INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

IT IS WELL RECOGNIZED that the endothelium is a major determinant of vascular tone through release of different relaxing and constricting factors that modulate the contractile activity of the underlying smooth muscle (7, 22). Abnormal endothelium-dependent coronary vasomotion has been demonstrated with the development of atherosclerosis (30), heart failure (27), and hypertension (6, 21). Impairment of endothelial function results in abnormal dilator or constrictor responses to humoral agents or during stress (8, 26, 29). These alterations could enhance the susceptibility of the myocardium to ischemic injury.

It is assumed that left ventricular (LV) hypertrophy (LVH) is characterized by coronary vascular dysfunction that renders the myocardium more susceptible to ischemia (18), since coronary reserve is reduced in LVH (5, 13, 14, 17). Furthermore, studies in systemic hypertension-induced LVH demonstrate impaired coronary endothelial regulation (24). To date, no information is available on the ability of the coronary endothelium to regulate the coronary vasculature in pressure overload-induced LVH in the absence of systemic hypertension, which often involves neurohumoral changes, most importantly in the renin-angiotensin system. The present study was designed to investigate coronary endothelial control in conscious dogs with supravalvular aortic banding-induced chronic coronary pressure overload (CCPO). The following two approaches were used: assessment of large coronary reactivity with measurement of coronary artery diameter and assessment of resistance vessel reactivity with measurement of coronary blood flow. We employed the following three stimuli: 1) intravenous administration of acetylcholine to assess endothelial receptor-mediated dilation; 2) a 20-s coronary artery occlusion, which, upon release, elicits reactive hyperemia and reactive dilation (12), i.e., endothelial flow-mediated dilation; and 3) nitroglycerin, which should dilate the coronary vessels independently of the endothelium. The physiological approach was complemented by pathological analyses to determine more precisely if histological lesions were associated with the deranged physiological responses.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Development of the model. Mongrel puppies of either sex at 8-10 wk of age were anesthetized with 12.5 mg/kg sodium thiamylal maintained with halothane anesthesia (1-2 vol/vol) and were ventilated with a respirator (Harvard Apparatus, South Natick, MA). A right thoracotomy was performed through the fourth intercostal space by use of a sterile surgical technique. The ascending aorta above the coronary arteries was isolated and dissected free of surrounding tissue. A 1-cm-wide Teflon cuff was placed around the aorta and tightened until a thrill was palpable over the aortic arch. Next, the chest was closed. The Teflon band created a fixed supravalvular aortic lesion, which became relatively more stenotic as the puppies grew.

Implantation of instrumentation. Seven adult aortic-banded dogs and seven control dogs (2 nonbanded but sham-operated littermates and 5 mongrel nonoperated dogs) were instrumented at 12 ± 1 mo of age. After induction with sodium thiamylal (12.5 mg/kg) and maintenance with halothane anesthesia (1-2 vol/vol), an incision was made in the fifth left intercostal space by use of a sterile surgical technique. All dogs were instrumented with Tygon catheters (Norton Elastics and Synthetic Division, Akron, OH) implanted in the descending thoracic aorta and the left atrium and in the LV apex. A pair of miniature 7-MHz ultrasonic transducers was attached to Dacron backing and sutured using Prolene 5-0 suture (Ethicon) to opposing surfaces of the left circumflex coronary artery, 3-6 cm from its origin. Alignment of the crystals was ensured at surgery by monitoring the ultrasonic signal with an oscilloscope. A Doppler flow transducer and a hydraulic occluder were implanted on the same artery. A solid-state miniature pressure transducer (model P22; Konigsberg Instruments, Pasadena, CA) was implanted in the apex to measure LV pressure. In one control dog and two CCPO dogs, an indwelling catheter was implanted in the left circumflex coronary artery. The catheter was constructed of two pieces of Silastic tubing (0.012 and 0.040 ID, respectively) bonded together with a Silastic monomer. All transducer leads and catheters were passed through the thoracic wall and positioned exteriorly between the scapulae. The thoracotomy incision was closed in layers, and the animals were allowed to recover for 2 wk before the study. The animals used in this study were maintained in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (NIH) [DHHS Publication No. (NIH) 85-23, Revised 1985, Office of Science and Health Reports, Bethesda, MD 20892].

