INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

RESULTS FROM DIFFERENT LABORATORIES (7, 11, 17, 20, 24-28) have indicated that collateral-dependent coronary arteries exhibit altered vasomotor reactivity compared with nonoccluded control arteries. Furthermore, experiments evaluating simultaneous changes in contractile tension and myoplasmic free Ca²⁺ have revealed enhanced constriction and impaired relaxation of coronary arterial rings from the collateral-dependent vasculature associated with increased myoplasmic free Ca²⁺ concentrations (11, 20). Recent studies also indicate that impaired adenosine-induced relaxation and the associated increase in myoplasmic free Ca²⁺ levels in collateral-dependent coronary arteries are reversed with exercise training (11). The mechanisms responsible for altered Ca²⁺ regulation in the collateral-dependent coronary vasculature and its reversal with exercise training remain undefined.

Our preliminary experiments demonstrated that after the sarcoplasmic reticulum (SR) had been partially depleted of Ca²⁺ by exposure to caffeine, KCl-induced membrane depolarization resulted in a greater rate of myoplasmic free Ca²⁺ accumulation in smooth muscle cells isolated from collateral-dependent left circumflex (LCx) versus nonoccluded left anterior descending (LAD) coronary arteries. In contrast, in the absence of prior SR depletion, the rate of myoplasmic free Ca²⁺ accumulation was not different between smooth muscle cells of collateral-dependent LCx and nonoccluded LAD arteries. On the basis of these findings, we hypothesized that chronic coronary occlusion impairs SR Ca²⁺ uptake, leading in turn to altered myoplasmic free Ca²⁺ regulation in the smooth muscle of collateral-dependent arteries. We also hypothesized that exercise training would correct the impaired SR Ca²⁺ uptake proposed in collateral-dependent vasculature. The present studies were designed to evaluate potential Ca²⁺ regulatory mechanisms that contribute to altered vasomotor reactivity in coronary arteries distal to chronic occlusion and the effects of exercise training on these mechanisms.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals and surgical procedures. All animal protocols were in accordance with the "Principles for the Utilization and Care of Vertebrate Animals Used in Testing, Research and Training" and approved by the University of Missouri Animal Care and Use Committee. Thirty-two adult female Yucatan miniature swine (Charles River; Wilmington, MA) were surgically instrumented with Ameroid constrictors around the proximal LCx artery as described previously (7, 11). Animals were preanesthetized with glycopyrrolate (0.004 mg/kg im) and midazolam (0.5 mg/kg im). Anesthesia was induced with ketamine (20 mg/kg im) and maintained with 3% isoflurane-97% O₂ throughout the aseptic surgery. Animals recovered from the surgery for 8 wk before the experimental protocols were initiated. Twelve additional control animals did not undergo surgery.

Exercise training. Ameroid-occluded animals were randomly assigned to either a sedentary or an exercise-training group. All control (nonoccluded) pigs used for this study remained sedentary. Exercise-trained pigs underwent a progressive treadmill exercise-training program (5 days/wk for 16 wk) as described previously (7, 11). Effectiveness of the exercise-training program was determined by comparing the heart weight-to-body weight ratio and skeletal muscle citrate synthase activity as previously described (7, 11).

Preparation of coronary arteries. After the 16-wk exercise-training protocol or sedentary confinement was completed, animals were anesthetized with ketamine (35 mg/kg im), rompun (2.25 mg/kg im), and pentothal sodium (10 mg/kg iv), followed by administration of heparin (1,000 U/kg iv). Animals were euthanized by removal of the heart, which was immediately placed in cold Krebs bicarbonate buffer (4°C) and weighed. With the aid of a dissection microscope, segments of the LCx and LAD arteries from both occluded and control pigs were trimmed of fat and connective tissue, and vessel diameter was measured with a calibrated Filar micrometer eyepiece (Hitschfel Instruments; St. Louis, MO). Visual inspection of the Ameroid occluder during dissection of the LCx artery indicated 100% occlusion in all animals that had undergone surgery for this study; the LCx was thus termed a "collateral-dependent" coronary artery.

