Targeting p300/CBP destabilizes HIF-2α and suppresses tumor growth in cell and animal models of clear cell renal cell carcinoma

1. Introduction

Clear cell renal cell carcinoma (ccRCC) is the most common (75–80%) subtype of RCC followed by papillary RCC (7–14%) and chromophobe RCC (6–11%) [1]. Inactivation of the von Hippel-Lindau (VHL) gene is a hallmark of ccRCC, occurring in up to 90% of sporadic and familial cases [2]. VHL directly interacts with hypoxia-inducible factors (HIFs), facilitating polyubiquitination and subsequent proteasomal degradation of their alpha subunits. Loss of VHL function therefore results in elevated expression of HIF-1α and HIF-2α. While HIF-1α generally acts as a tumor suppressor, HIF-2α functions as a potent oncogene and is a key driver of renal tumorigenesis [3]. Several HIF-2α inhibitors are currently under clinical investigation for the treatment of advanced ccRCC, including NKT2152 (NCT05119335, NCT05935748), DFF332 (NCT04895748) and AB521 (NCT05891171, NCT03629756). On December 14, 2023, the U.S. FDA approved Belzutifan (Welireg) for patients with advanced RCC whose disease progressed after treatment with VEGF-TKIs and PD-1/PD-L1 inhibitors.

p300 (EP300) and CBP (CREBBP) are paralogous lysine acetyltransferases involved in post-translational modifications (PTMs) of both histone and non-histone proteins, as well as co-activation of various transcription factors [4]. p300/CBP are frequently overexpressed in cancer cells, where they support oncogene transcription, proliferation, immune evasion and drug-resistance across multiple solid and hematological malignancies [5]. CCS1477 (Inobrodib), a first in class, selective p300/CBP bromodomain inhibitor, is currently under evaluation in phase 1 and 2 clinical trials involving patients with advanced solid/metastatic tumors including castration-resistant prostate cancer, breast cancer and non-small cell lung cancer (NCT03568656) as well as hematological malignancies (NCT04068597 and NCT07096778) as a monotherapy and in combination with other agents.

Our studies demonstrate that pharmacological or genetic inhibition of p300/CBP reduces HIF-2α protein levels in cell and animal models of ccRCC. Moreover, CCS1477 significantly suppresses ccRCC xenograft growth, and its combination with cabozantinib induces complete tumor regression.

2. Materials and Methods

Materials and methods are described in the Supplementary Information.

3. Results and Discussion

3.1. Inhibition of p300/CBP reduces expression of HIF-2α protein in ccRCC cells.

Given the clinical relevance of p300/CBP inhibition as an emerging therapeutic strategy in cancer treatment, and the established role of p300/CBP in regulating hypoxia-inducible transcriptional programs, we examined whether treatment with CCS1477, a clinical-stage bromodomain p300/CBP inhibitor, affects HIF-2α protein expression in ccRCC cells. As shown in Figure 1A, treatment with CCS1477 markedly decreased expression of HIF-2α protein in all tested ccRCC cell lines, i.e. SK-RC-45 (SK-45), RCC4, and 786-O. In contrast, HIF-2α mRNA levels were enhanced following treatment with CCS1477 (Fig. 1B). Similar reduction in HIF-2α protein expression, accompanied by enhanced HIF-2α mRNA levels, was observed following treatment with A-485 and C646, inhibitors targeting the catalytic activity of p300/CBP (Fig. 1C–D). To corroborate the role of p300/CBP in maintaining HIF-2α homeostasis in ccRCC cells, we performed siRNA-mediated knockdown of p300 and CBP in ccRCC cells. Consistent with the pharmacological inhibition data, HIF-2α protein, but not mRNA levels, were significantly reduced following either individual or combined knockdown of p300 and CBP (Fig. 1E–F). An observed increase in HIF-2α mRNA may potentially serve as a compensatory mechanism reflecting the cellular attempt to restore HIF-2α protein expression. HIF-2α is a well-established player in resistance chemo- and targeted therapeutic agents in a range of tumor types [6]. Treatment with cabozantinib, a standard-of-care tyrosine kinase inhibitor (TKI) used in both first- and second-line treatment settings in ccRCC, notably induced HIF-2α protein expression in ccRCC cells that was almost completely blocked by CCS1477 (Fig. 1G). Importantly, co-treatment with the proteasome inhibitor bortezomib restored HIF-2α protein levels in cells exposed to CCS1477 (Fig. 1H), suggesting that p300/CBP inhibition promotes HIF-2α degradation via the proteasome-mediated mechanism. Consistent with this, co-treatment with CCS1477 accelerated HIF-2α protein degradation in the presence of cycloheximide, a protein synthesis inhibitor, collectively indicating that p300/CBP regulates HIF-2α abundance at the level of post-translational stability (Fig. 1I). Next, we examined whether CCS1477 affects HIF-2α acetylation and found that treatment with CCS1477 notably reduced this modification (Figure 1J), suggesting that p300/CBP-mediated acetylation may stabilize HIF-2α against proteasomal degradation.