Experimental measurements. Statham strain gauge manometers (model P23 ID; Statham Instruments, Oxnard, CA) connected to the chronically implanted catheters were calibrated in vitro with a mercury manometer and used to measure aortic, left atrial, and LV pressures. LV pressure was also measured by use of the solid-state miniature pressure gauge calibrated in vitro with a mercury manometer and in vivo with the LV catheter and Statham strain gauge manometer. In the dogs with chronic indwelling coronary catheters, coronary perfusion pressure was measured by a miniature high-fidelity catheter pressure transducer (Mikro-tip 1.8F; Millar Instruments, Houston, TX) advanced to the coronary artery through the indwelling coronary artery catheter. Coronary artery diameter was measured with an ultrasonic transit-time dimension gauge. The dimension gauge generates a voltage linearly proportional to the transit time of the ultrasonic impulses traveling the velocity of 1.58 × 10⁶ mm/s between the pair of crystals. The frequency response of the dimension gauge is flat to 60 Hz. At constant room temperature, the thermal drift of the instrumentation is minimal, i.e., <0.02 mm in 6 h. Coronary blood flow velocity was measured by a Doppler flowmeter (model 100; Triton, San Diego, CA). Any drift in the measurement system was eliminated during the experiment by periodic calibration accomplished by substituting impulses of known duration from a pulse generator. The position of all catheters and crystals was confirmed at autopsy.

Experimental protocol. Experiments were performed in a quiet laboratory with the unsedated, conscious dogs resting comfortably in the right lateral position. The effects of bolus, intravenous injections of acetylcholine (3 µg/kg; Sigma, St. Louis, MO) and nitroglycerin (25 µg/kg; Parke, Davis, Morris Plains, NJ) and the effects of the release of 20-s coronary artery occlusion were examined. These experiments were conducted in seven CCPO and seven control dogs, except for the 20-s coronary artery occlusion, which could not be accomplished in one control dog due to failure of the hydraulic occluder.

In five of the control dogs and five of the dogs with CCPO, plasma renin activity was measured using the method of Haber et al. (9). The blood samples were taken from a peripheral vein with the dogs standing.

Data analysis. The data were recorded on a multichannel tape recorder (Honeywell, Denver, CO) and played back on a direct-writing oscillograph (Gould-Brush, Cleveland, OH). A cardiotachometer (model 9857B; Beckman Instrument, Fullerton, CA) triggered by the LV pressure pulse provided instantaneous records of heart rate. Continuous records of the LV rate of change in pressure (dP/dt) were derived from the LV pressure signal using operational amplifiers connected as differentiators and having a frequency of response of 700 Hz. A triangular wave signal was substituted for the pressure signals to directly calibrate the differentiator. Mean arterial pressure, mean coronary blood flow velocity, and mean coronary artery diameter were derived using resistor-capicator filters with 2-s time constants. The tracings corresponding to the coronary blood flow responses before and after 20-s coronary artery occlusion were digitized using a scanner interfaced to a computer. The area under the curve representing the volume and duration of the coronary blood flow deficit, i.e., the flow debt, and the excess of coronary blood flow that followed the release of the coronary artery occlusion, i.e., flow repayment, were quantitated with image analysis software (NIH Image, 1.5, Bethesda, MD).

In the control dogs, mean aortic pressure was used for the assessment of coronary perfusion pressure. In the dogs with CCPO for which coronary pressure was not measured directly, mean coronary perfusion pressure was assessed according to the technique described by Bache et al. (1), which averaged as systolic pressure, the LV pressure from the beginning of the upstroke of the aortic pressure to the dicrotic notch, and as diastolic pressure, the aortic pressure from the dicrotic notch to the beginning of the next upstroke. The analyses were performed on 10 consecutive beats with data acquisition software HEM 1.5 (Notocord, Croissy-Sur-Seine, France). In the dogs with coronary arterial catheters, coronary artery pressure was measured directly and indirectly using the previous analysis. Both measured and calculated mean coronary perfusion pressures were correlated from 140 data points. The slope, y-intercept, and r² were 0.87, 21 mmHg, and 0.979, respectively.