Myoplasmic free Ca²+ measurement. Smooth muscle cells from segments of the LCx and LAD arteries were enzymatically dissociated and loaded with the fluorescent Ca²⁺ indicator fura 2-AM (2.5 µM) as previously described (29). Fura 2-loaded cells were placed in a superfusion chamber and observed using an epifluorescence microscopy system (Nikon; Garden City, NY). Excitation light from a 300-W xenon arc lamp, passed via a liquid light guide, was directed through alternating 340- and 380-nm band-pass filters. Fluorescence emission (510 nm) from user-selected smooth muscle cells was synchronized with the appropriate excitation wavelength and reflected to an integrating charge-coupled device monochrome video camera (Cohu; San Diego, CA) with a dichroic mirror. Fluorescent images were acquired using InCa dual-wavelength calcium imaging software 2.1 (Intracellular Imaging; Cincinnati, OH). This microfluorometry system permits simultaneous evaluation of fura 2 fluorescence from multiple smooth muscle cells throughout an experimental protocol. Background fluorescence was determined before the experiment start for on-line subtraction during data collection. After the background fluorescence was subtracted, images obtained at 340 and 380 nm were ratioed on a pixel-by-pixel basis. Cells were continually superfused (~2.7 ml/min) under gravity flow. All experiments were conducted at room temperature (22-25°C), and fluorescence data were sampled every 2 s.

Sarco(endo)plasmic reticulum Ca²+-ATPase immunoblots. After the LCx and LAD coronary arteries were isolated from occluded hearts, arterial rings (inner diameter 0.8-1.2 mm; length ~0.5 mm) were cut, quick-frozen, and stored at 80°C for later immunoblot analysis of sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA)2b. Identification of SERCA2b was performed using a site-directed anti-peptide monoclonal antibody (Affinity BioReagents; Golden, CO) specifically against the SERCA2b isoform of the protein. Previous studies have documented that the SERCA of smooth muscle is transcribed exclusively from the SERCA2 isoform (8, 15), with SERCA2b accounting for ~80% of the total SERCA2 mRNA and the remainder encoding SERCA2a (5). Individual arterial rings were homogenized in 60 µl of SDS-polyacrylamide gel loading buffer containing 62.5 mM Tris · HCl (pH 6.8), 6 M urea, 2% SDS, 150 mM dithiothreitol, and 0.001% bromophenol blue. Samples were centrifuged at 14,000 g for 10 min at room temperature to remove insoluble material. Total protein content in the extract was assessed using the Bradford assay (Bio-Rad; Hercules, CA). Equal protein amounts (15 µg) were loaded onto each lane for SDS-PAGE (4-21% polyacrylamide precast gels; Bio-Rad). Direct electrophoretic transfer of proteins from the gel to a polyvinylidene difluoride membrane (Amersham; Arlington Heights, IL) was completed using a transfer buffer containing 0.7 M glycine and 25 mM Tris base. The membrane was blocked for 1 h in Tris-buffered saline (TBS; 50 mM Tris · HCl and 0.2 M NaCl; pH 7.4), 0.1% Tween, and 5% nonfat milk to reduce nonspecific binding and then incubated overnight in blocking buffer plus the SERCA2b antibody. After this incubation, the membrane was washed (TBS and 0.1% Tween) and incubated with a secondary anti-mouse antibody conjugated to horseradish peroxidase (Sigma; St. Louis, MO) for 1 h. The membrane was washed again, and the signal was developed using enhanced chemiluminescence (ECL; Amersham). Scanning densitometry (NIH Image 1.55) was used to quantify signal density from luminograms.

