SEARCH

SEARCH BY CITATION

Keywords:

  • cisplatin;
  • c-Abl;
  • MKK6;
  • p38 MAPK;
  • imatinib mesylate

Abstract

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Activation of p38 MAPK is a critical requisite for the therapeutics activity of the antitumor agent cisplatin. In this sense, a growing body of evidences supports the role of c-Abl as a major determinant of p38 MAPK activation, especially in response to genotoxic stress when triggered by cisplatin. Here, we demonstrate that p38 MAPK activation in response to cisplatin does not require the tyrosine kinase activity of c-Abl. Indeed, c-Abl can activate the p38 MAPK signaling pathway by a mechanism that is independent of its tyrosine kinase activity, but that instead involves the ability of c-Abl to increase the stability of MKK6. Similar results were obtained in chronic myeloid leukemia-derived cell lines, in which a chimeric Bcr/Abl protein mimics the effects of c-Abl overexpression on p38 MAPK activation. These findings may explain why a clinically used c-Abl inhibitor, imatinib mesylate, fails to inhibit the p38 MAPK pathway alone or in combination with cisplatin, and provide evidence of a novel signaling mechanism in which these antitumor agents act. © 2007 Wiley-Liss, Inc.

The p38 mitogen-activated protein kinase (MAPK) signaling pathway is implicated in a wide variety of cell functions,1 including inflammation,2 proliferation and cancer,3 and also the response to antitumor treatments.4 p38 MAPK is controlled by 2 mitogen-activated protein kinase kinases (MAPKKs), known as MKK3 and MKK6, although in response to some stimuli MKK4 can also be implicated.5 In this regard, a key molecule linking oncogenesis and genotoxic stress to the p38 MAPK pathway activation is c-Abl. This nonreceptor tyrosine (tyr) kinase, which contributes to the development of chronic myeloid leukemia (CML) when expressed as a chimeric protein Bcr/Abl, is a key molecule in the cellular response to genotoxic stress.6 In fact, c-Abl has been linked to the cellular response to several anticancer treatments.6, 7, 8, 9 Interestingly, one of the most widely used chemotherapeutic agent, cis-diamminedichloroplatinum(II), known as cisplatin (CDDP) (for a review see Ref.10), has been shown to be an excellent stimuli to study c-Abl and its role in DNA damage response.6, 11, 12, 13 In this regard, activation of the p38 MAPK and JNK signaling pathways by CDDP, a critical event for the antitumor effects of this drug,14, 15, 16, 17 has been reported to be c-Abl-dependent.18, 19, 20, 21 c-Abl has been also implicated in the cellular response to DNA damage through the control exerted on molecules such as p53 or p73. Specifically, c-Abl can control stability of these proteins by several mechanisms such as direct phosphorylation, as in the case of p73,11, 13, 22, 23 or by interference with ubiquitination machinery and the subcellular localization, as in the case of p53.24 The development of a specific c-Abl inhibitor, imatinib mesylate,25, 26 that has clearly improved the therapy of CML,27, 28 provides a unique opportunity to evaluate the role of c-Abl tyr kinase activity in cancer therapy.

In this scenario, we decided to study the molecular basis for the link between c-Abl tyr kinase activity and the p38 MAPK signaling pathway and the implications in the cellular response to CDDP. Our data demonstrate that p38 MAPK can be activated by CDDP, without requiring the c-Abl tyr kinase activity. The effect of c-Abl on the p38 MAPK pathways is not mediated by conventional mechanisms, such as the direct phosphorylation of upstream components of the MKK6-p38 MAPK kinase cascade. In fact, the specific stabilization of MKK6, which leads to p38 MAPK activation, is mediated by an independent tyr kinase mechanism. Therefore, our data provide evidences for a new model in which c-Abl can exert biochemical effects because of its expression level rather than its tyr kinase activity, which has direct implications in cancer therapy.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Cell lines, plasmid and transfections

