Skip main navigation

A Novel Angiotensin-Converting Enzyme–Related Carboxypeptidase (ACE2) Converts Angiotensin I to Angiotensin 1-9

Originally published Research. 2000;87:e1–e9


    Abstract—ACE2, the first known human homologue of angiotensin-converting enzyme (ACE), was identified from 5′ sequencing of a human heart failure ventricle cDNA library. ACE2 has an apparent signal peptide, a single metalloprotease active site, and a transmembrane domain. The metalloprotease catalytic domains of ACE2 and ACE are 42% identical, and comparison of the genomic structures indicates that the two genes arose through duplication. In contrast to the more ubiquitous ACE, ACE2 transcripts are found only in heart, kidney, and testis of 23 human tissues examined. Immunohistochemistry shows ACE2 protein predominantly in the endothelium of coronary and intrarenal vessels and in renal tubular epithelium. Active ACE2 enzyme is secreted from transfected cells by cleavage N-terminal to the transmembrane domain. Recombinant ACE2 hydrolyzes the carboxy terminal leucine from angiotensin I to generate angiotensin 1-9, which is converted to smaller angiotensin peptides by ACE in vitro and by cardiomyocytes in culture. ACE2 can also cleave des-Arg bradykinin and neurotensin but not bradykinin or 15 other vasoactive and hormonal peptides tested. ACE2 is not inhibited by lisinopril or captopril. The organ- and cell-specific expression of ACE2 and its unique cleavage of key vasoactive peptides suggest an essential role for ACE2 in the local renin-angiotensin system of the heart and kidney. The full text of this article is available at

    Angiotensin-converting enzyme (ACE) is a pivotal component of the renin-angiotensin system (RAS), mediating numerous systemic and local effects in the cardiovascular system. ACE is produced in the endothelium of somatic tissues as a transmembrane protein containing two active domains, both of which are inhibited by ACE inhibitors.1 Analysis of the genomic sequence suggests that the two active domains appear to have come from a common ancestor.2 ACE is also produced in the testis via an alternate promoter, generating a shorter testis-specific form with only one active site.2 Both the somatic endothelial ACE and the testicular form can be cleaved to generate soluble active enzymes. ACE removes the carboxy terminal dipeptide from the decapeptide angiotensin I (Ang I) to generate angiotensin II (Ang II), a potent vasoconstrictor, and degrades bradykinin, a vasodilator. Thus, ACE activity serves to increase systemic vascular tone, and pharmacological ACE inhibition is effective in the treatment of hypertension.

    In addition to the systemic RAS, a local RAS is known to operate within certain tissues (for review, see Reference 3 ). Cardiac myocytes express Ang II receptors and undergo hypertrophy in response to Ang II in vitro. Ang II also induces cardiac fibroblast proliferation and collagen production. In vivo, ACE inhibition reverses cardiac myocyte hypertrophy and fibrosis associated with ventricular remodeling and heart failure, presumably by reducing local generation of Ang II. The tissue RAS has additional complexity, however. At least one other enzyme, a heart chymase that is secreted by mast cells, is capable of converting Ang I to Ang II.4 Furthermore, several other Ang I cleavage products have been found, for example, Ang1-9, Ang1-7, Ang III,2345678 and Ang IV.345678 Although Ang1-7 has been shown to have physiological effects such as diuresis and vasodilation (for review, see Reference 5 ), in general the activities of these and other angiotensin peptides are not well understood.

    In the present study, we report the first human homologue of ACE, identified in an ongoing search for novel genes related to heart failure. ACE-related carboxypeptidase (ACE2), like ACE, is a membrane-associated and secreted enzyme expressed predominantly on endothelium, but unlike ACE, it is highly restricted in humans to heart, kidney, and testis. ACE2 catalyzes the cleavage of Ang I to Ang1-9 and may have a unique role in the local RAS in heart and kidney.

    Materials and Methods

    cDNA Library Construction, Sequencing, Extension, and Phylogenetic Analysis

    A human cardiac left ventricle cDNA library was prepared from the explant of a heart transplant recipient, a 43-year-old white woman with idiopathic dilated cardiomyopathy (sample obtained from G. Sandusky and N. Bowling, Eli Lilly & Co, Indianapolis, Ind), using standard directional cloning techniques. Nineteen thousand clones were selected at random for high throughput 5′ end automated sequencing. Homology searching was done using TBLASTN and Hidden Markov models from PFAM (see Rapid amplification of cDNA ends (RACE) was done using the Marathon cDNA amplification kit (Clontech), the 5′ primer 5′-GSP-ACE2 (CAC AGG TTC CAC CAC CCC AAC TAT CTC), and heart mRNA from which the original cDNA library was generated. The sequence can be found at GenBank No. AF291820. A similar sequence was submitted to GenBank (No. AF241254) by independent investigators shortly before submission of this manuscript. For sequence alignment and phylogenetic tree analysis, the CLUSTAL V method was used to group sequences into clusters by examining the distances between all pairs.6

