A frameshift mutation drives divergent biosynthesis of metallophores in Methylobacterium extorquens

Iron is widely considered the first metallocofactor, evolving as iron-sulfur clusters in early life. While iron-chelating siderophores have been widely characterized across microbial life, the lanthanide-chelating metallophore, methylolanthanin, has only recently been described in Methylobacterium extorquens AM1. Methylolanthanin shares structural similarities to the siderophore rhodopetrobactin but contains 4-hydroxybenzoate chelating moieties in place of canonical 3,4-dihydroxybenzoates. Here we compare M. extorquens AM1, which produces methylolanthanin, and the closely related M. extorquens PA1, which produces rhodopetrobactin. We present a pathway for the biosynthesis of both metallophores and describe the unusual synthesis of methylolanthanin's 4-HB moieties from tyrosine. We uncover a frameshift mutation in the predicted 3-dehydroshikimate dehydratase, mllF, that prevents production of rhodopetrobactin in AM1 through truncation of the catalytically essential N-terminus. We find that deletion of the uncharacterized gene mllG reveals a cryptic branch of the pathway, leading to production of both methylolanthanin and rhodopetrobactin. Finally, we discover that rhodopetrobactin production in this mutant is enabled through the activity of a 3-dehydroshikimate dehydratase in a separate biosynthetic gene cluster. These insights highlight an evolutionary mechanism for metallophore diversification through pseudogenization and regulation of distinct biosynthetic gene clusters with shared aromatic intermediates.