CK1δ Kinase Activity Is Modulated by Chk1-Mediated Phosphorylation

CK1δ, a member of the casein kinase 1 family, is involved in the regulation of various cellular processes and has been associated with the pathophysiology of neurodegenerative diseases and cancer. Therefore recently, interest in generating highly specific inhibitors for personalized therapy has increased enormously. However, the efficacy of newly developed inhibitors is affected by the phosphorylation state of CK1δ. Cellular kinases phosphorylating CK1δ within its C-terminal domain have been identified but still more information regarding the role of site-specific phosphorylation in modulating the activity of CK1δ is required. Here we show that Chk1 phosphorylates rat CK1δ at serine residues 328, 331, 370, and threonine residue 397 as well as the human CK1δ transcription variants 1 and 2. CK1δ mutant proteins bearing one, two or three mutations at these identified phosphorylation sites exhibited significant differences in their kinetic properties compared to wild-type CK1δ. Additionally, CK1δ co-precipitates with Chk1 from HT1080 cell extracts and activation of cellular Chk1 resulted in a significant decrease in cellular CK1δ kinase activity. Taken together, these data point towards a possible regulatory relationship between Chk1 and CK1δ.

In the cell CK1 expression and activity is tightly regulated. Processes regulating CK1 activity include stimulation of cells with insulin, viral transformation, treatment of cells with topoisomerase inhibitors or c-irradiation, all resulting in elevated CK1 activity [50,[60][61][62]. Furthermore, alternative splicing of CK1 isoforms is involved in modulating their subcellular localization, turnover and substrate binding [63][64][65]. CK1 can also be regulated by interaction with small molecules and other proteins. Interaction with cellular proteins has been shown to influence the localization of CK1 [66][67][68] and to either enhance or inhibit its activity [21,69]. Furthermore, intramolecular autophosphorylation downregulates CK1 activity [70][71][72]. However, several mechanisms exist to overcome the inhibitory effect of autophosphorylation [70,71,73].
Differences in the activity of CK1d in different tissues and cell lines have been suggested to be partially caused by site-specific phosphorylation by cellular kinases. One of these kinases has been identified as PKA which phosphorylates CK1d preferentially at Ser-370 in vitro and in vivo [74]. As consequence of PKA-mediated phosphorylation, CK1d kinase activity is reduced while in turn mutation of Ser-370 to alanine enhances CK1d-mediated substrate phosphorylation. Inhibition of the interaction of PKA with A-kinase anchoring proteins (AKAPs) in MiaPaCa-2 cells led to significantly increased CK1d kinase activity and overexpression of the CK1d S370A mutant enhanced the formation of an ectopic dorsal axis during Xenopus laevis development [74]. Whereas the phosphorylation site mainly targeted by PKA has been identified, only evidence exists that additional kinases, among them AKT, Clk2, and PKC-a, are able to phosphorylate CK1d within its Cterminal domain [74]. Given the fact that CK1d is also involved in regulating cell cycle progression, checkpoint kinase 1 (Chk1) could also be involved in the regulation of CK1d kinase activity by sitespecific phosphorylation. This assumption is supported by the fact that several minimal consensus sequences for Chk1 are present within the C-terminal regulatory domain of CK1d [75]. Here we show that Chk1 is able to phosphorylate rat CK1d at Ser-328, Ser-331, Ser-370, and Thr-397 as well as human CK1d transcription variants (TV) 1 and 2 in vitro. CK1d mutant proteins bearing one, two or three mutations at the identified phosphorylation sites exhibit altered kinetic characteristics compared to wild-type CK1d. Furthermore we show an interaction of CK1d and Chk1 by co-immunoprecipitation analyses and that activation of cellular Chk1 by hydroxyurea results in a significant decrease in the activity of CK1d. Therefore our data suggest a regulatory effect of Chk1 on the activity of CK1d.

Extraction of RNA and cDNA Synthesis
First, RNA was extracted from the human fibrosarcoma cell line HT1080 [76] using the peqGOLD RNAPure system (Peqlab, Germany) according to the manufacturer's instructions. 1 mg of total RNA was then reverse transcribed using oligo(dT) primers and the AffinityScript Multiple Temperature cDNA synthesis kit (Agilent Technologies, USA). For subsequent amplification of human CK1d TV1 and 2 fragments 1 ml of cDNA synthesis reaction was used in PCR.

