These authors contributed equally to this work.
Cyclin-dependent kinase 1 plays a critical role in DNA replication control during rat liver regeneration†
Article first published online: 10 AUG 2009
Copyright © 2009 American Association for the Study of Liver Diseases
Volume 50, Issue 6, pages 1946–1956, December 2009
How to Cite
Garnier, D., Loyer, P., Ribault, C., Guguen-Guillouzo, C. and Corlu, A. (2009), Cyclin-dependent kinase 1 plays a critical role in DNA replication control during rat liver regeneration. Hepatology, 50: 1946–1956. doi: 10.1002/hep.23225
Potential conflict of interest: Nothing to report.
- Issue published online: 20 NOV 2009
- Article first published online: 10 AUG 2009
- Accepted manuscript online: 10 AUG 2009 12:00AM EST
- Manuscript Accepted: 21 JUL 2009
- Manuscript Received: 17 JUN 2009
- Doctoral fellowship from ANRT and Fondation pour la Recherche Médicale
- Institut National de la Santé et de la Recherche Médicale
- Centre National de la Recherche Scientifique
- Ministère de l'Education Nationale de la Recherche et des Technologies. Grant Number: ACI, Jeune chercheur no. 04 5 191
- GIS de thérapie cellulaire de Rennes
- EEC FP7 program (proposal 223317 LIV-ES)
- Top of page
- Materials and Methods
- Supporting Information
Liver regeneration is a unique process to restore hepatic homeostasis through rapid and synchronous proliferation of differentiated hepatocytes. Previous studies have shown that hepatocyte proliferation is characterized by high expression levels of the “mitotic” cyclin-dependent kinase 1 (Cdk1) during S-phase compared to other mammalian cells. In the light of findings showing that Cdk1 compensates for the loss of Cdk2 and drives S-phase in Cdk2-deficient cells derived from Cdk2 knockout mice, we took advantage of the models of liver regeneration following partial hepatectomy and primary cultures of normal rat hepatocytes to further examine the involvement of Cdk1 during DNA replication in hepatocytes and to dissect specific cell cycle regulation in hepatocytes compared to control human foreskin fibroblasts. In hepatocytes, Cdk1 exhibited a biphasic activation pattern correlating S-phase and G2/M transition, bound to cyclin A or B1 and localized to the nucleus during DNA replication. Importantly, small interfering RNA (siRNA)-mediated silencing of Cdk1 led to a strong decrease in DNA synthesis without affecting centrosome duplication. Furthermore, in hepatocytes arrested by the iron chelator O-Trensox in early S-phase prior to DNA replication, Cdk1/cyclin complexes were active, while replication initiation components such as the minichromosome maintenance 7 (Mcm7) protein were loaded onto DNA. Moreover, Mcm7 expression and loading onto DNA were not modified by Cdk1 silencing. Conversely, in fibroblasts, Cdk1 expression and activation were low in S-phase and its silencing did not reduce DNA synthesis. Conclusion: Cdk1 is essential for DNA replication downstream formation of replication initiation complexes in hepatocytes but not in fibroblasts and, as such, our data exemplify crucial differences in the cell cycle regulation between various mammalian cell types. (HEPATOLOGY 2009.)
Cell cycle progression is regulated by the sequential formation, activation, and subsequent inactivation of complexes composed of structurally related serine/threonine protein kinases, the cyclin-dependent kinases (Cdks), and cyclins. In mammalian cells at least 14 Cdks and more than 30 cyclins form multiple Cdk/cyclin complexes.1 In mid-late G1 phase the commitment to S-phase is controlled by the phosphorylation status of the retinoblastoma protein (Rb) under the regulation of Cdk4/6-cyclin D complexes.2 Then, at the G1/S transition and in S-phase, Cdk2, successively associated with cyclins E and A, completes phosphorylation of Rb, promotes activation of the DNA replication machinery,3 and governs centrosome duplication.4 In contrast, the activity of Cdk1 associated with A- and B-type cyclins is required for mitosis. Therefore, Cdk2 and Cdk1 were thought to function independently at the G1/S and G2/M transitions, respectively, without functional redundancy.5 This model of cell cycle control was challenged by the finding that cancer cells proliferate despite Cdk2 inhibition,6 that knockout mice for Cdk2 are viable, and that the cell cycle of cultured Cdk2−/− mouse embryonic fibroblasts (MEFs) does not show major alterations.7, 8 These data indicated that Cdk2/cyclin E complexes were dispensable for commitment to S-phase and that other Cdk(s) compensate for the loss of Cdk2. Single and combined alterations in mice of Cdk4/6-cyclin D, Cdk2-cyclin E, p27Kip1, and Rb have been further investigated. These deletions did not affect early embryogenesis, demonstrating multiple compensatory mechanisms and the overlapping role of these genes.9, 10 Moreover, analysis of the cell cycle in MEFs derived from these knockout mice evidenced compensatory mechanisms between positive and negative regulators at the G1/S transition and highlighted a complex network regulating the expression and activation of these cell cycle regulators in the progression from G1 to S-phase. In contrast, Cdk1−/− mouse embryo development was arrested at a very early stage11 and knockdown of Cdk1 expression by short hairpin RNA (shRNA) in synchronized Cdk2−/− MEFs strongly reduced S-phase entry.12 A Cdk1-dependent compensatory mechanism controlling S-phase was also demonstrated in DT40 chicken cells lacking Cdk2.13 Together, these data have led to the proposal of a revised model of cell cycle in which Cdk1 compensates for Cdk2 ablation by controlling initiation of DNA replication and centrosome duplication.5, 14 Interestingly, it was recently demonstrated that both Cdk1 and Cdk2 were required for efficient DNA replication in Xenopus egg extracts,15 suggesting that at least in some eukaryotic cell types, Cdk1 contributes to S-phase initiation and/or DNA replication.
