S100A11 is involved in the regulation of the stability of cell cycle regulator p21CIP1/WAF1 in human keratinocyte HaCaT cells

Authors

  • Franziska Foertsch,

    1. Biomolecular Photonics Group, Jena University Hospital, Germany
    Search for more papers by this author
    • These authors contributed equally to this work
  • Nicole Teichmann,

    1. Institute of Human Genetics, Jena University Hospital, Germany
    Current affiliation:
    1. Second Department of Internal Medicine, Technical University of Munich, Germany
    Search for more papers by this author
    • These authors contributed equally to this work
  • Robert Kob,

    1. Institute of Human Genetics, Jena University Hospital, Germany
    Current affiliation:
    1. Institut für Biomedizin des Alterns, Friedrich-Alexander-Universität Erlangen-Nürnberg, Germany
    Search for more papers by this author
  • Julia Hentschel,

    1. Institute of Human Genetics, Jena University Hospital, Germany
    Search for more papers by this author
  • Ulrike Laubscher,

    1. Institute of Human Genetics, Jena University Hospital, Germany
    Current affiliation:
    1. Membrane Trafficking Group, Fritz Lipmann Institut (FLI) – Leibniz Institute for Age Research, Jena, Germany
    Search for more papers by this author
  • Christian Melle

    Corresponding author
    1. Biomolecular Photonics Group, Jena University Hospital, Germany
    • Correspondence

      C. Melle, AG Biomolekulare Photonik, Universitätsklinikum Jena, 07740 Jena, Germany

      Fax: +49 (0)3641 935518

      Tel.: +49 (0)3641 934849

      E-mail: christian.melle@mti.uni-jena.de

    Search for more papers by this author

Abstract

The cyclin-dependent kinase inhibitor p21CIP1/WAF1 is a regulatory factor of the cell cycle. Its transcriptional activation and protein stability are tightly controlled by several distinct mechanisms. S100A11 is a member of the S100 family of Ca2+-binding proteins involved in several biological processes, including cell cycle progression and signal transduction. In the present study, we show that down-regulation of S100A11 results in the reduction of p21 protein in human HaCaT keratinocytes. It appears that a ubiquitin-independent proteasomal degradation process is involved in p21 degradation in S100A11 down-regulated cells. The application of a proteasome inhibitor stabilized p21 protein in these cells. Analysis of distinct signal transduction pathways revealed a disturbed phosphatidylinositol-3-kinase/Akt pathway after S100A11 knockdown. We determined that the glycogen synthase kinase-3, which is negatively regulated by phosphatidylinositol 3-kinase/Akt, was activated in cells possessing knocked-down S100A11 and appears to be involved in p21 protein destabilization. The application of a specific inhibitor of glycogen synthase kinase 3 resulted in an increase of the p21 protein level in S100A11 down-regulated HaCaT cells. Glycogen synthase kinase 3 is able to phosphorylate p21 at T57, which induces p21 proteasomal turnover. Mutation of the glycogen synthase kinase 3 site threonine 57 into alanine (T57A) stabilizes p21 in HaCaT cells lacking S100A11. Beside decreased p21 protein, down-regulation of S100A11 triggered the induction of apoptosis in HaCaT cells. These observations suggest that S100A11 is involved in the maintenance of p21 protein stability and appears to function as an inhibitor of apoptosis in human HaCaT keratinocyte cells. Thus, the data shed light on a novel pathway regulating p21 protein stability.

Abbreviations
caspase inhibitor Z-Asp-CH2-DCB

Z-Asp-2,6-di-chlorebenzoyloxymethylketone

CDI

cyclin-dependent kinase inhibitor

CDK

cyclin-dependent kinase

FACS

fluorescence activated cell sorting

GSK3

glycogen synthase kinase 3

nsc siRNA

nonsilencing control small interfering RNA

PARP

poly(ADP-ribose) polymerase

PI3K

phosphatidylinositol 3-kinase

Introduction

The cell cycle is controlled by several cyclin-dependent kinases (CDKs). CDKs are governed by both activating cyclins and suppressing CDK inhibitors (CDIs). Thereby, the dynamic activities of CDKs in the cell cycle are affected by the expression and degradation of cyclins and CDIs. When the precise interplay between CDK activators and inhibitors is disturbed, unscheduled cell cycle progression occurs [1]. A main regulator of the cell cycle is the CDK inhibitor p21CIP1/WAF1 (CDKN1A), which is able to arrest cells in the G1 and G2 phases of the cell cycle [2, 3]. Thereby, p21CIP1/WAF1 (hereafter referred to as p21) contributes to temporal pauses that facilitate the integration of critical checkpoint signals for regulated entry into S phase or mitosis, respectively. The expression of p21 can be triggered by several cellular stress stimuli, such as DNA damage. In general, the tumour suppressor protein p53 is responsible for the transcriptional activation of p21/CDKN1A [4]. Induced expression of p21 inhibits DNA replication by binding of proliferating cell nuclear antigen and impedes entry into mitosis [5]. In addition to transcriptional activation of CDKN1A, regulation of p21 also occurs in a post-transcriptional manner [6]. The protein stability of p21 can be tightly controlled by ubiquitin-dependent proteasomal degradation, involving the E3 ligases SCFSKP2 and APC/CCDC20 in different cell cycle phases [7, 8]. Turnover of p21 protein also occurs through additional specific processes, such as activation of caspases or ubiquitin-independent proteasomal degradation [9-11]. This involves the direct binding of the C-terminal region of p21 to the C8α subunit of the 20S proteasome [12]. Additionally, phosphorylation of p21 by various kinases at several consensus sites also regulates the stability of the protein. Phosphorylation events are also responsible for the subcellular localization of p21. Extracellular signal-regulated kinase 2 phosphorylates p21 at threonine 57 and serine 130, which results in cytoplasmic translocation and Ub-dependent degradation of p21 triggering cell cycle progression [13]. By contrast, the phosphorylation of p21 at serine 130 by p38 MAPK and c-Jun N-terminal kinase 1 leads to its stabilization [14]. A phosphorylation of p21 threonine 145 by Akt leads to the translocation of p21 into the cytoplasm, whereas Akt-dependent phosphorylation of p21 at serine 146 increases its stability [15, 16]. A further contrary effect with respect to p21 protein stability appears when p21 is phosphorylated at identical sites by distinct kinases. Both glycogen synthase kinase 3 (GSK3) and c-Jun N-terminal kinase phosphorylate p21 at threonine 57, which results in the degradation or stabilization of p21, respectively [14, 17, 18]. Furthermore, GSK3β can be activated by UV irradiation and phosphorylates p21 at serine 114, triggering its degradation by the proteasome [19].

