ERK1/2 and p38 cooperate to induce a p21CIP1-dependent G1 cell cycle arrest

Abstract

To study the mechanisms by which mitogen- and stress-activated protein kinases regulate cell cycle re-entry, we have used a panel of conditional kinases that stimulate defined MAPK or SAPK cascades. Activation of ΔMEKK3:ER* during serum restimulation of quiescent cells causes a strong activation of JNK1 and p38α but only a modest potentiation of serum-stimulated ERK1/2 activity. In CCl39 cells this promoted a sustained G1 arrest that correlated with decreased expression of cyclin D1 and Cdc25A, increased expression of p21CIP1 and inhibition of CDK2 activity. In Rat-1 cells, in which p21CIP1 expression is silenced by methylation, ΔMEKK3:ER* activation caused only a transient delay in the S phase entry rather than a sustained G1 arrest. Furthermore, p21CIP1−/− 3T3 cells were defective for the ΔMEKK3:ER*-induced G1 cell cycle arrest compared to their wild-type counterparts. These results suggest that activated ΔMEKK3:ER* inhibits the G1 → S progression by two kinetically distinct mechanisms, with expression of p21CIP1 being required to ensure a sustained G1 cell cycle arrest. The ERK1/2 and p38αβ pathways cooperated to induce p21CIP1 expression and inhibition of p38αβ caused a partial reversal of the cell cycle arrest. In contrast, selective activation of ERK1/2 by ΔRaf-1:ER* did not inhibit serum stimulated cell cycle re-entry. Finally, selective activation of JNK by ΔMEKK1:ER* failed to inhibit cell cycle re-entry, even in cells that retained wild-type p53, arguing against a major role for JNK alone in antagonizing the G1 → S transition.

Introduction

The progress of cells through G1 and into S phase is regulated by the activity of the retinoblastoma tumour suppressor protein (pRb) (Hiebert et al., 1992; Weinberg, 1995; Hall and Peters, 1996; Kaelin, 1999). Mitogenic stimulation of quiescent cells promotes the expression of D-type cyclins, which assemble with their catalytic partners, CDK4 and CDK6 (Hall and Peters, 1996; Ekholm and Reed, 2000). These cyclin D/CDK complexes initiate the phosphorylation of pRb, which relieves the transcriptional repression of the cyclin E gene (Zhang et al., 2000). Subsequent expression of cyclin E and assembly of cyclin E/CDK2 complexes promotes further phosphorylation of pRb, causing the release of E2F and expression of E2F-regulated genes that are required for the initiation of S phase, including cyclin A (Knudsen and Wang, 1997; Dyson, 1998; Helin, 1998; Knudsen et al., 1998).

CDKs are also regulated by phosphorylation (Gu et al., 1992; Clarke, 1995). Activation of CDK4 and CDK2 requires the phosphorylation of a conserved threonine residue within the catalytic loop and dephosphorylation of Thr14 and Tyr15 in the ATP-binding pocket. Dephosphorylation of CDK2 is catalysed by Cdc25A, a dual specificity phosphatase, that is expressed during the G1 and S phases (Jinno et al., 1994; Blomberg and Hoffmann, 1999).

CDK activity is inhibited by the actions of small molecular weight CDK inhibitors (CDKIs) (Sherr and Roberts, 1999). The INK4 family CDKIs (e.g. p16INK4a) specifically inhibit CDK4 and CDK6, whereas the CIP/KIP family members (e.g. p27KIP1 and p21CIP1) bind and inhibit cyclin/CDK2 complexes. In quiescent cells, high p27KIP1 levels restrain cell cycle re-entry by inhibiting cyclin E/CDK2 complexes, whereas mitogenic signals promote ubiquitin-mediated degradation or proteolytic processing of p27KIP1 (Pagano et al., 1995; Shirane et al., 1999). p21CIP1 can exert opposing effects on the G1 → S transition (Sherr and Roberts, 1999). For example, mitogens stimulate a modest induction of p21CIP1, which may facilitate the assembly of cyclin D/CDK4 complexes and promote the G1 phase progression (LaBaer et al., 1997; Cheng et al., 1999). However, in response to DNA damage, p53 mediates a strong induction of p21CIP1 that exceeds an inhibitory threshold, abolishes CDK2 activity and is required for the DNA damage-induced G1 cell cycle arrest (Vogelstein et al., 2000).

The mitogen-activated protein kinase (MAPK) signalling pathways play a major role in regulating cell cycle re-entry (Marshall, 1995; Roovers and Assoian, 2000). Growth factors activate the Ras-regulated Raf → MEK1/2 → ERK1/2 cascade and both the duration and strength of ERK activation determines the proliferative response (Marshall, 1995; Woods et al., 1997; Balmanno and Cook, 1999; Cook et al., 1999). For example, moderate ERK1/2 activation is required for the expression of cyclin D1 and cell cycle re-entry, whereas high-intensity ERK1/2 activation can induce cell cycle arrest or senescence as a result of the high expression of p21CIP1 (Woods et al., 1997) or p16INK4a (Zhu et al., 1998).

c-Jun N-terminal kinase (JNK) and p38 are stress-activated protein kinases (SAPKs) that are activated weakly, if at all, by proliferative stimuli but strongly in response to cellular stresses, including many DNA damaging chemotherapeutic drugs (Davis, 2000). The precise role of the SAPKs in damage-induced cell cycle regulation is unclear as they have been proposed to be both positive and negative regulators of cell cycle progression (Molnar et al., 1997; Wang et al., 1999; Wisdom et al., 1999; Tournier et al., 2000). Apart from the possible differences in cellular context, this confusion may arise from the inability to separate SAPK activation from cellular damage, which may activate other components involved in cell cycle arrest such as p53 (Vogelstein et al., 2000).

To resolve this problem we have used a panel of conditional MAP kinase kinase kinase:ER fusions. Activation of ΔRaf-1:ER* results in selective activation of the ERK1/2 pathway (Woods et al 1997; Weston et al 2003). Studies indicate that MEKK1 shows selectivity for the JNK pathway (Xia et al., 1998; Yujiri et al., 2000) and we have recently shown that ΔMEKK1:ER* does indeed selectively activate the JNK pathway (Molton et al. 2003). Finally, MEKK3 can activate the ERK, JNK and p38 pathways when overexpressed (Deacon and Blank, 1999) and ΔMEKK3:ER* exhibits the same selectivity (Garner et al., 2002). Using this panel of conditional kinases we show that the ERK1/2 and p38 pathways cooperate to cause a sustained G1 cell cycle arrest, which requires p21CIP1 expression whereas persistent activation of JNK or ERK alone fails to inhibit the G1 → S transition.

Results

ΔMEKK3:ER* enhances ERK1 activity and induces JNK and p38 activity during serum-stimulated cell cycle re-entry

CCl39 fibroblasts expressing ΔMEKK3:ER* (CM3 cells) or a catalytically inactive variant (CM3.KD cells) were described previously (Garner et al., 2002). Activation of ΔMEKK3:ER* by 4-hydroxytamoxifen (4-HT) causes a strong activation of ERK1/2, JNK and p38 in serum-starved cells. However, since we were studying the effect of activating ΔMEKK3:ER* during serum restimulation of quiescent cells, it was important to determine the effect of activating ΔMEKK3:ER* under the same conditions. We stimulated quiescent CM3.6 cells with 10% FBS in the presence of 100 nM 4-HT or an ethanol vehicle control and assayed the activation of endogenous ERK1, JNK1 and p38α by immune complex kinase assays. Foetal bovine serum (FBS) caused a robust, sustained activation of ERK1 and this response was only moderately (30%) enhanced by the activation of ΔMEKK3:ER* (Figure 1a). In contrast, FBS stimulation caused little or no activation of JNK or p38, whereas FBS+4-HT caused a strong and sustained eight- to 12-fold activation of both pathways (Figure 1b).

