INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

A LIFESTYLE INCORPORATING moderate levels of physical activity is recognized as therapeutic in prevention and treatment of coronary heart disease (CHD) (15, 36). Although exercise training improves transport function in the coronary circulation (29, 31, 32), the reasons exercise training is beneficial in preventing and treating CHD have not been established. Recent evidence indicates that these beneficial effects of exercise training are partially the result of training-induced improvements in endothelial function in the coronary circulation (27, 34, 41, 48, 53) and/or peripheral arteries (7, 11, 12). The endothelium modulates both vascular resistance and arterial vascular compliance (43, 44) through vasoconstrictor and vasodilator responses (9, 10, 13, 14). Endothelium-mediated vasodilation is the result of release of a number of dilator substances, including prostacyclin (PGI₂), hyperpolarizing factors, and nitric oxide (NO) (8, 16). In the endothelium, NO is produced by the endothelial isoform of NO synthase (eNOS) (8, 16, 54). Our laboratory previously reported enhanced endothelium-mediated dilation in coronary arterioles from exercise trained pigs that was blocked by NO synthase inhibitors (arginine analogs) (34). In contrast, endothelium-mediated vasodilation is not altered in conduit coronary arteries in this model of exercise training (39). Thus the purpose of this study was to test the hypothesis that exercise training increases endothelium-mediated dilation in coronary arterioles (34), but not in conduit arteries (39), because eNOS protein content is increased in the coronary arterial microcirculation [i.e., all arteries with internal diameter (ID) <300 µm] but not in larger coronary arteries.

The relative importance of endothelium-mediated control has been shown to differ along the length of the coronary arterial network (23-25). In arteries and large arterioles, the endothelium may play a greater role in vascular control than in smaller arterioles. This fact combined with evidence that training increases endothelium-mediated dilation in coronary arterioles but not in large coronary arteries suggests that exercise training increases endothelial function in some coronary arteries but not in others. Therefore, these experiments were designed to determine where, in the progression from arteriole to conduit artery, endothelial adaptations occur. We measured eNOS protein content in six specific sizes of coronary arteries [3 sizes of coronary artery: conduit coronary arteries, small coronary arteries (301- to 1,000-µm ID), and resistance arteries (151- to 300-µm ID); and 3 sizes of coronary arteriole: large (101- to 150-µm ID), intermediate (51-to 100-µm ID), and small (<50-µm ID)]. Content of eNOS protein was measured in conduit coronary arteries, small coronary arteries, and resistance arteries using standard immunoblot techniques (42). Procedures for immunoblot analysis of eNOS protein in single arterioles were developed because preliminary results indicated that standard approaches lacked the sensitivity to measure eNOS protein in single coronary arterioles.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Experimental Animals

Exercise training program. During the first week, Ex pigs ran on the treadmill at 3 miles/h (mph), 0% grade for 20 to 30 min (endurance) and at 5 mph for 15 min (sprint). The speed and duration of running were progressively increased at a rate dependent on the tolerance of each pig. During the 12th wk of training, a typical training session consisted of the following 85-min workout: 1) 5-min warm-up run at 2.5 mph, 2) 15-min sprint at speeds of 5-8 mph, 3) 60-min endurance run at 4-5 mph, and 4) 5-min warm-down run at 2 mph. Ranges of running speed are presented because the training program was customized to each pig's exercise ability. The Ex pigs were given positive reinforcement for exercise by being fed after each training bout. Treadmill performance tests were again administered to the Sed and Ex pigs at the completion of the pen confinement or exercise training periods.

Treadmill performance test. The performance test consisted of four stages of exercise (38). During stage 1, the pigs ran at 3.1 mph and 0% grade for 5 min. The pigs ran for 10 min at stage 2 (speed = 3.1 mph and grade = 10%). The pigs then ran for 10 min at stage 3 (speed = 4.3 mph and grade = 10%). Finally, the pigs ran at stage 4 (speed = 6 mph, grade = 10%) until exhaustion. Heart rates were recorded throughout the treadmill performance test along with total duration of exercise.