Recently, we detected a marked decrease of p21 in HaCaT cells upon RNA interference-mediated knockdown of the Ca2+-binding protein S100A11 [20]. S100A11 belongs to the family of S100 proteins that are considered to be multitasking proteins involved in several biological processes, such as the Ca2+-signalling network, cell growth and motility, cell cycle progression, transcription and cell differentiation [21, 22]. S100 proteins are expressed in a cell- and tissue-specific manner. In several studies, S100A11 has been detected to be up- or down-regulated in different tumour entities [23]. S100A11 plays a dual role in growth regulation of human keratinocytes because it is able to mediate a Ca2+-induced growth inhibition, as well as growth stimulation, by enhancement of the level of epidermal growth factor protein family members [24, 25]. In the present study, we determined the underlying cellular mechanisms through which S100A11 appears to influence p21 protein turnover. We have identified that S100A11 knockdown deregulates Akt, which leads to GSK3 activation, resulting in p21 proteasomal degradation.

Results

S100A11 down-regulation induces a decrease in the p21 protein level in HaCaT cells

In a recent study, we detected a correlation between down-regulation of the Ca2+-binding protein S100A11 and a decrease of the p21 protein signal in HaCaT keratinocytes [20]. To determine the underlying cellular mechanism, we treated human HaCaT cells, carrying mutations in both p53 alleles [26], with several specific small interfering RNAs (siRNAs) against S100A11 and analyzed the protein level of p21 in these cells. Thereby, we detected a significant decrease of endogenous p21 protein abundance in S100A11-depleted cells compared to controls (Fig. 1A). All three single S100A11 siRNAs (as used to exclude off-target effects) caused a strong decrease in the p21 protein signal (Fig. S1). We exclude any protein interaction between S100A11 and p21 as a reason for p21 stabilization because no S100A11/p21 complex was detectable in co-immunoprecipitation experiments using either a specific S100A11 antibody or a specific p21 antibody for protein complex precipitation (data not shown). Because HaCaT cells carried mutations in both alleles of TP53 encoding the tumour suppressor protein p53, which is a well known transcriptional activator of the p21 gene, we assumed that the alteration of the amount of p21 must be controlled by post-transcriptional mechanisms. Consistent with our assumption, analysis of the p21 mRNA level of these cells revealed that transcriptional expression of CDKN1A was unaffected by S100A11 siRNA treatment in HaCaT cells (Fig. 1B). To exclude the possibility of translational effects on the deregulation of the p21 protein level in HaCaT cells after S100A11 knockdown, we carried out a cycloheximide chase experiment to inhibit translation. In the absence of translation, siRNA-mediated knockdown of S100A11 in HaCaT cells still caused a reduction in p21 protein levels (Fig. 1C), indicating that S100A11-dependent deregulation of p21 abundance is not a result of a decreased efficiency of protein synthesis. The reduction in p21 levels observed in nonsilencing control siRNA (nsc siRNA)-treated samples in the presence versus the absence of cycloheximide is likely to be a consequence of the half-life of p21, previously reported to be 30 min by Lee et al. [19].

Figure 1.

Down-regulation of S100A11 induces a decreased p21 protein abundance in HaCaT keratinocytes. (A) Left: Protein extracts of HaCaT cells treated with a specific siRNA against S100A11 (lane 2) or nonspecific nsc siRNA (lane 1) were analyzed by immunoblotting against S100A11 and p21 using specific antibodies. As a control for equal protein loading, corresponding actin levels were shown by immunoblotting. Right: Intensities of the protein bands corresponding to p21 relative to actin were densitometrically evaluated in four independent immunoblots. The mean of the p21 band intensities measured in control cells transfected with nsc siRNA was set to 100%. Data are shown as the mean ± SD (**P ≤ 0.01). (B) Total RNA of HaCaT cells treated with specific S100A11 siRNA or nonspecific nsc siRNA was extracted from the cells and p21 gene expression was analyzed by real-time PCR. The CT values of p21 were normalized to actin. Three independent experiments were performed and data are shown as the mean ± SD. (C) To inhibit translational mechanisms in HaCaT cells transfected with S100A11 siRNA or nsc siRNA, cells were treated with 50 μg·mL−1 cycloheximide for the indicated time periods. Protein extracts of treated cells were subjected to immunoblotting and analyzed against S100A11 and p21 using specific antibodies. Actin was used in the immunoblot as a control for equal protein loading.

Knockdown of S100A11 activates caspase 3/7

Because the p21 protein level is massively decreased and, additionally, the proliferation capacity is impeded in S100A11 down-regulated HaCaT cells [20], we suspected a deregulated cell cycle in HaCaT cells lacking S100A11. For this reason, we assessed the distribution of cells in distinct cell cycle phases by fluorescence activated cell sorting (FACS) analysis. When HaCaT cells were subjected to specific S100A11 siRNA, a significant increase in the subG1 cell fraction appeared, which was approximately five-fold higher compared to control cells treated with nsc siRNA (Fig. 2A). One explanation for cell cycle progression failing after S100A11 down-regulation might be the deregulation of several additional cell cycle regulators beside p21. Factors such as distinct cyclins, CDK4 and E2F3, which are commonly involved in G1 transition, were decreased in HaCaT cells lacking S100A11 (Fig. S2).