Figure 1
figure1

ΔMEKK3:ER* enhances ERK1/2 activity but strongly activates JNK and p38 during cell cycle re-entry. Quiescent CM3.6 cells were restimulated with 10% FBS for 1, 4 or 8 h in the presence of 0.04% ethanol (vehicle control) or 100 nM 4-HT. (a) Endogenous ERK1 activity was assayed by immune complex kinase assay with MBP as a substrate. (b) Endogenous JNK1 and p38α activities were assayed using GST-c-Jun (JNK) or His-MAPKAP-K2 (p38) as substrates. (c) IκB phosphorylation status was assessed by subjecting whole-cell lysates to Western blot analysis with phospho-IκB or total IκB antibodies; lysates from cells treated with 10 nM calyculin A for 1 h served as a positive control. Results are taken from a single experiment; similar results were observed in a total of three independent experiments

Full-length MEKK3 can activate NFκB when overexpressed (Yang et al., 2001) so we assayed IκB phosphorylation and degradation as a surrogate marker for NFκB activation. FBS stimulation, with or without 4-HT, had no effect on IκB whereas the phosphatase inhibitor calyculin A, used as a positive control, caused phosphorylation and degradation of IκB (Figure 1c); thus, it is unlikely that ΔMEKK3:ER* activates NFκB in these cells. These data indicate that activation of ΔMEKK3:ER* during serum-stimulated cell cycle re-entry causes a strong activation of JNK and p38 but only a modest enhancement of ERK1/2 activity.

ΔMEKK3:ER* induces a sustained G1 cell cycle arrest

We next examined the consequences of activating ΔMEKK3:ER* on serum-stimulated cell cycle re-entry in quiescent CM3.6 cells. In control cells, the first peak of FBS-stimulated [3H]thymidine incorporation was observed after 15 h, and was followed by a second round of S phase at 25–30 h (Figure 2a). In contrast, 4-HT-treated cells exhibited only a slight increase in [3H]thymidine incorporation at 15 h (25% of that seen in ethanol-treated cells) and complete inhibition to the basal level at all other time points. This inhibition of DNA synthesis was dose-dependent: an IC50 of 5–10 nM was typically observed (data not shown) agreeing well with the EC50 for the activation of ERK, JNK and p38 in these cells (Garner et al., 2002). As a specificity control CM3.KD cells, expressing catalytically inactive ΔMEKK3:ER*, were unaffected by 4-HT (Figure 2b).

Figure 2
figure2

ΔMEKK3:ER* induces a sustained G1 cell cycle arrest. (a) Quiescent CM3.6 cells were treated with 10% FBS in the presence of ethanol or 100 nM 4-HT for up to 30 h. For each time point [3H]thymidine was added for the final 2 h prior to sample preparation and assay of DNA synthesis by [3H]thymidine incorporation. (b) Quiescent CM3 (black bars) and CM3.KD (hatched bars) cells were treated with 10% FBS in the presence of ethanol or 100 nM 4-HT for 24 h and assayed for [3H]thymidine incorporation. (c) Quiescent CM3 cells were treated with 10% FBS in the presence of ethanol or 100 nM 4-HT for 24 h before being fixed, stained with propidium iodide and analysed by flow cytometry. For [3H]thymidine incorporation, data are expressed as mean±range of duplicate samples. Results are taken from a single experiment; similar results were observed in a total of three independent experiments

Flow cytometry was used to confirm at what stage in the cell cycle CM3 cells arrested. The majority of serum-starved cells were in G0/G1, but after 24 h with FBS significant numbers were in S and G2. In contrast, cells treated with FBS+4-HT arrested predominantly at G1; very few cells were able to pass through the first S phase and reach G2/M (Figure 2c). Similar results were obtained in two other CM3 clones (see also Garner et al., 2002). Thus, the activation of ΔMEKK3:ER* prevents FBS-stimulated cell cycle re-entry and CM3 cells arrest in G1.

Activation of ΔMEKK3:ER* promotes p21CIP1 expression, inhibition of CDK2 and prevents phosphorylation of pRb

We next analysed the expression of cell cycle regulatory proteins involved in the G1 → S progression. Western blot analysis revealed that activation of ΔMEKK3:ER* delayed the FBS-stimulated increase in cyclin D1 expression and inhibited both Cdc25A and cyclin A expression, but had little effect on the CDK4 and CDK2 expression or p27KIP1 degradation (Figure 3a). A selection of cyclin E antibodies exhibited poor crossreactivity with the hamster equivalent and as a consequence cyclin E protein levels could not be monitored. However, 4-HT treatment did cause a very strong induction of p21CIP1 expression. Activation of ΔMEKK3:ER* greatly enhanced the normally modest serum-stimulated increase in p21CIP1 expression at all time points tested (two examples, over different time points, are shown in Figure 3a). The increase in p21CIP1 levels was not accompanied by an increase in p53 expression. p53 levels increased in response to FBS stimulation, as seen previously (Reich and Levine, 1984), and the activation of ΔMEKK3:ER* actually reduced this increase in p53 levels.

Figure 3
figure3

ΔMEKK3:ER* induces p21Cip1 expression, inhibition of CDK2 and inhibits pRb phosphorylation. Quiescent CM3.6 cells were treated with 10% FBS in the presence of ethanol or 100 nM 4-HT for the times indicated. (a) Whole lysates were prepared, separated by SDS–PAGE, transferred to PVDF, and probed with antibodies to the indicated proteins. (b) (Left) Cell lysates from CM3.3 or CM3.6 cells were assayed for CDK2 activity by immune complex kinase assay and activity is expressed as radioactivity incorporated into Histone H1 in phosphorimager (PI) units. (Right) The CDK2 immunopreciptates were also analysed for the presence of p21CIP1 by Western blot. (a) Results are pooled from three different experiments representative of a total of six, while in (b) results are taken from a single experiment representative of three giving similar results

CDK2-associated H1 kinase activity increased rapidly after 12 h of FBS stimulation in two independently derived clones (CM3.3 and CM3.6; Garner et al., 2002). However, the activation of CDK2 was almost completely abolished when ΔMEKK3:ER* was activated (Figure 3b, left), and this was accompanied by a 4-HT-dependent recruitment of p21CIP1 into CDK2 complexes (Figure 3b, right). Activation of ΔMEKK3:ER* also inhibited cyclin A-associated kinase activity, and this was also accompanied by the recruitment of p21CIP1 to cyclin A immuno-precipitates (D. Todd and S Cook, unpublished results). Finally, 4-HT treatment inhibited the FBS-stimulated accumulation of hyper-phosphorylated pRb, and the expression of the related pocket protein p107 (Figure 3a). Collectively, these data are consistent with ΔMEKK3:ER* inducing a G1 arrest by increasing p21CIP1 levels, inhibiting CDK2 and so preventing the phosphorylation and inactivation of pRb. In such a scenario, the loss of Cdc25A, cyclin A and p107 expression would presumably reflect the fact that they are E2F target genes (Dyson, 1998).

ΔMEKK3:ER* fails to increase p21CIP1 expression and causes only a transient delay in cell cycle re-entry in Rat-1 cells

To investigate the potential role of p21CIP1 we examined Rat-1 cells, in which the p21CIP1 promoter is silenced by methylation (Allan et al., 2000). We have previously shown that ΔMEKK3:ER* also activates ERK, JNK and p38 in Rat-1 cells (called RM3 cells) (Garner et al., 2002). Furthermore, control experiments confirmed that activation of ΔMEKK3:ER* during FBS stimulation of RM3 cells resulted in the same modest enhancement of ERK1 activity and the strong activation of JNK1 and p38α, as seen in CM3 cells (D Todd, K Balmanno and S Cook, unpublished results).

