selective inhibition of casein kinase 1 epsilon minimally...
TRANSCRIPT
JPET #151415
1
Selective Inhibition of Casein Kinase 1 Epsilon Minimally Alters Circadian Clock
Period
Kevin M. Walton, Katherine Fisher, David Rubitski, Michael Marconi, Qing-Jun Meng,
Martin Sládek, Jessica Adams, Michael Bass, Rama Chandrasekaran, Todd Butler, Matt
Griffor, Francis Rajamohan, Megan Serpa, Yuhpyng Chen, Michelle Claffey, Michael
Hastings, Andrew Loudon, Elizabeth Maywood, Jeffrey Ohren, Angela Doran and Travis
T. Wager
Neuroscience Discovery (J.A., K.F., M.M., D.R., K.M.W. ), Medicinal Chemistry (T.B.,
R.C., M.C., Y.C., T.T.W.), Department of Pharmacokinetics, Dynamics, and Metabolism
(M.B., A.D., M.S.), Exploratory Medicinal Sciences (M.G., J.O., F.R.), Pfizer Global
Research and Development, Groton, CT; Faculty of Life Sciences, University of
Manchester, U.K. (Q-J.M, A.L.); Institute of Physiology, Academy of Sciences of the
Czech Republic, Prague, CZ (M.S.); Medical Research Council Laboratory of Molecular
Biology, Cambridge, U.K. (M.H., E.M)
JPET Fast Forward. Published on May 19, 2009 as DOI:10.1124/jpet.109.151415
Copyright 2009 by the American Society for Pharmacology and Experimental Therapeutics.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
2
Running Title: Selective inhibition of CK1ε
Corresponding Author:
Dr. Travis T. Wager
Medicinal Chemistry
MS 8220-4045
Pfizer Global Research and Development
Groton, CT 06340
Email: [email protected]
Tel: 860-715-4059
Fax: 860-686-5552
Number of text pages: 25
Number of tables: 5
Number of figures: 8
Number of references: 34
Number of words in the Abstract: 229
Number of words in the Introduction: 696
Number of words in the Discussion: 1182
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
3
ABBREVIATIONS: SCN, suprachiasmatic nucleus; CK1, casein kinase 1; FASPS,
familial advanced sleep phase syndrome; DSPS, delayed sleep phase syndrome; PF-
4800567, 3-(3-chloro-phenoxymethyl)-1-(tetrahydro-pyran-4-yl)-1H-pyrazolo[3,4-
d]pyrimidin-4-ylamine; PF-670462, 4-[3-cyclohexyl-5-(4-fluoro-phenyl)-3H-imidazol-4-
yl]-pyrimidin-2-ylamine; GFP, green fluorescent protein; EGFR, epidermal growth factor
receptor; PK, protein kinase; D4476, 4-[4-(2,3-dihydro-benzo[1,4]dioxin-6-yl)-5-pyridin-
2-yl-1H-imidazol-2-yl] benzamide; TCEP, tris(2-carboxyethyl)phosphine; EDTA,
ethylenediaminetetraacetic acid; PMSF, phenylmethanesulphonylfluoride.
Recommended section assignment: Cellular and Molecular
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
4
Abstract
The circadian clock links our daily cycles of sleep and activity to the external
environment. Deregulation of the clock is implicated in a number of human disorders
including depression, seasonal affective disorder, and metabolic disorders. Casein kinase
1 epsilon (CK1ε) and casein kinase 1 delta (CK1δ) are closely related ser-thr protein
kinases that serve as key clock regulators as demonstrated by mammalian mutations in
each that dramatically alter the circadian period. Therefore, inhibitors of CK1δ/ε may
have utility in treating circadian disorders. Though we previously demonstrated that a
pan-CK1δ/ε inhibitor, PF-670462, causes a significant phase delay in animal models of
circadian rhythm, it remains unclear whether one of the kinases has a predominant role in
regulating the circadian clock. To test this, we have characterized PF-4800567, a novel
and potent inhibitor of CK1ε (IC50 = 32 nM) with greater than 20-fold selectivity over
CK1δ. PF-4800567 completely blocks CK1ε-mediated PER3 nuclear localization and
PER2 degradation. In cycling Rat1 fibroblasts and a mouse model of circadian rhythm,
however, PF-4800567 has only a minimal effect on the circadian clock at concentrations
substantially over its CK1ε IC50. This is in contrast to the pan-CK1δ/ε inhibitor PF-
670462 which robustly alters the circadian clock under similar conditions. These data
indicate that CK1ε is not the predominant mediator of circadian timing relative to CK1δ.
PF-4800567 should prove useful in probing unique roles between these two kinases in
multiple signaling pathways.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
5
Introduction
All living things, from fungi to humans, have regular cycles aligning them to the
daily events in their environment. These cycles, known as circadian rhythms, are
controlled in mammals by the master clock located in the suprachiasmatic nucleus (SCN)
of the hypothalamus (Antle and Silver, 2005; Gallego and Virshup, 2007). At the cellular
level, the molecular events behind clock cycling are described by the regular increase and
decrease in mRNAs and proteins that define feedback loops, resulting in roughly 24-hour
cycles. The SCN is primarily regulated, or entrained, directly by light via the
retinohypothalamic tract. The cycling outputs of the SCN, not fully identified, regulates
multiple downstream rhythms, such as those in sleep and awakening, body temperature
and hormone secretion (Ko and Takahashi, 2006; Schibler et al, 2003). As anyone who
has experienced jet lag knows, misalignment of the internal clock with the external
environment profoundly affects well-being. Furthermore, diseases such as depression,
seasonal affective disorder and metabolic disorders may have a circadian origin (Barnard
and Nolan, 2008; Green et al, 2008).
The Drosophila kinase doubletime (Dbt) was among the earliest proteins
identified as having a molecular role in the clock, with point mutants producing either
short or long daily cycles (Kloss et al., 1998; Price et al., 1998). Soon after the
identification of Dbt, its mammalian homologue, CK1ε, was identified as mediating the
phenotype of the hamster with the tau mutation (Ralph and Menaker, 1988; Lowrey et al.,
2000). While a normal hamster has a 24 h period, i.e., the time from the beginning of one
active cycle to the next, the tau hamster has a markedly short period of 20 h. This
phenotype, caused by a point mutation in the gene for CK1ε, indicates that the role of this
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
6
kinase in the circadian clock is conserved across evolution. Subsequently, a mutation in
CK1δ, a kinase highly related to CK1ε, was found to mediate a human genetic trait
surprisingly similar to the hamster tau phenotype (Xu et al., 2005). Called familial
advanced sleep phase syndrome (FASPS), this trait is characterized by the “morning
lark” phenotype, wherein individuals exhibit early sleep times, early-morning awakening
and a short period (Jones et al., 1999). This suggests that CK1δ has as prominent a role
in the mammalian circadian clock as CK1ε.
