BET inhibition suppresses cells growth and induces cells apoptosis in TNBC by attenuating NF-κB signaling
To determine the effects of BET inhibition on human TNBC cell viability and apoptosis, MDA-MB-231 (231), MDA-MB-468 (468) and BT549 were exposed to different concentrations of BET inhibitors JQ1 and I-BET151 for 48 h. And then, cell viability and apoptosis were detected by CCK-8 and Annexin-V/PI assay, respectively. As shown in Fig. 1A and B, both JQ1 and I-BET151 caused a concentration dependent inhibition of cell viability and induction of cell apoptosis in all three cell lines. Additionally, to examine the possible role of BET inhibitors in NF-κB activation in TNBC, NF-κB activity assays were measured with a dual-luciferase reporter system. As shown in Fig. 1C, both BET inhibitors obviously decreased the NF-κB luciferase reporter activity in a dose-dependent manner in three TNBC cells. In 231 cells, we showed that both JQ1 and I-BET151 decreased the NF-κB luciferase reporter activity in a time-dependent manner (Fig. 1D).
To further confirm attenuation of NF-κB signaling by BET inhibition, we detected the expression of p-p65 (active form of NF-κB subunit p65) and NF-κB upstream regulators IKBα in TNBC cells treated by BET inhibitor by Western blot assay. As shown in Fig. 1E and G, JQ1 and I-BET151 significantly increased IKBα levels and decreased p-p65 expression. The nuclear translocation of NF-κB p65 represents the induction of NF-κB signal activation. By extracting nuclear protein for Western blot assay, we found that JQ1 and I-BET151 significantly decreased p65 expression in nuclear and increased p65 expression in cytoplasm (Fig. 1F and H). These data suggested that BET inhibitors suppressed cells growth and induced cells apoptosis in TNBC by attenuating NF-κB signaling.
BET inhibition attenuates NF-κB signaling in TNBC by decreasing IKBKE expression
To explore the molecular mechanism of NF-κB signaling inhibition by BET inhibitor, we performed PCR array to detect the mRNA levels of NF-κB cytoplasmic sequestering molecule including BCL3, CHUK (IKKα), IKBKB (IKKβ), IKBKE, IKBKG, NFKBIA (IKBα, MAD3), NFKBIB (TRIP9), NFKBIE. As shown in Fig. 2A, we found BET inhibitors JQ1 and I-BET151 observably decreased the IKBKE mRNA levels in 231 cells. To simultaneously correlate the fold change on gene expression induced by JQ1 and the statistical significance at the global level, volcano plot was delineated by log2-fold change as x axis and a log10 P value as y axis. As shown in Fig. 2B, volcano plot showed the difference of gene expression in different dose JQ1 treated groups. The four red dots represented the selected IKBKE gene in four doses of JQ1 treated groups. These results showed IKBKE was the most significant change gene regulated by JQ1 in 231 cells.
To further confirm expression of IKBKE regulated by BET inhibition, we treated 231 and BT549 cells with JQ1 and I-BET151. IKBKE mRNA and protein levels were detected by RT-PCR and Western blot respectively. As shown in Fig. 2C, D and E, F, JQ1 and I-BET151 markedly reduced IKBKE levels in a concentration demand manner. In addition, we also showed overexpression of IKBKE obviously blocked the suppression of NF-κB signal by BET inhibitor in TNBC cells. As shown in Fig. 2G and H, in 231 and BT549 cells transfected IKBKE overexpression vector, the IKBα expression induced by BET inhibitors was clearly downregulated, while p-p65 expression inhibited by BET inhibitors was clearly upregulated. Above data showed BET inhibition decreased IKBKE expression, and thus attenuating NF-κB signaling in TNBC.