Statistical analysis. Heart weight-to-body weight ratio and citrate synthase activity in sedentary and exercise-trained animals were evaluated using unpaired Student's t-tests. Data for all fura 2 experiments were analyzed as a randomized complete block design in which the treatments were arranged in a 2 × 2 factorial. Mean differences were ascertained using Fisher's least significant difference test. Comparisons of immunoblot band density were accomplished using paired Student's t-tests in which responses from sedentary and exercise-trained LAD coronary arteries served as controls for the respective LCx arteries. Analyses for all fura 2 experiments and immunoblot band density were performed on a per vessel basis, where each animal was represented by a single LAD and LCx arterial segment. For all analyses, a P value 0.05 was considered significant. Data are presented as means ± SE, and the n values presented reflect the number of animals and the number of cells.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Training status. The effectiveness of the 16-wk exercise-training program was demonstrated by significant increases in skeletal muscle oxidative enzyme activity and increased heart weight-to-body weight ratios in exercise-trained pigs (Table 1).

Baseline myoplasmic free Ca²+ concentrations. Baseline fura 2 ratios, indicative of myoplasmic free Ca²⁺ concentrations, were not significantly different between smooth muscle cells isolated from the occluded LCx and nonoccluded LAD arteries of sedentary and exercise-trained pigs (sedentary LAD 0.92 ± 0.07, n = 14 vessels; sedentary LCx 0.91 ± 0.06, n = 16 vessels; exercise LAD 0.89 ± 0.07, n = 15 vessels; exercise LCx 0.84 ± 0.07, n = 15 vessels).

Fura 2 measurements of myoplasmic free Ca²+ accumulation. Figure 1A illustrates our protocol for evaluation of myoplasmic free Ca²⁺ accumulation in response to membrane depolarization with 80 mM KCl and a representative recording from a single sedentary control (LAD) cell. As shown in Fig. 1B, during membrane depolarization, the time to half-maximal (t_1/2) myoplasmic free Ca²⁺ accumulation was significantly prolonged in the presence (80K2) versus the absence (80K1) of partially depleted caffeine-sensitive SR Ca²⁺ stores. Furthermore, the initial maximal rate (slope, 10 s) of free Ca²⁺ accumulation in response to membrane depolarization was significantly greater during 80K1 versus 80K2 for both the LAD and LCx. These findings indicate that partial depletion of the SR Ca²⁺ store with caffeine increases the buffering capacity of the SR adequately to significantly decrease the rate of myoplasmic free Ca²⁺ accumulation in response to membrane depolarization (29).

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Ca2+ accumulation during membrane depolarization (80 mM KCl) using fura 2 microfluorometry. A: experimental protocol and a representative recording from a single sedentary (SED) control [left anterior descending (LAD) coronary artery] cell are presented, showing the change in the fluorescence ratio at wavelengths of 340 and 380 nm (F340/F380). Cells were superfused with physiological saline solution unless otherwise specified. Cells were exposed to 80 mM KCl (80K) and caffeine (CAF, 5 mM) in the presence of 2 mM extracellular Ca2+ for the durations indicated by the horizontal line. The initial 80K1 and CAF1 exposures served as control responses for the subsequent 80K2 and CAF2 responses. B: time to half-maximal (t1/2) Ca2+ accumulation in the absence (80K1) and presence (80K2) of a partially depleted SR Ca2+ store in cells isolated from the nonoccluded LAD (n = 10 animals, 78 cells) and collateral-dependent left circumflex (LCx) artery (n = 11 animals, 89 cells) of sedentary animals. Data are means ± SE. *P < 0.05 vs. 80K1; P < 0.05 vs. 80K2 LAD.

Inhibition of SR Ca²+ sequestration. In additional experiments, we used the SERCA inhibitor CPA (10 µM) to further investigate SR Ca²⁺ uptake in smooth muscle cells from the collateral-dependent vasculature of sedentary animals. The protocol for these experiments and a representative recording from a single sedentary control (LAD) cell are illustrated in Fig. 2A. As presented in Fig. 2B, in the absence of CPA, t_1/2 myoplasmic free Ca²⁺ accumulation in response to high KCl-induced membrane depolarization (80K2) was significantly reduced in smooth muscle cells from the collateral-dependent LCx when compared with the nonoccluded LAD artery. The addition of CPA significantly reduced t_1/2 myoplasmic free Ca²⁺ accumulation in the nonoccluded LAD but had little effect on t_1/2 myoplasmic free Ca²⁺ accumulation in the collateral-dependent LCx (59 ± 7 vs. 54 ± 6 s). Thapsigargin (1 µM) also decreased t_1/2 myoplasmic free Ca²⁺ from control values in nonoccluded LAD but not collateral-dependent LCx smooth muscle cells (58 ± 3 vs. 62 ± 4 s, respectively). Thus both SERCA inhibitors had minimal effect on the rate of myoplasmic free Ca²⁺ accumulation in LCx smooth muscle cells, indicating reduced SR Ca²⁺ uptake in smooth muscle distal to chronic occlusion. Additional experiments in control nonoccluded animals conducted in our laboratory (Fig. 2C) indicate that normal pigs do not demonstrate regional differences (LCx vs. LAD) in t_1/2 free Ca²⁺ accumulation in response to high KCl-induced membrane depolarization in the absence or presence of SERCA inhibition.