CK1ε and CK1δ control the timing of the clock via phosphorylation of several
proteins, including PER1-3 (Vielhaber et al., 2000; Camacho et al., 2001). PER
phosphorylation by CK1ε and/or CK1δ in the cytoplasm induce their translocation to the
nucleus. Although the action of PER3 in the circadian clock is unclear, PER1 and PER2
bind to the co-regulators CRY1 and CRY2. In the nucleus, these multimers repress the
activity of the CLOCK/BMAL1 transcriptional complex. This results in decreased
expression of the circadian output genes, including repression of the Per and Cry genes
themselves. As the PER proteins collect in the nucleus, CK1ε and/or CK1δ
phosphorylate them further, leading to their ubiquitin-dependent degradation.
Ultimately, the transcriptional repression is fully released and the genes induced by
CLOCK/BMAL1, including Per and Cry, start to express at high levels again. In this
way, the circadian cycle begins anew.
Until recently, unique roles for CK1ε and CK1δ in the circadian clock were
unclear. Although the CK1ε tau mutation dramatically shortens period, recent
investigations with a CK1ε knockout mouse indicate the clock runs well without the
kinase (Meng et al., 2008a). While this suggests the kinases may be redundant, the
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
7
potential exists for compensation by CK1δ in an animal that never expressed CK1ε. An
alternative to address this issue is acute, pharmacologic inhibition. Several previously
identified CK1δ/ε inhibitors, however, have only modest potency and cell penetration
and none exhibit sufficient selectivity to be useful in identifying unique functions
(Badura et al, 2007). This is not surprising as the kinases are 85% similar overall with
98% similarity in the kinase domain (Knippschild et al, 2005). However, we now report
the development of a selective CK1ε inhibitor, PF-4800567. Our studies demonstrate
that acute CK1ε inhibition has little affect on circadian period while a pan-CK1δ/ε
inhibitor dramatically delays the clock. Thus, in the presence of a fully functioning
CK1δ, CK1ε activity is superfluous for establishing the period of the oscillator.
Methods
CK1δ/ε cloning and purification. Full length human CK1δ isoform 1 gene
(accession NP_001884A), mammalian codon optimized, was synthesized by DNA2.0
(Menlo Park, CA) and cloned in-frame into pcDNA4/HisA (Invitrogen, Carlsbad, CA) at
the PstI and ApaI sites. This expressed residues 2 to 415 of the full length CK1δ protein.
A COOH-terminally truncated kinase domain construct of human CK1δ isoform
including residues 1-317 was sub-cloned into the PET15 vector (Novagen, Gibbstown,
NJ) containing an NH3-terminal 6X-His tag and thrombin cleavage site and was
recombinantly expressed in E. coli at 20 °C. The resulting cell pellet was solubilized at 4
mL/g cells in a lysis buffer consisting of 100 mM Tris, pH 7.5, 0.5 M NaCl, 2 mM TCEP,
EDTA-free protease inhibitors (Roche Biochemical, Indianapolis, IN), 1 mM PMSF, and
10 mM imidazole and lysed with a single pass through a pre-chilled microfluidizer at 18
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
8
Kpsi. The lysate was clarified by centrifugation and loaded at 2 mL/min onto a 5 mL Hi
Trap Ni2+-Sepharose Fast-Flow column (GE Healthcare, Chalfont St. Giles, UK) that was
pre-equilibrated with a wash buffer consisting of 50 mM Tris-HCl, 0.5 M NaCl, 1 mM
TCEP, and 100 mM imidazole, pH 7.8. The column was washed with ~50 column
volumes (CV) of wash buffer and then eluted with a 15 CV gradient of wash buffer
containing 500 mM imidazole. The fractions containing CK1δ based on SDS-PAGE
analysis were pooled and then dialyzed into a size exclusion chromatography (SEC)
buffer consisting of 30 mM Tris-HCl, 0.3 M NaCl, 5 mM β-octylglucoside, 1 mM TCEP,
and 1 mM EDTA, pH 7.8. The dialyzed fractions were then treated with 1 μL of bovine
pancreatic high activity thrombin (cat# BCT-1020, Haematologic Technologies, Inc.,
Essex Junction, VT) for 2 h at room temperature or overnight at 4 oC. The cleavage
reaction was halted by slow addition of a 200 mM stock solution of PMSF to a final
concentration of 1 mM. The sample was concentrated, loaded onto a S200 2660 column
equilibrated in SEC buffer and purified at a flow rate of 1 mL/min. Finally, the SEC
fractions containing CK1δ based on SDS-PAGE analysis were pooled, concentrated to ~1
mg/mL, aliquoted and flash-cooled in liquid nitrogen for storage at -80 °C. All
purification steps were conducted at 4 °C unless otherwise noted.
CK1ε cloning and purification was described previously (Badura et al, 2007).
Kinase Assays. The CK1ε and δ kinase assays were performed in a 20 μL
volume in buffer containing 50 mM Tris, pH 7.5, 10 mM MgCl2, 1 mM DTT, 10 μM
ATP and 42 μM of a peptide substrate, PLSRTLpSVASLPGL (Flotow et. al., 1990). The
final enzyme concentrations were 2.5 nM for CK1ε and 2 nM for CK1δ. Assays were
run in a panel format in the presence 1 μL of CK1 inhibitor or 5% DMSO. The reactions
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
9
were incubated for 2 h at room temperature, followed by detection using 20 μL of the
Kinase-Glo Plus Assay reagent (Promega, Madison, WI) according to the manufacturer’s
instructions. Luminescence was measured using Enhanced Lum detection on an Envision
plate reader (Perkin Elmer, Waltham, MA).
Nuclear Translocation. CK1ε and CK1δ dependent PER3 nuclear translocation
was run as described previously for CK1ε (Badura et. al. 2007) with alterations as
indicated below. CK1δ dependent PER3 nuclear translocation was performed with
human full-length CKIδ-pcDNA4/HisA co-transfected into COS-7 cells with the GFP
tagged murine PER3. The sensitivity of the assay was increased by incorporating an anti-
GFP antibody staining step. After the plates were fixed with 4% paraformaldehyde and
washed in PBS, cells were blocked and permeabilized in PBS with 4% goat serum and
0.1% Triton X-100 for 1 h at room temperature. Blocking buffer was then replaced with
a polyclonal anti-GFP antibody (cat. # A11122, Invitrogen) diluted 1:1000 in blocking
buffer and incubated for 2 h. Following 3 washes in PBS, goat anti-rabbit alexa fluor 488
secondary antibody (cat. # A11008, Invitrogen), diluted 1:1000 in Hoechst staining
solution, was added and incubated for 1 h before 3 final washes in PBS.
