Renal cell carcinomas (RCCs) are highly hypoxic tumors characterized by dysregulated, arborizing vasculature. This aberrant vasculature is typically attributed to the hyperactive signaling from hypoxia-inducible factors (HIFs) common in these tumors. The primary regulator of HIFs, the protein product of the von Hippel Lindau (pVHL) gene, is commonly mutated in RCCs. While HIFs are responsible for activating a myriad of proangiogenic genes, the tumor suppressor p53 is key for the transcription of many antiangiogenic genes.
The gene that encodes p53, TP53, is rarely mutated in RCCs, suggesting that there must be another mechanism leading to its inactivation. It has previously been shown that pVHL binds p53, which increases p53 stability and activity in response to genotoxic stress. A protein that downregulates p53 is the oncoprotein murine double minute 2 (Mdm2), which is activated in response to genotoxic stress. We and others have shown that Mdm2 is detected in many tumor types and is correlated with high-grade, metastatic tumors. While the downregulation of p53 by Mdm2 is typically considered to be the primary function of Mdm2, our group has shown that Mdm2 can function independently of p53 to regulate the early stages of metastasis as well as increase HIFs. Collectively these observations raised the question: does Mdm2 have an indirect effect on the activity of p53 in the hypoxic microenvironment of RCCs?
Our work shows there is a complex interplay between pVHL and Mdm2. We found that Mdm2 will bind directly to pVHL and conjugate a ubiquitin-like polypeptide, a protein called neural precursor cell expressed developmentally downregulated 8 (nedd8), under hypoxic conditions. This is the first time anyone has shown that Mdm2 can neddylate pVHL and identifies another tumor suppressor that Mdm2 regulates. The lysine that Mdm2 conjugates nedd8 to is in the region required for binding to p53 and therefore prevents the formation of a p53-pVHL complex. We also show that the binding of pVHL to p53 is needed for the activation of antiangiogenic factors in response to hypoxia. By blocking the neddylation of pVHL (genetically or pharmacologically with pevonedistat/MLN4924), the antiangiogenic factors can be induced which requires wild type p53. Re-engaging the antiangiogenic pathway diminished angiogenesis. Interestingly, using the small molecule Nutlin3, which targets the binding of Mdm2 to p53, increased p53 and Mdm2, but induction of antiangiogenic genes was absent. These data demonstrate that p53 activity is highly regulated through interaction with other proteins to induce specific genes that regulate a myriad of cellular functions. Thus, it cannot be assumed that increased p53 will retain ability to activate all target genes. Our overall finding is that under conditions that induce the neddylating activity of Mdm2, there is an indirect impact on p53 to promote angiogenesis through Mdm2 regulation of pVHL.