Following chemical-induced and viral-induced cell death or surgical partial hepatectomy (PH), liver exhibits a peculiar tissue restoration by proliferation of differentiated cells. In normal liver, mature hepatocytes are quiescent but can reenter the cell cycle after liver injury. Cytokines and growth factors activate signaling pathways and induce the entry into and progression through the G1 phase. Interestingly, the first round of hepatocyte division occurs in a synchronous manner, allowing investigation of the expression and/or activation of cell cycle regulators.16 Progression of hepatocytes through the G1 and S-phases is controlled by sequential activation of cyclins D/Cdk4, cyclin E/Cdk2, and cyclin A/Cdk2.17, 18, 19 In Cdk2−/− mice, liver regeneration occurred, although a slight decrease in the percentage of replicating hepatocytes20 and/or a delayed resumption of DNA replication were observed.21 These results demonstrated that Cdk1 also compensates for Cdk2 gene elimination in hepatocytes. However, in wildtype hepatocytes, transcriptional activation of Cdk1 gene and Cdk1 protein induction were reported during S-phase20, 22 despite Cdk2 expression and activation,18, 19 indicating that Cdk1 and Cdk2 are coexpressed during DNA replication in normal hepatocytes. This is a major difference from most mammalian cell types in which Cdk2 expression is predominant over Cdk1.5 Thus, the aim of this work was to further compare Cdk1 and Cdk2 expression and activation in normal rat hepatocytes versus human foreskin fibroblasts (HFFs) and to address the involvement of Cdk1 during DNA replication in both cell types. Here, we demonstrate that both Cdk1 and Cdk2, associated with cyclins A and/or B, are activated prior to DNA replication in regenerating hepatocytes. Moreover, knockdown experiments of Cdk1 and Cdk2 in isolated hepatocytes and HFFs demonstrate a critical role of Cdk1 in early S-phase and/or during DNA replication in hepatocytes, but not in HFFs.
Materials and Methods
- Top of page
- Materials and Methods
- Supporting Information
A two-thirds hepatectomy was performed on male Sprague-Dawley rats (200–225 g, Janvier) anesthetized by intraperitoneal administration of ketamine 100 mg/kg and xylazine 10 mg/kg bodyweight followed by injection of 1 mL of a 10% glucose solution containing buprenorphine (0.03 mg/kg) in accordance with national animal welfare regulations. They were killed under pentobarbital anesthesia. Livers were frozen in liquid nitrogen and kept at −80°C.
Cell Isolation and Culture.
Hepatocytes, isolated from adult male Sprague-Dawley rats by a two-step collagenase perfusion, were seeded at 7 × 104 cells/cm2 in complete medium (75% minimum essential medium and 25% medium 199, 10% fetal bovine serum [FBS], 100 IU/mL penicillin, 100 μg/mL streptomycin, 1 mg/mL bovine serum albumin [BSA], 2 mM L-glutamine, 5 μg/mL bovine insulin, 7 × 10−7M hydrocortisone hemisuccinate).17 After cell spreading, the culture medium was deprived of FBS and supplemented with 7 × 10−7M hydrocortisone hemisuccinate and proliferation was induced with 50 ng/mL recombinant human epidermal growth factor (EGF) (Tebu-Bio). Cultures were treated or not with 100 μM O-Trensox (TRX). HFFs maintained in Dubelcco's modified Eagle's medium (DMEM) containing 10% FBS were synchronized by FBS starvation for 3 days before release with DMEM containing 20% FBS.
For rat hepatocytes, small interfering RNAs (siRNAs) (Eurogentec) were: siCtrl (sense 5′-GAC-CAU-GGA-CAA-CCA-UAU-G99-3′), siCdk1 (duplex 1) (sense 5′-GUA-CGG-CAA-UCC-GGG-AAA-U99-3′), siCdk1 (duplex 2) (sense 5′-AGA-GUU-ACU-UGU-ACC-AAA-U99-3′), siCdk2 (duplex 1) (sense 5′-CCA-ACU-CUU-CCG-GAU-CUU-U99-3′), siCdk2 (duplex 2) (sense 5′-CUU-CUA-UGC-CUG-AUU-AUA-A99-3′). siCdk1 (duplex 2) and siCdk2 (duplex 2) were more efficient (Supporting Fig. 4). For human fibroblasts, siRNAs (Sigma-Aldrich/Proligo validated siRNA) were: siCdk1 (SASA-00044049), siCdk2 (SASI-00060175) and siCtrl (sense 5′-GUG-GCA-UGA-CUU-CAA-GAG-C99-3′).
Freshly isolated hepatocytes were reverse-transfected for 4 hours with 100 nM siRNA using Transfectin reagent (Bio-Rad). Then the medium was renewed with EGF-supplemented medium. HFFs were transfected 3 hours after releasing from FBS starvation with 100 nM siRNA using SiPORTAmine transfection reagent (Ambion). Six hours later, medium with 20% FBS was renewed.
