A Molecular Basis for the C-Domain Selectivity of Angiotensin-Converting Enzyme

Angiotensin-Converting Enzyme (ACE) plays an essential role in blood pressure regulationand ACE inhibitors are widely used to treat cardiovascular disease. Two isoforms exist,somatic ACE (sACE) consisting of two homologous domains, N- and C-domain, and testisACE (tACE), corresponding to the C-domain of sACE. Despite a high degree of sequenceidentity, these two domains display marked differences in substrate and inhibitor specificity.Furthermore, the C-domain of ACE has been implicated to play a dominant role in bloodpressure control. It has therefore been suggested that development of ACE inhibitortreatments that selectively block the C-domain will result in decreased side-effects comparedto current therapies. Analysis of three-dimensional structures of tACE in complex withdomain-specific inhibitors has enabled the identification of key active-site residues potentiallyplaying a role in domain selectivity. To investigate the contribution of such residues, a seriesof C-domain mutants was generated containing single and multiple N-domain active-sitesubstitutions. These constructs were used to characterise specific interactions using domainselectiveinhibitors and fluorogenic peptides. Mutants tested with the fluorogenic peptidesdisplayed minimal, if any, acquisition of N-domain-like catalytic properties. Of the singlemutations, S2 (F391Y, tACE numbering) and S1 (V518T) pocket substitutions caused thelargest decreases in affinity for the C-selective phosphinic inhibitor RXPA380 (34-fold) andketo-ACE derivatives (14-26 fold), respectively. The V379S mutation caused an unexpectedincrease in affinity (2-10 fold) for C-selective inhibitors containing a P2’ Trp that could beexplained by the formation of a water-mediated hydrogen bond interaction resulting fromrearrangement of inhibitor and protein side-chains within the S2’ pocket. Multiple mutantscontaining an N-domain-like S2’ pocket combined with the S2 F391Y substitution (S2’F)caused the most notable shift in Ki from that of tACE for the highly selective phosphinicinhibitors, RXPA380 (Ki’s tACE = 69 nM; S2’F = 5300 nM) and N-specific RXP407 (Ki’stACE = 2800 nM; S2’F Ki = 16.1 nM). This work identifies key residues contributing to thedomain selectivity of ACE, and highlights the complex combination of effects involved in thisphenomenon. Furthermore, it provides useful insight for the further design of domainselectiveinhibitors