Following FBS stimulation in the presence or absence of 4-HT, we again observed that activation of ΔMEKK3:ER* caused a super-induction of p21CIP1 in CM3 cells whereas RM3 cells were unable to express p21CIP1 at any time point (Figure 4a). This lack of p21CIP1 induction was attributable to DNA methylation, as treatment with 5-Aza-2-deoxycytidine (an inhibitor of DNA methyltransferases) restores p21CIP1 expression (Allan et al., 2000; Garner et al., 2002). In control RM3 cells, CDK2-associated kinase activity increased after 12 h of serum stimulation. In contrast to the sustained inhibition of CDK2 activity seen in CM3 cells, 4-HT treatment of RM3 cells caused only a transient 5–10 h delay before CDK2 activation proceeded normally (Figure 4b, left). The delayed induction of CDK2 activity was reflected in the phosphorylation status of pRb, the expression of p107 and cyclin A which all proceeded normally after a delay of several hours in the presence of 4-HT compared to cells treated with FBS alone (Figure 4b, right).

Figure 4
figure4

ΔMEKK3:ER* fails to induce p21CIP1 expression and causes only a transient delay in cell cyle re-entry in Rat-1 cells. (a) Quiescent CM3 and RM3 cells were treated with 10% FBS in the presence of ethanol or 100 nM 4-HT for the indicated times and whole-cell lysates were prepared, normalized for protein content and probed for p21CIP1 expression (b) (Left) Quiescent RM3.21 cells were treated with 10% FBS in the presence of ethanol or 100 nM 4-HT for the indicated times. CDK2 activity was measured by immune complex kinase assay and was expressed as radioactivity incorporated into histone H1 in phosphorimager (PI) units. Results are mean±range of n=2 experiments; similar results were obtained in a third experiment. Right: in addition, whole-cell lysates were analysed by immunoblot for phosphorylation of pRb and expression of p107 and cyclin A. Results are taken from a single experiment; similar results were observed in a total of three independent experiments. (c,d) Quiescent RM3.16 cells were stimulated with 10% FBS in the presence of ethanol or 100 nM 4-HT for increasing periods of time. (c) DNA synthesis was assayed by [3H]thymidine incorporation (mean±range of duplicate samples) while in (d) cells were stained with propidium iodide and their DNA content was measured by flow cytometry. Results are taken from a single experiment; similar results were observed in a total of three independent experiments

These biochemical changes were in turn reflected in the kinetics of the S phase entry. In 4-HT-treated RM3 cells, the increase in [3H]thymidine incorporation was delayed by approximately 5–10 h, but the maximal response was the same and the rate of [3H]thymidine incorporation was essentially identical in control cells (10–20 h) and 4-HT-treated cells (15–25 h) (Figure 4c). These results were confirmed by flow cytometry (Figure 4d). For example, 4-HT caused a delay in cell cycle re-entry at 20 h, but after 25 h 4-HT-treated cells were progressing through S phase and G2 and cells with a 4n DNA content represented the largest fraction of the population after 30 h of FBS+4-HT (Figure 4d). This probably reflects the ΔMEKK3:ER*-induced G2 arrest that is independent of p21CIP1 status (Garner et al., 2002). Thus in contrast to CM3 cells (Figure 2), RM3 cells were unable to sustain a ΔMEKK3:ER*-induced G1 arrest and ultimately arrested at G2.

p21CIP1 expression is required for the ΔMEKK3:ER*-induced G1 cell cycle arrest

To confirm that p21CIP1 expression was required for the ΔMEKK3:ER*-induced sustained G1 arrest we expressed Myc-tagged ΔMEKK3:ER* in 3T3 cell lines derived from WT or p21CIP1−/− mouse embryo fibroblasts (Figure 5a and b). 4-HT treatment caused an increase in c-Jun phosphorylation in both cell lines, confirming that the ΔMEKK3:ER* construct was functional (Figure 5a). ΔMEKK3:ER*-induced expression of p21CIP1 was observed in the WT 3T3 cell line but not in the p21CIP1−/− cells. In WT 3T3 cells, 4-HT treatment caused a strong expression of p21CIP1, inhibited the FBS-stimulated expression of cyclin A and p107 (Figure 5b) and caused a strong, 84% inhibition of FBS-stimulated DNA synthesis (Figure 5c). In contrast, activation of ΔMEKK3:ER* in p21CIP1−/− 3T3 cells caused only a weak 25–30% inhibition of DNA synthesis indicating that the ΔMEKK3:ER*-induced G1 arrest was defective. Flow cytometry confirmed that the p21CIP1−/− 3T3 cells were defective for the G1 arrest so that after 24 h of FBS+4-HT the majority of cells were in S phase or G2 (Figure 5d). Thus, p21CIP1 is required for a sustained G1 arrest in response to the activation of ΔMEKK3:ER*; in the absence of p21CIP1 cells can still enter S phase but arrest at G2 (Garner et al., 2002).

Figure 5
figure5

p21CIP1 is required for ΔMEKK3:ER* to induce an optimal G1 arrest. (a) WT or p21CIP1−/− 3T3 cells expressing Myc-ΔMEKK3:ER* were kept in complete medium (C) or serum starved overnight (SF) before restimulating with FBS in the presence of ethanol (F) or 100 nM 4-HT (F+4-HT) for 24 h. Whole-cell lysates were prepared, fractionated by SDS–PAGE and blotted with antibodies for the Myc tag on ΔMEKK3:ER*, phospho-c-Jun or p21CIP1. Similar results were seen in two independent experiments. (b) WT 3T3 cells were treated with 10% serum in the presence of ethanol or 100 nM 4-HT for the indicated times and whole-cell lysates were prepared, normalized for protein content and blotted for p21CIP1, cyclin A and p107. (c,d) WT and p21CIP1−/− cells treated as above (a) were assayed for [3H]thymidine incorporation (c) or stained with propidium iodide and analysed for cell cycle distribution by flow cytometry (d). (c) Results are taken from a single experiment and represent the mean±s.d. of triplicate cell samples. The figures indicate the % inhibition of FBS-stimulated DNA synthesis by 4-HT in each cell line. (c) The percentage of cells in each phase of the cell cycle is indicated

ΔMEKK3:ER*-stimulated p21CIP1 expression requires the ERK1/2 and p38 pathways

Pharmacological inhibitors of the ERK1/2 pathway (PD184352, Sebolt-Leopold et al., 1999) and p38 (SB203580, Cuenda et al., 1995) were used to investigate which pathways were responsible for ΔMEKK3:ER*-mediated p21CIP1 expression and cell cycle arrest (Figure 6a). 4-HT-induced p21CIP1 expression was partially inhibited by treatment with either PD184352 or SB203580 and completely abolished when the two drugs were used in combination. A control blot for phospho-ERK1/2 confirmed that PD184352, but not SB203580, was preventing the activation of ERK1/2 under these conditions.

Figure 6
figure6

ΔMEKK3:ER*-stimulated p21CIP1 induction is mediated by the ERK1/2 and p38 pathways. (a) Quiescent CM3.6 cells were stimulated with either ethanol, 4-HT (100 nM), 4-HT+PD184352 (10 μ M), 4-HT+SB203580 (10 μ M) or 4-HT+PD+SB for the indicated times. Whole-cell lysates were prepared, normalized for protein and immunoblotted for p21CIP1 and phosphorylated ERK1/2 to confirm the efficacy of PD184352 and selectivity of SB203580. (b) Quiescent CM3.6 cells were left unstimulated or stimulated with 10 % FBS in the presence of either ethanol, ethanol+SB203580 (10 μ M), 4-HT (100 nM) or 4-HT+SB203580 for 24 h. DNA synthesis was assayed by [3H]thymidine incorporation and data are expressed as mean±range of duplicate samples. Results are taken from a single experiment; similar results were observed in a total of three independent experiments. (c) Quiescent CM3.6 cells were left unstimulated or stimulated with 10% FBS in the presence or absence of 10 μ M SB203580 as indicated. Whole-cell lysates were blotted for p21CIP1 expression. (d) Quiescent CM3.6 cells were left untreated or stimulated with 4-HT in the presence or absence of 10 μ M SB203580. Whole-cell lysates were blotted for (phospho)P-MAPKAK-K2, total MAPKAP-K2, P-ERK1/2, ERK1/2, P-c-Jun and total c-Jun to confirm the efficacy and selectivity of SB203580

We could only assess the effects of the p38 inhibitor (SB203580) on ΔMEKK3:ER*-mediated cell cycle arrest as the ERK1/2 pathway inhibitor PD184352 was itself very effective at inhibiting serum-stimulated cell cycle re-entry (Squires et al., 2002). Following release from quiescence, CM3.3 cells were labelled with [3H]thymidine to monitor their progress into S phase (Figure 6b). In the absence of 4-HT, serum-stimulated DNA synthesis was not affected by SB203580 and this agreed with the lack of effect of SB203580 on the modest serum-stimulated expression of p21CIP1 (Figure 6c). Since serum causes little if any increase in p38 activity these results demonstrate that there is no role for p38 in the normal FBS-stimulated expression of p21CIP1 or cell cycle re-entry. However, SB203580 was able to partially reverse the inhibition of serum-stimulated DNA synthesis by 4-HT (Figure 6b), suggesting that the p38α/β pathway was required for optimal p21CIP1 expression and cell cycle arrest induced by activated ΔMEKK3:ER*. Control immunoblots (Figure 6d) confirmed the selectivity of SB203580, which prevented phosphorylation of the p38 substrate, MAPKAP-K2, but did not prevent the phosphorylation of ERK1/2 or c-Jun (a surrogate assay for JNK).