DNA Synthesis and Mitotic Index.
Rats were injected i.p. with 5-bromo 2′deoxyuridine (BrdU) at 50 mg/kg bodyweight 2 hours before killing. BrdU-positive cells were detected using mouse anti-BrdU primary antibody (Amersham) followed by incubation with fluorescein isothiocyanate (FITC)-conjugated secondary antibody and counterstaining with Hoechst. In all, 4,000 hepatocytes/liver were scored on 10 adjacent fields and quantification of BrdU-positive cells was performed with Simple PCI 6.2.1 software. The mitotic index was determined with Hoechst staining. HFFs and hepatocytes in vitro were incubated with BrdU as indicated in the figures. Positive cells were detected using a cell proliferation kit (Amersham). For mitotic index, ≈600 cells were stained with methylene blue and Giemsa and scored.
Western Blot, Immunoprecipitation, Biochemical Fractionation, and H1 Kinase Assay.
Total proteins, immunoprecipitates or Cdk1/Cdk2 purified on p9CKShs1-sepharose-beads (p9CKShs1-beads) were resolved on 12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), transferred onto nitrocellulose membrane, and blotted using antibodies against Cdk1,22 Cdk2 (sc-163), cyclin A (sc-596), cyclin B1 (sc-245), Hsc70 (sc-7298) (from Santa Cruz), P-Tyr15Cdk1 (Stressgen, KAP-CC015), and cyclin E (US Biological, C8455-22). Minichromosome maintenance 7 (Mcm7) was detected (Santa Cruz, sc-9966) in cytosolic, soluble (nucleoplasm) and insoluble (chromatin enriched) nuclear fractions as previously described.23 Fractionation was verified by detecting nuclear serine-arginine rich proteins (SR proteins, Zymed Laboratories) and cytoplasmic GST Alpha 1/2 (Biotrin International). Cdk1, Cdk2 were purified with p9CKShs1-beads or specific immunoprecipitation (protein A-Sepharose beads, Amersham) and assayed for histone H1 kinase (Histone H1, type IIIS, Sigma) activity22 or association with cyclins by western blot.
Immunofluorescence of Cdk1 (Becton Dickinson, 610037), cyclin A (Santa Cruz, sc-751), and/or cyclin B1 (Santa Cruz, sc-595) were performed with 4% paraformaldehyde-fixed hepatocytes and γ-tubulin immunofluorescence (Sigma, GTU88) with methanol-fixed samples and using secondary rhodamine-conjugated and/or FITC-conjugated antibodies (Jackson Laboratories). Samples were counterstained with Hoechst and 200 to 500 cells were examined.
Flow Cytometry Analysis.
Cells were detached with trypsin and nuclei were stained using propidium iodide (Cycletest Plus Kit, Becton Dickinson). Data were acquired and analyzed with FACSCalibur flow cytometer (Becton Dickinson) using Cell Quest software (Becton Dickinson).
Statistical analyses were performed with SPSS software using the Mann-Whitney nonparametric test.
- Top of page
- Materials and Methods
- Supporting Information
Cdk1 Associates with Cyclins A and B1 During S-phase in Regenerating Hepatocytes.
Expressions of Cdk1, Cdk2, and their cyclin partners were studied during rat liver regeneration following PH in a detailed time-course covering the first wave of proliferation, which mainly involves hepatocytes. As reported,16, 24 DNA replication evidenced by BrdU incorporation took place between 18 and 28 hours, with a peak at 24–26 hours (Fig. 1A). A second burst of DNA replication beginning at 30–32 hours and lasting for several days involved hepatocytes and nonparenchymal cells. The low mitotic index (<1‰) in hepatectomized livers until 26 hours abruptly increased between 28 and 30 hours after PH (Fig. 1A).
Cdk1 undetectable in normal liver and during the G1 phase of hepatocytes was sharply induced between 18 and 24 hours during S-phase, followed by a transient decrease at 26–28 hours before another increase at 30 hours (Fig. 1A). Quantification of Cdk1 protein amounts indicated that at 24 hours during S-phase its expression level reached ≈80% of that found in mitosis at 30–32 hours (Fig. 1B). In contrast, Cdk2 was expressed in normal liver and during G1 phase and was slightly increased from 22 to 36 hours. Importantly, cyclin A as well as cyclin B1 showed biphasic expression peaking first in S-phase and then during mitosis. Cyclin E, detected at low levels in quiescent liver and during G1 phase, was induced at the onset of DNA replication.
Interestingly, Cdk1 exhibited two electrophoretic mobility forms during S-phase, whereas during G2/M phases the fastest migrating form was predominant. Knowing that phosphorylations affect Cdk1 electrophoretic mobility and kinase activity, expression of one of these phospho-Cdk1 forms was examined. Using an anti-P-Tyr15 Cdk1, a single band was mainly detected during S-phase with an opposite pattern to that of the lowest band detected with the anti-total Cdk1 (Fig. 1A). Because of the high conservation between Cdk1 and Cdk2 amino acid sequences around the Tyr,15 immunoprecipitations (IPs) of Cdk1 and Cdk2 were analyzed by immunoblotting with the anti-P-Tyr15 Cdk1. In 24-hour regenerating livers, anti-P-Tyr15 Cdk1 detected a single band in Cdk1 but not in Cdk2 IPs (Fig. 1C). Moreover, immunoblotting of lysates from 24 and 30 hours regenerating livers with anti-Cdk1 or -P-Tyr15 Cdk1 demonstrated that the slow migrating form of Cdk1 comigrated with P-Tyr15 Cdk1 (Fig. 1D).
