Angiotensin-converting enzyme (ACE) is present on the luminal surface of the coronary vessels, mostly on capillary endothelium. ACE is also expressed on coronary smooth muscle cells and on plaque lipid-laden macrophages. Excessive coronary circulation (CC)-ACE activity might be linked to plaque progression. Here we used the biologically inactive ACE substrate3H-labeled benzoyl-Phe-Ala-Pro ([3H]BPAP) to quantify CC-ACE activity in 10 patients by means of the indicator-dilution technique. The results were compared with atherosclerotic burden determined by coronary angiography. There was a wide range of CC-ACE activity as revealed by percent [3H]BPAP hydrolysis (30–74%). The atherosclerotic extent scores ranged from 0.0 to 66.97, and the plaque area scores ranged from 0 to 80 mm2. CC-ACE activity per unit extracellular space (V max/K m V i), an index of metabolically active vascular surface area, was correlated with myocardial blood flow (r = 0.738;P = 0.03) but not with measures of the atherosclerotic burden. These results show that CC-ACE activity can be safely measured in humans and that it is a good marker of the vascular area of the perfused myocardium. It does not, however, reflect epicardial atherosclerotic burden, suggesting that local tissue ACE may be more important in plaque development.
- coronary disease
- endothelial function
large randomized trials have established that angiotensin-converting enzyme (ACE) inhibitors (ACEIs) reduce recurrent myocardial infarctions (18). ACEIs confer similar benefits on patients with evidence of vascular disease or risk factors for coronary artery disease (CAD) (31). At least part of these benefits may be related to improvement of coronary endothelial function (15).
ACE, a dipeptidyl carboxypeptidase that converts ANG I to ANG II, is mostly present on the luminal surface of the vascular endothelium, particularly in the lung (3, 24). In the atherosclerotic heart, ACE is present not only on the coronary endothelium (11,19) but also on coronary medial smooth muscle cells and plaque lipid-laden macrophages (8). In the noninfarcted myocardium of both humans and rats, ACE is prominent on the endothelium of arterioles and capillaries, with almost no activity in venous vessels (11). Coronary circulation (CC)-ACE may contribute to plaque development because there is experimental evidence to support the hypothesis that ACEIs have antiatherogenic properties when administered to hypercholesterolemic animals (5, 25). It may also contribute to acute plaque instability because coronary atherectomy specimens from patients with acute coronary syndromes display higher ACE activity (14). Because coronary endothelial dysfunction occurs early in the development of CAD (28) and correlates with the number of risk factors for CAD (29), it is possible that CC-ACE activity would be modified by the severity of the disease process or could itself modulate the development of CAD. Little is known, however, about the role of CC-ACE in the pathophysiology of CAD. Such an evaluation could help in establishing the relative importance of endothelium-bound versus tissue ACE activity in CAD.
The hemodynamically inactive ACE substrate 3H-labeled benzoyl-Phe-Ala-Pro ([3H]BPAP) has already been used in the quantification of endothelium-bound ACE activity of human pulmonary vascular bed (17) as well as in the coronary vasculature of animals (13) by means of the indicator-dilution technique. This technique has been well validated in measuring pulmonary capillary endothelium-bound ACE activity that relates the perfused capillary surface area in animals (20) and in humans (17). In dogs, this approach has been suggested as a means of evaluating the area of the perfused coronary capillary surface (13).
The aim of this study was to quantify CC-ACE activity in humans and to compare it with the coronary atherosclerotic burden by measuring [3H]BPAP kinetics combined with quantitative coronary angiography (QCA) analysis.
MATERIALS AND METHODS
Ten patients participated in a protocol approved by the Research and Ethics Committee of the Montreal Heart Institute after written informed consent was obtained. They were recruited from a group of patients already scheduled for diagnostic coronary angiography. Patients with known ejection fraction of <40% or valvular heart disease, as well as those receiving ACEI or ANG II receptor antagonist therapy, were excluded. However, one subject taking the ACEI perindopril (2 mg b.i.d.) was mistakenly studied. We nevertheless report the results of the ACE kinetic analysis from this patient for informative purposes but have excluded it from the correlation analysis.
Venous and arterial femoral sheaths were inserted. Standard projection coronarographies were performed after the administration of intracoronary nitroglycerin. At the end of the procedure, a left Judkins catheter was positioned in the left main coronary artery to serve as the injection site for the indicator-dilution curve. A multipurpose catheter was then positioned in the coronary sinus and connected to a collection pump.