IKBKE prevents cells apoptosis and growth arrest by BET inhibition in TNBC
To determine if IKBKE could prevent the cytotoxic effects of BET inhibitor, 231 and BT549 cells with IKBKE overexpression were treated with JQ1 and I-BET151 for 48 h. The overexpressed IKBKE in TNBC cells was displayed in Fig. 3A. And then cell viability and apoptosis were determined by CCK-8 assay, hoechst 33258 staining and Annexin V-FITC and PI staining respectively. By Hoechst staining, we found BET inhibitors induced nuclear condensation and fragmentation in 231 cells (Fig. 3B). By CCK-8 assay, IKBKE was also found to significantly inhibit the cytotoxic effects of JQ1 and I-BET151 in two types of TNBC cells (Fig. 3C and D). Additionally, Fig. 3E-I indicated that two BET inhibitors induced cell apoptosis was significantly attenuated by IKBKE in both 231 and BT549 cells.
BET inhibition downregulates IKBKE levels through BRD2/E2F1 axis
Previous studies have demonstrated that IKBKE is transcriptionally upregulated by E2F1 [23]. Therefore, we investigated whether BET inhibition downregulates E2F1 levels. Our results showed that both BET inhibitors JQ1 and I-BET151 decreased E2F1 protein expression and E2F1 mRNA levels in 231 cells (Fig. 4A and B). To directly verify that IKBKE induced by E2F1, we used siRNA to knock down the E2F1 expression and detected the IKBKE expression in 231 cells. As shown in Fig. 4C, E2F1 siRNA significantly decreased IKBKE expression. To further determine that BET inhibition attenuated IKBKE levels was involved in repressing E2F1, 231 cells were transfected E2F1-HA overexpression constructs, subsequently cells were treated with BET inhibitor JQ1. Results showed that overexpression of E2F1 promoted the IKBKE expression (Fig. 4D). Moreover, attenuation of IKBKE by JQ1 was remarkably rescued by E2F1-HA overexpression vectors (Fig. 4D).
The BET protein BRD2 has been linked to promote E2F1 transcription [24]. To investigate whether BET inhibitor suppressed IKBKE levels through inhibiting BRD2/E2F1 pathway. The BRD2 was depleted by transfecting siRNA in 231 cells. We found BRD2 siRNA significantly decreased both E2F1 and IKBKE expression (Fig. 4E). Furthermore, attenuation of E2F1 and IKBKE by JQ1 was remarkably rescued by BRD2-HA overexpression vectors (Fig. 4F). To further examine the association between BRD2, E2F1 and IKBKE, the mRNA levels of these three genes in 1104 breast cancer tissues were extracted from TCGA database. By Pearson correlation analysis, we found significantly positive correlations between E2F1 and IKBKE expression (r = 0.15, P < 0.001; Fig. 4G), between E2F1 and BRD2 expression (r = 0.11, P < 0.001; Fig. 4H).
TNBC-stimulated TAMs activate NF-κB signaling to enhance breast cancer resistance to BET inhibition
Previous studies have showed that TAMs are associated with chemoresistance in breast cancer [10]. We thus supposed that the TAMs might be associated with the TNBC cells resistance to BET inhibitor. To assess potential effects of macrophage on the BET inhibitor resistance of TNBC, we mimicked in vivo model by co-culturing TNBC cells with macrophages derived from PBMC or THP1 cells. Subsequently, these TNBC cells were subjected to BET inhibitors treatment and then to perform drug sensitivity and apoptosis analysis. As shown in Fig. 5A, the IC50 of JQ1 in TNBC cells co-cultured with PBMC-TAMs was significantly higher than those of their corresponding control cells. Moreover, apoptosis assay showed that the ratios of apoptosis induced by BET inhibitors were significantly lower in 231 and BT549 cells co-cultured with PBMC- or THP1-derived macrophages than control cells (Fig. 5B-F). These results suggested TAMs promoted TNBC cells resistance to BET inhibitor.
