INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

SUBSTANTIAL EVIDENCE EXISTS supporting the role of chronic endurance exercise training in reducing the incidence of coronary heart disease (6, 39), yet the mechanism(s) of exercise-related cardioprotection remains unknown. Coronary heart disease can be subdivided into two general categories: diseases of coronary vessel tone, i.e., acute vasospasm and hypertension, and diseases of vessel injury, i.e., atherosclerosis. Several studies from our laboratory have addressed mechanisms associated with exercise training adaptations (for reviews, see Refs. 10 and 43). However, there are few studies (48) addressing exercise-related coronary smooth muscle adaptations and their relationship to resistance to the development of vascular disease.

Atherosclerosis is a complex disease involving an inflammatory-fibroproliferative response that develops from various forms of insult to the endothelium and smooth muscle cells of the artery (38). The phenotypic modulation of smooth muscle cells plays a key role in the development of atherosclerosis and is characterized in part by increased DNA synthesis (34); altered functional receptor expression, including receptors involved in growth factor signaling (29); and subcellular/ultrastructure morphology (35, 38, 47). In cells from diseased vessels, Ca²⁺ regulation is also altered (18) and implicated in increased vessel tone and vasospasm (27, 28). We recently reported that localized Ca²⁺ signaling events, specifically nuclear Ca²⁺ (Ca_n) signaling in response to endothelin-1 (a potent vasoconstrictor and mitogen), are also altered in cells from atherosclerotic vessels (47). In contrast, we have shown that exercise training increases the overall regulation of bulk myoplasmic Ca²⁺ (Ca_m) in coronary smooth muscle, ultimately decreasing constriction to endothelin-1 and improving endothelium-dependent relaxation, consistent with reducing coronary vessel tone (10, 43). Thus, in this study, we hypothesized that coronary smooth muscle after exercise training would be more resistant to endothelin-1-induced phenotypic modulation and that enhanced Ca_n regulation may, in part, play a role in this resistance.

In smooth muscle, a portion of the Ca²⁺ that enters the cell can be buffered by sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA) on the sarcoplasmic reticulum (SR) and preferentially unloaded toward the sarcolemma for extrusion from the cell by sarcolemmal Ca²⁺-ATPase and the Na⁺/Ca²⁺ exchanger. A similar mechanism is enhanced after exercise training and may underlie the decreased vasoconstrictor responses associated with exercise training (7-9, 17, 33, 41, 42). Thus, in this study, we extended this mechanism deeper into the cell and hypothesized that a target for enhanced Ca_n regulation associated with exercise training is the specific subcellular localization of SERCA in relation to the nucleus.

The present study determined 1) the effects of exercise training on endothelin-1-induced coronary smooth muscle phenotypic modulation and Ca_n regulation and 2) the subcellular localization of SERCA. In support of our hypotheses, we found novel evidence that exercise training decreases endothelin-1-induced smooth muscle phenotypic modulation and increases the ability of the cell to buffer Ca_n, possibly via SERCA localization to the nuclear envelope. This is the first report of altered Ca_n regulation associated with an intrinsic resistance to coronary smooth muscle phenotypic modulation produced by exercise training.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Exercise Training Procedures

Isolation of Coronary Arteries

Skeletal Muscle Oxidative Enzyme Activity

Single Smooth Muscle Cell Isolation

Endothelin-1-Induced Smooth Muscle Cell Phenotypic Modulation

5-Bromo-2'-deoxyuridine labeling. Our lab has previously described an organ culture model of smooth muscle cell phenotypic modulation characterized, in part, by increased myoplasmic calcium responsiveness to the P2 nucleotide receptor agonist UTP and increased single cell DNA content as determined by the DNA-specific fluorescent probe 4',6-diamidino-2-phenylindole dihydrochloride (DAPI) (19, 20). To confirm that smooth muscle cells in this model are undergoing DNA synthesis and that the DAPI measures are consistent with DNA synthesis, right coronary artery segments from domestic swine were opened longitudinally, placed with the lumen facing up in a 100-mm petri dish in serum-free RPMI 1640 (Life Technologies; Grand Island, NY), and either cold stored at 5°C or organ cultured for 2 days. 5-Bromo-2'-deoxyuridine (BrdU; 20 µmol/l) was added to the culture medium 6 h before the tissue was fixed in 10% phosphate-buffered formalin. Tissue samples were paraffin embedded, and 6-µm-thick cross sections were processed for BrdU labeling in nuclei employing a horseradish peroxidase detection system (BrdU Labeling and Detection Kit II, Roche). Note that utilization of BrdU incorporation would provide a precise measure of DNA synthesis rates (3) compared with DAPI measures of DNA content. The thymidine analog BrdU stays in the cell and is recognized as different than thymidine by analytic methods (i.e., antibodies) after cell division. In contrast, DAPI fluorescence increases only before cell division and then decreases to normal after division because it binds native DNA. Thus DAPI may underestimate the full extent of DNA synthesis.