    Expression Vector Construction

    Full-length human ACE2 expression vector (pACE2) was generated by ligating the 1.4-kb EcoRI/AflIII fragment of a 5′ cDNA extension clone (nucleotides 1 to 1381, see GenBank No. AF291820) and the 1.9-kb AflIII/NotI fragment of the original library clone (nucleotides 1381 to 3325, see GenBank No. AF291820) into the EcoRI/NotI sites of pcDNA3.1 (Invitrogen). An expression vector for secreted human ACE2 (psACE2) was generated by replacing the 3′ end of full-length human ACE2 with a polymerase chain reaction fragment containing a stop codon inserted after the serine immediately preceding the predicted transmembrane domain (amino acid 740).

    Northern Blot Analysis

    A 636-bp probe of the 3′ untranslated region of human ACE2 cDNA containing no homology to ACE (MscI/KpnI fragment, nucleotides 2583 to 3219, GenBank No. AF291820) was labeled with 32P using the Multiprime labeling system (Amersham) and hybridized to human multiple tissue Northern blots (Clontech) overnight at 65°C in Nylon Wash (128 mmol/L Na2HPO4 · 7H2O, 14 mmol/L EDTA, 0.2% Triton X-100, 14% SDS, pH 7.2). The blots were then washed three times for 30 minutes each at 65°C in 0.5× Nylon Wash. (Independent experiments on different Northern blots yielded identical results.) A BstXI fragment of testicular ACE comprising base pairs 578 to 1736 (100% homologous to endothelial ACE) was hybridized to the identical blots above after stripping ACE2 signal by washing with boiling 0.5% SDS. Testicular ACE cDNA was the generous gift of the James Riordan Laboratory, Harvard Medical School, Boston, Mass.

    Cell Culture and Transfections

    Chinese hamster ovary (CHO) K1 cells were maintained in serum-free Ultra CHO medium (Biowhittaker) at 37°C in a humidified 5% CO2 incubator. For transfections, cells were seeded at 1×106 cells/100-mm dish on day 0 and transfected with Lipofectamine (Gibco BRL) on day 1 as per the manufacturer’s instructions. Briefly, 10 μg of DNA was combined with 40 μL of Lipofectamine in Opti-MEM medium (Gibco BRL) and the mixture was incubated on the cells for 5 to 6 hours. Ultra CHO medium was then added to the cells and they were incubated overnight at 37°C. On day 2, the medium was changed and on day 4 the conditioned media were collected. For some experiments, the conditioned media were concentrated in Centriplus 30 concentrators (Amicon) in 10 mmol/L Tris-Cl, pH 7.0. After the conditioned media were harvested from cells, the cells were washed twice with PBS, and 2 mL of lysis buffer was added (50 mmol/L Tris-Cl, pH 7.5, 150 mmol/L NaCl, 0.02% NaN3, and 1% NP-40 with Complete protease inhibitors [Boehringer Mannheim]). The cells were then incubated for 20 minutes on ice. The cells were scraped, transferred to microfuge tubes, and the nuclei were pelleted 10 minutes in a microfuge. The remaining cell lysate was used immediately or frozen at −20°C. Primary rat neonatal ventricular cardiomyocytes were prepared using 1- to 3-day-old neonates according to published methods.7 The cells were plated in modified Eagle’s medium containing 5% FCS, 1% penicillin/streptomycin, and 0.1 mmol/L bromodeoxyuridine for 24 hours and then switched to serum-free medium supplemented with insulin-transferrin-sodium selenite (Sigma) for an additional 24 hours. Synthetic sarcosyl-blocked peptides (Ang1-9, Ang1-8, Ang1-7, and Ang1-5), which are resistant to N-terminal degradation by aminopeptidases, were then added individually at concentrations of 1 to 10 μmol/L for periods of up to 48 hours. Aliquots of conditioned media were collected at different time points for analysis by mass spectrometry.