Re-optimization of the mPer3::GFP cDNA to CK1δ/ε cDNA ratio used in the
transient transfections resulted in a 1:9 ratio for both kinases. Utilizing the Molecular
Translocation Algorithm, High Content analysis was performed on the ArrayScan VTI
(Thermo-Fisher, Thermo Scientific, Pittsburgh, PA). The new scanning parameters
included a 0.035 second exposure for the detection of the antibody stained mPER3-GFP,
an isodata threshold of 0.8, a cytoplasmic ring width of 5 pixels and zero Nuc-Cyto ring
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
10
separation. Output features and data analysis remained as described previously (Badura
et. al. 2007).
Protein Degradation. Real-time monitoring of PER2::YFP degradation in COS-
7 cells was performed essentially as described previously (Meng et al, 2008a). Cells co-
transfected with PER2::YFP and CK1ε expression plasmids were treated first with
compound and subsequently (1 h later) 20 μg/mL cycloheximide (Sigma). Fluorescence
recording was initiated 30 min later, at 10 min intervals, for 8 h by time-lapse microscopy
using Openlab software (Improvision, Warwick, UK). Single cell fluorescent data were
analyzed by IPlab software (Scanalytics, B.D. Bioscience, Rockville, MD) and
background corrected. Overall, 25-30 separate cells in three independent experiments
were measured for each experimental condition. Normalization, one phase exponential
decay curve fitting and statistics were performed in Prism (GraphPad Software, La Jolla,
CA). Data were plotted as best-fit decay curves or as calculated degradation rate
constants K ± SEM.
In Vitro Circadian Bioluminescence Real-Time Recording. To report in real-
time the transcriptional rhythms of the mouse Per2 gene, Rat1 cells were stably
transfected with an mPer2::luc expression plasmid as described previously (Meng,
2008b). Confluent Rat1 stable cells in 35 mm dishes were synchronized by treatment
with 200 nM dexamethasone (DEX) for 1 h. The medium was then changed to non-
phenol red DMEM supplemented with 0.1 mM Luciferin substrate (Yamazaki and
Takahashi, 2005; Meng et al, 2008a). Each 35 mm dish was sealed with vacuum grease
and placed in a light-tight and temperature controlled environment at 37 °C. Light
emission (bioluminescence) was measured continuously using Photomultiplier tubes
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
11
(PMT, H6240 MOD1, Hamamatsu Photonics, Hertfordshire, UK). Data are presented as
photon counts per minute.
For drug treatment, 24 h after synchronization, individual dishes of cells under
PMT recording were treated with CK1 inhibitors (PF-4800567 or PF-670462 at a range
of doses) or DMSO (vehicle control). The compounds were left continuously with the
samples thereafter while the luminescence patterns were recorded for at least 6 days.
Periods were analyzed using RAP software (Okamoto, 2005). Data are expressed as
means ± SEM.
In Vivo Circadian Timing. All procedures were approved by Pfizer’s
Institutional Animal Care and Use Committee before implementation and they were in
accordance with federal policies on animal care and handling procedures. Adult male
C57BL/6J mice (initial weight 25-30 g; Jackson Laboratories, Inc., Jackson, ME) were
used for all experiments. After acclimating for 1 week, the mice were anesthetized with
2.5% isoflurane and prepped for surgical implantation of an indwelling telemeter unit. A
midline incision through skin and muscle was made in the ventral abdominal region, and
a sterile telemeter (TA10TA-F40; Data Sciences International, St. Paul, MN) was
inserted into the peritoneal cavity. The muscle was closed with absorbable suture and the
skin was closed with wound clips. The animals were given 5 mg/kg s.c. carprofen, and
were allowed to recover in a clean cage beneath a heating lamp. Once ambulatory, the
animals were placed in their home cages, the cages were placed on matrixed receivers,
and the telemeters were activated and tested on the data collection system (Dataquest
Acquisition Software; Data Sciences International). Telemeters were programmed to
record temperature for 10 s every 5 min, and the collected data were uploaded into
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
12
Matlab-based (Mathworks Inc., Natick, MA) analysis software (ClockLab; Actimetrics,
Evanston, IL) containing algorithms to determine and predict activity onsets and
associated circadian parameters. Each cage contained a running wheel to encourage
activity and was housed within a light-tight isolation chamber (four cages/box; Plastic
Design, Inc., Burlington, MA) with constant ventilation. Lighting for each box was set at
250 to 300 lux and the light cycle was controlled with a timer (Chrontrol, San Diego,
CA). The timing of lights on and off in each box was also recorded and checked daily.
The mice were kept in 12:12 LD (light:dark cycle of 12 h each) for at least 2 weeks after
recovery from transmitter implantation surgery prior to initiation of the study to ensure
their sleep/wake cycles were entrained to the LD cycle.
Prior to dosing, baseline data were reviewed to ensure animals met exclusion
criteria. Animals were excluded if their baseline tau (daily period) was <23.90 or >24.10
h. Animals were also excluded if the characteristics of their data were insufficient to
allow for reliable determination of activity onsets. Treatment groups were randomized
across isolation chambers such that each chamber contained animals from different
treatment groups. On the day before dosing began, the lights were disconnected once
they had turned off for that day. Dosing was done at circadian time 11 (CT11), which
was 1 h before the lights would normally have turned off. The mice were dosed for 3
days at CT11 and after the completion of dosing the mice remained in DD (24 h dark,
i.e., with the lights off) conditions for 7 days post-dosing. Data were then downloaded
and phase shifts, i.e., the time difference between activity onset pre-dosing vs post-
dosing, were analyzed by ANOVA.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
13
Pharmacokinetic Characterization. The concentration-time profiles of PF-
4800567 and PF-670462 in plasma and whole brain were determined in male C57Bl/6J
mice following administration of a single subcutaneous dose. PF-4800567 was
administered at a dose of 100 mg/kg using 20% hydroxypropyl β-cyclodextrin as dosing
vehicle. PF-670462 was administered at a dose of 32 mg/kg using 5/5/90
DMSO:cremophor:20% hydroxypropyl β-cyclodextrin as dosing vehicle. The dosing
volume for both compounds was 10 mL/kg.
Plasma and whole brains were collected from mice while under isoflurane
treatment at 0.5, 1, 2, 4, 8, and 24 h post dose (N=4 per time point). Plasma was obtained
following centrifugation of whole blood. Whole brains were diluted in a 4x volume
(w/v) 60% isopropyl alcohol and homogenized using a Mini-Beadbeater 96 (Biospec
Products, Bartlesville, OK). Both plasma and brain homogenate samples were quantified
using an LC/MS method following protein precipitation. Concentration and
pharmacokinetic results are presented as mean ± SEM. AUC0-Tlast values were calculated
using linear trapezoidal rule.
Protein binding experiments were conducted with mouse plasma and brain
homogenate using high throughput 96-well equilibrium dialysis (Banker et al, 2003).
