The -Cys-X1-X2-Cys- motif of reduced glutaredoxins adopts a consensus structure that explains the low pK(a) of its catalytic cysteine.

The -Cys-X1-X2-Cys- active site motif is central to the function of enzymes of the thioredoxin superfamily, including glutaredoxins. Their chemistry depends on the lowered pK(a) of the N-terminal thiolate cysteine of the -Cys-X1-X2-Cys- sequence; therefore its structure, dynamics, and electrostatics matter. Much information about the glutaredoxin structures was obtained by nuclear magnetic resonance (NMR), yet these various NMR structures produced heterogeneous and discordant views of the -Cys-X1-X2-Cys- motifs. This study addresses these inconsistencies by a computational and experimental investigation of three diverse reduced -Cys-X1-X2-Cys- motifs, from human glutaredoxin 1 (hGrx1), Escherichia coli glutaredoxin 2 (EcGrx2), and T4 virus glutaredoxin (T4Grx). The NMR models do not account for the low pK(a) of the N-terminal cysteine. However, extensive investigations of the NMR conformers by simulations yielded consensus structures for the -Cys-X1-X2-Cys- motif, with well-defined orientations for the cysteines. pK(a) calculations indicated that the consensus structure stabilizes the thiolate by local hydrogen bonds. The consensus structures of EcGrx2 and T4Grx formed the basis for predicting low pK(a) values for their N-terminal cysteines. Subsequent experimental titrations showed that these pK(a) values are <5, supporting the validity of the consensus structure. The simulations also revisited the conformational dynamics of side chains around the -Cys-X1-X2-Cys- motif, which allowed reconciliation of calculated and measured pK(a) values for important hGrx1 mutants. The conformational spread of these side chains, which differs between NMR and molecular dynamics models, is likely to be relevant to substrate recognition. The new structural models determined in this work should prove to be valuable in future molecular studies of the glutaredoxins.