The engineering of cGAS phase separation ability alters downstream immune signaling

In this study, we developed a novel machine learning tool, PScalpel, that guide us to engineer the protein phase separation (PS) ability with amino acid mutation. cGAS was one of the most crucial nucleic acid sensors in immune system, and its PS ability was widely recognized that executed significant role in both anti-viral and anti-tumor immunity. Taking cGAS as an application object for PScalpel, we successfully modulated its PS ability in the anticipated direction by altering single amino acid, which subsequently optimized its immune function and impacted the activity of engineered macrophages. We got two mutants that significantly change the PS with higher and weaker ability. To further investigate the specific genes regulated by PS of cGAS, we performed transcriptome sequencing in cGAS WT, cGAS H5- and cGAS G6N expressed THP-1-derived macrophages. With the infection of HSV-1, we have sequenced the RNA difference and analyzed the genes associated with cGAS PS ability. Therefore, we manipulated cGAS's PS ability based on relative consistent protein expression to achieve the precise regulation of targeted signaling network, which inspired us the significance of PScalpel in medical biomolecular engineering.