were treated with 5 Amol/L H-89(Upstate) or 50 Amol/L forskolin (Sigma) for 12 hours. For half-life studies,

MEFs were treated with cycloheximide for the indicated times and whole-

cell lysates were made and then immunoblotted with cyclin D1 antibodies.

For analysis of cyclin D1 degradation, asynchronous MEFs were treatedwith MG132 (30 Amol/L; Sigma) or calpain inhibitor 1 (100 Amol/L; Sigma)

for the indicated times, and cells were harvested for preparation of whole-

cell lysates. Western blots of the lysates were probed with the cyclin D1

antibody and quantitated as above.Transient transfections. For transfections, 293T cells were seeded into

100-mm Petri dishes 1 day previously and transfected with either

constitutively active PKA-C or dominant-negative PKA plasmids using

Superfect (Qiagen) according to manufacturer’s recommendations. After 24hours of transfection, cells were stimulated with zinc sulfate (70 Amol/L)

for further 24 hours to induce the expression of R1A after which cells were

harvested for protein preparation. All transfections were done a minimumof three times before calculating means and SDs.

Treatment with DNA-damaging agents. MEFs were expanded and

plated onto 100-mm plates and grew them overnight to a confluency of

60% to 70%. The cells were washed with PBS and either exposed to UV-Cradiation at 40 J/m2 or treated with 3.5 Amol/L doxorubicin for 8 hours.

MEFs were harvested after 8 hours of treatment for protein extraction.

Results

Generation and characterization of Prkar1a�/�, Prkar1a+/�,and Prkar1a+/+ mouse embryonic fibroblasts. To explore themolecular basis for Prkar1a’s role in tumorigenesis, mice carryinga conditional null allele were bred and MEFs were prepared.Prkar1a�/�, Prkar1a+/�, and Prkar1a+/+ cells were obtained inculture by infecting Prkar1a loxP/loxP or Prkar1a loxP/+ cells with aretrovirus carrying Cre recombinase or a control virus lackingCre. Genotyping and Western blotting confirmed the loss ofPrkar1a coding sequences and protein (Fig. 1A and B). AlthoughPCR was able to detect the intact conditional allele in Cre-treatedcells at passage 5, there was no detectable Prkar1a protein byWestern blotting. The cells exhibited a null genotype by passage10. As expected, control virus-treated cells retained the intactconditional allele and the Prkar1a protein throughout the study,whereas Prkar1a+/� cells had intermediate Prkar1a protein levels.Interestingly, the PKA catalytic subunit was increased 2.5-fold inPrkar1a�/� cells, whereas the other regulatory subunits remainedunchanged (Fig. 1B). To address the functional effects of thesealterations, we measured total and free PKA activity in the cellsusing a novel nonradioactive PKA assay (see Materials andMethods). In agreement with the published data from Carneycomplex tumors (2), we observed an increased total and free PKAactivity in Prkar1a�/� cells compared with heterozygote and theWT counterparts (Fig. 1C).The effect of loss of Prkar1a on cell proliferation was assessed

by 3T3 assay (Fig. 1D). Although WT MEFs underwent typicalsenescence after 10 to 12 generations, Prkar1a�/� cells proliferatedunabatedly and continued to do so for >50 passages. Similar resultswere obtained in more than six independently replicated experi-ments. Furthermore, early-passage Prkar1a�/� MEFs were able toproliferate in serum deprived conditions, whereas control cellsarrested within 24 hours of serum removal (Fig. 1E). Despite theirimmortality, Prkar1a�/� MEFs were not transformed, as theyretained contact inhibition, did not exhibit anchorage-independentgrowth, and did not form tumors in scid mice (data not shown).Loss of Prkar1a is associated with up-regulation of cyclin

D1. Owing to the functional association of cyclins with cellproliferation and cell cycle regulation, we evaluated cyclin levels in

Cancer Research

Cancer Res 2005; 65: (22). November 15, 2005 10308 www.aacrjournals.org



asynchronous populations of cells. Unexpectedly, we observed astriking (3.5-fold) up-regulation of cyclin D1 in Prkar1a�/� MEFscompared with the heterozygous and WT cells (Fig. 2A). Thiselevation in cyclin D1 was observed at early passages (passage<10), at a time when there was no significant difference in thegrowth rate between knockout and WT cells. Confocal microscopyconfirmed the presence of increased cyclin D1 and showed that theprotein has increased nuclear retention (Fig. 2B). Consistent withthese observations, tumors obtained from Prkar1a+/� miceshowed increased nuclear localization of cyclin D1 by immuno-histochemistry (Fig. 2C). Other cyclins were not significantlyaltered in the knockout cells either at the mRNA level (data notshown) or at the protein level (Fig. 2A). In concert with theelevated cyclin D1 levels, an intact RB pathway was demonstrableby normal expression and phosphorylation of RB and normal S-phase reentry kinetics after serum starvation and release as shownby Western blotting (see below).To explore the mechanism of up-regulation of cyclin D1 levels,

we first examined cyclin D1 mRNA levels. Neither microarrayhybridization3 nor qRT-PCR (Fig. 2D) showed a significantchange in the levels of cyclin D1 mRNA, suggesting that post-transcriptional mechanisms might account for the change.Because cyclin D1 levels are tightly regulated by proteolysis, wesought to determine if there was an alteration in proteinstability. Cycloheximide treatment (Fig. 2E) showed that cyclin