Histopathology. All dogs were anesthetized with pentobarbital sodium anesthesia (30 mg/kg iv) and were maintained on positive pressure respiration while the chest was opened, and the heart was arrested with potassium chloride and excised. The heart was rinsed in cold phosphate-buffered saline and perfused with saline followed by 2% phosphate-buffered glutaraldehyde through a large-bore cannula in the thoracic aorta. The perfusion pressure was maintained at ~90 mmHg by gravity flow, and the right atrium was opened to allow efflux of fluid. To preserve fresh tissue for other studies, in eight controls and eight CCPO dogs, the left anterior descending coronary artery was cannulated at the aortic ostium, and 15-30 g of the anterior wall region were perfusion fixed with glutaraldehyde after a saline wash. In the regionally perfused hearts, portions of the left circumflex and right coronary arteries were immersion fixed in 10% phosphate-buffered Formalin for histopathological evaluation. For light microscopy, the proximal coronary arteries were embedded in paraffin. Sections were cut at 6 µm thickness and stained with hematoxylin and eosin and Gomori aldehyde trichrome fuchsin. Additional coronary artery sections were obtained from the proximal 2 cm of the left anterior descending, the left circumflex, and the right coronary artery of all dogs; from the mid and distal portions of the three major arteries; and from selected marginal and diagonal branches. These arterial tissues were embedded in glycol methacrylate, sectioned at 1 µm, and stained with toluidine blue. Multiple blocks of tissue were embedded in Spurr epoxy resin, sectioned at 1 µm thickness, and stained with toluidine blue for light microscopic examination. Selected tissues were thin sectioned at silver-gray interference color, stained with lead citrate and osmium, and examined on a Philips 400 electron microscope. For scanning electron microscopy of large epicardial coronary arteries, perfusion-fixed vessels were cut longitudinally, dehydrated, and critical point dried, mounted on aluminum stubs, and coated with gold palladium. Tissues were examined using a Philips 515 scanning electron microscope.

Morphometric analysis. Morphometry of the large epicardial coronary arteries and small intramyocardial coronary arteries was performed on the glycol methacrylate 1-µm-thick sections using perfusion-fixed tissues. Small intramyocardial arteries were defined as vessels >100 µm in diameter. Arterioles were those vessels with an outer medial diameter <100 µm and one to three layers of smooth muscle in the media. Veins and lymphatics were excluded from analysis by the thinner wall relative to lumen size. Transversely sectioned vessels were imaged on a video screen, and video prints were made on a Seikosha video printer (final magnification varied from approximately ×20 for large coronary arteries to ×2,000 for the small coronary arteries and arterioles). Video prints of all transversely sectioned vessels were measured with a sonic digitizer system (Graf Bar; Science Accessories, Southport, CT) using computer programs developed in the laboratory. The system was calibrated with a stage micrometer. Long and short lumen diameters and the shortest outer medial diameter were measured, and the medial area was determined by digitizing the inner and outer medial wall. Medial thickness was determined as (shortest outer medial diameter  shortest lumen diameter)/2. Intramyocardial vessels with a long/short lumen diameter >1.8 (~55° sectioning angle) were excluded from analysis. Measured medial area was corrected for oblique sectioning using the long and short lumen diameters as follows: corrected area = measured area × (short axis/long axis).