Figure 2.

Activation of caspase 3/7 by S100A11 knockdown in HaCaT cells. (A) Distribution of untreated HaCaT cells or cells transfected with S100A11 siRNA or nsc siRNA in cell cycle phases was determined by FACS analysis 72 h after transfection. Distribution of cells in cell cycle phases was assessed in three independent experiments and is shown as the mean ± SD. (B) HaCaT keratinocytes treated with S100A11 siRNA or nsc siRNA were lysed and subjected to immunoblotting against PARP and S100A11. A signal corresponding to cleaved PARP (labelled by an arrow) appeared in cells transfected with S100A11 siRNA (lane 2) that is absent in nsc siRNA treated control cells (lane 1). Actin was used as a loading control. (C) The induction of caspase 3/7 activity in HaCaT cells transfected with S100A11 siRNA was analyzed. Caspase 3/7 activity measured in control cells treated with nsc siRNA was set to 100%. Data of three independent experiments are shown as the mean ± SD. (D) Left: HaCaT cells transfected with S100A11 siRNA (lanes 2 and 4) or nsc siRNA (lanes 1 and 3) were treated with the specific caspase inhibitor Z-Asp-CH2-DCB (lanes 3 and 4) or dimethylsulfoxide (lanes 1 and 2) and analyzed by immunoblotting against S100A11, p21 and PARP using specific antibodies. As a control of equal protein loading, the corresponding actin level is shown. In S100A11 down-regulated cells treated with the caspase inhibitor for 8 h (lane 4), only a slightly increased p21 signal was detectable compared to control cells lacking S100A11 treated with dimethylsulfoxide alone (lane 2). In control cells transfected with nsc siRNA, a signal corresponding to p21 was clearly detectable independent of treatment with the caspase inhibitor (lanes 1 and 3). Right: The distribution of the cells in distinct cell cycle phases was assessed in three independent experiments by FACS analysis.

Because the subG1 cell fraction contains both necrotic and apoptotic cells, we were interested in whether caspase activity (an inducer of apoptosis) is increased in S100A11 down-regulated HaCaT cells. A substrate for caspase 3 is the poly(ADP-ribose) polymerase (PARP), specific cleavage of which produces a distinct product [27], allowing PARP to be used as an apoptotic marker [28]. To test a possible induction of caspase activity by S100A11 down-regulation, we analyzed crude extracts of cells treated with S100A11 siRNA using an antibody against PARP that is also able to detect cleaved PARP product. Full-length PARP was found to disappear almost completely. Concurrently, a specific product of caspase 3/7 activity, the cleaved PARP product, appeared in the immunoblots (Fig. 2B). When S100A11 knockdown cells were tested directly for caspase capacity using a caspase activity assay, a significant increase in caspase 3/7 activity was detectable compared to controls treated with nsc siRNA (Fig. 2C). The activation of caspase 3/7 was reversed using a specific caspase 3/7 inhibitor (Z-Asp-2,6-di-chlorebenzoyloxymethylketone; Z-Asp-CH2-DCB). No increase of the p21 level was detected in S100A11 down-regulated HaCaT cells treated with the caspase inhibitor dissolved in dimethylsulfoxide compared to control cells that were only treated with S100A11 siRNA and dimethylsulfoxide alone without the caspase 3/7 inhibitor Z-Asp-CH2-DCB (Fig. 2D). The treated samples, analyzed by FACS to determine the distribution of cells in distinct cell cycle phases, revealed a clear decrease of the subG1 cell fraction in the sample treated with both S100A11 siRNA and the caspase inhibitor compared to the control sample that was only treated with S100A11 siRNA. The results of the immunoblot analysis of p21 suggest that, despite Z-Asp-CH2-DCB-mediated inhibition of caspase activity being reflected by a decreased subG1 cell fraction, the inhibition of caspase failed to stabilize the p21 protein. It should be noted that the content of the subG1 cell fractions in control samples is unexpectedly high. We suspect that the treatment with both dimethylsulfoxide and siRNA is responsible for this phenomenon.

Hence, S100A11 appears to impede the activation of caspases, which might explain both caspase activation and apoptosis induction in S100A11 knockdown cells. By contrast to the previously described involvement of caspase 3 in p21 protein destabilization reflected in the appearance of a 15-kDa p21 cleavage product in HEK293 and A549 cells [9], which was never detected in the cells that we used, caspase 3/7 activity does not appear to be important for destabilization of p21 after down-regulation of S100A11 in HaCaT keratinocytes.

Proteasomal activity participates in p21 degradation after S100A11 down-regulation in a ubiquitin-independent manner

As shown above, a reconstitution of the p21 level was not possible by caspase inhibition in S100A11 down-regulated HaCaT cells. For this reason, at least one further mechanism is likely to be involved in p21 degradation in HaCaT keratinocytes possessing knocked-down S100A11. In addition to the proteasomal degradation of p21 involving E3 ubiquitin ligases, an ubiquitin-independent proteasomal degradation mechanism exists [29]. Hence, we assessed the p21 protein level in S100A11 down-regulated HaCaT cells when the proteasomal system was impeded. MG-132, which was used in experiments, is a specific inhibitor of the protease activity of both the 20S and 26S proteasome [30]. For this reason, HaCaT cells transfected with specific S100A11 siRNA or nonspecific nsc siRNA were treated 64 h post-transfection with 10 μm MG-132 for 8 h to impede proteasomal activity. Interestingly, a significant increase in the p21 level occurred when S100A11 knockdown cells were treated with MG-132 compared to S100A11 down-regulated control cells (Fig. 3A,B).

Figure 3.