D1 had a significantly longer half-life in the knockout cells (76minutes) compared with WT cells (23 minutes). Treatment ofcells with the proteasome inhibitor MG132 led to a modestincrease in cyclin D1 in the cells, whereas treatment with acalpain inhibitor led to a marked increase in levels (Fig. 2F).These data indicate that calpain is primarily responsible forcyclin D1 degradation in these cells, as has been observed inother mouse fibroblasts (17), and that protein stabilization isprimarily responsible for the observed increase in cyclin D1protein levels.Up-regulation of cyclin D1 in Prkar1a�/� mouse embryonic

fibroblasts is independent of other signaling pathways.Because GSK3h is known to phosphorylate cyclin D1 at theG1-S transition and target it to the cytoplasm for furtherdegradation (18), we examined its activity in the WT andknockout cells. There was no detectable difference betweenknockout and WT cells, indicating that the alterations in stabilityand increased nuclear retention are not due to changes in GSK3hactivity (Fig. 3A).To assess the potential contributions of other (non-PKA-

mediated) pathways to the elevated cyclin D1 levels, we measuredactivity of other pathways shown previously to regulate levels ofthis protein, including Ras/Raf/ERK, phosphatidylinositol 3-kinase/Akt, and nuclear factor-nB (NF-nB; refs. 19–22). GTP-loaded Raswas measured by a pull-down assay as a direct measurement ofRas activity (Fig. 3A), which showed that the active Ras levels wereequivalent or even slightly decreased in Prkar1a�/� cells whencompared with their WT or heterozygous counterparts. We also

Figure 1. Targeted disruption of Prkar1a leads to up-regulation of PKA activity and an immortalized phenotype. A, Prkar1a loxP/loxP MEFs were infected with the controlor Cre retrovirus at passage 1 and PCR genotyped at passage 5 (P5) or passage 10 (P10). B, Western blot of the same MEFs with the indicated antibodies.C, measurement of total and free PKA activity in Prkar1a�/�, Prkar1a+/�, and Prkar1a+/+ MEFs at passage 8. D, 3T3 assay for measuring rate of proliferation inMEFs infected with control or a Cre retrovirus. E, assessment of serum-independent growth in Prkar1a�/� and Prkar1a+/+ MEFs at passage 8.

3 L.S. Kirschner, in preparation.

Loss of Prkar1a Causes Immortalization




measured the levels of the activated (phosphorylated) form of theERKs p42 and p44. In agreement with the Ras data, no excessaccumulation of phosphorylated ERK was detected. Furthermore,changes in phosphorylated p38 levels did not account for thechanges in cyclin D1, excluding the possible involvement ofmitogen-activated protein kinases (MAPK) in this process (Fig. 3B).Similarly, immunoblotting with an antibody that recognizes allforms of phosphorylated Akt revealed that levels of total and phos-phorylated Akt were unchanged in the knockout cells (Fig. 3B).Finally, to determine if increases in cyclin D1 could be attributedto NF-nB activation, we used a luciferase reporter gene driven byNF-nB response elements to assess the activity of this pathway.This experiment revealed that NF-nB activity was not increasedin the knockout cells and was, in fact, somewhat reduced.4

At the functional level, we tested if elevated cyclin D1 levels wereassociated with increased CDK activity. CDK4 and CDK2 activitieswere measured in WT and knockout cells and found to beessentially unchanged (Fig. 3C). This observation prompted us tolook at the amounts of CDK4 physically associated with cyclin D1by coimmunoprecipitation experiments. This experiment, doneeither by immunoprecipitation with anti-CDK4 and immunoblot-ting with cyclin D1 or in the converse fashion, showed that theamount of CDK4 bound to cyclin D1 did not change with loss of

Prkar1a (data not shown). Interestingly, there is a consistent up-regulation of p16INK4a levels, a known CDK4/CDK6 inhibitor (seebelow; ref. 23).Regulation of cyclin D1 by protein kinase A. To confirm that

the increases in cyclin D1 were directly due to increases in PKAactivity, we treated WT and knockout cells with the PKA inhibitorH-89 or with forskolin, a stimulator of adenylate cyclase. In bothWT and Prkar1a�/� cells, H-89 produced a small but consistentdecrease in total cyclin D1 levels at doses where the compoundpredominantly inhibits PKA (24), whereas forskolin did not altertotal cyclin D1 levels (Fig. 4A). Interestingly, we noted a shift in therelative abundance of the cyclin D1 bands, with a decrease in theupper band noted in response to H-89, and an obvious increase inthe slower migrating form of cyclin D1 on forskolin treatment. Toconfirm the identity of the slower migrating form of the protein asa phosphorylated isoform, protein lysates were treated in vitrowith E phosphatase to induce nonspecific protein dephosphory-lation (Fig. 4A , rightmost lanes). This treatment abolished theupper cyclin D1 band with only minimal effects on the lower(presumably unphosphorylated) form of the protein.To further confirm the role of PKA in regulation of cyclin D1, we

