Determination of the 1-ethyl-3-[(3-dimethylamino)propyl]-carbodiimide- induced cross-link between the beta and epsilon subunits of Escherichia coli F1-ATPase.

Abstract

The zero-length cross-link between the inhibitory epsilon subunit and one of three catalytic beta subunits of Escherichia coli F1-ATPase (alpha 3 beta 3 gamma delta epsilon), induced by a water-soluble carbodiimide, 1-ethyl-3-[(3-dimethylamino) propyl]-carbodiimide (EDC), has been determined at the amino acid level. Lability of cross-linked beta-epsilon to base suggested an ester cross-link rather than the expected amide. A 10-kDa cross-linked CNBr fragment derived from beta-epsilon was identified by electrophoresis on high percentage polyacrylamide gels. Sequence analysis of this peptide revealed the constituent peptides to be Asp-380 to Met-431 of beta and Glu-96 to Met-138 of epsilon. Glu-381 of beta was absent from cycle 2 indicating that it was one of the cross-linked residues, but no potential cross-linked residue in epsilon was identified in this analysis. A form of epsilon containing a methionine residue in place of Val-112 (epsilon V112M) was produced by site-directed mutagenesis. epsilon V112M was incorporated into F1-ATPase which was then cross-linked with EDC. An 8-kDa cross-linked CNBr fragment of beta-epsilon V112M was shown to contain the peptide of epsilon between residues Glu-96 and Met-112 and the peptide of beta between residues Asp-380 and Met-431. Again residue Glu-381 of beta was notably reduced and no missing residue from the epsilon peptide could be identified, but the peptide sequence limited the possible choices to Ser-106, Ser-107, or Ser-108. Furthermore, an epsilon mutant in which Ser-108 was replaced by cysteine could no longer be cross-linked to a beta subunit in F1-ATPase by EDC. Both mutant forms of epsilon supported growth of an uncC-deficient E. coli strain and inhibited F1-ATPase. These results indicate that the EDC-induced cross-link between the beta and epsilon subunits of F1-ATPase is an ester linkage between beta-Glu-381 and, likely, epsilon-Ser-108. As these residues must be located immediately adjacent to one another in F1-ATPase, our results define a site of subunit-subunit contact between beta and epsilon.