IPs of Cdk1, cyclins A or B1 from regenerating liver extracts at 24 and 30 hours corresponding to S and M phases, respectively, analyzed by immunoblotting with anti-Cdk2 and anti-Cdk1 showed that Cdk2 was only associated with cyclin A at these timepoints. In contrast, Cdk1 formed complexes mainly with cyclin B1 and to a much lesser extent with cyclin A during S and M phases (Fig. 1E).
Cdk1 Is Expressed During S-phase in Mitogen-Stimulated Hepatocytes but Not in Fibroblasts.
In cultured hepatocytes stimulated with EGF the cell cycle was much longer and less synchronous compared to liver regeneration.17 However, a Cdk1 expression pattern was studied to determine whether this in vitro model could be further used for knocking down Cdk1 by siRNA. DNA replication began between 36 and 48 hours after plating and reached a maximum of ≈40% BrdU-positive hepatocytes between 60 and 72 hours, whereas mitotic activity sharply increased between 60 and 72 hours (Fig. 2A). Cdk1 expression began at 36 hours at the G1/S transition, increased at 48 hours when ≈20% of the hepatocytes replicated DNA, whereas the mitotic index was ≈5‰, and was maximal at 72 hours concomitant to the peak of mitosis. Expression of the P-Tyr15 Cdk1 form was maximal at 60 hours, correlating with active BrdU incorporation, whereas Cdk2 expression showed little variation.
In synchronized normal HFFs, Cdk1 expression was undetectable during G1 phase, low in S-phase, and strongly increased at 28 hours when cells reached G2/M phases (Fig. 2B). Cdk1 expression in S-phase represented only ≈15% of that observed in mitosis in HFFs (Fig. 2C). In contrast, Cdk2 showed little variation throughout the cell cycle. Cyclins A and B1 were not expressed in quiescent cells and during G1 phase, appearing in S and G2/M phases, respectively.
Both Cdk1 and Cdk2 Are Active During S-phase in Hepatocytes but Not in Fibroblasts.
IPs of Cdk1, Cdk2, and cyclins A, B1, or E using regenerating liver extracts were assayed for histone H1 kinase activity and compared with those obtained using synchronized HFF extracts. Cdk1 and Cdk2 activities exhibited a biphasic pattern in regenerating liver, with peaks taking place at 22–26 hours and 30–32 hours corresponding to DNA replication and mitosis, respectively (Fig. 3A). In contrast, in HFFs, Cdk1 activity was extremely low during S-phase, with a unique peak at G2/M transition, whereas Cdk2 activity was detectable as early as 20 hours, with a strong induction at 22 hours and a steady-state level thereafter (Fig. 3B). In both cell types, cyclin A-associated kinase activity appeared during S-phase with a significant increase during mitosis but cyclin B1-associated kinase activity was strictly found in G2/M phases in HFFs.
To compare the relative activities of Cdk1 and Cdk2 during S-phase, the histone H1 kinase activity in total 24-hour regenerating liver extract, measured after binding of both Cdk1 and Cdk2 to p9CKS1-beads, was compared to the remaining activity after binding to p9CKS1-beads in extracts immunodepleted in Cdk1 and/or Cdk2 (Fig. 3C and Supporting Fig. 1). Depletion in Cdk1 or both Cdk1 and Cdk2 resulted in ≈70% and ≈95% decreases in histone H1 phosphorylation level, respectively, compared to control IPs (Fig. 3C). These data demonstrate that total p9CKShs1 beads-associated kinase activity found at 24 hours in regenerating liver is due to both Cdk1 and Cdk2 in a 2:1 ratio.
Cdk1 and Cyclin B1 Are Localized in Nuclei of Diploid and Polyploid Replicating Hepatocytes.
Because polyploidization is a feature of parenchymal liver cells, Cdk1 expression during proliferation of the total hepatocyte population was questioned. In normal livers, five-cell subpopulations were identified based on DNA content (Fig. 4A and Supporting Fig. 2). Cells in S-phase represented ≈0.4%, suggesting that hepatocytes in G2 or M phases were rare and that the majority of the hepatocytes in normal livers were quiescent (G0/G1) cells either diploid (≈43%), tetraploid (≈56%), or octoploid (≈1%). In EGF-stimulated hepatocyte cultures, both diploid and tetraploid cells entered S-phase 48 hours after seeding (Fig. 4B and Supporting Fig. 2). Interestingly, nuclear localization of Cdk1 and cyclin B1 was observed in both mononuclear and binuclear hepatocytes at 48 hours (Fig. 4C), with over 60% of Cdk1 and cyclin B1-positive cells at 72 hours (Fig. 4D).
Cdk1 and Cdk2 Are Active Prior to DNA Replication in Hepatocytes Arrested in S-phase with the Iron Chelator O-Trensox.