Cells were maintained in 5% CO2 and 37°C. HEK-293T, Hop 62, HaCaT, A431, HN19 and HeLa were maintained in DMEM (Invitrogen, Carlsbad, CA) supplemented with 10% FCS plus antibiotics (Biowhittaker, Verviers, Belgium). K562, Wehi, Baf3 and Baf3 cells stably transfected with Bcr/Abl (Baf3 p210) were maintained in RPMI (Invitrogen, Carlsbad, CA), supplemented with 10% FCS plus antibiotics (Biowhittaker, Verviers, Belgium). Wehi cells media was used as a source of IL3 for Baf3 cells maintenance. Flag tag wild-type (WT) form of p38 in pCDNA3 was a generous gift from Dr. Han (The Scripps Research Institute, California). Expression vector for Flag tag MKK6, Flag MKK6 inactive (K82A) and MKK3 were kindly supplied by Dr. R. Davis (H.H.M.I and Program in Molecular Medicine, University of Massachusetts Medical School, MA). Expression vectors for HA-P38 MAPK, green fluorescence protein (GFP), HA-p73, Bcr/abl and HA-Erk5 have been previously described.9, 29, 30 HA-Src active forms in pCEFL were kindly suplied by Dr. JM Rojas (ISCIII, Madrid, Spain). GFP tag c-Abl WT and inactive kinase mutant (kinase death, KD) in PCFL were obtained by PCR from pSG5 Abl WT and KD, kindly supplied by Dr. Y. Shaul (Weizman Institute of Science, Rehovot, Israel). Briefly, primers used were forward 5′-ATAGGATCCATGGGGCAGCAGCCTGGAAAA-3′ and reverse 5′-ATAGAATTCTCAGCTAGCCCTCCGGACAATGTCGC TGAT-3′. PCR conditions were 94°C 30 sec during the first cycle and then, 30 cycles (94°C 30 sec, 52°C 1 min and 72°C 3 min) with a final extension of 72°C during 10 min. The PCR products were cloned in pCEFL-GFP vectors using the BamH1/EcoR1 sites. DNAs were confirmed by automatic sequencing. HEK 293T cells were transfected using calcium phosphate technique following standard procedures. For RNA interference, cells were transfected by lipofectamine 2000 (Invitrogen, Carlsbad, CA) following manufacturer instructions.

Western blotting and immunoprecipitation procedures

Cells were treated and collected in lysis buffer (25 mM HEPES pH 7.5, 0.3 M NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 1% Triton X-100, 0.1% SDS, 0.5% deoxycholic acid, 20 mM B-glycerophosphate) plus protease and phosphatase inhibitors (2 μg/ml leupeptin, 2 μg/ml aprotinin, 1 mM PMSF and 0.1 mM Na3VO4). Then, the indicated amount was loaded onto 8–10% SDS-Page and blotted against the different antibodies. In the cases of immunoprecipitation assays, extracts were washed and then, soluble fraction was incubated with the indicated antibody (1 μg/sample for 2 hr), incubated for 45 min in the presence of protein G (gamma Bind Sepharose, Amersham) and then washed 3 times in the same lysis buffer. Antibody detection was achieved by enhanced chemiluminescence (Amersham Pharmacia, Uppsala, Sweden). Protein quantification was performed using BCA Protein Assay Kit (Pierce, Rockford, IL) following manufacturer instruction. Quantification of autoradiograms was performed by using WCIF Image J. software. Tubulin was used as an internal loading control. The fold change of activation was estimated by densitometric analysis of filters. The values (given as arbitrary units) were normalized taking into consideration the total protein levels of each sample as determined by immunoblot analysis. Results show a representative blot out of 3.

Treatment, chemicals and antibodies

Antibodies against activated forms of p38 MAPK, MKK3/6 and total Erk5 were from (Cell Signaling Technologies, Beverly, MA) Antibodies against total p38 MAPK, GFP and MKK3 were from Santa Cruz Biotechnology. Anti phospho-tyr clone 4G10 was from Upstate. Cycloheximide (CHX) and antibodies against Flag, hemagglutinin (HA) and tubulin were from Sigma. Antibodies against MKK6 and c-Abl were from Stratagene and Pharmingen, respectively. UV irradiation was performed at 120 mJ, using a UV-Stratalinker 1800 (Stratagene) and collected 30 min after UV exposure. imatinib mesylate was kindly supplied by Elisabeth Buchdunger (Novartis-Pharma, Basel, Switzerland). Cisplatin was purchased from Chiesi.

RNA interference assays

shRNA against c-Abl was purchased from Sigma (Catalog number SHDNA-NM_009594, mission RNAi). We selected the best performing shRNA for further analysis as judged for functional interference with endogenous c-Abl by Western blotting. Cells were transfected and 72 hr later were tested for endogenous expression of c-Abl, MKK3 and MKK6.

Viability assays

Viability was evaluated by the crystal violet method.14 Cells were seeded 24-hr prior drug treatment at 4 × 104 cells/well. In the case of imatinib, cells were pretreated for 45 min prior to CDDP. Colorant was recovered using 1% acetic acid and optical density was evaluated at 590 nm. Values were reoffered to untreated controls. Data are the average of at least 3 independent experiments performed in triplicates cultures.

