Southern California, I want to speak to you today about the large number of non-coding RNAs we were able to identify with functional relevance in glioblastoma invasion using our CRISPR functional screen. I want to thank Kathleen Sung for her significant contribution to this project, and I want to start by acknowledging the clear role of glioblastoma invasion and the aggressive behavior of this tumor. And while studies have really identified multiple protein target candidates for treatment, gains have been modest. It's our feeling that this modest gain might be improved by focusing on the 80% of the genome that has not been really studied in the context of glioblastomas. And these are long non-coding RNAs, and they weren't studied because it was felt that this represented junk, as proteins are classically thought to modulate disease. But it turns out these non-coding RNAs have very significant function in multiple cancers using the mechanism shown to the right. So you have a large untapped resource for the diagnosis and treatment of these tumors. The issue is that bioinformatically, while we can identify candidates of non-coding RNAs, it's much harder to determine their function. CRISPR-I is a perfect way to assess this, and that's largely because of its efficiency and very good specificity in targeting the transcriptional start site of these genes. And what we're able to do is target these at the nucleus, which is where they act in many cases alone, which is an advantage over the classical shRNA systems. And finally, we're able to repress transcription entirely, which is important with our interference system rather than the classical CRISPR cutting, which doesn't always affect link RNAs. Now here, we show using qPCR that we are able to repress link RNA expression significantly in both coding and non-coding genes. So what we did was use a massively parallel screen to assess thousands of candidate genes. And here I have an example of an individual knockdown of a red, a green, and a yellow non-coding RNA. Now these colors are just representative, they're not fluorophores. And this knockdown is also tagged with the same identity. So at the beginning, you see its presence in your pool, but at the end of whatever phenotype screen you're doing, in our case, invasion, you see the presence of red here suggests that knockdown of red, again, tagged as red, does not repress its presence, the cell presence, but knockdown of green and yellow is no longer present, which does give you a sense of the functional relevance of that knockdown. So we did this in 2,300 genes simultaneously with 250 negative controls using a massive pool lentivirally transduced into our system. And we screened this through a matrigel invasion chamber, which allows cells to invade through a top chamber into a bottom chamber, assessing the invasion as those cells that reach the bottom. And we validated those targeted knockdowns and those non-coding RNAs as hits by individually testing them, knocking them out individually and assessing the migration using crystal violet stain, which is gold standard and qPCR to ensure no off-target effects. Ultimately, we used antisensibly nucleotides as a pharmacologic means of seeing if we could translate this to bedside. Now, our outcomes are shown in the top right where you see 48 positive hits, which are outside of this little volcano in the top right. And assessing these one by one, you see these non-annotated genes, which is why they don't have names, having significant knockdown and the qPCR at the left of 50%. When you look at the invasion, specifically the cells that have made it all the way past to the bottom of the chamber in the control, you see a lot of purple cells, but in the knockdown, you see an 80s for 7% repression of that ability to migrate through the matrigel. Other candidates showed both invasion and growth phenotypes. And one of our more promising candidates showed a 16-fold decrease in invasion. Finally, antisensibly nucleotides, our pharmacologic agent, was able to repress gene expression up to 95% with a phenotypic corresponding result on the right. So I wanna thank my lab as well, significant funding from the NIH-KL2 as well as the Margaret Early Foundation that has supported our work. Thank you.

non-coding RNAs

functional relevance

glioblastoma invasion

long non-coding RNAs

CRISPR-I