transfected 293T cells with the PKA catalytic subunit or with adominant-negative Prkar1a isoform that does not bind cAMP (25).Western blot analysis showed that dominant-negative PKAconsistently down-regulated endogenous cyclin D1 levels (Fig. 4B),although no significant changes were seen with overexpression of

Figure 2. Loss of Prkar1a leads to elevation of cyclin D1 by a post-transcriptional mechanism. A, immunoblotting of MEF lysates for cyclins and downstreamtargets. PTEN levels are included as a loading control. B, increased levels of cyclin D1 are observed by confocal analysis of cyclin D1 staining in Prkar1a+/+ andPrkar1a�/� MEFs. C, immunohistochemical staining for cyclin D1 of a schwannoma from a Prkar1a+/� mouse (8). D, qRT-PCR analysis of cyclin D1 mRNA inWT and knockout MEFs. E, Prkar1a�/� and Prkar1a+/+ MEFs were treated with cycloheximide for indicated times and probed for cyclin D1. Normalized levels ofcyclin D1 protein are plotted versus time. F, Prkar1a�/� and WT MEFs were treated with calpain inhibitor 1 (Clp Inh ; 100 Amol/L) and MG132 (30 Amol/L) forindicated times and probed with the cyclin D1 (CycD1 ) antibody. Actin levels are shown as a loading control (ctl ).

4 K.S. Nadella and L.S. Kirschner, unpublished observations.

Cancer Research




the active catalytic subunit. The induction of dominant-negativePKA was confirmed by increased expression of R1A, providingevidence that the observed changes in cyclin D1 were effectsspecific for PKA action.Immortalization of Prkar1a�/� mouse embryonic fibro-

blasts is independent of p53/p19ARF and other pathways. Toassess for possible non–cyclin D1–mediated effects and to furtherexplore the mechanism of cellular immortalization, we sought todetermine if senescence pathways remained intact. Surprisingly,measurement of p53 activity by a luciferase reporter showed thatthe p53 signaling pathway did not differ significantly between WTand knockout cells (Fig. 5A). Serially passaged Prkar1a�/� andPrkar1a+/+ MEFs showed intact p53, p19ARF, and p16INK4a

compared with the tail tumors obtained from Prkar1a+/� mice,which showed evidence for loss of p53 and p19ARF (Fig. 5B). Thelevels of p16INK4a consistently remained up-regulated in knockoutMEFs compared with the WT MEFs, where the levels were seento decrease with increasing passages. Notably, even in the late-passage (passage 22) Prkar1a�/� MEFS, the p53 pathwayremained intact as judged by p53 stabilization and growth arrest(shown by induction of p21Cip1) after exposure to UV and DNA-damaging agents, such as doxorubicin (Fig. 5C). Examination ofthe tumor suppressor genes involved in cellular senescencedetected only moderate down-regulation of p21Cip1 and a markedup-regulation of p27Kip1 (Fig. 5D). Confocal microscopy of p27Kip1

confirmed the notable up-regulation of the protein and furthershowed that the protein remained localized in the nucleus (Fig. 5E).A similar analysis of p21Cip1 confirmed slight quantitative reduc-tions as well as a redistribution from a predominantly nuclearlocalization to a mixed cytoplasmic-nuclear distribution (Fig. 5F).

Cell cycle–dependent regulation of cyclin D1 in Prkar1a�/�

and Prkar1a+/+ mouse embryonic fibroblasts. To betterunderstand the regulation of cyclin D1, cells synchronized for 72hours by serum deprivation were stimulated to proliferate byaddition of serum. As expected, cyclin D1 accumulated after 9hours in WT cells, whereas in Prkar1a-null MEFs, levels did notvary significantly over the experimental time course. Examinationof the pattern of band mobility also showed differences: a doubletof cyclin D1 was observed only from 0 to 6 hours in the WT cells,but two bands were seen at all times in the Prkar1a�/� MEFs(Fig. 6A and B). This observation implies a cell cycle–dependentregulation of cyclin D1 synthesis, which is abolished in thepresence of elevated PKA activity.Cyclin D1 levels rose during G1 coincident with the decrease of

p27Kip1 levels in control cells, whereas the oscillations of cyclin D1,p27Kip1, and CDK4 roughly paralleled each other in Prkar1a�/�

Figure 3. Other mediators of cyclin D1 increases are not altered inPrkar1a�/� MEFs. A, measurement of GSK3h activity in MEFs by Westernblotting using phosphorylated GSK3h antibody. The genotypes of the cells isshown on top of each set of lanes. B, measurement of GTP-loaded Ras, totaland phosphorylated ERK, Akt, and p38MAPK in MEFs. C, measurement ofCDK4 and CDK2 kinase activities in MEFs of the indicated genotypes.

