Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Angiotensin-converting enzyme (ACE) is a member of the superfamily of zinc metalloproteinases and a critical participant in several important physiological systems (Corvol et al., 1995). ACE is a very promiscuous enzyme that is involved in catabolism of numerous substrates. In addition to acting on a wide variety of substrates, ACE is exceptional in that it can function as an aminopeptidase and endopeptidase as well as a carboxypeptidase (Skidgel and Erods, 1985). Depending on the substrate, ACE can hydrolyze di- and tripeptides. A free C terminus and the absence of a proline in the penultimate position are the structural requirements that a substrate must possess to be susceptible to the carboxypeptidase action of ACE (Kim et al., 1983).

ACE inhibitors, which prevent the conversion of angiotensin I (AngI) to angiotensin II (AngII), have an extensive therapeutic history and have been used successfully for the treatment of hypertension (Faxon et al., 1982) and left ventricular remodeling (Konstam, 1995). The mechanism by which ACE inhibitors acutely lower blood pressure appears to involve a decrease in circulating AngII. Nevertheless, Bao et al. (1992) provided evidence that the fall in blood pressure following acute treatment with ACE inhibitors is accompanied by an increase in bradykinin, a potent vasodilator. The mechanism by which ACE inhibitors produce long-term reductions in blood pressure is less clear (Brunner et al., 1988) because circulating AngII levels return to normal in the face of lowered blood pressure. Determining the exact mechanism responsible for the effectiveness of chronic ACE treatment is complicated by the multiple functions of AngII and the promiscuity of ACE with regards to substrate and catalytic specificity.

The construction of ACE inhibitors was based primarily on structure-activity relationships that began with the isolation of several peptides from snake venom that inhibited the metabolism of bradykinin and AngI. All present-day ACE inhibitors rely on the C-terminal structure of BPP_5a, a bradykinin-potentiating pentapeptide (Cushman et al., 1987) that was originally isolated from venom and was determined to have strong interactions within the catalytic site of ACE (Cushman et al., 1977). Given the success of generating ACE inhibitors based on BPP_5a, the utilization of endogenous substrates to examine the catalytic sites has been largely ignored. In addition, endogenous substrates possess very low affinities for ACE, which makes them undesirable as models.

Angiotensinogen 3-11(Lys¹¹) was first synthesized in our laboratory to be utilized as a ligand to the recently discovered angiotensin IV (AT₄) receptor (Swanson et al., 1992). Traditionally, angiotensin fragments smaller than angiotensin III (AngIII) were considered to be biologically inactive. This was primarily due to low affinity for the AngII receptors (AT₁ and AT₂) and poor ability to elicit classic angiotensin-dependent physiological responses (Blair-West et al., 1971; Fitzsimons, 1971; Tonnaer et al., 1982). The AT₄ receptor was shown to preferentially bind the hexapeptide angiotensin IV (AngIV) (VYIHPF) (Swanson et al., 1992). Unlike the AngII receptors, which have been thoroughly characterized, much less is known about the AT₄ receptor. Therefore, our laboratory set out to determine the structure-binding characteristics of the AT₄ receptor by examining the binding properties of various AngIV analogs. These experiments revealed that the first three amino acids of AngIV (VYI) were critical for receptor binding. Furthermore, extending the C terminus with the three amino acids normally present in the angiotensinogen sequence (His⁹-Leu¹⁰-Val¹¹) had no significant effect on the binding. Using this information, it was hypothesized that a radioactive peptide (¹²⁵I-Y) could be modified at its C terminus to facilitate chemical cross-linking to the AT₄ receptor. This would allow the receptor to be visualized on gels and through various purification procedures. A lysine (Lys) was substituted for Val¹¹ (angiotensinogen numbering) so that the epsilon amine group could be used with the chemical cross-linker disuccinimidyl suberate (DSS) to covalently link ¹²⁵I-angiotensinogen 3-11(Lys¹¹) to the AT₄ receptor. Before this analog was utilized for tracing the purification of the AT₄ receptor, the specificity was examined. Surprisingly, it was determined that ¹²⁵I-angiotensinogen 3-11(Lys¹¹) bound to a second protein in addition to the AT₄ receptor. Further analysis suggested that this additional protein could be ACE. Thus, establishing that angiotensinogen 3-11(Lys¹¹) does bind to ACE was the focus of the present study.