Free Ca_m measurement with fura 2. To determine the effect of endothelin-1 on smooth muscle phenotypic modulation in organ culture, Ca_m responsiveness to the P2 nucleotide receptor agonist UTP was determined in cells from arterial segments that were cold stored at 5°C or organ cultured for 2 days in the presence or absence of 5 × 10⁸ M endothelin-1 at 37°C in a 95% O₂-5% CO₂ incubator. We chose to assay for Ca_m responsiveness to UTP because the Ca²⁺ measurements utilized allow for a quantitative assessment in response to enothelin-1.

6

DNA imaging. Distal right coronary artery segments (>1.0 mm) from SED and EX swine were sectioned into 3-mm rings and placed in a 60-mm petri dish in RPMI 1640 (Life Technologies; Grand Island, NY). Arterial segments were then cold stored for 2 days at 5°C or organ cultured for 2 days in the presence or absence of 5 × 10⁸ M endothelin-1 at 37°C in a 95% O₂-5% CO₂ incubator.

7

Regional Ca²+ Imaging Using Confocal Microscopy

Confocal Ca²+ measurement with fluo 3 and analysis. Cells from SED and EX pigs were loaded with 5 µM fluo 3-AM (Molecular Probes) for 30 min at 37°C. A drop of the fluo 3-loaded cellular suspension was placed on a coverslip inside a constant-flow superfusion chamber. Observation of intracellular Ca²⁺ was made using a Noran Oz laser scanning confocal system through an oil immersion ×60 (NA 1.4) objective on an inverted Nikon microscope. A cell was illuminated using the 488-nm line of an argon-krypton laser, and a high-gain photomultiplier tube collected the emission after it had passed through a 500- to 550-nm bandpass filter. All parameters (i.e., laser intensity, gain, and slit size) were kept constant for all experiments. Generally, acquisition speed was set to 8 ms/frame with a four-frame integration (jump averaging method) resulting in an effective frame rate of 32 ms/frame. Images were taken from the middle focal plane of the cell as determined by equally dividing the distance from preexperimental images taken from the bottom and top of the cell. Cells were constantly superfused with PSS. Images were acquired at 30-s intervals during PSS perfusion and 80 mM K depolarization, 10-s intervals during endothelin-1 exposure, and continuously during caffeine exposure.

Localization of SERCA

BODIPY FL-thapsigargin. Smooth muscle cells were dispersed from right coronary artery segments of domestic swine as described above. For localization of SERCA in living smooth muscle cells, the cellular suspension was incubated with PSS containing a sarcolemmal dye, Di-8-ANEPPS (10 µM, 6 min, excitation wavelength 465 nm, emission wavelength 635 nm, Molecular Probes), and a fluorescent-tagged SERCA inhibitor, BODIPY FL-thapsigargin (B-TSG; 10 µM, 6 min, excitation wavelength 503 nm, emission wavelength 511 nm, Molecular Probes). Images were acquired on a Bio-Rad Lasersharp MRC-600 laser scanning confocal system (Bio-Rad Laboratories; Hercules, CA) through an oil immersion ×60 (NA 1.4) objective mounted on an invert microscope. The confocal pinhole aperture and the photomultiplier gain were kept constant. For control experiments, cells were incubated with unconjugated BODIPY C₃ (1 µM, 15 min, excitation wavelength 493 nm, emission wavelength 503 nm, Molecular Probes). All digital images were merged and pseudocolored in Adobe Photoshop.

