Ponatinib is a potential therapeutic approach for malignant pleural mesothelioma
Introduction
Malignant pleural mesothelioma (MPM) is a dis- mal neoplasm with an abysmal prognosis and is highly resistant to chemotherapy. It originates from mesothelial surfaces of the pleural cavity, pericardium, or peritoneal cavity. MPM repre- sents 85% of the overall mesothelioma cases.1 MPM’s predominant cause is asbestos inhala- tional, with roughly 70–80% of cases linked to asbestos exposure.2–4 The development of tumors may occur up to 45 years after exposure to asbes- tos fibers.5 There are three morphohistological MPM subgroups: epithelial, sarcomatoid, and biphasic. MPM is a relatively uncommon, but inexorably fatal carcinoma with about 3300 new cases in the US every year.6 The prevalence of the US’s disease is increasing and could reach 85,000 new MPM cases in the next decade.7 MPM is generally diagnosed in the advanced stage, with a poor survival outcome of 4–13 months when untreated8 and 6–18 months when treated with any of the currently approved MPM therapy.
Current standard therapies for MPM are inef- fective. Notwithstanding the encouraging Phase 2 results of improved median survival with plat- inum and pemetrexed combination,11 long-term survival despite these promising results, long- term survival is rarely attained with available treatments. Hence, new effective treatments for MPM patients are much needed.
cAbl is a multifaceted tyrosine kinase regulat- ing numerous pathophysiological cellular proc- esses, including adhesion, cell division, apoptosis, differentiation, and migration.12 cAbl is innately inactive but can be stimulated by several intra- and extra-cellular factors, including cells exposure to DNA damaging agents.13 In various solid tumor cells, cAbl stimulation is associated with activated tyrosine kinases and chemokine recep- tors such as EGFR, HER2, PDGFR, cMet, IGFR, and CXCR4.13–15 Also, cAbl may be activated via the silencing of its inhibitory proteins such as tumor suppressor protein FUS1,16 which is downregulated in major subsets of many cancers, including MPM.17 Subsequently, cAbl activates various signaling pathways involved in cancer growth or metastatic progression, including increased tumor cell survival, proliferation, migration, and invasion.
cAbl is strongly inhibited by ponatinib (Iclusig, AP24534), a small molecule inhibitor of the BCR/ Abl fusion protein. BCR/Abl fusion, initially dis- covered in chronic myeloid leukemia,8 leads to a constitutively active cAbl kinase and promotes cancer cell growth and survival. cAbl abnormal activity in cancer cells can be blocked with pona- tinib, an ATP-competitive inhibitor.Here, we hypothesize that inhibition of cAbl activity with ponatinib will increase DNA damage and cAbl-dependent apoptosis in mesothelioma cells, which may offer new possibilities for a novel translational therapeutic approach to MPM, based on inhibition of the cAbl signaling pathway.
Results
cAbl is expressed in human MPM tumors
Firstly, we examined cAbl gene expression in human MPM by mining public microarray data. Compared to normal lung, cAbl gene expression was amplified about four-fold in MPM compared to normal (P 0.001) (Figure 1a) (http://www. oncomine.org).Additionally, we performed immunohisto- chemistry for cAbl in 10 cases of human MPM tumors selected from our UCSF archived surgical specimens. All 10 cases of MPM showed con- spicuous and varying degrees of staining for cAbl in the tumor cells. The signal was enriched in the nucleus (Figure 1b).
Ponatinib, a cAbl potent inhibitor, reduces the growth of MPM cell lines in a dose- dependent manner
Next, we sought to determine the effect of pona- tinib on the in vitro growth of four mesothelioma cell lines, MSTO-211H (epithelial), H2452 (epi- thelial), H28 (biphasic), and H2052 (biphasic). We used the cell titer Glo assays to assess the tumor cell lines viability treated with ponatinib or control vehicle for 72 h. We chose to perform these assays under the standard culture condition with 5% serum (fetal bovine serum) since ponatinib’s effect on tumor cells should be eval- uated under the standard, not a serum-free con- dition. Our result showed that increasing concentrations of ponatinib in the culture medium markedly inhibited the growth of all four mesothelioma lines, in a dose-dependent manner, as compared with control-treated cells (Figure 2a). The concentration of ponatinib required to inhibit 50% cell growth (IC50) of the mesothelioma lines was determined and ranged between 90 and 200 nM (Figure 2b). Ponatinib differentially inhibited the cell growth of all four mesothelioma lines tested with higher potency in MSTO-211H and H2452 compared to H2052 and H28. These results showed that ponatinib alone could inhibit the growth of mesothelioma cells. Notably, ponatinib’s inhibitory effect in all four mesothelioma cells tested appeared to be at pharmacologically relevant concentrations (50–200 nM) of the drug, as demonstrated in chronic myelogenous leukemia cells from patients.21 Interestingly, ponatinib was more effective in the epithelial subtype cell lines (MSTO-211H and H2452) than the biphasic sub- type cell lines (H2052 and H28). These results suggest that the sensitivity of MPM cells to pona- tinib is subtype dependent.