Construction of Expression Vectors
Full length CK1d and C-terminal fragments of rat CK1d (accession ID: L07578 and [77]) were amplified by PCR using the primers listed in table 1. Resulting PCR products were first ligated into the pSC-A cloning vector (Agilent Technologies, USA) followed by subsequent cloning into pGEX expression vectors (GE Healthcare, UK) to generate the GST-CK1d fusion proteins shown in table 2. The C-terminal fragments of human CK1d TV1 (NM_001893) and TV2 (NM_139062) were amplified from HT1080 cDNA.
Phosphorylation site mutants of wild-type full length rat CK1d and CK1d C-terminal fragments were created by using the QuickChange site-directed mutagenesis kit (Agilent Technologies, USA) and the SDM primer pairs as indicated in tables 1 and 2.

Overproduction and Purification of Glutathione Stransferase Fusion Proteins
The production and purification of glutathione S-transferase (GST) fusion proteins were carried out as described previously [61].

Immunoprecipitation Kinase Assay
Using 2 mg of specific antibodies Chk1 or CK1d were precipitated from 250 mg of HT1080 protein extracts after cells were treated with hydroxyurea. The phosphatase inhibitors sodium fluoride (50 mM) and sodium orthovanadate (0.2 mM) were added to the lysis buffer to protect proteins from getting dephosphorylated. Antibody complexes were captured using 30 ml of protein G sepharose (50% (v/v) protein G sepharose in PBS; GE Healthcare, USA). Sepharose-bound complexes were washed three times with lysis buffer and once with kinase buffer without ATP. Determination of Chk1-or CK1d-speicific kinase activity was carried out in kinase buffer using 2 mCi [c-32 P]-ATP per reaction and either 1 mg CHKtide peptide substrate (Millipore, Billerica, USA) or 0.5 mg GST-p53 1-64 (FP267) were used as substrate. Reactions were incubated for 30 min at 30uC and reactions using CHKtide peptide substrate were stopped by absorption of the complete reaction volume on P81 filter paper. Papers were washed three times with 75 mM orthophosphoric acid and once in acetone. Dried papers were used for Cherenkov counting to quantify phosphate incorporation. Reactions containing FP267 were separated by SDS-PAGE and phosphate incorporation into FP267 was determined by Cherenkov counting.
For co-incubation experiments activated, precipitated Chk1 was incubated with GST-CK1d for 10 min at 30uC prior to addition of GST-b-catenin 1-181 (FP1355; 0.5 mg/reaction) and another incubation step at 30uC for 30 min.

Two-dimensional Phosphopeptide Analysis
In vitro phosphorylated GST-CK1d fusion proteins were analyzed by two-dimensional phosphopeptide analysis using a standard protocol described earlier [83]. Briefly, in vitro phosphorylated proteins were separated by SDS-PAGE and blotted onto a PVDF membrane (Hybond-P, GE Healthcare, UK). Membrane fragments containing labeled proteins were incubated in 100 mM acetic acid containing 5% (w/v) polyvinylpyrrolidone at 37uC for 30 min, extensively washed with 50 mM ammonium bicarbonate buffer, digested with TPCK-trypsin (Thermo Fisher Scientific, Germany), and oxidized with performic acid on ice for two hours. Phosphopeptides were analyzed on cellulose TLC plates (Merck, Germany) by electrophoresis at pH 1.9, followed by ascending chromatography in a buffer containing butanol/pyridine/acetic acid/H 2 O, 15:10:3:12 (v/v).

