Metabolic module exchange in plant-endophyte coevolution: mechanisms and implications

BACKGROUND: Endophytes establish intimate symbioses with their hosts through long-term coevolution. They are known to promote plant growth, enhance stress resistance, and contribute to the biosynthesis of bioactive compounds. However, despite extensive research on these beneficial roles, a systematic synthesis of their distribution patterns, functional differentiation, and mechanisms of metabolite production remains lacking. This knowledge gap limits our understanding of plant-endophyte coadaptation and hinders its applications in agriculture and medicine. PURPOSE OF REVIEW: This review critically synthesizes major studies on endophyte diversity, functional specialization, and metabolite biosynthesis. By integrating bibliometric and systematic evidence, we aim to elucidate the essential patterns and possible mechanisms underlying these phenomena-specifically, how endophytes and their host plants achieve functional complementarity and metabolic congruence. We further evaluate the roles of horizontal gene transfer and convergent evolution in enabling both partners to produce identical or structurally similar bioactive metabolites. Significant Scientific Ideas: Endophyte communities are highly structured, forming the basis for their interactions with host plants. Their distribution and functions are host‑ and tissue‑specific: strains from the same tissue of congeneric species show closer phylogenetic relatedness. A clear functional division exists in phytohormone production-bacteria predominantly produce indole-3-acetic acid (IAA), while fungi mainly produce gibberellins (GAs). Roots represent the dominant and most diverse colonization site. Horizontal gene transfer and convergent evolution serve as critical mechanisms enabling metabolic complementation between endophytes and plants. Together, these findings indicate that plant-endophyte coadaptation is underpinned by structured community assembly and functional specialization, rather than by random association.