Abstract
Disclosure: J. Hsieh: Gordian Biotechnology. N. Muraoka: Gordian Biotechnology. J. Chen: Gordian Biotechnology. S. Wang: Gordian Biotechnology. G. Pharaoh: Gordian Biotechnology. I. Driver: Gordian Biotechnology. V. Kartha: Gordian Biotechnology. D. Fuentes: Gordian Biotechnology. D. Popov: Gordian Biotechnology. L. Fung: Gordian Biotechnology. L. Chio: Gordian Biotechnology. C. Carrico: Gordian Biotechnology. H. Miyazaki: Gordian Biotechnology. S. Shambhu: Gordian Biotechnology. A. Gupta: Gordian Biotechnology. C. Bowman: Gordian Biotechnology. A. Alex: Gordian Biotechnology. J. Nonora: Gordian Biotechnology. S. Ruiz: Gordian Biotechnology. A. Vaidyanathan: Gordian Biotechnology. D. Garcia: Gordian Biotechnology. J. Cruz: Gordian Biotechnology. C. Towne: Gordian Biotechnology. F. LePort: Gordian Biotechnology. M. Borch Jensen: Gordian Biotechnology, Siren Biotechnology. Background: Obesity, heart failure with preserved ejection fraction (HFpEF) and metabolic dysfunction-associated steatohepatitis (MASH) share many common comorbidities, including diabetes, hypertension, and chronic kidney disease. We sought to identify new therapeutic targets by screening hundreds of interventions in parallel in animal models with high proximity to the human diseases. Methods: We delivered cocktails of hundreds of barcoded adeno-associated virus (AAV) therapies to various animal models of disease at a low dose. High fat, cholesterol, and fructose (Gubra-Amylin/GAN) diet-fed male C57Bl/6NTac mice were used to model obesity and MASH. MASH was additionally modelled with a middle-aged male African Green monkey fed a similar diet for 82 weeks. 16 weeks old female ZSF1 obese rats were used to model HFpEF. Expression of the interventions were driven by promoters specific to adipocytes, hepatocytes, or cardiomyocytes. Transcriptomic shifts in many biological pathways were measured by single nucleus RNA-sequencing to infer therapeutic effect. The dose of the AAV cocktail was titrated to reduce the likelihood that perturbed cells were adjacent to each other, ensuring that the transcriptomic readout reflects the effects of an isolated intervention. Candidate interventions were then administered as an AAV therapy at a saturating dose to preclinical models to validate the screen predictions. Results: In adipocytes and adipose stem and progenitor cells of GAN diet-fed mice, the effect of knockdown interventions were evaluated for beigeing, insulin signaling, among others. In the hepatocytes of a MASH nonhuman primate, overexpression of the primate-specific alcohol dehydrogenase ADH1B downregulated multiple liver tumor signatures and decreased ER stress. In 62-week GAN diet-fed mice, knockdown of Scap decreased lipid biosynthetic pathways and increased a number of beneficial features. 6 out of 10 tested targets reduced ALT relative to baseline within 8 weeks of AAV administration in mice fed the GAN diet for 44-48 weeks. One novel target reduced fibrosis biomarkers in both short-term and long-term GAN diet-fed mice, with significant reductions in hepatic extracellular matrix gene expression. Screening in the cardiomyocytes of ZSF1 obese rats indicated that overexpression of NPPA suppressed YAP/TAZ-driven hypertrophy, NPPB activated calcium handling in the sarcoplasmic reticulum, and constitutively active PPP1R1A suppressed both ET1-signaling hypertrophy and oxidative stress. Saturation studies in female ZSF1 obese rats showed improvements in diastolic function, including increased e’, and reduced E/e’ and isovolume relaxation time with a novel target. Conclusions: In vivo mosaic screening yields mechanistic insights into clinical and GWAS targets, in addition to identifying novel therapeutic targets for metabolic disease across multiple organs. Presentation: Monday, July 14, 2025