Western blot analysis. To determine the immunospecificity of monoclonal (mouse) anti-SERCA2 ATPase antibody (IgG1, SERCA2b, Affinity Bioreagents), protein extracts were collected from the domestic porcine right coronary artery. Sample protein concentrations were measured using the Bradford protein assay (Bio-Rad). A canine cardiac muscle protein extract (Affinity Bioreagents) was also prepared as a positive control. Extracts were resolved via SDS-PAGE on a 4-20% acrylamide gradient gel and transferred to a polyvinylidene difluoride membrane. Blots were incubated in blocking solution (50 mM Tris, 200 mM NaCl, 5% dry milk, and 0.01% Tween 20). Blots were exposed to the SERCA2b antibody (1:1,000) for 1 h in the blocking solution, washed, and incubated with an alkaline phosphatase-conjugated anti-mouse IgG secondary antibody for 1 h. Blots were washed and visualized using chemiluminescence. As seen in Fig. 6, this antibody was specific for SERCA2b (~115 kDa) in porcine coronary smooth muscle.

Immunofluorescence labeling. Smooth muscle cells were dispersed from right coronary artery segments of domestic swine as described above. Cells were fixed in 2% paraformaldehyde and washed with SSC buffer (150 mM NaCl and 1.4 mM sodium citrate; pH 7.2). Cells were permeablized in SSC buffer containing 0.1% saponin and 0.1% BSA. Cells were incubated with either monoclonal (mouse) anti-SERCA2b antibody (1:500), monoclonal (mouse) nuclear transport factor p97 antibody (IgG2a, 1:2,000, Affinity Bioreagents), or monoclonal (mouse) anti-smoothelin antibody (IgG1, 1:500, Affinity Bioreagents) in antibody buffer (SSC containing 5% BSA, 2% goat serum, 0.1% saponin, and 0.02% NaN₃) for 12-16 h at 5°C. Cells were washed for 3 h in SSC containing 0.1% saponin and 0.1% BSA and incubated with goat anti-mouse IgG antibody conjugated with Alexa fluor 488 (IgG, 1:500, excitation wavelength 494 nm, emission wavelength 517 nm, Molecular Probes) in antibody buffer for 4 h. The cells were washed for 3 h in SSC containing 0.1% saponin and 0.1% BSA, incubated with the DNA indicator DAPI (2.5 × 10⁷ M, 15 min), washed, and mounted onto positively charged slides.

Staining of subcellular membranes for electron microscopy. Right coronary artery segments from domestic swine were stained similar to methods described by Forbes et al. (14). Briefly, proximal coronary arteries were cut into 3-mm rings and fixed in a 0.1 M cacodylate buffer solution that contained 2% glutaraldehyde, 2% paraformaldehyde, and 5 mM CaCl₂. The ring segments were postfixed in 1% OsO₄ and 1.5% potassium ferricyanide for 2 h at room temperature followed by a buffer wash. Samples were incubated in 2% aqueous uranyl acetate at room temperature for 30 min and dehydrated in a graded series of ethanol concentrations up to 100%. The ring segments were then placed in EMBED araldite resin overnight, embedded in resin, and polymerized at 60°C. Sections were cut at <100-nm thickness (LKB Ultratome III) using diamond knives (Diatome), mounted on 200-mesh copper grids, and poststained with lead citrate and 3% uranyl acetate. Images were acquired on a Jeol 1200EX transmission electron microscope at 80 keV.

Immunogold labeling for electron microscopy. Domestic swine right coronary arteries were cut into 3-mm rings and fixed in a 0.1 M cacodylate buffer solution that contained 0.1% glutaraldehyde and 4% paraformaldehyde for 3 h. The samples were rinsed three times in 0.1 M cacodylate buffer and then dehydrated in a graded series of ethanol concentrations up to 100%. Embedding was by transfer into a 2:1 ratio of ethanol-LR White resin overnight at 5°C. Samples were then placed into pure LR White for 3 days, embedded in fresh LR White resin, and polymerized at 60°C. Sections were cut at <100-nm thickness, mounted on 200-mesh nickel grids, and incubated for 1 h in antibody buffer (20 mM Tris, 20 mM NaN₃, 150 mM NaCl, 1.0% BSA, and 0.15% glycine). The samples were incubated with monoclonal (mouse) anti-SERCA2b antibody (1:10, Affinity Bioreagents) in antibody buffer overnight at 5°C. All grids were washed 10 times in blocking buffer and then incubated with goat anti-mouse IgG antibody conjugated with 10 nm gold (1:10; Sigma) in blocking buffer overnight at 5°C. The grids were washed 5 times in blocking buffer, 10 times in distilled deionized H₂O, and stained for 5 min in uranyl acetate and 5 min in lead citrate. Images were acquired on a Jeol 1200EX transmission electron microscope.

