←
Jason Sheltzer4,785 posts
Jason Sheltzer avatar

Jason Sheltzer

@JSheltzer

Assistant prof at @StanfordMed. Interested in aneuploidy, mitotic kinases, cancer therapeutics, and drug development. Co-founder x2.

◷ Joined February 2016▧ 1,016 Photos and videos
Jason Sheltzer avatar
Jason Sheltzer@JSheltzer·6 Jul 2023

Check out our new study in @ScienceMagazine, where we take on a 100-year-old debate: what’s the role of aneuploidy in cancer? We discovered that genetically removing extra chromosomes blocks cancer growth - a phenomenon we call “aneuploidy addiction”. science.org/doi/10.1126/scie…

Matthew Yglesias avatar
⟳ Jason Sheltzer retweeted
Matthew Yglesias@mattyglesias·Oct 30

Jean-Luc Picard remains the single most compelling vision of progressive politics I am familiar with — secular, rational, compassionate, egalitarian; a believer in progress & abundance but also duty, hard work, and a constant battle for self-improvement and improving the world.

Post attachment
Jason Sheltzer avatar
Jason Sheltzer@JSheltzer·Oct 29

So, we’re excited about DDX3X. The therapeutic hypothesis is that an inhibitor that’s selective for DDX3X over DDX3Y would be lethal in XO tumors (about 30% of cancers in adult men) but spare healthy XY tissue.

Jason Sheltzer avatar
Jason Sheltzer@JSheltzer·Oct 29

You can even put this next to clinical vulnerabilities. KRAS is king, but DDX3X-LOY is comparable to biomarkers like BRAF mutations for BRAF dependency and HER2 amplifications for HER2 dependency. It’s much stronger than MTAP/PRMT5 and MSI/WRN.

Post attachment
Jason Sheltzer avatar
Jason Sheltzer@JSheltzer·Oct 29

Within the limitations of DepMap - it’s genetic not pharmacological perturbation, there’s no healthy tissue control, it’s simply measuring cell fitness in vitro, etc etc - here are 23 reported aneuploidy vulnerabilities. The DDX3X-LOY vulnerability is by far the strongest.

Post attachment
Jason Sheltzer avatar
Jason Sheltzer@JSheltzer·Oct 29

Now, returning to the first tweet - why we think DDX3X may be the best approach for targeting aneuploidy. Many vulnerabilities in aneuploid cells have been described, including my lab’s work on UBE2H. We conducted a uniform analysis of each biomarker/vulnerability pair in DepMap.

Jason Sheltzer avatar
Jason Sheltzer@JSheltzer·Oct 29

Next, we wanted to evaluate the suitability of DDX3X for therapeutic targeting. We generated LOY cells with degron-tagged DDX3X, and we found that ablation of DDX3X triggers apoptosis in vitro and tumor regression in mice.

Post attachmentPost attachmentPost attachment
Jason Sheltzer avatar
Jason Sheltzer@JSheltzer·Oct 29

DDX3X is an RNA helicase. WT but not catalytically-inactive DDX3X rescues viability, demonstrating a potentially druggable enzymatic dependency. Also, unlike PPP2CA, prolonged DDX3X suppression remains lethal - in the XO state, there are no compensatory copies of DDX3Y remaining.

Post attachment
Jason Sheltzer avatar
Jason Sheltzer@JSheltzer·Oct 29

Redundancy between X and Y paralogs has been previously observed, including by the Page, Floor, and Neumüller groups. We generated five isogenic XY and XO clones and validated that DDX3X ablation is tolerated in the XY but not the XO state.

Post attachmentPost attachment
Jason Sheltzer avatar
Jason Sheltzer@JSheltzer·Oct 29

The answer was already in our data. The top hit in our screen was DDX3X, which has a Y chromosome paralog (DDX3Y) and is essential in cells that have lost Y. Unlike 8p-loss, which creates hemizygosity, Y-loss creates nullizygosity. There is no DDX3Y remaining to compensate.

Post attachment
Jason Sheltzer avatar
Jason Sheltzer@JSheltzer·Oct 29

This seemed like a problem, for PPP2CA and for potentially any aneuploidy-targeting therapy. There’s still a copy of PPP2CB remaining. If you put these cells under cytotoxic stress, then they can just evolve to compensate. Which aneuploidy could we target to avoid this problem?

Jason Sheltzer avatar
Jason Sheltzer@JSheltzer·Oct 29

However, we ran into an issue. We used CRISPRi to achieve partial PPP2CA inhibition, and we found that 8p-loss cells were able to evolve to compensate. After 20 days of PPP2CA suppression, they had upregulated PPP2CB and they were no longer dependent on PPP2CA.

Post attachmentPost attachmentPost attachment
Jason Sheltzer avatar
Jason Sheltzer@JSheltzer·Oct 29

PPP2CB over-expression in 8p-loss cells rescues the dependency, validating the paralog-compensation mechanism.

Post attachment
Jason Sheltzer avatar
Jason Sheltzer@JSheltzer·Oct 29

Next, we sought to validate this dependency. With the Taylor & Roessler labs, we generated isogenic cells with one or two copies of chromosome 8p. Consistent with the whole-genome screening data, we found that PPP2CA was only essential in the 8p-loss cells.

Post attachmentPost attachment