In the present study we demonstrate that angiotensinogen 3-11(Lys¹¹) (VYIHPFHLK) possesses subnanomolar affinity for ACE and suggest that this analog could be a valuable tool for studying the physiochemical characteristics of the substrate binding domain of ACE. Ultimately, these studies could lead to the generation of highly specific ACE inhibitors.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents. All chemicals were purchased from Sigma Chemical Company (St. Louis, MO) except for Plummer's inhibitor (Calbiochem, La Jolla, CA) and bestatin (Peninsula Laboratories, Belmont, CA). AngI and bradykinin were also purchased from Sigma, whereas angiotensinogen 3-11(Lys¹¹) was synthesized in our laboratory. ¹²⁵I-351A was obtained from The Peptide Radioiodination Service Center (Washington State University, Pullman, WA). Guinea pig lungs, testes, and serum were purchased from Pel-Freez Biologicals (Rogers, AR).

Tissue Preparation. A modification of a previously described protocol was used for membrane preparation of guinea pig lungs for binding assays (Hanesworth et al., 1993). Briefly, the guinea pig lungs (5 g) were cut into several pieces and placed into 20 ml of hypotonic buffer (50 mM Tris and 1 mM EDTA, pH 7.4, at 4°C). The tissue was homogenized (Polytron; Brinkman Instruments, Westbury, NY) for 2 s/ml and centrifuged at 500g for 10 min at 4°C. The supernatant was decanted into a 40-ml centrifuge tube on ice, and the pellet was rehomogenized (2 s/ml) with 20 ml of hypotonic buffer. After centrifuging again at 500g for 10 min, the supernatants were combined and centrifuged at 40,000g for 20 min at 4°C. The supernatant was discarded, and the pellet was rehomogenized (2 s/ml) and then centrifuged a final time at 40,000g for 20 min at 4°C. The resulting pellet was homogenized and diluted to a final protein concentration of 1.25 mg of protein/ml using a standard protocol (Lowry et al., 1951).

Iodination of Angiotensinogen 3-11(Lys¹¹) and 351A. Angiotensinogen 3-11(Lys¹¹) and 351A were iodinated using a standard chloramine T method. The mono-[¹²⁵I]-angiotensinogen 3-11(Lys¹¹) and mono-[¹²⁵I]-351A were separated from the unlabeled and di-iodinated compounds using HPLC reversed-phase C₁₈ column (25-cm × 4.6-mm Microsorb-MV; Rainin Instrument Co., Woburn, MA). The mobile phase consisted of 250 mM phosphate buffered to pH 3.0 with triethylamine. A linear gradient (9-25%) using acetonitrile was used to separate the mono-iodinated from the di-iodinated compounds and for the unlabeled angiotensinogen 3-11(Lys¹¹). The separation of the mono-[¹²⁵I]-351A was achieved isocratically with 14% acetonitrile.

Binding Assays. Binding assays were performed at 22°C in duplicate with 250 µl of isotonic buffer (150 mM NaCl, 50 mM Tris, 5 mM EDTA, 50 µM Plummer's inhibitor, 20 µM bestatin, 0.1% heat-treated BSA, pH 7.4, at 22°C). Assays with guinea pig lung membranes contained 12.5 µg of protein and were terminated by vacuum filtration (Cell Harvester; Brandel Laboratories, Gaithersburg, MD), which separated the bound from the free ligands. Filters (no. 32; Schleicher & Schuell, Keene, NH) were washed three times with 4 ml of PBS, and the retained radioactivity was determined using a gamma counter (77% efficiency).