View larger version (34K): [in this window] [in a new window]   Fig. 2.   Evaluation of Ca2+ accumulation in the presence of sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA) inhibition during membrane depolarization (80 mM KCl). A: experimental protocols are similar to that used in Fig. 1A with the addition of cyclopiazonic acid (CPA) for the 11-min period designated by the horizontal line. A representative recording from a single sedentary control (LAD) cell is presented, showing the change in F340/ F380. B: t1/2 Ca2+ accumulation in the absence and presence of SERCA inhibition for cells isolated from LAD and LCx arteries of sedentary animals. n values for the control protocol (without CPA) are as presented in the Fig. 1 legend; n values for the CPA protocol are LAD = 8 animals, 77 cells and LCx = 8 animals, 60 cells. C: t1/2 Ca2+ accumulation in the absence or presence of SERCA inhibition in cells isolated from control nonoccluded animals indicate that normal pigs do not demonstrate regional differences between LAD (n = 6 animals, 67 cells) and LCx (n = 6 animals, 66 cells) arteries in Ca2+ regulation in response to high KCl-induced membrane depolarization. D: t1/2 Ca2+ accumulation in the absence and presence of SERCA inhibition for cells isolated from LAD and LCx arteries of exercise-trained (EX) animals. n values for the control protocol are LAD = 8 animals, 73 cells and LCx = 8 animals, 67 cells. n values for the CPA protocol are LAD = 8 animals, 78 cells and LCx = 7 animals, 64 cells. Data are means ± SE. *P < 0.05 vs. LAD without CPA; P < 0.05 vs. LCx without CPA.

Exercise training and SR Ca²+ sequestration. We also evaluated SR Ca²⁺ uptake in smooth muscle cells isolated from collateral-dependent LCx and nonoccluded LAD coronary arteries of exercise-trained animals. Similar to our findings in sedentary animals, in the presence of partially depleted caffeine-sensitive SR stores, t_1/2 myoplasmic free Ca²⁺ accumulation in response to high KCl-induced membrane depolarization was significantly reduced in smooth muscle cells from the collateral-dependent LCx compared with the nonoccluded LAD artery (Fig. 2D). Furthermore, the addition of CPA significantly reduced t_1/2 myoplasmic free Ca²⁺ accumulation in the nonoccluded LAD but had little effect on t_1/2 myoplasmic free Ca²⁺ accumulation in the collateral-dependent LCx (Fig. 2D). These data are contrary to our hypothesis and suggest exercise training does not correct impaired SR Ca²⁺ uptake in collateral-dependent coronary smooth muscle cells.

