Reversion of resistance to oxaliplatin by inhibition of p38 MAPK in colorectal cancer cell lines: involvement of the calpain / Nox1 pathway

Abstract

Oxaliplatin is a major treatment for metastatic colorectal cancer, however its effectiveness is greatly diminished by the development of resistances. Our previous work has shown that oxaliplatin efficacy depends on the reactive oxygen species (ROS) produced by Nox1. In this report, we investigated Nox1 involvement in the survival mechanisms of oxaliplatin resistant cell lines that we have selected. Our results show that basal ROS production by Nox1 is increased in resistant cells. Whereas the transitory Nox1-dependent production of superoxide contributes to the cytotoxicity of oxaliplatin in sensitive cells, oxaliplatin treatment of resistant cells leads to a decrease in the production of superoxide associated with an increase of H₂O₂ and a decreased cytotoxicity of oxaliplatin. We have shown that calpains regulate differently Nox1 according to the sensitivity of the cells to oxaliplatin. In sensitive cells, calpains inhibit Nox1 by cleaving NoxA1 leading to a transient ROS production necessary for oxaliplatin cytotoxic effects. In contrast, in resistant cells calpain activation is associated with an increase of Nox1 activity through Src kinases, inducing a strong and maintained ROS production responsible for cell survival. Using a kinomic study we have shown that this overactivation of Nox1 results in an increase of p38 MAPK activity allowing the resistant cells to escape apoptosis. Our results show that the modulation of Nox1 activity in the context of anticancer treatment remains complex. However, a strategy to maximize Nox1 activation while inhibiting the p38 MAPK-dependent escape routes appears to be an option of choice to optimize oxaliplatin efficiency.

Keywords: NADPH oxidase; calpain; chemoresistance; colorectal cancer; oxaliplatin.

Figure 1. Validation of the resistance status of the selected cells

The HT29-D4, HT29-D4-Rox1 (Rox1) and HT29-D4-Rox2 (Rox2) cells were submitted to a 72-hour cytotoxicity assay in 2 dimensions (2D, A) and in 3 dimensions (3D, spheroids, D). The IC50 of oxaliplatin were then calculated using the Chou and Talalay’s method in 2D (B) and 3D models (E). The effects of oxaliplatin on spheroids are illustrated in (C). Asteriks indicate a statistical significance with p<0.05.

Figure 2. Implication of Nox1 in oxaliplatin-induced ROS production and cytotoxicity

The effects of oxaliplatin on cell viability were studied with HT29-D4, Rox1 and Rox2 cells treated with apocynin (A) or transfected with control siRNA (si Control) or Nox1 specific siRNA (si Nox1) (B). HT29-D4, Rox1 and Rox2 cells were lysed, and equal amounts of cellular proteins were processed for immunoblotting using the antibodies against Nox1, NoxA1, NoxO1 and GAPDH (C). HT29-D4, Rox1 and Rox2 cells were transfected with control siRNA (si Control) and Nox1 specific siRNA (si Nox1). The cells were lysed, and equal amounts of cellular protein were processed for immunoblotting using the antibodies against Nox1 (D). Transfected cells were also seeded in white 96-well plates to perform lucigenin assays (E) and in black 96-well plates to perform Amplex red assay (F). These cells were treated with 2 μM oxaliplatin over time (- untreated, 45 minutes (45 min), 4 hours (4h) and 24 hours (24h)). Asteriks indicate a statistical significance with p<0.05.

Figure 3. Study of calpain expression, activity and implication in oxaliplatin-induced cytotoxicity

HT29-D4, Rox1 and Rox2 cells were lysed and equal amounts of proteins were processed for immunoblotting using the antibodies against calpain 1, calpain 2 and GAPDH (A). HT29-D4, Rox1 and Rox2 cells were seeded in black 96-well plates to perform calpain activity assays with (Oxaliplatin) or without oxaliplatin (Control) (B). HT29-D4, Rox1 and Rox2 cells were transfected with control siRNA (si Control), calpain-1 specific siRNA (si Calpain 1), calpain 2 specific siRNA (si Calpain 2) or both siRNAs (si Calpain 1/2). The cells were lysed and equal amounts of proteins were processed for immunoblotting using antibodies against calpain 1 and calpain 2 (C). The transfected cells were also seeded to perform 72-hour cytotoxicity assays (C). Asteriks indicate a statistical significance with p<0.05.

Figure 4. Regulation of calpain activity

HT29-D4, Rox1 and Rox2 cells were seeded in black 96-well plates to perform calpain activity assays in the absence (control) or in the presence of a MEK inhibitor (PD98058, 2.5 μM) (A). Calpain activity was measured like previously (A) with cells incubated in the absence (Control) or in the presence of an inhibitor of ROS (DPI, 5 μM) (B). HT29-D4, Rox1 and Rox2 cells were treated with (Oxaliplatin, 2 μM) or without oxaliplatin (Control) and incubated with 10 μM of FURA-2-AM to measure the intracellular concentration of calcium (C). HT29-D4, Rox1 and Rox2 cells were transfected with control siRNA (si CTRL) or PKC d specific siRNA (si PKC delta) and seeded in black 96-well plates to perform calpain activity assays (D). Asteriks indicate a statistical significance with p<0.05.