Aliquots (N=6 per matrix) of plasma and brain homogenate spiked with 1 μM compound
were dialyzed against an equal volume of buffer for 6 h at 37 °C. The dialysis
membranes had a molecular cutoff of 12 to 14 kDa (Spectrum Laboratories Inc., Rancho
Dominguez, CA). Following incubation, donor and receiver samples were transferred to
a 96-well block containing an equal volume of the opposite matrix and internal standard.
The diluted samples were then analyzed by LC/MS following protein precipitation. Free
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
14
fraction in plasma was calculated as the ratio of instrument response between
donor/receiver samples. Determination of undiluted brain free fraction was calculated as
described previously (Kalvass and Maurer, 2002).
Drugs
Synthesis of PF-4800567. The synthesis of PF-4800567 begins with
commercially available starting materials, as illustrated (Supplemental Materials;
Supplementary Fig. 1). Condensation of 2-(benzyloxy)acetyl chloride with malononitrile
using NaH as the base gave 2 in a 75% isolated yield (1.8 kg). Methylation of 2 with
trimethylorthoformate in the presence of PTSA gave 3 in a 44% isolated yield (500g).
Reaction of 3 with known hydrazine, (tetrahydro-pyran-4-yl)-hydrazine (4) (Ranatunge et
al, 2004) as the dihydrochloride in the presence of triethyl amine in methanol gave 5-
amino-3-benzyloxymethyl-1-(tetrahydro-pyran-4-yl)-1H-pyrazole-4-carbonitrile (5) in a
80% crude yield (190 g). Reaction of 5 with formamide at elevated temperature effected
a second ring closure to give the pyrazolpyrimidine (6) in a 73 % (crude) yield (110 g).
Reaction of 6 under standard hydrogenation conditions removed the benzyl group to yield
[4-amino-1-(tetrahydro-pyran-4-yl)-1H-pyrazolo[3,4-d]pyrimidin-3-yl]-methanol (7) in a
73 % crude yield (73 g). Completion of the synthesis was accomplished by reaction of 7
with 3-chlorophenol under Mitsunobu conditions yielding the title compound, 3-(3-
chloro-phenoxymethyl)-1-(tetrahydro-pyran-4-yl)-1H-pyrazolo[3,4-d]pyrimidin-4-
ylamine (PF-4800567) in a 77% isolated yield. See Supplemental Information for the
detailed synthetic scheme.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
15
PF-670462 was synthesized by the Department of Medicinal Chemistry (Pfizer
Global Research and Development). Through Pfizer’s Gift program, both PF-670462
and PF-4800567 may be available for those interested in using them as tools to better
understand the role of CK1ε and CK1δ for in vitro and in vivo systems. If interested,
please contact the authors for details on the program.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
16
Results
Selectivity of PF-4800567. Recently published data suggest that the kinases
CK1ε and CK1δ may have roles distinct from each other (Meng, 2008a). This prompted
the initiation of efforts to identify compounds that preferentially inhibited one of these
kinases. Via directed chemistry efforts, PF-4800567 (Fig. 1) was identified. When tested
in an in vitro assay against purified CK1ε and CK1δ, this compound had an IC50 of 32
nM for CK1ε, while its CK1δ IC50 was 711 nM (Fig. 2; Table 1). This revealed a 22-fold
greater potency towards CK1ε. In contrast, PF-670462 (Fig. 1), a compound we
described previously as a potent inhibitor of both CK1ε and CK1δ, showed a small
preference in the opposite direction and was approximately 6-fold more potent towards
CK1δ. Both compounds were determined to be ATP competitive (data not shown). PF-
4800567 was tested against 50 additional kinases at 1 μM and/or 10 μM (Fig. 3; Table 2).
These kinases represent a highly diverse subset of the major branches of the kinome tree.
At the lower concentration, PF-4800567 showed significant activity, though less potent,
towards only one additional kinase in this panel, EGFR. The broad nature of this kinase
panel suggests that a similar degree of selectivity for CK1ε exists in the rest of the
kinome. PF-4800567 has greater selectivity against several kinases compared to PF-
670462, including PKA, PKC, p38 and GSK3β.
Inhibition of PER3 nuclear translocation. CK1ε and CK1δ have multiple roles
in the circadian clock. One of these is to phosphorylate the PER proteins in the
cytoplasm, resulting in their translocation into the nucleus. We wanted to test whether
the in vitro preference of PF-4800567 for CK1ε would translate to selective inhibition of
this kinase in intact cells. Each kinase effectively promotes the nuclear localization of
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
17
PER3 (Fig. 4A, 4B, DMSO control). At the highest concentration of each compound, 10
μM, the nuclear localization of the GFP-tagged PER3 by each kinase is effectively
blocked, resulting in the absence of nuclear fluorescence (Fig. 4A, 4B). For PF-670462,
the nuclei are clearly visible down to 0.1 μM, and even down to 0.01 μM the nuclei are
more visible than in the DMSO control (Fig. 4B). Similarly, PF-4800567 effectively
blocks nuclear translocation mediated by CK1ε down to 0.01 μM (Fig4A, top).
However, it is less effective against CK1δ at this concentration, and even at 0.1 μM does
not block CK1δ mediated translocation of PER3 (Fig. 4A, bottom). Quantitation of the
relative PER3 nuclear intensity shows that PF-670462 has comparable IC50’s for both
kinases, while PF-4800567 is over 20-fold more potent for CK1ε (Fig. 4C, 4D; Table 1).
For both inhibitors, the ratio of the kinase IC50’s against the purified enzymes was
essentially the same as the ratio of the kinase IC50’s determined in the whole cell assay.
These data support the translatability of the selectivity data measured with the enzymes to
the more complex whole cell environment.
Inhibition of PER2 degradation. Another key role for CK1ε in the circadian
clock is the phosphorylation-dependent degradation of the PER proteins. Subsequent to
the phosphorylation event that occurs in the cytoplasm to induce PER translocation to the
nucleus and repression of CLOCK/BMAL1, this second phosphorylation takes place in
the nucleus to promote ubiquitin-mediated degradation of PER and relief of the
CLOCK/BMAL1 transcriptional repression (Eide et al, 2005). Coexpression of CK1ε
with PER2-YFP significantly increases the rate of PER2 degradation by about 4-fold (Fig
5). Both PF-4800567 and PF-670462 can completely block the enhanced PER2
degradation. These data, along with the nuclear localization studies, show that each of
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
18
the modulatory roles that CK1ε has on the PER proteins can be potently inhibited by
these compounds.
CK1ε inhibition by PF-4800567 does not alter circadian period in vitro. With
confirmation of the selective inhibition of CK1ε by PF-4800567 and its ability to block
the roles of CK1ε in the clock, we then wanted to investigate the effect of CK1ε
inhibition on the cycling clock itself. Extensive studies of the circadian clock have been
done in Rat1 cells (Ripperger and Schibler, 2006; Meng et al, 2008b). Following the
quiescence of these cells by serum starvation, addition of dexamethasone initiates a
synchronized circadian rhythm. In the cells used here, a stable line expressing PER2-
luciferase shows the rhythmic cycling of this signal over several days (Fig. 6A, 6B).
