Fig. 5: Optimizing panicle architecture to enhance rice yield by suppressing ligand–receptor pairs.

a–g Rice panicles from FAZ1 (a) and the osepfl6;7;8 (b), osepfl6;7;9 (c), osepfl6;8;9 (d), osepfl7;8;9 (e), osepfl6;7;8;9 (f), and oser1 (g) mutants. Scale bar = 5 cm. h–j Comparison of average spikelet number per panicle (h), setting percentage (i), and yield per plant (j) between FAZ1, osepfl6;7;8, osepfl6;7;9, osepfl6;8;9, osepfl7;8;9, osepfl6;7;8;9, and oser1. Values are given as the mean ± SD (n = 10 plants). Different letters indicate statistical significance groups at p < 0.05 (one-way ANOVA with post-hoc Tukey’s multiple comparison test). k Comparison of average plot yield between FAZ1 and osepfl6;7;9. Values are given as the mean ± SD (n = 3 plots). *p < 0.05 indicates significant difference compared with FAZ1 by two-sided Student’s t test. l, m Proposed working model for optimizing rice panicle architecture to improve grain yield by specifically suppressing ligand–receptor pairs. During young panicle morphogenesis, the active reproductive meristem produces the small peptides OsEPFL6, OsEPFL7, OsEPFL8, and OsEPFL9, which act as ligands to recognize the OsER1 receptor. This triggers the OsMKKK10–OsMKK4–OsMPK6 cascade to phosphorylate downstream substrates. The spatiotemporally activated GSN1 acts as a molecular “brake” to negatively regulate the OsER1–OsMKKK10–OsMKK4–OsMPK6 pathway by inactivating OsMPK6, which controls rice panicle architecture (l). The active reproductive meristem of the osepfl6;7;9 triple mutant failed to produce OsEPFL6, OsEPFL7, and OsEPFL9. Only OsEPFL8 could thus be perceived by OsER1, alleviating activation of the OsMKKK10–OsMKK4–OsMPK6 cascade and altering the phosphorylation status of the substrates. These integrated signal outputs thus generated optimized rice panicle architecture without affecting spikelet fertility, further enhancing grain yield (m). The source data underlying the statistical analysis in h–k are provided in the Source Data file.