To semiquantitatively evaluate lesions in the intima of epicardial coronary arteries, a score representing the grade of intimal proliferation was established as follows: 0, no intimal proliferation and an intact internal elastic lamina; 1, presence of one to two cell layers beneath the endothelium of up to 0.25 of the circumference, the internal elastic lamina being intact; 2, presence of one to two cell layers beneath the endothelium on >0.25 of the circumference or two to five cell layers beneath the endothelium on <0.25 of the circumference, associated with one to three areas with focal loss or splitting of the internal elastic lamina; 3, presence of more than five cell layers on <0.25 of the circumference associated with focal loss or splitting of the internal elastic lamina; and 4, presence of more than five cell layers on >0.25 of the circumference associated with focal loss or splitting of the internal elastic lamina. Grading was done blinded to the animal identification. Because a variable number of epicardial vessel sections was examined (3-15) for different animals, the grade reported is the maximum lesion grade found for each dog for all vessels examined. The total number of large arterial samples was similar for control (n = 74) and CCPO (n = 76) dogs. For intramyocardial small arteries and arterioles, all vessels present in three to eight methacrylate sections were examined, regardless of sectioning angle, to determine if intimal thickening was present.

Statistical analysis. Statistical analysis was performed using StatView and Super analysis of variance (ANOVA) software (Abacus Concepts, Berkeley, CA) on a Macintosh computer. The data are reported as means ± SE. The comparisons betweens groups and between baseline and responses were analyzed by a 2-way ANOVA for repeated measures, followed, if necessary, by a Student's t-test or a paired t-test. The comparison of the variations between groups was performed using Student's t-test. The arbitrary pathological grades for intimal proliferation were compared by a test of Mann and Whitney. A value of P < 0.05 was considered significant.

The present study consisted of several groups of dogs. There were four CCPO dogs that could be included in both the physiological and pathological analyses. Three additional CCPO dogs were used to complete each subgroup. There were two control groups with seven dogs in the physiology group and eight dogs in the pathology group.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Heart weight. The LV and right ventricular masses and their relation to body weight are shown in Table 1. Chronic pressure overload induced by aortic banding caused an 82% increase (P < 0.01) in the LV weight-to-body weight ratio in CCPO dogs compared with control dogs, without any changes in the right ventricular weight-to-body weight ratio.

Baseline hemodynamics. Baseline LV function and systemic hemodynamics are noted in Table 2. There were no significant differences in heart rate or LV dP/dt between CCPO dogs and control dogs, but LV systolic pressure was almost doubled in dogs with CCPO. As indicated in Tables 3-5, mean coronary artery diameter was slightly but not significantly elevated in CCPO dogs, and mean coronary blood flow velocity was similar between the two groups. However, systolic coronary blood flow velocity was greater, P < 0.01, in CCPO (94 ± 17 cm/s) than in control dogs (35 ± 7 cm/s), which was consistent with greater increases in systolic coronary artery pressure in CCPO (215 mmHg) compared with control dogs (130 mmHg; Fig. 1). Mean coronary artery pressure was also higher in CCPO (119 ± 6 mmHg, P < 0.01) than in control dogs (98 ± 2 mmHg).

View larger version (108K): [in this window] [in a new window]   Fig. 1.   Representative waveforms of left ventricular (LV) pressure, aortic pressure, coronary perfusion pressure, coronary blood flow velocity, and coronary diameter in a control dog (A) and a dog with chronic coronary pressure overload (CCPO; B). In the control dog, at midsystolic LV pressure, aortic pressure and coronary perfusion pressure were matched. In contrast, peak systolic coronary perfusion pressure was increased compared with aortic pressure in a CCPO dog. Note also that peak systolic blood flow velocity in a CCPO dog was augmented compared with control. Calculated and measured coronary perfusion pressures were similar, i.e., 106 and 103 mmHg in CCPO, respectively.

Plasma renin activity was not different in control (1.3 ± 0.4 ng · ml¹ · h¹) and CCPO (1.0 ± 0.4 ng · ml¹ · h¹) dogs.

Effects of acetylcholine, reactive hyperemia, and nitroglycerin. As indicated in Tables 3 and 4, both administration of acetylcholine and nitroglycerin decreased mean arterial pressure and increased heart rate and LV dP/dt. These effects were similar in control and CCPO dogs, except for the fall in LV systolic pressure, which was greater in CCPO. No significant changes in mean arterial pressure, heart rate, and LV dP/dt were noted during reactive hyperemia/dilation (Table 5).