RNA isolation, reverse transcription and real-time quantitative PCR

Total RNA was obtained with the Qiagen RNA isolation kit, following the manufacturer protocol. The isolated RNA was subsequently treated with DNase (Promega, Madison, WI) to remove any contaminating genomic DNA. The integrity of RNA was always checked by running an aliquot in an agarose gel. Reverse transcription was performed using 1 μg of DNase-treated RNA in 20 μl of reaction volume (Fermentas, Glen Burnie, MD). Samples were stored at 20°C until their utilization. Changes in the mRNA expression of MKK6 were examined by real-time quantitative PCR using an ABIPrism 7000 Sequence Detection System (Applied Biosystems, Foster City, CA). cDNA (5 μl of diluted reverse transcription product) was amplified using SYBR1 Green PCRMaster Mix (Applied Biosystems) in the presence of primer oligonucleotides specific for MKK6 and GADPH. The PCR conditions were as follows: 95°C for 10 min, followed by 40 cycles consisting of 95°C for 15 sec and 60°C for 1 min. The quantification was performed by the comparative cycle threshold method, using the GADPH RNA expression level as internal control. The presence of c-Abl did not modify the GADPH mRNA levels. Primers for all target sequences were designed using the computer Primer Express software program specially provided with the 7000 Sequence Detection System (Applied Biosystems, Foster City, CA). In all the cases, only one amplification product was obtained. Chosen PCR primers were as follows:

MKK6 sense 5′-GGCCCCTGAAAGAATAAACCC-3′, antisense 5′-CGAAGGATGGCCAACTCAATC-3′ and for GADPH sense 5′-TCGTGGAAGGACTCATGACCA-3′ antisense 5′-CAGTCTTCTGGGTGGCAGTGA-3′.

Results

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

p38 MAPK is activated by CDDP in the presence of c-Abl tyrosine kinase activity inhibitor imatinib mesylate

Several evidences support p38 MAPK as a key mediator in the response to CDDP (for a review Ref.31). Among the several molecules that control p38 MAPK activation in response to genotoxic stress, c-Abl seems to be a critical player.9, 18, 21, 32 To further support this idea, we took advantage of the availability of imatinib mesylate, a specific c-Abl tyrosin kinase inhibitor, formerly known as STI571.25 Therefore, several human cell lines, representatives of tumors usually treated with CDDP, such as epidermoid carcinoma (A431), head and neck (HN19), cervix (Hela) or lung (Hop62), were exposed to CDDP with or without imatinib mesylate pretreatment. Contrary to expectation, p38 MAPK activation by CDDP was unaffected by c-Abl inhibitor (Fig. 1a). In fact, we also did not detect any effect in terms of viability due to the presence of imatinib (Fig. 1a). As a control, a nontumorigenic cell line, such as HaCaT, was also tested for p38 MAPK activation and viability, showing the same result (Fig. 1b). Expression levels of c-Abl were confirmed in all the cell lines tested (Fig. 1c). The functionality of imatinib mesylate was tested in an experimental system of HEK 293T overexpressing c-Abl by using a well-characterized substrate of this tyr kinase, such as p73, and testing the overall tyr phosphorylation pattern (Fig. 1d). In fact, in the same cell line, p38 MAPK activation by CDDP was also unaffected by the presence of imatinib mesylate (Fig. 1e). Although these results suggested a lack of involvement of c-Abl in the activation of p38 MAPK by CDDP, we decided to verify them by knocking down c-Abl expression by RNA interference approaches. Interestingly, a marked decrease in the p38 MAPK activation by CDDP was observed in cells transfected with the shRNA for c-Abl when compared to those transfected with a control vector (Fig. 1f). Taken together, these findings suggest that the expression levels of c-Abl, rather than its kinase activity, could be major determinants of p38 MAPK activation in response to CDDP.

thumbnail image

Figure 1. Inhibition of c-Abl tyr kinase activity by imatinib mersylate treatment does not modify the p38 MAPK activation. (a) Left panel: A431, HN19, HeLa and Hop62 were incubated with CDDP (20 μg/ml) with or without pretreatment of imatinib mesylate for 45 min. After 2 hr, samples were collected and p38 MAPK activation was evaluated by Western blotting by using 20 μg of total cell lysates (TCL). As a control, membranes were reproved with anti-tubuline. Right panel: A431, HN19, HeLa and Hop62 were treated during 24 hr with the indicated doses of CDDP in the presence/absence of imatinib mesylate (10 μM). Viability was referred to untreated cells. Bars mean standard deviation (SD). ♦ means CDDP alone and ▪ CDDP plus imatinib mesylate pretreatment. (b) Same experimental conditions but in HaCat Cells. (c) HEK 293T cells were transfected with 2 μg of a plasmid coding for c-Abl (left panel) for being used as a positive control for immunoblotting detection of endogenous c-Abl in HaCaT, A431, HN19, Hop 62 and HeLa using 100 μg of TCL. (d) HEK-293T cells were transfected with HA-p73 plus the indicated amounts of plasmid coding for c-Abl or GFP. Twenty-four hours after transfection, cells were incubated with the indicated doses and times of imatinib mesylate and p73 tyr phosphorylation, and total amount of protein and overall tyr phosphorylation pattern were evaluated as in (a). (e) HEK-293T cells were incubated with CDDP (20 μg/ml) with or without pretreatment of imatinib mesylate for 45 min. After 2 hr, samples were collected and p38 MAPK activation was evaluated by Western blotting as in (a). Tubulin was used as a loading control. (f) HEK-293T cells were transfected with a shRNA for c-Abl or the shRNA empty vector (3 μg). Thirty-six hours after transfection, cells were splitted and 36 hr later exposed to CDDP (20 μg/ml) during 2 hr to evaluated p38 MAPK activation. Samples were collected as in (a) and blotted against the indicated proteins using 60 μg of TCL. Tubulin was used as a loading control. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Download figure to PowerPoint