    Antibodies, Immunohistochemistry, and Immunoblot Analysis

    Two rabbit polyclonal anti-ACE2 antibodies (anti-ACE251 and anti-ACE2489) were raised against synthetic peptides representing human ACE2 residues 51 to 69 (NTNITEENVQNMNNAGDKW) and residues 489 to 508 (EPVPHDETYCDPASLFHVSN) (Research Genetics). These were affinity-purified on peptide columns and concentrated according to standard methods. For immunohistochemistry, frozen sections of human heart and kidney were cut at 6 μm thickness, mounted on charged slides, fixed in cold acetone (Sigma), and immersed in 0.3% H2O2 in methanol for 10 minutes to block endogenous peroxidase. A 1:2000 dilution of anti-ACE251 was applied in 5% BSA/PBS for 30 minutes at room temperature. Slides were then incubated with an alkaline phosphatase–labeled polymer (DAKO) for 30 minutes at room temperature. After multiple PBS washes, Fast Red substrate-chromogen solution was applied for 10 minutes and the slides were then counterstained with hematoxylin. To determine the specificity of the antibody, peptide competition was performed on serial sections. ACE2 peptide 51 to 69 or a nonspecific peptide (not shown) was incubated for 2 hours at room temperature at a 1:10 ratio (by weight) with anti-ACE251. The solution was then applied at the step of the primary antibody addition above. For Western blot analysis, conditioned medium from transfected cells was loaded onto 10% to 20% precast SDS-polyacrylamide gels (Bio-Rad), followed by separation at 100 V for 2 hours and transferred to Hybond P using a semidry transfer system (Amersham). The blots were incubated in blocking buffer (5% nonfat milk, 0.1% Tween-20, in PBS) for 1 hour at room temperature or overnight at 4°C. The blot was then incubated with anti-ACE2489 at a 1:5000 dilution for 30 minutes in PBS with 0.1% Tween-20, then washed 1 time for 15 minutes and three times for 5 minutes in PBS containing 0.1% Tween-20. The blot was then incubated with donkey anti-rabbit HRP (Amersham) in PBS with 0.1% Tween-20 for 30 minutes, followed by washing as above. Signal was detected using the ECL Plus kit (Amersham). Antibody specificity was tested against ACE2 and testicular ACE produced by transfected CHO cells in an immunoblot analysis. None of the antibodies cross-reacted with ACE.

    Purification of Recombinant ACE2 Enzyme

    Conditioned media from CHO cells transfected with soluble human ACE2 (see above) were concentrated 10-fold in an Amicon 10K molecular weight cutoff stirred cell apparatus (Amicon), then desalted by injection on a Pharmacia PC 3.2-mm/100-mm 800-μL-bed volume Fast Desalting column (Pharmacia) in bis-Tris buffer (25 mmol/L bis-Tris-propane, 25 mmol/L Tris HCl, pH 6.5). The resulting protein mixture was injected onto a Pharmacia PC 1.6-mm/50-mm 100-μL-bed volume MonoQ anion exchange column in bis-Tris buffer. The column was washed with bis-Tris buffer and then eluted with a 0 to 250 mmol/L NaCl gradient over 20 column volumes. The ACE2-containing fractions, as determined by Western blot, were pooled. (NH4)2SO4 was added to a 1 mol/L final concentration. This mixture was injected onto a Pharmacia PC 1.6-mm/50-mm 100-μL-bed volume Phenyl Superose HIC column in 1 mol/L (NH4)2SO4 and 100 mmol/L NaPi, pH 7.0. The column was washed with (NH4)2SO4/NaPi, pH 7.0, loading buffer and eluted with a 1 to 0.5 mol/L (NH4)2SO4 reverse gradient in 100 mmol/L NaPi, pH 7.0, over 10 column volumes. ACE2-containing fractions, as determined by Coomassie staining of a reduced SDS-PAGE gel, were pooled and desalted by injection on a Pharmacia PC 3.2-mm/100-mm 800-μL-bed volume Fast Desalting column in 100 mmol/L NaPi, pH 7.0. ACE2-containing fractions, as resolved by reduced SDS-PAGE, were used for assays. Conditioned medium from empty vector–transfected cells was purified as above to produce the appropriate negative control.

    Analysis of ACE and ACE2 Activity by Mass Spectrometry

    Enzymatic reactions were performed in 15 μL. To each tube at room temperature was added 10 μL of buffer (10 mmol/L Tris, pH 7.0) with or without enzyme. Five microliters of purified Ang I (DRVYIHPFHL) (Sigma) or other peptide substrates were added to each tube for a final concentration of 5 μmol/L. Lisinopril or captopril (Sigma) was added to some reactions at final concentrations of 6.6 μmol/L. For reactions and control experiments, the tubes were incubated at 37°C for 30 minutes. A portion (1 μL) of each reaction was quenched by the addition of 1 μL of a low-pH MALDI matrix compound (10 g/L α-cyano-4 hydroxycinnamic acid in a 1:1 mixture of acetonitrile and water). One microliter of the resulting solution was applied to the surface of a MALDI plate. The plate was then air-dried and inserted into the sample introduction port of the Voyager Elite biospectrometry MALDI time-of-flight (TOF) mass spectrometer (PerSeptive Biosystems). The resulting signal was digitized at a frequency of 1 GHz and accumulated for 64 scans. Purified conditioned medium from empty vector transfections was used to control individual experiments for variability in extent of substrate conversion to product. For tandem mass spectrometry sequencing, a hybrid quadrupole time-of-flight mass spectrometer (Q-TOF MS) (Micromass UK Limited) equipped with an orthogonal electrospray source (Z-spray) was used. The quadrupole was set up to pass precursor ions of selected m/z to the hexapole collision cell (Q2), and product ion spectra were acquired with the TOF analyzer. Argon was introduced into the Q2 with a collision energy of 35 eV and cone energy of 25 V.