S100A11 down-regulation induces ubiquitin-independent proteasomal degradation of p21. HaCaT cells were transfected with S100A11 siRNA (lanes 2 and 4) or nsc siRNA (lanes 1 and 3) for 64 h and subsequently subjected to the proteasome inhibitor MG-132 (10 μm) (lanes 1 and 2) or, as a control, to dimethylsulfoxide alone (lanes 3 and 4) for 8 h. Protein extracts of treated cells were analyzed by western blotting against S100A11 and p21 using specific antibodies. As a control for equal protein loading, corresponding actin levels were shown by immunoblotting. (B) Densitometric analysis of the impact of proteasomal mechanisms with respect to p21 protein level relative to actin in HaCaT cells transfected with S100A11 siRNA or nsc siRNA and treated with the proteasome inhibitor MG-132 or dimethylsulfoxide alone in three independent experiments. Data are shown as the mean ± SD (*P ≤ 0.05). (C) Left: Protein extracts of HaCaT cells transfected with S100A11 siRNA (lanes 2 and 4) or nsc siRNA (lanes 1 and 3) were generated and subsequently used for precipitation of p21 by a specific anti-p21 serum. As a negative control, normal mouse IgG was used for precipitation. Samples precipitated by p21 antibody (lanes 3 and 4) or IgG (lanes 1 and 2), respectively, were subjected to western blotting and analyzed by anti-ubiquitin serum. Loaded p21 was detected in the immunoblot by anti-p21 serum. Right: p21 was immunoprecipitated from S100A11 knocked-down HaCaT cells treated with MG-132 or, as a control, with dimethylsulfoxide for 4 h. The precipitated p21 was analyzed by immunoblotting using anti-ubiquitin serum to determine the level of possible accumulation of ubiquitinated p21. As a loading control, precipitated p21 is shown. (D) HaCaT cells were transfected with S100A11 siRNA or nsc siRNA. Treated cells were harvested 72 h after transfection and the resulting cell crude extracts were analyzed by immunoblotting against S100A11, p21, SKP2 and CDC20 with specific antibodies. As a control for equal protein loading, corresponding actin levels were shown by immunoblotting.

Both ubiquitin-dependent and -independent proteasomal processes are involved in p21 protein degradation. To determine whether p21 protein destabilization after S100A11 knockdown requires p21 ubiquitination, p21 was precipitated from cell extracts derived from HaCaT cells treated with S100A11 siRNA or nsc siRNA, respectively, followed by immunoblotting against ubiquitin. As shown in Fig. 3C, there was hardly any accumulation of ubiquitinated p21 in S100A11 knocked-down HaCaT cells compared to control cells treated with nsc siRNA, suggesting that the proteasomal degradation of p21 induced by S100A11 down-regulation occurs in an ubiquitin-independent manner. Furthermore, when p21 was precipitated from protein extract derived from S100A11 down-regulated HaCaT cells treated with the proteasome inhibitor MG-132, it appears that a slight accumulation of ubiquitinated p21 appeared in contrast to control cells lacking S100A11. Because p21 appears to be stabilized in cells pretreated with MG-132, an increased amount of p21 compared to control cells was precipitated using the same volume of anti-p21 serum-loaded beads in the precipitation approaches. This is reflected by different p21 signal intensities in the western blot (Fig. 3C, right), which might be one reason for the obvious accumulation of ubiquitinated p21 in S100A11 knocked-down HaCaT cells treated with MG-132. These results are consistent with the observation that E3 ligases SCFSKP2 and APC/CCDC20, which are commonly involved in normal turnover of p21, were also down-regulated in HaCaT cells possessing down-regulated S100A11 (Fig. 3D).

S100A11 mediates p21 protein stability by influencing the phosphatidylinositol 3-kinase (PI3K)/Akt pathway

To determine the molecular mechanism through which S100A11 is involved in p21 protein stabilization, we examined several distinct signal transduction pathways. When we assessed the PI3K/Akt pathway for possible effects of S100A11 down-regulation in HaCaT keratinocytes, we detected a significant, albeit not complete, dephosphorylation of Akt (also named protein kinase B) in cells possessing down-regulated S100A11. In HaCaT cells treated with the nonspecific nsc siRNA, Akt was phosphorylated reflecting the active state of Akt (Fig. 4). The level of overall Akt protein in S100A11 knockdown cells and in control cells was comparable. Because Akt regulates glycogen synthase kinase-3 (GSK3) by phosphorylation [31], the finding of dephosphorylated GSK3 was understandable. The amount of total GSK3 signal was similar in both S100A11 siRNA-treated and control cells (Fig. 4).

Figure 4.

Down-regulation of S100A11 influenced PI3K/Akt signalling. Left: HaCaT cells were transfected with S100A11 siRNA (lane 2) or nsc siRNA (lane 1) for 72 h. Protein extracts of treated cells were subjected to immunoblotting against S100A11, p21, Akt and GSK3 using specific antibodies. Additionally, signals corresponding to phosphorylated Akt (P-Akt) or GSK3 (P-GSK3), respectively, were detected by specific P-site recognizing antibodies. Actin was used in the immunoblot as a control for equal protein loading. Middle: Intensities of the protein bands corresponding to P-Akt relative to actin and Akt were densitometrically evaluated in three independent immunoblots. The mean of the signal intensities corresponding to P-Akt measured in control cells transfected with nsc siRNA was set to 100%. Data are shown as the mean ± SD (*P ≤ 0.05). Right: Intensities of the protein bands corresponding to P-GSK3 relative to actin and GSK3 were densitometrically evaluated in three independent experiments. The mean of the P-GSK3 band intensities measured in control cells (nsc siRNA) was set as 100%. Data are shown as the mean ± SD (*P ≤ 0.05).

To strengthen our data indicating that S100A11 knockdown appears to be involved in p21 destabilization, we additionally analyzed the impact of S100A11 down-regulation in the human epidermoid carcinoma cell line A431, a further cell line possessing functionally inactive p53 [32]. The results achieved by S100A11 knockdown in A431 cells with respect to p21 protein status were comparable to the data generated in HaCaT cells possessing down-regulated S100A11 (Fig. S3). The p21 protein level was significantly decreased in S100A11 knockdown cells compared to control cells treated with nonspecific nsc siRNA. In addition to a reduced p21 protein level, S100A11 knockdown also induced a decreased phosphorylation status of Akt without alteration of total abundance of Akt protein. At the same time, activated GSK3, which is dephosphorylated, appeared in S100A11 knockdown A431 cells. A reduction of the p21 protein abundance was also detectable when A431 cells were treated with the PI3K inhibitor wortmannin.

