GSDMB promotes bladder cancer progression in vitro and in vivo
First, we explored the clinical biological effects of GSDMB in bladder cancer. It is worth noting that the mRNA expression level of GSDMB in bladder cancer tissues was higher than that in adjacent normal tissues according to the bioinformatics analysis of the TCGA data set (n =19, P = 1.98e-2; Fig. 1A–C). To further study the biological role of GSDMB in bladder cancer, GSDMB was knocked down by two different short harpin RNAs (shRNAs) in both T24 and 5637 cell lines (Fig. 1D). In vitro cell proliferation assays showed that the growth activity of bladder cancer cells was significantly inhibited after GSDMB was knocked down (Fig. 1E–G). We also found that the apoptotic rate of bladder cancer cells in the GSDMB silencing group was significantly higher than that in the control group in both 5637 and T24 cancer cells, as analyzed by the Annexin V-APC assay (Fig. 1H). The cell cycle analysis also showed that the number of bladder cancer cells in the S phase decreased when GSDMB was knocked down (Supplementary Fig. 1A), indicating slow cell division and proliferation. In addition to the in vitro experiments, we also employed the nude mice xenograft tumor model after inhibition of the expression of GSDMB to study the growth-promoting effect of GSDMB in bladder cancer in vivo. The results revealed significantly reduced tumor volume and mass and even no tumor after knocking down GSDMB, which was consistent with the results of the in vitro experiments (Fig. 1I–K). Furthermore, the invasion and migration ability of the cells was also significantly weakened after knocking down GSDMB (Supplementary Fig. 1B). Concluding, GSDMB promotes bladder cancer progression in cells and in mice.
GSDMB regulates the glycolysis of bladder cancer cells
Given the critical role of GSDMB in bladder cancer, we investigated the mechanism by which GSDMB promotes the progression of bladder cancer. We conducted RNA-Seq analysis after using small interfering RNA (siRNA) to inhibit the expression of GSDMB (Fig. 2A, B). Then, we performed GO and KEGG enrichment analysis, which indicated that GSDMB had a significant influence on glycolysis and gluconeogenesis (Fig. 2C, D). To further confirm this mechanism, we knocked down or overexpressed GSDMB in T24 cells (Fig. 2E, I, respectively). The mRNA expression levels of the genes HK2, LDHA, and ENO2, which are related to the glucose metabolism pathway, were markedly decreased when GSDMB was knocked down (Fig. 2F). Moreover, GSDMB knock down conveniently reduced glucose consumption and lactate production corresponding to the amount of GSDMB knock down (Fig. 2G, H). In contrast, overexpression of GSDMB had the inverse effect (Fig. 2J–L). Collectively, our data suggest that GSDMB can regulate the glycolysis of bladder cancer cells.
GSDMB regulates the glycolysis via activating STAT3 in bladder cancer cells
In the following, the mechanism underlying the regulation of the glycolysis of bladder cancer cells by GSDMB was studied. We first performed mass spectrometry to identify the potential binding partners of GSDMB (Fig. 3A and Table S1), revealing that GSDMB may bind to STAT3. This finding was confirmed by immunoprecipitation in both T24 and 5637 bladder cancer cells (Fig. 3B). Multiple studies showed that STAT3 exerts a significant effect on the glycolysis and glucose-dependent addiction of tumor cells, which is known as the Warburg effect 16, 17. Specifically, it has been reported that STAT3 regulates the glucose metabolism through increasing the HK2 expression in cancer cells 18. Furthermore, STAT3 up-regulates LDHA expression to promote the proliferation of urinary bladder cancer cells 19. Interestingly, GSDMB and STAT3 proteins could precipitate each other (Fig. 3A, B). Moreover, GSDMB modulates the glycolysis by up-regulating the expression of HK2 in bladder cancer cells (Fig. 2E–2L), suggesting that GSDMB may be involved in regulating the glucose metabolism of bladder cancer cells through STAT3-related signaling pathways. To further verify that GSDMB regulates the glucose metabolism via the STAT3 pathway, we suppressed the expression of STAT3 while GSDMB was knocked down or overexpressed. The results showed that the protein and mRNA levels of HK2 remained constant after knocking down STAT3 whether GSDMB was inhibited or overexpressed (Fig. 3C–F), indicating that STAT3 is an important intermediate mediator for GSDMB in regulating the glucose metabolism. Furthermore, it has been reported that Tyr 705 STAT3 phosphorylation leads to downstream glucose metabolism disorders 20, 21. Importantly, the protein level of p-STAT3 Tyr705 was down-regulated when GSDMB was knocked down (Fig. 3G). On the contrary, p-STAT3 Tyr705 increased as the overexpression level of GSDMB increased (Fig. 3H). However, the total STAT3 level did not change whether GSDMB was knocked down or overexpressed. Furthermore, we used cryptotanshinone to inhibit the phosphorylation of Tyr 705 STAT3 in bladder cancer cells with knocked down or overexpressed GSDMB, resulting in constant protein and mRNA levels of HK2 (Fig. 3I–L). These data suggested that GSDMB regulates the glucose metabolism via promoting the phosphorylation of Tyr 705 STAT3 in bladder cancer cells.