The iron chelator TRX, a trimer of 8-hydroxyquinoline, induces cell cycle arrest in G1/S transition and/or early S-phase in hepatoma cells.25 We demonstrated that TRX also strongly inhibited DNA replication in normal rat hepatocytes in a dose-dependent manner (Fig. 5A and Supporting Fig. 3C) without affecting their progression through G1 phase. Indeed, sequential inductions of c-myc and cyclin D1 in early and mid-late G1 phase, respectively, were similar in EGF-stimulated cells treated or not by TRX (Fig. 5B). Although induction of Cdk1 was slightly delayed at 48 hours in TRX-treated cells, its expression was observed at 72 hours, demonstrating that hepatocytes progressed up to the G1/S transition in the presence of TRX. This growth arrest was not related to a toxic effect of the TRX but to its iron chelating property because inhibition of DNA replication was reversible following drug removal (Supporting Fig. 3B). To determine whether hepatocytes treated by TRX were arrested in early S-phase at 72 hours, we investigated the recruitment onto chromatin of the protein Mcm7, an essential component of the prereplication complex because Mcm2-7 proteins are loaded onto chromatin during S-phase and excluded from chromatin in G2 phase. Mcm7 was predominantly expressed and localized in the nuclear fraction of EGF-stimulated hepatocytes independently of the presence of TRX (Fig. 5C), demonstrating that TRX arrested hepatocytes in early S-phase following prereplication complex formation but prior to DNA synthesis.
To investigate whether Cdk1 was activated concomitantly to replication origin firing or during DNA synthesis, we analyzed Cdk1 protein expression and kinase activity in hepatocytes treated with EGF and TRX. Cdk1, Cdk2, and cyclin A were expressed at the same levels and the kinase activities of p9CKShs1 affinity-purified Cdk1 and Cdk2 were similar (Fig. 5D). Moreover, cyclins A, B1, and Cdk1 were localized in the nuclei of TRX-arrested hepatocytes (Fig. 5E). These data demonstrated that both Cdk1 and Cdk2 were activated at an early stage during S-phase prior to DNA replication in hepatocytes.
Cdk1 Silencing in Hepatocytes Decreases DNA Replication Without Affecting Mcm7 Loading onto DNA.
To determine the respective involvement of Cdk1 and Cdk2 in DNA replication, we transfected freshly isolated hepatocytes and synchronized HFFs with siRNAs directed against Cdk1 (siCdk1) and/or Cdk2 (siCdk2) or control siRNA (siCtrl).
In hepatocytes at 48 and 72 hours, siCdk1 and siCdk2 specifically inhibited protein expressions compared to control experiments and significantly decreased the percentage of BrdU-positive cells (Fig. 6A). Interestingly, Cdk1 repression induced a greater decrease in DNA replication than Cdk2 silencing: at 48 hours when the cell population contained no mitotic hepatocytes, DNA synthesis was reduced by 70% and 40% in siCdk1- and siCdk2-treated cultures, respectively, compared to their control counterparts. Inhibition of both Cdk1 and Cdk2 led to a decrease in DNA replication similar to that observed in siCdk1-treated cells. The inhibition of DNA synthesis by Cdk1 silencing abolished hepatocyte expansion without loss of cell viability at 72 hours (Fig. 6D) or modification of the binuclear cell number (Fig. 6E). However, important cell damage in siCdk1-treated cultures was detected at 96 hours, with numerous refringent cells and hepatocytes coiling up in condensed colonies (Supporting Fig. 5), but no variation either in caspase-3 or extracellular lactate dehydrogenase activity was noticed (data not shown).
In HFFs at 24 and 32 hours, siCdk1 and siCdk2 specifically inhibited protein expressions compared to control experiments. However, knocking-down Cdk1 alone in these cells did not reduce BrdU incorporation but affected the mitotic activity (Fig. 6B,C). DNA synthesis was decreased in HFFs after extinction of Cdk2 alone or both Cdk1 and Cdk2.
In order to determine whether prereplication complex formation were dependent on Cdk1 and/or Cdk2 activation, cytosol, soluble (nucleoplasm), and insoluble (chromatin enriched) nuclear fractions were prepared from siCdk1-treated and siCdk2-treated hepatocytes and Mcm7 loading onto chromatin was investigated. Despite the marked inhibition of Cdk1 or Cdk2 expression compared to control culture (Fig. 7A), Mcm7 levels remained unchanged in the fractions and recruitment onto chromatin was still observed (Fig. 7B). Moreover, in siCdk1/TRX-treated hepatocytes, Mcm7 expression and subcellular localization were not modified by Cdk1 silencing, confirming that Cdk1 acts downstream the formation of prereplication and/or preinitiation complexes.
S-phase Arrest by Cdk1 Silencing or TRX Treatment Is Not Associated with Inhibition of Centrosome Duplication in Hepatocytes.
It was previously shown that Cdk1 compensates for Cdk2 ablation by controlling the G1/S transition, initiation of DNA replication, and centrosome duplication.13 To explore whether Cdk1 could influence hepatocyte centrosome duplication occurring in S-phase, we established the number of centrosomes in mononuclear and binuclear hepatocytes transfected with either siCtrl or siCdk1. Following EGF stimulation the number of centrosomes increased at 72 hours compared to the one in nonstimulated cells (Fig. 8A-C). Moreover, centrosome numbers were similar after transfection with siCtrl or siCdk1 (Fig. 8A–C), demonstrating that Cdk1 extinction did not impair centrosome duplication.