ERK1/2 and p38 cooperate to increase p21CIP1 mRNA levels in response to activation of ΔMEKK3:ER*

To determine at what level p21CIP1 expression was regulated we used HM3 cells, HEK293 cells expressing ΔMEKK3:ER*, since we could use the published sequence of human p21CIP1 to develop oligos for quantitative real-time RT–PCR. Activation of ΔMEKK3:ER in HM3 cells (confirmed by the appearance of phosphorylated ERK1/2) caused a striking increase in p21CIP1 protein expression within 4 h (Figure 7a, left), which was precisely mirrored by a substantial increase in p21CIP1 mRNA (Figure 7a, right). In common with CM3 cells, there was no accompanying increase in p53 levels; indeed, HM3 cells failed to express p21CIP1 or arrest at G1 in response to doxorubicin (R Densham and S Cook, unpublished results) indicating that p53 is nonfunctional in these cells. The combination of U0126 and SB203580 abolished the ΔMEKK3:ER*-induced expression of p21CIP1 protein (Figure 7b, left) and identical results were again observed when p21CIP1 mRNA levels were assayed in parallel by (Q)RT–PCR (Figure 7b, right). Finally, p21CIP1 expression was also abolished by the combination of PD184352 and SB203580 (Figure 7c). Since PD184352 selectively inhibits the ERK1/2 pathway (Squires et al 2002) whereas U0126 inhibits both the ERK1/2 and ERK5 pathways (Kamakura et al., 1999), these results indicate that ΔMEKK3:ER* promotes the expression of p21CIP1 mRNA and protein through the combined effects of the ERK1/2 and p38 pathways. Furthermore, this expression of p21CIP1 is independent of p53 since it takes place in the absence of p53 stabilization and in cells (CCl39 and HEK293) in which p53 appears to be nonfunctional.

Figure 7
figure7

The ERK1/2 and p38 pathways are required for ΔMEKK3:ER*-stimulated p21CIP1 mRNA accumulation in HEK293 cells. (a) HM3 cells (HEK293 cells expressing ΔMEKK3:ER*) were stimulated with 4-HT for the times indicated. (Left) Whole-cell lysates were prepared and blotted for expression of p21CIP1, p53, P-ERK1/2 and ERK1/2. (Right) Total RNA was prepared and used to assay p21CIP1 mRNA levels by (Q)RT–PCR. Levels of p21CIP1 mRNA were normalized to the GAPDH internal control and are expressed as fold increase over basal. In both cases results are shown from a single experiment representative of three. (b) HM3 cells were treated with 10 μ M SB203580, 10 μ M U0126 and 100 nM 4-HT as indicated. (Left) Whole-cell lysates were prepared and blotted for expression of p21CIP1, p53, P-ERK1/2. Results are shown from a single experiment representative of three. (Right) Total RNA was prepared and used to assay p21CIP1 mRNA levels by (Q)RT–PCR. Levels of p21CIP1 mRNA were normalized to the GAPDH internal control, expressed as a percentage of the maximum and are pooled from two independent experiments (mean±range). (c) HM3 cells were treated with 10 μ M SB203580, 5 μ M PD184352 and 100 nM 4-HT as indicated. (Left) Whole-cell lysates were prepared and blotted for expression of p21CIP1, p53, P-ERK1/2. Results are shown from a single experiment representative of two

Selective activation of the ERK1/2 pathway by ΔRaf-1:ER* does not induce p21CIP1 and fails to inhibit the G1 → S transition

Since PD184352 and U0126 both inhibit cell cycle re-entry it was not possible to use these drugs to assess the role of the ERK1/2 pathway in the ΔMEKK3:ER*-induced cell cycle arrest. As an alternative we examined the impact of selective activation of the ERK1/2 on FBS-stimulated cell cycle re-entry. We have previously described CR1-11 cells (Weston et al., 2003); these are CCl39 cells stably expressing ΔRaf-1:ER*, which selectively activates the ERK1/2 pathway, but not JNK, p38 or PKB. When quiescent CR1-11 cells were stimulated with FBS±4-HT we observed that activation of ΔRaf-1:ER* had no effect on the early peak of ERK1 activity but potentiated the smaller sustained phase of ERK1 activity by between 30 and 50% (Figure 8a). This was slightly greater than the effect of ΔMEKK3:ER* on FBS-stimulated ERK1 activity (Figure 1a). However, this was not sufficient to increase p21CIP1 levels over that seen in response to FBS (Figure 8b) and did not inhibit FBS-stimulated DNA synthesis (Figure 8c).

Figure 8
figure8

Selective activation of the ERK1/2 pathway by ΔRaf-1:ER* does not induce p21CIP1 and fails to inhibit the G1 → S transition. (a) Quiescent CR1-11 cells (CCl39 cells expressing ΔRaf-1:ER*) were stimulated with 10% FBS±100 nM 4-HT for the indicated time points before assaying ERK1 activity by immune complex kinase assay. (b) Quiescent CR1-11 cells were stimulated with 10% FBS±100 nM 4-HT for the indicated time points and whole-cell lysates were blotted for the expression of p21CIP1. (c) Quiescent CR1-11 cells were stimulated with 10% FBS±10 nM or 100 nM 4-HT and assayed for [3H]thymidine incorporation. (a,b) Results are from a single experiment representative of three while in (c) results are mean±range of duplicate cell samples from a single experiment representative of three

Selective activation of the JNK pathway by ΔMEKK1:ER* does not induce p21CIP1 and fails to inhibit the G1 → S transition

Activation of JNK is frequently associated with stressful stimuli that can cause cell cycle arrest; therefore, we wished to determine what contribution JNK signalling made to the G1 arrest. To date we have failed to demonstrate significant, reproducible efficacy with the commercially available JNK inhibitor SP600125 (K Balmanno and S Cook, unpublished results). Furthermore, this drug has recently been shown to exhibit poor selectivity (Bain et al., 2003). Therefore, as an alternative strategy for examining the role of JNK in the G1 → S transition we used the conditional protein kinase ΔMEKK1:ER*, which selectively activates the JNK pathway without activating ERK1/2, ERK5, p38, PKB or IκB kinase (Molton et al. 2003). For these experiments we used CCl39 and Rat-1 cells stably expressing ΔMEKK1:ER* (CM1 and RM1 cells respectively).