c-Abl overexpression induces p38 MAPK activation by a specific increase of MKK6

In light of our findings we decided to study the role of c-Abl tyrosin kinase activity in the activation of p38 MAPK pathway. We first verified that the overexpression of c-Abl was sufficient to promote p38 MAPK activation. Therefore, HEK 293T cells were transfected with Flag-tagged p38 MAPK (Flag-p38 MAPK) plus increasing amounts of GFP-tagged c-Abl (GFP-c-Abl), detecting a clear activation of the MAPK (Fig. 2a). Hence, both natural upstream molecules of p38 MAPK were analysed in the same conditions. No variation in terms of activation was observed for MKK3 (Fig. 2b), while a marked increase in the phosphorylation was detected in the case of MKK6 (Fig. 2c). As a loading control, membranes were reproved against Flag. While protein levels of MKK3 remained unaffected (Fig. 2b), in the case of MKK6 a robust increase in the amount of Flag-MKK6 was detected (Fig. 2c). We used UV light, which induces a marked increase in MKK6 phosphorylation, to exclude that activation of MKK6 per se can affect the stability of MKK6 (Fig. 2b). To determine whether the increase in MKK6 expression level is specifically mediated by c-Abl, a similar approach was performed in the presence of another member of this nonreceptor tyr kinase super family. Overexpression of a hyperactive form of c-Src has no effect on MKK6 or MKK3 (Fig. 2d). c-Src tyr kinase activity was evaluated by using 4G10 antibody (data not shown). Next, we decided to prove if the increased levels of MKK6 can mediate p38 MAPK activation. As it is shown, HA-tagged p38 MAPK was clearly activated by the presence of increasing amounts of Flag-MKK6, in the absence of any other stimuli (Fig. 2e). To evaluate the role of MKK6 kinase activity in the c-Abl-mediated activation of p38 MAPK, we used an inactive form of this MAPKK, whose overexpression was not able to activate p38 (data not shown). This nonfunctional kinase suppressed p38 MAPK activation mediated by c-Abl (Fig. 2f), supporting the critical role of MKK6 activity. However, the functionality of MKK6 does not interfere with the effect that c-Abl has on this MAPKK, whose total amount was increased by the overexpression of c-Abl in a similar manner to the WT form (Fig. 2g).

thumbnail image

Figure 2. c-Abl overexpression induces phosphorylation of p38 MAPK. (a) HEK-293T cells were transfected with 0.2 μg Flag-p38 MAPK alone or plus the indicated amounts of GFP-c-Abl. Cells were collected 36 hr after transfection and immunoprecipitated (IP) against Flag. Immunocomplexes were blotted against activated p38 (P-p38) and then, against Flag as a loading control. Twenty microgram of TCL was blotted against Flag and GFP. (b) HEK-293T cells were transfected with 0.2 μg of Flag-MKK3 alone or plus the indicated amounts of GFP-c-Abl. Samples were processed as in (a). UV treatment was used as a positive activation control. (c) Cells were transfected and processed as in (b), but using 0.2 μg Flag-MKK6. (d) HEK-293T cells were transfected with 0.2 μg Flag-MKK6 or Flag-MKK3 alone or plus the indicated amounts of hyper HA-Src. Total cell lysate (20 μg) was blotted against Flag and HA. (e) HEK-293T cells were transfected with 0.2 μg HA-p38 MAPK alone or plus the indicated amounts of Flag MKK6. Cells were collected 36 hr after transfection and IP against HA. Inmunocomplexes were blotted against P-p38 and then, against HA as a loading control. (f) HEK-293T cells were transfected with 0.2 μg HA-p38 MAPK alone or with 2 μg of GFP-c-Abl in the absence and presence of an inactive form of MKK6 (MKK6 KD). Cells were collected 36 hr after transfection and IP against HA. Inmunocomplexes were blotted against P-p38 and then against HA as a loading control. (g) Cells were transfected and processed as in (b), but using 0.2 μg Flag-MKK6 KD. (h) HEK-293T cells were transfected with GFP alone or the indicated amounts of GFP-c-Abl. Endogenous levels of MKK3/6 were analysed by inmunoblotting using 20 μg of TLC. Standard deviation (SD).