Then, we determined the number of centrosomes in TRX-treated hepatocytes arrested in early S-phase but expressing active Cdk1 and Cdk2 (Fig. 8A,B). The number of centrosomes increased in a similar manner in both mononuclear and binuclear cells as early as 48 hours of treatment, confirming that in normal hepatocytes, Cdk1 was not involved in centrosome duplication.
- Top of page
- Materials and Methods
- Supporting Information
Several studies have reported that in Cdk2−/− cells Cdk1 compensates for Cdk2 ablation by controlling commitment to S-phase and/or DNA replication and centrosome duplication. However, the involvement of Cdk1 in S-phase of normal and nongenetically modified mammalian cells remains unclear. Here we show the crucial role of Cdk1 during DNA replication in normal mammalian hepatocytes containing wildtype Cdk2 alleles but not in HFFs.
In regenerating rat liver and in proliferating cultured hepatocytes, both Cdk1 and Cdk2 proteins accumulated and exhibited high levels of kinase activity in S-phase. This pattern contrasted with the low expression and kinase activity levels of Cdk1 detected during DNA replication in foreskin fibroblasts. Assuming that Cdk1 and Cdk2 kinase activities toward the substrate histone H1 were identical, we provided evidence that Cdk1 activity was twice as high as Cdk2 activity during S-phase in hepatocytes. siRNA-mediated repression of Cdk1 significantly decreased DNA replication in hepatocytes, whereas in HFFs, expressing low levels of Cdk1 during S-phase, repression of Cdk1 had no effect on the rate of DNA replication but, as expected, reduced the mitotic index. Interestingly, the greatest decrease in DNA synthesis resulted from Cdk1 rather than Cdk2 silencing in hepatocytes. These data further support and extend the conclusion that Cdk1 compensates for Cdk2 gene ablation in genetically modified mice. Indeed, in regenerating livers of Cdk2−/− mice the timing of S-phase was not altered, although the percentage of BrdU-positive cells slightly decreased compared to wild-type livers.20
Importantly, involvement of Cdk1 in early S-phase was evidenced by showing that hepatocytes arrested after G1/S transition but prior to DNA replication by the iron chelator TRX expressed fully active Cdk1 and Cdk2. Moreover, the decrease in DNA replication after Cdk1 or Cdk2 silencing was not linked to impaired formation of the prereplication complex because Mcm7 nuclear localization and loading onto chromatin were not altered. Therefore, Cdk1 may be involved in the origin of firing events downstream of the formation of replication complexes in hepatocytes, in agreement with a recent study showing that enforced expression of constitutively active Cdk1 mutant in HeLa cells resulted in abnormal origin firing and premature DNA replication in early S-phase.26
In a wild-type background, Cdk1 forms active complexes with cyclins A and B1 and controls mitosis.27 Given the expression and activation of Cdk1 during DNA replication in hepatocytes, an earlier nuclear localization of Cdk1 and its associated cyclins was postulated. We showed that Cdk1, cyclins A, and, unexpectedly, cyclin B1 were indeed localized in the nucleus of replicating hepatocytes and formed active complexes during S-phase, supporting their implication in DNA replication. In mammalian cells, Cdk1/cyclin B1 complexes localize in the cytoplasm during G2 phase,28 phosphorylate cytoplasmic substrates, and trigger nuclear envelop breakdown and mitosis. Although the absolute requirement of cytosolic cyclin B1 during initiation of mitosis remains questioned, it has been postulated that relocating cyclin B1 to the nucleus in S-phase might compromise entry into mitosis.29 This would explain why the accumulation of nuclear Cdk1/cyclin B1 complexes during DNA replication does not trigger premature mitosis in hepatocytes. Moreover, P-Tyr15 Cdk1 found in replicating hepatocytes and known to be an inactive form of Cdk1 could also participate in this control. Indeed, a recent report showed that Tyr15 phosphorylation of Cdk1 is important to avoid premature entry into mitosis.30 Regulation of the ratio between pools of active and inactive Cdk1 in hepatocytes might be essential to allow S-phase initiation while preventing premature mitosis. Taken together, our data dispel any doubt on the contribution of Cdk1 to promote S-phase along with Cdk2 in normal adult hepatocytes, supporting the idea that Cdk1 might contribute to initiation of DNA replication at least in some wildtype mammalian cells. In contrast, in presence of Cdk2, Cdk1 is not directly implicated in the control of centrosome duplication in hepatocytes, as Cdk1 silencing and treatment by TRX resulted in normal centrosome duplication.
Our data also highlight the conclusion that Cdk1 plays a critical role during S-phase in hepatocytes but not in HFFs. Indeed, a major difference between these two cell types is the level of Cdk1 expression during S-phase. The peculiar biphasic pattern of Cdk1 activity in hepatocytes in which evenly active Cdk1 and Cdk2 coexist during S-phase contrasts with most mammalian cell types in which active Cdk2 is highly predominant over Cdk1.5 Moreover, in DT40 chicken cells expressing low levels of active Cdk1 in S-phase, elimination of Cdk2 induced a Cdk1-dependent S-phase but the presence of a single Cdk2 allele rendered the S-phase independent of Cdk113 suggesting that Cdk1 and Cdk2 are functionally exclusive at the level of kinase activity in such eukaryotic cells.