Interestingly, it appears that the regulation of the p21 protein abundance by S100A11 is dependent on the kind of p53 activity status because, in contrast to HaCaT and A431 cells possessing inactive p53, the p21 protein status is unchanged after S100A11 knockdown in the colon carcinoma cell line HCT116, as well as in MCF-7 cells, a breast adenocarcinoma cell line, in which both cell lines contain functionally active p53 [33, 34] (Fig. S4).

The function of the PI3K/Akt pathway in the maintenance of the p21 protein stability in HaCaT cells reflects the results obtained by the application of wortmannin, a specific inhibitor of PI3K/Akt. When cells were treated with wortmannin, the level of phosphorylated, active Akt was significantly reduced, leading to an increase of dephosphorylated GSK3, which is active in this state. A very significant decrease of the p21 signal occurred concurrently in HaCaT cells treated with 500 nm wortmannin for 16 h compared to untreated control cells (Fig. 5A). Because the activation of GSK3 via the inhibition of Akt induced a reduction of p21 protein abundance, we were interested in whether direct inhibition of GSK3 also stabilizes p21 in cells possessing down-regulated S100A11. Accordingly, we treated HaCaT keratinocytes with specific S100A11 siRNA and the GSK3-inhibitor SB216763. Here, we again observed a significantly decreased p21 protein level in S100A11 knockdown cells compared to cells treated with control nsc siRNA. When cells were treated with SB216763 in addition to S100A11 siRNA, the p21 protein was significantly stabilized compared to cells without GSK3 inhibitor incubation despite down-regulation of S100A11. Although GSK3 appears dephosphorylated corresponding to its active state, GSK3 is inhibited by the addition of SB216763, which is reflected by an increased p21 protein signal in S100A11 knocked-down cells (Fig. 5B). GSK3 phosphorylates p21 at threonine 57 to label it for proteasomal turnover [17]. To determine whether a modification at this site is involved in p21 protein destabilization after S100A11 knockdown, we generated an expression construct containing p21 T57A mutant that prevents phosphorylation at residue 57 of p21. Additionally, we generated a p21 T57D expression construct that mimics a permanent phosphorylation at aspartate 57. When HaCaT cells were transiently transfected with the p21 wild-type (p21 wt) or the p21 T57D expression construct and subsequently treated with nsc siRNA or S100A11 siRNA, a significant decrease of both p21 wt and the p21 T57D mutant protein levels occurred in cells lacking S100A11 compared to control cells (Fig. 5C). The extent of the protein reduction of p21 wt as well as of the p21 T57D mutant was comparable to the reduction of the protein abundance of endogenous p21 after S100A11 down-regulation. By contrast, the mutant p21 T57A protein was stabilized also in HaCaT cells possessing knocked-down S100A11 (Fig. 5C). These data suggest that S100A11 is involved in controlling the PI3K/Akt–GSK3 axis that regulates p21 protein stability.

Figure 5.

Manipulation of the PI3K/Akt pathway regulates p21 protein abundance in HaCaT keratinocytes. (A) Inhibition of PI3K/Akt signalling triggers p21 degradation. Left: Protein extracts of HaCaT cells treated with 500 nm of the PI3K/Akt inhibitor wortmannin (lane 2) for 16 h and, as a control, protein extracts of cells treated with dimethylsulfoxide alone (lane 1) were analyzed by western blot experiments to determine the protein signal level of p21, S100A11, P-Akt, Akt, P-GSK3 and GSK using specific antibodies. Analysis of actin signals in the immunoblots was used as a control for equal sample loading. Right: Protein signals corresponding to p21 relative to actin were evaluated densitometrically in three independent experiments. The mean of the p21 band intensities measured in control cells transfected with nsc siRNA was set to 100%. Data are shown as the mean ± SD (***P ≤ 0.005). (B) Inhibition of GSK3 activity impedes p21 degradation in S100A11 down-regulated HaCaT keratinocytes. Left: HaCat cells transfected with S100A11 siRNA were treated twice with 330 nm GSK3 inhibitor SB216763 (lane 3) 24 and 48 h after initial siRNA transfection or with dimethylsulfoxide alone (lane 2). As a control, HaCaT cells transfected with nsc siRNA were treated with dimethylsulfoxide alone (lane 1). Protein extracts of treated cells were subjected to immunoblotting and analyzed using specific antibodies against S100A11, p21, P-Akt, Akt, P-GSK3 and GSK3. As a control for equal protein loading, corresponding actin levels were shown by immunoblotting. Right: Intensities of protein bands corresponding to p21 relative to actin were densitometrically evaluated in four independent immunoblots. The mean of the p21 band intensities measured in control cells transfected with nsc siRNA was set to 100%. Data are shown as the mean ± SD (*P ≤ 0.05; **P ≤ 0.01). (C) p21 T57A mutant is stabilized in HaCaT cells lacking S100A11. HaCaT cells were treated with S100A11 siRNA (lanes 4–6) or nsc siRNA (lanes 1–3) 24 h after initial transfection with p21 wt (lanes 1 and 4), p21 T57A mutant (lanes 2 and 5) or p21 T57D mutant (lanes 3 and 6) expression construct, respectively. Cells were harvested after a further 72 h and the derived protein extracts were subjected to immunoblot experiments and analyzed using specific antibodies against S100A11 and p21. As a control for equal protein loading, corresponding GAPDH levels were shown by immunoblotting. The relative signal intensity (RSI) of the bands corresponding to p21 relative to GAPDH was densitometrically evaluated in two independent immunoblots. Mean ± SD RSI values are indicated below the respective bands.

