All right. Our next talk is from Arman Jahangiri. His abstract presentation is on CRISPR-improved inducible CRISPR interference achieves specific and reversible PIK-A silencing in glioblastoma. Good afternoon. My name is Arman Jahangiri. I'm an MD-PhD student from the Agilab at UCSF. As she mentioned, my talk is entitled CRISPR-improved inducible CRISPR interference achieves specific and reversible PIK-A silencing in glioblastoma. The funding for this was from an F31 grant of the NCI and Howard Hughes Fellowship. So many of you in this room have tried to study glioblastoma cells in your lab, and sometimes the essential genes can be quite difficult to focus on. As we use traditional knockout or knockdown methods such as siRSH RNA and conventional cloning, unfortunately, a lot of these can prove to be nonspecific and irreversible, often selecting against targeted cells through antibiotic selection. A specific gene that was of interest to us in the Agilab was a PI3 kinase enhancer protein A gene or PIK-A. It's also called Agab2 in the literature, but for the remainder of this talk, I will refer to it as PIK-A. PIK-A is a proto-oncogene and a GTPase. It's also an important component of the CDK4 amplicon. The way PIK-A functions is through binding to PI3 kinase, and what that does is it enhances the AKT's kinase activity and then promotes cancer progression. Now, GTPases usually are not good drugable targets, however, and the reason being that they have identical GTP binding sites, they usually have picomolar affinity for nucleotides, and basically, small molecules have a hard time competing with them. PIK-A, on the other hand, is a very appealing target because it has a very weak affinity for nucleotides. We're talking in the range of protein kinases, and it has the ability to hydrolyze multiple nucleotides in a selective fashion, and most importantly, because we are collaborating with a chemistry lab, we have a crystal structure of it, which helps us in coming up with small molecules. The literature, there's not too much on it in terms of preclinical studies. It is a very difficult target to repress genetically. So, what we did was we used a CRISPR-I or CRISPR interference tool, which is a tunable system able to downregulate individual alleles, and we repressed PIK gene expression in human glial blastoma cells. Now, the way CRISPR-I works is you have a deactivated Cas9 protein bound to a Crab protein, and then you have your guide RNA that comes and binds to the DNA targeting region, and through an inducible fashion, you're able to repress your target. So, at baseline, the cells are expressing target gene, and with addition of doxycycline, which we used, you're able to deactivate that. You have to first create the cells with the machinery in them, and we utilized U87 and GBM43 glial blastoma cells for this, and selected three guide RNAs, small guide RNAs, utilizing the MIT website that were best capable of knocking down our gene of interest. We also used additional siRNA and shRNA modalities to use as a control and utilize qPCR immunoblotting and biological assays to assess PIK's ability or impact once it's knocked down. So, basically, to take advantage of this CRISPR-I in GBM cells, we started by generating a stable GBM cell line. It's not advancing. There we go. And basically, it has deactivated Cas9 and CRAB that's inducible under doxycycline inducible promoter from a safe harbor locus, and then we introduced the guide RNAs of interest through a plasmid transfection, and by targeting the gRNA of interest near the transcription start site, you can then activate the knockdown of your gene of interest if this would advance. So, the gRNA binds, and you're able to then there's a very long delay. So, basically, we created cell lines that express the dCas9 and CRAB fusion protein. This is essentially the CRISPR-I machinery that was designed by Stan Chiad of UCSF. He's now at Stanford, and it's under a doxycycline inducible promoter. This construct was then integrated in a single copy vector into AAVS safe harbor locus, but it remains transcriptionally active. So, in the presence of doxycycline, the transactivator is then able to bind to the TRE, or the tetracycline response element, and it promotes and starts transcribing dCas9. So, in our study, again, we created these GBM cell lines that had the machinery, and then we took the PIKeA or AgAP2 guide RNA, which had a fluorescent marker, and were able to introduce them, and through cell sorting, we were able to select them as yielded a higher yield as compared to using antibiotic selection. So, what did we find, and did this work for us with PIKeA? As you can see, utilizing CRISPR-I, the knockdown levels were greater than 99 percent as compared to using regular CRISPR-S control, and much of that reason might be that the cells that have the PIKeA knockdown might die, so the expression is much lower. When you compare it to shRNA transfection as controls, our shRNA transfections led to cell death during clonal selection. So, if you look at lanes 1 and 2, they're expressing PIKeA fully, but our guide RNA 1 and 7, from the 20-something numbers that we selected, have an almost complete knockdown in U87 cells. And then, looking at our biologic assays, here in the Augie lab, we created a three-dimensional agros assay to study invasion, and you can see that the control cells are able to invade across the red line, and the regular CRISPRs have some invasion, but when you look at the CRISPR-I of PIKeA knockdown, the amount of invasion is almost reduced to zero. And lastly, looking at proliferation levels, on the far right, we have the CRISPR-I knockdown of PIKeA versus the CRISPR-induced ones, and you can see the difference between the proliferation rates after 48 hours using a cyclone proliferation assay. So, I hope that I've been able to demonstrate to you that CRISPR-I is a valuable tool allowing for highly specific and inducible genetic repression of genes essential for cancer cell survival, such as PIKeA, and it is a more sensitive technology than regular CRISPR technology, allowing verification of PIKeA's role in glioblastoma proliferation, migration, and validating this target as a drugable GTPase target in glioblastoma. We're currently working on understanding more about this specific gene, and also working with the Conklin Lab to do whole genome CRISPR-I screening libraries that will hopefully identify additional novel drugable targets that can yield clinical benefits for GBM patients. I'd like to thank my PIs, Dr. Bill Weiss and Dr. Manish Aghi, and a special thanks to Dr. Mandagar, who's the postdoc from the Conklin Lab, helped create the machinery, members of the lab, Max Zetter of NAMCO Chandra. Thank you very much.

CRISPR-improved inducible CRISPR interference

PIK-A gene

glioblastoma

gene knockout

gene knockdown