Efficacy of training. The effectiveness of the training program was determined by comparing the exercise tolerance (as reflected in the treadmill performance test), heart weight-to-body weight ratios, and skeletal muscle oxidative capacity of trained and control groups. At the time of death, samples were taken from the middle of 1) lateral head of triceps brachii, 2) long head of triceps brachii, and 3) deltoid muscles. Muscle samples were frozen and stored at 70°C until processed. Citrate synthase activity was measured from whole muscle homogenate using the spectrophotometric method of Srere (50).

Isolation of Coronary Arteries and Arterioles

Isolation of coronary arteries. After completion of exercise training or sedentary confinement, the pigs were sedated with ketamine (30 mg/kg), anesthetized with pentobarbital sodium (35 mg/kg; Fort Dodge), and heparinized (1,000 U/kg body wt). Hearts were removed and placed in iced (4°C) physiological salt solution during vessel isolation. Segments of conduit coronary arteries [2- to 3-mm external diameter (OD)] and small coronary arteries (300- to 1,000-µm ID) were carefully dissected from the myocardium and trimmed of fat and connective tissue. Vessel samples were taken from similar sites in hearts from Sed and Ex pigs. Segments of left anterior descending (LAD) and left circumflex coronary artery were cut into rings. The ID and OD of the arteries were measured with a Filar-calibrated micrometer eyepiece, and the arteries were placed in microcentrifuge tubes on dry ice. At the time of analysis, arteries were thawed and added to 4 vol of 0.05 M Tris extraction buffer (pH 7.4) with 0.1 mM EDTA, 0.01 mM EGTA, 0.01% -mercaptoethanol, 10% vol/vol glycerol, 20 mM 3-[(3-cholamidopropyl) dimethylammonium]-1-propanesulfonate (CHAPS), 2 µM leupeptin, l µM pepstatin A, and 1 mM phenylmethylsulfonyl fluoride (PMSF) and were homogenized (Tekmar Tissumizer). Samples were centrifuged at 100,000 g for 30 min at 4°C, and the supernatant used for analysis of eNOS protein content as described in eNOS Protein Content of Arteries and Arterioles.

Isolation of arterioles and resistance arteries. With the aid of an Olympus dissecting microscope coronary arteries 300-µm ID (measured with a calibrated eyepiece) were identified. We dissected along the length of these 300-µm-internal-diameter arterioles, out to the smallest branches we could accurately see (7- to 10-µm ID). These arteriolar trees were then carefully isolated and the portions of the tree with IDs of 151-300 were placed in the resistance artery sample. Arteries were frozen in microcentrifuge tubes maintained at 70°C as they were dissected from the myocardium. Sampling was completed in 4 h (17). At the time of analysis, arteries were thawed, separated from excess dissection buffer, and sonicated with three 15-s bursts in 200-300 µl of extraction buffer (pH 7.4; 50 mM Tris · HCl, 0.1 mM EDTA, 0.1 mM EGTA, 0.01% -mercaptoethanol, 20 mM CHAPS, 2 µM leupeptin, l µM pepstatin A, and 1 mM PMSF). Samples were centrifuged at 100,000 g for 30 min at 4°C, and the supernatant was used for analysis of eNOS protein content.

eNOS Protein Content of Arteries and Arterioles

Standard immunoblot analysis. Artery eNOS protein content was measured with immunoblots using two distinct monoclonal antibodies on separate samples (H32, BIOMOL; and Transduction Laboratories) for eNOS. Samples (30 µg total protein) were loaded and electrophoresed on 7.5% SDS-PAGE gels and electroeluted onto nitrocellulose membranes. Each gel was loaded with equal numbers of Sed and Ex pig samples of equal-sized arteries <150-µm ID, 151- to 300-µm ID, 301- to 1,000-µm ID, or conduit arteries) to facilitate comparison of bands between groups. Blots were air dried for 1 h and blocked with 6% nonfat dry milk (Blotto) for 1 h at room temperature while being gently agitated. Blots were washed two times with Tris-buffered saline (TBS) and incubated for 2 h in primary eNOS antibody (1:1,000 in 1% Blotto) at room temperature. Blots were rinsed four times in TBS and incubated with a secondary antibody (sheep horseradish peroxidase-conjugated anti-mouse) at 1:5,000 in 1% Blotto for 1 h. The blots were rinsed four times for 10-15 s, and eNOS protein content was determined by means of the Pierce Super Signal chemiluminescence system. Estimates of the amount of eNOS protein in the immunoreactive bands were determined using scanning densitometry applying the NIH Image software.