Ponatinib inhibits cell migration of MPM cells in vitro
The ability to migrate and invade surrounding tissues is the main hallmark of aggressive tumor cells. The ability of ponatinib to inhibit cell migration, which may be an essential process in tumor metastasis, was tested using in vitro scratch assay that assesses cell migration by the recovery of the scratch wound in two mesotheli- oma cell lines: MSTO-211H (most sensitive) and H28 (least sensitive) (Figure 3a). Twenty-four hours post-treatment, ponatinib effectively inhib- ited scratch wound recovery in both cell lines (P- value <0.05) (Figure 3b, c). Interestingly, 24 h of treatment with ponatinib significantly inhibited cell migration of H28 compared to MSTO-211H. In cell viability assay, H28 was the least sensitive cell line to ponatinib, and MSTO-211H was the most sensitive cell line. The difference observed between cell viability and cell migration assays might be due to ponatinib’s differential effect on various pathways associated with cell survival (cell viability assay) and cell motility (cell migra- tion assay) in either of these lines. Albeit, these results suggest a requirement for cAbl activity in mesothelioma cancer cell growth and motility. Figure 1. cABL transcript and protein expression in malignant pleural mesothelioma tumors. (a) DNA microarray analysis of cABL expression in normal lung normal pleura and malignant pleural mesothelioma. The data were mined (www.oncomine.org from the analysis Gordon et al. (Gordon, Am J Pathol, PNAS).20 The upper panel shows boxplots for the expression of cABL transcripts in normal lung samples (4 cases) and normal pleura samples (5 cases) vs. malignant pleural mesothelioma (40 cases). The horizontal lines within the boxes indicate median values. cABL transcripts were markedly upregulated in both malignant pleural mesothelioma relative to normal lung control and normal pleura samples. (b) cAbl protein expression in MPM tumors: representative sections of MPM tumors with anti-cAbl antibody (B–G) or a class-matched (mIgG1) (A) followed by AP-conjugated secondary antibody and substrate. Panels A, B, D, F, and H are low-power micrographs (10 , scale bar 100lm). Panels C, E, G, and I are higher-power micrographs (40 , scale bar 20lm). MPM stained with anti-cAbl antibody demonstrates positive tumor cells staining for cAbl (Brown staining) surrounded by weakly staining desmoplastic stroma (Panels B–I). All tumor sections were counterstained with H&E to simultaneously visualize the cells’ morphology and where cAbl protein is detected. Figure 2. Ponatinib suppresses in vitro cell growth of MPM cell lines. Four MPM cell lines were seeded (1000 cells per well) to 96- well opaque bottom plates and cultured overnight. 200 ponatinib DMSO solution was diluted with 5% FBS media and added to cells. Cell growth was monitored for 72 h post-drug treatment using Promega Cell Titer Glo reagent. (a) A representative normal- ized growth curve of four mesothelioma cell lines, 72 h post-ponatinib treatment. Growth curves were normalized to day 0 as 0% growth. (b) Average of the normalized IC50 means from three independent experiments for each of MPM tested. cAbl/Arg signaling pathway is activated in MPM cells Ponatinib is an orally active multi-tyrosine kinase inhibitor (TKI) with potent activity against Abl-family kinases. Ponatinib, an FDA-approved drug for patients with leukemia, explicitly targets BCR- Abl oncogene with an IC50 of 0.37 nM.18 We hypothesized that the growth inhibition of meso- thelioma cell lines by ponatinib treatment might be due, at least in part, to the cAbl signaling pathway disruption. We first sought to determine whether the cAbl signaling pathway is activated in the mesothelioma cell lines to investigate this hypothesis. All four mesothelioma cell lines showed high protein levels of both cAbl1 and active cAbl1 kinase (pcAbl) phosphorylated at Y412 tyrosine residue in the activation loop (Figure 4a). Moreover, all four MPM lines exhib- ited significant Arg tyrosine kinase levels, an Abl- related gene known as Abl2.12,21 Abl2 was also expressed and phosphorylated, thus activated in all four assessed mesothelioma lines (Figure 4b). In keeping, all four lines displayed substantial levels of phosphorylated CrkL(pCrlk) (v-crk sar- coma virus CT10 oncogene homolog (avian-like), a cAbl kinase substrate of22,23 (Figure 4a). Also, important levels of phosphorylated AKT (pAkt) protein, an indirect cAbl-downstream effector24,25 (Figure 4a), whose activation depends on Abl/ PI3K signaling pathway, were detected in all MPM lines except H2452. However, the basal AKT phosphorylation level was deficient in H2452 (Figure 4a). Together, these results dem- onstrate that the cAbl signaling pathway is acti- vated in the MPM cells. Figure 3. Ponatinib inhibits MPM cell migration. MSTO-211H and H28 cells were seeded to 12-well plates and grew until 95–100% confluent. Scratches were made using a 20 ll tip, and the media were changed immediately to either DMSO control or ponatinib (0.5 lM) containing media. Wound gap size was assessed both at time 0 h and 24 h post-drug treatment. The gap dis- tance was measured using Image J. (a) Representative images of cells at 0 h and 24 h. (b) Mean of the change in gap distance 24 h post-DMSO or ponatinib (0.5 lM) treatment. (c) Percentage recovery of MSTO-211H and H28 lines 24 h post-treatment