Differential effects of HIF2α antagonist and HIF2α silencing in renal cancer and sensitivity to repurposed drugs

Abstract

Background: In clear cell renal cell carcinoma, 80% of cases have biallelic inactivation of the VHL gene, leading to constitutive activation of both HIF1α and HIF2α. As HIF2α is the driver of the disease promoting tumour growth and metastasis, drugs targeting HIF2α have been developed. However, resistance is common, therefore new therapies are needed.

Methods: We assessed the effect of the HIF2α antagonist PT2385 in several steps of tumour development and performed RNAseq to identify genes differentially expressed upon treatment. A drug screening was used to identify drugs with antiproliferative effects on VHL-mutated HIF2α-expressing cells and could increase effectiveness of PT2385.

Results: PT2385 did not reduce cell proliferation or clonogenicity but, in contrast to the genetic silencing of HIF2α, it reduced in vitro cell invasion. Many HIF-inducible genes were down-regulated upon PT2385 treatment, whereas some genes involved in cell migration or extracellular matrix were up-regulated. HIF2α was associated with resistance to statins, addition to PT2385 did not increase the sensitivity.

Conclusions: this study shows key differences between inhibiting a target versus knockdown, which are potentially targetable.

Keywords: Drug resistance; EMT; HIF2α; Renal cancer.

Fig. 1

786–0 and RCC4 cell migration. A) Representative images of 786–0 WT and 786–0 WT + 1 μM PT2385 (PT) cells migrating 12 h after the scratch and wound confluence and wound width measure until the complete closure of the wound. n = 3. B) Representative images of RCC4 WT, RCC4 WT + 1 μM PT2385 (PT) and RCC4 VHL cells migrating 24 h after the scratch and wound confluence and wound width measure until the complete closure of the wound. n = 3. * represents comparison between RCC4 WT and RCC4 VHL and # represents comparison between RCC4 WT and RCC4 WT + 1 μM PT2385. Purple region corresponds to the initial scratch mask, whereas blue region represents the wound lacking cells at the compared moment. * p < 0.05, ** p < 0.01, *** p < 0.001. Errors bars depict standard error of the mean

Fig. 2

786–0 and RCC4 cell invasion. A) Representative images of 786–0 WT and 786–0 WT + 1/10 μM PT2385 (PT) cells invading 48 h after the scratch and RWD measurement until 96 h after wound performing. n = 3. * represents comparison between 786-0 WT and 786–0 WT + 1 μM PT2385 and # represents comparison between 786-0 WT and 786–0 WT + 10 μM PT2385. B) Representative images of RCC4 WT, RCC4 WT + 1 μM PT2385 (PT) and RCC4 VHL cells invading the wound 48 h after the scratch and RWD measurement until 96 h after wound performing. Purple region corresponds to the initial scratch mask, whereas blue region represents the wound lacking cells at the compared moment. n = 3. * p < 0.05, ** p < 0.01, *** p < 0.001. Errors bars depict standard error of the mean

Fig. 3

RNAseq results. Heatmap of the top 50 variant significant genes in 786–0 WT cells vs PT2385 treatment

Fig. 4

Drug screening of 786–0 WT cells. A) Diagram showing the drug screening. B) Viability of 786–0 WT and 786–0 WT cells transfected with siCON or siHIF2α and treated with 2.5 μM statins or 50 μM terbinafine +/− 10 μM PT2385 (PT) during 5 days relative to cells treated with DMSO. n = 3. * p < 0.05, ** p < 0.01, *** p < 0.001. Errors bars depict standard error of the mean