Modified immunoblot analysis of eNOS protein in arterioles. In preliminary experiments, eNOS protein content of arterioles was determined with standard immunoblot techniques as described in Standard immunoblot analysis, applied to samples that combined all coronary arteries with diameters <150-µm ID isolated from each heart (pooled into 1 sample) (28). It was necessary to pool all arterioles from each heart to obtain sufficient protein to assay for protein content and have 30 µg total protein remaining to load for the immunoblot. We were concerned that pooling samples of all arterioles <150-µm ID could mask differential effects of exercise training on eNOS expression in arterioles of different diameter. Therefore, we modified these procedures to allow estimates of eNOS protein content in single arterioles. On the basis of the functional results of Kuo et al. (25), we applied this method of immunoblot analysis of eNOS protein in single arterioles in the following size categories: <50-µm ID, 51- to 100-µm ID, and 101- to 150-µm ID from individual hearts of Ex and Sed pigs.

Immunohistochemistry. To ensure representative samples including the full range of coronary arteries in the study, two 1-cm³ regions of the heart were sampled: 1) myocardium containing a cross section of the LAD and 2) myocardium from the apical region of the left ventricle. In addition, brachial artery was collected as a positive control. Tissues were fixed by immersion in aqueous-buffered zinc Formalin (Anatech) for at least 6 h and processed through standard paraffin embedding.

Data Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Efficacy of Exercise Training

Amount of eNOS Protein in Coronary Arteries

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Sample immunoblots of resistance arteries isolated from the hearts of sedentary (Sed) and exercise-trained (Ex) pigs. Proteins were electrophoretically separated and probed for immunoreactivity to H32 antibody for endothelial nitric oxide synthase (eNOS) used at a 1:3,000 dilution (top) or Transduction Laboratories antibody for eNOS used at a 1:1,000 dilution for 1 h (bottom). Secondary anti-mouse antibody was used at 1:3,000 (top) and 1:5,000 (bottom). RM, rainbow marker; BAE, cultured bovine aortic endothelial cells. eNOS can be most clearly seen in lane 10 of the blot shown at top, the BAE sample. The primary band of eNOS consistently is observed with a molecular weight of ~135.

View larger version (39K): [in this window] [in a new window]   Fig. 2.   Mean ratios of Ex/Sed eNOS levels measured by scanning densitometry of individual autoradiograms. Values are means ± SE. Lanes were loaded with equal amounts of total protein as described previously (42). Data were collected from blots containing at least 2 paired samples from Ex and Sed pigs of a given-size artery. RA, resistance arteries (151- to 300-µm ID, n = 15 Sed and 15 Ex paired samples on 9 blots); SA, small arteries (301- to 1,000-µm ID, n = 8 Sed and 8 Ex paired samples on 5 blots); CC, conduit arteries (>1,000-µm ID, n = 6 Sed and 6 Ex paired samples on 4 blots). Both RA and SA from Ex pigs have significantly more eNOS protein than comparable arteries from Sed pigs.

Amount of eNOS Protein in Arterioles

View larger version (56K): [in this window] [in a new window]   Fig. 3.   Immunoblots from single arterioles 35- to 40-µm internal diameter. Each lane represents material from a single coronary arteriole. Three Sed and Ex samples shown were matched for diameter and length. Each arteriole was dissolved in 20 µl Laemmli sample buffer, boiled and sonicated, and probed for eNOS (top) and glyceraldehyde 3-phosphate-dehydrogenase (GAPDH; bottom) immunoreactivity. The material was analyzed via SDS-PAGE. The luminograms were generated with chemiluminescence.