Statistics

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Efficacy of Exercise Training

Endothelin-1 Effect in Organ Culture

View larger version (77K): [in this window] [in a new window]   Fig. 1.   Endothelin-1 (ET-1) is a mitogen and induces phenotypic modulation in organ cultures. The images in A and B represent 5-bromo-2'-uridine (BrdU) labeling in right coronary artery cross sections either cold-stored (A) or organ-cultured vessels (B) for 2 days. L, lumen; M, medial smooth muscle; A, adventitia; arrowheads, BrdU-labeled nuclei. C: typical tracings from cold-stored cells (dashed line) and organ-cultured plus ET-1 cells (solid line) where the fura 2 ratio is a relative indicator of myoplasmic Ca2+ (Cam). Smooth muscle cells were continuously superfused with physiological saline solution (PSS) unless otherwise indicated by the horizontal lines, which represent exposure times to the agonists: 80 mM KCl (80K) and 10 µM UTP. * Significant difference (P < 0.05). The bar graph in D represents the percentage of cells responding to the nucleotide receptor agonist UTP (10 µM) in cold-stored cells (2 responder cells of 27 total cells), organ-cultured cells (7 responder cells of 62 total cells), and organ-cultured cells in the presence of ET-1 (5 × 108 M, 24 responder cells of 56 total cells). Artery segments were cultured for 2 days. Cells were defined as responders to UTP if the fura 2 response was at least 3 SD above baseline for 5% of the time during exposure.

To test the hypothesis that endothelin-1 accelerates the phenotypic modulation in response to organ culture, right coronary artery segments were either cold stored, organ cultured, or organ cultured in the presence of endothelin-1 (5 × 10⁸ M) for 2 days. The percentage of cells that responded to UTP (10 µM) increased fourfold (P < 0.05) in organ culture in the presence of endothelin-1 compared with organ culture alone or cold-stored arteries (Fig. 1D). The Ca_m responses to UTP were increased fivefold (P < 0.05) in cells from vessels organ cultured in the presence of endothelin-1 (0.89 ± 0.13 ratio units above baseline) compared with cold-stored (0.16 ± 0.02 ratio units) vessels; there was no difference compared with organ-cultured vessels (1.13 ± 0.28 ratio units; Fig. 1C). These results confirm that endothelin-1 increases UTP responsiveness in organ culture, also a marker for smooth muscle cell phenotypic modulation in this model (20).

Endothelin-1-Induced Coronary Smooth Muscle Phenotypic Modulation

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Exercise training attenuates the ET-1-induced increase in DNA content. Immunocytochemistry for smoothelin revealed that both the elongated cell phenotype from cold-stored arteries (A) and the rounded cell phenotype from organ-cultured arteries (B) stained positive for smoothelin. A-D, insets: light images for each cell. C and D: DNA staining with 4,6'-diamidino-2-phenylindole dihydrochloride (DAPI) in typical cells from 2-day cold-stored and organ-cultured arteries, respectively. The green line around the cell wall represents the area of interest (AOI) that was set before quantification of the mean pixel intensity of DAPI fluorescence. Notice the rounded morphology and relative increase in DAPI fluorescence in the organ-cultured cell (D) compared with the elongated cold-stored cell (C). E: the y-axis of the bar graph represents the relative DAPI fluorescence (0-255 pixel intensity units) in coronary smooth muscle cells from sedentary (SED) and exercise-trained (EX) animals in the following conditions: cold stored (n = 88 cells from 7 SED animals and 66 cells from 6 EX animals), organ cultured (n = 88 cells from 7 SED animals and 57 cells from 6 EX animals), and organ cultured in the presence of ET-1 (5 × 108 M; n = 96 cells from 7 SED animals and 81 cells from 6 EX animals). * Significant difference (P < 0.05).