Download figure to PowerPoint

Finally, to challenge our observation, mainly based in the overexpression of the aforementioned molecules, we decided to evaluate the effect of c-Abl on both endogenous MAPKKs (MKK6 and MKK3). HEK 293T cells, which express detectable protein levels of endogenous MKK6 and MKK3, were transfected with increasing amounts of GFP-c-Abl. A clear increase was observed on the endogenous MKK6, with no effect on MKK3 (Fig. 2g). In summary, this set of experiments demonstrates that overexpression of c-Abl specifically increases the levels of MKK6 mediating p38 MAPK activation.

In this context, we decided to study the possibility that c-Abl could specifically increase the stability of MKK6. HEK 293T cells were transfected with Flag-MKK6 or Flag-MKK3 in the presence or absence of GFP-c-Abl. Twelve hours after transfection, cells were incubated for different lengths of time with CHX, ranging from 6 up to 36 hr. No differences were detected in the first 6 hr of treatment with CHX for any of the MAPKK (data not shown). At 12 and 24 hr, the CHX effect was clearly diminished by the presence of c-Abl in the case of MKK6 when compared with MKK3 (Fig. 3a). Similar results were obtained at 30 and 36 hr time-points (data not shown). Note that the small difference observed between cells transfected with MKK6 and c-Abl in the presence/absence of CHX is due to the effect of CHX on the transfected c-Abl (Fig. 3b). However, to exclude any effect at the transcriptional level, real-time PCR assays were performed in the presence of increased amounts of c-Abl. No increment was observed in MKK6 RNA levels due to the overexpression of GFP-c-Abl (Fig. 3c). As a control, c-Abl and MKK6 protein levels were evaluated by Western Blotting as in Figure 2h (data not shown). Therefore, this set of experiments supports the role of c-Abl on MKK6 at the protein level. After this, and to corroborate our observations based on overexpression approaches, we decided to use RNAi technology to knock down endogenous c-Abl. Therefore, HEK 293T cells were transfected with a plasmid coding for a shRNA against c-Abl or a control vector, and endogenous MKK6 and MKK3 protein levels were evaluated (Fig. 3d, left panel). Decreased c-Abl levels induced by shRNA transfection correlated with lower MKK6 expression (>50%) and a slight decrease in p38 MAPK basal phosphorylation (<15%), while MKK3 remained unaffected. Furthermore, the same result was obtained in Hela cells, which are also known to have a functional c-Abl (Fig. 3d right panel). These experiments support our previous data and demonstrate that c-Abl expression is a key player in controlling the levels of MKK6, with obvious implications in the p38 MAPK activation.

thumbnail image

Figure 3. c-Abl overexpression increases specifically the stability of MKK6. (a) HEK-293T cells were transfected with 0.2 μg of MKK6 or MKK3 with or without GFP-c-Abl. Sixteen hours after transfection, cells were exposed of 40 μM cycloheximide (CHX) for the indicated times. Twenty microgram of TLC was immunoblotted against Flag. (b) Same extracts were blotted against GFP to detect GFP-c-Abl. Standard deviation (SD). (c) RNA obtained from control cells and cells transfected with the indicated amounts of c-Abl were used for the quantification of MKK6 mRNA expression levels. After reverse transcription, samples were analyzed by quantitative real-time PCR. GAPDH expression levels were used as internal control. Figures represent the relative expression referred to control. Bars mean SD. The graphic shows the average of 3 independent experiments. (d) HEK-293T (left panel) and Hela cells (right panel) were transfected with 5 μg of control plasmid or plasmid coding for shRNA for c-Abl. Seventy-two hours after transfection cells were collected and immunoblotted against endogenous c-Abl, MKK6, MKK3, active form of p38 MAPK 60 μg of TCL and, as a loading control, tubulin.

Download figure to PowerPoint

The c-Abl tyr kinase activity is not required for MKK6 stabilization

In order to gain further insight into the molecular mechanism of this effect, we investigate the possibility of a direct binding and/or phosphorylation, as it has been described for p73.13 Although no binding was detected between both molecules, as judged by coimmunoprecipitation assays (data not shown), we decided to challenge the possibility that c-Abl may directly phosphorylate specifically MKK6. HEK 293T cells were transfected with Flag-MKK6 or Flag-MKK3 alone or in the presence of GFP-c-Abl. Samples were IP with Flag and inmunoblotted against tyr phosphorylation. No tyr phosphorylation was detected on any MAPKKs (Fig. 4a), while a marked and specific tyr phosphorylation was detected on HA-p73 in the presence of c-Abl (Fig. 4b).

thumbnail image

Figure 4. Direct phosphorylation does not mediate the MKK6 stabilization by c-Abl. (a) HEK-293T cells-transfected with 0.2 μg of Flag-MKK3 or Flag-MKK6 alone or plus 2 μg of GFP-c-Abl. Cells were collected, IP and immunoblotted against P-Tyr. (b) HEK-293T cells were transfected and processed as in (a), but using 0.2 μg of HA-p73.