A total of 32 arterioles were examined from Sed pigs and 26 from Ex pigs in the size categories summarized in Table 1. When the pooled arteriolar data (all 3 sizes combined) were analyzed, results revealed that there were no between-group differences in mean arteriolar diameter (Sed = 69 ± 6 µm, Ex = 73 ± 6 µm) or arteriolar length (Sed = 1,838 ± 131 µm, Ex = 1,670 ± 135 µm). However, the amount of eNOS protein content per arteriole was 37% greater in the arterioles from Ex pigs (44,402 ± 5,283 units) than from Sed pigs (32,355 ± 4,917 units).

To determine whether arteriolar size influenced these results, we examined eNOS content within arteriolar sizes as shown in Fig. 4. For comparison between samples, eNOS protein was expressed as arbitrary densitometric units without correction (assuming that arterioles of equal diameter and length contain equal amounts of total protein). These results indicate that small (P < 0.05) and large (P < 0.08) sizes of arterioles isolated from Ex pigs have more eNOS protein per arteriole than size-matched arterioles from Sed pigs (Fig. 4).

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Immunoreactive eNOS protein levels of electrophoretically separated eNOS protein from coronary arterioles isolated from Sed and Ex pigs. eNOS levels were quantified by scanning densitometry of individual autoradiograms for arterioles from 13 Sed and 10 Ex pig hearts for small (top) and intermediate (middle) sizes and from 6 Sed and 6 Ex pig hearts for larger arterioles (bottom). Actual diameters and lengths of these arterioles are presented in Table 1. Values are means ± SE for relative densitometric units. *Ex greater than Sed, P < 0.05. # Ex greater than Sed, P < 0.08.

Table 1 presents the dimensions (diameters and lengths) of arterioles and eNOS protein content adjusted for gel-loading differences among the samples by correcting to the intensity of the GAPDH signal within individual samples. These data indicate that the smallest Ex pig arterioles have a 64% increase in eNOS protein, the intermediate-size arterioles from Ex pigs have 9% more eNOS protein, and the large arterioles from Ex and Sed pigs have similar amounts of eNOS protein relative to the amount of GAPDH protein. Thus Ex arterioles <50-µm ID have increased eNOS content, independent of how the eNOS protein content data are expressed (per size-matched arteriole or relative to GAPDH protein content). The total amount of eNOS protein per arteriole is increased in the larger sized arterioles from Ex pigs (P < 0.08; Fig. 4). However, when the data are expressed relative to the GAPDH content, large arterioles from Ex and Sed have similar amounts of eNOS (Table 1).