To further determine TAMs increased JQ1 resistance in TNBC, we performed tumor formation assay in BalB/C nude mice. TAMs were induced from THP1 cells and then co-cultured with 231 cells for 72 h in vitro. Subsequently, 231 cells mixed with THP1 were injected into flanks of nude mice. Treatment schedule was shown in Fig. 5J. We found the tumors of the 231 cells mixed with THP1 cells were more resistant to JQ1 therapy than those tumors of the 231 alone (Fig. 5G and I). In addition, no significant loss in body weight was observed in these mice treated with JQ1 (Fig. 5H), suggesting the side effects of JQ1 were minimal in vivo.
In addition, THP1-derived macrophages were also co-cultured with non-TNBC cell line MCF7. Subsequently, these macrophages were used to co-culture with 231 cells. After 72 h, 231 cells were treated with different dose of JQ1 for 48 h. And then, cell apoptosis was evaluated. Results showed macrophages co-cultured by MCF7 failed to decrease the cytotoxic effects of JQ1 in 231 cells (Fig. S1).
Since TNBC cells apoptosis induced by BET inhibitor was associated with suppression of NF-κB activity, we thus explored whether BET inhibitors reduced NF-κB activity to provoke cells apoptosis. By immunofluorescence assay, we showed the translocation of NF-κB p65 from the cytoplasm to the nucleus was notably blocked by both JQ1 and I-BET151 in 231 cells. However, the levels of NF-κB p65 in the nucleus were obviously increased in 231 cells treated with PBMC-TAM-CM or co-cultured with PBMC-TAMs. Furthermore, the reduction of p65 levels in nucleus by BET inhibitors was significantly blocked by PBMC-TAM-CM or PBMC-TAMs (Fig. 5K). Besides, by Western blot, we also found that PBMC-TAMs or PBMC-TAM-CM attenuated the upregulation of IKBα levels and downregulation of p-p65 expression elicited by BET inhibitors (Fig. 5L).
TNBC-treated TAMs promote IKBKE expression to prevent BET inhibitor-induced apoptosis by activating STAT3 signaling
Above data have indicated that BET inhibitor induced TNBC cell apoptosis was associated with inhibition of IKBKE/NF-κB signal. Moreover, TAMs have been shown to activate NF-κB signaling in breast cancer cells in this study. We thus explored whether TAMs regulated IKBKE expression of TNBC cells. The results in Fig. 6A indicated that TNBC cells-treated TAMs or TAM-CM significantly increased IKBKE protein levels, while BET inhibitors restrained IKBKE expression in 231 cells. Moreover, the reduction of IKBKE by BET inhibitors in 231 cells was obviously alleviated by TAMs or TAM-CM (Fig. 6A). By RT-PCR assay, we showed that the mRNA levels of IKBKE were observably higher in 231 cells co-cultured with TNBC cells-activated TAMs than those cells co-cultured with Non-TNBC cells MCF7-activated TAMs (Fig. S2A). Similarly, in 231 cells treated with TNBC cells-activated TAMs CM, the mRNA levels of IKBKE were significantly higher (Fig. S2B).
Clinical studies have found that a large number of macrophages are infiltrated into the tumor tissue in breast cancer [25]. Thus, we speculated that TAMs in TNBC induced higher levels of IKBKE of TNBC cells relative to TAMs in Non-TNBC. By RT-PCR analysis, IKBKE expression was performed in 19 TNBC tissues and 45 Non-TNBC tissues from breast cancer patients. Results indicated that the IKBKE levels were higher in TNBC tissues than Non-TNBC tissues (Fig. 6B). In addition, IKBKE expression was analysed in 93 TNBC tissues and 898 Non-TNBC tissues from breast cancer patients in TCGA data sets. We found the IKBKE levels were also higher in TNBC tissues than Non-TNBC tissues (Fig. 6C).