Phosphoamino Acid Analysis
Phosphoamino acid analysis was performed as described earlier [83]. Phosphopeptides were eluted from the TLC plate and subsequently hydrolyzed in 6 N HCl at 120uC. Resulting amino acids were mixed with phosphoamino acid standards (Sigma-Aldrich, USA) and analyzed by two-dimensional electrophoresis at pH 1.9 and 3.6. Phosphoamino acid positions were determined after staining of the standards with ninhydrin staining solution (0.2% (w/v) ninhydrin and 4% (v/v) acetic acid in acetone). [ 32 P] labeled amino acids were visible after autoradiography.
10% of the digest was directly analyzed by LC-MS/MS, acetonitrile was added to 90% of the peptide mixture to a final concentration of 30% and the pH was adjusted to 2-3. Enrichment of phosphopeptides by titanium dioxide chromatography was done as described previously [85] with the following modifications: phosphopeptide elution from the beads was performed three times with 100 ml 40% (v/v) ammonium hydroxide solution in 60% (v/v) acetonitrile at a pH of .10.5.
Analyses of the peptides were done on a Proxeon Easy-LC system (Proxeon Biosystems, Denmark) coupled to a LTQ-Orbitrap-XL (Thermo Fisher Scientific, Germany) equipped with a nanoelectrospray ion source (Proxeon Biosystems, Denmark) as described previously [86]. The five most intense precursor ions were fragmented by activation of neutral loss ions at 298, 249, and 232.6 relative to the precursor ion (multistage activation). Mass spectra were analyzed using the software suite MaxQuant, version 1.0.14.3 [87]. The data were searched against a targetdecoy Escherichia coli database containing 4163 forward protein sequences, the sequence of the GST-CK1d 305-428 fragment, and 262 frequently observed contaminants. Trypsin and GluC, respectively, were set as proteases in which two missed cleavage sites were allowed. Beside acetylation at the N-terminus and oxidation of methionine, phosphorylation of serine, threonine, and tyrosine were set as variable modifications. Carbamidomethylation of cysteine was set as fixed modification. Initial precursor mass tolerance was set to 7 parts per million (ppm) at the precursor ion and 0.5 Da at the fragment ion level. Spectra of modified peptides were manually validated. MASCOT scores are given for probabilistic evaluation of mass spectrometric results. Total obtained sequence coverage was 47.6%.

Immunoprecipitation
For immunoprecipitation 250 mg whole cell protein extract was diluted to a final volume of 500 ml and incubated with 2 mg of Chk1-specific antibody or control serum overnight under slow rotation at 4uC. Antibodies were captured with 40 ml protein G sepharose (50% (v/v) protein G sepharose in PBS) under continued rotation at 4uC for 5 hours. Immunocomplexes were collected by pulse centrifugation and washed three times with 250 ml NP40 lysis buffer. Finally, collected complexes were mixed with 26SDS sample buffer and incubated for 10 min at 37uC and 750 rpm in a thermal mixer. All samples were pulse-centrifuged, the supernatant was used for SDS-PAGE and subsequent Western blotting.

SDS-PAGE and Western Blot Analysis
Cell lysates used for Western blot analyses were prepared in NP40 lysis buffer. 80 mg of cleared extracts were separated on 12.5% (v/v) gels in SDS-PAGE and transferred to a PVDF membrane (Hybond-P, GE Healthcare, UK). For immunodetection of Chk1, pChk1 S345 , CK1d, and b-actin membranes were probed with appropriate antibodies overnight. Immunocomplexes were detected using peroxidase conjugated anti-mouse or antirabbit IgG and chemiluminescence detection.

Fractionation of Cell Extracts
For subsequent fractionation cells were lysed in sucrose lysis buffer. Cell lysates were passed through a 0.4 mm filter and equal amounts of total protein were applied to an anion exchange column (Resource Q, GE Healthcare, UK) attached to an Ettan LC purifier (GE Healthcare, UK). Bound proteins were eluted with a linear ascending gradient between 0 and 1000 mM NaCl in 50 mM Tris-HCl pH 7.5, 1 mM EDTA, 5% (v/v) glycerol, 0.04% (v/v) Brij, 1 mM benzamidine, 25 mg/ml aprotinin, and

