Hypoxia represses early responses of prostate and renal cancer cells to YM155 independent of HIF-1α and HIF-2α

Abstract

The imidazolium compound Sepantronium Bromide (YM155) successfully promotes tumor regression in various pre-clinical models but has shown modest responses in human clinical trials. We provide evidence to support that the hypoxic milieu of tumors may limit the clinical usefulness of YM155. Hypoxia (1% O₂) strongly (>16-fold) represses the cytotoxic activity of YM155 on prostate and renal cancer cells in vitro. Hypoxia also represses all early signaling responses associated with YM155, including activation of AMPK and retinoblastoma protein (Rb), inactivation of the mechanistic target of rapamycin complex 1 (mTORC1), inhibition of phospho-ribosomal protein S6 (rS6), and suppression of the expression of Cyclin Ds, Mcl-1 and Survivin. Cells pre-incubated with hypoxia for 24 ​h are desensitized to YM155 even when they are treated with YM155 under atmospheric oxygen conditions, supporting that cells at least temporarily retain hypoxia-induced resistance to YM155. We tested the role of hypoxia-inducible factor (HIF)-1α and HIF-2α in the hypoxia-induced resistance to YM155 by comparing responses of YM155 in VHL-proficient versus VHL-deficient RCC4 and 786-O renal cancer cells and silencing HIF expression in PC-3 prostate cancer cells. Those studies suggested that hypoxia-induced resistance to YM155 occurs independent of HIF-1α and HIF-2α. Moreover, the hypoxia mimetics deferoxamine and dimethyloxalylglycine, which robustly induce HIF-1α levels in PC-3 ​cells under atmospheric oxygen, did not diminish their early cellular responses to YM155. Collectively, our data support that hypoxia induces resistance of cells to YM155 through a HIF-1α and HIF-2α-independent mechanism. We hypothesize that a hypothetical hypoxia-inducer factor (HIF-X) represses early signaling responses to YM155.

Keywords: AMPK; Drug resistance; Genitourinary cancers; Hypoxia; VHL; mTORC1.

Graphical abstract

Fig. 1

Effect of hypoxia on the suppression of PC-3 ​cell growth. PC-3 ​cells were allowed to attach to 12-well dishes for 24 ​h under normal atmospheric oxygen (21% O₂) and then treated with various doses of YM155 or vehicle for 72 ​h in 1% or 21% O₂(A), pre-incubated for 24 ​h in 1% or 21% O₂ before treatment with various doses of YM155 or vehicle for 72 ​h in 1% or 21% O₂(B), pre-incubated for 24 ​h in 2.5% or 21% O₂ before treatment with various doses of YM155 or vehicle for 72 ​h in 2.5% or 21% O₂(C), and pre-incubated for 24 ​h in 5% or 21% O₂ before treatment with various doses of YM155 or vehicle for 72 ​h in 5% or 21% O₂(D). Cell growth was assessed by crystal violet staining (A570 nm) as described in Materials and Methods. Data points represent the average of triplicate determinations (biological replicates) ​± ​SE. Statistical significance (p-values) was determined by Student’s t-test (two-tailed). ∗p ​< ​0.01.

Fig. 2

Effect of hypoxia on the ability of YM155 control cell cycle regulators in PC-3 ​cells. PC-3 ​cells were allowed to attach to 6-well dishes for 24 ​h under normal atmospheric oxygen (21% O₂), the cells were pre-incubated for 24 ​h in 1% or 21% O₂ before treatment with various doses of YM155 or vehicle for 8 ​h continuously under the same conditions. Expression of P–Rb (S807/811), Cyclin D1, Cyclin D2, Survivin, HIF-1α, and β-Actin were determined by Western blot analysis, as described in Materials and Methods. The data shown are representative of three independent experiments.

Fig. 3

Pre-treatment with hypoxia alone suppresses early signaling responses in PC-3 elicited by YM155 in 21% O₂. PC-3 ​cells were allowed to attach to 6-well dishes for 24 ​h under normal atmospheric oxygen (21% O₂), the cells were pre-incubated for 24 ​h in 1% or 21% O₂ before treatment with various doses of YM155 or vehicle for 4 ​h in 21% O₂ (A). Expression of Mcl-1, Cyclin D2, P-Raptor (S792), Raptor, P-rS6 (S240), rS6, P-AMPKα (T172) Survivin, P-p70S6K (T389), and β-Actin were determined by Western blot analysis (B), as described in Materials and Methods. C) The ED₅₀ of YM155 on the phosphorylation of Raptor and AMPK in cells pre-treated for 24 ​h with 21% O₂ versus 1% O₂ was quantified by Image J. The data shown are representative of two independently conducted experiments. The graphs in C were generated by normalization to β-actin values after the background phosphorylation induced by hypoxia was subtracted for the comparative analysis of ED₅₀s of YM155. Each data point represents the average of 5 replicates (5 composite scans from 3 different blots representing biological replicates) ​± ​SE. Statistical significance (p-values) was determined by Student’s t-test (two-tailed). ∗p ​< ​0.05.