with either DMSO control or ponatinib. The values shown in the graphs are means ± SDs. (ω indicates P < 0.05 relative to controls). Ponatinib treatment inhibits cAbl signaling in MPM cell lines Next, we examined ponatinib’s effect on cAbl sig- naling in the MPM cell lines, MSTO-211H, H28, H2052, and H2452. Cell lysates were prepared from ponatinib- or control-treated cells and sub- jected to Western blot analysis. We found that 2 h of treatment with ponatinib induced a dose- dependent inhibition of cAbl (Figure 5). The pcAbl (Tyr412) level was reduced significantly and dose-dependent by ponatinib treatment (Figure 5). In agreement, the activation of Crkl was dramatically inhibited by ponatinib. In the ponatinib treated cells, the phosphorylation of CrkL at tyrosine 207 was reduced below the detectable level (Figure 5). Ponatinib-induced inhibition of the cAbl signaling was also con- firmed by the reduced phosphorylated Akt levels in the mesothelioma lines (Figure 5). These results corroborate that the ponatinib effect on MPM cell lines growth is associated with its inhibition of the cAbl kinase signaling pathway. Ponatinib treatment increases the double-stranded DNA damage in MPM cells Cells are constantly exposed to a number of DNA-damaging factors, including endogenous and environmental factors. Thus, cells regularly rely for their survival on the DNA repair path- ways. DNA repair mechanisms play a central role in carcinogenesis and are potential therapeutic targets.26 A growing body of data implicates DNA repair mechanisms in MPM pathogenesis. Several studies have suggested that cAbl is required to properly activate critical elements in the double strand DNA breaks (DSBs) repair pathways, including activation of ATM/ATR and Rad51. Thus, we hypothesized that ponatinib- induced reduction in cell growth might be associ- ated with an amplification of DNA DSBs in MPM cells due to cAbl inhibition. To address this question, we have performed c-H2AX ana- lysis, a universal pharmacodynamic surrogate marker for DSBs. MSTO-211H cells were treated with ponatinib for 14 h, and a commercially available ELISA-based assay was used to analyze the cH2AX levels in the cells. Figure 4. cAbl/Arg signaling pathway is activated in MPM cell lines. Basal expression levels (total and active) of cAbl/Arg sig- naling pathway elements in four mesothelioma cell line (1: MSTO-211H, 2: H28, 3: H2052, 4: H2452). (a) Immunoblotting of pcAbl, cAbl, pAKT, AKT, pCrkl, proteins, and GAPDH protein as a loading control in MPM cell lysates. (b) Arg and pArg expression levels determined by immunoprecipitation (IP) of Arg protein from MPM cell lysates followed by Western blot (WB) detection with anti-Arg antibody or anti-phosphotyrosine antibody (Anti-pTyr), respectively. Our analysis showed that ponatinib treatment resulted in a marked and dose-dependent increase in DSBs damage, as evidenced by an increase in the levels of gamma-H2AX (cH2AX) in ponatinib-treated as compared to vehicle- treated MPM cells (Figure 6). Ponatinib treatment alters the activity of important receptor tyrosine kinases in MPM cells Owing to ponatinib’s multi-targeted properties,16 and its inhibitory activity against various essential tyrosine kinases including PDGFR, FGFR, RET, SRC, FLT1, and KIT,17,18 we have assessed the potential that ponatinib may inhibit other tyro- sine kinases in mesothelioma cells. Using com- mercially available human phospho-receptor tyrosine kinase (RTK) array, we evaluated the effect of ponatinib treatment on 49 RTKs in two MPM cell lines, MSTO-211H and H28, the most and the least ponatinib-sensitive cell lines, respectively, in the cell viability assay (Figure 2). Cells were starved overnight and treated for 2 h with either medium control or medium contain- ing 5 lM ponatinib, and then cell lysates were prepared and used for the array screening. Our results revealed differential basal levels of activa- tion in various RTKs between the two assessed MPM lines (Figure 7a). Two hours of treatment of MSTO-211H and H28 lines with 5 lM ponati- nib induced a significant inhibition of EGFR and HGFR activation in both cell lines (Figure 7a, b). ErB4 and IGF-1R activation were also markedly reduced in the H28 cell line (Figure 7a, b). In contrast, in H28 cells, ponatinib induced a mild increase in Axl kinase activity, a known driver of diverse cellular processes, including growth, pro- liferation, and migration (Figure 7a, b). Thus, the increased Axl kinase activity may contribute to the weaker sensitivity of H28 cells to ponatinib. A previous report by Marek et al.27 showed that FGFR1 is expressed and activated in MSTO- 211H but not in H28, and the expression level of FGF1 is associated with the sensitivity of the MSTO-211H cell line to ponatinib. In the present study, we did not observe the same expression pattern in MSTO-211H using either Western blot (Figure 7c) or the RTK array (Figure 7a) to ana- lyze the basal expression levels of FGFR1. We also did not observe an alteration in FGFR1 acti- vation in response to ponatinib treatment using the RTK array assay (Figure 7a). The difference could be due to the assay condition or the experiment design. Figure 5. Ponatinib inhibits cAbl signaling in MPM cell lines. MSTO-211H, H2052, H28, and H2452 cell lines were treated with DMSO control or ponatinib for 2 h. Cell lysates were prepared from treated cells, and 