Figure 4. Elevations of cyclin D1 are a direct effect of elevated PKA activity.A, representative immunoblot showing cyclin D1 and CDK4 levels in H-89 orforskolin (FSK )–treated MEFs. Right, Prkar1a�/� cell lysates treated with andwithout E phosphatase and probed for cyclin D1; bottom, quantitation of thebands for total cyclin D1 normalized against CDK4 loading control (left ) as wellas the ratio between upper and lower cyclin D1–specific bands (right ).B, representative Western blot of 293T cell lysates transiently transfected withconstitutively active (CA ) and dominant-negative (DN ) PKA plasmids.Bottom, quantitation of the cyclin D1 and Prkar1a levels, normalized to actin.*, P = 0.01, compared with untransfected cells.





MEFs, suggesting that a similar mechanism (i.e., phosphorylation)may also be involved in their up-regulation. We observed similarderegulation of cyclin D3 levels in R1A-null MEFs (Fig. 6A and B),whereas cyclin D2 levels were essentially undetectable in the cells(data not shown), which is in agreement with the data reportedin MEFs (26). Timely synthesis of different cell cycle markers wasobserved in WT cells, suggesting a stringent regulation of cell cycle.In contrast to the dysregulation of D-type cyclins, the expression ofother cell cycle markers was not significantly different betweenknockout and WT MEFs.To confirm that loss of Prkar1a was the primary factor

responsible for the observed changes, we reintroduced HA-taggedPRKAR1A (13) into null and WT MEFs to see if the defect couldbe rescued. Prkar1a�/� cells into which HA-R1A had beenreintroduced were synchronized by serum deprivation for 72hours and harvested at different time points after addition ofserum-containing medium. Although not fully reverted to normal,the kinetics of cyclin D1, p27Kip1, and CDK4 in HA-R1A cells werecloser to those observed in WT cells (Fig. 6C). Specifically, cyclinD1 levels increased starting from 6 hours, coincident with adecrease of p27Kip1 levels; in addition, the cyclin D1 doublet wasdetected only at early times after release into the cell cycle,confirming a partial restoration of normal cyclin D1 dynamics.

We ascribe the observation that we were only able to observepartial restoration of normal cell cycle regulation to the fact thatHA-tagged R1A expression was generally quite poor in theknockout compared with robust expression in WT cells (data notshown).

Discussion

In this article, we report the initial phases of our work tounderstand the molecular basis by which patients with the Carneycomplex develop tumors. Compared with the study of human ormouse tumors in vivo , the in vitro system used in this article hasthe advantages that the cells are uniform and that they canbe analyzed immediately after the relevant genetic event hasoccurred. In this way, the changes responsible for molecularalterations in the early phases of tumor initiation can be moreeasily studied.The role of increased protein kinase A activity in tumor

formation. The most striking observation made in this report isthat removal of Prkar1a from primary cells leads to immortaliza-tion without transformation. This finding is in contrast to priorstudies indicating that down-regulation of PRKAR1A suppressedcell proliferation in a variety of cell types, including prostate andpancreatic cancer cells (27, 28). Similarly, as has been observed in

Figure 5. Analysis of cell cycle/cell senescence regulators in Prkar1a�/� cells. A, p53 activity in Prkar1a+/+ and Prkar1a�/� MEFs was measured by luciferaseassay using the p21 promoter. Tp53�/� MEFs are shown as a negative control (right ). B, immunoblot of tail tumors obtained from Prkar1a+/� mice comparedwith WT tail samples and MEFs from passages 7 and 14 probed for p53, p19ARF, p16INK4a, and CDK4. C, Western blotting of late-passage (P22) MEF lysates treatedwith DNA-damaging agents (UV and DOX ) probed for R1A, p53, p19ARF, p21Cip1, and CDK4. D, immunoblot of MEFs probed for CDK inhibitors as indicated.CDK4 is included as a loading control. E, confocal analysis of p27Kip1 immunofluorescence staining in Prkar1a�/� and Prkar1a+/+ MEFs. DAPI immunofluorescencefrom the same field is shown. F, as in (E ), except staining for p21Cip1. Cell nuclei are well visualized, so DAPI staining is not shown.