Figure 5. Regulation of Nox1 by calpain

HT29-D4, Rox1 and Rox2 cells were transfected with control siRNA (si Control), calpain 1 specific siRNA (si Calpain 1), calpain 2 specific siRNA (si Calpain 2) or with both siRNA (si Calpain 1/2). The cells were seeded in white 96-well plates to perform lucigenin assays (A), and in 6-well plates to perform Western blots after oxaliplatin treatment (2 μM) (C). HT29-D4, Rox1 and Rox2 cells were transfected with control siRNA (si CTRL) or PKC d specific siRNA (si PKC delta) and seeded in white 96-well plates to perform lucigenin assays (B). NoxA1 was immunoprecipitated from HT29-D4 lysates and incubated with purified calpain 1. The samples were then processed for immunoblotting using antibodies against calpain 1, NoxA1 and GAPDH (Control: total lysate before IP, (D). HT29-D4 cells were transfected with a plasmid encoding NoxA1-DDK. The cells were lysed, and NoxA1 was immunoprecipitated (IP NoxA1) and incubated with purified calpain 1. The samples were then processed for immunoblotting using antibodies against calpain 1, DDK tag and GAPDH (E). HT29-D4, Rox1 and Rox2 cells were seeded in white 96-well plates to perform lucigenin assays in the absence (Control) or in the presence of Src inhibitor (F). HT29-D4 (HT29), Rox1 and Rox2 cells were transfected with control siRNA (si Control), or both calpain 1 and calpain 2 specific siRNAs (si Calpain 1/2). The cells were then lysed and the proteins were processed for immunoblotting using antibodies against Src, p-Src (tyrosine 527) and GAPDH (G). Asteriks indicate a statistical significance with p<0.05.

Figure 6. Comparison of HT29-D4, Rox1 and Rox2 signaling pathways

HT29-D4 and Rox1 cells were seeded in 6-well plates and incubated in the absence or in the presence of 2 μM of oxaliplatin for 4 hours. The cells were lysed and 0.5 μg of proteins were used for Pamgene kinase activity assay. The data were analyzed using the Bionavigator software to compare the kinase activity of HT29-D4 and Rox1. The top kinase lists obtained in the absence of in the presence of oxaliplatin are presented in (A and B), respectively. A positive normalized kinase statistic value indicate a kinase activity higher for Rox1 than for HT29-D4.HT29-D4, Rox1 and Rox2 cells were seeded in 6-well plates and were treated in the absence (Control) or in the presence of 100 μM of oxaliplatin (Oxaliplatin) for 24 hours. The cells were then lysed and 37.5 μg of proteins were used for the PathScan assay. The data were analyzed and the phosphorylation levels of p53 (C) and p38 (E) as well as the cleavage of PARP (D) were compared. The results obtained for PARP were also confirmed by Western blots using lysates of HT29-D4, Rox1 and Rox2 cells treated with or without 100 μM of oxaliplatin (D). The effects of oxaliplatin were visualized by calculating the ratio between the treated and untreated cells (F). Asteriks indicate a statistical significance with p<0.05.

Figure 7. Implication of p38 in the resistance to oxaliplatin

HT29-D4, Rox1 and Rox2 cells were seeded in 6-well plates and incubated in the absence (control) or in the presence of 2 μM of oxaliplatin. The cells were lysed and equal amounts of proteins were processed for immunoblotting using antibodies against vinculin, p38 and phospho-p38 (thr180/tyr 182) (A). HT29-D4, Rox1 and Rox2 cells were transfected with control siRNA (si Control) and Nox1 specific siRNA (si Nox1). The cells were incubated in the absence (CTRL) or in the presence of 2 μM of oxaliplatin for 45 minutes (45m), 4 hours (4h), 24 hours (24h). The cells were lysed and equal amounts of proteins were processed for immunoblotting using antibodies against vinculin, p38 and phospho-p38 (thr180/tyr 182) (B to D). Cytotoxicity assays were performed with HT29-D4, Rox1 and Rox2 treated with oxaliplatin and incubated in the absence (Control) or in the presence of SB203580, a specific inhibitor of p38 (5 μM) (E). The cleavage of PARP was studied by Western blot performed with lysates from HT29-D4, Rox1 and Rox2 cells treated with or without oxaliplatin (2 μM) +/- SB203580 (10 μM) (F). Asteriks indicate a statistical significance with p<0.05.

Figure 8. Proposed model for the regulation of oxaliplatin effects by calpains, Nox1 and p38 in sensitive and resistant colorectal cancer cells