Download figure to PowerPoint

In this scenario, we explored the contribution of the c-Abl tyr kinase activity in c-Abl mediated MKK6 stabilization. As a first step, we overexpressed Flag-MKK6 or Flag-MKK3 in the presence of c-Abl WT or an inactive form (KD). Surprisingly, a marked increase in Flag-MKK6 was detected in the presence of both forms of GFP-c-Abl (Fig. 5a), although the effect of this KD form was slightly lower (15%). No effect was observed on Flag-MKK3 (Fig. 5a). To corroborate this observation, we also evaluated endogenous MKK6 and MKK3, with nearly identical results (Fig. 5b). To demonstrate the lack of functional activity of our c-Abl KD form, we measured its effect on p73, MKK6 and also the overall pattern of tyr phosphorylation. While the c-Abl WT form induces a marked increase in all the cases, the c-Abl KD form only showed a marked increase on Flag-MKK6 (Fig. 5c). Furthermore, we challenge whether this KD was able to activate p38 MAPK in transient transfection assays. This inactive form showed a strong activation of the MAPK (Fig. 5d), but to a lower extent than WT (30% less, see Fig. 2a). This last result suggests that another mechanism related to the tyr kinase activity could also act for the full p38 MAPK activation mediated by c-Abl, in addition to the independent-kinase MKK6 stabilization. In order to abolish any c-Abl tyr kinase activity, endogenous or exogenous, we decided to use imatinib mesylate. A minimum effect was detected on Flag-MKK6 expression when cells were exposed to this c-Abl inhibitor, further supporting the limited role of the c-Abl tyr kinase activity in the activation of the p38 MAPK pathway (Fig. 5e). Again, p73 was used as an internal control. HA-p73 levels showed a dramatic decrease (Fig. 5e) due to the critical role of the c-Abl tyr kinase activity in the stabilization of this particular protein. In fact, expression levels of MKK6 after CDDP exposure were tested in the same experimental systems used in Figure 1 and no difference was observed (Supplementary Information 1).

thumbnail image

Figure 5. Tyrosine kinase activity of c-Abl is not required for MKK6 stabilization. (a) HEK-293T cells were transfected as in Fig. 1b and 1c, using the indicated amount of GFP-c-Abl (WT) or GFP-c-Abl inactive form (KD). Flag MKK3/6 levels were evaluated by immunoblotting using TCL (20 μg). As a loading control, membranes were reproved against tubulin. (b) Endogenous level of MKK3 and MKK6 was evaluated using 20 μg of TCL, after transfection of the indicated amounts of c-Abl KD form. As a loading control, membranes were reproved against tubulin. (c) HEK-293T cells were transiently transfected as in Figure 4a using 0.2 μg HA-p73 or Flag MKK6 with 2 μg of plasmid coding for GFP, WT or KD forms. Tyr phosphorylation, HA-p73 and Flag-MKK6 were evaluated by immunoblotting in TCL (20 μg). Tubulin was used as a loading control. (d) HEK-293T cells were transfected with 0.5 μg of HA-p38 MAPK and 2 or 4 μg of KD form. Samples were processed as in Figure 1a (e) HEK-293T cells were transfected as in panel (c). Twelve hours after transfection, cells were treated with imatinib mesylate (10 μM) for the indicated times. Expressions of HA-p73 or Flag MKK6 were evaluated in TCL (20 μg). As a loading control, TCL (20 μg) was inmunoblotted in a separate membrane against tubuline. Standard deviation (SD).

Download figure to PowerPoint

In summary, all the evidences using genetic and pharmacological approaches indicate that the tyr kinase activity is not strictly required for c-Abl-mediated MKK6 stabilization and the subsequent p38 MAPK activation.

The oncogenic form of c-Abl, Bcr/Abl, also promotes MKK6 stabilization

Finally, we decided to study the effect of oncogenic form of c-Abl, Bcr/Abl, on MKK6 expression levels. HEK 293T cells were transfected with Flag-MKK6 and increasing amounts of Bcr/Abl expression vector. As expected, we detected an increase in the activation of MKK6 (Fig. 6a). Blots were reproved against Flag, showing a robust increase on Flag-MKK6 expression in those cells transfected with Bcr/Abl (Fig. 6a). In fact, an increase close to 100% in MKK6 expression level was also observed in Baf 3 cells stably transfected with Bcr/abl (Baf3 p210) when compared with Baf3 cells (Fig. 6b).

