Read the Full Video Transcript

Andrea Miyahira: Hi, I'm Andrea Miyahira at the Prostate Cancer Foundation. Today, I'm joined by Dr. Shaghayegh Nouruzi of the University of British Columbia and Vancouver Prostate Center to discuss her recent paper, ASXL1 regulates and cooperates with FOXA2 to drive terminal neuroendocrine phenotype in prostate cancer, published in JCI Insight. Dr. Nouruzi, thank you for joining.

Shaghayegh Nouruzi: Thank you, Dr. Miyahira, for your invitation today. I'm very excited to present our work to your audience today. Yeah. So today, I'm going to talk about three points: the stage-specific role of transcription factors, new potential therapeutic targets in metastatic CRPC, and the broader implication of our findings.

So what we're interested in studying is this metastable, highly plastic state that we see emerge after targeted therapies directed at AR. This lineage plasticity allows the tumor to shift between different phenotypes. There are very limited mutations compared to adenocarcinoma, and this lineage plasticity can develop very fast in response to AR pathway inhibitor.

And this is something that we see in the clinic. Data from the [INAUDIBLE] trial, for example, showed that high-risk localized prostate cancer patients treated with neoadjuvant AR pathway inhibitors can develop lineage plasticity as early as three months. So all these data together suggest that this is an epigenetically driven state. Eventually, we see that some tumor cells can acquire other characteristics, for example, neuronal characteristics, and lose the AR expression.

So in the metaplastic state, often AR remains expressed while we see an increase in stem cell and neuronal programs. But later on, we see subclones or populations that commit to these terminal lineages. And we're interested in studying these different states because they do have different transcriptomes and epigenomes.

But understanding what their underlying molecular mechanisms are is important because it can influence treatment strategies and decision making. So we have these models in our lab to study these different stages. And we see that they can reflect what we see in the clinic.

So in our previous work published in 2022, we identified ASCL1, which is a neuronal transcription factor, as a driver of lineage plasticity and neuronal differentiation. And at that time, we were interested in studying what ASCL1 does in prostate cancer, regardless of the disease stage.

And we saw that it regulates the PRC2 complex and EZH2 activity. But what came out of that work that was really interesting was that in this metastable plastic state, where AR is still expressed, if we target ASCL1, not only do we lose neuronal characteristics, but we reactivate an AR-driven luminal signaling. And this has been since validated and shown by others.

So this suggests that lineage plasticity is dynamic and reversible. And in this metaplastic state, we have this opportunity to reverse the phenotype and potentially resensitize tumors to further AR inhibition. But this strategy won't work if the tumor cells are already committed to a terminal phenotype and lost their expression. So we need to think of what else we can do.

So in our recent work, we focused on studying what ASCL1 does specifically in the terminal stage. How does it work differently? And we see ASCL1 actually does have unique binding sites and drives unique programs in this setting. One of the transcription factors that we followed up on was FOXA2. So not only does ASCL1 regulate FOXA2, but we actually saw that they form a complex together.

And they co-bound to these specific genomic loci or regions that are marked by an active histone mark within an accessible chromatin region and are transcriptionally active only in this terminal setting and not earlier in the disease. So there are these unique programs and unique complexes that only become activated and form in this late stage, which are not present earlier in the disease.

And that led us to identify PROX1 as their downstream target, which is another transcription factor. And we show that if we target PROX1 in NEPC, we can influence downregulation of neuronal characteristics and neuronal phenotype, but it also results in reduced cell proliferation and tumor growth, suggesting that it can be a potential target in this setting.

So to summarize, our work showed that ASCL1 can have a specific role early on—it can influence lineage plasticity and neuronal differentiation—and then later on cooperates with other factors, such as FOXA2, to maintain the neuronal phenotype. And we identified PROX1 as a driver, as well as a potential therapeutic target for NEPC.

So the take-home message is that there are powerful transcription factors, such as ASCL1, that can influence lineage decision-making. And even though they may be expressed at different stages of the disease, they will have distinct roles in this setting. And why it is important is because if we then target ASCL1 at different stages, it will have different outcomes.