Endothelin-1-Induced Changes in Regional Calcium

8

View larger version (33K): [in this window] [in a new window]   Fig. 3.   Determining regional Ca2+ responses in coronary smooth muscle. After an experiment, cells were superfused with SYTO-64 (500 nM). A and B: three-dimensional reconstruction of SYTO-64. To confirm that the experimental focal plane lay within the nucleus (N), a confocal fluo 3 image attained during the experiment was inserted into the SYTO-64 volume data set. Only cells where the experimental fluo 3 focal plane lay within the nucleus (A) were used for off-line fluo 3 data analysis. A, inset: two-dimensional image of SYTO-64 and fluo 3 in the experimental focal plane. As depicted in B, the experimental focal plane could lay above or below the nucleus. C-G: high-resolution fluo 3 (C) and SYTO-64 (D) images were collected after the an experiment, and regional Ca2+ responses were analyzed by generating an AOI for the whole cell (WC AOI; E), the nucleus (NUC AOI; F), and the cytosol (CYT AOI; G). H: raw fluo 3 intensity (F) data generated from the regional AOIs (top) and regional fluorescence changes (F/F0; bottom). Caw, whole cell Ca2+; Cac, cytosolic Ca2+; Can, nuclear Ca2+; BG, background. Smooth muscle cells were continuously superfused with PSS unless otherwise indicated by the horizontal lines, which represent exposure times to the agonists: 80 mM KCl and 5 × 108 M ET-1. The arrows on the x-axis in H denote the time points for the images in I depicting the transient increase in Caw in response to ET-1.

In smooth muscle cells from SED pigs, there were no differences in the peak Ca_c and Ca_n responses to endothelin-1 (Fig. 4, A and B), and the times to half-minimum (t_1/2) response from peak and relative duration of the Ca_c and Ca_n responses were not different (Table 2); t_1/2 to minimum is the time required for the F/F₀ signal to reach one-half of the absolute peak response from peak to baseline (minimum F/F₀). In contrast, smooth muscle cells from EX pigs exhibited a 20% decrease in the peak endothelin-1-induced Ca_n response compared with the Ca_c response (Fig. 4, A and C). The relative Ca_n t_1/2 to minimum response from peak and the relative response duration were decreased compared with Ca_c (Table 2).

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Exercise training attenuates ET-1-induced Cac and Can responses. A: bar graph representing the mean peak ET-1-induced Cac and Can responses to ET-1 (5 × 108M) in coronary smooth muscle cells from SED and EX pigs. The y-axis represents the change in F/F0 from baseline to maximum. The Cac and Can responses were decreased in EX compared with SED pigs, and, in EX, the Can response was less than the Cac response. * Significant difference (P < 0.05). B and C: typical ET-1-induced Cac (solid line) and Can (dashed line) transients in SED (B) and EX (C) pigs.

When peak endothelin-1-induced Ca_c and Ca_n responses were compared between SED and EX swine, exercise training significantly decreased the Ca_c response by 16% and the Ca_n response by 33% (Fig. 4A). Furthermore, exercise training significantly decreased the relative Ca_n t_1/2 to minimum response from peak by ~14 s and the relative duration of the response by ~31 s compared with SED animals (Table 2). Notice that the typical Ca_c and Ca_n responses in a cell from a SED animal (Fig. 4B) are nearly superimposable, whereas there are distinct differences in the Ca_c and Ca_n responses in a typical cell from an EX animal (Fig. 4C). In both SED and EX animals, the increase in Ca_n neither preceded nor followed the changes in Ca_c but occurred along the same time course. Our findings from these experiments demonstrate that coronary smooth muscle cells from EX pigs have decreased endothelin-1-induced Ca_c and Ca_n responses compared with SED pigs.

Caffeine-Induced Changes in Regional Calcium

View larger version (42K): [in this window] [in a new window]   Fig. 5.   Exercise training has minor effects on Can responses to caffeine. A: time lapse series of images (33 ms between images) in response to 5 mM caffeine in a coronary smooth muscle cell from an EX pig. Notice the delay in the Can response relative to the Cac response. B: percent cells in SED and EX pigs showing the Can delay. C: average delay (in ms) of the Can response relative to the Cac response.

Localization of SERCA and Coronary Smooth Muscle Ultrastructure

In "living" smooth muscle cells, identification of SERCA with B-TSG revealed a punctate distribution near the sarcolemma (Fig. 6A, arrows) and a weblike distribution in the central region of the cell (Fig. 6A, arrowheads). The subsarcolemmal distribution of B-TSG appeared to colocalize or lie in proximity to the sarcolemma, as indicated by the distribution of Di-8-ANEPPS and the yellow fluorescence where B-TSG and Di-8-ANEPPS overlap. The weblike distribution of B-TSG is consistent with the transnuclear distribution of the Ca²⁺ indicator fluo 3 (Fig. 6A, inset, arrows) and the transnuclear distribution of the SR probe 3',3'-dihexylocarbocyanine iodide (DiOC₆) as previously described by our lab (20, 47) and others (15). To negate the possibility of nonspecific binding of the BODIPY analog, cells were exposed to unconjugated BODIPY C₃, and no fluorescent distribution of the dye was observed (results not shown).