Immunochemistry

View larger version (132K): [in this window] [in a new window]   Fig. 5.   eNOS immunoreactivity demonstrated in coronary arteries of 4 different internal diameters. Data for Ex pigs are presented on the left and for Sed pigs on the right. Inserts at top left corner of each panel demonstrate eNOS immunoreactivity of the endothelial cells of each artery taken with a ×100 objective under oil immersion. Paraffin sections (5 µm) immunostained using the Vectastain avidin-biotin-peroxidase complex method with biotin- and casein-blocking steps using Transduction Laboratory eNOS antibody (2 µg/ml). Endothelial fixation was assessed by immunoperoxidase staining using purified monoclonal mouse anti-bovine vimentin IgG2a (Dako) on an adjacent serial section. Sections were counterstained with eosin, dehydrated, cleared, and coverslipped with Permount. Immunoreactivity was observed using an Olympus BX60 microscope with a tungsten-halogen light source.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The purpose of this study was to test the hypothesis that exercise training causes increased expression of eNOS protein throughout the coronary arteriolar microcirculation but not in large coronary arteries. This hypothesis emerged from results demonstrating that endothelium-mediated dilation is increased in arterioles (34) and coronary resistance arteries (41), but not changed in conduit coronary arteries (38, 39), of trained miniature swine. We measured eNOS protein content in six specific sizes of coronary artery to obtain an appreciation for the spatial pattern of training-induced changes in eNOS expression throughout the coronary arterial tree. Results indicate that conduit coronary arteries from Ex pigs have similar amounts of eNOS immunoreactivity compared with those from Sed (Fig. 2). In contrast, small coronary arteries, coronary resistance arteries, and large and small coronary arterioles from Ex pigs have increased eNOS immunoreactivity compared with similar-sized arteries from Sed pigs (Figs. 2 and 4). Finally, immunohistochemistry revealed that eNOS immunoreactivity is located in the endothelium of the coronary arteries of all sizes. We conclude that exercise training induces increased eNOS protein content in a nonuniform pattern. Training did not alter eNOS content in conduit coronary arteries or intermediate-sized arterioles, but it produced 60-80% increases in eNOS content in small (<50-µm ID; P < 0.05) and large arterioles (101- to 150-µm ID; P < 0.08) and in coronary arteries 151- to 1,000-µm ID (P < 0.05). The nonuniform pattern of increased eNOS protein content in the coronary arterial tree produced by exercise training is illustrated in Fig. 6 as the amount of eNOS protein in each size coronary artery from Ex pigs expressed as a percentage of the eNOS content of the same size artery of Sed pigs. It is fascinating that exercise training increased eNOS protein content in some coronary arteries but not in others.

View larger version (43K): [in this window] [in a new window]   Fig. 6.   Relative increase in eNOS protein content in coronary arteries from Ex pigs expressed as percentage of control (Sed values). For the 3 sizes of arterioles [large (La), intermediate (Ia), and small (Sa)], eNOS protein content per arteriole (Fig. 4) for Ex pigs is expressed as a percentage of Sed eNOS protein content per arteriole. For RA, SA, and CC, the amount of eNOS protein is expressed per µg artery protein relative to eNOS protein content in Sed arteries of the same size.

Changes in eNOS content may result from increases or decreases in eNOS gene expression in endothelial cells and/or from changes in the amount of total arterial protein present. To determine whether eNOS protein content increased because of increased eNOS gene expression, it is necessary to measure eNOS content of the endothelial cells. This is not currently possible for coronary arterioles. We considered using a constitutively expressed endothelial-specific gene such as Factor VIII or von Willebrand factor (vWF) to adjust eNOS protein data to endothelial cell protein content. However, vWF is a secreted protein (52), the expression of which by endothelial cells can be increased by a number of signals, including estrogen (19), interleukin-1 and other lymphocyte products (5), endothelin-1 (18), thrombin (47), and, importantly for our study, acute exercise bouts (6, 21, 49). In addition, vWF expression in endothelial cells varies throughout the vascular tree within tissues and among different tissues as well as from one endothelial cell to the next within an artery (1). Thus, whereas vWF is an excellent endothelial cell-specific marker (all endothelial cells express mRNA and protein for this gene), its expression is highly variable, and it therefore is not good for our use to evaluate whether eNOS expression is increased or other mechanisms result in increased eNOS protein content in some sizes of coronary artery in trained pigs. We are not aware of an endothelial specific gene having an expression that has been demonstrated to be unchanged by exercise.

Although our results do not allow us to determine whether eNOS content is increased by exercise training due to increased expression of the eNOS gene, we believe there are two sets of observations that suggest these changes in eNOS content of coronary arteries are the result of increased expression of the eNOS gene. First, eNOS mRNA expression is increased in coronary arterioles isolated from exercise trained pigs (55). Second, because length-tension relationships, wall thickness, and maximal contractile tension induced by KCl are not altered in coronary arteries of Ex pigs, it is likely that only negligible changes occur in vascular smooth muscle or connective tissue content of these arteries (31, 38, 39). Similarly, consistent with the notion that exercise training does not alter the amount of vascular smooth muscle or connective tissue of arterioles, passive pressure-diameter relationships and vasoconstrictor responses (observed at constant intraluminal pressures) are similar in arterioles of Sed and Ex pigs (30, 34). If exercise bouts produce signals known to result in increased eNOS gene expression, this may provide further evidence of how eNOS content is increased.