    Identification of ACE2

    A novel human cDNA homologous to ACE was identified among 19 000 5′ end sequences obtained from a heart failure ventricle cDNA library. Complete sequencing of the 2.3-kb partial cDNA clone and a 1178-bp 5′ extension yielded an 805-codon open reading frame encoding ACE2 (see GenBank No. AF291820 and Figure 1 for amino acid sequence). The sequence includes an initiator methionine codon within a Kozak consensus sequence,8 a putative signal sequence (amino acids 1 to 18), a potential transmembrane domain (amino acids 740 to 763), and a potential metalloprotease zinc binding site (amino acids 374 to 378, HEMGH).

    Comparison with ACE proteins indicated that ACE2 contains a single catalytic domain (amino acids 147 to 555) that is 42% identical to each of the two catalytic domains in endothelial ACE.19 Human ACE2 is 33% identical overall to human testicular ACE, the alternatively spliced product of the ACE gene that also contains a single catalytic site, although there is considerable divergence in C-terminal regions (Figure 1). Phylogenetic analysis indicated that the mammalian ACE family members are more similar to each other than they are to Drosophila ACE (Figure 2).

    Comparison of the human ACE2 cDNA with previously unannotated human genomic sequences (GenBank No. AC003669, BAC GS-594A7) revealed an ACE2 gene comprising 19 exons (not shown). Relative conservation of exon/intron organization between the ACE2 and ACE catalytic domains2 further suggests that these two genes arose by duplication of a common ancestor (Table 1).

    ACE2 Tissue Distribution

    Northern blot analysis revealed a ≈3.4-kb ACE2 transcript expressed only in heart, kidney, and testis of 23 human tissues examined; a minor ≈8-kb band was also detected in kidney (Figure 3A). In contrast, a doublet of endothelial ACE mRNA was detected in 16 of 23 tissues examined, and the much shorter testicular isoform was also detected in testis (Figure 3B). Immunohistochemical analysis of the ventricular myocardium identified ACE2 localized to the endothelium of most intramyocardial vessels including capillaries, venules, and medium-sized coronary arteries and arterioles. Immunostaining was apparent, but to a lesser extent, in vascular smooth muscle cells and focally in the adventitia of some larger vessels (Figures 4A through 4F). There was no observable difference in the distribution or intensity of protein expression in sections from failing and nonfailing heart samples. In the kidney, ACE2 protein was again present throughout the endothelium and focally in rare smooth muscle cells of medium-sized vessels. It was also identified in proximal tubule epithelial cells (Figures 4G and 4H).

    Expression of Recombinant ACE2

    ACE is made as a transmembrane protein, some of which is cleaved posttranslationally to generate a secreted form in vivo and in cell culture.101112 To determine whether ACE2 is processed in a similar fashion, CHO cells were transiently transfected with expression plasmids containing either ACE2 cDNA (pACE2) or no insert (pcDNA3.1). Conditioned medium and whole-cell lysates were compared in Western analyses using polyclonal antiserum raised to an ACE2 peptide (amino acids 489 to 508; see Materials and Methods) present in both full-length and putative cleaved secreted forms (Figure 5). An approximately 90-kDa immunoreactive band was present in the whole-cell lysate, and a slightly smaller band was detected in the conditioned medium of ACE2-transfected cells indicating that full-length ACE2 is processed in CHO cells to generate a secreted form.