Fig. 4

Impact of HIF-1α on the ability of hypoxia to YM155 to control growth regulators and suppress the growth of PC-3 ​cells. PC-3 ​cells stably silenced for the expression of HIF-1α by lentiviral mediated transduction of HIF-1α shRNAs or scrambled control shRNA were allowed to attach overnight to 6-well (A) and 12-well (B) dishes in 21% O₂. Cells were then pre-incubated for 24 ​h with 1% or 21% O₂ before treatment with various doses of YM155 or vehicle for 4 ​h (A) or 72 ​h (B) with 1% or 21% O₂. The dependence of HIF-1α on the impact of hypoxia on YM155 responses was assessed by Western blot analysis (A) and crystal violet growth assay (B). Data points in panel B represent the average of biological triplicates ​± ​SE.

Fig. 5

Impact of hypoxia on YM155 responses in renal carcinoma cells deficient or proficient in pVHL. A) VHL-null RCC4 cells were first allowed to attach to 6-well dishes for 24 ​h under normal atmospheric oxygen (21% O₂), they were then pre-incubated for 24 ​h in 1% or 21% O₂ and thereafter treated with various doses of YM155 or vehicle for 24 ​h in 1% or 21% O₂. B) VHL null RCC4 cells and VHL expressing RCC4 cells were first allowed to attach to 12-well dishes for 24 ​h under normal atmospheric oxygen (21% O₂), they were then pre-incubated for 24 ​h in 1% or 21% O₂ and subsequently treated with YM155 or vehicle for 72 ​h with their respective levels of O₂. C, D) VLH proficient and null 786-O cells were first allowed to attach to 12-well (C) or 6-well (D) dishes for 24 ​h under normal atmospheric oxygen (21% O₂), then pre-incubated for 24 ​h in 1% or 21% O₂ and subsequently treated with YM155 or vehicle for 72 ​h (C) or 24 ​h (D) in their respective levels of O₂. E, F) VHL-null 786-O cells stably silenced for HIF-2α by lentiviral-mediated transduction of HIF-2α shRNAs versus scramble control shRNA were first allowed to attach to 12-well (E) or 6-well (F) dishes for 24 ​h under normal atmospheric oxygen (21% O₂), they were then pre-incubated for 24 ​h in 1% or 21% O₂ and subsequently treated with YM155 or vehicle for 72 ​h (E) or 24 ​h (F) with their respective levels of O₂. Cells were analyzed for cell growth by crystal violet growth assay (B, C, E) or alteration of signaling responses by Western blot (A, D, F). Data points in panels B, D, and E represent the average of triplicate biological determinations ​± ​SE. Statistical significance (p-values) was determined by Student’s t-test. ∗p ​< ​0.01 comparisons are for differences between 1% versus 21% O₂ for both VHL proficient and deficient cells.

Fig. 6

YM155 alters early cellular responses independent of cellular delivery of oxygen. PC-3 ​cells were allowed to attach to 6-well dishes for 24 ​h under normal culture conditions, the cells were then treated with 0.2 ​mM deferoxamine mesylate (DFX), 1 ​mM dimethyloxalylglycine (DMOG), or water (vehicle). Following 180 ​min treatment cells received 100 ​nM YM155 or vehicle, and after an additional 120 ​min, cells were harvested for Western blot analysis (A). Expression of P-Raptor (S792), HIF-1α, VEGFR2, P-rS6 (S240), P-AMPKα (T172), AMPKα1, Cyclin D1, and GAPDH was determined by Western blot analysis. The data shown are representative of two independent experiments.

Fig. 7

Summary of a model describing the potential mechanism by which hypoxia suppresses cellular responses to YM155. Our study support that hypoxia, through a HIF-1α and HIF-2α-independent mechanism, inhibits the cytostatic/cytotoxic responses to YM155 in part by inhibiting the activation of AMPK and subsequently reversing YM155’s suppression of a previously reported downstream signaling cascade, including the activity of mTORC1, S6K1, and rS6 and translation of Survivin, Mcl-1, and Cyclin Ds (Danielpour et al., 2019). The ability of hypoxia to repress YM155-mediated loss of phosphorylated pRB can be explained by changes in Cyclin Ds that partner with cyclin-dependent kinases to inactivate pRB through the phosphorylation of pRB at S807/811 (Danielpour et al., 2019). The iron chelator DFX, which induces the expression of HIF-1α by repressing proline hydroxylation of HIF-1α, does not antagonize early responses to YM155, suggesting that YM155 activates AMPK through an iron-independent mechanism. We hypothesize that an unknown hypoxia-inducible factor (designated HIF-X) suppresses cellular response to YM155.