Addition of the pan-CK1δ/ε inhibitor PF-670462 causes a robust lengthening of period,
as shown by the shifting of the peak of activity over several days (Fig. 6B). Under the
same conditions, PF-4800567 causes a minimal shift in the cycling of PER2-luciferase,
even at the highest doses (Fig. 6A). The period of the cycle in the cells can be calculated
be measuring the time peak to peak. The dose responsive effect on the period shows a
dramatic increase by PF-670462 at concentrations as low as 1 μM, but PF-4800567
shows only a minor effect at concentrations up to 30 μM (Fig. 6C). When these data are
analyzed as a function of the compound concentration relative to the whole-cell IC50 for
each kinase, the ability of the compound to dissect the role of each kinase on the cycling
of the circadian clock is highlighted (Fig. 6D, 6E). PF-670462 begins to increase the
period of the clock at about 3- to10-fold over the whole-cell CK1δ IC50 and about 3-fold
lower for CK1ε. PF-4800567 parallels this effect on period up to about 3-fold the CK1δ
IC50 though it falls off at the highest concentration. However, at the doses where PF-
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
19
4800567 shows a small increase in period, its concentration is nearly 100-fold over its
whole-cell CK1ε IC50 (Fig. 6D). These data indicate that in this circadian model, CK1ε
inhibition alone does not drive changes in the period of the circadian clock.
Characterization of PF-4800567 as an in vivo tool. In order to use PF-4800567
to test the role of CK1e in the clock, its in vivo exposure must be sufficient to inhibit the
kinase in the brain. To assess this, we determined in C57BL/6J mice the
pharmacokinetics of this compound after a single 100 mg/kg subcutaneous dose and
compared it to the pan-CK1δ/ε inhibitor PF-670462 dosed at 32 mg/kg (Fig. 7; Table 3).
The Tmax for both compounds in plasma and brain were observed at the first time point,
0.5 h, indicating rapid absorption and distribution. The brain to plasma ratios were
generally constant throughout the 24 h time course, with an average value of 2.1 and 1.3
for PF-4800567 and PF-670462, respectively.
Conversion of the time course data to reflect free tissue concentrations showed
that free brain Cmax of PF-4800567 was nearly 3-fold over the whole-cell CK1ε IC50
(Fig. 7, Table 4). For CK1δ, however, the free Cmax in the brain is below its whole-cell
CK1δ IC50, reaching only 0.2-fold. PF-670462, in contrast, was 3-fold above its CK1δ
whole cell IC50, the same level as PF-4800567 was for CK1ε, along with a 0.7-fold CK1ε
whole cell IC50. This indicates that in vivo, at the doses tested, PF-670462 should inhibit
most of the CK1δ and CK1ε activity, while PF-4800567 should only reach levels to
selectively inhibit CK1ε. Similar studies at both lower and higher doses indicate the
central and peripheral exposure is dose related for both compounds, and that for PF-
4800567, the higher dose would significantly inhibit CK1δ and thereby lose the selective
nature of the compound (data not shown). At the 24 h time point, PF-4800567 is
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
20
undetectable in the brain and free compound in the plasma is below 1 nM. Therefore, no
significant accumulation of the compound would be expected with daily dosing of 100
mg/kg. Likewise, no significant accumulation of PF-670462 dosed at 32 mg/kg is
expected with daily dosing (Badura et al, 2007; data not shown).
Selective inhibition of CK1ε has little effect on circadian timing in vivo.
Confirmation that PF-4800567 and PF-670462 can be dosed in mice to similar multiples
of their whole-cell CK1ε and CK1δ IC50‘s, respectively, allowed us to test the effects of
this inhibition on in vivo circadian rhythm. Mice were kept in a controlled 12:12 LD
(light and dark cycles of 12 h each) environment for several weeks to ensure their
circadian rhythms were synchronized. They were subsequently shifted into DD (24 h,
dark) to isolate them from the entraining effects light has on circadian rhythm, thereby
providing the greatest sensitivity for the effects of the compounds on the endogenous
circadian rhythm. Following the first 24 h shift into DD, the mice were dosed daily for
three days with either 100 mg/kg PF-4800567 or 32 mg/kg PF-670462 and their
respective vehicles at CT11, i.e., one hour before the time that the lights had previously
been turned off, which is also the beginning of their active period. Following the third
day of dosing, the mice were maintained in DD to determine the effects of the
compounds on their endogenous circadian clock, as indicated by the shift in the start of
their active period post-dosing versus pre-dosing. PF-4800567 delayed the active period
by only 0.5 h when dosed at 3- and 0.2-fold its whole-cell CK1ε and CK1δ IC50’s,
respectively (Fig. 8, Table 5). In contrast, PF-670462 delayed the active period by 6.5 h
when dosed to 0.7- and 3-fold its whole-cell CK1ε and CK1δ IC50‘s, respectively. The
compound-induced delay of the circadian clock did not have any post-dose effects on the
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
21
animals’ tau (period), i.e., the time from the start of one active period to the start of the
next active period (Table 5). This suggests that the changes are confined to an acute
effect on the clock only when the compounds are present.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
22
Discussion
Phosphorylation of circadian clock proteins is an essential element in controlling
the clock’s cyclical rhythm. Genetic analyses of the hamster tau and the human FASPS
mutations confirmed CK1ε and CK1δ, respectively, as primary drivers of the mammalian
clock (Lowrey et al., 2000; Xu et al., 2005). Consistent with the central role of these
kinases in circadian rhythm, chronic administration of a potent, dual inhibitor of CK1ε
and CK1δ, PF-670462, dramatically slows down the clock in cycling fibroblasts in vitro
(Fig. 6). More significantly, acute treatment in vivo induces phase delays in a rodent
model of circadian rhythm (Fig. 8; Badura et al., 2007). These data, however, do not
clarify whether CK1ε and CK1δ are redundant or whether one is the primary driver of the
mammalian circadian clock. The study presented here demonstrates that selective
inhibition of CK1ε by PF-4800567 does not affect the circadian clock in vitro and yields
only a small phase shift in vivo. At 3-fold its whole-cell CK1ε IC50, PF-4800567 caused
less than one-tenth the phase shift that was induced by PF-670462 at 3- and 0.7-fold its
whole-cell CK1δ and CK1ε IC50’s, respectively. It is interesting to note that the large in
vivo effect by PF-670462 is at a concentration that has only a very small effect on Rat1
cells in vitro. This may indicate that the in vivo clock is much more sensitive to changes
in CK1δ/ε inhibition. If so, then the in vivo shift by PF-4800567 may be due to a small
but direct effect on CK1δ added on to its CK1ε inhibition. At the dose tested, PF-
4800567 is at 0.2-fold its whole-cell CK1δ IC50. A 0.5 h phase shift via this low level
CK1δ inhibition is consistent with our observations that phase shift responds in a linear
fashion to compound concentration (data not shown). Additional support for the minor in
vivo activity being due to CK1δ inhibition is that the kinase panel used to characterize
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
23
the selectivity of PF-4800567 indicates it has little activity against most other kinases
tested, including other kinases implicated in circadian function, such as GSK3β and
ERK2 (Gallego and Virshup, 2007).