As shown in Tables 3-5, acetylcholine (Figs. 2 and 3), reactive dilation (Fig. 3), and nitroglycerin (Fig. 3) increased coronary artery diameter similarly in control dogs. In CCPO dogs, the increases in coronary artery diameter with acetylcholine and reactive dilation were blunted (P < 0.05), whereas responses to nitroglycerin were not significantly different from those in control dogs. As shown in Tables 3-5 and illustrated in Fig. 4, the increases in coronary blood flow velocity were similar in control and CCPO dogs in response to acetylcholine, reactive hyperemia, and nitroglycerin. Furthermore, after release of the 20-s coronary artery occlusion, the ratio of flow repayment to flow debt was similar between the two groups (4.2 ± 0.5 vs. 4.1 ± 0.7 for control and CCPO dogs, respectively).

View larger version (64K): [in this window] [in a new window]   Fig. 2.   Representative waveforms of the coronary effects of the administration of acetylcholine in control (A) and CCPO (B) dogs. Increase in coronary diameter observed is reduced in the CCPO dog, despite a preserved response of blood flow. Acetylcholine was administered at the arrows.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Increases in coronary diameter in response to acetylcholine, 20-s coronary artery occlusion, and nitroglycerin in control (open bars) and CCPO (hatched bars) dogs. Responses to acetylcholine and reactive dilation were depressed (* P < 0.05), whereas responses to nitroglycerin were not different.

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Increases in coronary blood flow velocity in response to acetylcholine, which peak during reactive hyperemia, and in response to nitroglycerin in control (open bars) and CCPO (hatched bars) dogs. There were no differences between control and CCPO dogs.

Morphological evaluation. In control dogs, scanning electron microscopy of large coronary arteries revealed a normal, smooth, and regular luminal surface covered by endothelial cells oriented in a longitudinal direction in line with blood flow (Fig. 5). In the CCPO dogs, multifocal areas of endothelial irregularity with raised areas and disoriented alignment of endothelial cells were observed (Fig. 6).

View larger version (134K): [in this window] [in a new window]   Fig. 5.   Scanning electronic micrograph of the endothelial surface of the proximal left anterior descending coronary artery of a control dog. Endothelial cell nuclei are aligned parallel to the direction of blood flow (long arrow). Ostium of a branch artery is in the bottom right. Bar = 50 µm.

View larger version (155K): [in this window] [in a new window]   Fig. 6.   Scanning electronic micrograph of the endothelial surface of the proximal left anterior descending coronary artery of a CCPO dog. Focal irregularities of the surface endothelium are apparent, with endothelial cells oriented in several directions. Large arrow indicates direction of blood flow. Bar = 100 µm.

The large coronary arteries of the control dogs had an intact layer of endothelium, which by light microscopy appeared to be closely aligned to the underlying internal elastic membrane, with never more than an occasional single cell in the intima between the internal elastic membrane and the endothelium (Fig. 7A). In the CCPO dogs, both left and right coronary arteries and their major diagonal and marginal branches had multifocal areas with intimal thickening. The thickened intima was characterized by infiltration of cells, one to six or more cell layers thick, and increased amounts of interstitial matrix (Figs. 7B and 8). The internal elastic membrane was often split, fragmented, or duplicated. The focal areas of thickening, up to 60 µm thick, ranged in size up to the entire luminal circumference in a few animals, although more often, one-fourth to one-half of the circumference was affected. Lesions were found in at least one section of all CCPO dogs, with a minimum of one section each from the right, left anterior descending and circumflex coronary arteries. Intimal lesions were most common in the proximal portions of the large coronary arteries and less common in mid and distal portions, although lesions were found in the smallest epicardial diagonal and marginal branches on both ventricles. No intimal lesions were found in any intramyocardial coronary arteries. The subjective scoring system used identified lesions in all epicardial arteries in CCPO dogs (maximum grade = 3.1 ± 0.3) but little intimal thickening in control dogs (lesion grade 0 to 1 in all dogs; Table 6). Cells with large nuclei and abundant cytoplasm were often found in the vicinity of the internal elastic membrane (Fig. 8), interpreted as recently migrated activated smooth muscle cells. More mature cells with smaller nuclei were closer to the endothelium. By electron microscopy, the infiltrating cells were smaller and darker staining than medial smooth muscle cells and contained fewer contractile filaments, characteristic of activated smooth muscle cells (Fig. 9). There was abundant fibrillar collagen surrounding these intimal smooth muscle cells.