10 lg of proteins were subjected to Western blot analysis for pcAbl, cAbl, pAKT, AKT, pCrkl, Arg proteins, and GAPDH protein as a loading control. Ponatinib resulted in a dose- dependent inhibition of pAbl, pCrkl, and pAkt signals, as well as Arg protein expression in MPM cell lines. Figure 6. Ponatinib increases the cellular level of double- stranded DNA break marker gamma-H2AX (rH2AX) in MSTO- 211H cells. MSTO-211H cells were treated for 14 h with either medium alone control or medium containing DMSO control (DMSO) or medium containing increasing ponatinib concentra- tions. Cell lysates were prepared from treated cells, and rH2AX levels in the lysates were measured using Trevigen gamma H2AX pharmacodynamic assay. The results were normalized to 107 cells. Increased rH2AX levels in the treated MSTO-211H cells correlated with an increase in ponatinib concentration. Ponatinib inhibits in vivo tumor growth and cAbl signaling pathway in the MSTO-211H-derived xenograft model The effect of ponatinib on MPM cell growth was further analyzed in vivo. MSTO-211H xenografts tumors were established by inoculation of female NSG immunodeficient mice with 107 MSTO- 211H cells subcutaneously. When the tumors reached approximately 450 mm3, mice were randomized to two treatment groups (5 animals per group): (1) vehicle control (25 mmol/L citrate buffer, pH 2.75), oral; (2) 30 mg/kg PO BID ponatinib on 5-day ON/2-day OFF cycles, oral. All animals were treated over a 21-day dosing period. Tumor size for each animal was measured twice a week (Figure 8a). Tumors were harvested at the end of the treatment period. Figure 7. Ponatinib regulates the activity of important RTKs in MPM cells. MSTO-211H and H28 were starved and treated with either DMSO control or 5 lM ponatinib for 2 h. Cell lysates were prepared and analyzed using the Proteome Profiler Human Phospho-RTK Array Kit (R&D system) to assess the phosphorylation status of 42 different RTKs. (a) The RTK array blots. The intensity of each RTK phosphorylation was quantified using ImageJ, averaged, and normalized to that of the internal reference dots (Ctrl). (b) Fold decrease vs. DMSO control in the RTKs phosphorylation levels with ponatinib treatment were calculated from the array blots. EGFR and HGFR (cMet) phosphorylation levels and their activation were reduced in MSTO-211H cells treated with ponatinib for 2 h. EGFR, ErbB4, IGF-1R, and HGFR activation were reduced in the ponatinib-treated H28 cell line. Axl activation was slightly increased in pona- tinib-treated H28 cells. (c) The basal expression levels of FGFR1 and pFGFR1 in four mesothelioma cell lines. 1: MSTO-211H, 2: H28, 3: H2052, 4: H2452 were assessed by western blot assay. GAPDH protein expression level was used as a loading control. Ponatinib treatment exhibited a clear and significant tumor growth inhibition compared to the vehicle-treated group (Figure 8a). Also, ponatinib treatment’s pharmacodynamic param- eters were determined using lysate prepared with xenograft tumor specimens from each group in this study. Ponatinib treatment at 30 mg/kg inhibited dramatically cAbl phosphor- ylation level, and thus, its activation in tumors derived from mice treated with ponatinib compared tumors derived from vehicle-treated mice (Figure 8b). Additionally, pCrkl and pSTAT5 levels, two downstream targets of the cAbl signaling pathway, were significantly reduced in ponatinib-treated tumors compared to vehicle-treated tumors (Figure 8b, c). Together, these results indicate that in vivo, ponatinib treatment inhibits the cAbl signaling pathway and blunts the growth of MSTO- 211H-derived tumors. Figure 8. Ponatinib and inhibits the cAbl signaling pathway and blunts in vivo tumor growth of MSTO-211H-derived xenograft. NSG immunodeficient mice were injected subcutaneously with MSTO-211H cells. The tumors were allowed to grow to a palpable size, at which point mice were treated with 30 mg/kg ponatinib or the diluent control (Control) orally once per day for 21 days. Tumor volumes were measured every 2–3 days at the indicated times. Means ± SEMs based on five determinations are shown. a) Administration of ponatinib to NSG mice markedly inhibited the growth of MSTO-211H as xenograft tumors as compared to the diluent control-treated mice. (b, c) Tumor lysates were prepared from tumors derived from either ponatinib-treated or control- treated mice and assessed by western blot analysis for cAbl pathway activation. (b) pCrkl and pSTAT5 levels, two downstream tar- gets of the cAbl signaling pathway, were significantly reduced in ponatinib-treated tumors than vehicle-treated tumors. (c) The intensity of pCrkl and pSTAT5 protein levels were quantified using ImageJ, averaged, and normalized to the GAPDH level. The values shown in the graphs are means ± SDs. (ωindicates P < 0.05 relative to controls). Ponatinib and cisplatin combination therapy showed a synergistic effect in MPM cells growth in vitro Next, we evaluated ponatinib’s potential syner- gistic effect on in vitro MSTO-211H cell viabil- ity, combined with some selected clinically approved anticancer therapies to include DNA damaging chemotherapies gemcitabine, cisplatin, and pemetrexed. Cells were treated with a serial dilution of pemetrexed, cisplatin, or gemcitabine alone or with a fixed concentration of ponatinib in 5% serum media for 72 h, and the cell viabil- ity was measured using