thumbnail image

Figure 6. Bcr/Abl also induces MKK6 stability without tyr kinase implication. (a) HEK 293T cells were transfected and processed as in Figure 1b, but using a plasmid coding for Bcr/Abl. (b) Expression levels of Bcr/Abl and MKK6 were evaluated in Baf3 and Baf3 p210 using TCL (50 μg). As a loading control, tubulin expression level was assessed. (c) HEK 293T cells were transfected with the indicated amount of Bcr/abl plus Flag-MKK6 or HA-ERK5. Twelve hours later, cells were exposed to imatinib mesylate for 16 hr and the level of Flag MKK6 or HA-ERK5 were evaluated in TCL (20 μg) (d) K562 cells were exposed to imatinib mesylate for 16 hr at the indicated doses. Samples were lysed and 50 μg from TCL were immunoblotted with the indicated antibodies. (e) K562 cells were untreated or exposed to CDDP (40 μg/ml) for 2 hr in the presence/ absence of imatinib mesylate (1 μM) pretreatment. Samples were collected and 50 μg from TCL was analyzed by Western blotting for p38 MAPK activation, MKK6 and overall tyr phosphorylation pattern. Membranes were reblotted against tubulin as a loading control.

Download figure to PowerPoint

The next step was to evaluate whether the effect of Bcr/Abl overexpression on MKK6 was also independent of the tyr kinase activity. Cells were transfected with Flag-MKK6 or HA-tagged extracellular regulated kinase 5 (Erk5), a protein whose stability is affected by the tyr kinase activity of Bcr/Abl,33 plus GFP or Bcr/abl and then incubated with or without imatinib mesylate. While Flag-MKK6 remained unaffected by the presence of imatinib mesylate, HA-Erk5 showed a marked decrease in the total amount of protein (Fig. 6c). Taking into account our last results, we decided to investigate the effect of imatinib mesylate on MKK6 and Erk5 molecules in a CML-derived cell line, such as K562, with a natural translocation Bcr/abl. Thus, K562 cells were exposed to imatinib mesylate and the levels of both proteins were evaluated. The levels of endogenous MKK6 were almost identical regardless of the presence of imatinib mesylate while Erk5 underwent a dramatic decrease (Fig. 6d) supporting our observation in HEK293T cells. Finally, to confirm our initial observation, lack of involvement for c-Abl tyr kinase activity in the p38 MAPK activation by CDDP, K562 cells were exposed to CDDP in the presence/absence of imatinib and p38 MAPK activation was evaluated. While Bcr/abl tyrosine activity was almost abolished by the presence of imatinib mesylate, as indicated by the tyrosine phosphorylation pattern, no effect on MKK6 levels was detected and only a slight decrease was detected in the p38 MAPK activation by the presence of imatinib mesylate (Fig. 6e). All this data together confirm the marginal of c-Abl tyr kinase activity even in the context of Bcr/Abl overexpression.

Discussion

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Several conclusions can be drawn from the present study. Our initial objective was to study the role of c-Abl in the p38 MAPK activation mediated by CDDP. However, we observed that the tyr kinase activity of c-Abl is not implicated in the activation of p38 MAPK in response to this anticancer agent. This first observation is apparently contradictory to other reports, which proposed that c-Abl is a critical mediator in the cellular response to genotoxic stress through the control exerted on molecules such as p73,11, 13 p53,12 JNK6, 19 or p38 MAPK.18, 21 Interestingly, most of the experimental data linking p38 MAPK with c-Abl, including those in which CDDP has been used, were inferred from knock out models or by overexpression of c-Abl or Bcr/abl, assuming that the effects observed in the p38 MAPK were due to the lack or abundance of the tyr kinase activity of c-Abl. The fact that MKK6 levels are not affected by CDDP in the variety of experimental systems used supports the lack of involvement of c-Abl tyrosine kinase in the p38 MAPK activation by CDDP. In fact, this antitumor agent does not affect c-Abl expression levels (i.e. this report and Ref.11), and thus it is not expected to affect MKK6 levels, as we demonstrate here. Furthermore, this observation also indicates the existence of a c-Abl-independent mechanism activating p38 MAPK in response to CDDP, in which MKK3 could play a role, as we have previously reported.14 Therefore, the main role of c-Abl in p38 MAPK activation by CDDP seems to be the maintenance of the expression levels of MKK6, but not the direct activation of the MAPKK for p38 MAPK in response to CDDP. Hence, the present study provides evidences that demonstrate how the enzymatic activity of c-Abl is not implicated in the activation of p38 MAPK by neither CDDP nor the overexpression of c-Abl, and it gives a biochemical explanation for recent clinical evidences that demonstrate a lack of effect for the therapeutic combination of imatinib mesylate with CDDP in patients with small-cell lung carcinoma.34