Cancer Research




Carney complex tumors (2), we found increased basal and totalPKA activity in the Prkar1a knockout cells (Fig. 1C). In theliterature, compounds that stimulate PKA in vitro have generatedconflicting reports regarding proliferation, with some investigatorsreporting growth-inhibitory effects, whereas others describegrowth promotion (29). The explanation for these apparentdiscrepancies may lie in the fact that the prior studies were donein immortalized cell lines in tissue culture, whereas we have usedprimary cells. Observations on cells that have undergone multiplemutations to generate a continuous cell line may not accuratelyreflect the situation in tumor initiation in vivo .The fact that we are able to recapitulate one important aspect of

tumorigenesis (i.e., immortalization) in this primary cell culturemodel bespeaks the importance of the proper model system forunderstanding biological observations. Consistent with the tumorsuppressor function in vivo, Prkar1a deficiency prevented culture-induced senescence. In addition to the loss of replicativesenescence, knockout of Prkar1a also allows the cells a gain ofgrowth factor independence, a second key feature in thedevelopment of cancer (30). However, in order for Carney complexpatients to develop aggressive tumors (e.g., thyroid cancer),secondary and/or tertiary mutations must occur in the immortal-ized cells. This is also modeled nicely in vitro by the fact thatintroduction of activated Ras into the Prkar1a-null MEFs leads to atransformed phenotype.5

The increase in PKA activity caused by loss of Prkar1a may alsoprovide a model with which to understand benign tumorigenesisinduced by activation of PKA caused by mutations in G-protein-coupled receptors (e.g., the thyrotrophin receptor in thyroidadenomas) or in G proteins themselves (e.g., McCune-Albrightsyndrome or somatotroph pituitary adenomas; ref. 31). In each ofthese cases, the establishment of autonomous hyperproliferativecells may be due to the immortalizing effect of increased PKAsignaling, although, as above, there is little impetus towardmalignant transformation in the absence of additional mutations.In fact, down-regulation of PRKAR1A has recently been describedin sporadic growth hormone–producing adenomas (32), although

mutations in PRKAR1A are rarely, if ever, observed in these tumors(33, 34).Effects on modulators of cell senescence. Cellular immortal-

ization is an essential step toward malignant transformation ofnormal cells. In primary MEFs, cellular senescence is mediated viaregulators of cell cycle progression, most prominently p53 andp19ARF (35, 36). In Prkar1a knockout cells, p53/p19ARF and RBactivity remained intact. We found that p16INK4a levels were up-regulated in the knockout MEFs. p16INK4a can compete with cyclinD1 for binding to CDK4, which may account for the lack of changeof CDK4 activity in spite of the increased cyclin D1 levels (37).Together, these observations suggest that loss of Prkar1a immortal-izes MEFs without directly impairing the function of p53/p19ARF

or the RB pathway. Interestingly, we have observed a consistentup-regulation of p27Kip1, which may also occur at the post-transcriptional level (data not shown). As levels of p27 and cyclinD1 have been shown to correlate with each other (38), this finding issomewhat unsurprising. This observation is likely based on theknown association of these proteins, leading to p27Kip1 inhibition ofthe activity of cyclin D-CDK4 and cyclin E-CDK2 complexes (39, 40).Furthermore, it has also been shown (26) that p27Kip1 stabilizescyclin D1 and results in the formation of inactive stable complexeswith CDK4; thus, up-regulation of p27Kip1 in Prkar1a�/� MEFs maypartially explain the increased stability of cyclin D1 in these cells.It is somewhat unusual, however, that p27 retains its nuclearlocalization, as high levels of nuclear p27 have generally been notedto be growth inhibitory. The relationship between elevation incyclin D1 and p27 are currently under further investigation.Relationship between protein kinase A and D-type cyclins in

neoplastic growth. D-type cyclins are the major common targetsby which cells proliferate in response to environmental signals.Intriguingly, MEF cells that carry a knockout of the Men1 gene alsoexhibit increased levels of cyclin D1 (41), although if this was atranscriptional or post-transcriptional event is currently unknown.This common pathway provides a molecular link to the over-lapping spectrum of tumors in patients with MEN1 and Carneycomplex, which includes pituitary and adrenal lesions, as well asskin abnormalities (42).The observation that PKA can affect protein levels adds a new

level of complexity to understanding cyclin-mediated cell cycle

Figure 6. Dysregulation of D-type cyclins and p27Kip1 in Prkar1a�/� MEFs. A, serum-starved R1A+/+ MEFs were induced to reenter the cell cycle by additionof serum, and protein samples were made at time points shown and probed for indicated cell cycle regulators. B, Western blot of Prkar1a�/� MEFs treated as in (A ).C, Western blot of Prkar1a�/� MEFs into which HA-tagged Prkar1a had been reintroduced and treated as in (A).