CK1d is Phosphorylated by Chk1 in vitro
Phosphorylation of the C-terminal regulatory domain of CK1d by inhibitory autocatalysis or cellular kinases plays an important role in regulating CK1d kinase activity. We previously reported that several cellular kinases are able to phosphorylate this Cterminal region and that phosphorylation of CK1d by PKA at Ser-370 significantly affects CK1d activity [74]. This site is also a potential target for phosphorylation by Chk1. Since both kinases, CK1d and Chk1, are involved in regulating cell cycle progression and p53 functions, the present study intends to demonstrate a possible interaction between Chk1 and CK1d. According to the Chk1 consensus motif published by O'Neill and colleagues [75], the C-terminal regulatory domain of CK1d contains several putative target sites for Chk1-mediated phosphorylation ( figure 1A and B). Initially, mass spectrometric examination of trypsin or GluC digested GST-CK1d 305-428 (FP1181), which was phosphorylated by Chk1 in vitro, revealed that CK1d can be phosphorylated by Chk1 at positions Ser-328, Thr-329, Ser-331, Ser-361, and Ser-382 (table 3). However, regarding the weaker results for Ser-361 and Ser-382, phosphorylation of these sites has to be confirmed by further methods.
In order to validate the potential phosphorylation sites predicted by the consensus motif or detected by the mass spectrometric analysis three sets of fusion proteins encompassing amino acids 305-375, 353-375, and 375-428 of the rat CK1d sequence were designed and created by using the templates and primers given in tables 1 and 2. Each fusion protein sequence contained potential Chk1 phosphorylation sites ( figure 1B and C). Mutations of the respective positions were introduced into the wild-type CK1d fragments of each set of fusion proteins in order to detect residues which can be phosphorylated by Chk1 in vitro (see figures 1C and 2). In order to eliminate the effects of intramolecular autophosphorylation, all fragments were created without containing any parts of the CK1d kinase domain. Quantification of Chk1mediated phosphorylation of the fusion proteins (wild-type vs. mutant) in each set provided strong evidence for phosphorylation of Ser-328, Ser-331, Ser-370, and Thr-397 while data for Thr-329 and Ser-361 are less striking. Phosphorylation of Ser-382, Thr-392, and Thr-427 could not be verified by these experiments ( figure 2).
Additionally, the significance of these findings was improved by phosphopeptide analyses using the GST-wt CK1d fusion proteins GST- In conclusion, combined data of mass spectrometric and classical biochemical approaches provide evidence for the CK1d C-terminal residues Ser-328, Ser-331, Ser-370, and Thr-397 to be targeted by Chk1 in vitro.

Chk1-mediated Phosphorylation of CK1d Shows Speciesspecific Differences
The GST-CK1d fusion proteins used in the previous section all originate from the rat CK1d sequence (FP449, [78]). However, when we try to apply the described results to human CK1d isoforms, we have to pay attention on the different C-terminal sequences. Up to amino acid position 399, the two human CK1d transcription variants are identical to the rat CK1d sequence except one exchange from isoleucine to valine at position 381. From amino acid (aa) 400 the sequences differ significantly (figure 4A). As a consequence, C-terminal GST-CK1d fusion proteins for both human CK1d transcription variants starting from aa 375 were constructed and phosphorylation intensity and phosphorylation patterns of these GST-CK1d fusion proteins upon Chk1-mediated phosphorylation were compared to that of rat CK1d. The intensity of phosphorylation by Chk1 is decreased for CK1d fusion proteins of human origin (FP1341 and FP1343) compared to that of the GST-rat CK1d fusion protein FP1183 ( figure 4B). Likewise, the phosphopeptide patterns of the analyzed variants are quite different ( figure 4C). However, further analyses are needed to determine the phosphorylated positions and to describe the functional differences between these CK1d variants.