GSDMB-STAT3 signaling regulates IGFBP3 expression in bladder cancer
The RNA-Seq assay revealed 27 up-regulated genes and 110 down-regulated genes with IGFBP3 as the most significantly down-regulated gene after silencing GSDMB in T24 cells. To verify the accuracy of these results, we knocked down or overexpressed GSDMB and found a positive correlation with the change in the protein and mRNA expression of IGFBP3 in both T24 and 5637 cells (Fig. 4A–D). IGFBP3 is closely related to the glucose metabolism in cells (Saukkonen et al., 2006; Wang et al., 2013). The above data indicated that GSDMB may activate STAT3 to regulate the glycolysis in bladder cancer cells. Therefore, we speculated that GSDMB increased IGFBP3 in a STAT3-dependent manner. Notably, the knockdown of STAT3 inhibited the expression of IGFBP3 at the transcription and translation levels (Fig. 4E, F). Furthermore, we found that STAT3 could bind to the promoter of IGFBP3 (Fig. 4G, H), indicating that STAT3 acted as a transcription factor to initiate the transcription of IGFBP3 in bladder cancer cells. Furthermore, we knocked down the expression of STAT3 while GSDMB was knocked down or overexpressed, and the results showed that IGFBP3 remained constant (Fig. 4I–L), which confirmed our speculation. In conclusion, GSDMB regulates IGFBP3 expression through the STAT3 pathway in bladder cancer.
Moreover, the rate of cell proliferation did not significantly change after the simultaneous knock down of GSDMB and STAT3, while it decreased after STAT3 knock down alone (Fig. 4M, N). The results of the in vivo experiments were also completely consistent with these results, and no statistical difference in the tumor mass or volume was detected between the two groups, which further demonstrated that GSDMB promoted the progression of bladder cancer by regulating STAT3.
USP24 interacts with GSDMB to up-regulate the GSDMB protein level in bladder cancer
After revealing the important role of GSDMB in bladder cancer, we studied how GSDMB was regulated in bladder cancer for the targeted treatment of bladder cancer. Mass spectrometry of GSDMB indicated that GSDMB may interact with USP24 (Table S1), which was verified in bladder cancer cells through immunoprecipitation (Fig. 5A). Interestingly, GEPIA network tool analysis showed that there was no correlation between GSDMB and USP24 in their RNA levels (Fig. 5B). To analyze the protein expression levels of GSDMB and USP24, a bladder cancer tissue microarray (n = 80) was performed, and obtained typical images are displayed in Fig. 5C. The results showed a significant positive correlation between GSDMB and USP24 proteins in clinical specimens (Fig. 5D). As a deubiquitinating enzyme, ubiquitin-specific peptidase USP24 could deubiquitinate downstream proteins to maintain their stability 22, 23. Thus, USP24 may mediate the deubiquitination of GSDMB and maintain its stability. Consistently, the protein expression level of GSDMB also decreased or increased after knock down or overexpression of USP24, respectively, while the mRNA level of GSDMB remained constant (Fig. 5E–H). Furthermore, the changes in GSDMB observed after treatment with USP24 inhibitor EOAI3402143 were exactly the same as those observed after knocking down USP24 (Fig. 5I, J). To sum up, USP24 interacts with GSDMB to up-regulate the GSDMB protein level in bladder cancer.
USP24 stabilizes GSDMB to promote STAT3 phosphorylation in bladder cancer cells
Given that USP24 promoted the expression of the GSDMB protein levels rather than that of the mRNA levels in bladder cancer cells, we hypothesized that USP24 may regulate the stability of GSDMB through the ubiquitinated proteasome pathway. We have shown that knocking down USP24 using shUSP24 or USP24 inhibitor EOAI3402143 could reduce the protein expression of GSDMB, but the process was inhibited by proteasome inhibitor MG132 (Fig. 6A, B, E, F). Moreover, the protein half-life of GSDMB was significantly decreased when USP24 was knocked out or inhibited, while overexpression of USP24 showed the opposite effect (Fig. 6C, G). The proteasome recognized and degraded ubiquitinated proteins 24. We proved that knocking down or inhibiting USP24 increased the polyubiquitination of GSDMB, but overexpressing USP24 decreased the polyubiquitination of GSDMB in bladder cancer cells (Fig. 6D, H). These data suggest that USP24 promoted the stability of the GSDMB protein in bladder cancer by deubiquitinating GSDMB.
Moreover, after USP24 was inhibited or overexpressed, the expression level of p-STAT3 Y705 also decreased or increased, respectively (Fig. 6I–K). This raised the question if USP24 regulates the activation of the STAT3 signaling pathway through GSDMB. To verify this hypothesis, we knocked down the expression of GSDMB while knocking down or suppressing the expression of USP24. The results showed that the p-STAT3 Y705 protein level remained stable, while it decreased when GSDMB was knocked down alone (Fig. 6I, J). Similarly, overexpressing USP24 after knocking down GSDMB also resulted in a constant p-STAT3 Y705 protein level (Fig. 6K). Therefore, the USP24–GSDMB pathway axis played a key role in modulating the phosphorylation of STAT3. In conclusion, USP24 stabilizes GSDMB to promote STAT3 phosphorylation in bladder cancer cells (Fig. 6L).