11

5

Purification of Somatic ACE from Guinea Pig Serum. The purification of guinea pig serum ACE was modified from a previously described protocol (Ryan, 1993). All of the steps were performed at room temperature except where noted. Briefly, 20 ml of guinea pig serum was dialyzed in 1.6 liters of 5 mM potassium phosphate buffer, pH 6.8, for 4 h. Precipitate formed from the dialysis was removed by centrifugation at 5000 g for 30 min at 5°C. The supernatant from the centrifugation was applied to a 35-ml column of Fast-Flow-Cibacron Blue 3GA equilibrated with 5 mM potassium phosphate buffer, pH 6.8. The eluant, which contained ACE, was mixed with hydroxylapatite equilibrated with 5 mM potassium phosphate buffer, pH 6.8, and separated by centrifugation (500g for 10 min at 5°C). The resin was washed with 10 ml of 5 mM potassium phosphate buffer, pH 6.8, and recentrifuged for 10 min at 500g, 5°C. The supernatant and wash were combined and applied directly to a 35-ml DEAE-cellulose column equilibrated with the same buffer. After eluting unretained proteins, a linear gradient was employed using 0.5 M potassium phosphate, 1.0 M NaCl, pH 6.8, as the limiting buffer to elute ACE. Fractions containing ACE were pooled and dialyzed in 0.5 M potassium phosphate, 1.0 M NaCl, pH 6.8, at 5°C. The dialysate was applied to a 10-ml phenyl-agarose column equilibrated with 0.5 M potassium phosphate, 1.0 M NaCl, pH 6.8. After eluting unretained proteins, a descending gradient was employed to elute ACE using 5 mM potassium phosphate buffer, pH 6.8, as the limiting buffer. The fractions containing ACE were pooled and concentrated using a spin concentrator (Macrosep 30). Concentrate was applied to an 88-ml Sephacryl S-200 HR column equilibrated with 10 mM HEPES buffer, 0.15 M NaCl, pH 7.4. ACE-containing fractions were pooled, and protein concentration was determined using a modified Lowry protein assay (Peterson, 1977). The eluant from this final column was assayed for its ability to compete for binding with ¹²⁵I-angiotensinogen 3-11(Lys¹¹) and ¹²⁵I-351A binding. Bound and free ligands were separated using Bio-Gel P-6 extra fine (Bio-Rad, Hercules, CA) (0.4-ml resin in a 1-ml syringe) spun at 500g for 10 min, 4°C. The unretained radioactivity (bound ligand) was determined with an ICN 10/880 gamma counter.

In Vitro Autoradiography. The protocol used for the in vitro autoradiography was previously described (Rowe et al., 1991). Briefly, the tissues were serially sectioned at a thickness of 20 mm using a cryostat (2800 E; Jung-Reichert) and were thaw-mounted onto gelatin-subbed slides. After preincubating the tissues for 30 min in the appropriate isotonic buffer [¹²⁵I-angiotensinogen 3-11(Lys¹¹): 150 mM, 50 mM Tris, 5 mM EDTA, 50 mM Plummer's inhibitor, 20 mM bestatin, 0.1% heat-treated BSA, pH 7.4, at 22°C; ¹²⁵I-351A: 50 mM HEPES, 150 mM NaCl, 10 mM ZnSO₄, pH 7.5 at 22°C] the tissues were placed in incubation jars containing labeled ligand at a concentration of 250 to 300 pM for 60 min at room temperature. Nonspecific binding was determined by adding unlabeled ligands at a concentration of 1 mM (guinea pig lung and testes). After incubation, the slides were washed three times, with 2 min for each washing, in isotonic buffer, dried, and exposed to X-ray film (Kodak SB-5). Film was exposed at 80°C and developed with D-19 developer (Kodak).

Solubilization and Cross-Linking. After following the tissue preparation protocol described above, solubilization of guinea pig lung ACE was performed at a tissue concentration of 10 mg/ml. Lung membranes were solubilized using 0.1% CHAPS and mixing for 2 h at 4°C. Solubilized lung membranes were then centrifuged at 60,000g for 1 h at 4°C. The supernatant was passed through a 0.22-mm filter, and aliquots were stored at 80°C until used. Binding of solubilized lung membranes was performed following the binding assay protocol described above, except that approximately 7 nM ¹²⁵I-angiotensinogen 3-11(Lys¹¹) was used to approach saturable conditions. The bound and free ligand were separated with Bio-Gel P-6 extra fine [in 10 mM Na₂HPO₄, 5 mM EDTA, 0.01% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid (CHAPS), and 0.02% NaN₃] and spun for 10 min at 500g at 4°C. After separating the bound and free ligand, cross-linking of bound ¹²⁵I-angiotensinogen 3-11(Lys¹¹) was performed with 1 mM DSS (in dimethyl sulfoxide) and incubated for 15 min on ice. The reaction was quenched by incubating 5 min on ice with 200 mM Tris, pH 7.4. ¹²⁵I-Angiotensinogen 3-11(Lys¹¹)-DSS cross-linked to ACE was separated from the excess DSS, Tris, dimethyl sulfoxide, and free ligand by recentrifugation with Bio-Gel P-6 extra fine. Samples were then methanol-precipitated, and the pellets were solubilized with loading buffer and analyzed by SDS-polyacrylamide gel electrophoresis (PAGE). Gels were dried and exposed to hyperfilm (Kodak) at 80°C and developed.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