    ACE2 Catalytic Activity

    We tested a variety of vasoactive and hormonal peptides as candidate substrates using conditioned media from transfected CHO cells as a source of secreted recombinant ACE2. We compared ACE2 activity with that of ACE. The enzymes were incubated with synthetically prepared peptides in vitro, and reaction products were analyzed by mass spectrometry. ACE converted the decapeptide Ang I (Ang1-10; m/z 1296.68) to Ang II (Ang1-8; m/z 1046.54), as expected, by cleavage of the C-terminal His-Leu dipeptide (Figure 6). ACE2, in contrast, converted Ang I to a new species of m/z 1183.60, which when sequenced by tandem mass spectrometry proved to be the nonapeptide Ang1-9 (data not shown). Thus, ACE2, apparently a carboxypeptidase, quantitatively cleaved the C-terminal Leu from this substrate. No further cleavage of Ang1-9 by ACE2 was evident on prolonged incubation (not shown). In similar experiments, ACE2 removed the C-terminal residue from three other vasoactive peptides, neurotensin, kinetensin, and des-Arg bradykinin (Table 2). However, it failed to cleave bradykinin and 15 other unrelated vasoactive and hormonal peptides, indicating considerable specificity.

    We tested the effects of ACE inhibitors on ACE2 activity in vitro. Conversion of Ang I to Ang1-9 by ACE2 was not inhibited by lisinopril under conditions that completely inhibited the generation of Ang II by ACE (Figure 7). Identical results were obtained using another ACE inhibitor, captopril (not shown). Thus, despite their homologous catalytic domains, ACE2 and ACE are biochemically and pharmacologically distinct.

    Metabolism of Ang1-9

    Ang1-9 is found in vivo and accumulates in animals treated with ACE inhibitors.1314 On the basis of the present studies, Ang1-9 may be generated in heart and kidney by endothelium-associated ACE2. To explore possible fates for this peptide, we tested its conversion in vitro and in serum-free cell culture. Incubation of Ang1-9 with ACE in vitro generated Ang1-7 (m/z 899.47) and Ang1-5 (m/z 665.36), apparently by sequential cleavage of C-terminal dipeptides (Figure 8). Conversion to Ang II (Ang1-8) was not detected. Incubation of Ang1-9 with primary rat neonatal cardiac myocytes, which produce ACE,15 also yielded Ang1-7 and Ang1-5, as well as Ang1-4 (m/z 508.28) (Figure 9). Addition of lisinopril blocked only the generation of Ang1-5, indicating that in these cultures generation of Ang1-5 requires ACE, whereas generation of Ang1-7 and Ang1-4 do not. Other cardiomyocyte peptidases that might cleave Ang1-9 are unknown at present.


    The present study identifies ACE2 as the first known human homologue of ACE, an enzyme that plays a central role in vascular, renal, and myocardial physiology. The ACE2 and ACE catalytic domains are 42% identical in amino acid sequence, and conservation of exon/intron organization further indicates that the two genes evolved from a common ancestor. Also like ACE, ACE2 is expressed in endothelium and can be secreted in vitro through proteolytic cleavage of a putative membrane-bound form. In contrast to ACE, however, ACE2 is highly tissue-specific: whereas ACE is expressed ubiquitously in the vasculature, human ACE2 is restricted to heart, kidney, and testis. In addition to endothelial expression, ACE2 is present in smooth muscle in some coronary vessels and focally in tubular epithelium of the kidney. Thus, ACE2 is particularly well poised to participate in both cardiac and renal physiology.

    Both ACE and ACE2 cleave Ang I, but their activities are distinct. Whereas ACE is a dipeptidase, ACE2 removes the single C-terminal Leu residue to generate Ang1-9. Ang1-9 has been identified in vivo in rat and human plasma, but its function is unknown.1314 Some studies have suggested that it may be an endogenous inhibitor of ACE.16 Our findings indicate that Ang1-9 is a competitive inhibitor of ACE because it is itself an ACE substrate. Under conditions in which ACE is inhibited, such as after long-term administration of ACE inhibitors in rats, Ang1-9 levels have been shown to be increased in plasma and kidney.1314 This increase in Ang1-9 steady-state levels could be due to decreased catabolism of Ang1-9 by ACE. Conversely, the increased levels of Ang1-9 could be due to increased production by ACE2 as a result of increased availability of Ang I substrate. Together these results indicate that alternate pathways of Ang I metabolism exist and that these pathways may be amplified in the presence of ACE inhibitors.

    Other investigators have found that Ang II can be generated in human renal extracts in a two-step process with Ang1-9 as an intermediate.1417 The enzymes responsible for this activity have not been definitively determined, but cathepsin A has been proposed.18 ACE2 is most highly expressed in kidney and may be responsible for some or all of the Ang1-9 generated from Ang I in these studies. ACE2 does not convert Ang1-9 further to Ang II, at least under our in vitro conditions. Thus, ACE2 would not directly be responsible for Ang II generation in the presence of ACE inhibitors.

