Manipulating rice canonical Gα and extra-large G protein subunits for improved agronomic traits

Rice productivity is fundamentally linked to its architecture, governed by signaling networks including those based on heterotrimeric G proteins. In this study, we investigated the individual gene impacts and genetic interactions of the canonical G gene (RGA1), and the non-canonical extra-large G genes (OsXLG1, OsXLG3a, OsXLG3b, OsXLG4) in controlling plant architecture. We generated OsXLG mutants using CRISPR/Cpf1 gene editing in Nipponbare (WT) and d1, a Nipponbare null mutant of RGA1. We then phenotyped 25 different genotypes in the greenhouse for 19 different agronomic traits. In wild type (WT), mutations in RGA1, OsXLG3a, OsXLG3b, or OsXLG4, as well as any combination of G genes, resulted in a shorter stature, a desirable trait. Mutations in OsXLG1 and OsXLG4 increased the number of spikelets and grains per panicle, showcasing advantageous traits that led to higher yield. Mutations in OsXLG3a, OsXLG3b, any combination of OsXLGs, or any OsXLG combined with the d1 mutation, reduced seed production and yield. Flag leaf width was the only trait influenced solely by RGA1. RGA1 transcript abundance in the osxlg mutants was positively correlated with height, culm length, panicle exsertion, and harvest index, implicating OsXLG regulation of RGA1 expression as an underlying mechanism. Overall, increased RGA1 expression is correlated with more favorable reproductive traits but less favorable vegetative traits. Our study reveals the complex interaction of RGA1 and OsXLGs within the signaling networks that shape rice architecture, from vegetative to post-harvest stages. Our results suggest modulation of RGA1, OsXLG1, OsXLG3a, or OsXLG4 expression as strategies to enhance yield.