¹²⁵I-Angiotensinogen 3-11(Lys¹¹) binding experiments were performed in the guinea pig lung membranes because of the large concentration of ACE in the lung. Figure 1 is a representation of the equilibrium saturation isotherms obtained with ¹²⁵I-angiotensinogen 3-11(Lys¹¹). Curves best fit a one-site binding model (r² = 0.975 ± 0.011) and demonstrate that the ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding is saturable in the presence of EDTA. It is important to note that no ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding is evident without EDTA. HPLC analysis indicated that without EDTA ¹²⁵I-angiotensinogen 3-11(Lys¹¹) is rapidly metabolized (data not shown). These data indicate that ¹²⁵I-angiotensinogen 3-11(Lys¹¹) possesses a high affinity, with K_d = 0.15 ± 0.02 nM and B_max = 4.30 ± 0.54 pmol/mg (n = 5, mean ± S.E.).

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Saturation isotherm performed in guinea pig lung membranes with 125I-angiotensinogen 3-11(Lys11). This representative curve is from five experiments. Assays contained 12.5 µg of total protein incubated for 120 min at 22°C. Nonspecific binding was determined by incubating 125I-angiotensinogen 3-11(Lys11) in the presence of 1 µM angiotensinogen 3-11(Lys11).

Competition curves were also performed using lung membranes and ¹²⁵I-angiotensinogen 3-11(Lys¹¹) with angiotensinogen 3-11(Lys¹¹), AngI, and bradykinin as competitors (Fig. 2). The rank order of affinity for the competitors was angiotensinogen 3-11(Lys¹¹) bradykinin angiotensin I. These data demonstrate that ACE substrates can compete, albeit poorly, for ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding. Bradykinin best fit a single-site binding model (bradykinin: r² = 0.997 ± 0.001). The concentrations used in the competition studies were insufficient to generate a curve with AngI. Unlike AngI and bradykinin, angiotensinogen 3-11(Lys¹¹) best fit a two-site binding model (f < 0.001; r² = 1.00 ± 0.001) and both sites were approximately 50% of the total bound protein (K_i values are listed in Table 1). Captopril was unable to compete for ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding at concentrations up to 10⁴ M. 

View larger version (20K): [in this window] [in a new window]   Fig. 2.   125I-Angiotensinogen 3-11(Lys11) competition curves with guinea pig lungs. Representative curves of angiotensinogen 3-11(Lys11) (*) and ACE substrates: , bradykinin; , angiotensin I). Guinea pig lung membranes (12.5 µg) were incubated for 60 min at 22°C with 0.6 nM 125I-angiotensinogen 3-11(Lys11). Competitor concentrations ranged from 1.00 × 1011 M to 3.16 × 105 M and were constructed by half-log dilutions (n = 3). See Table 1 for Ki values.

View this table: [in this window] [in a new window]   TABLE 1 Competition studies using 125I-angiotensinogen 3-11(Lys11) in guinea pig lung membranes Concentration of competitors ranged from 3.16 × 105 to 1 × 1011 M and were constructed by half-log dilutions. Angiotensinogen 3-11(Lys11) best fit a two-site binding model (r2 = 1.000 ± 0.0001; P < .0005). The percentage of the total bound receptor is recorded next to each site. The below data are the mean ± S.E. (n = 3).

