Stomatal development in plants is regulated by defensin-like secretory epidermal patterning factor (EPF) peptide hormones. Only one of these, stomagen, is a positive regulator, whereas EPF1, EPF2, and possibly others are negative regulators. Here we explore the structure-function relationships of EPFs, by integrating NMR and semi-in vitro stomagen experiments. We show that stomagen is composed of a loop and a scaffold containing three disulphide bonds. A mutant composed of the stomagen loop and the EPF2 scaffold positively regulates the stomatal density on Arabidopsis cotyledons. The reciprocal mutant composed of the EPF2 loop and the stomagen scaffold acts negatively. Deletion of the disulphide bond introduces unfolding and inactivity. Our results suggest that the loop confers the functional specificity of EPFs and that the scaffold is structurally required for their activity. This structural decomposition approach to elucidating the functional site could be adapted for the analysis of other cysteine-rich peptide families.

Scorpion K⁺ channel toxins and insect defensins share a conserved three-dimensional structure and related biological activities (defense against competitors or invasive microbes by disrupting their membrane functions), which provides an ideal system to study how functional evolution occurs in a conserved structural scaffold. Using an experimental approach, we show that the deletion of a small loop of a parasitoid venom defensin possessing the “scorpion toxin signature” (STS) can remove steric hindrance of peptide-channel interactions and result in a neurotoxin selectively inhibiting K⁺ channels with high affinities. This insect defensin-derived toxin adopts a hallmark scorpion toxin fold with a common cysteine-stabilized α-helical and β-sheet motif, as determined by nuclear magnetic resonance analysis. Mutations of two key residues located in STS completely diminish or significantly decrease the affinity of the toxin on the channels, demonstrating that this toxin binds to K⁺ channels in the same manner as scorpion toxins. Taken together, these results provide new structural and functional evidence supporting the predictability of toxin evolution. The experimental strategy is the first employed to establish an evolutionary relationship of two distantly related protein families.

A model legume, Medicago truncatula, has over 600 nodule-specific cysteine-rich (NCR) peptides required for symbiosis with rhizobia. Among them, NCR169, an essential factor for establishing symbiosis, has four cysteine residues that are indispensable for its function. However, knowledge of NCR169 structure and mechanism of action is still lacking. In this study, we solved two NMR structures of NCR169 caused by different disulfide linkage patterns. We show that both structures have a consensus C-terminal β-sheet attached to an extended N-terminal region with dissimilar features; one moves widely, whereas the other is relatively stapled. We further revealed that the disulfide bonds of NCR169 contribute to its structural stability and solubility. Regarding the function, one of the NCR169 oxidized forms could bind to negatively charged bacterial phospholipids. Furthermore, the positively charged lysine-rich region of NCR169 may be responsible for its antimicrobial activity against Escherichia coli and Sinorhizobium meliloti. This active region was disordered even in the phospholipid bound state, suggesting that the disordered conformation of this region is key to its function. Morphological observations suggested the mechanism of action of NCR169 on bacteria. The present study on NCR169 provides new insights into the structure and function of NCR peptides.