Figure 1. Specificity-determining residues that dictate toxin-antitoxin interactions.
(A) Schematic representing how antitoxin mutations in specificity-determining residues might affect binding to the cognate toxin (blue) or the non-cognate toxin (gold). A wider gap in binding preference reflects greater specificity. The effects of mutating a specificity residue that serves a positive specificity role (a), a negative specificity role (b), or both roles (c) are shown. Mutations that represent possible trade-offs between binding the cognate toxin and discriminating against the non-cognate toxin are also shown (d–e). (B) Models for how positive and negative specificity determinants are distributed across the interface. Model 1: positive and negative design are accomplished by distinct interface residues. Model 2: individual residues can serve both roles. (C) Sequence alignment of three ParD-ParE systems from Mesorhizobium opportunistum. Coevolving residues are highlighted in purple (ParD) or orange (ParE). (D) Phylogenetic tree inferred from protein sequences of 15 highly conserved genes. Distribution of ParDE2 systems and ParDE3 systems are indicated in gold and blue, respectively. (E) Toxicity-rescue assay for wild-type ParD2, wild-type ParD3, or the ParD3 variant indicated, each co-expressed with either wild-type ParE2 or ParE3 toxin. The ParD3 variants harbor subsets, as noted, of the mutations D61I, K64L, and E80K. (F) Residues in ParD and ParE that strongly coevolve (probability score >0.95) with lines connecting covarying pairs. Residues are numbered according to their position in the alignments in panel (C). Positions selected for the ParD3 saturation mutagenesis library are indicated in red.
Figure 1—figure supplement 1. ParE protein tree.