    Other angiotensin peptides generated by the combined activities of ACE2 and other peptidases, including ACE, may play key signaling roles in heart and kidney. Although the functions of Ang1-5 and Ang1-9 are unknown, Ang1-7 is known to function as a vasodilator (for review, see Reference 3 ). Thus, ACE2 could function to increase local vasodilation through its ability to produce a precursor to Ang1-7. In addition, ACE2 could also alter vasomotor tone by decreasing the availability of Ang I, a well-established substrate for ACE and precursor of the vasoconstrictor Ang II (Figure 10). Thus, ACE2 activity in the heart could induce local vasodilation, thereby maintaining myocardial perfusion under conditions of generalized systemic vasoconstriction such as acute blood loss. Local ACE2 activity in the endothelium and tubular epithelium of the kidney may modulate renal blood flow distribution and salt and water handling and thereby regulate blood pressure.

    It remains to be determined whether Ang I is indeed a physiological substrate for ACE2 in vivo and whether other substrates for ACE2 exist. Our in vitro studies demonstrate that, in addition to Ang I, ACE2 is capable of cleaving at least three other vasoactive peptides, des-Arg bradykinin, neurotensin, and kinetensin. Des-Arg bradykinin is involved locally in vessel dilation through binding to the B1 receptor that is expressed under conditions of tissue damage or inflammation.19 This is consistent with a role for ACE2 in local regulation of vasomotor tone, through both Ang I and des-Arg bradykinin cleavage. In contrast, ACE cleaves the vasodilator bradykinin, which acts systemically through the B2 receptor. Neurotensin has diverse effects in the cardiovascular, nervous, and digestive systems (for review, see Reference 20 ), whereas kinetensin, a related peptide derived in vitro, stimulates mast cell degranulation and vascular permeability.21 Degradation of these peptides by ACE2 may serve to modulate these activities.

    The elucidation of ACE2 adds a potential new dimension to the complexity of the RAS. Enzymes other than ACE, eg, heart chymase, convert Ang I to Ang II (for review, see References 2 and 22 ), and Ang II can act through at least three distinct receptors (for review, see References 23 and 24 ). In addition, multiple other angiotensin-related peptides including Ang1-9 and Ang1-7 are present and may have unique effects. Targeting of the cardiac RAS at several of these points is a mainstay of drug treatment for hypertension and heart failure. ACE2, expressed specifically in heart and kidney and capable of cleaving Ang I and other key vasoactive peptides, provides additional possibilities for the development of novel therapeutics.

          Figure 1.
          Figure 1.
          Figure 1.

    Figure 1. Comparison of ACE2 and ACE protein sequences. The amino acid sequences of human ACE2, rat-secreted ACE2 (Rat-sACE2), human testicular ACE (human T-ACE), murine testicular ACE (murine T-ACE), rabbit testicular ACE (rabbit T-ACE), and Drosophila melanogaster ACE are aligned. Residues that do not match the consensus are shaded.

          Figure 2.

    Figure 2. Phylogenetic tree of ACE family members. The phylogenetic relationships among the several ACE and ACE2 sequences were determined by the Clustal V method and the Neighbor-joining method (see Materials and Methods). The phylogenies are based on divergence found in pair distances (numbers are based on substitution events).

          Figure 3.

    Figure 3. Tissue-restricted expression of human ACE2. Northern blots representing multiple human tissues were hybridized with probes from either ACE2 (A) or ACE (B) (see Materials and Methods). RNA standards (kb) are shown at left. sk. muscle indicates skeletal muscle; sm. intestine, small intestine; and pbl, peripheral blood leukocyte.

          Figure 4.

    Figure 4. Cell-specific expression of ACE2 protein in heart and kidney. Human tissue sections were probed with antibody to human ACE2 (Ab ACE251) in the presence or absence of peptide antigen (see Materials and Methods). Arrows indicate endothelial cells; a, artery; and t, proximal tubule. A, C, and E, Three sections of human left ventricle stained with Ab ACE251. B, D, and F, Corresponding serial sections incubated with Ab ACE251 in the presence of competing peptide ACE251. G and H, Serial sections of human kidney incubated with Ab ACE251 in the absence and presence, respectively, of competing peptide ACE251.

          Figure 5.

    Figure 5. Recombinant ACE2 protein is secreted by transfected CHO cells. Conditioned media and lysates from cells transfected with full-length ACE2 or empty vector control were run on SDS-PAGE and immunoblotted using an antibody to human ACE2 (Ab ACE2489). Molecular mass standards are shown at left.

          Figure 6.

    Figure 6. ACE2 converts Ang I to Ang1-9. Ang I was incubated with ACE2, ACE, or no enzyme, as indicated, for 30 minutes at 37°C, and the products were analyzed by mass spectrometry: no enzyme control (purified fraction from conditioned medium of CHO cells transfected with empty vector) showing no conversion (A), purified porcine kidney ACE showing conversion to Ang1-8 (Ang II) (B), and purified ACE2 showing conversion to Ang1-9 (C).