Cell Titer Glo reagent. Our results showed that ponatinib sensitized MSTO-211H cells to cisplatin, resulting in a more potent anti-proliferative effect in the MSTO-211H cell line than either single agents alone (Figure 9a, b). The drug combination dose-effect analysis with isobolograms, shown in Figure 9b, depicts ponatinib and cisplatin’s opti- mal combination ratio that results in a synergis- tic or additive cytotoxic effect in the MSTO- 211H cell line. In contrast, a combination of ponatinib with either pemetrexed or gemcitabine did not demonstrate any significant synergistic or additive anti-growth effect in the MSTO- 211H cell line at the tested combination ratios in this study (not shown). Figure 9. Ponatinib increased the sensitization of MSTO-211H cells to cisplatin therapy. Effects of ponatinib on responses of MSTO-211H cells to cisplatin agent. (a) MSTO-211H were exposed to varying doses of ponatinib, or varying doses of cisplatin com- bined with or without a fixed dose of ponatinib as indicated. Cell growth was measured after 72 h by the cell titer glow assay. (b) Normalized isobolograms depict ponatinib and cisplatin’s optimal combination ratios that result in a synergistic cytotoxic effect of cisplatin in MSTO-211H cells. Discussion Malignant pleural mesothelioma is a rare but deadly tumor, with about 3300 new cases annually in the US. With a dismal survival rate of 7–9 months for untreated MPM and 5-year overall survival of about 5%, there is a pressing need to develop an effective anti-MPM treatment modalities. The pathogenesis of mesothelioma remains unclear, and the oncogenic pathways involved in the neoplastic transformation of mesothelial cells are mostly undefined. DNA repair systems play an essential role in carcinogenesis and are potential cancer tar- gets.26,28 A growing body of evidence implicates DNA repair mechanisms in MPM pathogenesis. Defective repair mechanisms can increase muta- genesis and act synergistically with asbestos exposure. Furthermore, increased expression and activity of DNA repair enzymes such as UNG2, PARP, cAbl, BRCA2, XRCC4, KU80, and others29 may protect mesothelial cells from asbes- tos-induced DNA damage but augments resist- ance to chemotherapy and radiotherapy of MPM tumor cells. Blocking the activated DNA repair enzymes, or extending DNA repair deficiency in tumors already presenting DNA repair defects, could provide novel targets for the treatment of MPM as a single therapy or in combination with current chemotherapy to include DNA damag- ing drugs. Previous studies have demonstrated that cAbl inhibition causes altered responses to DNA dam- age.30 cAbl is a non-receptor tyrosine kinase that plays a role in differentiation, adhesion, cell migra- tion, cell division, death, and cellular stress responses by binding to several proteins involved in apoptosis pathways.30 Following DNA damage, cAbl is activated and binds and activates several DNA damage repair proteins, including ATM, ATR, Rad51, BRCA1, DNA-PK, RFX1, and p73.23 In various cancers, cAbl stimulation is associ- ated with hyperactive receptor tyrosine kinases (RTKs) and chemokine receptors such as EGFR, HER2, PDGFR, cMet, IGFR, and CXCR4.Also, cAbl kinases can be stimulated by the silencing of its inhibitory protein and tumor sup- pressor protein FUS1,16 which is downregulated in major subsets of many cancers, including MPM.17 Through these mechanisms, cAbl acti- vates various intracellular signaling pathways and stimulates several processes involved in tumor growth and metastatic progression. Ponatinib (Iclusig, AP24534) is a clinically approved multi- tyrosine kinase inhibitor (TKI) with potent activ- ity against cAbl. In the present study, we show a dysregulated cAbl pathway in human MPM. Furthermore, we demonstrate that ponatinib treatment in 4 MPM cell lines reverted these cells’ malignant pheno- type in vitro and reduced MPM cell growth and cell migration (Figures 2 and 3). All MPM lines showed considerable pCrkl protein expression demonstrating a cAbl pathway activation (Figure 4). Additionally, we have found that the ponati- nib treatment of MPM cell lines results in a sig- nificant reduction in the phosphorylation and activation levels of cAbl and its downstream effectors CrkL, ATK, and STAT5 (Figures 5 and 8). Also, ponatinib treatment amplified double- strand DNA breaks (DSDBs) revealed by an increased accumulation of c-H2AX in MPM cells (Figure 6). In vivo, ponatinib reduced tumor growth (Figure 8). Moreover, ponatinib decreased pCrkl expres- sion in tumors (Figure 8b). Together, these results suggest ponatinib’s potential utility in treating Mesothelioma patients through inhibition of cAbl kinase activity. These results also high- light pCrkl tumor levels’ potential as a biomarker in determining the eligibility of mesothelioma patients for ponatinib-based therapy. DNA repair enzymes are critical for maintain- ing the genome,32 and defects in the repair pro- cess lead to genome instability and promote tumorigenesis.33,34 Conversely, good DNA repair provides resistance to therapeutic radiation and some cytotoxic chemotherapy.26 Moreover, com- bined impairment in different DNA repair sys- tems results in gross genomic instability and death of tumor cells, a principle called “synthetic lethality.”35,36 Radiation therapy, as well as the majority of chemotherapy, weaken cancer cells by damaging