While it may be that both CK1ε and CK1δ need to be inhibited in order to delay
the clock, the effects of PF-670462 and its preferential inhibition of CK1δ over CK1ε
suggest that CK1δ is the predominant mediator of clock period. The limited separation
between its potency for these kinases, however, makes this claim uncertain. Evaluating
the role of CK1δ in the clock has been difficult, as the knockout of this kinase is
embryonic lethal (Xu et al., 2005). Although it is now clear through selective inhibition
of CK1ε that CK1δ can set clock period by itself, the question remains whether it is also
true for CK1ε in the absence of CK1δ, or is CK1δ the primary controller? Answering
this question requires the development of additional pharmacologic or genetic tools.
Both CK1ε and CK1δ phosphorylate the PER and CRY proteins and CK1ε also
phosphorylates BMAL1, though CK1δ has yet to be tested against BMAL1. Among
these clock pathway substrates, PER phosphorylation is the most extensively studied and
has several specific effects: 1) phosphorylation is required for PER translocation to the
nucleus; 2) in the nucleus, the phosphorylated PERs in combination with the CRYs
inhibit transcription of CLOCK/BMAL1 induced genes, and; 3) further phosphorylation
leads to PER degradation and the restart of transcriptional induction by the
CLOCK/BMAL1 complex (Vielhaber et al., 2000; Akashi et al., 2002; Eide et al., 2005).
In vitro, the lengthening of period by inhibiting both CK1ε and CK1δ with PF-670462
(Fig. 6) is consistent with the slowing of the cycle via chronic inhibition of PER
phosphorylation. This would slow down the rate of movement through each step, but still
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
24
allow the cycle to continue. In vivo, the half-lives of the compounds indicate they would
only slow down the cycle for the short time the compound is present in sufficient
concentration. This would yield an acute delay in the clock each day, with a cumulative
shift over the treatment days. This aligns with the observed effect of PF-670462 (Fig. 8).
The effect of these compounds on the clock is consistent with the characterization
of the hamster tau mutation as having a gain of function CK1ε phenotype (Gallego et al,
2006; Meng et al., 2008a). If the tau mutation induced shorter periods via a kinase with
decreased activity, then it would be expected that a CK1ε kinase inhibitor would have the
same effect. On the contrary, CK1ε inhibition has little or no effect on period. The
lengthening of period by CK1δ/ε inhibition, via decreased PER phosphorylation, aligns
with the shortened period of the tau mutation being mediated by increased PER
phosphorylation. With regards to the lack of an effect with CK1ε inhibition alone, loss of
CK1ε phosphorylation of PER could be compensated simply by the increased
presentation of unphosphorylated PER to CK1δ. In the tau mutation, however,
compensation of increased PER phoshorylation by CK1ε would require the concomitant
decrease of CK1δ activity and the kinase may not be able to respond in this manner.
Despite the observation that selective inhibition of CK1ε has little effect on
setting period, it should not be construed to mean that CK1ε is dispensable in circadian
biology. With the central role that phosphorylation plays in the clock (Gallego and
Virshup, 2007), there are other aspects of circadian function not tested here that CK1ε
may yet reveal a primary role. A recently reported genetic association between CK1ε
and stimulant sensitivity suggests the kinase may have a significant role in the ability of
stimulants to entrain behavior (Falcon and McClung, 2008; Bryant et al, 2009).
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
25
The identification of a selective CK1ε inhibitor opens up other key regulatory
pathways for exploration into the function of this kinase. For example, CK1ε has a key
role in Wnt signaling. The Wnt pathway stimulates changes in gene expression through
the induction of β-catenin translocation to the nucleus, leading to effects on proliferation,
survival and cell fate (Price, 2006). Addition of the peptide Wnt to cells activates CK1ε
(Swiatek et al., 2004). This results in phosphorylation of Dishevelled and translocation of
β-catenin to the nucleus (McKay et al., 2001; Price, 2006). It is unclear whether both
CK1δ and CK1ε have a role in the Wnt pathway. It was claimed that the CK1ε tau
mutation has a loss of function in this pathway, but no related effects were reported in the
tau hamster or tau transgenic mice. It is possible, however, that the loss of activity is
compensated by CK1δ. Although the CK1 inhibitor D4476 can block Wnt signaling, it
does not have selectivity between CK1ε and CK1δ (Bryja et al., 2007). PF-4800567
provides a tool to probe for separate roles of these kinases in this pathway.
CK1ε and CK1δ have generally been treated as interchangeable due to their high
degree of sequence identity and the lack of tools to tease apart their unique functions.
The development of PF-4800567 as a pharmacological tool that can distinguish between
these two kinases now creates an opportunity to explore the differences between these
kinases in greater detail.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
26
References
Akashi M, Tsuchiya Y, Yoshino T, and Nishida E (2002) Control of intracellular
dynamics of mammalian period proteins by casein kinase I ε (CK1ε) and CK1δ in
cultured cells. Mol Cell Biol 22:1693–1703.
Antle MC and Silver R (2005) Orchestrating time: arrangements of the brain circadian
clock. Trends Neurosci 28:145–151.
Badura L, Swanson T, Adamowicz W, Adams J, Cianfrogna J, Fisher K, Holland J,
Kleiman R, Nelson F, Reynolds L, et al. (2007) An inhibitor of casein kinase Iε
induces phase delays in circadian rhythms under free-running and entrained
conditions. J Pharm Exp Ther 322:730-738.
Banker MJ, Clark TH, and Williams JA (2003) Development and validation of a 96-well
equilibrium dialysis apparatus for measuring plasma protein binding. J Pharm Sci 92:
967-974.
Barnard AR and Nolan PM (2008) When Clocks Go Bad: Neurobehavioural
Consequences of Disrupted Circadian Timing. PLOS Genet 4:e1000040.
Bryant CD, Graham ME, Distler MG, Munoz MB, Li D, Vezina P, Sokoloff G and
Palmer AA (2009) A role for casein kinase 1 epsilon in the locomotor stimulant
response to methamphetamine. Psychopharm doi 10.1007/s00213-008-1417-z.
Bryja V, Schulte G, Rawal N, Grahn A, and Arenas E. (2007) Wnt -5a induces
Dishevelled phosphorylation and dopaminergic differentiation via a CK1-dependent
mechanism. J Cell Sci 120:586-595.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
27
Camacho F, Cilio M, Guo Y, Virshup DM, Patel K, Khorkova O, Styren S, Morse B, Yao
Z, and Keesler GA (2001) Human casein kinase Iδ phosphorylation of human
circadian clock proteins period 1 and 2. FEBS Lett 489:159-165.