5 L.S. Kirschner, unpublished observations.





control. In vitro studies of cyclin D1 have indicated that the proteincan be phosphorylated by PKA and that there are three potentialPKA phosphorylation sites in the protein, including one in thecyclin box (43). Our data suggest that PKA activation leads tophosphorylation of cyclin D1, although whether this is a directeffect or the result of PKA activation of a secondary kinase isunknown at present. Our interpretation of the data presented hereis that the majority of cyclin D1 runs as a singlet at Mr 36,000,whereas the PKA-mediated phosphoisoform of the protein runs ata slightly reduced mobility. The exact nature of this isoform and itsimmunologic characteristics are under active investigation.Although cyclin D1 levels have been shown to be elevated in

breast cancers, there has been, in general, poor association betweenlevels of elevation and tumor progression and/or prognosis. Withthe finding that cyclin D1 may be subject to post-translationalmodification affecting its activity, this question may need to bereconsidered. Interaction between cyclin D1 and estrogen receptor(ER) has been studied in the past (44, 45). However, phosphorylationcertainly has the possibility to affect this interaction, and studieshave already suggested that phosphorylation of the ER can affect itsestrogen responsiveness (46). The implication of these observationsfor clinical cancers has yet to be addressed experimentally.Cyclin D1 and cellular immortalization. Overexpression of

cyclin D1 is associated with human cancer and is the most commongenetic alteration seen in many types of cancers, including breast,head and neck squamous cell carcinomas, and mantle celllymphomas (47, 48). Although cyclin D1 is overexpressed in oralesophageal squamous cell carcinomas, inactivating mutations ofpRB were not observed. Furthermore, cyclin D1 overexpressionalone has extended the replicative life span of primary humankeratinocytes up to 80 PDs (49). In early studies of immortalizedfibroblast lines, cyclin D1 overexpression was reported to enhance

cell cycle transit times, although, as seen here, no effect ontransformation was observed (50). Intriguingly, in that study,overexpression of cyclin D1 was sufficient to induce partial growthfactor independence, similar to what we have observed for thePrkar1a knockout cells. The fact that increased cyclin D1 levelswere seen in the early passages (passages 6-8) suggests thatup-regulation of cyclin D1 could be an early event in immortali-zation, although further experiments are needed to determine ifthis finding is a cyclin D1 effect or is a direct result from PKAdysregulation.Summary. The data presented in this article provide a good

in vitro model for tumor initiation as observed in patients with theCarney complex. The observations presented here indicate anunexpected role for PKA in modulating cell cycle progressionthrough regulation of cyclin D1 and p27Kip1 protein levels.Deregulation of these key cell cycle proteins represents a previouslyunexplored PKA dependent mechanism for regulation of cell cycleat G0/G1 phase. The findings presented here also describe a novelrelationship between PKA and D-type cyclins, with implications forboth normal and tumor cell biology.

Acknowledgments

Received 9/5/2005; accepted 9/7/2005.Grant support: Grants HD01323 and CA112268-02 (L.S. Kirschner) and National

Cancer Institute grant CA16058 (Ohio State University Comprehensive Cancer Center).The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

We thank William H. Towns for expert technical assistance, Drs. G. Leone andD. Guttridge for sharing reagents and guidance, Drs. L. Wu and C. Timmers fortechnical advice, Dr. J. Boss for 6� His-Creb, Dr. M-D. Tsai for GST-Rb, Dr. J. Bertheratfor pREP4-HA-R1A, Drs. Stanley Mc Knight and Constantine Stratakis for thedominant-negative and constitutively active PKA plasmids, and Drs S. Jhiang andM. Ringel for critical review of the article.

References1. Carney JA, Gordon H, Carpenter PC, Shenoy BV, GoVL. The complex of myxomas, spotty pigmentation, andendocrine overactivity. Medicine (Baltimore) 1985;64:270–83.

2. Kirschner LS, Carney JA, Pack SD, et al. Mutations ofthe gene encoding the protein kinase A type I-aregulatory subunit in patients with the Carney complex.Nat Genet 2000;26:89–92.

3. Kirschner LS, Sandrini F, Monbo J, Lin JP, Carney JA,Stratakis CA. Genetic heterogeneity and spectrum ofmutations of the PRKAR1A gene in patients with theCarney complex. Hum Mol Genet 2000;9:3037–46.

4. Casey M, Vaughan CJ, He J, et al. Mutations in theprotein kinase A R1a regulatory subunit cause familialcardiac myxomas and Carney complex. J Clin Invest2000;106:R31–8.

5. Sandrini F, Matyakhina L, Sarlis NJ, et al. Regulatorysubunit type I-a of protein kinase A (PRKAR1A): atumor-suppressor gene for sporadic thyroid cancer.Genes Chromosomes Cancer 2002;35:182–92.

6. Bertherat J, Groussin L, Sandrini F, et al. Molecular andfunctional analysis of PRKAR1A and its locus (17q22–24) in sporadic adrenocortical tumors: 17q losses,somatic mutations, and protein kinase A expressionand activity. Cancer Res 2003;63:5308–19.

7. Amieux PS, Howe DG, Knickerbocker H, et al.Increased basal cAMP-dependent protein kinase activityinhibits the formation of mesoderm-derived structuresin the developing mouse embryo. J Biol Chem2002;277:27294–304.

8. Kirschner LS, Kusewitt DF, Matyakhina L, et al. Amouse model for the Carney complex tumor syndrome

develops neoplasia in cAMP-responsive tissues. CancerRes 2005;65:4506–14.