Chk1-targeted Phosphorylation Sites Play an Important Role in Modulating CK1d Kinase Activity in vitro
In order to investigate the importance of the so far identified target sites for Chk1 within the C-terminal domain of CK1d, residues in GST-CK1d (FP449) which have been identified as targets for Chk1-mediated phosphorylation were exchanged to alanine by site-directed mutagenesis and the kinetic parameters of wild-type CK1d and the resulting CK1d single and double phosphorylation-site mutants or the triple mutant were experimentally determined using a-casein as substrate. Data shown in figure 5A and table 4 clearly show that both maximum velocity (V max ) and K m values of all analyzed phosphorylation-site mutants increased compared to those of wild-type GST-CK1d. GST-CK1d S370A and GST-CK1d S328A, S370A, T397A presented the highest K m values in this test (7.378 and 7.739 mM, respectively), indicating lowest substrate affinity. Among the tested single-site mutants the most significant change in V max could be obtained for GST-CK1d S370A .
The relevance of the identified target sites for CK1d kinase activity was furthermore indicated by co-incubating full length wild-type CK1d with activated, precipitated Chk1 (Chk1(IP)) prior to addition of the non-canonical CK1d substrate GST-b-catenin 1-181 (FP1355). Chk1(IP) alone shows no notable phosphorylation of this substrate. In contrast, GST-b-catenin 1-181 can be intensely phosphorylated by GST-CK1d. When GST-CK1d is preincubated with Chk1(IP) the GST-CK1d-mediated phosphorylation of GST-b-catenin 1-181 is markedly decreased by about 30% (figure 5B). However, this difference did not achieve statistical significance (p = 0.0565).

CK1d Phosphorylation State Influences the Effects of CK1d-specific Inhibitors
Because site-specific phosphorylation of CK1d has been shown to influence the effects of CK1-specific inhibitors [58,59] kinase assays were performed in the presence or absence of either 5 nM of compound 17 [82] or 20 nM of compound 8 [58] using GST-p53 1-64 (FP267) as substrate and GST-wt CK1d or GST-CK1d S328A, S370A, T397A as enzymes. Both wild-type and mutant GST-CK1d were inhibited to a similar extend by compounds 17 and 8. However, slightly stronger inhibitory effects for both small Next we analyzed if Chk1-mediated phosphorylation of GST-CK1d modulates the effects of the CK1d-and e-specific inhibitors D4476 [81], compound 17 [82] and compound 8 [58]. For this purpose kinase assays were performed in the presence or absence of either D4476 (300 nM), compound 17 (10 nM), or compound 8 (20 nM) using GST-p53 1-64 (FP267) as substrate and GST-wt CK1d alone or in combination with Chk1 as enzymes. Whereas Chk1-mediated phosphorylation of CK1d significantly enhanced the ability of compound 17 to inhibit CK1d kinase activity, the efficacy of D4476 was decreased in reactions containing Chk1 and CK1d. The ability of compound 8 to inhibit CK1d was only modestly affected in the presence of Chk1 (figure 6B) and did not lead to a statistically significant difference.

CK1d co-precipitates with Chk1 from HT1080 Protein Extracts
In order to analyze the relevance of the Chk1/CK1d interplay in cultured cells, the fibrosarcoma cell line HT1080 was treated with hydroxyurea (HU) to activate Chk1 [88]. As a first demonstration of a functional connection between Chk1 and CK1d a Chk1-specific antibody was used to precipitate Chk1 from protein extracts prepared from HT1080 cells which were either untreated or treated with 2.5 mM HU for 2 h. Precipitated protein was separated in SDS-PAGE and used for Western blotting. The detection of CK1d showed successful co-precipitation of CK1d with Chk1 ( figure 7). Specificity of binding was demonstrated by including a no-antibody control as well as a nonspecific control. Precipitation of Chk1 was confirmed using antibodies against total Chk1 and activated Chk1 (pChk1 S345 ). According to Walker and co-workers [89] phosphorylation of Ser-345 is mandatory for the activation of Chk1 after DNA damage