Why do hepatocytes exhibit an unusual pattern of active Cdk1? One feature of the liver is the tetraploid status of more than 50% of the hepatocytes.31 Here we demonstrated that Cdk1 was active in all hepatocytes regardless of their ploidy status, excluding a peculiar regulation or role of Cdk1 during S-phase related to their polyploid status. Another characteristic of hepatocytes is their singular capacity to proliferate despite their high differentiated status. This ability has been related to the low expression level of p21CIP1 Cdk-inhibitors in adult liver and primary hepatocytes in vitro,32 which could explain the rapid exit from quiescence and high level of active Cdk1 following G1/S transition. In addition, it was also recently shown that, whereas Cdk1 compensates for Cdk2 ablation during DNA synthesis, Cdk2 was important for proper repair of DNA damage in MEFs and regenerating livers.21 The question is thus raised as to whether the ability of Cdk1 to stimulate DNA replication is a peculiar molecular mechanism of cell cycle control during proliferation of some mature differentiated cells following stress stimuli. Indeed, in quiescent T lymphocytes, which reenter the cell cycle after specific immune stimulation, inhibition of Cdk1 expression by antisense oligodeoxynucleotides also decreased DNA synthesis.33 Moreover, we previously demonstrated that in a long-term coculture model of adult hepatocytes, Cdk1 expression at the G1/S transition was triggered by a transient extracellular matrix remodeling induced by the proinflammatory cytokines TNFα and interleukin-6.34 In conclusion, we hypothesize that the ability of Cdk1 kinase to regulate S-phase initiation prior to DNA replication might serve liver regeneration by allowing rapid and efficient hepatocyte proliferation in order to maintain liver homeostasis and metabolic requirements following injury signals and tissue loss.
- Top of page
- Materials and Methods
- Supporting Information
We thank the imaging cellchip platform (ImpacCell) from Biogenouest and the cytometry platform from IFR 140. We also thank Drs. G. Lescoat (INSERM U522, Rennes, France) and P. Baret (CNRS UMR5616, Grenoble, France) for providing the iron chelator O-Trensox.
- Top of page
- Materials and Methods
- Supporting Information
- 1Mammalian cyclin-dependent kinases. Trends Biochem Sci 2005; 30: 630–641., .
- 2Cyclin-dependent kinases: engines, clocks, and microprocessors. Annu Rev Cell Dev Biol 1997; 13: 261–291..
- 3G1 phase progression: cycling on cue. Cell 1994; 79: 551–555..
- 4Cyclin-dependent kinase 2 (Cdk2) is required for centrosome duplication in mammalian cells. Curr Biol 1999; 9: 429–432., , .
- 5Cdk1: the dominant sibling of Cdk2. Nat Cell Biol 2005; 7: 779–781., .
- 6Proliferation of cancer cells despite CDK2 inhibition. Cancer Cell 2003; 3: 233–245., .
- 7Cyclin-dependent kinase 2 is essential for meiosis but not for mitotic cell division in mice. Nat Genet 2003; 35: 25–31., , , , , , et al.
- 8Cdk2 knockout mice are viable. Curr Biol 2003; 13: 1775–1785., , , , .
- 9Rb/Cdk2/Cdk4 triple mutant mice elicit an alternative mechanism for regulation of the G1/S transition. Proc Natl Acad Sci U S A 2009; 106: 486–491., , , , .
- 10Cell cycle, CDKs and cancer: a changing paradigm. Nat Rev Cancer 2009; 9: 153–166., .
- 11Cdk1 is sufficient to drive the mammalian cell cycle. Nature 2007; 448: 811–815., , , , , , et al.
- 12Cdc2-cyclin E complexes regulate the G1/S phase transition. Nat Cell Biol 2005; 7: 831–836., , .
- 13An essential role for Cdk1 in S phase control is revealed via chemical genetics in vertebrate cells. J Cell Biol 2007; 178: 257–268., , , , , , et al.
- 14Cell cycle sibling rivalry: Cdc2 vs. Cdk2. Cell Cycle 2005; 4: 1491–1494., .
- 15Cdk1 and Cdk2 activity levels determine the efficiency of replication origin firing in Xenopus. EMBO J 2008; 27: 758–769., , , , , , et al.
- 16Liver regeneration. Science 1997; 276: 60–66., .
- 17Growth factor dependence of progression through G1 and S phases of adult rat hepatocytes in vitro. Evidence of a mitogen restriction point in mid-late G1. J Biol Chem 1996; 271: 11484–11492., , , , , .
- 18Involvement of p21 and p27 in the regulation of CDK activity and cell cycle progression in the regenerating liver. Oncogene 1998; 16: 2141–2150., , , , , .
- 19Activation of cdk4 and cdk2 during rat liver regeneration is associated with intranuclear rearrangements of cyclin-cdk complexes. HEPATOLOGY 1999; 29: 385–395., , , , .
- 20Genetic substitution of Cdk1 by Cdk2 leads to embryonic lethality and loss of meiotic function of Cdk2. Development 2008; 135: 3389–3400., , , , , .
- 21p21 inhibits Cdk1 in the absence of Cdk2 to maintain the G1/S phase DNA damage checkpoint. Mol Biol Cell 2008; 19: 65–77., , .
- 22Expression and activation of cdks (1 and 2) and cyclins in the cell cycle progression during liver regeneration. J Biol Chem 1994; 269: 2491–2500., , , , , .