Figure 1. Inhibition of p300/CBP reduces HIF-2α protein levels in ccRCC cells.

(A) Western blot and (B) qRT-PCR analyses of HIF-2α expression in ccRCC cells treated with CCS1477 (10 μM) for 24 hr. Results are expressed as the mean (n=3)±SD. * P<0.05; ** P<0.001; *** P<0.0001 (compared to medium). (C) Western blot and (D) qRT-PCR analyses of HIF-2α expression in ccRCC cells treated with A-485 (25 μM) or C646 (10 μM) for 24 hr. Results are expressed as the mean (n=3)±SD. * P<0.05; *** P<0.0001 (compared to medium). (E) Western blot and (F) qRT-PCR analyses of p300, CBP and HIF-2α expression in SK-45 and RCC4 cells following individual or combined siRNA-mediated knockdown of p300 and CBP. Results are expressed as the mean (n=3)±SD. ** P<0.001; *** P<0.0001 (compared to medium). (G) Western blot analysis of HIF-2α expression in SK-45 and RCC4 cells treated with cabozantinib (2.5μM) with or without CCS1477 (10μM) for 24 h. (H) Western blot analysis of HIF-2α expression in SK-45 cells treated with CCS1477 (10 μM) with or without bortezomib (50 nM) for 24 hr. (I) Western blot analysis of HIF-2α expression in SK-45 cells treated with cycloheximide (CHX) (50 μg/ml) with or without CCS1477 (10 μM) for the indicated periods of time. (J) Analysis of HIF-2α acetylation in 786-O cells. Cells were cultured in medium alone or in the presence of CCS1477 (10 μM for 12 hr). MG-132 (1 μM) was added to both cell cultures to prevent proteasomal degradation of HIF-2α. HIF-2α was immunoprecipitated from cell lysates using an anti-HIF-2α antibody, and acetylation was assessed by immunoblotting with an anti-acetyl-lysine antibody.

3.2. CCS1477 suppresses ccRCC xenograft tumor growth and exhibits synergy with cabozantinib.

To extend our in vitro findings to an in vivo setting, PNX0010 xenograft tumors were established in NSG mice. The PNX0010 cell line was derived from a lung metastatic lesion of a ccRCC patient who underwent nephron-sparing surgery at our institution, and it represents an aggressive, VHL-deficient variant of ccRCC [7]. As shown in Figure 2A, treatment with either CCS1477 or cabozantinib significantly reduced PNX0010 xenograft tumor growth, while the combination of CCS1477 and cabozantinib resulted in complete tumor regression. All treatment regimens were well tolerated by the experimental animals, with no notable weight loss observed (Fig. 2B). Consistent with our in vitro findings, cabozantinib treatment led to increased expression of HIF-2α protein in xenograft tumors, whereas CCS1477 treatment markedly reduced its expression (Fig. 2C–D). Future studies are needed to delineate the downstream signaling pathways by which p300/CBP inhibition enhances the antitumor activity of cabozantinib, identify potential mechanisms of acquired resistance, and evaluate the durability of tumor responses to combination therapy. Although CCS1477 monotherapy produced a greater reduction in HIF-2α protein levels than the combination treatment, the CCS1477/cabozantinib combination achieved superior antitumor efficacy, suggesting that CCS1477-mediated HIF-2α downregulation alone does not fully account for the enhanced therapeutic response. In addition to suppressing HIF-2α, the complementary inhibition of VEGFR, MET, AXL, and other receptor tyrosine kinase signaling pathways by cabozantinib is likely to contribute to the enhanced antitumor activity. Although p300/CBP inhibition may affect multiple cellular pathways beyond HIF-2α, HIF-2α depletion alone was sufficient to almost completely block PNX0010 xenograft tumor growth (Figure 2E and F). The more pronounced inhibitory effect of genetic HIF-2α knockout compared to CCS1477 treatment on xenograft growth likely reflects the difference between complete, constitutive loss of HIF-2α from the outset of tumor formation versus partial, pharmacologic suppression of HIF-2α protein levels in an established tumor. Additionally, CCS1477 may achieve incomplete target engagement due to pharmacokinetic and dosing limitations, whereas genetic knockout eliminates HIF-2α protein entirely and irreversibly.

