Hypoxia-inducible factor (HIF)-prolyl hydroxylase 3 (PHD3) maintains high HIF2A mRNA levels in clear cell renal cell carcinoma

Abstract

Most clear cell renal cell carcinomas (ccRCCs) have inactivation of the von Hippel-Lindau tumor suppressor protein (pVHL), resulting in the accumulation of hypoxia-inducible factor α-subunits (HIF-α) and their downstream targets. HIF-2α expression is particularly high in ccRCC and is associated with increased ccRCC growth and aggressiveness. In the canonical HIF signaling pathway, HIF-prolyl hydroxylase 3 (PHD3) suppresses HIF-2α protein by post-translational hydroxylation under sufficient oxygen availability. Here, using immunoblotting and immunofluorescence staining, qRT-PCR, and siRNA-mediated gene silencing, we show that unlike in the canonical pathway, PHD3 silencing in ccRCC cells leads to down-regulation of HIF-2α protein and mRNA. Depletion of other PHD family members had no effect on HIF-2α expression, and PHD3 knockdown in non-RCC cells resulted in the expected increase in HIF-2α protein expression. Accordingly, PHD3 knockdown decreased HIF-2α target gene expression in ccRCC cells and expression was restored upon forced HIF-2α expression. The effect of PHD3 depletion was pinpointed to HIF2A mRNA stability. In line with these in vitro results, a strong positive correlation of PHD3 and HIF2A mRNA expression in ccRCC tumors was detected. Our results suggest that in contrast to the known negative regulation of HIF-2α in most cell types, high PHD3 expression in ccRCC cells maintains elevated HIF-2α expression and that of its target genes, which may enhance kidney cancer aggressiveness.

Keywords: Egl-9 family hypoxia-inducible factor 3 (EGLN3); cancer; cancer biology; ccRCC; hypoxia; hypoxia-inducible factor (HIF); mRNA decay; post-transcriptional regulation; prolyl hydroxylase PHD3; renal cell carcinoma.

Figure 1.

PHD3 knockdown results in decreased GLUT1 and LDHA mRNA expression in ccRCC cells. A, GLUT1 and LDHA expression in 786-O cells transfected with control (scr) or PHD3-targeted siRNA (siPHD3#1, siPHD3#2) followed by normoxic (NOX, 21% oxygen) or hypoxic (HOX, 1% oxygen) exposure. Quantification of five (GLUT1) or six (LDHA) individual experiments, -fold change to scr. Bar represents mean ± S.D.; ***, p <0.001; **, p <0.01; *, p <0.05. B, GLUT1 and LDHA expression in RCC4 cells transfected with control (scr) or siPHD3#1, siPHD3#2 followed by normoxic (NOX, 21% oxygen) or hypoxic (HOX, 1% oxygen) exposure. Quantification of four individual experiments, -fold change to scr. Bar represents mean ± S.D.; ns, not significant; **, p <0.01; *, p <0.05.

Figure 2.

PHD3 silencing leads to down-regulation of HIF-2α protein expression in ccRCC cells. A, HIF-2α protein expression in 786-O cells transfected with control (scr), siPHD3#1 or siPHD3#2 followed by normoxic (NOX, 21% oxygen) or hypoxic (HOX, 1% oxygen) exposure. Quantification of five independent experiments, -fold change to scr. Bar represents mean ± S.D.; ***, p <0.001. B, HIF-2α protein expression in RCC4 cells transfected with control (scr), siPHD3#1, siPHD3#2 followed by normoxic (NOX, 21% oxygen) or hypoxic (HOX, 1% oxygen) exposure. Quantification of at least three independent experiments, -fold change to scr. Bar represents mean ± S.D.; **, p <0.01; ***, p <0.001. C, 786-O cells infected with adenoviral control (shCtrl) or PHD3-targeted shRNA (shPHD3) in normoxia. Quantification of five independent experiments, -fold change to scr. Bar represents mean ± S.D.; ***, p < 0.001. D, immunostaining of HIF-2α in 786-O cells transfected with control (scr) or siPHD3. Representative images are shown, scale bar, 100 μm. E, quantification of the HIF-2α immunostainings, mean intensity normalized to cell count. Bar represents mean ± S.D.; n = 7 fields of views from two individual experiments; **, p <0.01. F, primary head and neck squamous cell carcinoma cells (UT-SCC-34) or breast cancer cells T47D cells were transfected with scr or siPHD3#1 followed by hypoxic (HOX, 1% oxygen) exposure. Quantification of four (UT-SCC34) or three (T47D) independent experiments, bar represents mean ± S.D.; -fold change to scr; ns = not significant.

Figure 3.

Down-regulation of HIF-2α protein is specific to PHD3 silencing and independent from hydroxylase activity. A, 786-O cells transfected with control (scr), siPHD1, siPHD2, or siPHD3. No decrease in HIF-2α protein expression was observed in response to PHD1 or PHD2 depletion. Quantification of three independent experiments, -fold change to scr. Bar represents mean ± S.D.; ns = not significant. B, 786-O, RCC4, and RCC4+VHL cells were treated with panhydroxylase inhibitor DMOG for 8 h under normoxia or hypoxia followed by western blot analysis of HIF-2α expression. In VHL mutated 786-O and RCC4 cells DMOG has no effect on HIF-2α protein expression. Representative blots from two individual experiments. C, 786-O, RCC4, and RCC4+VHL cells were treated with cobalt chloride (CoCl₂) for 6 h under normoxia or hypoxia followed by western blot analysis of HIF-2α expression. Representative blots from four individual experiments. D, HIF-2α protein decay was studied using 4 h CHX treatment. HIF-2α protein decay remains unchanged with PHD3 depletion. Quantification of three individual experiments, individual data points are shown. -Fold change to untreated sample (0 h).

