Varadha Balaji Venkadakrishnan: Thanks for having me. To provide some context, androgen receptor signaling has been the key target in advanced prostate cancer, acting as the engine driving this aggressive disease. However, prostate cancer can acquire resistance by essentially pulling into the pit stop and re-engineering itself. In up to 15% of cases, it undergoes a complete identity change, losing androgen receptor expression entirely. This process, known as lineage plasticity, involves prostate adenocarcinoma histologically transforming into small-cell-like neuroendocrine prostate cancer, or NEPC. NEPC is associated with poor prognosis, marked by the loss of androgen receptor expression. Unfortunately, there are currently no approved targeted therapies for NEPC. This lineage transformation is accompanied by significant epigenetic reprogramming. My role as a postdoc in Himisha's lab has been focused on understanding how epigenetic modulators drive NEPC transition, and whether these modulators can be exploited as therapeutic targets. Today, I'll highlight an example of how we can harness epigenetic dysregulation by targeting genes that are epigenetically derepressed in NEPC. In our previous work published last year, I found that repressive histone marks are distinctly distributed in NEPC versus prostate adenocarcinoma.
Specifically, these marks silence neuroendocrine lineage genes in prostate adenocarcinoma and luminal lineage genes in NEPC. Building on this work, we asked, what are the key drivers that are epigenetically derepressed during the NEPC transition, and can they be therapeutically exploited? One such driver that is derepressed is PROX1, or prospero homeobox 1. It is a transcription factor that plays a critical role in endothelial cell differentiation, and recent studies have implicated a role for PROX1 in NEPC. We found that NEPC exhibits a high dependency on PROX1 expression. In fact, PROX1 emerged as one of the highest top hits in CRISPR screens we performed using two independent NEPC organoid models. This dependency was validated in vivo, where PROX1 knockout significantly reduced tumor growth. Interestingly, when I overexpressed PROX1 in 22Rv1 prostate adenocarcinoma model and implanted the cells subcutaneously into castrated mice, I observed spontaneous metastasis to the liver, and this also was accompanied with significantly increased tumor growth. Pathologically, these tumors were poorly differentiated, with acquisition of neuroendocrine markers such as chromogranin A, further supporting PROX1's role in driving NEPC progression. At the transcriptomic level, PROX1 knockout in NEPC organoid models significantly reduced the expression of other neuroendocrine drivers. Using a recombinant degron tag system, I was able to dynamically degrade PROX1 and identify its direct targets.
Short-term degradation of PROX1 combined with PROX1 CUT&RUN revealed its recruitment to the promoters of other neuroendocrine lineage transcription factors, validating its regulation of neuroendocrine lineage program. To explore strategies for targeting PROX1 in NEPC, I immunoprecipitated PROX1 and subjected it to mass spectrometry, which revealed previously unreported phosphorylation sites on its DNA-binding domain. I generated phospho-null and phospho-mimetic mutations, and found that serine-719 (Ser719) residue significantly impacted PROX1 stability and function. Using in silico library analysis, I predicted CHK1 and CDK2 as potential upstream kinases, and my initial experiments suggest that targeting these kinases could block PROX1 activity in NEPC. So overall, this is just one example of how we can leverage epigenetic dysregulation to identify and target derepressed drivers in NEPC. With that, I would like to thank my mentor, Dr. Himisha Beltran, for her support in this study, as well as my lab members, collaborators, and funding sources, including DoD and the National Cancer Center, and especially Prostate Cancer Foundation. Thank you for your attention, and thank you, Andrea, for hosting me on this digital platform.
Andrea Miyahira: Well, thank you so much for coming on and sharing this with us today. So there have been several NEPC subtypes described. Which subtypes have you seen a role for PROX1 in?
Varadha Balaji Venkadakrishnan: Yeah, that's a great question. PROX1 expression correlated with expression of other neuroendocrine drivers, like ASCL1, INSM1, and neuroendocrine markers, like chromogranin A. When we reviewed our RNA-seq data from clinical samples, PROX1 appears to be ubiquitously expressed in all NEPCs that are histologically classified as transformed small-cell phenotype. So it may play a broad role across NEPC subtypes.
Andrea Miyahira: Okay, thank you. And what is the normal role of PROX1? Is it expressed in any adult tissues? And what side effects might be anticipated by targeting it?
Varadha Balaji Venkadakrishnan: Yeah, that's a really important question. PROX1 is critical for cell fate specification, particularly in lymphatic endothelial cells. And during development, it plays an essential role, and complete knockout of PROX1 is embryonically lethal. However, tissue-specific conditional knockouts of PROX1 are viable, which is encouraging. NEPC shows high expression of PROX1, and in adult tissues, PROX1 is expressed in liver and brain, so systemic targeting of PROX1 would need to be carefully evaluated to avoid any off-target effects. So I think this is an area that requires further exploration.
Andrea Miyahira: Okay. You identified three different phosphorylation sites on PROX1, each with a different candidate kinase. What do you think this means in terms of how and when PROX1 is activated, and how to optimally target PROX1? And are all of these phosphorylation sites and kinases required for its function?
Varadha Balaji Venkadakrishnan: Yeah, again, great question. Our initial experiments show that mutating these phosphorylation sites impact PROX1 expression. In particular, mutating the serine-719 site had a profound effect on PROX1 expression and its downstream target genes. We don't yet fully understand how and when these phosphorylation events occur, and it's likely that they are regulated by different upstream signals. So more work is needed to determine whether all three phosphorylation sites are required for PROX1 activation or if one site plays a dominant role. This is quite an exciting area for further investigation, especially as we think about therapeutic strategies.
Andrea Miyahira: Thanks. And what are your next translational steps, and also your thoughts on pursuing further studies on repurposing prexasertib or tegtociclib in NEPC?
Varadha Balaji Venkadakrishnan: Yeah, thank you for that question. I have established reporter models, and currently working on identifying endogenous degraders of PROX1 to develop potentially a PROTAC or a molecular glue. As for repurposing prexasertib or tegtociclib in NEPC, I think it's a promising idea, but further mechanistic studies are definitely needed. We have previously shown that CDK2 inhibition is a favorable therapeutic strategy, so exploring these compounds in the context of NEPC could be worthwhile.
Andrea Miyahira: Thanks. And you also mentioned that PROX1 expression was associated with increased liver metastases. Do you have any further thoughts on this?
Varadha Balaji Venkadakrishnan: Yes, that was a very interesting observation. When we overexpressed PROX1 in prostate adenocarcinoma model and implanted them subcutaneously, we saw spontaneous metastasis to the liver, along with the acquisition of neuroendocrine features. So this is quite rare, because subcutaneous implantation of prostate cancer cell lines or organoids typically does not result in spontaneous metastasis. This finding suggests that PROX1 overexpression drives clinically aggressive features, consistent with NEPC favoring liver metastasis. I've now derived cell lines from these liver lesions, and I'm characterizing them further to gain deeper insights into PROX1's role in liver mets.
Andrea Miyahira: Okay. Well, thank you so much, Balaji, for sharing all of this with us. It looks like there's a lot of studies to be done, a lot of promise, so good luck and thanks again.
Varadha Balaji Venkadakrishnan: Thank you very much for having me. Thank you for your time.