Fura 2 measurements of Ba²+ influx. In additional experiments, we assessed divalent cation influx through voltage-gated Ca²⁺ channels using 2 mM Ba²⁺ in the presence of low extracellular Na⁺ (5 mM) to determine whether the enhanced rate of myoplasmic free Ca²⁺ accumulation in smooth muscle cells from the LCx artery was partially dependent on enhanced Ca²⁺ influx at the plasma membrane. Our protocol for these experiments and a representative recording from a single sedentary control (LAD) cell are presented in Fig. 3A. As shown in Fig. 3B, neither maximal rate nor net Ba²⁺ accumulation was significantly enhanced in smooth muscle cells isolated from the collateral-dependent LCx of sedentary animals. These findings suggest that alterations in Ca²⁺ influx do not contribute to the reduced t_1/2 myoplasmic free Ca²⁺ accumulation observed in response to membrane depolarization in smooth muscle cells isolated from the LCx artery.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Ba2+ accumulation during membrane depolarization (80 mM KCl) using fura 2 microfluorometry. A: experimental protocol and a representative recording from a single sedentary control (LAD) cell are presented, showing the change in F340/F380. The cell was exposed to 80 mM KCl (2Ca80K) and caffeine (5 mM) in the presence of 2 mM extracellular Ca2+ for the durations indicated by the horizontal line. The cell was then exposed to 80 mM KCl in the presence of 2 mM Ba2+ (2Ba80K5Na; equimolar substitution for Ca2+) for 7 min. B and C: maximal rate (slope, 10 s) and net (AUC) Ba2+ accumulation throughout membrane depolarization (7 min) for cells isolated from the nonoccluded LAD and collateral-dependent LCx from sedentary (B; LAD = 9 animals, 72 cells and LCx = 9 animals, 79 cells) and exercise-trained (C; LAD = 6 animals, 65 cells and LCx = 6 animals, 55 cells) animals. AUC was obtained by subtracting the baseline fura 2 ratio for each cell (average of 10 data points before 80K2 exposure) from each subsequent fura 2 data point throughout the 80 mM KCl exposure (minutes 11-18). These fura 2 ratio values for each data point were then added together to obtain AUC. Data are means ± SE. NS, not significant.

Myoplasmic free Ca²+ removal during repolarization. Myoplasmic free Ca²⁺ removal rate during recovery from 80 mM KCl-induced membrane depolarization was assessed to evaluate the function of Ca²⁺-removal mechanisms in smooth muscle cells isolated from collateral-dependent vasculature of sedentary and exercise-trained pigs. We further evaluated free Ca²⁺ removal both in the absence (Fig. 1A; minutes 18-20) and presence of CPA (Fig. 2A; minutes 18-20) to determine the specific contribution of SR sequestration in Ca²⁺ removal after depolarization. t_1/2 myoplasmic free Ca²⁺ removal during repolarization was not significantly different in the absence or presence of CPA in both sedentary and exercise-trained animals (Fig. 4, A and B, respectively). These data suggest that impaired SR Ca²⁺ uptake in LCx smooth muscle cells, evident during depolarization, was not detected during repolarization most likely because depolarization-induced refilling of the SR Ca²⁺ store limited subsequent Ca²⁺ sequestration during repolarization. Furthermore, t_1/2 myoplasmic free Ca²⁺ removal during repolarization was not significantly different between smooth muscle cells isolated from the collateral-dependent LCx and nonoccluded LAD of both sedentary and exercise-trained pigs (Fig. 4, A and B, respectively). Taken together, these findings indicate that the remaining Ca²⁺-removal mechanisms, primarily sarcolemmal Ca²⁺-ATPase and Na/Ca exchange, function similarly in smooth muscle cells isolated from the collateral-dependent LCx and nonoccluded LAD of both sedentary and exercise-trained animals.

View larger version (31K): [in this window] [in a new window]   Fig. 4.   Evaluation of Ca2+ removal during recovery from membrane depolarization (80 mM KCl) using fura 2 microfluorometry. Data were obtained for t1/2 Ca2+ removal during minutes 18-20 in the absence (Fig. 1A) and presence (Fig. 2A) of SERCA inhibition (10 µM CPA) for cells isolated from LAD and LCx arteries of sedentary (A) and exercise-trained animals (B). n values for protocols are as presented in the Fig. 2 legend. Data are means ± SE.

SERCA immunoblots. We also assessed SERCA2b protein content in arterial rings from collateral-dependent LCx and nonoccluded LAD coronary arteries of sedentary (n = 7) and exercise-trained (n = 7) animals to determine whether decreases in protein content are responsible for impaired Ca²⁺ uptake by the SR. Luminograms of SERCA2b protein content (Fig. 5) were quantified using densitometric analysis; band density of the LCx artery was expressed relative to the respective control LAD from the same heart. Our experiments indicated that the LCx artery demonstrated 99 ± 9 and 99 ± 8% of LAD SERCA2b protein content in sedentary and exercise-trained animals, respectively, suggesting decreased protein content did not contribute to impaired SR Ca²⁺ uptake.