DNA. Successful DNA repair is essen- tial for the normal cells to surmount the therapy’s adverse effects, but in the tumor can result in treatment resistance. cAbl inhibition has been recently shown to alter DNA damage response.30 DNA damage elicits cAbl activation, which then binds and activates DNA damage repair proteins, including ATM, ATR, Rad51, BRCA1, DNA-PK, RFX1, and p73.23 Moreover, activated cAbl increases the Rad51 gene. Our present study shows that the MSTO-211H cell line treatment with various combination ratios of ponatinib and cisplatin results in syner- gistic cytotoxicity with some of the combination ratios tested. This result strongly supports that ponatinib’s inhibition of cAbl activity may sensi- tize MPM cells to DNA damages induced by cisplatin. The genetic background of cancer cells is a critical factor determining cells’ susceptibility to drugs.37 Our results show that ponatinib is more effective in the epithelial subtype cell lines (MSTO-211H and H2452) than the biphasic sub- type cell lines (H2052 and H28), thus suggesting that the sensitivity of MPM cells to ponatinib is subtype dependent. Further investigations of the MPM cells’ genetic background role in their ponatinib-sensitivity may help determine which MPM patients are more likely to benefit from ponatinib therapy, either as a single agent or in combination with chemotherapy. High expression and activation of EGFR have been reported in MPM patients.38,39 Previous investigations also showed that cAbl phosphory- lates EGFR and reduces the endocytosis of EGFR in cells required for the degradation of EGFR.40 Thus, active cAbl may both stimulate and pro- long the activation of EGFR. The findings men- tioned above suggest that crosstalk between these two molecules may contribute to human cancer. Other reports also demonstrated that cAbl activ- ity is required for survival and tumorigenesis through HGFR/c-Met signaling.41 Furthermore, both cAbl and Arg (Abl2) can bind and phos- phorylate the ErbB4 tyrosine kinase receptor. Our RTK array result showed that the activa- tion of both EGFR and HGFR/cMet was down- regulated in both MSTO-211H and H28 cells in response to ponatinib (Figure 7), which is con- sistent with a decrease in pAbl level in these cells when treated with ponatinib (Figure 5). Moreover, H28 showed a lower expression of phospho-ErbB4. However, it is not clear whether these aforementioned events triggered by ponati- nib treatment were sequential or concomitant, whether they were primary or secondary effects and the impact of the downregulation of these phosphorylated RTKs on the ponatinib-induced cytotoxicity in MPM cells. Further investigations are needed to answer these questions. Interestingly, ponatinib slightly upregulated Axl activation in the H28 line. Axl is an oncogene reg- ulating cell proliferation, migration, angiogenesis, and immune modulation.43 Axl is often expressed and activated in mesothelioma.44 Upon Axl phos-phorylation/activation, downstream cell survival pathways such as AKT, NF-jB, and MAPK signal- ing pathways are activated.43,45–48 Thus, Axl activa- tion in the H28 cell line might explain its relative weaker sensitivity to ponatinib than the other tested MPM line in this study. Per the finding mentioned above, Axl activation has been sug- gested as a driver of resistance to chemotherapy, targeted therapy, and immunotherapy in solid tumors.43 Further investigations are required to determine the role of Axl expression and activa- tion in the MPM cells’ sensitivity to ponatinib. Conclusion Collectively our findings provide a rationale for a novel translational therapeutic approach to MPM.Our results suggest a potential utility for ponati- nib in MPM patients through inhibition of cAbl kinase. Furthermore, finding a role for ponatinib in MPM enhanced sensitivity to anticancer thera- pies would provide an impetus for integrating this targeted therapy into treatment regimens with approved chemotherapies for patients with mesothelioma. Materials and methods Cell lines and reagent Mesothelioma cell line MSTO-211H, H28, H2052, and H2452 were culture in RPMI1640 media supplied with 10% fetal bovine serum (Gemini Bio-Products, West Sacramento, CA). Ponatinib (SelleckChem, Houston, TX) stock solution was prepared in DMSO as 20 mM and stored at 80 ◦C. PathScanVRBcr/Abl Activity Assay (1:250; Cell Signaling, Danvers, MA) and anti-Arg antibody (1:400; Santa Cruz, Dallas, TX) were diluted with 5% BSA in Tris-buffered saline supplied with 0.1% Tween-20 (TBST). Anti-cAbl (ab15130) (1:1000; Abcam, Cambridge, MA), anti-cAbl (Tyr-412), phospho-specific Rabbit Polyclonal (Tyr412) (1:1000; ECM Biosciences, Versailles, KY), Anti-FGFR1 antibody (ab10646) (1:1000, Abcam), Anti-FGFR1 (phospho Y653 Y654) antibody (ab111124) (1:1000, Abcam), Akt (pan) (11E7) Rabbit mAb (1:1000, Cell Signaling Technology), phospho-Akt (Ser473) (D9E) XPVR Rabbit mAb #4060 (1:2000, Cell Signaling Technology), phospho-CrkL (pTyr207), rabbit polyclonal (PA517741) (1:1000; Thermo, Waltham, MA) antibodies were diluted with 5% nonfat milk in TBST. Horseradish per- oxidase-conjugated anti-phospho-tyrosine anti- body (1:5000; Proteome Profiler Human Phospho-RTK Array Kit, R&D system, Minneapolis, MN) were diluted with 1 Array Buffer 2 (Proteome Profiler Human Phospho- RTK Array Kit, R&D system). cAbl expression in published gene microarray To investigate