DSS chemical cross-linking of ¹²⁵I-angiotensinogen 3-11(Lys¹¹) to solubilized lung membranes followed by SDS-PAGE allowed for the visualization of the proteins bound to ¹²⁵I-angiotensinogen 3-11(Lys¹¹). Figure 3 shows a typical autoradiogram. DSS-¹²⁵I-angiotensinogen 3-11(Lys¹¹) revealed a major band with a molecular weight of 173 ± 7.1 × 10³ (n = 3, mean ± S.E.). This band represented a specific ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding protein because excess unlabeled angiotensinogen 3-11(Lys¹¹) effectively competed with ¹²⁵I-angiotensinogen 3-11(Lys¹¹) for binding. In addition to the M_r 173,000 band, a minor band with an approximate molecular weight of 120,000 was also observed. Additional analysis indicated that the mobility of the M_r 173,000 band was unaffected by reducing agents, suggesting that it did not represent a multimeric protein complex (data not shown).

View larger version (59K): [in this window] [in a new window]   Fig. 3.   SDS-PAGE lung membranes. Representative gel from DSS cross-linking 125I-angiotensinogen 3-11(Lys11) in solubilized guinea pig lung membranes (lane 1, total binding). Nonspecific binding was determined by incubating in the presence of 1 µM angiotensinogen 3-11(Lys11) (lane 2, nonspecific binding). Solubilized lung membranes (50 µg) were incubated for 60 min at 22°C with approximately 7 nM 125I-angiotensinogen 3-11(Lys11). See text for additional details.

A comparison of the distribution of ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding to the ACE inhibitor ¹²⁵I-351A was performed using in vitro autoradiography with serial sections of guinea pig lung (Fig. 4) and testes (Fig. 5). Autoradiograms demonstrate that ¹²⁵I-angiotensinogen 3-11(Lys¹¹) and ¹²⁵I-351A have identical distributions in the lung (Figs. 4A and 4E, respectively) and testes (Fig. 5, A and E, respectively). Specifically, ¹²⁵I-angiotensinogen 3-11(Lys¹¹) and ¹²⁵I-351A binding were evenly distributed throughout the lung. Nonspecific binding within the lung was negligible (Fig. 4, B and F). As expected, bradykinin was able to compete for ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding in the lung but to a lesser extent than angiotensinogen 3-11(Lys¹¹) (Fig. 4C). Furthermore, the localization of ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding did not appear to change in the presence of bradykinin. In addition, angiotensinogen 3-11(Lys¹¹) at a concentration of 1 µM was able to compete for ¹²⁵I-351A without an apparent effect on the localization of its binding (Fig. 4G). Finally, AngIV (1 µM) did not compete or change the localization of ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding, indicating that there are very few AT₄ receptors in the guinea pig lung (Fig. 4D).

View larger version (65K): [in this window] [in a new window]   Fig. 4.   Autoradiographic distribution of 125I-angiotensinogen 3-11(Lys11) and 125I-351A in guinea pig lung. Serial sections (20 mm) were incubated for 60 min at 22°C with 250 to 300 pM radiolabel. A, total binding of 125I-angiotensinogen 3-11(Lys11); B, 125I-angiotensinogen 3-11(Lys11) in the presence of 1 µM angiotensinogen 3-11(Lys11); C, 125I-angiotensinogen 3-11(Lys11) in the presence of 1 µM bradykinin; D, 125I-angiotensinogen 3-11(Lys11) in the presence of 1 µM angiotensin IV; E, total binding of 125I-351A; F, 125I-351A in the presence of 1 µM captopril; and G, 125I-351A in the presence of 1 µM angiotensinogen 3-11(Lys11).

View larger version (104K): [in this window] [in a new window]   Fig. 5.   Autoradographic distribution of 125I-angiotensinogen 3-11(Lys11) and 125I-351A in guinea pig testis. Serial sections (20 mm) were incubated for 60 min at 22°C with 250 to 300 pM radiolabel. A, total binding of 125I-angiotensinogen 3-11(Lys11); B, 125I-angiotensinogen 3-11(Lys11) in the presence of 1 µM angiotensinogen 3-11(Lys11); C, 125I-angiotensinogen 3-11(Lys11) in the presence of 1 µM bradykinin; D, 125I-angiotensinogen 3-11(Lys11) in the presence of 1 µM angiotensin IV; E, total binding of 125I-351A; F, 125I-351A in the presence of 1 µM captopril; G, 125I-351A in the presence of 1 µM angiotensinogen 3-11(Lys11).

