AACR 2021 | Mutant p53 and oncogenic KRAS drive metastasis in pancreatic cancer

The mutational landscape of pancreatic cancer is fairly bland, and the two mountains within that landscape are oncogenic KRAS and p53. Both of those mutations co-occur at a very high rate, about 60%, 70% in all of pancreatic cancer patients, and they’re recognized as the top genetic drivers. So our goal was to figure out, given that high incidence of co-occurrence and the fact that the mutant form of p53 is more prevalent than the null or loss form of p53, is there some biologic advantage conferred when you have mutant p53 in the setting of oncogenic KRAS, relative to that simple loss of p53?

Since mutant p53 is the most commonly altered tumor suppressor gene in human cancer, period, we hypothesized that the mutant forms of these proteins, which tumor suppressors are generally lost, but mutant p53 generates abnormal physical mutant proteins. And so it’s been known for some time, through the work of many others, that those mutant p53 proteins interact with other proteins within the cell and confer oncogenic properties, such as increased migration, invasion, metastasis, proliferation. The list is extensive. And so, we hypothesize that was somehow working with oncogenic KRAS to confer increased fitness traits to pancreatic cancer cells that allows them to metastasize better, more efficiently.

We generated a mouse model system, a new mouse model, a KBC mouse model, that had somatic mutant p53. We compared it to the same mouse model that had deletion of p53, and then we identified transcription factors that mutant p53 was associated with. We went to our human patient-derived xenograft systems and repeated that study, this time with p53 null and mutant p53 high, and also got a list of transcription factors. And then we took the overlap, which was on a protein named FOXA1. FOXA1 has been implicated in metastasis, more so in the last couple of years, and even in pancreatic cancer metastasis, so that got our attention because our mutant p53 model had more metastasis relative to the loss of p53 model.

From there, we did some deep science to confirm that that protein FOXA1 was regulated by mutant p53, and it was doing it through a separate protein called CREB1. The unique thing about CREB1 is it generally has to get activated in order for mutant p53 to bind to it. What activates CREB1? A lot of KRAS effectors. So that got our attention. Then, we’re faced with a situation where there’s this protein that mutant p53 binds to, to upregulate FOXA1, is activated by a lot of the mitogenic genes that KRAS activates.

So then through that level of inquiry, we confirmed that oncogenic KRAS effectors phosphorylate and activate CREB1 and allow mutant p53 to bind to it, which then go and upregulate FOXA1 transcriptional networks and Wnt/beta-catenin signaling as well. It’s that multiplexed activation of transcriptional networks in a KRAS-dependent manner, because KRAS is enabling the mutant p53 gain of function to occur. And so that’s the deep mechanistic work we did.

We also found the one CREB inhibitor we can identify, and that was previously published, and found that when we treated pancreatic cancer cells with that drug, Wnt/beta-catenin signaling went away down, FOXA1 went way down, CREB phosphorylation went away down, and that’s also what we saw when we silenced KRAS. So a lot of those things fit. Fundamentally, we identify CREB1 as this signaling node between oncogenic KRAS and mutant p53, and as the link between those two axes that’s targetable with this drug.