Discussion

In the present study, we determined a cellular mechanism of p21 degradation in human HaCaT keratinocytes after knockdown of S100A11. S100A11 belongs to the S100 family of Ca2+-binding proteins. The genes of several members of the S100 protein family including S100A11 are clustered on chromosome 1q21 within a region named epidermal differentiation complex [35]. Many of the genes encoded in the epidermal differentiation complex are expressed in human keratinocytes and appear to possess a function in the epidermis [22, 36]. The HaCaT cell line used in the present study contains transcriptionally inactive p53 caused by mutations that are located in the DNA-binding domain of p53 [26, 37]. Despite p53 being a transcriptional activator of p21/CDKN1A, there are several distinct mechanisms that activate p21 in a p53-independent manner [38, 39].

Several distinct S100 proteins are involved in cell growth and cell cycle by influencing cell cycle regulators such as p53, p21 or CDKs and cyclins. A prominent example is S100A4, which targets p53 and interferes with the DNA-binding capacity of p53, resulting in THE suppression of p21 expression [40]. The expression of S100B can induce an increased p21 level in PC12 neuronal cells [41]. Very recently, it was reported that S100A14 can affect both the transactivation activity and the stability of p53 [42]. Additionally, the down-regulation of a further member of the S100 family, S100A6, induces cell cycle arrest that is mediated by decreased level of CDK1, cyclin A and cyclin B [43]. In the present study, we show that the down-regulation of S100A11 triggers apoptosis and induces p21 destabilization.

Both an intracellular and a secreted subfraction of S100A11 exists that functions as growth inhibitor or stimulates cell proliferation in human keratinocytes, respectively [24, 25]. Our data suggest that S100A11 might be a stimulator of survival because it impeded the activation of caspases, and an apoptosis induction occurred in S100A11 down-regulated HaCaT keratinocyte cells. In addition, S100A11 appears to function in parallel as an inhibitor of inordinate cell proliferation because it controls the stability of p21 and further cell cycle regulators (Fig. S2). These data confirm a former study reporting a decrease of p21 signal after down-regulation of S100A11 in HaCaT cells [44]. Interestingly, there might be differences between the activities of wild-type S100A11 and S100A11-derived peptides because an N-terminal peptide of S100A11 appears to be able to induce apoptotic cell death that is independent of caspase activity, p53 and p21 [45]. S100A11 is differentially expressed in several tumours. The expression of S100A11 can correlate with both tumour suppression and higher tumour grading depending on the specificity of the tumour entity [46, 47]. Furthermore, S100A11 expression appears to be able to discriminate between distinct tumour entities [48].

Several E3 ligases are reported to be commonly involved in the normal turnover of p21 at different phases of the cell cycle using the proteasome in a ubiquitin-dependent manner [7, 8]. Additionally, ionizing and UV radiation-induced proteasome-dependent p21 degradation processes reveal that p21 stability is regulated by further mechanisms in response to extracellular stimuli [19, 49]. It has been shown previously that S100A11 might be involved in the regulation of p21 stability after triggering extracellular stimuli. When a cellular influx of Ca2+ or DNA damage was induced in cells, a directed nuclear translocation of S100A11 in response to these stimuli correlated with an increase of the p21 protein level [24, 50]. Vice versa, a reduction of the p21 protein abundance was detectable when S100A11 was down-regulated by RNA interference, as shown in the present study. The decrease of p21 protein abundance induced by S100A11 knockdown appears to be very specific because the protein status of other members of the kinase inhibitor protein group and the inhibitor of CDKs group, p27Kip1 or p16INK4A, respectively, is unaltered in HaCaT cells lacking S100A11 (Fig. S2). In the present study, we reveal a cellular mechanism that appears to be involved in p21 degradation triggered by S100A11 down-regulation. Interestingly, our results suggest that, in cells containing a mutated form of p53, alternative regulation pathways governing p21 activity exist in addition to p53-dependent transcriptional activation, with expression being regulated through several p53-independent mechanisms [51, 52].

In the present study, we show that S100A11 down-regulation is able to trigger activation of caspase 3/7. It has been shown previously that a caspase cascade is involved in the cleavage of p21 [53]. Despite the presence of activated caspases in S100A11 knockdown cells, it is not conceivable that caspases are involved in destabilization of p21 protein in the HaCaT cells that we used because a described [9] specific p21 cleavage product generated by caspase 3 did not appear, nor did the application of a caspase inhibitor achieve a reconstitution of initial p21 protein abundance after S100A11 down-regulation. The activation of caspase 3/7 resulted in an increase of the apoptotic fraction of S100A11-depleted cells. This is not unexpected because caspases are common inducers of apoptosis (i.e. programmed cell death). Furthermore, with respect to its cell cycle regulatory function, p21 can act as an inhibitor of apoptosis in a number of systems [54]. Thus, p21 protein degradation triggered by S100A11 knockdown is able to induce apoptosis in HaCaT cells, as shown in the present study. These data confirm previous results showing that the induction of apoptosis correlates with the down-regulation of p21 in HaCaT cells [55].

A further mechanism for p21 degradation can be achieved by the proteasome in a ubiquitin-independent manner [11, 29]. In the present study, we discovered a proteasomal mechanism that appears to be required for degradation of p21 in S100A11 knocked-down HaCaT cells. The degradation of p21 was significantly reduced if a specific proteasome inhibitor was employed. A proteasomal-dependent turnover of p21 that is independent of SCFSKP2 was reported recently [56]. The results of the present study suggest that p21 is degraded by a ubiquitin-independent proteasomal process in HaCaT cells lacking S100A11 because: (a) no significant accumulation of ubiquitinated p21 was detectable and (b) E3 ligases involved in normal p21 turnover are also decreased in S100A11 down-regulated HaCaT cells. Very recently, it was shown that Akt is required for the mRNA translation of E3 ligase SCFSKP2 [57]. Therefore, as shown in the present study, it is conceivable that S100A11 knockdown induces a change in Akt activity that results in a negative effect on the translation of SKP2 mRNA in HaCaT cells.