Autoradographic analysis of ¹²⁵I-angiotensinogen 3-11(Lys¹¹) and ¹²⁵I-351A in the testis revealed nearly identical results as the lung. The distribution was identical for both ligands with intense binding in the body of testis containing the seminiferous tubules. Nonspecific binding was negligible for both ligands (Fig. 5, B and F). As with the lung, bradykinin (1 µM) was able to compete with ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding, but to a lesser extent (Fig. 5C). The displacement of ¹²⁵I-angiotensinogen 3-11(Lys¹¹) by bradykinin did not alter the localization as similarly seen in the lung. AngIV (1 µM) competed minimally with ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding (Fig. 6D), indicating that little of the ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding was to AT₄ receptors. Furthermore, angiotensinogen 3-11(Lys¹¹) was also able to compete for ¹²⁵I-351A, although to a lesser extent than observed in the lung.

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Purification of somatic ACE from guinea pig serum. A, representation of the binding elution profile for 125I-angiotensinogen 3-11(Lys11) and 125I-351A from a DEAE-cellulose column. B, representation of the binding elution profile for 125I-angiotensinogen 3-11(Lys11) and 125I-351A from a phenyl agarose column. C, representation of the binding elution profile for 125I-angiotensinogen 3-11(Lys11) and 125I-351A from an S200 HR column. A sample from each fraction was used to test for binding of 125I-angiotensinogen 3-11(Lys11) and 125I-351A. Samples were incubated with 0.6 nM radiolabel for 60 min at 22°C. Nonspecific binding was determined by incubating radiolabel with the presence of 1 µM competitor. See text for additional details.

Purification of guinea pig serum ACE was performed to demonstrate that the M_r 173,000 protein that bound ¹²⁵I-angiotensinogen 3-11(Lys¹¹) copurified with ACE. The purification protocol typically yielded a 170-fold purification with a 15% recovery of ACE (Table 2). The purification of ACE was monitored with ¹²⁵I-351A and ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding. Figure 6 provides examples of elution profiles from several columns demonstrating that the elution of binding activity for ¹²⁵I-angiotensinogen 3-11(Lys¹¹) and ¹²⁵I-351A were nearly identical.

View this table: [in this window] [in a new window]   TABLE 2 Purification of serum ACE Yield was based on saturation isotherms using 125I-angiotensinogen 3-11(Lys11) in serum and purified ACE. Nonspecific binding was determined using 1 µM angiotensinogen 3-11(Lys11). Protein concentration of serum was determined using the method of Lowry (1951). Protein concentration of purified ACE was determined using a modified Lowry protein assay (Peterson, 1977).

Figure 7 shows an SDS-PAGE gel of the partially purified ACE from guinea pig serum. The gel demonstrates that the purification protocol used was sufficient to remove the majority of smaller molecular weight proteins. Several proteins of high molecular weight, however, were copurified and appear to have equivalent concentrations. Furthermore, this purification protocol was sufficient to remove the M_r 120,000 ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding protein. This M_r 120,000 band was evident in the lung membranes (Fig. 3) and DSS-¹²⁵I-angiotensinogen 3-11(Lys¹¹) experiments in whole serum (Fig. 8).

View larger version (72K): [in this window] [in a new window]   Fig. 7.   SDS-PAGE of purified ACE. Representative gel assessing the purification of serum ACE. Lane 1, 2.0 µg of purified ACE; lane 2, 1.0 µg of purified ACE; lane 3, 0.5 µg of purified ACE. Protein was loaded onto a 6% polyacrylamide gel and visualized by silver staining.