View larger version (134K): [in this window] [in a new window]   Fig. 7.   Proximal coronary artery sections from a control (A) and CCPO (B) dog. Media of the artery from the CCPO dog is more than two times the thickness of that in the control dog, and there is focal thickening of the intima in the CCPO dog coronary artery. There is malalignment of the smooth muscle cells in the outer medial layers in the CCPO vessel. Internal elastica (arrowhead) is pale unstained line separating the media and the intima. Methacrylate embedded 1-µm-thick sections stained with toluidine blue. Bar in A = 100 µm; magnification is the same in A and B.

View larger version (97K): [in this window] [in a new window]   Fig. 8.   Proximal left anterior descending coronary artery from a CCPO dog. Several cells with large nuclei in the base of the intima (arrowheads) apparently are modified smooth muscle cells that have migrated from the media. There is minor fragmentation of the internal elastica (double arrowheads) in this section. Methacrylate embedded 1-µm-thick sections stained with toluidine blue. Bar = 50 µm.

View larger version (147K): [in this window] [in a new window]   View larger version (169K): [in this window] [in a new window]   Fig. 9.   Transmission electron micrographs of the proximal left anterior descending coronary artery of a control dog (A) and a CCPO dog (B). Endothelial cells (E) in the control dog form a uniform layer, closely applied to the internal elastic membrane (IEL). Smooth muscle cells (SM) are present in the media underlying the internal elastic membrane. In the CCPO dog, the endothelial cells have an irregular appearance, with some dark and light staining, suggesting variable states of viability, and there are apparent breaks in the luminal endothelial plasma membrane (arrowhead). Internal elastic membrane is disrupted. In the enlarged space between the endothelium and the medial smooth muscle cells, there are smaller, darkly staining activated smooth muscle cells (ASM) in a bed of collagenous matrix. Magnification bar in A and B = 5 µm.

In control dogs, by electron microscopy of the large coronary arteries, the endothelium was uniformly intact and evenly stained and separated from the internal elastica by a thin matrix layer (Fig. 9A). In the CCPO dogs, the endothelial cells often had abnormal structure, consisting of pale staining of some cells, dark staining of others, and increased cellular fragility as evidenced by apparent rupture of the luminal plasma membrane in some cells (Fig. 9B). The internal elastic membrane was usually multiply fragmented into small pieces. In contrast to the large epicardial coronary arteries, the intramyocardial coronary arteries and arterioles of CCPO dogs had no alterations of the endothelium or intima by light microscopy, and there was significantly less medial thickening (Table 6). Electron microscopy confirmed the presence of an intact endothelium closely applied to the underlying medial smooth muscle cells, with a normal acellular thin intima and variable amounts of internal elastica in small arteries and arterioles in both groups (Fig. 10).

View larger version (117K): [in this window] [in a new window]   View larger version (146K): [in this window] [in a new window]   Fig. 10.   Transmission electron micrographs of coronary arterioles of a control dog (A) and a CCPO dog (B). Endothelium (E) is intact in both vessels, and 1-2 layers of smooth muscle cells (SM) are present in the media. Bars in A and B = 2 µm.