Exercise produces many stimuli that could signal adaptive changes in gene expression in the vascular endothelium, including increased levels of metabolic signals (e.g., adenosine, hypoxia, and so forth), increased mechanical forces on the endothelium (e.g., shear stress and pressure), and altered neurohumoral milieu (e.g., increased catecholamine levels) (29). A number of studies support the hypothesis that increases in coronary blood flow associated with each bout of exercise generate a "shear stress" signal for vascular adaptation (13, 27, 35, 45, 48, 53). Shear stress has been reported to increase transcription of the eNOS gene and increase eNOS protein expression in cultured endothelial cells (37, 45). Chronic increases in blood flow produced by arteriovenous fistulas increase expression of eNOS mRNA and protein and enhances endothelium-mediated vasodilation in conduit arteries (33, 35). Increases in blood flow through conduit arteries caused by exercise have also been reported to increase eNOS protein content and endothelium-mediated vasodilation in conduit arteries (7, 11, 12). Furthermore, increased shear stress in isolated coronary arterioles has been shown to produce increased eNOS mRNA (55). We are not aware of measures of shear stress throughout the coronary arterial tree at rest or during exercise. However, it seems likely that shear stress is increased in conduit coronary arteries during exercise because coronary blood flow increases four- to sixfold and these arteries only increase diameter by 4-10% (4, 20, 53). Therefore, it is surprising that eNOS content was not increased in conduit coronary arteries of these trained pigs (Fig. 2). As discussed above, it is possible that eNOS expression was increased in conduit coronary arteries but that this was matched by increased total protein. Either way, the lack of change in eNOS protein content in conduit coronary arteries of Ex pigs is consistent with reports that endothelium-mediated control of proximal coronary arteries is not altered by exercise training (38-40, 46).

Short-term exercise training (7-10 days) has been reported to produce increased eNOS expression and endothelium-dependent dilation of conduit coronary arteries of dogs (53). The reason that our laboratory and others have not found similar changes in conduit coronary arteries of trained animals (38-40, 46) appears to be the duration of training (27). Endothelium-mediated dilation of conduit coronary arteries appears to be enhanced early in the exercise-adaptive process (48, 53) but returns to normal later in the training period (27). It is possible that after the subject is fully adapted to exercise training, structural adaptations decrease shear stress in the coronary artery toward normal values during exercise so increased eNOS protein content is no longer necessary (27, 39, 40)

We expected adaptations in small coronary arteries (301-to 1,000-µm ID) would resemble those in conduit arteries, although we had not examined endothelium-dependent vasodilation in these arteries. Thus we were surprised that eNOS protein content was increased in small arteries from Ex pigs (Figs. 2 and 6). Resistance arteries from Ex pigs also exhibited increased eNOS content (Figs. 2 and 6), which is consistent with enhanced endothelium-dependent vasodilation of coronary arteries of this size (41).

It is important to relate the data of the present study to those of Muller et al. (34) showing enhanced endothelium-dependent vasodilation in coronary arterioles of Ex pigs. This requires careful consideration of the conditions under which arteriolar diameters were measured in the two studies. In the present study, arteriolar diameters were measured with the arterioles in the dissection bath. Muller et al. measured diameters after arterioles had been cannulated, pressurized, and exposed to sodium nitroprusside. In our hands, diameters measured in passive, pressurized coronary arterioles are 1.35 times the diameter measured during dissection with the arteriole in the bath. That is, passive diameter generally increases by 35% on cannulating and pressurizing of the arteriole. Muller et al. reported that the arterioles included in their study were 65- to 157-µm passive ID (cannulated and exposed to 40 mmHg luminal pressure) with those from Ex pigs averaging 117-µm ID and those from Sed pigs averaging of 114-µm ID. Conversion of these diameters to those that existed at dissection indicates that the range of diameters for the arterioles used by Muller et al. was 48- to 116-µm ID with the average Ex and Sed values of 87- and 84-µm ID, respectively, under conditions used to measure diameter in the present study. The range of diameters included in the present study was 25- to 120-µm ID. Thus, by design, the smallest arterioles (in which we measured eNOS protein content) were smaller and the largest arterioles were larger than those used by Muller et al. Our results indicate that there was 37% more eNOS protein per arteriole in arterioles from Ex pigs (pooled results of all 3 sizes of arterioles), which confirms previous results from pooled arteriolar samples (28) and increased levels of eNOS mRNA in coronary arterioles of similar size, isolated from pigs trained with a similar training program (55). These results are consistent with the hypothesis that the enhanced responses of arterioles from Ex pigs reported by Muller et al.(34) were the result of an increased amount of eNOS protein.