It also means that ASCL1 or this transcription factor might regulate distinct programs, so these different tumor states will have different vulnerabilities. And thinking of it therapeutically means that if we have one strategy, it won't be enough, so we can think about different stages. We'll need a unique approach.

And this is important if we want to have precise and more effective treatment. So in the metaplastic state, we have the opportunity to leverage the reversible nature of lineage plasticity, restore AR activity, and resensitize to AR inhibition, but later on, we need to think about different strategies: disrupt transcriptional complexes that are in this setting or inhibit drivers that are important for neuronal characteristics or proliferation, such as what we identified here, which was PROX1.

And with that, I want to thank my mentor, Amina Zoubeidi, and the Zoubeidi lab members, but most importantly, the PCF and the donors who supported my PCF Young Investigator Award. And if you have questions, you can reach me.

Andrea Miyahira: Thank you so much, Dr. Nouruzi, for sharing that. So based on this data, are ASCL1, FOXA2, or PROX1 possible therapeutic targets in MPC? Which do you think has the best properties for inhibition?

Shaghayegh Nouruzi: So yes, based on our work, I think all three are very good potential targets. But we need to also think about different stages of the disease. ASCL1 is something that comes earlier during lineage plasticity and is there during the transition to the terminal phenotype, while FOXA2 and PROX1 are later genes and markers that come.

But targeting transcription factors is challenging. It's not impossible. So there are possibilities in the sense that ASCL1 forms homo- or heterodimers; it forms complexes, like the one with FOXA2 that we identified. There is potential to have small molecules disrupting these interactions and their programs. But what I think we're also focused on is to identify vulnerabilities that are associated with these factors and complexes, and then indirectly target them.

Andrea Miyahira: And as you showed, ASCL1 expression is induced by AR pathway inhibitors. So how would you envision treating patients with NEPC considering that many have heterogeneous diseases, including AR-driven subclones?

Shaghayegh Nouruzi: Yeah. So I think this is where the power of combination therapy comes in. If we think about earlier in the disease, in the metaplastic state, if we combine an isolation inhibitor with an AR pathway inhibitor, we have the opportunity to stop or reverse the tumors that are going through lineage plasticity, while at the same time targeting the ones that are driven by AR.

And then there are other factors that were described by other groups, like combining TET2 with an AR pathway inhibitor, or the work that is coming from Dr. Li and Dr. Nelson's lab looking at ADCs to target these different surface markers from different subclones.

And studying these different settings, what we do is then identify targets from these specific subclones, and then combine them to tackle this heterogeneous nature of the disease.

Andrea Miyahira: OK, thank you. And is it known what other genes are regulated by ASCL1 and FOXA2 that drive the lethal NEPC phenotype?

Shaghayegh Nouruzi: Yeah. So in this work, we specifically focus on transcription factors that are downstream targets of these two, ASCL1 and FOXA2. But this list is very extensive. It has kinases, epigenetic regulators, and chromatin remodelers.

And one of them that is notable is NSD2, which is a methyltransferase. It's important in adenocarcinoma as well as lineage plasticity. Dr. Michael Shen showed that inhibition of NSD2 can also reverse lineage plasticity. And Dr. Chinnaiyan is developing NSD2 PROTACs. And there's actually an NSD2 inhibitor in clinical trials for multiple myeloma. So these are things that we can follow up with and look further into.

Andrea Miyahira: And what are your next steps for your studies?

Shaghayegh Nouruzi: So we're looking then into, as I said, understanding how we can indirectly target these programs. So all these hits that came out from our list, such as NSD2 or other vulnerabilities that they would have, we want to study further. We want to think about drugs that are already available or repurpose them—things that are already in trials for other cancers—and see in which setting they will be most beneficial.

Andrea Miyahira: OK. Thank you so much, Dr. Nouruzi, for sharing this study with us.

Shaghayegh Nouruzi: Thank you for having me.