Activation of Chk1 Negatively Affects CK1d Kinase Activity in HT1080 Cells
Cellular Chk1 was activated by treating HT1080 cells with 2.5 mM HU for 2, 4, 6, or 8 h prior to the preparation of protein extracts. Control cells remained untreated. Activation of Chk1 was confirmed by detection of Ser-345 phosphorylation. Western blot analysis revealed that Ser-345 phosphorylation of Chk1 only emerged when HT1080 cells were treated with 2.5 mM HU, reaching a maximum pChk1 S345 level after four hours of treatment ( figure 8A). Expression levels of total Chk1 and CK1d remained nearly stable over the total time of the experiment. However, in vitro co-incubation of Chk1 and CK1d already demonstrated reduced activity of CK1d in reactions also containing Chk1 (see figure 5B). Activated cellular Chk1 could phosphorylate and thereby modulate CK1d activity in the same way. In order to examine this regulatory interaction in cultured cells, fractionated extracts of untreated and HU-treated cells were  analyzed for CK1-specific kinase activity using GST-p53 1-64 (FP267) as substrate for in vitro kinase reactions. Interestingly, after two hours of HU treatment CK1-specific kinase activity dramatically decreased by more than 60% compared to that of control cells ( figure 8B). At later time points, kinase activity again increased but after eight hours still exhibited only 75% of the activity detected in control extracts. Since the kinase activity in the activity peak fractions was inhibited by the CK1d-specific inhibitor compound 17 [82], the detected activity could clearly be assigned to the kinase activity of CK1d ( figure 8C). Furthermore, similar changes in cellular CK1d and Chk1 activities could be observed in immunoprecipitation kinase assays using Chk1-and CK1d-specific antibodies and specific peptide substrates (supporting figure S1). Verifying the interaction of Chk1 and CK1d within the cell, HT1080 cells were simultaneously exposed to Chk1-activating HU and the Chk1-specific inhibitor SB-218078 (10 mM; [90]) for two hours. Being compared to cells which were only treated with 2.5 mM HU the combined treatment with HU and SB-218078 showed higher CK1d-specific kinase activity possibly caused by inhibition of Chk1-mediated phosphorylation of CK1d ( figure 8D). However, treatment with 10 mM SB-218078 alone also already resulted in a decrease of CK1d-specific kinase activity (70% of control). This finding could be explained by an inhibitory effect on CK1d which could be observed at increased concentrations of SB-218078 in in vitro kinase reactions (supporting figure S2).

Discussion
CK1d, a key regulator of various cellular processes and an interesting new drug target, is tightly regulated on various levels ( [1] and references therein). On protein level reversible phosphorylation not only has a great impact on the activity of CK1d [70][71][72]74], it also modulates the effects of CK1-specific small molecule inhibitors [58,59]. The C-terminal regulatory domain of CK1d harbors phosphorylation sites, which can be targeted by several cellular kinases. Here, we provide evidence for the first time that Chk1 is one of these kinases specifically phosphorylating rat CK1d within the C-terminal domain on Ser-328, Ser-331, Ser-370, and Thr-397. Furthermore, we show that CK1d and Chk1 are cellular interaction partners.