Figure 2. CCS1477 suppresses ccRCC xenograft tumor growth and exhibits a synergistic effect with cabozantinib treatment.

(A) The effect of treatment with CCS1477 and cabozantinib on the growth of PNX0010 xenograft tumors in NSG mice. Values shown represent means (n=5) ± SD. (B) Percentage of body weight change in mice carrying PNX0010 xenografts during the treatment period. Values shown represent means (n=5) ± (magnification: 400x). (D) H-score quantitation of HIF-2α staining in the xenograft tumor specimens. Slides were scanned using Vectra 3 Automated Quantitative Pathology Imaging System (Akoya Biosciences). Images were analyzed using inForm Automated Image Analysis Software (Akoya Biosciences). (E) Western blot analyses of HIF-2α expression in wild-type PNX0010 cells (WT) and HIF-2α knockout PNX0010 cells (KO). (F) The effect of HIF-2α knockout on the growth of PNX0010 xenograft tumors in NSG mice. Values shown represent means (n=7) ± SD. (G) Representative images of WT and HIF-2α-KO PNX0010 xenograft tumors from the experiment shown in panel F stained with an anti-HIF-2α antibody (magnification, 400×).

HIF-2α inhibitors, such as belzutifan, selectively suppress HIF-2α transcriptional activity by disrupting HIF-2α /ARNT dimerization and have demonstrated significant clinical benefit in patients with ccRCC. However, tumors can ultimately acquire resistance to HIF-2α inhibitors through mutations in the HIF-2α drug-binding pocket [8]. This underscores the need for alternative therapeutic strategies targeting this oncogenic driver. In contrast, CCS1477 suppresses HIF-2α by reducing its protein expression, representing a distinct therapeutic strategy. This approach may retain activity in tumors harboring resistance to direct HIF-2α inhibitors. Although our findings identify HIF-2α as an important downstream target of p300/CBP inhibition in ccRCC, p300/CBP regulates multiple transcriptional programs, and additional pathways are likely to contribute to the antitumor activity of CCS1477. Defining the relative contribution of HIF-2α and other p300/CBP-regulated pathways to the therapeutic effects of p300/CBP inhibition remains an important direction for future investigation.

Acetylation of HIF-2α by CBP facilitates the formation of CBP/HIF-2α complexes and is essential for the induction of HIF-2α target genes [9, 10]. The importance of this modification is further underscored by recent studies showing that inhibition of acetyl-CoA synthetase 2 (ACSS2) reduces HIF-2α expression and stability by promoting its proteasomal degradation [11]. Our data show that the p300/CBP bromodomain inhibitor CCS1477, as well as inhibitors of p300/CBP lysine acetyltransferase activity, A-485 and C646, reduce HIF-2α protein expression in ccRCC cells. Based on these findings, we propose the following model: HIF-2α is acetylated by p300/CBP, and once acetylated, HIF-2α interacts with the bromodomain of p300/CBP, thereby preventing its proteasomal degradation. Indeed, there is substantial evidence that binding to p300/CBP can stabilize various proteins and protect them from proteasomal degradation. For instance, binding to p300/CBP stabilizes p53 protein [12]. This suggests two potential therapeutic approaches to destabilize HIF-2α: targeting either its acetylation by p300/CBP or its interaction with the p300/CBP bromodomain. Although p300 and CBP share extensive structural and functional homology, they may have both distinct and overlapping roles in ccRCC.

Recent studies provide rationale for targeting both epigenetic writers (p300/CBP) and readers (BET proteins) [13]. Due to the homology of the bromodomains present in p300/CBP and BET proteins, small molecules with dual inhibitory activities, such as NEO2734 and XP-524, were designed. CCS1477 is highly specific for the p300/CBP bromodomains, therefore it could provide some advantages over dual p300/CBP and BET bromodomain inhibitors if the BET pathway is not a primary therapeutic target. Also, combining two inhibitory mechanisms could lead to a narrower therapeutic window due to enhanced off-target effects associated with inhibiting multiple pathways. Yet, the dual inhibitory mechanism may help to overcome resistance to single-target inhibitors [13].

Collectively, these studies underscore the importance of p300 and CBP in renal tumorigenesis and provide a molecular rationale for investigating the therapeutic synergy between p300/CBP inhibitors and targeted therapeutics in patients with ccRCC.

Highlights.

• p300/CBP inhibition reduces HIF-2α via proteasome-mediated degradation.

• HIF-2α depletion alone is sufficient to block ccRCC xenograft tumor growth.

• CCS1477 plus cabozantinib induces complete regression of ccRCC xenografts.

Data availability

The original data included in this study are presented in this article. Further inquiries can be directed to the corresponding author.