- 23Cyclin A-CDK activity during G1 phase impairs MCM chromatin loading and inhibits DNA synthesis in mammalian cells. Cell Cycle 2008; 7: 2179–2188., , .
- 24Liver regeneration. J Hepatol 2000; 32: 19–31..
- 25Antiproliferative and apoptotic effects of O-Trensox, a new synthetic iron chelator, on differentiated human hepatoma cell lines. Carcinogenesis 2000; 21: 943–951., , , , , , et al.
- 26Cyclin A-Cdk1 regulates the origin firing program in mammalian cells. Proc Natl Acad Sci U S A 2009; 106: 3184–3189., , , , , , et al.
- 27Unmasking the S-phase-promoting potential of cyclin B1. Science 2003; 300: 987–990., , .
- 28The differential localization of human cyclins A and B is due to a cytoplasmic retention signal in cyclin B. EMBO J 1994; 13: 3772–3781., .
- 29Cyclin-dependent kinases and cell-cycle transitions: does one fit all? Nat Rev Mol Cell Biol 2008; 9: 910–916., , .
- 30Rapid cycling and precocious termination of G1 phase in cells expressing CDK1AF. Mol Biol Cell 2008; 19: 3426–3441., ,
- 31Liver cell polyploidization: a pivotal role for binuclear hepatocytes. J Biol Chem 2003; 278: 19095–19101., , , , , .
- 32Regulation and role of p21 and p27 cyclin-dependent kinase inhibitors during hepatocyte differentiation and growth. Am J Physiol Gastrointest Liver Physiol 2003; 285: G115–127., , , , .
- 33Cdc2 gene expression at the G1 to S transition in human T lymphocytes. Science 1990; 250: 805–808., , , , .
- 34TNFalpha-mediated extracellular matrix remodeling is required for multiple division cycles in rat hepatocytes. HEPATOLOGY 2005; 41: 478–486., , , , , , et al.
- Top of page
- Materials and Methods
- Supporting Information
Additional Supporting Information may be found in the online version of this article.
|HEP_23225_sm_SupFig1.tif||96K||Supporting Figure 1: Cdk1 immuno-depletion. Five successive immunoprecitations of Cdk1 were performed using liver extracts 24h post-PH (IP Cdk1, lanes 1 to 5), and lysates after immunodepletion of Cdk1 (Cdk1 dep.) were immunoblotted with anti-Cdk1 and anti-Cdk1+Cdk2 antibodies. Lysates from 24 and 36h post-PH were loaded as positive control.|
|HEP_23225_sm_SupFig2.tif||142K||Supporting Figure 2: DNA content in isolated and cultured hepatocytes. (A) Flow cytometry analysis of nuclear DNA stained with propidium iodide in freshly isolated rat hepatocytes (T0) and in hepatocytes stimulated (+) or not (-) with EGF for 48 and 72h. Single cells were gated on dot plots FL2-W versus FL2-A and DNA content of these single cells was determined on FL2-A histograms. Five regions (M) were selected according to DNA content in diploid and tetraploid hepatocytes. M1: diploid cells in G1 phase, M2: diploid cells in S phase, M3: diploid cells in G2/M and tetraploid cells in G1 phase, M4: tetraploid cells in S phase, M5: tetraploid cells in G2/M and octoploid cells in G1 phase. (B) Percentages of each sub-population (M1 to 5) in freshly isolated hepatocytes according to DNA content. Results represent means± SD (n=3).|
|HEP_23225_sm_SupFig3.tif||63K||Supporting Figure 3: Hepatocyte cell cycle arrest by the iron chelator O-Trensox. (A) [3Hmethyl]Thymidine (Amersham, 25 Ci/mmole) incorporation (24h incubation) was measured at 48 and 72h of culture in EGF-stimulated hepatocytes exposed or not to increasing concentrations (5 to 100?M) of O-Trensox (TRX). Results represent means± SD (n=3). (B) [3Hmethyl]Thymidine incorporation (6h incubations) was measured in EGF-stimulated hepatocytes exposed to 100μM of O-Trensox (TRX) for 48h before TRX removal. Untreated cultures were performed as control (CTRL). (n=3).|
|HEP_23225_sm_SupFig4.tif||77K||Supporting Figure 4: Identification of potent siRNAs targeting Cdk1 or Cdk2. Hepatocytes stimulated with EGF were transfected with control siRNA that did not affect Cdk1 and Cdk2 expression (siCtrl), or two siRNAs directed against Cdk1 (siCdk1 duplexes n°1 and 2) or Cdk2 (siCdk2 duplexes n°1 and 2). (A) Proliferation of hepatocytes inhibited for Cdk1 expression was measured at the indicated times after 24h BrdU incorporation was expressed as percentage of BrdU-positive cells reported to initial population (number of cells at 48h after transfection with siCtrl). Results represent means± SD (n=3). Cell extracts were immunoblotted for Cdk1 (A) or Cdk2 (B) at the indicated times to evaluate inhibition.|
|HEP_23225_sm_SupFig5.tif||491K||Supporting Figure 5: Hepatocyte morphology following transfection. Phase contrast photographs of hepatocyte cultures stimulated with EGF 48, 72 and 96h after transfection with control siRNA (siCtrl) or with siRNAs directed against Cdk1 (siCdk1) or Cdk2 (siCdk2). Scale bars: 100 μm.|
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.