Crystal structure of the N domain of human somatic angiotensin I-converting enzyme provides a structural basis for domain-specific inhibitor design

Human somatic angiotensin I-converting enzyme (sACE) is a key regulator of blood pressure and an important drug target for combating cardiovascular and renal disease. sACE comprises two homologous metallopeptidase domains, N and C, joined by an inter-domain linker. Both domains are capable of cleaving the two hemoregulatory peptides angiotensin I and bradykinin, but differ in their affinities for a range of other substrates and inhibitors. Previously we determined the structure of testis ACE (C domain); here we present the crystal structure of the N domain of sACE (both in the presence and absence of the antihypertensive drug lisinopril) in order to aid the understanding of how these two domains differ in specificity and function. In addition, the structure of most of the inter-domain linker allows us to propose relative domain positions for sACE that may contribute to the domain cooperativity. The structure now provides a platform for the design of "domain-specific" second-generation ACE inhibitors.     

Mapping of conformational mAb epitopes to the C domain of human angiotensin I-converting enzyme

Angiotensin I-converting enzyme (ACE, CD143) has two homologous domains, each having a functional active site. Fine epitope mapping of 8 mAbs to the C-terminal domain of human ACE was carried out using plate precipitation assays, mAbs' cross-reactivity with ACE from different species, site-directed mutagenesis, and antigen- and cell-based ELISAs. Almost all epitopes contained potential glycosylation sites. Therefore, these mAbs could be used to distinguish different glycoforms of ACE expressed in different tissues or cell lines. mAbs 1B8 and 3F10 were especially sensitive to the composition of the N-glycan attached to Asn 731; mAbs 2H9 and 3F11 detected the glycosylation status of the glycan attached to Asn 685 and perhaps Asn1162; and mAb 1E10 and 4E3 recognized the glycan on Asn 666. The epitope of mAb 1E10 is located at the N-terminal end of the C domain, close to the unique 36 amino acid residues of testicular ACE (tACE). Moreover, it binds preferentially to tACE on the surface of human spermatozoa and thus may find application as an immunocontraceptive drug. mAb 4E3 was the best mAb for quantification of ACE-expressing somatic cells by flow cytometry. In contrast to the other mAbs, binding of mAb 2B11 was not markedly influenced by ACE glycosylation or by the cell culture conditions or cell types, making this mAb a suitable reference antibody. Epitope mapping of these C-domain mAbs, particularly those that compete with N-domain mAbs, enabled us to propose a model of the two-domain somatic ACE that might explain the interdomain cooperativity. Our findings demonstrated that mAbs directed to conformational epitopes on the C-terminal domain of human ACE are very useful for the detection of testicular and somatic ACE, quantification using flow cytometry and ELISA assays, and for the study of different aspects of ACE biology.     

Purification of angiotensin I-converting enzyme from human lung.

Angiotensin I-converting enzyme (peptidyl dipeptide hydrolase, EC 3.4.15.1) was solubilized from the membrane fraction of human lung using trypsin treatment and purfied using columns of DE 52-cellulose, hydroxyapatite and Sephadex G-200. The purified enzyme was shown to convert angiotensin I to angiotensin II and also to inactivate bradykinin. The specific activity of the enzyme was 9.5 units/mg protein for Hippuryl-His-Leu-OH and 0.665 mumol/min per mg protein for angiotensin I. The enzymic activity obtained after trypsin treatment (1 mg/200 mg protein) for 2 h could be divided into three components: (i) an enzyme of molecular weight 290 000 (peak I), (ii) an enzyme of molecular weight 180 000 (peak II) and (iii) an enzyme of molecular weight 98 000 (peak III), by columns of DE 52-cellulose and Sephadex G-200. Km values of peak I, II and III fraction for Hippuryl-His-Leu-OH were identical at 1.1 mM. pH optimum of the enzyme was 8.3 for Hippuryl-His-Leu-OH.     

Tissue angiotensin I-converting enzyme activity in spontaneously hypertensive hamsters.

The purpose of this study was to measure angiotensin I-converting activity in heart, kidney, lung and cheek pouch tissue homogenates of spontaneously hypertensive and normotensive hamsters. We also determined inhibitor sensitivity and the effects of chloride anion concentration on kidney angiotensin I-converting activity in these animals. We found no significant differences in angiotensin I-converting activity between hypertensive and normotensive hamsters in all tissues tested. Inhibitor sensitivity of kidney angiotensin I-converting activity with captopril and lisonopril was similar in both groups. Finally, kidney angiotensin I-converting activity increased significantly in both groups as chloride anion concentration in the assay buffer increased. Substituting chloride anion for citrate abrogated the increase in angiotensin I-converting enzyme activity.     

Purification of human serum angiotensin I-converting enzyme by affinity chromatography

A relatively simple procedure is described for purifying human serum angiotensin-converting enzyme. The enzyme was purified 130,000-fold to electrophoretic homogeneity using affinity chromatography as the principal purification step. The ligand was an immobilized competitive inhibitor, d-cysteinyl-l-proline. A six-carbon spacer arm was satisfactory for trapping the enzyme; 80% of the bound enzyme was eluted with 3 m urea-1.0 m NaCl-0.1 m Tris, pH 8.3. The specific activity was 39 units/mg protein and the molecular weight (155,000), isoelectric point (4.7), kinetic properties, and the effect of various inhibitors are in agreement with published reports.     