          Figure 7.

    Figure 7. ACE2 conversion of Ang I is not blocked by lisinopril. Ang I was incubated for 30 minutes at 37°C in the presence of porcine kidney ACE (A), porcine kidney ACE+lisinopril (6.6 μmol/L) (B), purified ACE2 (C), and purified ACE2+lisinopril (6.6 μmol/L) (D), and the products were analyzed by mass spectrometry.

          Figure 8.

    Figure 8. Ang1-9 is a substrate for ACE. Ang1-9 was incubated alone (A, time=0) or with ACE (B, time=5 seconds at 20°C; C, time=30 minutes at 37°C), and products were analyzed by mass spectrometry.

          Figure 9.

    Figure 9. Conversion of Ang1-9 to products by rat neonatal cardiac myocyte culture. [Sar1]Ang1-9 was added to serum-free cultures of primary rat neonatal cardiac myocytes for 0 hours (A) and 48 hours in the absence (B) and presence (C) of lisinopril. The resulting products in the conditioned medium were analyzed by mass spectrometry.

          Figure 10.

    Figure 10. Schematic of Ang I and bradykinin peptide processing by ACE and ACE2. A, Proposed pathways of Ang I cleavage to products via ACE and ACE2. B, Proposed pathways of bradykinin and des-Arg bradykinin degradation to products by ACE and ACE2.

    Table 1. Comparison of the Exon Sizes Encoding the Putative Catalytic Domain of ACE2 and the Catalytic Domains of ACE

    Domain 1Domain 2
    Bases, bp (Exon)Bases, bp (Exon)Bases, bp (Exon)
    144 (5)144 (4)144 (17)
    113 (6) plus 106 (7)192 (5)192 (18)
    98 (8)98 (6)100 (19)
    170 (9)173 (7)171 (20)
    227 (10)224 (8)224 (21)
    145 (11)145 (9)145 (22)
    99 (12)99 (10)99 (23)
    123 (13)123 (11)122 (24)

    Table 2. Peptides Tested for Cleavage by ACE2

    C-Terminal 4-Amino Acids
    Angiotensin IP-F-H*L
    des-Arg bradykininF-S-P*F
    Not Cleaved
    Angiotensin (1-9)H-P-F-H
    Met enkephalinG-G-F-M
    Leu enkephalinG-G-F-L
    Atrial natriuretic peptideS-F-R-Y
    Adrenocorticotropic hormoneP-L-E-F
    Neuromedin CG-H-L-M
    Luteinizing hormone–releasing hormoneL-R-P-G
    Neuropeptide YR-Q-R-Y
    Mast cell–degranulating peptide HR1K-K-V-L

    *Site of cleavage.

    This work was supported by funding from Eli Lilly & Company, Indianapolis, Ind. We gratefully acknowledge the skilled technical assistance of Stephen Katz, Yong Yao Xu, Roslyn Feeley, Ron Meyer, and Lin Zhang. We appreciate the gift of a left ventricular heart failure tissue sample from George Sandusky and Nancy Bowling of Eli Lilly & Company, Indianapolis, Ind. The advice and support of Andrew Nichols and Geoff Ginsburg are also greatly appreciated.


    Correspondence to Susan Acton, Millennium Pharmaceuticals, Inc, 75 Sidney St, Cambridge, MA 02139. E-mail