All patients received one rapid bolus injection through the catheter placed in the left main coronary artery. The bolus was prepared by combining 30 μCi of [3H]BPAP (generous gift of Dr. J. W. Ryan) and 2 μCi of [14C]sucrose (NEN, Boston, MA) with 3 ml of 0.9% saline. Two milliliters of the bolus was used for injection and the remainder for the preparation of dilution curve standards. The 2-ml bolus was introduced into an extension line connected to the left coronary catheter. The bolus was injected, and the injection line was flushed with a syringe containing 10 ml of the patient's blood. Collection of serial blood samples from the coronary sinus was simultaneously initiated by a peristaltic pump at 80 ml/min (MasterFlex L/S) and a fraction collector (Pharmacia LKB Frac-100) advancing at one tube per 1.7 s. Each fraction collector tube contained 1 ml of “stop” solution to inactivate any serum ACE activity (6.8 mM 8-hydroxyquinoline-5-sulfonic acid) and 10 μl of 1,000 IU/ml heparin to prevent blood clotting. After centrifugation (3,000 rpm, 10 min), 0.3 ml of the clear supernatant was transferred into a scintillation vial containing 1 ml of HCl (0.1 M) and 1 ml of toluene omnifluor (0.4%). The 3H extracted into the toluene phase was measured. The next day, 14 ml of aqueous scintillation liquid (Ready Safe, Beckman Instruments, Fullerton, CA) was added to convert the two-phase toluene-aqueous mixture into a single-phase solution and total 3H and 14C activities were measured. The counts were corrected for background and crossover activities. With this technique, the fractional extraction of substrate or intact [3H]BPAP (f s) in the toluene phase is 5% and that of its product of hydrolysis (f p) [3H]benzoyl-Phe (3H-BP) is 60%. The amount of product in each sample can then be calculated as [3H]BP = ([3H]toluene − f s × [3H]total)/f p −f s, where f s andf p are the fractional extractions of substrate ([3H]BPAP) and product ([3H]BP), respectively, into the toluene layer.
Coronary atherosclerosis was quantitatively assessed by QCA. All analyzable arteries of >1.5 mm were evaluated. The number, location, and severity of lesions were recorded, and the data were analyzed to determine an extent score (ES). This ES was developed to indicate the angiographically detectable atherosclerotic burden of the coronary arterial tree. The coronary vessels were divided into segments as defined in the bypass angioplasty revascularization study (BARI; Ref.27). Each segment received a multiplication factor in proportion to its caliber and territory as inspired by the extent score described by Sullivan et al. (26). Practically, the first three segments of the right coronary artery (segments 1,2, 3), the proximal and mid-left anterior descending coronary artery (segments 12, 13), and the ramus intermedius (segment 28) each received a factor of 20. The proximal, mid-, and distal circumflex artery (segments 18, 19, 19.1), the first three diagonal branches (segments 15, 16, 29), all obtuse marginal branches (segments 20–22), and all posterolateral branches (left or right; segments 4–8,24–27) received a factor of 10. The acute marginal branches (segment 10) and ventricular branches (segment 23) received a factor of 5. The left main coronary artery (segment 11) received a factor of 5. The percent diameter stenosis computed by QCA for each segment was multiplied by the corresponding factor. Where multiple stenoses were present in one segment, each was multiplied by the same factor. When a vessel was occluded and the distal vessel was not visualized by collateral flow, the proportion of vessel not visualized was given the mean ES of the remaining vessels. Although the entire coronary tree was analyzed, only the territory irrigated by the left coronary artery was reported in this study, as the area of interest. All the values obtained from the left coronary system were therefore added up to determine the ES.
In addition, the total plaque area (PA) was reported for the entire left coronary tree. The PA was calculated as the integral of the distances between the luminal contours and the original size of the vessel at the stenosis before disease occurred (computer-derived interpolated reference technique), within the given obstruction limit (in mm2) (21).
Mathematical modeling for determination of coronary endothelium-bound ACE activity.
Using indicator-dilution techniques, we estimated under first-order reaction conditions the single-pass transcoronary hydrolysis of the synthetic ACE substrate [3H]BPAP by CC-ACE. Figure1 represents the model.
Modeling of sucrose kinetics.