View larger version (75K): [in this window] [in a new window]   Fig. 5.   Effect of chronic coronary artery occlusion on SERCA2b protein content in the smooth muscle of nonoccluded LAD and collateral-dependent LCx arteries from sedentary and exercise-trained swine. Representative luminogram of SERCA2b (~110 kDa; bold arrow) content in arterial rings of sedentary (lanes 1 and 2) and the exercise-trained (lanes 3 and 4) animals. LAD arteries (lanes 1 and 3) from sedentary and exercise-trained pigs served as controls for the respective LCx arteries (lanes 2 and 4). Chronic occlusion did not alter protein content of SERCA2b in arterial rings of sedentary and exercise-trained animals.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The current study demonstrates that SR Ca²⁺ uptake is impaired in intact smooth muscle cells isolated from coronary arteries distal to chronic occlusion. These data also suggest that exercise training does not correct the impaired SR Ca²⁺ sequestration in these smooth muscle cells. Furthermore, alterations in SERCA2b protein content were not responsible for impaired SR Ca²⁺ sequestration in smooth muscle cells isolated from the LCx artery of either sedentary or exercise-trained animals.

Collateral-dependent coronary arteries exhibit altered vasomotor reactivity in vitro that can be partially attributed to alterations in myoplasmic Ca²⁺ handling (11, 20). Impaired adenosine-induced relaxation in these collateral-dependent arteries is associated with an impaired adenosine-stimulated reduction in myoplasmic free Ca²⁺, which is corrected with exercise training (11). Increases in cAMP-dependent protein kinase, a product of adenosine stimulation, have been reported to affect multiple Ca²⁺ regulatory mechanisms to reduce intracellular Ca²⁺ levels, including SR Ca²⁺ uptake from the bulk myoplasm (19). We postulated that the impaired adenosine-mediated reduction in myoplasmic Ca²⁺ in the collateral-dependent artery (11) might be attributable to impaired SR Ca²⁺ sequestration. Although our present data suggest that SR Ca²⁺ uptake is impaired in the smooth muscle distal to chronic occlusion, the impaired SR Ca²⁺ sequestration was not corrected with exercise training. These results indicate that exercise training corrects impaired adenosine-stimulated reductions in myoplasmic free Ca²⁺ in collateral-dependent arteries (11) by other mechanisms. Figures 3C and 4B present results demonstrating that Ca²⁺ entry via voltage-gated Ca²⁺ channels and Ca²⁺ removal via Ca²⁺-ATPase and Na/Ca exchange are also not altered in cells from collateral-dependent arteries of exercise-trained pigs. Thus the available evidence indicates that the previously reported (11) training-induced correction of adenosine-mediated reduction in intracellular Ca²⁺ levels results from adaptation of other cellular mechanisms involved in Ca²⁺ regulation that are specific to adenosine-mediated decreases in Ca²⁺. For example, adenosine-induced relaxation is partly mediated by K⁺ channel activation (16), and increased K⁺ channel dependence of arterial tone has been demonstrated in coronary arteries isolated from exercise-trained versus sedentary pigs not subjected to coronary occlusion (2). Therefore, adenosine-mediated decreases in intracellular Ca²⁺ levels, reported in collateral-dependent arteries from trained pigs, may be the result of increased K⁺ channel activation.