Our data point to a new role for c-Abl as a potent inducer of MKK6 stability, with a marked selectivity for this particular MAPKK, as further supported by the data obtained with RNAi approaches. Indeed, this stabilization seems to be mediating the vast majority of p38 MAPK activation in response to c-Abl overexpression. It has also been reported that the expression levels of several proteins, such as p73, p53 and Erk5, are controlled by c-Abl through different mechanisms that require tyr kinase activity.13, 33, 35 However, our study provides evidences that c-Abl can also control protein stability without the requirement of thec-Abl tyr kinase activity, as we have shown for MKK6. This conclusion was inferred from the data obtained by the use of an inactive kinase form of c-Abl, which was also able to increase the levels of MKK6 in a similar fashion to the WT form, and the use of imatinib mesylate, which did not prevent the increase in MKK6 levels mediated by c-Abl. Interestingly, it has been recently reported that c-Abl promotes CUL-4A-mediated DDB ubiquitination and degradation in a tyr kinase-independent manner.36 In addition, blockage of the ubiquitination, has also been demonstrated to be the main mechanism for c-Abl-mediated p53 stabilization, although in this case, the tyr kinase activity of c-Abl is strictly required.24, 35 However, c-Abl did not modify MKK6 ubiquitination pattern (Supplementary Material 2), indicating that the c-Abl-mediated MKK6 stabilization seems to be not mediated by interference with the ubiquitination machinery. Thus, further work may be required to clarify how c-Abl mediates the stabilization of this particular MAPKK, including the possibility that c-Abl may cause the indirect serine/threonine phosphorylation of MKK6 or other posttranslational modifications, including methylation, acetylation, sumoylation and/or neddylation.

Although unexpected, the lack of a critical role for the enzymatic activity of c-Abl in biochemical processes, as we showed for p38 MAPK activation or in the case of CUL-4A, is not unique. In fact, another member of this tyr kinase family, c-Src, implicated in a wide variety of signal transduction pathways, has been previously demonstrated that can exert several effects in vitro and in vivo, such as cellular adhesion or osteoclast function, in a tyr kinase-independent manner.37, 38

Regarding specifically the p38 MAPK pathway activation, our results corroborate previous reports that proposes MKK6 as the key molecule linking c-Abl to p38 MAPK.18 However, it is important to say that the conclusion of this previous work was based on the use of a MKK6 dominant negative mutant, which supported that the activation of p38 MAPK by c-Abl involves an increase in the kinase activity of MKK6. However, in light of our present findings, we may be able to postulate that the stabilization of MKK6 promoted by c-Abl in a tyr kinase independent manner may represent a critical mechanism by which c-Abl stimulates p38 MAPK. It is noteworthy that hyperactive forms of c-Abl, lacking the SH3 domain, showed similar result in MKK6 stabilization and the subsequent p38 MAPK activation than the WT form (data not shown), while c-terminal region of c-Abl has been demonstrated to be strictly required to induce p38 MAPK activation.18 In summary, our observations can explain how the overexpression of an inactive kinase c-Abl can also activate p38 MAPK and also provide new insight in the MKK6 regulation.

Other interesting issue in the present study is the data obtained with the oncogenic form of c-Abl, Bcr/Abl. This chimeric protein is an excellent model for overactivity of c-Abl in human pathology. Our data could explain the lack of effect in the p38 MAPK pathway of imatinib mesylate described in K562 cells39 and also corroborate those obtained for Erk5.33 Although previous evidences consider p38 MAPK a key molecule in the toxic effect of imatinib mesylate,40 our data support the lack of a role for this signaling pathway, due to the fact that inhibition of c-Abl tyr kinase activity seems to have a marginal role in this signaling pathway. Therefore, the use of imatinib mesylate can block the effect of Bcr/Abl in some signaling pathways, for example Erk5, while others are minimally affected as we demonstrate for the p38 MAPK pathway.

In summary, the results shown here demonstrate that p38 MAPK activation induced by c-Abl is mediated through a specific stabilization of MKK6 independent of its tyr kinase activity, showing how c-Abl can exert biochemical effects in signal transduction without the requirement of its enzymatic activity. This observation may have important implications in cancer therapy, as we have shown for CDDP. Whether this model is applicable or not to other contexts, such as c-Abl-mediated apoptosis, are currently being studied.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We appreciate the comments and suggestions of Drs. A. Nebreda, I. Sanchez-Perez, M.J. Ruiz, M.J. Diaz-Guerra, V.J. Sanchez-Arevalo and A. Mas.

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

This article contains supplementary material available via the Internet at http://www.interscience.wiley.com/jpages/0020-7136/suppmat .

FilenameFormatSizeDescription
SM1.TIF113KSupporting Information file SM1.TIF
SM2.TIF240KSupporting Information file SM2.TIF

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.