When quiescent CM1 cells were restimulated with FBS we observed a strong increase in ERK1/2 phosphorylation, whereas p38 was not phosphorylated and FBS caused only a weak increase in phosphorylation of c-Jun, a surrogate assay for JNK (Figure 9a). When cells were stimulated with FBS+4-HT, activation of ΔMEKK1:ER* caused a striking increase in c-Jun phosphorylation which was apparent within 1 h; quantification of these blots indicated that this increase was of the order of 10-fold over that seen in response to FBS (Figure 9a). In contrast, ΔMEKK1:ER* had no effect on ERK1/2 phosphorylation and little, if any, effect on p38. These results show that the selectivity of ΔMEKK1:ER* previously described (Molton et al., 2003) also applies during FBS stimulation. Despite this strong, selective activation of JNK, activation of ΔMEKK1:ER* had no effect on FBS-stimulated expression of cyclin D1 or cyclin A (Figure 9b) or DNA synthesis assayed up to 15 h (Figure 9c). Thus, activation of JNK does not antagonize the G1 → S transition in CCl39 cells. However, in the course of these studies we noted that activation of ΔMEKK1:ER* delayed the progress of cells to the second round of DNA synthesis at 25–30 h (Figure 9c). Since activation of JNK did not delay the G1 → S transition per se, we reasoned that it might be exerting an effect elsewhere in the cell cycle. Indeed, flow cytometric analysis of CM1 cells revealed that activation of ΔMEKK1:ER* caused cells to accumulate at G2 with a 4n DNA content (Figure 9d), though this effect was not observed with a catalytically inactive variant of ΔMEKK1:ER* (KD in Figure 9d). Identical results were observed in RM1 cells. For example, activation ΔMEKK1:ER* caused a strong, 18-fold activation of JNK but had little effect on ERK1 or p38 (Figure 9e). Despite this, activation of ΔMEKK1:ER* failed to inhibit the G1 → S in RM1 cells (Figure 9f) though it did delay progression through to the second S phase.

Figure 9
figure9

Selective activation of the JNK pathway by ΔMEKK1:ER* does not induce p21CIP1 and fails to inhibit the G1 → S transition in CM1 and RM1 cells. (a) Quiescent CM1 cells (CCl39 cells expressing ΔMEKK1:ER*) were stimulated with 10% FBS±100 nM 4-HT for the times indicated or were treated with or without 50 μ M sodium arsenite (NaAs) for 1 h. Whole-cell lysates were immunoblotted with antibodies specific for P-ERK1/2, P-p38, P-c-Jun and total ERK1/2 (as a loading control). (b and c) Quiescent CM1 cells were stimulated with 10% FBS±100 nM 4-HT for the times indicated and (b) whole-cell lysates were immunoblotted with antibodies specific for cyclin D1, cyclin A, CDK2 and p21CIP1 or (c) cells were assayed for [3H]thymidine incorporation (mean±range of n=2 cell samples). (d) CM1 or CM1.KD cells (the latter expressing catalytically inactive ΔMEKK1:ER*) were stimulated with 100 nM 4-HT for the times indicated and were then stained with propidium iodide and analysed for cell cycle distribution by flow cytometry. (e) RM1 or RM1.KD cells (Rat-1 cells expressing WT or inactive ΔMEKK1:ER*) were stimulated with 100 nM 4-HT for 4 h and ERK1, JNK1 or p38α activity was assayed by immune complex kinase assay. (f) Quiescent RM1 cells were stimulated with 10% FBS±100 nM 4-HT for the times indicated before being assayed for [3H]thymidine incorporation (mean±range of n=2 cell samples)

Previous studies have suggested that JNK-dependent phosphorylation can stabilize p53 (Fuchs et al., 1998; Buschmann et al., 2001). Since CCl39 cells lack functional p53 and Rat-1 cells fail to express p21CIP1 (the major p53 target involved in the G1 checkpoint) it was important to examine the effect of ΔMEKK1:ER* in a cell line that contained wild-type p53 and p21. Accordingly we expressed Myc-ΔMEKK1:ER* in the HCT116 colorectal cancer cell line (Figure 10a). Control studies confirmed that these HCT116-M1 cells exhibited stabilization of p53, expression of p21CIP1 and G1 arrest in response to doxorubicin (Garner et al. 2002; C Newson, K Balmanno and S Cook, unpublished observations). Treatment of HCT116-M1 cells with 4-HT had no effect on the phosphorylation of ERK1/2 or p38 but caused a strong phosphorylation of c-Jun confirming that the ΔMEKK1:ER* was functional and selective. Furthermore, the magnitude of JNK activation was comparable to that seen with many commonly used stress agonists (Figure 10b). Despite this, ΔMEKK1:ER* had no effect on the expression of p53 and p21CIP1 expression (Figure 10a) or FBS-stimulated DNA synthesis in HCT116-M1 cells (Figure 10c). In contrast, both PD184352 and doxorubicin inhibited DNA synthesis; the latter confirming that the DNA damage-induced G1 checkpoint was intact in these cells. Thus, these results show that selective and persistent activation of the JNK pathway fails to inhibit the G1 → S transition even in cells with an intact p53 → p21CIP1 pathway.

Figure 10
figure10

Selective activation of the JNK pathway by ΔMEKK1:ER* fails to inhibit the G1 → S transition in HCT116 cells. (a) Quiescent HCT116-M1 cells (HCT116 cells expressing Myc-ΔMEKK1:ER*) were stimulated with 100 nM 4-HT for 4 h. Whole-cell lysates were then blotted with antibodies specific for P-c-Jun, c-Jun, P-ERK1/2, P-p38, Myc (9E10), p53 and p21CIP1. (b) Quiescent HCT116-M1 cells were stimulated with 100 nM 4-HT or the indicated stress agonists for 4 hours before JNK1 activity was assayed by immune complex kinase assay; results are mean±s.d. of triplicate cell samples. (c) Quiescent HCT116-M1 cells were stimulated with 10% FBS in the absence or presence of 100 nM 4-HT, 5 μ M PD184352 or 10 μ M doxorubicin. [3H]thymidine incorporation was assayed and the results are mean±range of n=2 cell samples. Results are from a single experiment representative of two (a) or three (b and c)

Discussion

In order to analyse the effects of MAPK or SAPK activation on cell cycle re-entry we felt that it was important to stimulate these pathways in the absence of cellular damage that might activate other cell cycle arrest pathways. The panel of kinase:ER* fusions used here appears to fulfil this aim. During FBS stimulation ΔRaf-1:ER* selectively enhances ERK1/2 activation, ΔMEKK1:ER* selectively activates JNK while ΔMEKK3:ER* enhances ERK1/2 but strongly activates JNK and p38. In the case of ΔMEKK1:ER* and ΔMEKK3:ER* we have previously shown that the magnitude of JNK and/or p38 activation is similar to that seen in response to either anisomycin or UV (Garner et al., 2002; Molton et al., 2003). These constructs should not be considered as mimetics of the corresponding full-length kinases but simply afford rapid, conditional, titratable activation of defined MAPK or SAPK pathways without overexpression or overt cellular stress. Using these systems we have shown that ERK1/2 and p38 cooperate to inhibit the G1 → S progression by at least two discrete mechanisms: a sustained p53-independent arrest that requires p21CIP1 expression and a more transient delay that is independent of p21CIP1.

In CM3 cells activation of ΔMEKK3:ER* induced a strong increase in p21CIP1 expression, inhibition of CDK2 and DNA synthesis (Figure 2 and 3). In contrast, in RM3 cells, in which the p21CIP1 promoter is silenced by methylation, activation of ΔMEKK3:ER* failed to increase p21CIP1 expression and caused only a delay in cell cycle re-entry. While these results were compelling, they did not provide proof of the requirement for p21CIP1. However, the demonstration that p21CIP1−/− 3T3 cells were defective in their G1 arrest in response to the activation of ΔMEKK3:ER* compared to their wild-type counterparts confirmed that p21CIP1 expression is required for ΔMEKK3:ER* to induce a sustained G1 arrest.

Our results differ fundamentally from a study in NIH3T3 cells using a related construct, MEKK3-ER, which failed to increase p21CIP1 expression (Ellinger-Ziegelbauer et al., 1999). This is unlikely to represent a cell type-specific difference as we have observed p21CIP1 induction in CM3 cells, HEK293 cells and 3T3-immortalized MEFs. A more likely explanation may be the way the experiments were performed. In NIH3T3 cells, MEKK3-ER was activated for 4 days in serum-free conditions before restimulating with FBS. In our own studies we have found that the expression of p21CIP1 is rapid (within 4 h) and maximal by 24 h; however, p21CIP1 levels then decline, probably due to the substantial apoptotic response observed from approximately 24 h onwards in all cells that we have examined to date (C Weston and S Cook, unpublished observations). Thus, by activating MEKK3-ER for 4 days rather than the detailed time courses employed here the previous study may simply have missed the increase in p21CIP1 expression. Certainly, they did not examine the requirement for p21CIP1.