Green CB, Takahashi JS and Bass J (2008) The meter of metabolism. Cell 134:728-742.
Eide EJ, Woolf MF, Kang H, Woolf P, Hurst W, Camacho F, Vielhaber EL, Giovanni A,
and Virshup DM (2005) Control of mammalian circadian rhythm by CK1ε-Regulated
Proteasome-Mediated PER2 Degradation. Mol Cell Biol 25:2795-2807.
Falcon E and McClung CA (2008) A role for the circadian genes in drug addiction.
Neuropharm 56:91-96
Flotow H, Graves PR, Wang AQ, Fiol CJ, Roeske RW, and Roach PJ (1990) Phosphate
groups as substrate determinants for casein kinase I action. J Biol Chem 265:14264–
14269.
Gallego M, Eide EJ, Woolf MF, Virshup DM, and Forger DB (2006) An opposite role for
tau in circadian rhythms revealed by mathematical modeling. Proc Natl Acad Sci USA
103:10618–10623.
Gallego M and Virshup DM (2007) Post-translational modifications regulate the tiCK1ng
of the circadian clock. Nat Rev Mol Cell Biol 8:139-148.
Jones CR, Campbell SS, Zone SE, Cooper F, DeSano A, Murphy PJ, Jones B,
Czajkowski L, and Ptacek LJ (1999) Familial advanced sleep-phase syndrome: A
short-period circadian rhythm variant in humans. Nat Med 5:1062-1065.
Kalvass JC and Maurer TS (2002) Influence of nonspecific brain and plasma binding on
CNS exposure: implications for rational drug discovery. Biopharm Drug Dispos 23:
327-338.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
28
Kloss B, Price JL, Saez L, Blau J, Rothenfluh A, Wesley CS, and Young MW (1998) The
Drosophila clock gene double-time encodes a protein closely related to human casein
kinase Iε. Cell 94:97-107.
Knippschild U, Gocht A, Wolff S, Huber N, Lohler J, and Stoter M (2005) The casein
kinase 1 family: participation in multiple cellular processes in eukaryotes. Cell Signal
17:675–689.
Ko CH and Takahashi JS (2006) Molecular components of the mammalian circadian
clock. Hum Mol Gen 15:R271-R277.
Lowrey PL, Swhimomura K, Antoch MP, Yamazaki S, Zemenides PD, Ralph MR,
Menaker M, and Takahashi JS (2000) Positional syntenic cloning and functional
characterization of the mammalian circadian mutation tau. Science 288: 483–490.
McKay RM, Peters JM, and Graff JM. (2001) The Casein Kinase I Family in Wnt
Signaling. Dev Biol 235:388-396.
Meng QJ, Logunova L, Maywood ES, Gallego M, Lebiecki J, Brown TM, Sladek M,
Semikhodskii AS, Glossop NRJ, Piggins HD, et al. (2008a) Setting clock speed in
mammals: The CK1e tau mutation in mice accelerates circadian pacemakers by
selectively destabilizing PERIOD proteins. Neuron 58:78-88.
Meng QJ, McMaster A, Beesley S, Lu WQ, Gibbs J, Parks D, Collins J, Farrow S, Donn
R, Ray D, and Loudon A (2008b) Ligand modulation of REV-ERBα function resets
the peripheral circadian clock in a phasic manner. J Cell Sci 121:3629-3635.
Okamoto K, Onai K, and Ishiura M (2005) RAP, an integrated program for monitoring
bioluminescence and analyzing circadian rhythms in real time. Anal Biochem 340:193-
200.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
29
Price JL, Blau J, Rothenfluh A, Abodeely M, Kloss B, and Young MW (1998) Double-
time is a novel Drosophila clock gene that regulates PERIOD protein accumulation.
Cell 94:83-95.
Price MA (2006) CK1, there's more than one: casein kinase I family members in Wnt and
Hedgehog signaling. Genes & Dev 20:399-410.
Ralph MR and Menaker M (1988) A mutation of the circadian system in golden
hamsters. Science 241:1225-1227.
Ranatunge RR, Augustyniak M, Bandarage UK, Earl RA, Ellis JL, Garvey DS, Janero
DR, L. Letts G, Martino AM, Murty MG, et al. (2004) Synthesis and Selective
Cyclooxygenase-2 Inhibitory Activity of a Series of Novel, Nitric Oxide Donor-
Containing Pyrazoles. J Med Chem 47:2180-2193.
Ripperger J and Schibler U (2006) Rhythmic CLOCK-BMAL1 binding to multiple E-box
motifs drives circadian Dbp transcription and chromatin transitions. Nat Gen 38:369-
374
Schibler U, Ripperger J and Brown SA (2003) Peripheral Circadian Oscillators in
Mammals: Time and Food. J Biol Rhythms 18:250-260
Swiatek W, Tsai I-C, Klimowski L, Pepler A, Barnette J, Yost HJ, and Virshup DM
(2004) Regulation of Casein Kinase Iε Activity by Wnt Signaling. J Biol Chem 279:
13011-13017.
Vielhaber E, Eide E, Gao Z-H, and Virshup DM (2000) Nuclear entry of the circadian
regulator mPER1 is controlled by mammalian casein kinase I ε. Mol Cell Biol
20:4888–4899.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
30
Xu Y, Padiath QS, Shapiro RE, Jones CR, Wu SC, Saigoh N, Saigoh K, Ptacek LJ, and
Fu Y-H (2005) Functional consequences of a CK1δ mutation causing familial
advanced sleep phase syndrome. Nature 434:640–644.
Yamazaki S and Takahashi JS (2005) Real-time luminescence reporting of circadian gene
expression in mammals. Methods Enzymol 393:288-301.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
31
Footnotes
Send reprint requests to Dr. Travis T. Wager, Medicinal Chemistry, MS 8220-4045,
Pfizer Global Research and Development, Groton, CT 06340
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
32
Legends for Figures
Figure 1. Structures of PF-4800567 and PF-670462.
Figure 2. PF-4800567 selectively inhibits CK1ε versus PF-670462 which is more potent
at CK1δ. Purified enzymes were assayed in vitro in the presence of increasing
concentrations of inhibitor. Activity was measured as a function of remaining ATP at the
termination of the kinase reaction and plotted as percent activity inhibited relative to
DMSO control. Data shown are from two separate experiments each performed in
triplicate, mean +/- SEM. IC50 values were determined by analysis in Graphpad PRISM
and are shown in Table 1.
Figure 3. Heat map of PF-4800567 selectivity. Compound was tested at 1 μM at the
indicated kinases using Invitrogen SelectScreen Kinase Profiling. All kinases were tested
at their Km for ATP.
Figure 4. PF-4800567 selectively inhibits CK1ε mediated PER3 nuclear translocation.