Previous study has reported that IKBKE expression was regulated by STAT3 signal [26]. To determine the mechanism by which TAMs regulated IKBKE expression, we used STAT3 inhibitor STAT3-IN-1 to treat TNBC cells. Western blot analysis indicated induction of IKBKE levels by TAM-CM were strikingly inhibited by STAT3-IN-1 (Fig. 6D). To investigate whether STAT3 inhibition influenced TAM-mediated TNBC cells resistance to BET inhibitor, we performed cell apoptosis and viability assay in 231 and BT549 cells. As shown in Fig. 6E and F, STAT3-IN-1 synergistically enhanced JQ1-induced cytotoxic effects in both 231 and BT549 cells. Moreover, STAT3-IN-1 completely abrogated the resistance to BET inhibitors induced by TAMs or TAM-CM in TNBC cells. In addition, to more directly assess the role of IKBKE in TAM-induced BET inhibitor resistance, cell apoptosis and viability were evaluated in JQ1-treated TNBC cells with siRNA-mediated knockdown of IKBKE. As shown in Fig. 6G-I, knockdown of IKBKE significantly enhanced cell apoptosis induced by JQ1 treatment in TNBC cells. Importantly, depletion of IKBKE significantly prevented the resistance to BET inhibitors induced by TAMs or TAM-CM in TNBC cells.
To test whether STAT3-IN-1 combined JQ1 was an effective strategy to overcome TAM-induced JQ1 resistance in vivo, we further assessed the antitumor activity of the two drugs in xenografts established in nude mice implanted with 231 mixed THP1 cells. Treatment schedule was shown in Fig. 6M. As shown in Fig. 6J, either JQ1 or STAT3-IN-1 alone inhibited the tumors growth; the two drugs in combination displayed a much stronger antitumor effects in the xenograft tumor models (P < 0.001). By TUNEL assay, we detect the apoptotic cells in tumor tissues treated by JQ1 and STAT3-IN-1 alone or in combination. Fig. 6L showed that the numbers of apoptotic cells in the tumors treated by JQ1 and STAT3-IN-1 in combination were significantly higher than other group. This result also suggested that JQ1 combined with STAT3-IN-1 displayed a much stronger antitumor effects in vivo. The body weight of mice recorded throughout the experiments was shown in Fig. 6K. Results showed no significant loss in body weight was found in the mice treated with the two drugs alone or in combination, suggesting the side effects were minimal in vivo.
TAMs promote IKBKE expression in TNBC by secreting IL-6 and IL-10.
STAT3 activation is responsible for IL-6 and IL-10. Therefore, we supposed the induction of IKBKE by TAMs in TNBC cells was relevant to IL-6 and IL-10 secreted by TAMs. The IL-6 and IL-10 in TAMs and breast cancer condition media were detected by ELISA assay. Fig. S3 indicated the levels of two interleukins were obviously higher in 231- and BT549-treated TAMs CM relative to others CM from MCF7-treated TAMs, THP1 macrophage and breast cancer cells. Subsequently, the IL-6 and IL-10 in TAMs CM was deprived by incubating with IL-6 and IL-10 neutralizing antibody alone or in combination. And then TAMs CM was used to treat 231 and BT549 cells. Subsequently, IKBKE protein levels were detected by Western blot in TNBC cells. As shown in Fig. 7A and B, IL-6 and IL-10 antibodies alone sharply decreased IKBKE levels in both TNBC cells respectively. Moreover, IL-6 combined with IL-10 antibodies synergistically inhibited IKBKE expression. Additionally, by RT-PCR experiment, we also showed that IL-6 and IL-10 antibodies significantly reduced IKBKE mRNA levels in both TNBC cells treated with TAMs CM. IKBKE mRNA levels was synergistically inhibited by IL-6 and IL-10 antibodies in combination (Fig. S4A and B).
More importantly, we also validated the IKBKE expression was induced by IL-6 and IL-10 in TNBC cells. As shown in Fig. 7C-F, IL-6 and IL-10 markedly increased IKBKE protein levels in both TNBC cells in concentration-dependent manner. Similarly, by RT-PCR assay, we also observed IKBKE mRNA levels were increased by IL-6 and IL-10 in dose-dependent manner in 231 and BT549 cells (Fig. S4C-F).