The use of state of the art mass spectrometry, CK1d phosphorylation site mutants, and classical biochemical methods revealed that Chk1 phosphorylates Ser-328, Ser-331, Ser-370, and Thr-397 within the C-terminal domain of rat CK1d. These findings highlight the central role of Ser-370, which can be phosphorylated by a number of cellular kinases, including PKA, AKT, Clk2, and PKC-a [74]. Our results also show that not all sites containing the Chk1 consensus motif suggested by O'Neill and co-workers [75] are phosphorylated by Chk1 in vitro. Furthermore, several Chk1-targeted posphorylation sites detected by mass spectrometry could not be verified in other in vitro experiments and vice versa. Ser-370 and Thr-397 were only identified by the use of phosphorylation-site mutants in kinase reactions and two-dimensional phosphopeptide analyses highlighting the potential of these two methods. The failure to detect phosphorylation of these sites by mass spectrometry could have several reasons: (i) peptides resulting from tryptic digestion (Ser-370: VSMR; Thr-397: MSTSQR) are too small to be resolved by Nano LC-MS/MS, (ii) GST-CK1d 305-428 used as substrate was incompletely phosphorylated, or (iii) loss of phosphorylation at these sites occurred during sample processing. Additional analysis of phosphopeptides derived from GluC digestion of GST-CK1d 305-428 also failed in confirming phosphorylation of Ser-370 and Thr-397 but revealed that aa 340 to 366 did not contain any modified residue. This is in line with our phosphorylation analyses showing that Chk1 does not phosphorylate any amino acid of CK1d between Ser-332 and Ser-369.
Since shorter CK1d phosphopeptides like those containing Ser-370 and Thr-397 are likely to escape detection by Nano LC/ES-MS/MS, data from mass spectrometry should always be reviewed carefully and confirmed by alternative methods. This escape from detection could also explain why so far none of the large-scale mass spectrometry studies has shown in vivo phosphorylation of CK1d at Ser-370, a site clearly identified as target for several cellular kinases using classical methods ( [74] and this manuscript).
Although Ser-328 could be identified as a target for Chk1 by all of the employed experimental approaches, the entirety of in vitro data could not clearly point out the role of Ser-331. On the one hand Chk1-mediated phosphorylation of GST-CK1d 305-375 by Chk1 was decreased when Ser-331 is mutated to alanine and mutation of both, Ser-328 and Ser-331, is able to decrease phosphorylation intensity even below the levels detected for the respective single-site mutants. Also mass spectrometric data strongly suggest Ser-331 to be phosphorylated by Chk1. On the   (B) CK1 kinase activity in fractionated extracts from HT1080 cells before and after treatment with 2.5 mM HU for 2, 4, 6 and 8 hours, respectively, was determined using GST-p53 1-64 (FP267) as substrate. The detected kinase activity was normalized towards the untreated control. (C) Presence of CK1d in the kinase peak fractions shown in (B) was confirmed by use of the CK1d-specific inhibitor compound 17 at 50 nM [82]. (D) HT1080 cells were treated with 2.5 mM HU and/or the Chk1-specific inhibitor SB-218078 for 2 h. CK1 kinase activity in fractionated extracts was determined using GST-p53 1-64 (FP267) as substrate. The detected kinase activity was normalized towards the untreated control. doi:10.1371/journal.pone.0068803.g008