9. Veugelers M, Wilkes D, Burton K, et al. ComparativePRKAR1A genotype-phenotype analyses in humans withCarney complex and prkar1a haploinsufficient mice.Proc Natl Acad Sci U S A 2004;101:14222–7.

10. Morgenstern JP, Land H. Advanced mammaliangene transfer: high titre retroviral vectors with multipledrug selection markers and a complementary helper-free packaging cell line. Nucleic Acids Res 1990;18:3587–96.

11. Saavedra HI, Maiti B, Timmers C, et al. Inactivationof E2F3 results in centrosome amplification. Cancer Cell2003;3:333–46.

12. Guttridge DC, Albanese C, Reuther JY, Pestell RG,Baldwin AS, Jr. NF-nB controls cell growth anddifferentiation through transcriptional regulation ofcyclin D1. Mol Cell Biol 1999;19:5785–99.

13. Groussin L, Kirschner LS, Vincent-Dejean C, et al.Molecular analysis of the cyclic AMP-dependentprotein kinase A (PKA) regulatory subunit 1A(PRKAR1A) gene in patients with Carney complexand primary pigmented nodular adrenocortical dis-ease (PPNAD) reveals novel mutations and clues forpathophysiology: augmented PKA signaling is associ-ated with adrenal tumorigenesis in PPNAD. Am JHum Genet 2002;71:1433–42.

14. Bardeesy N, Sinha M, Hezel AF, et al. Loss of the Lkb1tumour suppressor provokes intestinal polyposis butresistance to transformation. Nature 2002;419:162–7.

15. DeGregori J, Leone G, Ohtani K, Miron A, Nevins JR.E2F-1 accumulation bypasses a G1 arrest resulting fromthe inhibition of G1 cyclin-dependent kinase activity.Genes Dev 1995;9:2873–87.

16. Neuman E, Ladha MH, Lin N, et al. Cyclin D1stimulation of estrogen receptor transcriptional ac-tivity independent of cdk4. Mol Cell Biol 1997;17:5338–47.

17. Choi YH, Lee SJ, Nguyen P, et al. Regulation of cyclinD1 by calpain protease. J Biol Chem 1997;272:28479–84.

18. Diehl JA, Cheng M, Roussel MF, Sherr CJ. Glycogensynthase kinase-3h regulates cyclin D1 proteolysis andsubcellular localization. Genes Dev 1998;12:3499–511.

19. Nheu T, He H, Hirokawa Y, Walker F, Wood J, MarutaH. PAK is essential for RAS-induced upregulation ofcyclin D1 during the G1 to S transition. Cell Cycle 2004;3:71–4.

20. Marshall C. How do small GTPase signal transduc-tion pathways regulate cell cycle entry? Curr Opin CellBiol 1999;11:732–6.

21. Radu A, Neubauer V, Akagi T, Hanafusa H, GeorgescuMM. PTEN induces cell cycle arrest by decreasing thelevel and nuclear localization of cyclin D1. Mol Cell Biol2003;23:6139–49.

22. Zhong H, Voll RE, Ghosh S. Phosphorylation of NF-nB p65 by PKA stimulates transcriptional activity bypromoting a novel bivalent interaction with thecoactivator CBP/p300. Mol Cell 1998;1:661–71.

23. McConnell BB, Gregory FJ, Stott FJ, Hara E, Peters G.Induced expression of p16(INK4a) inhibits both CDK4-and CDK2-associated kinase activity by reassortment ofcyclin-CDK-inhibitor complexes. Mol Cell Biol 1999;19:1981–9.

24. Davies SP, Reddy H, Caivano M, Cohen P. Specificityand mechanism of action of some commonly usedprotein kinase inhibitors. Biochem J 2000;351:95–105.

25. Clegg CH, Correll LA, Cadd GG, McKnight GS.Inhibition of intracellular cAMP-dependent protein

Cancer Research




kinase using mutant genes of the regulatory type Isubunit. J Biol Chem 1987;262:13111–9.

26. Bagui TK, Mohapatra S, Haura E, Pledger WJ.P27Kip1 and p21Cip1 are not required for the formationof active D cyclin-cdk4 complexes. Mol Cell Biol2003;23:7285–90.

27. Cho-Chung YS, Nesterova M, Becker KG, et al.Dissecting the circuitry of protein kinase A and cAMPsignaling in cancer genesis: antisense, microarray, geneoverexpression, and transcription factor decoy. Ann N YAcad Sci 2002;968:22–36.

28. Farrow B, Rychahou P, Murillo C, O’Connor KL,Iwamura T, Evers BM. Inhibition of pancreatic cancercell growth and induction of apoptosis with noveltherapies directed against protein kinase A. Surgery2003;134:197–205.

29. Stork PJ, Schmitt JM. Crosstalk between cAMP andMAP kinase signaling in the regulation of cell prolifer-ation. Trends Cell Biol 2002;12:258–66.

30. Hanahan D, Weinberg RA. The hallmarks of cancer.Cell 2000;100:57–70.

31. Weinstein LS, Yu S, Warner DR, Liu J. Endocrinemanifestations of stimulatory G protein a-subunitmutations and the role of genomic imprinting. EndocrRev 2001;22:675–705.