Figure 4.

PHD3 silencing leads to down-regulation of HIF2A mRNA expression. A, qRT-PCR analysis of HIF2A expression in 786-O cells transfected with control (scr), siPHD3#1, siPHD3#2, or siPHD2 followed by normoxic (NOX, 21% oxygen) or hypoxic (HOX, 1% oxygen) exposure. Quantification of at least three independent experiments, -fold change to scr. Bar represents mean ± S.D.; ***, p <0.001; **, p <0.001; ns = not significant. B, HIF2A expression in RCC4 cells transfected with scr, siPHD3#1, siPHD3#2, or siPHD2 followed by normoxic or hypoxic exposure. Quantification of at least three independent experiments, -fold change to scr. Bar represents mean ± S.D.; *** p <0.001; ** p <0.01; ns = not significant. C, HIF1A expression in RCC4 cells remains unchained with PHD3 or PHD2 silencing. Quantification of three (siPHD3) or two (siPHD2) individual experiments, -fold change to scr. Bar represents mean ± S.D. D, VEGFA expression in 786-O cells transfected with scr, siPHD3#1, siPHD3#2 followed by normoxic or hypoxic exposure. Quantification of four individual experiments, -fold change to scr. Bar represents mean ± S.D.; ** p <0.01; * p <0.05. E, VEGFA expression in RCC4 cells transfected with scr, siPHD3#1, siPHD3#2 followed by normoxic or hypoxic exposure. Quantification of four individual experiments, -fold change to scr. Bar represents mean ± S.D.; *, p < 0.05. F, OCT4 expression in 786-O cells. Quantification of four individual experiments, -fold change to scr. Bar represents mean ± S.D.; **, p <0.01; *, p <0.05.

Figure 5.

Forced expression of HIF2A in PHD3-silenced cells restore GLUT1 and LDHA expression. A, 786-O cells were transfected with scr or siPHD3 followed by forced expression of HIF2A or empty vector for 24 h. LDHA and GLUT1 expressions were determined with qRT-PCR. Quantification of four individual experiments, -fold change to scr+vector. Bar represents mean ± S.D. Statistics: one-way ANOVA; post hoc: Tukey, **, p <0.01; *, p <0.05. B, PHD3 and HIF2A expression in 786-O cells with PHD3 silencing and forced expression of HIF2A. Quantification of four individual experiments, -fold change to scr+vector. Bar represents mean ± S.D. Statistics: one-way ANOVA; post hoc: Tukey, ***, p <0.001; **, p <0.01. C, LDHA protein level is rescued by restoring expression of HIF2A in PHD3-depleted cells. A representative blot. D, Incucyte^® Live Cell Analysis of 786-O cells treated with two distinct siRNA sequences targeting PHD3 and HIF2A overexpression or empty vector. Forced expression of HIF2A restores the proliferation of PHD3-silenced cells. Curves represent mean values of three individual biological experiments with eight replicate wells in each experiment.

Figure 6.

PHD3 silencing leads to attenuated HIF-2α mRNA stability. A, qRT-PCR analysis of HIF2A expression in 786-O cells transfected with control (scr), siPHD3#1, or siPHD3#2 and treated with transcription inhibitor actinomycin D (Act D) for 4 h. Act D treatment further reduces the HIF2A expression in PHD3-depleted cells. Quantification of at least three independent experiments, -fold change to untreated scr. Bar represents mean ± S.D.; Statistics: one-way ANOVA; post hoc: Tukey, **, p <0.01; *, p <0.05; ns = not significant. B, 786-O cells were transfected with scr or siPHD3#1 and treated with Act D for 4 h. Samples were collected at indicated time points and HIF2A decay was determined by qRT-PCR. GAPDH expression was used for normalization. Graph represents mean values of three individual experiments ± S.E., -fold change to untreated sample (0 h). C, 786-O cells were transfected with scr, siPHD3#1, siPHD3#2 and treated with protein translation inhibitor CHX for 4 h. Treatment with CHX restores the expression of HIF2A mRNA. Quantification of at least three independent experiments, -fold change to untreated scr. Bar represents mean ± S.D., Statistics: one-way ANOVA; post hoc: Tukey, **, p <0.01, *, p <0.05, ns = not significant.

Figure 7.

PHD3 and HIF2A mRNA expression correlate in clinical ccRCC tumor samples. A, distribution of HIF1A, HIF2A, and PHD1–3 expression levels across tumor (n = 71) and their adjacent normal (n = 71) samples. Expression levels (normalized, log transformed) of both HIF1A and HIF2A in addition to PHD2 and PHD3 were found significantly different between tumor and normal samples (p <0.001). B, hierarchical clustering of TCGA ccRCC VHL mutated (n = 217) and nonmutated (n = 194) samples based on the expression levels of HIF1A, HIF2A and PHD1–3 genes. C, correlation coefficients (Pearson correlation) with corresponding p values for HIF1A or HIF2A and PHD1–3 in the whole ccRCC tumor sample data set (n = 442). Correlation of expression between HIF2A and PHD3 was higher than the correlation between other gene pairs (Pearson correlation 0.196, p <0.001). D, correlation of the expression levels (normalized, log transformed) between HIF2A and PHD3 for the patients with the poorest survival (n = 40). A significant correlation of 0.526 was observed (Pearson correlation, p <0.001).