The kinetics of sucrose transport within the coronary circulation was described by a barrier-limited model, assuming heterogeneous transit times of large vessels and capillaries that are linearly related to each other (23). Sucrose is a diffusible substance that may leave the capillaries to enter the extravascular space and later return to the vasculature. We apply here a reformulation of the original equations recently described in an investigation of serotonin kinetics in the lung (10). The transit time of a path through the coronary circulation can be partitioned into a capillary part (τc), where exchange between the vascular and the extravascular space takes place, and a large-vessel part (τl), comprising arterioles and venules, where no such exchange occurs. The assumption of a linear relationship between the capillary transit time τc and the large-vessel transit time τl as a function of time of passage through the heart t is formulated by the following relationship where b and t 0 are constants, with 0 < b <1. The parametert 0 is related to the distribution of transit times, and b is an index of transit time heterogeneity through the heart.
The kinetic analysis requires data from a reference indicator that does not leave the vasculature. As discussed below, labeled [3H]BPAP and its labeled product, [3H]BP, remain in the vasculature, so that the sum of their concentrations can be used as the vascular reference. The fractional recovery of sucrose can then be expressed as Equation 1 where k pi and k ipare sucrose transfer coefficients for transfer from plasma to the interstitial space and return to the plasma space, respectively. In the above equation, variables b and k piappear as a product and cannot be identified separately.C Ref(t) is obtained by deconvoluting the catheter function from the measured reference curve, andC Suc(t) is convoluted with the catheter function before fitting to the measured sucrose curve, as previously developed (10).
Optimized parameters were obtained by a nonlinear least-squares procedure from Visual Numerics (Houston, TX). Data were weighted according to the inverse of the square root of their values.
Modeling of BPAP kinetics.
BPAP exists in the injection bolus as an equilibrium mixture ofcis- and trans-isomers. In the capillaries, thetrans-isomer is split by ACE, which is located at the endothelial surface (Ref. 16; Fig. 1).Cis-trans isomerization is slow compared with the time scale of the present data. Attempts to take into account isomerization within capillaries did not change the results of the analysis; hence, it was ignored in the present model. The outflow of [3H]BPAP can then be expressed as Equation 2where θ is the previously measured value for thecis-fraction (20) andk hyd is the transfer coefficient for enzymic conversion of the trans-isomer. Catheter distortion was taken into account, as in the case of sucrose, andt 0 was set to the value obtained from the sucrose analysis.
The [14C]sucrose diffuses into the interstitial space of the myocardium and is completely recovered without being metabolized. Its transit time relative to that of the tracer that remains within the vascular space was previously validated to derive a parameter (Φ) proportional to myocardial blood flow (22). Because the outflow profile for total 3H is identical to that of a vascular tracer like albumin, Φ was computed as flow per unit extracellular space as previously described (9) and is given by where Ref and Suc are the mean transit times of the vascular reference and sucrose, respectively. The curves were extrapolated with power law functions (1). Because in some experiments extrapolation led to large or infinite sucrose transit times (presumably because sucrose curves did not extend far enough in time to provide sufficient information on tail behavior), parameters from the following equation (23) were used to estimate sucrose mean transit times The rate of hydrolysis of BPAP is determined by the kinetic constants of the enzyme, i.e., V max andK m. At tracer concentrations of BPAP that are small compared with K m, the ratioV max/K m is equivalent to intrinsic clearance or to the permeability-surface area product for a permeating substance. In analogy to the latter case, the relation between V max/K m per unit extracellular space for BPAP hydrolysis and k hydcan than be found as follows (10) This can be expressed as a function of the parameters identified in the sucrose and BPAP curve analysis
Individual data are reported as well as means ± SD. Where appropriate, relationships between parameters were evaluated with Spearman's correlations. Differences were considered significant at a value of P < 0.05.
Representative myocardial outflow profiles of total3H, the surviving [3H]BPAP, and the corresponding ACE hydrolysis product [3H]BP are depicted in Fig. 2. Hydrolysis of [3H]BPAP is evident, with a progressively lower outflow profile compared with total 3H. In this example taken frompatient 4, percent [3H]BPAP hydrolysis by CC-ACE was 68%. The myocardial interstitial space tracer [14C]sucrose is delayed compared with total3H and reaches a smaller and later peak by virtue of its larger volume of distribution. Data derived from the indicator-dilution curves and optimized parameters derived from the modeling of [3H]BPAP kinetics are presented in Table1. The patient taking ACEI therapy (patient 8) demonstrated very low [3H]BPAP hydrolysis of 15%, whereas the others had variable degrees of hydrolysis ranging from 30 to 74%. Patient 8 was excluded from the correlation analysis. The individual coronary atherosclerosis burden scores derived from the QCA analysis are also included in Table 1. The ES ranged from 0.0 to 66.97, and the PA ranged from 0 to 80.