cAbl expression in published gene microarray studies, we used the oncoming data- base (www.oncomine.org). Immunohistochemistry Under UCSF CHR approval, archived paraffin- embedded tissue blocks were obtained for ten cases of MPM. 5 lM sections of paraffin- embedded human tumor tissue sections were de-paraffined in xylenes for 3 min, followed by rehydration in 100%, 95%, 75%, and 50% etha- nol for 3 min, twice. After one phosphate-buf- fered saline (PBS) wash, sections were incubated in Antigen Retrieval Citra Solution (BioGenex, Fremont, CA) at 92 ◦C for 8 min, followed by 30 min cooling down on a bench. The endogenous peroxidase activity was inhibited by incu- bating sections in 3% hydrogen peroxide (Fisher, Waltham, MA) for 10 min. After three 3-min PBS washes, sections were blocked with serum-free protein blocking agent (Dako North America, Inc., Carpinteria, CA) or 5% normal goat serum in Tris-buffered saline supplied with 0.1% Tween 20 (TBST) for 30 min. Primary antibodies were diluted with antibody diluent with reducing background (Dako) or 5% normal goat serum in TBST. Tissue sections were incu- bated with primary antibodies at 4 ◦C overnight, followed by three 5-min PBS washes. Two drops of Histostain-Plus Kit (DAB, Broad Spectrum) buffer B (ThermoFisher, Waltham, MA) were applied to tissues and incubated for 10 min. Tissues were washed in PBS for 2 min, three times, and two drops of Histostain-Plus buffer C were applied to tissue sections. After a 10- min incubation, tissues were washed in PBS for 2 min, three times. DAB substrate was prepared according to the manufacturer’s instruction. 100 ll of DAB was applied to the tissue section and incubated for 3 min. The slides were then washed with water and subsequently, counter- stained with hematoxylin. After water wash to remove remaining hematoxyline, sections were dehydrated by incubation in PBS (once), 50%, 75%, 95%, 100%, and xylene (twice in each solution), 3 min each. One drop of DPX mount- ing medium (Sigma, St. Louis, MO) was then applied to each tissue section and covered with a cover glass. Images were taken from three different fields in each section using Zeiss Axioimager2 and software (Zeiss, Oberkochen, Germany). Cell viability assay One thousand cells were plated to wells in Coster 3917 solid white flat bottom polystyrene TC- treated microplates (Corning, Corning, NY) and incubated in a 37 ◦C humidified incubator sup- plied with 5% CO2 overnight. The next day, a serial dilution of 200 drug solution was prepared freshly with DMSO and diluted with RPMI 1640 media supplied with 5% fetal bovine serum to make drug media. The culture media was replaced with drug media, and cells were incubated for 72 h. Cell viability was measured using the Cell Titer-Glo Luminescent Cell Viability Assay (Promega, Madison, WI) as the manufacturer’s instruction. Data were processed and analyzed using Graphpad Prism 6 (Graphpad, San Diego, CA). The mean from triplicate wells was calcu- lated and normalized to untreated cells at 0 h as 0% and DMSO-treated cells at 72 h as 100%. Wound healing assay MSTO-211H and H28 cells were seeded to 24- well plates and cultured in a 37 ◦C humidified incubator supplied with 5% CO2 overnight or until reaching 95–100% confluent. Wounds were generated using a 20 ll tip scrapping crossing wells vertically and horizontally. After wounds were made, the media were removed, and drug media (0.5 lM ponatinib in RPMI 1640/5% fetal bovine serum) or control media (DMSO only in RPMI 1640/5% fetal bovine serum) were added to wells immediately. Three wound images from different spots were taken from each well right after the media were changed 24 h after the drug treatment. The width of wounds was measured using ImageJ (NIH, Bethesda, MD) and averaged from three independent experiments. Western blot Cells grown on 6-well plates or 60-mm dishes were starved (RPMI 1640/0.5% fetal bovine serum) for 16–24 h before adding drug media (RPMI 1640/1% fetal bovine serum). Cells were treated with the drug for 2 h, and cell lysate was prepared using M-PER Mammalian Protein Extraction Reagent (Thermo, Waltham, MA) supplied with 1× Halt protease inhibitor cocktail (Thermo, Waltham, MA) and sodium orthovanadate. Protein concentration was measured using Pierce BCA Protein Assay Kit (Thermo, Waltham, MA), and 10 lg of cell lysate was loaded to wells of 4- 20% TGX gels (Bio-Rad, Hercules, CA). Proteins were separated using gel electrophoresis and trans- ferred to PVDF membranes (Millipore, Billerica, MA). Membranes were blocked with 5% nonfat milk/TBST buffer at room temperature for 1 h and incubated with the primary antibody in blocking buffer at 4 ◦C overnight. After three TBST washes, membranes were incubated with horseradish peroxidase-conjugated goat anti-mouse, donkey anti- rabbit or bovine anti-goat secondary antibody at 1:5000 in blocking buffer (Jackson Immuno Research, West Grove, PA) at room temperature for 1 h. Membranes were washed with TBST three times and incubated with Thermo Scientific Super Signal West Pico Chemiluminescent Substrate (Thermo, Waltham, MA). The chemiluminescent signals were exposed to X-ray films and analyzed using ImageJ. Immunoprecipitation 10 or 50 lg of cell lysate was pre-cleared with protein G beads (GE Healthcare, Little Chalfont, UK) in M-PER Mammalian Protein Extraction Reagent supplied with 1 Halt protease inhibitor and