Quantitative evaluation of the large epicardial coronary arteries revealed more than two times the normal medial area and medial thickness in dogs with CCPO compared with controls (Table 6 and Fig. 7). In the control dogs, the medial smooth muscle cells of large epicardial vessels formed a well-ordered layer, unidirectional and circumferentially oriented. In the CCPO dog arteries, the outer one-half of the media often formed intertwining layers, with the smooth muscle cells oriented at oblique angles to the adjacent layers. Occasionally, there was a moderate increase in interstitial matrix, staining as collagen, that separated these disordered layers, and rare fibrocytes were found in this outer medial interstitial collagenous matrix. As previously reported (2), there was medial thickening of the small intramyocardial arteries and arterioles (Table 6) as well as in the large coronary arteries, which became progressively less as vessels decreased in size, so that the smallest arterioles, 15-25 µm outer diameter, had only minimally thicker media in CCPO compared with controls.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The results of this study demonstrated both functional and morphological abnormality of the endothelium, intima, and medial smooth muscle of the large epicardial coronary arteries in CCPO, whereas the intramyocardial small arteries and resistance vessels, in spite of medial thickening, retained normal function and endothelial structure. In the model used in the present study with LV pressure overload induced by banding the ascending aortic arch above the coronary ostia, there is a sharp increase in the pulse pressure in the coronary arteries and normal pressure in the systemic circulation distal to the band. Peak systolic pressure gradient across the band in this model is usually in excess of 100 mmHg, whereas diastolic pressures are not different, resulting in marked flow disturbances in the coronary arteries. Thus, although there is systolic hypertension in the coronary arteries, the systemic vasculature is not hypertensive, and there is no increase in plasma renin activity.

Alterations in the coronary circulation play a major role in the enhanced susceptibility of the myocardium in pressure overload-induced LVH to ischemic injury. In this regard, several studies have shown that the coronary circulation of this model is characterized by impaired subendocardial coronary vasodilator reserve (5, 13, 14, 17, 19) without a change in the capillary density (2). However, it is now well established that endothelial cells are a major determinant of coronary vasoactivity (7, 22). Loss of the integrity of this structure leads to abnormal vasomotor responses and to exaggerated constrictor responses (29).

In the model of supravalvular aortic banding-induced CCPO used in the present investigation, the large epicardial coronary endothelial function was selectively impaired. Both receptor- and flow-mediated endothelium-dependent vasodilations were attenuated, as reflected by blunted increases in coronary artery diameter in response to acetylcholine and the release of a 20-s coronary artery occlusion, i.e., reactive dilation, interventions recognized as hallmarks of endothelial function in studies in conscious animals (11, 12, 27, 29). The attenuated large coronary vasodilation was selective to these endothelium-mediated responses, since vasodilation in response to nitroglycerin was preserved. Interestingly, the resistance coronary vessels were spared from the endothelial dysfunction in this model, as reflected by normal increases in coronary blood flow and decreases in coronary vascular resistance in response to acetylcholine and reactive hyperemia as well as nitroglycerin. The failure to observe impaired coronary blood flow responses to these interventions in this study may appear to be internally inconsistent with the well-recognized limitations in subendocardial flow reserve in LVH (13, 14, 19). However, the reduced flow reserve limited to the subendocardium in LVH is only manifest upon maximal vasodilation. None of the interventions employed in this study elicited maximal vasodilation, and subendocardial blood flow was not measured.

Our next goal was to determine if the pathological analysis of the coronary arteries could explain the dysfunction observed in CCPO dogs. Scanning electron microscopy revealed endothelial lesions in both left and right large coronary arteries of the CCPO dogs, with the primary lesion being malalignment of the endothelial cells with regard to the direction of blood flow. Light and transmission electron microscopy revealed these focal thickenings to be due to the presence of modified smooth muscle cells and matrix deposition in the intima, with disruptions of the internal elastic membrane. These lesions were nonuniformly distributed in all epicardial coronary arteries and are consistent with impaired flow-dependent vasodilation (11). It is interesting to speculate that the misalignment may be caused by turbulence and vortex patterns of blood flow induced by the abnormally high systolic coronary artery pressure and velocity (3). These alterations in endothelial structure were located specifically in the large epicardial coronary arteries, sparing small resistance arteries, consistent with the hemodynamic data demonstrating selective impairment of large coronary responses but preserved resistance vessel function in CCPO.

