The ST6Gal-I sialyltransferase protects tumor cells against hypoxia by enhancing HIF-1α signaling

Abstract

Aberrant cell surface glycosylation is prevalent in tumor cells, and there is ample evidence that glycans have functional roles in carcinogenesis. Nonetheless, many molecular details remain unclear. Tumor cells frequently exhibit increased α2-6 sialylation on N-glycans, a modification that is added by the ST6Gal-I sialyltransferase, and emerging evidence suggests that ST6Gal-I-mediated sialylation promotes the survival of tumor cells exposed to various cell stressors. Here we report that ST6Gal-I protects cancer cells from hypoxic stress. It is well known that hypoxia-inducible factor 1α (HIF-1α) is stabilized in hypoxic cells, and, in turn, HIF-1α directs the transcription of genes important for cell survival. To investigate a putative role for ST6Gal-I in the hypoxic response, we examined HIF-1α accumulation in ovarian and pancreatic cancer cells in ST6Gal-I overexpression or knockdown experiments. We found that ST6Gal-I activity augmented HIF-1α accumulation in cells grown in a hypoxic environment or treated with two chemical hypoxia mimetics, deferoxamine and dimethyloxalylglycine. Correspondingly, hypoxic cells with high ST6Gal-I expression had increased mRNA levels of HIF-1α transcriptional targets, including the glucose transporter genes GLUT1 and GLUT3 and the glycolytic enzyme gene PDHK1 Interestingly, high ST6Gal-I-expressing cells also had an increased pool of HIF-1α mRNA, suggesting that ST6Gal-I may influence HIF-1α expression. Finally, cells grown in hypoxia for several weeks displayed enriched ST6Gal-I expression, consistent with a pro-survival function. Taken together, these findings unravel a glycosylation-dependent mechanism that facilitates tumor cell adaptation to a hypoxic milieu.

Keywords: ST6Gal-I; anoxia; cancer stem cells; cell stress; cell surface glycosylation; glycan; glycosylation; hypoxia; hypoxia-inducible factor (HIF); sialyltransferase.

Figure 1.

Cells with high ST6Gal-I expression have increased accumulation of HIF-1α after treatment with DFO. A, OV4 cells were stably transduced with a lentivirus encoding ST6Gal-I, and ST6Gal-I OE was verified by immunoblotting. Control cells were generated by stable transduction of an EV. B, MiaPaCa-2 cells were stably transduced with ST6Gal-I–targeting shRNA, and ST6Gal-I KD was confirmed by immunoblotting. C, PA-1 cells were stably transduced with shRNA for ST6Gal-I, and ST6Gal-I KD was verified by immunoblotting. D–F, OV4 (D), MiaPaCa-2 (E), and PA-1 cells (F) were stained with SNA-DyLIght649 to detect surface α2–6 sialic acids and evaluated by flow cytometry. The graphs depict mean fluorescence intensity (MFI). G, OV4 EV and OE cells were cultured with the hypoxia mimetic DFO for 12 h, and then whole-cell lysates were immunoblotted for HIF-1α. H and I, MiaPaCa-2 EV and KD cells (H) and PA-1 EV and KD cells (I) were cultured with DFO for 24 h and probed for HIF-1α.

Figure 2.

Cells with high ST6Gal-I expression have increased HIF-1α accumulation after treatment with DMOG. A, OV4 EV and OE cells were cultured with the hypoxia mimetic DMOG for 24 h, and then cell lysates were immunoblotted for HIF-1α. B, MiaPaCa-2 EV and KD cells were cultured with DMOG, and lysates were probed for HIF-1α. C, OV4 and MiaPaCa-2 cell models were imaged 24 h after treatment with DMOG.

Figure 3.

ST6Gal-I activity promotes HIF-1α accumulation in cells exposed to physiologic hypoxia. A, OV4 EV and OE cells were cultured in either normoxia (Norm) or hypoxia (Hypox) for 24 h, and then cell lysates were immunoblotted for HIF-1α. B, PA-1 cells were cultured for 12 or 24 h in hypoxia or, alternatively, for 24 h in normoxia. EV, KD, and OE cell lysates were probed for HIF-1α. C, PA-1 ShC and KD cells were cultured in hypoxia for 12 or 24 h and immunoblotted for HIF-1α. D, HIF-1α immunoblots of MiaPaCa-2 EV and KD cells cultured in hypoxia or normoxia for 24, 48, and 72 h. In the blot for normoxia (bottom panel), a positive control (PC) sample was loaded onto the gel to provide a detectable HIF-1α signal. The positive control lysate was from EV cells exposed to hypoxia for 24 h. E, MiaPaCa-2 ShC and KD cells were cultured in hypoxia for 12 or 24 h and then blotted for HIF-1α. F, MiaPaCa-2 cells transduced with two different shRNA constructs (KD and KD-2) were cultured in hypoxia for 24 h and immunoblotted for HIF-1α. Lysates were also probed for ST6Gal-I to confirm knockdown.

Figure 4.

High ST6Gal-I expression provides a growth advantage. A, MiaPaCa-2 parental cells were cultured in hypoxia for 6 weeks to develop a hypoxia-adapted population. The original parental (Par) and hypoxia-adapted populations were immunoblotted for ST6Gal-I. B, MiaPaCa-2 EV and KD cells were seeded at equal density and allowed to adhere overnight in standard normoxic conditions. Cultures were then placed in hypoxia for 24, 48, or 72 h, and cell number was quantified by crystal violet staining. Values for hypoxic cultures were normalized to the overnight normoxic cultures. Values represent means and S.D. for three independent experiments. *, p < 0.05. C, representative images are shown for MiaPaCa-2 EV and KD cells grown in hypoxia for 72 h.

Figure 5.

ST6Gal-I overexpression in OV4 cells promotes the hypoxia-induced transcription of HIF-1α target genes. A–D, OV4 EV and OE cells were grown in normoxia or hypoxia for 12 or 24 h, and then RT-qPCR was used to measure the expression of GLUT1 (A), GLUT3 (B), PDHK1 (C), and VEGFα (D). Graphs depict means and S.D. from at least three independent experiments. *, p < 0.05.

Figure 6.

ST6Gal-I knockdown in MiaPaCa-2 cells decreases the hypoxia-induced transcription of HIF-1α target genes. A–D, MiaPaCa-2 ShC and KD cells were grown in normoxia or hypoxia for 12 or 24 h, and then RT-qPCR was used to measure the expression of GLUT1 (A), GLUT3 (B), PDHK1 (C), and VEGFα (D). Graphs depict means and S.D.s from at least three independent experiments. *, p < 0.05.

Figure 7.

ST6Gal activity increases HIF-1α mRNA levels. A and B, OV4 (A) and MiaPaCa-2 (B) cells were cultured in normoxia or hypoxia for 12 or 24 h, and then HIF-1α mRNA expression was analyzed by RT-qPCR. C and D, levels of HIF-2α mRNA (C) or Oct4 mRNA (D) were evaluated by RT-qPCR in OV4 EV and OE cells grown in normoxia or hypoxia for 12 or 24 h. Graphs depict means and S.D. from at least three independent experiments. *, p < 0.05.