In the absence of p21CIP1, RM3 cells progressed through G1 and entered S phase in the presence of 4-HT, albeit with delayed kinetics (Figure 4); similar results were observed in the p21CIP1−/− 3T3 cells (Figure 5). Thus, the activation of ΔMEKK3:ER* can clearly delay cell cycle re-entry via a p21CIP1-independent mechanism. In CM3 cells activation of ΔMEKK3:ER* was able to reduce the FBS-stimulated expression of cyclin D1 and Cdc25A (Figure 3) and a similar effect was observed in the RM3 cells (D Todd and S Cook, unpublished observations). Impaired expression of cyclin D1 and/or Cdc25A would both seem to be plausible mechanisms for a p21CIP1-independent delay in S phase entry by preventing the activation of CDK4, CDK2, pRb phosphorylation and the subsequent induction of E2F-responsive genes (Jinno et al., 1994; Schulze et al., 1995; Geng et al., 1996; Blomberg and Hoffmann, 1999).

p21CIP1−/− mouse embryo fibroblasts still exhibit some degree of p21CIP1-independent G1 arrest in response to DNA damage (Brugarolas et al., 1995, Deng et al., 1995) and recent reports have suggested that the p53-dependent upregulation of p21CIP1 is too slow to account for the rapid cell cycle block seen in response to DNA damage. Alternative mechanisms proposed for a rapid p53-independent initiation of G1 arrest include the proteasome-dependent destruction of either cyclin D1 (Agami and Bernards, 2000) or Cdc25A (Mailand et al., 2000) before the slower p53-dependent induction of p21CIP1 ensures a sustained G1 arrest. Activation of ΔMEKK3:ER* may trigger similar mechanisms to elicit a sustained G1 arrest, although p21CIP1 induction may be p53-independent in our system (see below). Activation of p38 has been shown to inhibit the transcription of cyclin D1 (Lavoie et al., 1996) and to increase the turnover of the cyclin D1 protein (Casanovas et al., 2000), and future studies will aim to elucidate the mechanism by which activation of ΔMEKK3:ER* downregulates cyclin D1 and Cdc25A expression.

ERK1/2 and p38 cooperate to promote a p53-independent expression of p21CIP1

Selective pharmacological inhibitors of the ERK1/2 pathway (PD184352) or of p38 (SB203580) revealed for the first time that ERK1/2 and p38 cooperate to induce p21CIP1 expression. The requirement for the ERK1/2 pathway in the ΔMEKK3:ER*-dependent expression of p21CIP1 is reminiscent of the response seen with ΔB-Raf:ER* (Woods et al., 1997; Park et al., 2000). However, the role of the ERK1/2 pathway in the ΔMEKK3:ER*-induced cell cycle arrest has proved difficult to resolve because MEK1/2 inhibitors inhibit cell cycle re-entry per se (for example see Balmanno and Cook, 1999; Squires et al., 2002). In CR1-11 cells, activation of ΔRaf-1:ER* caused a similar enhancement of serum-stimulated ERK1 activity, but this did not increase p21CIP1 expression or cause a G1 arrest (Figure 8). A role for p38 in the expression of p21CIP1 is not entirely without precedent. Some stresses or cytokines can increase p21CIP1 expression in a p38-dependent fashion, acting at the transcriptional level (Ussat et al., 2002) or by promoting stabilization of the p21CIP1 mRNA (Wang et al., 2000) or p21CIP1 protein (Kim et al., 2002). We have shown that ΔMEKK3:ER* can induce a substantial increase in p21CIP1 mRNA and protein in HEK293 cells and both protein and mRNA expression is abolished by the combination of ERK1/2 and p38 pathway inhibitors, underscoring the reproducibility of this novel cooperation between the ERK1/2 and p38 pathways (Figure 7). Whether this represents an increase in transcription or mRNA stability remains to be determined.

There is a pre-eminent role for p53 in stress-induced p21CIP1 expression (Kastan et al., 1991; El-Deiry et al., 1993; Tishler et al., 1993; Waldman et al., 1995; Vogelstein et al., 2000) but while p53 has been linked to the MAPKs/SAPKs (Bulavin et al., 1999; Fuchs et al., 1998; Persons et al., 2000; Sanchez-Prieto et al., 2000; She et al., 2000) p53 is unlikely to be involved in ΔMEKK3:ER*-dependent p21CIP1 expression. For example, activation of ΔMEKK3:ER* did not cause stabilization of p53 in CM3 or HM3 cells (Figure 3 and 7). Furthermore, while CM3 cells express immune-reactive p53, it seems to be nonfunctional since the protein is not stabilized by DNA damage and CM3 cells fail to express p21CIP1 or arrest at G1 following DNA damage (Garner et al., 2002); the same applies to p53 in HM3 cells (R Densham and S Cook, unpublished observations). The ERK1/2 pathway can induce p21CIP1 expression by p53-dependent and independent pathways (McMahon and Woods, 2001). While the strong activation of ERK1/2 seen with ΔB-Raf:ER* may be sufficient for the strong expression of p21CIP1 and cell cycle arrest (Woods et al., 1997), the more modest levels of ERK1/2 activity attained here with ΔRaf-1:ER* or ΔMEKK3:ER* may require cooperation from other pathways, such as p53 (missing in CCl39 and HEK293 cells) or p38 (provided by ΔMEKK3:ER* but not ΔMEKK1:ER*), to induce growth inhibitory levels of p21CIP1.

Activation of JNK by ΔMEKK1:ER* fails to inhibit the G1 → S transition

Interestingly, no obvious role for the JNK pathway in regulating the G1 → S transition emerges from this study. We tried using the ‘JNK inhibitor’ SP600125 to see if JNK was required for the effects of ΔMEKK3:ER* but doses of up to 30 μ M were required to cause even a partial reduction in c-Jun phosphorylation and resulted in the inhibition of DNA synthesis and induction of apoptosis (K Balmanno and S Cook, unpublished results), perhaps reflecting the nonspecific nature of this molecule (Bain et al., 2003). As an alternative strategy we used ΔMEKK1:ER*, which selectively activates the JNK pathway (Molton et al., 2003). In three different cell lines, sustained JNK activation failed to inhibit FBS-stimulated cell cycle re-entry; this included HCT116 cells, which possess a functional p53 → p21CIP1 pathway. While this conflicts with the notion that JNK-dependent phosphorylation stabilizes p53 (Fuchs et al., 1998) it may simply mean that JNK alone is not sufficient to activate p53. For example, both JNK and p38 have been shown to phosphorylate p53, allowing binding of the peptidyl-prolyl isomerase Pin1 (Zacchi et al., 2002) and their combined effects may be important for the activation of p53. Alternatively, JNK-dependent regulation of p53 may be linked to the proapoptotic functions of p53 but not its cell cycle arrest function. Indeed, Jnk1−/−Jnk2−/− MEFs exhibit a proliferative defect but still undergo a G1 arrest in response to DNA damage (Tournier et al., 2000), suggesting that JNK is required for normal cell proliferation but not for a p53-mediated G1 cell cycle arrest. Our results with ΔMEKK1:ER* are perhaps more in accord with this model. Activation of JNK by ΔMEKK1:ER* did delay completion of the cell cycle after the S phase, causing arrest at G2/M, and future studies should aim to address how JNK interacts with the G2 checkpoint machinery.