Characterization of angiotensin I-converting enzyme activity in the freshwater turtle, Amyda japonica

1. Angiotensin I-Converting Enzyme (ACE) activity has been characterized in the freshwater turtle, Amyda japonica. 2. Peak activity of ACE in plasma from the freshwater turtle was shown at pH 9.0, which was more alkaline compared to that of mammals. 3. Chloride requirements for the optimal ACE activity were different from species. 4. ACE inhibitors, EDTA, teprotide (SQ 20,881), Captopril (SQ 14,225) showed dose-dependent inhibitions of ACE activity in plasma from the freshwater turtle as well as mammals. 5. ACE activity was found in several different tissues with a different activity showing the highest activity in kidney homogenate from the freshwater turtle, Amyda japonica.     

Structural basis of proline-specific exopeptidase activity as observed in human dipeptidyl peptidase-IV

Inhibition of dipeptidyl peptidase IV (DPP-IV), the main glucagon-like peptide 1 (GLP1)-degrading enzyme, has been proposed for the treatment of type II diabetes. We expressed and purified the ectodomain of human DPP-IV in Pichia pastoris and determined the X-ray structure at 2.1 A resolution. The enzyme consists of two domains, the catalytic domain, with an alpha/beta hydrolase fold, and a beta propeller domain with an 8-fold repeat of a four-strand beta sheet motif. The beta propeller domain contributes two important functions to the molecule that have not been reported for such structures, an extra beta sheet motif that forms part of the dimerization interface and an additional short helix with a double Glu sequence motif. The Glu motif provides recognition and a binding site for the N terminus of the substrates, as revealed by the complex structure with diprotin A, a substrate with low turnover that is trapped in the tetrahedral intermediate of the reaction in the crystal.     

Regulation of angiotensin I-converting enzyme activity in serially cultivated bovine endothelial cells

Angiotensin I-converting enzyme (ACE) activity was measured in lysates of cloned and uncloned cultures of bovine fetal aortic endothelial cells. The expression of ACE activity in these cells was complex, and influenced by subcultivation, cell density, serum, cumulative population doublings, and clonal heterogeneity. The ACE specific activity at any point in the in vitro lifespan was determined, at least in part, by interaction of these culture variables. After subcultivation to subconfluent densities, cellular ACE specific activity decreased markedly and did not reach detectable levels until cells attained confluent densities. The use of different suppliers' lots of serum in the growth medium resulted in different cellular ACE specific activities. The ACE specific activity decreased as cultures were serially subcultivated, but remained detectable throughout the lifespan, suggesting a linkage between the proliferative history of an endothelial cell and its remaining capacity to express ACE. Increased ACE activity was observed when cells at the end of their lifespan were cultured at high densities. Cloned strains behaved similarly to the uncloned parent culture, except that they exhibited a wide range of ACE specific activities.     

Functional conservation of the active sites of human and Drosophila angiotensin I-converting enzyme

Human somatic angiotensin I-converting enzyme (sACE) has two active sites present in two homologous protein domains, resulting from a tandem gene duplication. It has been proposed that the N- and C-terminal active sites can have specific in vivo roles. In Drosophila melanogaster, Ance and Acercode for two ACE-like single-domain proteins, also predicted to have distinct physiological roles. We have investigated the relationship of Ance and Acer to the N- and C-domains of human sACE by genomic sequence analysis and by using domain-selective inhibitors, including RXP 407, a selective inhibitor of the human N-domain. These phosphinic peptides were potent inhibitors of Acer, but not of Ance. We conclude that the active sites of the N-domain and of Acer share structural features that permit the binding of the unusual RXP407 inhibitor and the hydrolysis of a broader range of peptide structures. In comparison, Ance, like the human C-domain of ACE, displays greater inhibitor selectivity. From the analysis of the published sequence of the Adh region of Drosophila chromosome 2, which carries Ance, Acer, and four additional ACE-like genes, we also suggest that this functional conservation is reflected in an ancestral gene structure identifiable in both protostome and deuterostome lineages and that the duplication seen in vertebrate genomes predates the divergence of these lineages. The conservation of ACE enzymes with distinct active sites in the evolution of both vertebrate and invertebrate species provides further evidence that these two kinds of active sites have different physiological functions.     

Dipeptide-hydroxamates are good inhibitors of the angiotensin I-converting enzyme

The inhibition constants (Ki) and modes of inhibition have been determined for a series of dipeptide-hydroxamate compounds with bovine lung parenchyma angiotensin I-converting enzyme (peptidyldipeptide carboxy-hydrolase, E.C. 3.4. 15.1). The hydroxamido function was borne by aspartic, glutamic, or aminoadipic acid and extended by 2, 3 or 4 bond lengths from the proline amide bond. L-glu(NHOH)-L-pro (Ki = 3.4 microM) and D,L-aminoadipicyl (NHOH)-L-pro (Ki = 1.2 microM) were the best competitive inhibitors of the hydrolysis of benzoyl-gly-his-gly but were not effective as affinity ligands for purification of the enzyme.     