    • 1 Ehlers MRW, Riordan JF. Angiotensin-converting enzyme. Zinc- and inhibitor-binding stoichiometries of the somatic and testis isozymes. Biochemistry.1991; 30:7118–7126.CrossrefMedlineGoogle Scholar
    • 2 Hubert C, Houot AM, Corvol P, Soubrier F. Structure of the angiotensin I-converting enzyme gene: two alternate promoters correspond to evolutionary steps of a duplicated gene. J Biol Chem.1991; 266:15377–15383.CrossrefMedlineGoogle Scholar
    • 3 Stock P, Liefeldt L, Paul M, Ganten D. Local renin-angiotensin systems in cardiovascular tissues: localization and functional role. Cardiology.1995; 86:2–8.CrossrefMedlineGoogle Scholar
    • 4 Urata H, Nishimura H, Ganten D, Chymase-dependent angiotensin II forming system in humans. Am J Hypertens.1996; 9:277–284.CrossrefMedlineGoogle Scholar
    • 5 Ferrario CM, Chappell MC, Dean RH, Iyer SN. Novel angiotensin peptides regulate blood pressure, endothelial function, and natriuresis. J Am Soc Nephrol.1998; 9:1716–1722.CrossrefMedlineGoogle Scholar
    • 6 Higgins DG, Sharp PM. Fast and sensitive multiple sequence alignments on a microcomputer. Comput Appl Biosci.1989; 5:151–153.MedlineGoogle Scholar
    • 7 Long CS, Henrich CJ, Simpson PC. A growth factor for cardiac myocytes is produced by cardiac nonmyocytes. Cell Regul.1991; 2:1081–1095.CrossrefMedlineGoogle Scholar
    • 8 Kozak M. Interpreting cDNA sequences: some insights from studies on translation. Mamm Genome.1996; 7:563–574.CrossrefMedlineGoogle Scholar
    • 9 Soubrier F, Alhenc-Gelas F, Hubert C, Allegrini J, John M, Tregear G, Corvol P. Two putative active centers in human angiotensin I-converting enzyme revealed by molecular cloning. Proc Natl Acad Sci U S A.1988; 85:9386–9390.CrossrefMedlineGoogle Scholar
    • 10 Ehlers MR, Schwager SL, Chubb AJ, Scholle RR, Brandt WF, Riordan JF. Proteolytic release of membrane proteins: studies on a membrane-protein-solubilizing activity in CHO cells. Immunopharmacology.1997; 36:271–278.CrossrefMedlineGoogle Scholar
    • 11 Sadhukhan R, Sen GC, Ramchandran R, Sen I. The distal ectodomain of angiotensin-converting enzyme regulates its cleavage-secretion from the cell surface. Proc Natl Acad SciU S A.1998; 95:138–143.CrossrefGoogle Scholar
    • 12 Skeggs L Jr. Historical overview of the renin-angiotensin system. In: Doyle AE, Bearn AG, eds. Hypertension and the Angiotensin System: Therapeutic Approaches. New York, NY: Raven Press; 1984:31–45.Google Scholar
    • 13 Johnson H, Kourtis S, Waters J, Drummer OH. Radioimmunoassay for immunoreactive [des-Leu10]-angiotensin I. Peptides.1989; 10:489–492.CrossrefMedlineGoogle Scholar
    • 14 Drummer OH, Kourtis S, Johnson H. Effect of chronic enalapril treatment on enzymes responsible for the catabolism of angiotensin I and formation of angiotensin II. Biochem Pharmacol.1990; 39:513–518.CrossrefMedlineGoogle Scholar
    • 15 Dostal DE, Rothblum KN, Conrad KM, Cooper GR, Baker KM. Detection of angiotensin I and II in cultured rat cardiac myocytes and fibroblasts. Am J Physiol.1992; 263:C851–C863.CrossrefMedlineGoogle Scholar
    • 16 Snyder RA, Wintroub BU. Inhibition of angiotensin-converting enzyme by des-Leu10-angiotensin I: a potential mechanism of endogenous angiotensin-converting enzyme regulation. Biochim Biophys Acta.1986; 871:1–5.CrossrefMedlineGoogle Scholar
    • 17 Changaris DG, Miller JJ, Levy RS. Angiotensin II generated by a human renal carboxypeptidase. Biochem Biophys Res Commun.1986; 138:573–579.CrossrefMedlineGoogle Scholar
    • 18 Miller JJ, Changaris DG, Levy RS. Conversion of angiotensin I to angiotensin II by cathepsin A isoenzymes of porcine kidney. Biochem Biophys Res Commun.1988; 154:1122–1129.CrossrefMedlineGoogle Scholar
    • 19 Marceau F. Kinin B1 receptors: a review. Immunopharmacology.1995; 30:1–26.CrossrefMedlineGoogle Scholar
    • 20 Vincent JP, Mazella J, Kitabgi P. Neurotensin and neurotensin receptors. Trends Pharmacol Sci.1999; 20:302–309.CrossrefMedlineGoogle Scholar
    • 21 Sydbom A, Ware J, Mogard MH. Stimulation of histamine release by the peptide kinetensin. Agents Actions.1989; 27:68–71.CrossrefMedlineGoogle Scholar
    • 22 Danser AH, Saris JJ, Schuijt MP, van Kats JP. Is there a local renin-angiotensin system in the heart? Cardiovasc Res.1999; 44:252–265.CrossrefMedlineGoogle Scholar
    • 23 Lorell BH. Role of angiotensin AT1, and AT2 receptors in cardiac hypertrophy and disease. Am J Cardiol.1999; 83:48H–52H.MedlineGoogle Scholar
    • 24 Riordan JF. Angiotensin II: biosynthesis, molecular recognition, and signal transduction. Cell Mol Neurobiol.1995; 15:637–651.CrossrefMedlineGoogle Scholar


    eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

    Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.