Representative examples of COS-7 cells cotransfected with cDNA expression vectors for
GFP-PER3 and CK1ε or CK1δ were incubated with increasing concentrations of either
PF-4800567 (A) or PF-670462 (B); a single DMSO control is used for each kinase for
both compounds. Percent inhibition of GFP signal localized to the nucleus relative to
DMSO control was quantitated and compiled from two separate experiments performed
in triplicate, mean +/- SEM (C, D). IC50 values were determined by analysis in Graphpad
PRISM and are shown in Table 1.
Figure 5. Increased PER2 degradation is inhibited by both PF-4800567 and PF-670462.
A) Decay of PER2::YFP fluorescence in COS-7 cells cotransfected with CK1ε kinase
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
33
dead or CK1ε wild-type incubated either with vehicle or 0.5 μM compound. Cells were
treated with cyclohexamide (20 mg/mL) 0.5 h before recording. B) Degradation rate
constant K for PER2::YFP in presence of CK1ε kinase-dead or wild-type treated with
vehicle or the indicated compounds (mean ± SEM, *p < 0.001 relative to all other
conditions).
Figure 6. Selective CK1ε inhibition by PF-4800567 does not alter the circadian period
while preferential CK1δ inhibition by PF-670462 does alter the period. Representative
PER2::LUC bioluminescence oscillations in Rat1 cells in the presence of increasing
concentrations of either PF-4800567 (A) or PF-670462 (B); arrow indicates time of
compound addition. (C) Quantitation of period as measured peak to peak, data are mean
+/- SEM from 5 independent experiments of single cultures. Data are replotted as fold
over whole cell CK1ε IC50 (D) or as fold over whole cell CK1δ IC50 (E).
Figure 7. CNS exposure of PF-4800567 is sufficient to selectively inhibit CK1ε in vivo.
Time course of PF-4800567 exposure in C57BL/6J mice following a single 100mg/kg
s.c. dose. Brain concentrations were below the limit of quantitation in all animals at the
24 hour time point.
Figure 8. Selective CK1ε inhibition by PF-4800567 has only minor effects on shifting in
vivo phase while preferential CK1δ inhibition by PF-670462 induces a large phase shift.
Mice (N=5) were maintained in 12:12 LD for several weeks to get a baseline of activity.
The lights were subsequently turned off and the mice kept in DD conditions for testing.
They were dosed with 100 mg/kg of PF-4800567 or 32 mg/kg PF-670462, or the
appropriate vehicle, for 3 days, q.d. They were maintained another week in DD and
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
34
shifts in the start of their active periods post-dose were determined relative to the starts
pre-dose; this is defined as the phase shift. A) quantitation of the phase shifts. B)
representative temperature analyses shown as double-plotted actograms from each of the
treatment groups. Asterisks indicate time of dosing. By convention, an advance in the
active period is positive and a delay is negative.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
35
Table 1: Enzyme and whole cell IC50’s for PF-4800567 and PF-670462. Values with
95% confidence intervals (CI) determined from data shown in Figs. 3 and 4.
ENZYME WHOLE CELL
IC50 (μM) Ratio IC50 (μM) Ratio
CK1ε (95% CI)
CK1δ (95% CI)
.CK1δ CK1ε
CK1ε (95% CI)
CK1δ (95% CI)
CK1δ CK1ε
PF-4800567 0.032 (0.025 to 0.042)
0.711 (0.566 to 0.893)
22.22 0.13 (0.10 to 0.17)
2.65 (1.35 to 5.19)
20.38
PF-670462 0.080 (0.069 to 0.092)
0.013 (0.011 to 0.017)
0.16 0.55 (0.41 to 0.74)
0.14 (0.10 to 0.20)
0.25
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
36
Table 2: Selectivity comparison between PF-4800567 and PF-670462
PF-4800567 PF-670462
1 μM 10 μM 1 μM 10 μM
CK1γ2 1 nd 2 nd
GSK3β nd 0 nd 57
p38 cascade 2 3 100 100
MLCK nd 4 nd 53
PRKC B2 2 7 29 75
PKACα 1 23 78 98
EphA2 16 39 20 68
SRC 9.5 47 nd 58
VEGFR2 11 51 20 70
LCK 32 57 55 88
HGK 20 75 97 99
EGFR 69 83 76 99
Data shown are per cent inhibition at the indicated concentration of each inhibitor.
CK1γ2, casein kinase 1γ2; GSK3β, glycogen synthase kinase 3β; p38 cascade, p38α
coupled to MAPKAPK2; MLCK, myosin light chain kinase; PRKC B2, protein kinase C
B2; PKAC α, protein kinase A catalytic subunit α; VEGFR2, vascular endothelial growth
factor receptor 2 tyrosine kinase; EGFR, epidermal growth factor receptor tyrosine
kinase.
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
37
Table 3: Pharmacokinetic parameters of PF-4800567 and PF-670462 following s.c.
administration to mice.
PF-4800567 PF-670462
Plasma Brain Plasma Brain
Dose (mg/kg) 100 100 32 32
AUC (ng*h/mL) 6030 11400 2880 3410
Total Cmax (ng/mL) 5250 9250 2760 2880
Tmax (h) 0.5 0.5 0.5 0.5
fu 0.069 0.016 0.12 0.046
Free Cmax (nM) 1007 411 982 393
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
38
Table 4: Comparison of the observed free Cmax concentrations in mice relative to
whole-cell CK1δ/ε IC50 values. PF-4800567 and PF-670462 were dosed at 100mpk and
32mpk, respectively.
PF-4800567 PF-670462
Fold over Whole Cell IC50 Free Plasma Free Brain Free Plasma Free Brain
CK1ε 8 3 2 0.7
CK1δ 0.4 0.2 7 3
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
JPET #151415
39
Table 5: Selective CK1ε inhibition by PF-4800567 has only a small effect on the
circadian clock in vivo while preferential CK1δ inhibition by PF-670462 dramatically
delays the clock. Phase shift indicates the difference in the time of day of the start of the
active period prior to treatment to the start of the active period following treatment. By
convention, an advance in the active period is positive and a delay is negative. Tau refers
to the time from the start of one active period to the start of the next active period.
Significance testing performed by one-way ANOVA.
Treatment Phase Shift (h) Pre-dose tau Post-dose tau
mean SEM p value mean SEM mean SEM
Veh1 0.17 0.15 24.00 0.01 23.81 0.03
PF-4800567 -0.51 0.08 <0.01 24.01 0.01 23.84 0.03
Veh2 0.19 0.13 24.00 0.01 23.81 0.08
PF-670462 -6.55 0.23 <0.001 23.99 0.01 23.80 0.11
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from
This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 19, 2009 as DOI: 10.1124/jpet.109.151415
at ASPE
T Journals on January 9, 2020
jpet.aspetjournals.orgD
ownloaded from