View larger version (58K): [in this window] [in a new window]   Fig. 8.   SDS-PAGE of whole guinea pig serum. Representative gel from DSS cross-linking 125I-angiotensinogen 3-11(Lys11) in whole guinea pig serum. Lane 1, nonspecific binding; lane 2, total binding. Nonspecific binding was determined by incubating in the presence of 1 µM angiotensinogen 3-11(Lys11). Serum (500 µg) was incubated for 60 min at 22°C with approximately 7 nM 125I-angiotensinogen 3-11(Lys11) for both total and nonspecific binding. See text for additional details.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The data presented in this paper describe a novel angiotensin I analog, angiotensinogen 3-11(Lys¹¹), that binds with high affinity to guinea pig lung ACE. Guinea pig lung membranes were chosen as a source of ACE for two reasons. First, lungs have been shown to contain a large concentration of ACE. Ng and Vane (1970) found that up to 80% of angiotensin I at physiological concentrations was converted to angiotensin II in the lung. Secondly, guinea pig lungs contain a small number of AT₄ receptors that also bind angiotensinogen 3-11(Lys¹¹) (see the Introduction).

Analysis of the competition curves using ¹²⁵I-angiotensinogen 3-11(Lys¹¹) revealed that the rank order of affinity of the peptides examined was angiotensinogen 3-11(Lys¹¹) bradykinin AngI. These data are consistent with the idea of angiotensinogen 3-11(Lys¹¹) binding ACE and, demonstrate that it possesses a higher affinity than the endogenous substrates. The observed rank order is supported by previous reports that have proposed that bradykinin, which possesses a 100-fold higher affinity than angiotensin I for ACE, is the preferred substrate (Bull et al., 1985). This is in good agreement with our observation that bradykinin has a 200-fold higher affinity than angiotensin I for the ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding site. The reason for the exceptionally high affinity of angiotensinogen 3-11(Lys¹¹) is presently unknown. Interesting, however, is the fact that angiotensinogen 3-11(Lys¹¹) possesses a combination of structural features found in either AngI or bradykinin. Although angiotensinogen 3-11(Lys¹¹) obviously contains eight amino acids common to AngI, it does have several bradykinin-like features. Both bradykinin and angiotensinogen 3-11(Lys¹¹) possess a positively charged polar amino acid at the C terminus. There are also three nonpolar uncharged amino acid side chains located in the same position in angiotensinogen 3-11(Lys¹¹) and bradykinin. Therefore, four of the nine amino acid side chains of angiotensinogen 3-11(Lys¹¹) have similar properties as bradykinin.

EDTA is a metal chelator and has been shown to inactivate ACE (Skeggs, 1956). Without EDTA in the binding assays, there is no specific binding of ¹²⁵I-angiotensinogen 3-11(Lys¹¹) to guinea pig lung membranes or purified ACE. This observation is consistent with the notion that angiotensinogen 3-11(Lys¹¹) is a substrate for active ACE but does not rule out the possibility that ¹²⁵I-angiotensinogen 3-11(Lys¹¹) is metabolized by other enzymes. The addition of EDTA to the isotonic buffer allowed for an assessment of ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding to ACE without the confounding action of active catalysis. It has been demonstrated that the structural integrity of the catalytic site is unaltered in the apoenzyme state, indicating that zinc is not involved in stabilizing the enzyme-substrate complex (Ryan, 1978). Furthermore, Odya et al. (1983) demonstrated that bradykinin can bind to the ACE apoenzyme. In fact, they reported that the IC₅₀ for bradykinin was 1.6 × 10⁸ M in the porcine kidney ACE. They also examined angiotensin I and found that the IC₅₀ was 3.2 × 10⁶ M. These data are in agreement with the rank order observed in this study. Absolute differences in IC₅₀ values may reflect the varied sources of enzyme used in this and other studies.

Analysis of the competition curves with angiotensinogen 3-11(Lys¹¹) revealed a two-site binding model with approximately 50% of the total bound protein attributed to either site. This observation encourages speculation that angiotensinogen 3-11(Lys¹¹) might possess different affinities for the two catalytic domains of ACE. Although there is a degree of high homology between the catalytic domains of ACE, evidence does exist demonstrating that there are structural differences in regards to the catalytic activity of ACE and substrate binding (Corvol et al., 1995). An alternative explanation for the observed two-site fit is that the preparation of guinea pig lung membranes used for binding studies may have resulted in the partial solubilization of ACE, which could have reduced affinity when compared with membrane-bound ACE. This possibility is supported by preliminary competition studies performed with purified serum (soluble) ACE where the K_i values obtained for angiotensinogen 3-11(Lys¹¹), as well as bradykinin, were lower than those observed with lung membranes and fit a single-site model (data not shown).