Post-transcriptional modifications by various kinases at several consensus sites of p21 appear to be involved in regulation of p21 protein stability [13, 14, 16]. To determine whether phosphorylation events of p21 are involved in its destabilization after S100A11 knockdown, we assessed several signalling pathways. By analysis of the PI3K/Akt pathway, in addition to a reduced p21 protein level, we detected a decreased phosphorylation level of Akt kinase, reflecting the reduced activity of Akt in S100A11 down-regulated cells. S100A11 is reported to be involved in the control of PI3K/Akt signalling because Akt can be activated by S100A11 [25]. In neuronal cells, it has been shown that the expression of S100B (another member of the S100 protein family) is also able to induce the activation of Akt [41]. Furthermore, it was shown that Akt is involved in the up-regulation of p21 [58]. Inactivated Akt results in GSK3 activation in S100A11 knocked-down HaCaT cells. GSK3 is able to phosphorylate p21 at threonine 57 (T57), triggering its proteasomal degradation in both DNA-damaged and unstressed cells [17]. When we used a p21 T57A mutant that prevented phosphorylation by GSK3 at threonine 57, this specific p21 mutant was even stabilized in cells lacking S100A11. GSK3, which is highly active in quiescent cells, can induce apoptosis and suppress cell proliferation, which is comparable to the cellular behaviour triggered by S100A11 down-regulation [59, 60]. The inhibition of GSK3 activity by a specific GSK3 inhibitor impeded p21 degradation despite S100A11 down-regulation in HaCaT cells as shown in the present study, suggesting a role of S100A11 in the control of both p21 protein stability and GSK3 activity.

In summary, our present data reveal that the down-regulation of S100A11 is responsible for the increased degradation of p21 triggered by activation of GSK3 after the PI3K/Akt pathway in p53-defective epidermal cells. It appears that S100A11 is able to control p21 protein abundance by a dual mechanism in human keratinocytes: activation of Sp1-dependent p21 expression [24] and, as shown in the present study, the control of p21 protein stability.

Experimental procedures

Cells and reagents

The human keratinocyte cell line HaCaT was cultured in DMEM supplemented with 10% fetal bovine serum. The epidermoid carcinoma cell line A431 was cultured in DMEM/Ham's F-12 supplemented with 10% fetal bovine serum. Cells were grown to 80% confluence and were passaged at a split ratio of 1 : 4. Cells were harvested at 70–90% confluence and lysed in a buffer containing 100 mm sodium phosphate pH 7.5, 5 mm EDTA, 2 mm MgCl2, 0.1% Chaps, 500 μm leupeptin and 0.1 mm phenylmethanesulfonyl fluoride. After centrifugation (15 min at 20 800 g), the supernatant was immediately processed further for western blotting. HaCaT keratinocytes and A431 cells possess functionally inactive p53 caused by mutations in triplet codons 179, 281 and 282 or triple codon 273, respectively, of the TP53 gene [26, 32]. These described mutations were confirmed in our used HaCaT cells by sequencing with the p53 sequencing primers: 5′-CCTGTGCAGCTGTGGGTTGATTCC-3′ (forward), 5′-TTAGTGCTCCCTGGGGGCAGCT'3′ (reverse). If siRNA transfected cells were treated with the specific caspase 3/7 inhibitor, cells grown in a six-well gap for 64 h were subsequently incubated with 100 μg·mL−1 Z-Asp-CH2-DCB caspase inhibitor (PeptaNova, Sandhausen, Germany) dissolved in dimethylsulfoxide or, as a control, with dimethylsulfoxide alone for 8 h. Afterwards, cells were harvested and analyzed. For assessment of the impact of a proteasome inhibitor, cells transfected with specific S100A11 siRNA or unspecific nsc siRNA grown in a six-well gap for 64 h and were subsequently incubated with 10 μm MG-132 (AG Scientific, San Diego, CA, USA) for 8 h. Cells were used to generate whole cell extract for analysis by western blot experiments. For inhibition of GSK3 activity, cells transfected with specific S100A11 siRNA were treated with the specific GSK3 inhibitor SB216763 (335 nm; ab120202; Abcam, Cambridge, UK) dissolved in dimethylsulfoxide or, as a control, with dimethylsulfoxide alone 24 h post-transfection for 48 h. A repeated addition of 335 nm SB216763 occurred 24 h after initial inhibitor incubation. For inhibition of PI3K/Akt, cells were treated with 500 nm wortmannin (W1628; Sigma-Aldrich, St Louis, MO, USA) for 16 h. As a control, dimethylsulfoxide alone was used.

Plasmids constructions and in vitro mutagenesis

Human p21 was PCR amplified from U2OS cDNA using the oligonucleotids: 5′- TATTGGAGGATCCCAATGTCAGAACCGGCTGGGGAT-3′ (sense) and 5′- CGGAGAAGGCTCGAGTTAGGGCTTCCTCTTGGAGAAG-3′ (antisense). The PCR fragment was cloned between the BamH1 and Xho1 restriction site of a pcDNA3 vector. Correct insertion of p21 was confirmed by sequencing. The putative GSK3β phosphorylation site (T57) was changed by site-directed mutagenesis (QickChange II Mutagenesis Kit; Agilent Technologies, Böblingen, Germany) using the primers: p21 T57A 5′-AGCCTTTGTCACCGAGGCACCACTGGAGGGTGAC-3′ (forward), 5′-GTCACCCTCCAGTGGTGCCTCGGTGACAAAGGCT-3′ (reverse); p21 T57D 5′-AGCCTTTGTCACCGAGGACCCACTGGAGGGTGAC-3′ (forward), 5′-AGCCTTTGTCACCGAGGACCCACTGGAGGGTGAC-3′ (reverse). All constructs were confirmed by sequencing. For transient transfection with the individual constructs (1 μg), HaCaT cells (1 × 105 cells per well/six-well plate) were plated 18 h before transfection. Transfection was performed using the DNA delivery system TurboFect transfection reagent (Thermo Scientific, St Leon-Rot, Germany) in accordance with the manufacturer's instructions.