The finding that the t_1/2 myoplasmic free Ca²⁺ accumulation was more rapid in the absence (80K1) versus the presence (80K2) of depleted caffeine-sensitive SR Ca²⁺ stores (Fig. 1B) confirms previous results in normal coronary arteries (29). Caffeine depletes the SR adequately to potentiate SERCA activity (5, 29) and, therefore, examine the role of Ca²⁺ uptake by SERCA in coronary smooth muscle cells. Furthermore, numerous studies have reported that the superficial SR sequesters a portion of the Ca²⁺ that enters through the plasmalemma of smooth muscle cells during high KCl-induced depolarization (14, 29, 31, 32). Accumulation of incoming Ca²⁺ by the SR near the site of Ca²⁺ entry functions to limit Ca²⁺ entry into the bulk myoplasm and subsequent activation of smooth muscle contractile proteins (14, 31, 32). Our data demonstrate that only after SR depletion (80K2), t_1/2 myoplasmic free Ca²⁺ accumulation in response to membrane depolarization was significantly reduced in smooth muscle cells isolated from the collateral-dependent LCx versus nonoccluded LAD (Fig. 1B). These findings suggest either impaired SERCA function or disruption of the structural/functional relationship between voltage-gated Ca²⁺ channels and SERCA in smooth muscle cells isolated from collateral-dependent vasculature. Furthermore, the prolonged t_1/2 myoplasmic free Ca²⁺ accumulation (Fig. 2, C vs. B) in the LAD of control nonoccluded (~110 s) versus occluded (~80 s) animals suggests that the presence of chronic occlusion may produce global alterations in Ca²⁺ handling in both collateral-dependent and nonoccluded vasculature of chronically occluded hearts.

Hypothetically, the enhanced rate of depolarization-induced free Ca²⁺ accumulation observed in the presence of SR depletion may also reflect changes in Ca²⁺ influx at the sarcolemma or alterations in Ca²⁺ extrusion via sarcolemmal Ca²⁺-ATPase or Na/Ca exchange. Evaluation of Ba²⁺ accumulation in the presence of low extracellular Na⁺ has been used previously as a measure of unidirectional divalent cation influx in smooth muscle cells (13). Ba²⁺ is transported via voltage-gated Ca²⁺ channels (1) and Na/Ca exchanger (3) but is a poor substrate for ATP-dependent Ca²⁺ pumps (23). Ba²⁺ produces a shift in the fura 2 excitation wavelength spectrum, with increasing concentrations similar to that observed with Ca²⁺, although fura 2 has a higher affinity for Ca²⁺ versus Ba²⁺ (12, 23). However, Ba²⁺ is not sequestered by intracellular organelles (3, 23) and, therefore, allows us to evaluate influx at the plasma membrane under conditions in which corresponding Ca²⁺ fluxes would be difficult to interpret due to multiple pathways for Ca²⁺ removal and sequestration. Depolarization-induced Ba²⁺ influx in the presence of low extracellular Na⁺ revealed that divalent cation influx at the sarcolemma is not different between smooth muscle cells isolated from nonoccluded LAD and collateral-dependent LCx arteries of both sedentary and exercise-trained pigs (Fig. 3, B and C, respectively).

Furthermore, one might speculate that partial depletion of the SR with caffeine may stimulate Ca²⁺ entry through store-operated channels, refilling the SR and thereby reducing t_1/2 myoplasmic free Ca²⁺ accumulation. However, previous experiments indicate that SR store depletion does not activate Ca²⁺ influx in our porcine coronary smooth muscle cell preparation (4). Furthermore, data from the present study demonstrate that myoplasmic free Ca²⁺ concentration returned to baseline levels (Fig. 1A; minute 11) after partial depletion of the SR in cells isolated from all groups (sedentary LAD 0.82 ± 0.10, n = 14 vessels; sedentary LCx 0.84 ± 0.08, n = 16 vessels; exercise LAD 0.91 ± 0.08, n = 15 vessels; exercise LCx 0.88 ± 0.08, n = 15 vessels), indicating neither exercise training nor coronary occlusion enhances Ca²⁺ influx through store-operated channels in these smooth muscle cells.

The rate of reduction in myoplasmic free Ca²⁺ during repolarization, measured after KCl-induced membrane depolarization, was not different between LAD and LCx smooth muscle cells of sedentary or exercise-trained pigs. These observations indicate that the Ca²⁺ extrusion mechanisms, sarcolemmal Ca²⁺-ATPase and Na/Ca exchange, were not discernibly different between cells of these arteries. These findings further support our hypothesis that impaired SR Ca²⁺ uptake contributes to altered myoplasmic free Ca²⁺ regulation in smooth muscle cells from collateral-dependent vasculature and potentially subsequent alterations in vasomotor reactivity distal to chronic occlusion.