To conclude, we have used conditional protein kinases as a means of activating defined MAPK and SAPK signalling pathways in the absence of cell stress. Activation of ΔMEKK3:ER* mimics the DNA damage response by inhibiting the G1 → S progression by two kinetically distinct mechanisms. Our data suggest that the ERK1/2 and p38 pathways cooperate to promote p21CIP1 induction and this is necessary to ensure a sustained G1 cell cycle arrest, whereas the more transient delay may reflect the inhibition of Cdc25A and cyclin D1 expression. In contrast, selective activation of the JNK pathway fails to impact on cell cycle re-entry. It is hoped that the combined use of a selection of kinase:ER* fusions and pharmacological inhibitors will help to further define the role of individual MAPK and SAPK signalling cassettes in regulating the G1 → S transition.

Materials and methods

Materials

4-HT was purchased from Sigma, SB203580 and PD184352 were from Calbiochem. 4-HT was prepared as a 1 mM stock in ethanol and stored at −70°C. SB203580 (2 mM) and PD184352 (20 mM) were made up in DMSO and stored at −20°C. Foetal bovine serum (FBS) and all other cell culture reagents were from Gibco Life Technologies. The following antibodies were used: Myc tag antibody (9E10); CDK2 (sc-163), CDK4 (sc-260), Cdc25A (sc-7389), cyclin E (sc-481) and p107 (sc-318) were from Santa Cruz; cyclin D1 (CC12) and p27Kip1 (NA35) from Calbiochem; cyclin A (MS-384-P) and p53 (MS-104-P) from Neo Markers; p21Cip1 (556431) from Pharmingen; Rb C-terminal control antibody (9302), phospho-IκB (9241), IκB (9242), MAPKAP-K2 (3042), phospho-MAPKAP-K2 (3044) and phospho-ERK1/2 (9106) were from Cell Signalling Technology. Horseradish peroxidase-conjugated secondary antibodies were from Bio-Rad, and detection was with the enhanced chemiluminescence (ECL) system (Amersham). Protease inhibitors were from Sigma. All other reagents were of the highest grade commercially available.

Cells, cell culture and cell synchronization

CCl39 and Rat-1 cells expressing ΔMEKK3:ER* (CM3 and RM3 respectively) were described previously (Garner et al.,. 2002). WT and p21CIP1−/− 3T3 cells, provided by Amer Mirza and Martin McMahon (UCSF Cancer Centre), were infected with ecotropic retrovirus generated by packaging pBABEPuro-ΔMEKK3:ER* through BOSC23 cells, followed by selection in 2 μg/ml puromycin. HM3 cells were derived by transfecting HEK293 cells with pBABEPuro-ΔMEKK3:ER* followed by limiting dilution, selection in 2 μg/ml puromycin and ring cloning. CCl39 cells expressing ΔRaf-1:ER* (CR1-11 cells) or ΔMEKK1:ER* (CM1 cells) were described previously (Molton et al., 2003; Weston et al., 2003). HCT116-M1 cells expressing ΔMEKK1:ER were derived by infecting with amphotropic retrovirus generated by packaging pBABEPuro-ΔMEKK1:ER* through Phoenix-Ampho cells; experiments were performed on a cell line derived from pools of drug-resistant cells. All cells were maintained in Dulbecco's modified Eagle medium DMEM (w/o phenol red) supplemented with glutamine, penicillin–streptomycin, 10% (v/v) FBS and 2 μg/ml puromycin. Cells were synchronized (G0/G1) by placing in serum-free DMEM for 20 h. The quiescent cells were stimulated to re-enter the cell cycle by addition of 10% FBS in the presence or absence of 4-HT; 0.04% (v/v) ethanol served as a vehicle control.

Preparation of cell lysates, Western blotting and assay of ERK, JNK, p38 and CDK2 kinase activities

These were performed as described previously (Balmanno and Cook, 1999; Garner et al.,. 2002; Weston et al., 2003). All cell extracts were normalized for cell protein content before proceeding to appropriate assays. For kinase assays phosphorylation signals were analysed using a phosphorimager and are expressed as radioactivity incorporated into appropriate substrate in phosphorimager units.

Real-time RT–PCR

Preparation of total RNA was performed as described previously (Weston et al.,. 2003). RT–PCR was performed according to the protocol supplied with the Taqman® Reverse Transcription reagents (Applied Biosystems) as described previously (Weston et al., 2003). For human p21CIP1 we used 5′-CCAGGTGGACCTGGAGACTCT-3′ as the forward primer and 5′-GGCTTCCTCTTGGAGAAGATCAG-3′ as the reverse primer. For human GAPDH we used 5′-AACAGCCTCAAGATCATCAGCAA-3′ as the forward primer and 5′-CATGAGTCCTTCCACGATACCAA-3′ as the reverse primer. Values for p21CIP1 were normalized to those for GAPDH.

Flow cytometry and assay of DNA synthesis

For analysis of cell cycle re-entry, quiescent cells were stimulated with 10% FBS in the presence of 100 nM 4-HT or 0.04% ethanol (vehicle control). At the indicated times, the cells were trypsinized, stained with propidium iodide (50 μg/ml) and analysed by flow cytometry using a FACSCalibur flow cytometer (Becton Dickinson) as described previously (Garner et al., 2002). [3H]thymidine incorporation was assayed as described previously (Balmanno and Cook, 1999).

References

  1. Agami R and Bernards R . (2000). Cell, 102, 55–66.

  2. Allan LA, Duhig T, Read M and Fried M . (2000). Mol. Cell. Biol., 20, 1291–1298.

  3. Bain J, McLauchlan H, Elliott M and Cohen P . (2003). Biochem. J., 371, 199–204.

  4. Balmanno K and Cook SJ . (1999). Oncogene, 18, 3085–3097.

  5. Blomberg I and Hoffmann I . (1999). Mol. Cell. Biol., 19, 6183–6194.

  6. Brugarolas J, Chandrasekaran C, Gordon JI, Beach D, Jacks T and Hannon GJ . (1995). Nature, 377, 552–557.

  7. Bulavin DV, Saito S, Hollander MC, Sakaguchi K, Anderson CW, Appella E and Fornace Jr AJ . (1999). EMBO J., 18, 6845–6854.

  8. Buschmann T, Potapova O, Bar-Shira A, Ivanov VN, Fuchs SY, Henderson S, Fried VA, Minamoto T, Alarcon-Vargas D, Pincus MR, Gaarde WA, Holbrook NJ, Shiloh Y and Ronai Z . (2001). Mol. Cell. Biol., 21, 2743–2754.

  9. Casanovas O, Miro F, Estanyol JM, Itarte E, Agell N and Bachs O . (2000). J. Biol. Chem., 275, 35091–35097.

  10. Cheng M, Olivier P, Diehl JA, Fero M, Roussel MF, Roberts JM and Sherr CJ . (1999). EMBO J., 18, 1571–1583.

  11. Clarke PR . (1995). Curr. Biol., 5, 40–42.

  12. Cook SJ, Aziz N and McMahon M . (1999). Mol. Cell. Biol., 19, 330–341.

  13. Cuenda A, Rouse J, Doza YN, Meier R, Cohen P, Gallagher TF, Young PR and Lee JC . (1995). FEBS Lett., 364, 229–233.

  14. Davis RJ . (2000). Cell, 103, 239–252.

  15. Deacon K and Blank JL . (1999). J Biol Chem., 274, 16604–16610.

  16. Deng C, Zhang P, Harper JW, Elledge SJ and Leder P . (1995). Cell, 82, 675–684.

  17. Dyson N . (1998). Genes Dev., 12, 2245–2262.

  18. Ekholm SV and Reed SI . (2000). Curr. Opin. Cell Biol., 12, 676–684.

  19. El-Deiry WS, Tokino T, Velculescu VE, Levy DB, Parsons R, Trent JM, Lin D, Mercer WE, Kinzler KW and Vogelstein B . (1993). Cell, 75, 817–825.

  20. Ellinger-Ziegelbauer H, Kelly K and Siebenlist U . (1999). Mol. Cell. Biol., 19, 3857–3868.

  21. Fuchs SY, Adler V, Pincus MR and Ronai Z . (1998). Proc. Natl. Acad. Sci. USA, 95, 10541–10546.

  22. Garner AP, Weston CR, Todd DE, Balmanno K and Cook SJ . (2002). Oncogene, 21, 8089–8104.

  23. Geng Y, Eaton EN, Picon M, Roberts JM, Lundberg AS, Gifford A, Sardet C and Weinberg RA . (1996). Oncogene, 12, 1173–1180.