RNA interference

The siRNA duplex oligonucleotides used in the present study are based on the human cDNAs encoding S100A11. S100A11 siRNAs, as well as a nsc siRNA, were obtained from Qiagen GmbH (Hilden, Germany). The siRNAs applied to S100A11 target sequence were A11si#5 5′-AAGGATGGTTATAACTACACT-3′, A11si#6 5′-CAGAACTAGCTGCCTTCACAA-3′, A11si#7 5′-CAGCTAGATTTCTCAAGAATTT-3′. The siRNA sequences employed as negative controls were 5′-UUCUCCGAACGUGUCACGUdTdT-3′ (sense) and 5′-ACGUGACACGUUCGGAGAAdTdT-3′ (antisense). Cells (1 × 105) were plated on six-well plates 18 h before transfection and were 50% confluent when siRNA was added. The amount of siRNA duplexes applied was 1.5 μg per well. Transfection was performed using the amphiphilic delivery system SAINT-RED (Synvolux Therapeutics BV, Groningen, The Netherlands) as described previously [20]. For the cycloheximide chase assay, 50 μg of cycloheximide per millilitre of medium was applied and cells were collected after the indicated times.

Western blotting and precipitation

Proteins of interest in crude extracts of HaCaT cells treated with specific S100A11 siRNA or unspecific control siRNA (nsc) and/or with specific reagents were verified by western blot assays using specific antibodies. The primary antibodies used were: anti-β-actin rabbit polyclonal (A2066; Sigma, St. Louis, MO, USA), anti-Akt rabbit monoclonal (4685; Cell Signaling, Danvers, MA, USA), anti-Phospho-Akt (Ser473) rabbit monoclonal (4060; Cell Signaling), anti-CDC20 rabbit polyclonal (sc-8358; Santa Cruz Biotechnology; Dallas, TX, USA), anti-GAPDH rabbit monoclonal (5174; Cell Signaling), anti-GSK3β rabbit monoclonal (9315; Cell Signaling), anti-Phospho-GSK3β (Ser9) rabbit polyclonal (9336; Cell Signaling), anti-p21 rabbit polyclonal (sc-469; Santa Cruz Biotechnology), anti-PARP mouse monoclonal (9546; Cell Signaling), anti-S100A11 rabbit polyclonal (10237-1-AP; Protein Tech Group, Chicago, IL, USA), anti-S100A11 sheep polyclonal (PAS9364; Randox Laboratories Inc., Crumlin, UK), anti-SKP2 rabbit polyclonal (sc-7164; Santa Cruz Biotechnology), anti-ubiquitin mouse monoclonal (P4D1; Cell Signaling). Secondary antibodies were: anti-mouse horseradish peroxidase-conjugated (HRP) (sc-2005; Santa Cruz Biotechnology), anti-rabbit HRP-conjugated (sc-2370; Santa Cruz Biotechnology) and anti-sheep HRP-conjugated (sc-2473; Santa Cruz Biotechnology). Western blots were densiometrically analyzed using labimage-1d (Kapelan Bio-Imaging, Leipzig, Germany). For immunoprecipitation of p21 from HaCaT cell extract, a mouse anti-p21 serum (610234; BD Bioscience, Franklin Lakes, NJ, USA) was used, with normal mouse IgG (500-M00; Peprotech, Rocky Hill, NJ, USA) used for control experiments.

FACS

The cell cycle distribution of HaCaT cells was shown by FACS. Cells treated with specific S100A11 siRNA or unspecific control siRNA were harvested and collected together with the medium to obtain all cell cycle phases and the apoptotic cell population. After centrifugation (5 min at 453 g. at room temperature), the resulting pellet was washed twice with NaCl/Pi. Afterwards, the samples were incubated for staining the DNA with 50 μg propidium iodide solution containing 200 μg·mL−1 RNase A for 20 min at room temperature. Fluorescent labelling was measured with a FAC-Scan using cell quest (BD Bioscience).

Quantitative real-time PCR

Cells were transfected with specific S100A11 siRNA or, as a control, with unspecific control siRNA (nsc). Total RNA was then extracted using an RNA Isolation Kit (Qiagen) and first-strand synthesis was synthesized using a Kit system (Fermentas, St Leon-Rot, Germany) in accordance with the manufacturer′s instructions. The mRNA level of the p21 gene was estimated by quantitative real-time PCR using specific primers: p21 5′- CTGTCACTGTCTTGTACCCTTGT-3′ (forward); p21 5′-CTTCCTGTGGGCGGATTAG-3′ (reverse). The CT value of the p21 gene was normalized to β-actin.

Caspase 3/7 activity assay

Changes of the apoptotic population of HaCaT cells treated with either specific S100A11 siRNA or unspecific control siRNA were analyzed with the Apo-ONE Homogeneous Caspase-3/7 Assay (Promega, Mannheim, Germany) in accordance with the manufacturer′s instructions. The amount of total protein of cell lysates was measured with a NanoDrop device (NanoDrop Technologies, Inc., Wilmington, DE, USA) and samples were diluted to the same concentrations. Afterwards, samples were measured on a fluorescence plate reader with an excitation maximum of 498 nm and an emission maximum of 521 nm.

Statistical analysis

Each experiment was repeated at least three times. Data are presented as the mean ± SD. Statistical comparisons were performed using Student's t-test. ≤ 0.05 was considered statistically significant.

Acknowledgements

We thank Dr Richard W. P. Smith for critically reading the manuscript, as well as Drs A. Baniahmad, C. Biskup, S. Franke, T. Ginter, U. R. Markert, I. Rubio, B. Schlott and F. von Eggeling for materials. This study was supported in part by a grant of the Wilhelm Sander-Stiftung to C.M (2006.042.2).

Ancillary

Advertisement