CK1d Is Phosphorylated by Chk1
We furthermore provide evidence that in addition to rat CK1d also human CK1d TV1 and 2 are phosphorylated by Chk1. Up to aa 399 all three proteins are identical and our in vitro data for rat CK1d can easily be applied to the human CK1d TV1 and 2 as well. Beyond aa 399 the primary sequences of human CK1d TV1 and 2 differ significantly from the rat sequence but also contain residues, which can be phosphorylated by cellular kinases [74,[91][92][93][94][95][96][97][98][99]. Therefore, additional analyses are necessary to identify all Chk1-targeted phosphorylation sites in the GST-human CK1d Cterminal fusion proteins GST-CK1d 375-415 TV1 and GST-CK1d 375-409 TV2 .
In order to examine the importance of Ser-328, Ser-370, and Thr-397 for the catalytic activity of CK1d in vitro, the kinetic constants of the phosphorylation of a-casein by GST-CK1d S328A , GST-CK1d S370A , GST-CK1d T397A , GST-CK1d S328A, S370A , GST-CK1d S328A, T397A , and GST-CK1d S328A, S370A, T397A were calculated experimentally. Especially GST-CK1d S370A and the tested multi-site mutants (GST-CK1d S328A, S370A , GST-CK1d S328A, T397A , and GST-CK1d S328A, S370A, T397A ) differed remarkably in their kinetic parameters compared to those of wildtype CK1d. The maximum catalytic velocity, as expressed by V max , is nearly doubled for these mutants. However, the K m values are also increasing, indicating decreased substrate affinity for these CK1d mutants. These findings illustrate a major influence of Ser-370 and Thr-397 on CK1d activity. The potential phosphosite Ser-331 was not taken into account in these experiments since its exact role has not finally been validated yet.
In a first approach to establish a direct functional connection between Chk1 and CK1d, phosphorylation of the non-canonical substrate GST-b-catenin 1-181 [79] was successfully decreased by co-incubating GST-CK1d with precipitated, activated Chk1. GST-b-catenin 1-181 in contrast to GST-p53 1-64 cannot be phosphorylated by Chk1 and the use of activated Chk1 furthermore allows to compare this result to previously published data generated using the active subunit of PKA [74]. Although Chk1 shows basal activity it can be activated five to ten-fold by phosphorylation of Ser-345 by upstream kinases (e.g. ATR) (reviewed by [100]). As published by Giamas and colleagues [74] the activity of GST-CK1d is reduced by almost 80% when being subjected to phosphorylation by the active subunit of PKA. Results obtained for the effects of activated Chk1 on CK1d kinase activity are in line with these data. However, the influence of Chk1 seems to be weaker since the observed reduction in GST-CK1dspecific activity in presence of Chk1 only accounts to approximately 30%.
In further analyses we demonstrate that mutation of Chk1 target sites in GST-CK1d as well as co-incubation of recombinant Chk1 and GST-CK1d results in modification of the effects of CK1-specific inhibitors. These differences are specific for the inhibition of CK1d because selectivity screening approaches revealed that the tested classes of inhibitors did not show any effects on the activity of Chk1 [58,82]. This finding is consistent with previous work reporting the impact of site-specific phosphorylation on inhibitor efficacy [58,59]. However, these effects are specifically dependent on the respective inhibitor molecule and its binding mode.
In addition we provide evidence that Chk1 influences CK1d activity in the cellular context. Results from co-immunoprecipitation and co-incubation experiments lead to the assumption, that a regulatory interaction between CK1d and Chk1 might also be observed in vivo. Actually, CK1d activity in HT1080 cells dramatically decreases after hydroxyurea treatment, which induces the depletion of dNTP pools resulting in S-phase arrest and Chk1 activation [88,101]. Simultaneous treatment with the Chk1-specific inhibitor SB-218078 partly reverses the HU-induced effects on cellular CK1d kinase activity. Here however, we also have to take into account the inhibitory effects of SB-218078 on CK1d activity we detected in vitro, which may falsify the presented result. Without this effect SB-218078 may potentially lead to complete reversal of the HU-induced effects.
Our cell culture results are further supported by recent phosphorylation analyses using large-scale mass spectrometry approaches showing that Ser-328, Ser-331, and Thr-397, which can be phosphorylated by Chk1 in vitro, are also phosphorylated in vivo and that phosphorylation of these sites can be observed in response to cell cycle related signaling [91][92][93][94][95][96][97][98][99].
A tight physiological interaction between Chk1 and CK1 isoforms is furthermore indicated by the fact that both kinases are implicated in regulation of DNA damage response and replication by modulating the activity of various regulatory proteins by sitespecific phosphorylation in concert. For example, phosphorylation of Cdc25A (cell division cycle 25 A) by both, CK1 (isoforms a and e) and Chk1, contributes to Cdc25A degradation and precise control of the cell cycle [102,103]. Furthermore, CK1c1 has been shown to phosphorylate residues in claspin which are important for Chk1 activation by ATR (ataxia telangiectasia and rad-3related kinase) [104]. Given these facts, a direct regulatory connection between Chk1 and CK1 in fine tuning of cell cycle regulation is quite plausible.
In summary, we clearly identified phosphorylation sites within the C-terminal domain of CK1d which can be targeted by Chk1. Furthermore, our results show for the first time that site-specific phosphorylation of CK1d by Chk1 plays an important role in modulating the activity of CK1d in vitro and most likely also within cellular signal transduction regulation. However, detailed information about this interaction remains to be determined by future work.

Supporting Information
Figure S1 Chk1-and CK1d-specific kinase activity in HT1080 cells is altered after treatment with hydroxyurea. Cellular Chk1 was activated by treating HT1080 cells with 2.5 mM hydroxyurea (HU) for the indicated periods of time. Immunoprecipitated Chk1 and CK1d were used to phosphorylate CHKtide substrate peptide (Chk1) or GST-p53 1-64 (FP267; CK1d). Substrate phosphorylation was quantified by Cherenkov counting. Data are presented as normalized bar graph. (TIF) Figure S2 CK1d is inhibited by SB-218078 in vitro. Kinase activity of rat CK1d (FP449) was assayed in presence of increasing concentrations of the Chk1-specific inhibitor SB-218078. Phosphorylation intensity of the substrate GST-p53 1-64 (FP267) was quantified by Cherenkov counting. Dose-response data were processed using GraphPad Prism 5. (TIF)