AlgenML: Drug target discovery platform for transcriptional reprogramming of MYCN-driven neuroblastoma

AlgenML: Drug target discovery platform for transcriptional reprogramming of MYCN-driven neuroblastoma PROJECT SUMMARY Drug discovery is a laborious, time-consuming, and expensive undertaking for biopharma. Oncology is especially difficult with new drugs in clinical trials having just 3.4% probability of success. This application addresses significant challenges of traditional drug target discovery in oncology that relies on cell viability or reporter assays which oversimplifies cell state. New advancements in single-cell RNA expression profiling allows us to overcome these challenges by quantitatively mapping transcriptional dependencies in cancer cells and rapidly probing vulnerabilities to reprogram the oncogenic signaling networks. Transcription factors MYCN and MYC are to date non-druggable by small molecules despite being high value cancer drug targets as they are frequently amplified genes and drive poor outcome across the cancer spectrum. Agents that block MYCN indirectly identified from synthetic lethal viability screens have resulted in only modest or short-lived responses in ongoing clinical trials. Algen’s proprietary machine learning platform (AlgenML) identifies targets that block oncogenic transcription addiction on MYCN using single-cell RNA expression of CRISPR interference (CRISPRi) gene knockdown. Genome-wide single-cell RNA expression profiling measures 10,000 genes per cell and each high- throughput assay routinely captures 160,000 cells at once. Using CRISPRi gene knockdown libraries and multiplexing the assays, hundreds of genes can be knocked down simultaneously and we single-cell RNA sequence 200 cells per CRISPRi gene knockdown. This makes for an extremely rich data set with over 400 million data points of RNA expression data which AlgenML analyzes. Our drug discovery approach is innovative because, unlike traditional approaches, the AlgenML platform does not identify essential genes that cause cell death, but rather selects drug targets in an unbiased manner whose suppression can reprogram the disease- related transcriptional dependencies. Resulting drugs should be safer and better tolerated. Here, our approach is to optimize AlgenML to monitor and reprogram MYCN transcriptional activity in new genetically defined models of MYCN-driven neuroblastoma. We focus on neuroblastoma because MYCN amplifications are common in the disease, and the genetically defined models allow detection of the precise contribution of MYCN oncogene compared to isogenic controls. In Aim 1, we define MYCN transcriptional signature, nominate target genes, and test target genes in vitro based on their ability to reprogram the MYCN transcriptional dependency. Aim 2 evaluates in vivo efficacy of target inhibition to shrink tumors and extend lifespan in new human induced pluripotent stem cell (iPSC) and rodent models of neuroblastoma from UCSF. Our team of investigators at Algen and UCSF has decades of experience in developing RNA signatures to indirectly targeting MYC and MYCN via synthetic lethality. Mature startup Algen Biotechnologies has a vibrant ecosystem of investors, innovators, and grew out of Nobel Laureate Dr. Jennifer Doudna’s lab at UC Berkeley, who has been a scientific advisor since its founding. With Dr. William Weiss at UCSF, we have assembled a powerhouse environment.PROJECT NARRATIVE Algen is developing a drug discovery platform to select and test drug targets that block the oncogenic transcriptional effects of non-druggable drivers such as MYCN. MYC and MYCN transcription factors are directly non-druggable by small molecules, despite being high value targets for drug discovery because their amplifications correlate with extremely poor prognosis across the cancer spectrum. MYCN is amplified in 25% of all neuroblastoma, the most common extracranial solid tumor of childhood, and marks high-risk disease with an aggressive phenotype that is resistant to chemotherapy and only a 40-50% 5-year survival rate.