An example of the fit for [3H]BPAP outflow profile from modeling of the experimental data is shown in Fig.3. There was a very good quality of fit except for the tail end of the curves, where we observed a small deviation from the experimental data. There was a direct correlation between the estimate of coronary blood flow (Φ) and CC-ACE activity as measured by the parameterV max/K m V i(r = 0.738, P = 0.033; Fig.4).
ES and PA of the left coronary tree, however, did not significantly correlate with CC-ACE activity (PA: r = −0.417,P = 0.238; ES: r = −0.400,P = 0.257).
BPAP is a pharmacologically inactive high-affinity ACE substrate. It is rapidly hydrolyzed by ACE present at the luminal surface of the vascular endothelium so that its volume of distribution is restricted to the vascular space during a single transcoronary passage (4). Using BPAP, we demonstrated the feasibility of measuring CC-ACE activity in human coronary vessels in vivo. In the rat myocardium, autoradiographic studies revealed linear streaks of high ACE activity corresponding to sections of coronary arteries (30); assuming that ACE is uniformly distributed throughout coronary small vessels and capillaries, measurements of CC-ACE activity could be a useful tool in estimating the perfused capillary bed in humans. The results from the patient (patient 8) mistakenly studied while receiving ACEI therapy not only validate our methodology but demonstrate for the first time the almost complete inhibition of CC-ACE activity in a human subject taking this therapy.
It has already been shown in several indicator-dilution experiments in animals and in humans that the capillary permeability-surface area products of select solutes and metabolizable tracers such as norepinephrine change in proportion to coronary flow changes (2,6, 7, 9, 22). This increase with flow of the coronary capillary exchange surface area reflects capillary recruitment and reaches an asymptotic maximum at the highest level of blood flows (22). Using BPAP as a marker of the metabolically active coronary vascular bed, Horvath et al. (13) demonstrated in dogs that the reduction of CC-ACE activity corresponded to reduction in myocardial blood flow induced by coronary artery ligation. Our data support this view in human coronary arteries, because we practically found a linear relationship between CC-ACE activity and coronary flow. Because the patients were in the resting state, however, we did not reach the asymptotic maximum in CC-ACE activity that could possibly be achieved with higher coronary blood flows. Our findings therefore support the concept that a reduction in coronary flow reduces the surface area available for microcirculatory exchanges as a consequence of capillary derecruitment (12).
In coronaries from patients with advanced atherosclerotic lesions, the inflammatory cells of the endothelium of the neovasculature present in the plaques exhibit substantial ACE activity (8). We found a wide variability in luminal CC-ACE activity, but this variability did not significantly correlate with the atherosclerotic burden measured by QCA. Although the sample size in our study is too small to allow definite conclusions, the data suggest that luminal ACE expression is determined by factors other than atherosclerotic burden. Because atherosclerosis is mainly a disease of large- and medium-sized vessels, its presence does not seem to alter the activity of CC-ACE at the level of small vessels and capillaries. The linear correlation found between the CC-ACE activity and the myocardial blood flow, and the lack of influence by the atherosclerotic burden, suggest that CC-ACE is practically uniformly distributed in the coronary microcirculation and is not influenced by the disease process. The known increase in endothelial plaque ACE may only represent a minor contribution to overall myocardial CC-ACE activity, and because this activity may be restricted to plaque content it is likely that it is not accessible to a vascular substrate and hence was not measured in the present study.
In conclusion, we showed that CC-ACE activity could be safely measured in humans. This activity is proportional to myocardial blood flow and, presumably, the metabolically active surface of perfused coronary capillaries. However, we found that the variability in CC-ACE activity could not be related to the coronary atherosclerotic burden. This suggests that coronary luminal ACE activity may be less important than tissue ACE activity in the pathophysiology of coronary atherosclerosis.
Address for reprint requests and other correspondence: J. Dupuis, Montreal Heart Institute, Research Center, 5000 Belanger St. East, Montreal, Quebec, Canada H1T 1C8 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2003 the American Physiological Society