sodium orthovanadate rocking at 4 ◦C for 1 h. After centrifuging at 14,000 g at 4 ◦C for 1 min, the supernatant was transferred to clean tubes and incubated with 1 lg of anti-Arg anti- body at 4 ◦C overnight with gentle rocking. The next day, protein G beads were added to tubes and rocking at 4 ◦C for 3–4 h to capture antibod- ies, followed by centrifugation at 14,000 g at 4 ◦C for 1 min and washed with ice-cold calcium and magnesium-free phosphate-buffered saline three times. Proteins were eluted from beads using Laemmli Sample Buffer (Bio-Rad, Hercules, CA) supplied with 50 mM DTT and boiled at 95–100 ◦C for 5 min. Beads were pelleted at 14,000 g at 4 ◦C for 1 min, and soluble proteins in the supernatant were analyzed using Western blot assays. The membrane was blocked with Array Buffer 1 (Proteome Profiler Human Phospho-RTK Array Kit, R&D system) at 4 ◦C overnight, and phospho-Arg levels were analyzed horseradish peroxidase-conjugated anti-phospho- tyrosine antibody (room temperature, 1 h). The membrane was washed with a 1 washing buffer (Proteome Profiler Human Phospho-RTK Array Kit) 10 min three times, followed by incubating with Thermo Scientific SuperSignal West Femto Chemiluminescent Substrate (Thermo, Waltham, MA). The chemiluminescent signals were exposed to X-ray films. Animal study The UCSF Preclinical Therapeutics Core con- ducted the animal study. In brief, 107 MSTO- 211H cells (HBSS:Matrigel 1:1) were injected subcutaneously per mice (NOD/SCID (NSG), female). When tumors reached a volume of 450 mm3, the mice were randomized to receive either (1) ponatinib in aqueous 25 mmol/L citrate buffer (pH 2.75): 30 mg/kg gavage (5 mice), or (2) 25 mmol/L citrate buffer (pH 2.75) (control, five mice) by gavage. The mice were treated once per day for 21 days, and the tumor sizes were measured every 2 or 3 days. After the final dosing and measurement, the mice were sacrificed, and tumors were dissected out. Half of the tissues were fixed in paraformaldehyde for immunohis- tochemistry, and the other half of the tissues were frozen at —80 ◦C for protein extraction. Protein extraction from frozen mouse tissue A small piece of frozen tissue was cut and ground to a fine powder. The protein was extracted by incubating the tissue powder in M-PER protein extraction reagent (Thermo Fisher, Waltham, MA) supplied with 1 Halt protease inhibitor cocktail and sodium orthovanadate at 4 ◦C for 1 h with gentle rocking, followed by centrifugation at 14,000 g at 4 ◦C for 15 min. The supernatant was collected and stored at 80 ◦C. The Pierce BCA Protein Assay kit measured the protein concentra- tion, and the lysate was subjected to Western blot analysis as described above. RTK profiling For basal expression, cells grown on 60-mm dishes were starved (RPMI 1640/0.5% fetal bovine serum) for 16–24 h, followed by changing to complete media (RPMI 1640/10% fetal bovine serum) and cultured for 24 h before preparing cell lysate. For expression levels in response to drug treatment, cells grown on 6-well plates or 60-mm dishes were starved (RPMI 1640/0.5% fetal bovine serum) for 16–24 h before adding drug media (DMSO control or 5 lM ponatinib in RPMI1640/1% fetal bovine serum). Cells were treated with the drug for 2 h. The cell lysate was prepared using lysis buffer from the Proteome Profiler Human Phospho-RTK Array Kit (R&D system) supplied with a 1 Halt protease inhibi- tor cocktail and sodium orthovanadate. Protein concentration was measured using the Pierce BCA Protein Assay kit, and 300 lg of protein was used for RTK assay according to manufacture instruction. The expression levels were quantified using ImageJ and normalized to the internal con- trols on the same blots. Combination therapy One thousand cells were plated to wells on Coster 3917 solid white flat bottom polystyrene TC-treated microplates and incubated in a 37 ◦C incubator supplied with 5% CO2 overnight. Cells were treated with cisplatin (100, 10, 1, 0.1 and 0.01 lM), gemcitabine (0.1, 0.02, 0.04, 0.008 and 0.0016 lM) and pemetrexed (10, 1, 0.1, 0.01 and 0.001) alone or combining with four different concentrations of ponatinib (0.01, 0.05, 0.1 and 0.2 lM) for 72 h in RPMI1640 media supplied with 5% FBS. Cell viability was measured using Cell Titer-Glo Luminescent Cell Viability Assay (Promega, Madison, WI) according to manufac- ture instruction. Data were processed and ana- lyzed using Graphpad Prism 6 (Graphpad, San Diego, CA). The mean from triplicate wells was calculated and normalized to untreated cells at 0 h as 0% and DMSO-treated cells at 72 h as 100%. The combination index was calculated using CompuSyn software (ComboSyn, Inc., Paramus, NJ).
Gamma-H2AX assay
MSTO-211H cells were seeded to a 6-well plate and culture in complete media overnight. A serial dilution of 200× drug solution was prepared freshly with DMSO and diluted 1 to 200 with RPMI 1640 media supplied with 5% fetal bovine serum. Cells were cultured in drug media for 14 h, and the cells were harvested using trypsin. After washed with ice-cold DPBS twice, cell num- bers were counted. 107 cells were lysed in 1 ml of the lysis buffer from the Trevigen gamma H2AX Pharmacodynamic assay kit (Trevigen, Gaithersburg, MD). According to the man- ufacturer’s instruction, the lysate was prepared, diluted 1 to 10 with assay buffer, and the rH2AX levels were measured using the Trevigen gamma H2AX Pharmacodynamic assay kit.