Autophagy inhibition by sustained overproduction of IL6 contributes to arsenic carcinogenesis

Abstract

Chronic inflammation has been implicated as an etiologic factor in cancer, whereas autophagy may help preserve cancer cell survival but exert anti-inflammatory effects. How these phenomenas interact during carcinogenesis remains unclear. We explored this question in a human bronchial epithelial cell-based model of lung carcinogenesis that is mediated by subchronic exposure to arsenic. We found that sustained overexpression of the pro-inflammatory IL6 promoted arsenic-induced cell transformation by inhibiting autophagy. Conversely, strategies to enhance autophagy counteracted the effect of IL6 in the model. These findings were confirmed and extended in a mouse model of arsenic-induced lung cancer. Mechanistic investigations suggested that mTOR inhibition contributed to the activation of autophagy, whereas IL6 overexpression was sufficient to block autophagy by supporting Beclin-1/Mcl-1 interaction. Overall, our findings argued that chronic inflammatory states driven by IL6 could antagonize autophagic states that may help preserve cancer cell survival and promote malignant progression, suggesting a need to uncouple inflammation and autophagy controls to enable tumor progression.

Figure 1

(A) BEAS-2B cells were treated with 0.25 μM sodium arsenite for 24 hours (24 h), 1 week (1 w) and 16 weeks (16 w). The cytokines in the supernatants of arsenic-treated or passage-matched controlled cells were determined by ELISA. Data are mean ± SEM of three experiments, * p<0.05 vs. matched passage controlled cells, # p<0.05. (B) The mRNA levels of IL6 and GM-CSF in arsenic-treated BEAS-2B cells were determined by qRT-PCR and were normalized to matched actin mRNA levels. The data were shown as relative expression folds compared to passage controlled cells. Data are mean ± SEM of three experiments, * p<0.05.

Figure 2

(A) The down-regulation of IL6 or GM-CSF in BEAS-2B cells stably transfected with IL6 or GM-CSF shRNA was determined by immunoblotting and quantified by densitometry. Data are mean ± SEM of three experiments, * p<0.05. (B) Soft agar assays were performed after IL6 knocking-down (IL6 KD) cells, GM-CSF KD cells or passage-matched control cells were exposed to arsenic (As) for 16 weeks. The numbers of colonies in each well were counted and quantified. Data are mean ± SEM of three experiments, * p<0.05. (C) After BEAS-2B cells were treated as in B for 16 weeks, 2 × 10⁶ of these cells were injected sc into the nude mice (n=6). Tumor volumes and incidence rates were determined 3 weeks after injection, (n=6, mean ± SEM), * p<0.05. Scale bar length: 5 mm.

Figure 3

(A) BEAS-2B cells, (B) SAEC cells and (C) NL20 cells were treated with arsenic (As) or recombinant IL6 (IL6, 10 ng/ml) alone or combined; IL6 KD cells were exposed to arsenic (As) for 24 hours. The levels of LC3-II and p62 were measured by immunoblotting. The decreased levels of IL6 by transfection of IL6 shRNA in SAEC or NL20 cells were shown in the upper levels of (B) and (C), respectively. Data are mean ± SEM of three experiments, * p<0.05.

Figure 4

(A) Immunoblotting shows the increased level of Beclin-1 in BEAS-2B cells stably transfected with Beclin-1 cDNA. (B) The levels of LC3-II and P62 in Beclin-1 over-expressed cells (Becl) and vector control cells (vector) with or without 24 hours arsenic treatment were shown by immunoblotting. (C) Regular (control) or Beclin-1 over-expressed (Becl) BEAS-2B cells were exposed to arsenic (As), human recombinant IL6 (IL6, 10 ng/ml) or rapamycin (Rap, 10 nM) alone or combined for 16 weeks, then soft agar assays were performed. The numbers of the colonies in each well were counted and quantified. Data are mean ± SEM of three experiments, * p<0.05.

Figure 5

(A) The levels of p-MTOR, p-P70S6K and p-4E-BP1 were determined by immunoblotting after regular BEAS-2B cells (BEAS-2B) or IL6 KD BEAS-2B cells (IL6 KD) were exposed to arsenic for 24 hours (24 h), 1 week (1 w) and 16 weeks (16 w). (B) The levels of Bcl-2, Bcl-xL, Mcl-1 and Bcl-w as well as Bad and tBid were determined by immunoblotting after BEAS-2B cells were treated with arsenic for 24 hours, 1 week and 16 weeks. (C) The expression levels of Mcl-1 were determined after SAEC (upper panel) and NL20 cells (lower panel) were treated with arsenic for 24 hours, 1 week and 3 weeks. (D) The levels of Mcl-1 were determined by immunoblotting after BEAS-2B cells (BEAS-2B) or IL6 KD BEAS-2B cells (IL6KD) were treated with arsenic, or BEAS-2B cells were co-treated with arsenic and recombinant IL6 (IL6) for 24 hours, 1 week and 16 weeks, * p<0.05 vs matched control groups, # p<0.05 vs the same treatment groups in BEAS-2B or IL6 KD cells. (E) Upper panel: the levels of Mcl-1 in BEAS-2B cells stably transfected with Mcl-1 siRNA or scrambled siRNA were determined by immunoblotting. Lower panel: scrambled control (Scrambled Ct) BEAS-2B cells and Mcl-1 knocking-down (Mcl-1 KD) cells were exposed to arsenic or arsenic plus recombinant IL6 (10 ng/ml) for 24 hours. The levels of LC3-II and p62 were measured by immunoblotting. (F) The binding of Beclin 1/Mcl-1 was determined by immunoprecipitation after BEAS-2B cells were treated with arsenic for 24 hours, 1 week and 16 weeks. Relative levels of Mcl-1 or Beclin 1 in arsenic treated samples were compared with that in their matched control groups, respectively. (G). BEAS-2B, SAEC and NL20 cells were treated with recombinant IL6 (10 ng/ml) for 1 week, the association of Mcl-1 and Beclin 1 was determined by immunoprecipitation. Data are mean ± SEM of three experiments, * p<0.05.

Figure 6

(A) The levels of Mcl-1 were determined by immunoblotting after BEAS-2B cells were treated with arsenic and rapamycin (Rap, 10 nM) for 24 hours, 1 week and 16 weeks. (B) The expression level of STAT3 was determined in BEAS-2B cells stably transfected with STAT3 shRNA. (C) The Mcl-1 levels were determined by immunoblotting after STAT3 KD BEAS-2B cells were treated with arsenic (upper panel), arsenic plus IL6 (10 ng/ml, middle panel) or arsenic plus rapamycin (Rap, 10 nM, lower panel) for 24 hours, 1 week and 3 weeks. Data are mean ± SEM, of three experiments, * p<0.05.