Molecular dynamics simulations of human and dog gastric lipases: insights into domain movements

Mammalian gastric lipases are stable and active under acidic conditions and also in the duodenal lumen. There has been considerable interest in acid stable lipases owing to their potential application in the treatment of pancreatic exocrine insufficiency. In order to gain insights into the domain movements of these enzymes, molecular dynamics simulations of human gastric lipase was performed at an acidic pH and under neutral conditions. For comparative studies, simulation of dog gastric lipase was also performed at an acidic pH. Analyses show, that in addition to the lid region, there is another region of high mobility in these lipases. The potential role of this novel region is discussed.     

Defining the boundaries of the testis angiotensin I-converting enzyme ectodomain

Numerous cytokines, receptors, and ectoenzymes, including angiotensin I-converting enzyme (ACE), are shed from the cell surface by limited proteolysis at the juxtamembrane stalk region. The membrane-proximal C domain of ACE has been implicated in sheddase-substrate recognition. We mapped the functional boundaries of the testis ACE ectodomain (identical to the C domain of somatic ACE) by progressive deletions from the N- and C-termini and analysing the effects on catalytic activity, stability, and shedding in transfected cells. We found that deletions extending beyond Leu37 at the N-terminus and Trp616 at the C-terminus abolished catalytic activity and shedding, either by disturbing the ectodomain conformation or by inhibiting maturation and surface expression. Based on these data and on sequence alignments, we propose that the boundaries of the ACE ectodomain are Asp40 at the N-terminus and Gly615 at the C-terminus.     

Physicochemical characteristics of homogeneous bovine lung angiotensin I-converting enzyme. Comparison with human serum enzyme

Angiotensin I-converting enzyme was purified to electrophoretic homogeneity (12 units/mg) from bovine lung tissue and from human serum using an affinity gel described previously (Harris et al., (1981) Anal. Biochem. 111, 227-234). The isoelectric point (4.5), molecular weight (145 000), S20,W (8.1), amino acid composition and carbohydrate content of the lung enzyme are all similar to the values obtained for the human serum enzyme. The NH2-terminus of the lung enzyme (Ala) is different from that of the serum enzyme (Tyr) but the COOH-terminal sequences are identical (-Leu-Ser-OH). Pure bovine lung enzyme was reduced and carboxyamidomethylated with iodo (14C1) acetamide to the extent predicted by the number of cysteine residues. Since no radioactivity was incorporated into denatured enzyme that was not reduced, all of the cysteine residues must be in the form of disulfide bonds. Reverse-phase HPLC was used to separate peptides obtained from the lung enzyme after degradation with either trypsin or cyanogen bromide. The number of peptides resolved (42 after trypsin, 31 after cyanogen bromide), were only 20% fewer than the number predicted from the amino acid analysis and therefore the possibility that the converting enzyme (a single polypeptide chain) might be a fused dimer is excluded.     

Arg(1098) is critical for the chloride dependence of human angiotensin I-converting enzyme C-domain catalytic activity.

Angiotensin (Ang) I-converting enzyme (ACE) is a Zn(2+) metalloprotease with two homologous catalytic domains. Both the N- and C-terminal domains are peptidyl dipeptidases. Hydrolysis by ACE of its decapeptide substrate Ang I is increased by Cl(-), but the molecular mechanism of this regulation is unclear. A search for single substitutions to Gln among all conserved basic residues (Lys/Arg) in human ACE C-domain identified R1098Q as the sole mutant that lacked Cl(-) dependence. Cl(-) dependence is also lost when the equivalent Arg in the N-domain, Arg(500), is substituted with Gln. The Arg(1098) to Lys substitution reduced Cl(-) binding affinity by approximately 100-fold. In the absence of Cl(-), substrate binding affinity (1/K(m)) of and catalytic efficiency (k(cat)/K(m)) for Ang I hydrolysis are increased 6.9- and 32-fold, respectively, by the Arg(1098) to Gln substitution, and are similar (<2-fold difference) to the respective wild-type C-domain catalytic constants in the presence of optimal [Cl(-)]. The Arg(1098) to Gln substitution also eliminates Cl(-) dependence for hydrolysis of tetrapeptide substrates, but activity toward these substrates is similar to that of the wild-type C-domain in the absence of Cl(-). These findings indicate that: 1) Arg(1098) is a critical residue of the C-domain Cl(-)-binding site and 2) a basic side chain is necessary for Cl(-) dependence. For tetrapeptide substrates, the inability of R1098Q to recreate the high affinity state generated by the Cl(-)-C-domain interaction suggests that substrate interactions with the enzyme-bound Cl(-) are much more important for the hydrolysis of short substrates than for Ang I. Since Cl(-) concentrations are saturating under physiological conditions and Arg(1098) is not critical for Ang I hydrolysis, we speculate that the evolutionary pressure for the maintenance of the Cl(-)-binding site is its ability to allow cleavage of short cognate peptide substrates at high catalytic efficiencies.