BAG2, MAD2L1, and MDK are cancer-driver genes and candidate targets for novel therapies in malignant pleural mesothelioma

Abstract

Malignant pleural mesothelioma (MPM) is an aggressive cancer with a poor prognosis and the identification of novel druggable targets is urgently needed. In previous work, we identified 15 deregulated genes highly expressed in MPM tissues and correlated with a poor prognosis. Here, we validated these findings on an independent dataset of 211 MPM patients (EGA, EGAD00001001915) and on a panel of MPM cell lines. Furthermore, we carried out in vitro gene silencing followed by proliferation, cytotoxicity, caspase, and migration assays to define whether these targets could be cancer-driver genes. We ended up with three novel candidates (i.e., BAG2, MAD2L1, and MDK), whose encoded proteins could be exploited as druggable targets. Moreover, of novelty, immunohistochemistry analysis on tissues revealed that the overexpression of BAG2 and MAD2L1 could differentiate MPM from RMP patients. Furthermore, when we tested Neratinib (an inhibitor of MAD2L1) and iMDK (an inhibitor of MDK) we found that they are effective on MPM cells, in part phenocopying the effects of MAD2L1 and MDK gene silencing. In summary, in the present work, we report that BAG2, MAD2L1, and MDK are bona fide cancer-driver genes for MPM worth of further studies.

Fig. 1. Box plots of the patients stratified by histology.

A–O In green the epithelioid, in orange the biphasic, and in purple the sarcomatoid MPM cases. p value = 0.0332(*), 0.0021(**), 0.0002(***).

Fig. 2. KM only on the epithelioid MPM cases.

In red the high-expression group, in blue the low-expression group. The time is expressed in days. Log-rank test was used to calculate the significant difference between the two curves.

Fig. 3. A representative western blot of the siRNA transfection experiment.

A The blot of siRNA BAG2 and MDK, C the blot of siRNA MAD2L1. β-Tubulin was used as a loading control and GAPDH was the positive control of the siRNA transfection. As shown in the blots and also in graphs (B) and (D) there was a significant reduction of the protein level in the silenced cells, as compared to the negative CTRL. E Graph showing the comparison of protein expression in MeT-5A and MSTO cell lines of BAG2, MAD2L1, and MDK (Mann–Whitney test; p value < 0.05 were considered significant, p value = 0.0332(*), 0.0021(**), 0.0002(***), <0.0001(****)).

Fig. 4. Viability, proliferation and caspase assays after gene silencing.

A AOC of MeT-5A and MSTO at 72 h after treatment with siRNA-negative-control (siRNA CTRL−), siRNA BAG2, siRNA MAD2L1, siRNA MDK. The AOC is normalized to the time of transfection (t₀). B Overall, FI of MeT-5A and MSTO was measured at 72 h after silencing; the graphs show the comparison between MeT-5A and MSTO for the three-siRNA transfection. ANOVA test; p value = 0.0332(*), 0.0021(**), 0.0002(***), <0.0001(****). C–E End point of apoptosis analysis in MSTO and MeT-5A cells in the presence of 5 µM caspase 3/7 Green Dye and siRNA BAG2, siRNA MAD2L1, and siRNA MDK, respectively. Mann–Whitney test; p value 0.0332(*), 0.0021(**), 0.0002(***), <0.0001(****).

Fig. 5. Quantitative treatment efficacy of Neratinib and iMDK.

A Fluorescence intensity (%) of the cell lines 72 h after the treatment with Neratinib at increasing doses (10, 20, 40 µM). B Fluorescence intensity (%) of the cell lines 72 h after the iMDK treatment at increasing doses (7.5, 15, 50 µM). Red object count for Neratinib (C) and iMDK (D) treatment showing cytotoxicity. Green object count for Neratinib (E) and iMDK (F) treatment. These overview bar graphs indicate that overall, the expression of caspase 3/7 has not increased significantly when compared to relevant controls. The vehicle control (DMSO) is indicated in purple. MSTO and MeT-5A treated cells are indicated in blue and green. ANOVA test; p value 0.0332(*), 0.0021(**), 0.0002(***), <0.0001(****).

Fig. 6. Time course analysis of the wound closure.

A, B MeT-5A and C, D MSTO cells were treated just with DMSO (in blue), or with Neratinib (10 or 20 µM, in yellow). E, F MeT-5A and G, H MSTO cells were treated just with DMSO (in blue), or with iMDK (7.5 or 50 µM, in yellow). Mann–Whitney test; p value 0.0332(*), 0.0021(**), 0.0002(***), <0.0001(****).

Fig. 7. Representative immunohistochemical staining of MPM samples showing intense expression of MDK, MAD2L1, and BAG2 proteins in tumor cells (brown staining).

A–D Representative BAG2 immunostainings in RMP, EMM, BMM, and SMM, respectively. Moderate to strong expression in the three types of MPM, while no expression is visible in the hyperplastic mesothelium in RMP (red arrow). E–H Representative MAD2L1 immunostainings in RMP, EMM, BMM, and SMM, respectively. Strong expression is present in the three types of MPM, but there are areas of discernable expression in the mesothelial cells on the pleural surface in RMP as well (blue arrow positive mesothelial area, red arrow negative area). I–L Representative MDK immunostainings in RMP, EMM, BMM, and SMM, respectively. Strong expression is seen in the three types of MPM; however, some focal expression is also detectable in the mesothelial cells on the pleural surface in RMP (blue arrow focal staining, red arrow negative staining). Magnification: RMPs ×400, MPMs ×200.

Fig. 8. Quantification of protein expression in RMP vs MPM as assessed by IHC and H-score.

Graphs for BAG2 immunostaining in RMP vs MPM regardless of subtype (A) and RMP vs the different subtypes of MPM (B), respectively. Graphs for MAD2L1 immunostaining in RMP vs MPM regardless of subtype (C) and RMP vs the different subtypes of MPM (D), respectively. Graphs for MDK immunostaining in RMP vs MPM regardless of subtype (E) and RMP vs the different subtypes of MPM (F), respectively. Kruskal–Wallis test followed by a post hoc Dunn’s test with Bonferroni correction was used to evaluate the statistical difference between the median of each group, p value <0.001(***), <0.01(**), <0.05(*).