32. Lania AG, Mantovani G, Ferrero S, et al. Proliferationof transformed somatotroph cells related to low orabsent expression of protein kinase a regulatory subunit1A protein. Cancer Res 2004;64:9193–8.

33. Kaltsas GA, Kola B, Borboli N, et al. Sequenceanalysis of the PRKAR1A gene in sporadic somatotroph

and other pituitary tumours. Clin Endocrinol (Oxf)2002;57:443–8.

34. Sandrini F, Kirschner LS, Bei T, et al. PRKAR1A, oneof the Carney complex genes, and its locus (17q22-24)are rarely altered in pituitary tumours outside theCarney complex. J Med Genet 2002;39:e78.

35. Serrano M, Lee H, Chin L, Cordon-Cardo C, Beach D,DePinho RA. Role of the INK4a locus in tumorsuppression and cell mortality. Cell 1996;85:27–37.

36. Sharpless NE, Alson S, Chan S, Silver DP, CastrillonDH, DePinho RA. p16(INK4a) and p53 deficiencycooperate in tumorigenesis. Cancer Res 2002;62:2761–5.

37. Byeon IJ, Li J, Ericson K, et al. Tumor suppressorp16INK4A: determination of solution structure and ana-lyses of its interaction with cyclin-dependent kinase 4.Mol Cell 1998;1:421–31.

38. Bryja V, Pachernik J, Faldikova L, et al. The role ofp27(Kip1) in maintaining the levels of D-type cyclinsin vivo . Biochim Biophys Acta 2004;1691:105–16.

39. LaBaer J, Garrett MD, Stevenson LF, et al. Newfunctional activities for the p21 family of CDKinhibitors. Genes Dev 1997;11:847–62.

40. Bagui TK, Jackson RJ, Agrawal D, Pledger WJ.Analysis of cyclin D3-cdk4 complexes in fibroblastsexpressing and lacking p27(kip1) and p21(cip1). Mol CellBiol 2000;20:8748–57.

41. Agarwal SK, Novotny EA, Crabtree JS, et al.Transcription factor JunD, deprived of menin, switchesfrom growth suppressor to growth promoter. Proc NatlAcad Sci U S A 2003;100:10770–5.

42. Brandi ML, Gagel RF, Angeli A, et al. Guidelines fordiagnosis and therapy of MEN type 1 and type 2. J ClinEndocrinol Metab 2001;86:5658–71.

43. Sewing A, Muller R. Protein kinase A phosphor-ylates cyclin D1 at three distinct sites within thecyclin box and at the C-terminus. Oncogene 1994;9:2733–6.

44. Doisneau-Sixou SF, Sergio CM, Carroll JS, Hui R,Musgrove EA, Sutherland RL. Estrogen and antiestrogenregulation of cell cycle progression in breast cancercells. Endocr Relat Cancer 2003;10:179–86.

45. Foster JS, Henley DC, Ahamed S, Wimalasena J.Estrogens and cell-cycle regulation in breast cancer.Trends Endocrinol Metab 2001;12:320–7.

46. Michalides R, Griekspoor A, Balkenende A, et al.Tamoxifen resistance by a conformational arrest of theestrogen receptor a after PKA activation in breastcancer. Cancer Cell 2004;5:597–605.

47. Diehl JA. Cyclin g to cancer with cyclin D1. CancerBiol Ther 2002;1:226–31.

48. Fu M, Wang C, Li Z, Sakamaki T, Pestell RG.Minireview: cyclin D1: normal and abnormal functions.Endocrinology 2004;145:5439–47.

49. Opitz OG, Suliman Y, Hahn WC, Harada H, Blum HE,Rustgi AK. Cyclin D1 overexpression and p53 inactiva-tion immortalize primary oral keratinocytes by atelomerase-independent mechanism. J Clin Invest2001;108:725–32.

50. Quelle DE, Ashmun RA, Shurtleff SA, et al. Over-expression of mouse D-type cyclins accelerates G1 phasein rodent fibroblasts. Genes Dev 1993;7:1559–71.





2005;65:10307-10315. Cancer Res Kiran S. Nadella and Lawrence S. Kirschner Immortalization and Dysregulation of D-Type CyclinsDisruption of Protein Kinase A Regulation Causes

Updated version

http://cancerres.aacrjournals.org/content/65/22/10307

Access the most recent version of this article at:

Cited articles

http://cancerres.aacrjournals.org/content/65/22/10307.full#ref-list-1

This article cites 49 articles, 21 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/65/22/10307.full#related-urls

This article has been cited by 10 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's

.http://cancerres.aacrjournals.org/content/65/22/10307To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/content/65/22/10307.full#ref-list-1

http://cancerres.aacrjournals.org/content/65/22/10307.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



disruption of protein kinase a regulation causes ...disruption of protein kinase a regulation causes...

Documents