  24. Gu Y, Rosenblatt J and Morgan DO . (1992). EMBO J., 11, 3995–4005.

  25. Hall M and Peters G . (1996). Adv. Cancer Res., 68, 67–108.

  26. Helin K . (1998). Curr. Opin. Genet. Dev., 8, 28–35.

  27. Hiebert SW, Chellappan SP, Horowitz JM and Nevins JR . (1992). Genes Dev., 6, 177–185.

  28. Jinno S, Suto K, Nagata A, Igarashi M, Kanaoka Y, Nojima H and Okayama H . (1994). EMBO J., 13, 1549–1556.

  29. Kaelin WG . (1999). BioEssays, 21, 950–958.

  30. Kamakura S, Moriguchi T and Nishida E . (1999). J. Biol. Chem., 274, 26563–26571.

  31. Kastan MB, Onyekwere O, Sidransky D, Vogelstein B and Craig RW . (1991). Cancer Res., 51, 6304–6311.

  32. Kim GY, Mercer SE, Ewton DZ, Yan Z, Jin K and Friedman E . (2002). J. Biol. Chem., 277, 29792–29802.

  33. Knudsen ES, Buckmaster C, Chen T, Feramisco JR and Wang JYJ . (1998). Genes Dev., 12, 2278–2292.

  34. Knudsen ES and Wang JYJ . (1997). Mol. Cell. Biol., 17, 5771–5783.

  35. LaBaer J, Garret MD, Stevenson LF, Slingerland JM, Sandhu C, Chou HS, Fattaey A and Harlow E . (1997). Genes Dev., 11, 847–862.

  36. Lavoie JN, L'Allemain G, Brunet A, Muller R and Pouyssegur J . (1996). J. Biol. Chem., 271, 20608–20616.

  37. Mailand N, Falck J, Lukas C, Syljuasen RG, Welcker M, Bartek J and Lukas J . (2000). Science., 288, 1425–1429.

  38. Marshall CJ . (1995). Cell, 80, 179–185.

  39. McMahon M and Woods D . (2001). Biochim. Biophys. Acta, 1471, M63–M71.

  40. Molnar A, Theodoras AM, Zon LI and Kyriakis JM . (1997). J. Biol. Chem., 272, 13229–13235.

  41. Molton SA, Todd DE and Cook SJ . (2003). Oncogene, 22, 4690–4701.

  42. Pagano M, Tam SW, Theodoras AM, Beer-Romero P, Del Sal G, Chau V, Yew PR, Draetta GF and Rolfe M . (1995). Science, 269, 682–685.

  43. Park J, Qiao L, Gilfor D, Yang MY, Hylemon PB, Benz C, Darlington G, Firestone G, Fisher PB and Dent P . (2000). Mol. Biol. Cell., 10, 2915–2932.

  44. Persons DL, Yazlovitskaya EM and Pelling JC . (2000). J. Biol. Chem., 275, 35778–35785.

  45. Reich NC and Levine AJ . (1984). Nature, 308, 199–201.

  46. Roovers K and Assoian RK . (2000). BioEssays, 22, 818–826.

  47. Sanchez-Prieto R, Rojas JM, Taya Y and Gutkind JS . (2000). Cancer Res., 60, 2464–2472.

  48. Schulze A, Zerfass K, Spitkovsky D, Middendorp S, Berges J, Helin K, Jansen-Durr P and Henglein B . (1995). Proc. Natl. Acad. Sci. USA, 92, 11264–11268.

  49. Sebolt-Leopold JS, Dudley DT, Herrera R, Van Becelaere K, Wiland A, Gowan RC, Tecle H, Barrett SD, Bridges A, Przybranowski S, Leopold WR and Saltiel AR . (1999). Nat. Med., 5, 810–816.

  50. She QB, Chen N and Dong Z . (2000). J. Biol. Chem., 275, 20444–20449.

  51. Sherr CJ and Roberts JM . (1999). Genes Dev., 13, 1501–1512.

  52. Shirane M, Harumiya Y, Ishida N, Hirai A, Miyamoto C, Hatakeyama S, Nakayama K and Kitagawa M . (1999). J. Biol. Chem., 274, 13886–13893.

  53. Squires MS, Nixon PM and Cook SJ . (2002). Biochem. J., 366, 673–680.

  54. Tishler RB, Calderwood SK, Coleman CN and Price BD . (1993). Cancer Res., 53, 2212–2216.

  55. Tournier C, Hess P, Yang DD, Xu J, Turner TK, Nimnual A, Bar-Sagi D, Jones NS, Flavell RA and Davis RJ . (2000). Science, 288, 870–874.

  56. Ussat S, Werner UE, Kruse ML, Luschen S, Scherer G, Kabelitz D and Adam-Klages S . (2002). Biochem. Biophys. Res. Commun., 294, 672–679.

  57. Vogelstein B, Lane D and Levine AJ . (2000). Nature, 408, 307–310.

  58. Waldman T, Kinzler KW and Vogelstein B . (1995). Cancer Res., 55, 5187–5190.

  59. Wang S, Nath N, Minden A and Chellappan S . (1999). EMBO J., 18, 1559–1570.

  60. Wang W, Furneaux H, Cheng H, Caldwell MC, Hutter D, Liu Y, Holbrook N and Gorospe M . (2000). Mol. Cell. Biol., 20, 760–769.

  61. Weinberg RA . (1995). Cell, 81, 323–330.

  62. Weston CR, Balmanno K, Chalmers C, Hadfield K, Molton SA, Ley R, Wagner EF and Cook SJ . (2003). Oncogene, 22, 1281–1293.

  63. Wisdom R, Johnson RS and Moore C . (1999). EMBO J., 18, 188–197.

  64. Woods D, Parry D, Cherwinski H, Bosch E, Lees E and McMahon M . (1997). Mol. Cell. Biol., 17, 5598–5611.

  65. Xia Y, Wu Z, Su B, Murray B and Karin M . (1998). Genes Dev., 12, 3369–3381.

  66. Yang J, Lin Y, Guo Z, Cheng J, Huang J, Deng L, Liao W, Chen Z, Liu Z and Su B . (2001). Nat. Immunol., 2, 620–624.

  67. Yujiri T, Ware M, Widmann C, Oyer R, Russell D, Chan E, Zaitsu Y, Clarke P, Tyler K, Oka Y, Fanger GR, Henson P and Johnson GL . (2000). Proc. Natl. Acad. Sci., 97, 7272–7277.

  68. Zacchi P, Gostissa M, Uchida T, Salvagno C, Avolio F, Volinia S, Ronai Z, Blandino G, Schneider C and Del Sal G . (2002). Nature, 419, 853–857.

  69. Zhang HS, Gavin M, Dahiya A, Postigo AA, Ma D, Luo RX, Harbour JW and Dean DC . (2000). Cell, 101, 79–89.

  70. Zhu J, Woods D, McMahon M and Bishop JM . (1998). Genes Dev., 12, 2997–3007.

Download references

Acknowledgements

We thank Amer Mirza and Martin McMahon for provision of WT and p21CIP1−/− 3T3 cell lines and members of the Cook Lab for useful discussions and comments. We are grateful to Geoff Morgan for maintenance of the Babraham Institute FACS facility and for timely advice and help. This work was supported by Cancer Research UK (SP2458/0201), the BBSRC Science of Ageing Initiative (202/SAG10012) and a Competitive Strategic Grant from the BBSRC. CRW was supported by a MRC PhD studentship, SAM by a BBSRC PhD studentship and RMD by a BBSRC/CASE studentship with GSK. SJC is a Senior Cancer Research Fellow of Cancer Research UK.

Author information

Correspondence to Simon J Cook.

Rights and permissions

Reprints and Permissions

About this article

Keywords

  • ERK
  • JNK
  • MEKK
  • p21CIP1
  • p38
  • Raf

Further reading