We also emphasize that, whereas present results indicate that changes in eNOS protein content are involved in enhanced vasodilator responses of coronary arterioles, available information indicates involvement of other endothelium-derived factors. For example, increased production of PGI₂ appears to contribute to the enhanced vasodilation of arterioles from Ex pigs. Muller et al. (34) reported that blockade of cyclooxygenase activity with indomethacin (10⁵ M) produced a dramatic decrease in bradykinin sensitivity of coronary arterioles from both Ex (IC₅₀ increased 50-fold) and Sed (IC₅₀ increased 5-fold) pigs, indicating that production of PGI₂ is important in bradykinin-induced, endothelium-mediated dilation in porcine coronary arterioles. Furthermore, because indomethacin produced a 10-fold greater decrease in bradykinin IC₅₀ in arterioles from Ex pigs than in arterioles from Sed pigs these results suggest that training increases the contribution of this pathway to endothelium-mediated dilation.

Finally, because the relative importance of endothelium-mediated vascular control varies throughout the coronary arterial tree, it is of interest to know whether endothelial cell eNOS content normally differs throughout the coronary arterial tree (22, 24, 25, 41). For example, Kuo et al. (25) compared the sensitivity of porcine coronary arterioles to shear stress and found that shear stress sensitivity increased as a function of coronary arteriole diameter in the range of 40- to 100-µm diameter. It is not known whether these functional differences among arterioles result from differing mechanisms that couple shear stress to endothelium-derived mediators in arteries of different size or from other variations of endothelial phenotype throughout the coronary tree. Ando et al. (3) recently reported that eNOS expression and eNOS activity are greater in cultured aortic endothelial cells than in cultured cardiac microvascular endothelial cells. Also, data for Sed pigs presented in Table 1 suggest that the ratio of eNOS protein content to GAPDH protein content is not constant but shows the following pattern: small arterioles < large arterioles < intermediate arterioles. It is interesting that these results are consistent with functional evidence of Kuo et al.(25) and biochemical evidence (3) suggesting that eNOS protein content is greater in endothelial cells of large arteries than of small coronary arterioles. However, our results cannot be used for comparisons of eNOS content among arterioles of different size because large arterioles contain more total protein than small (Fig. 4). We could not express eNOS protein content relative to total protein of single arterioles because there is just too little protein in one arteriole to measure protein content and have sufficient sample to run on an SDS gel. Accordingly, our results do not allow conclusions concerning the relative amounts of eNOS protein among coronary arteries of different size in the normal coronary tree.

In summary, results of this study indicate that in porcine conduit coronary arteries training does not increase eNOS protein content as reflected in immunoblot and immunohistochemistry data. In the smaller coronary arteries, we found that exercise training produces increased eNOS protein content in a nonuniform manner. Content of eNOS protein was increased 60-80% in small arteries (301- to 1,000-µm ID), resistance arteries (151- to 300-µm ID), and large (101- to 150-µm ID) and small-sized arterioles (< 50-µm ID). Intermediate arterioles do not appear to have a change in eNOS protein content after training. We conclude that the increased endothelium-mediated vasodilation observed in coronary arterioles is partially the result of increased eNOS protein content. Increased eNOS activity, increased release of PGI₂, and changes in other signaling pathways for endothelium-mediated vasodilation may also be modified by exercise training.