Angiotensinogen 3-11(Lys¹¹) could also be binding to another protein. The cross-linking experiments with solubilized lung membranes revealed a major band that had a molecular weight of M_r 173,000, which is very similar to the M_r 171,000 that has been previously reported for guinea pig serum ACE (Ripka et al., 1993). However, an additional band with a lower molecular weight, M_r 120,000, was also present, indicating that there is more than one protein that specifically binds ¹²⁵I-angiotensinogen 3-11(Lys¹¹). Cross-linking experiments with whole serum also revealed a similar lower molecular weight band.

The ¹²⁵I-angiotensinogen 3-11(Lys¹¹) competition data are different from the results obtained from the equilibrium saturation isotherms. The competition curves with angiotensinogen 3-11(Lys¹¹) exhibited a two-site model, whereas the saturation isotherms best fit a single-site model. This suggests the possibility that the ¹²⁵I label on the tyrosine improves the affinity for only one of the two ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding sites. This idea is supported by the demonstration that the affinity of the high-affinity site from the competition curves (K_i = 2.20 ± 0.82 × 10¹⁰ M) is similar to that obtained from saturation isotherms (K_d = 1.46 ± 0.20 × 10¹⁰ M).

The in vitro autoradiography presented in this paper demonstrates that ¹²⁵I-angiotensinogen 3-11(Lys¹¹) binding sites have a similar distribution to those observed for the ACE inhibitor ¹²⁵I-351A in the guinea pig lung and testis. Furthermore, angiotensinogen 3-11(Lys¹¹) was able to compete for the binding of ¹²⁵I-351A, indicating that they interact with the same protein. The incomplete displacement of ¹²⁵I-351A could be due to the degradation of angiotensinogen 3-11(Lys¹¹) because the binding conditions (no EDTA) that were employed supported ACE in the haloenzyme state. The absence of EDTA in the incubation buffer used for ¹²⁵I-351A autoradiography were necessary because the binding of 351A to ACE requires zinc at the catalytic site (Bull et al., 1985). As expected from its lower affinity, bradykinin also was shown to compete for binding, but to a lesser extent than angiotensinogen 3-11(Lys¹¹). Similar results were also observed in the guinea pig testes (data not shown). These data are consistent with results from the competition studies that demonstrate that angiotensinogen 3-11(Lys¹¹) possesses a greater affinity than bradykinin to ACE.

The purification of guinea pig serum ACE revealed nearly identical elution profiles for proteins that bound ¹²⁵I-angiotensinogen 3-11(Lys¹¹) and ¹²⁵I-351A consistent with both ligands binding the same target, ACE. The marginal difference that was observed in the elution profile of the phenyl-agarose column could be due to the chloride (which varies during elution) differentially affecting the binding of both ligands. Because chloride concentration has effects on catalytic activity and substrate binding by acting at the zinc site, it would be expected to differentially effect ¹²⁵I-351A binding (Skeggs, 1956; Wei, 1991, 1992). Another plausible explanation for the marginal difference is the presence of isozymes. Ryan et al. (1993) observed several isozymes during the purification of guinea pig serum ACE.

The major goal of this research is to understand how substrates interact with ACE. Ultimately, the objective is to define the importance of specific amino acid side chains in determining substrate affinity and ACE specificity. The development of angiotensinogen 3-11(Lys¹¹) as a high-affinity probe is a first step in this endeavor. Next, the individual amino acids of angiotensinogen 3-11(Lys¹¹) will be systematically altered to establish their contribution to overall substrate affinity. Data already available from our research group indicate that modifications of the valine and tyrosine in angiotensinogen 3-11(Lys¹¹), which are outside the cleavage site region, dramatically alter affinity for ACE. Therefore, these data suggest that the interaction of ACE and substrate extends beyond its catalytic site to other regions of the protein. Once the optimal structure of an ACE ligand has been deduced, further replacement of metabolically susceptible bonds, like the Phe-His cleavage site, with nonpeptide linkages could yield high-affinity, specific, and metabolically resistant ACE ligands.