Our findings that t_1/2 Ca²⁺ removal (time for half-recovery of free Ca²⁺ to minimum levels) and net Ca²⁺ removal after CAF1 exposure (minutes 9-11) were not significantly different between smooth muscle cells of the LAD and LCx do not appear to support the hypothesis of reduced SR Ca²⁺ uptake. Generally, after washout of caffeine, recovery of myoplasmic Ca²⁺ is largely attributed to Ca²⁺ uptake by the SR (6). However, one might speculate that SR sequestration of Ca²⁺ entering the cell (e.g., membrane depolarization) may be a different cellular process than SR sequestration of myoplasmic free Ca²⁺ elevated by intracellular release (e.g., caffeine). Alternatively, disruption of the structural/functional relationship between voltage-gated Ca²⁺ channels and SERCA in smooth muscle cells isolated from collateral-dependent vasculature may explain the discrepancy between SR sequestration of Ca²⁺ entering the cell versus myoplasmic free Ca²⁺ elevated by intracellular release. We also report that after partial depletion of caffeine-sensitive SR Ca²⁺ stores (CAF1), subsequent membrane depolarization (80K2) replenished SR Ca²⁺ stores (CAF2) similarly in LCx and LAD smooth muscle cells. Although these data do not support our hypothesis of impaired SR Ca²⁺ uptake, the long duration of membrane depolarization (7 min) may have been adequate for complete refilling of the SR stores even in the presence of a reduced rate of Ca²⁺ sequestration by the SR of LCx smooth muscle cells.

Thus several sets of data from the present study indicate decreased SR Ca²⁺ uptake in vascular smooth muscle of collateral-dependent coronary arteries. Our results indicate that this decreased sequestration is present with normal amounts of coronary smooth muscle SERCA protein. Our results do not reveal the cause of decreased SR Ca²⁺ uptake; however, there is evidence in the literature that hearts subjected to global ischemia demonstrate impaired SR Ca²⁺ uptake and reduced SERCA mRNA and protein levels in cardiac muscle cells (30), changes that may be associated with ischemia-mediated production of oxygen-derived free radicals (18, 33). In turn, the production of reactive oxygen species has been documented to inactivate SERCA and increase SR membrane permeability to Ca²⁺ in porcine coronary smooth muscle (9, 10). These studies (9, 10) demonstrate that even low concentrations of reactive oxygen species (e.g., 3 × 10⁶ M hydrogen peroxide) can reduce SR Ca²⁺ uptake by ~20%, a reduction similar to that observed in our study. Furthermore, studies (21, 22) using a porcine model of chronic occlusion have demonstrated that exercise-induced myocardial ischemia persists in these animals long after coronary collateral development is adequate to restore blood flow and regional myocardial function to normal resting levels in the collateral-dependent region. Thus available evidence supports the notion that SERCA activity in smooth muscle cells isolated from collateral-dependent vasculature in our model of chronic coronary occlusion may be altered by reactive oxygen species generated by ischemic conditions. Reactive oxygen species-induced injury to SERCA of the collateral-dependent LCx artery may impair the activity of this Ca²⁺-transport mechanism.

We conclude that these data provide strong evidence for impairment of SR Ca²⁺ uptake at the level of the single smooth muscle cell in collateral-dependent coronary arteries. Impaired SERCA activity or disruption of the structural/functional relationship between voltage-gated Ca²⁺ channels and SERCA in coronary smooth muscle of collateral-dependent vasculature could explain a reduced ability of the SR to effectively sequester incoming Ca²⁺, thus leading to increased availability of Ca²⁺ to contractile proteins in the bulk myoplasm. The resulting enhanced contractile responsiveness of vasculature distal to chronic occlusion would further compromise blood flow to collateral-dependent myocardium.