It is important to note that the pathological changes in large coronary arteries extended well beneath the endothelium. As would be expected, there was a marked increase in the medial smooth muscle, most likely secondary to the systolic coronary artery hypertension due to the supravalvular aortic banding. The medial increase was not as profound in 50- to 100-µm intramyocardial arterioles. The increases in medial thickening may have been slightly overestimated in our study, since the coronary arteries were perfused at 90 mmHg postmortem, significantly lower than in vivo coronary perfusion pressure (119 mmHg). Because increasing arterial pressure from 90 to 120 mmHg in the conscious animal does not result in a large increase in arterial diameter, due to the exponential relationship of the pressure-diameter curve, the lower postmortem perfusion pressures most likely did not affect the data substantially. However, it is conceivable that the relatively minor, although statistically significant, increases in medial thickness in the smaller intramyocardial arterioles would not have been observed with physiological levels (120 mmHg) of coronary perfusion postmortem in the CCPO dogs.

In the large coronary arteries in CCPO, there were also severe intimal lesions, characterized by accumulation of smooth muscle cells in the subendothelium and alteration of the internal elastic lamina. Importantly, these intimal lesions, as well as the irregularities in the endothelium, were not found in the resistance vessels. Interestingly, the vascular response to nitroglycerin was not affected significantly, despite the alterations in the intima and media, suggesting that the disruption of the endothelium and intimal thickening was responsible for mediating the abnormal vasodilation to acetylcholine and reactive hyperemia.

Endothelial irregularities and intimal proliferation have been observed by others in hypertensive disease states. Alterations of the endothelium have been reported in aortas from rats with aortic coarctation (20) and systemic hypertension (10). The intimal thickening and infiltration of smooth muscle cells described in our study are similar to that observed in pressure overload secondary to aortic coarctation (20, 25) and systemic hypertension whether or not accompanied by LVH (10, 23, 28). Prior studies have described intimal lesions as aggregates of infiltrating smooth muscle cells (10, 23, 28) as well as collections of mononuclear cells from the lumen migrating through the endothelium (28). It is important to note, however, that these lesions have not all been observed previously in coronary arteries, particularly in the absence of systemic hypertension. This latter point is particularly important since these lesions at the coronary artery level might be involved in the process of heart disease. It is not surprising that disease processes, such as systemic hypertension, atherosclerosis, or congestive heart failure, may cause major changes in vascular regulation and structure (21, 27, 30). Although atherosclerosis is a primary vascular disease, hypertension and heart failure affect neurohormonal balance, e.g., activate the renin-angiotensin system, which exerts trophic influences on blood vessels (4). In the current model of CCPO, plasma renin activity is normal, suggesting that the effects on the coronary vasculature were primarily mechanical rather than neurohormonal, although an effect of the local renin-angiotensin system cannot be excluded.

Differences in coronary hemodynamics between control and CCPO dogs rather than hypertrophy per se could explain the differences in the physiology and pathology observed in the large coronary arteries vs. resistance vessels, since endothelial lesions were observed in the right coronary artery not subjected to ventricular hypertension. Specifically, systolic coronary artery pressure was increased in the CCPO dogs, but diastolic pressure was not (Fig. 1). Furthermore, it is known that there is a pressure drop between the large arteries and the resistance vessels (15), suggesting that the systolic pressure difference would be less marked in the resistance vessels, which would be consistent with the physiological and pathological observations. It appears likely that the abnormally high systolic coronary arterial pressure induces abnormally high systolic flow velocity with disruption of laminar flow (16) in CCPO, resulting in the morphological changes. Although the abnormal flow patterns in CCPO (high systolic flow velocity and turbulence) are most likely secondary to the elevated systolic coronary arterial pressure, it is not clear whether just one of these factors or their combination is responsible for the morphological changes.

In conclusion, the present study demonstrates selective impairment of the large coronary artery endothelial function after CCPO in the absence of systemic hypertension. The in vivo alteration in the endothelium-mediated dilation was associated with morphological abnormalities of the endothelium and the intima. The lesions are most likely the consequence of systolic coronary arterial hypertension, which induces abnormally high systolic flow velocity and disruption of laminar flow patterns. The clinical implications extend not only to the corollary of patients with supravalvular aortic stenosis and coarctation of the aorta but to all patients with coronary artery hypertension.