View larger version (176K): [in this window] [in a new window]   Fig. 6.   Localization of sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA) and coronary smooth muscle ultrastructure. A: sarcolemmal staining with Di-8-ANEPPS (red) and SERCA localization with BODIPY FL-thapsigargin (green) in living coronary smooth muscle cells. A, inset: DNA staining with SYTO-64 (red) and Ca2+ detection with fluo 3 (green) in the nuclear region of a cell. B-E: electronmicrographs of potassium ferricyanide-stained ultrathin sections of the pig right coronary artery. The images are of the same region with increasing magnification from B to E. F: immunofluorescence detection of SERCA2b (green) and DNA staining with DAPI (blue) in a single coronary smooth muscle cell. F, inset: immunofluorescence detection of the nuclear transport factor p97 (green) and DNA staining with DAPI (blue) in the nuclear region of a single coronary smooth muscle cell. G-I: immunogold detection of SERCA2b by electron microscopy. The small black box in G represents the sarcolemma and is presented at higher magnification in H. The large black box in G represents a nuclear cleft or invagination and is presented at higher magnification in I. In A and F, arrowheads denote the transnuclear region and arrows denote the sarcolemma or region near the sarcolemma. In B-E and G-I, the following are indicated: S, sarcolemma; N1 and N2, nucleus 1 and 2; M, mitochondria; Co, collagen; NE, nuclear envelope; INM, inner nuclear membrane; and ONM, outer nuclear membrane. Brackets denote the perinuclear space, arrowheads denote nuclear clefts and invaginations, and solid triangles denote immunogold labeling for SERCA2b. All images were generated from experiments in domestic swine right coronary artery smooth muscle cells.

We resolved the subcellular structures that may represent the transnuclear pattern of B-TSG using transmission electron microscopy of intact coronary artery segments. In the distal region of the right porcine coronary artery, there are two distinct layers of smooth muscle cells in the tunica media. The region of the tunica media nearest the lumen contains smooth muscle cells that run longitudinally in the vessel, and the region nearest the adventitia contains cells that run circumferentially, or around the vessel. The images depicted in Fig. 6, B-E, represent a cross section of the vessel and two distinct smooth muscle cells that run longitudinally in the vessel (the images are of the same region with increasing magnification from B to E).

The sacrolemma of each cell is not smooth or flat in nature but contains many infolds. These infolds are also depicted by the distribution of Di-8-ANEPPS that appears to lie below the edge of the cell, regions in the proximity of B-TSG distribution (Fig. 6A). The nuclei (labeled N1 and N2) of both cells are fusiform in shape, containing many clefts and invaginations (Fig. 6B, arrowheads). Much of the typical nucleus is taken up by heterochromatic regions, the opaque regions appearing near the nuclear rim and scattered throughout the nucleoplasm. Surrounding the nucleus and extending from its poles are mitochondria. The nuclear envelope, which is a double-unit membrane composed of the inner nuclear membrane (INM) bilayer and the outer nuclear membrane (ONM) bilayer, can be traced to continuities with the SR. Higher magnification of the nucleus labeled N2 resolved several mitochondria near the site of the nuclear invagination and membranes representing the nuclear envelope (Fig. 6, C and D). Figure 6E identified these membranes as the INM and ONM of the nuclear envelope; the brackets in Fig. 6E identify the perinuclear space or the space between the INM and ONM, ~30-50 nm. Thus it appears that the weblike or transnuclear distribution of B-TSG is localized to the nuclear envelope and fluo 3 is retained in the perinuclear space (Fig. 6A, inset).

Immunolocalization experiments were performed to further identify the cellular distribution of SERCA. As depicted by the Western blot in Fig. 7, the SERCA2b antibody resolves a band of the expected molecular mass, ~115 kDa. The peripheral (Fig. 6F, arrows) and nuclear weblike (Fig. 6F, arrowheads) distribution of SERCA2b (Fig. 6F) was very similar to the localization pattern observed with B-TSG (Fig. 6A). Identification of the nucleus with DAPI (blue) confirmed that the central, weblike distribution of SERCA2b is transnuclear. The perinuclear and transnuclear distribution of SERCA2b is very similar to the distribution the nuclear transport factor p97, a component of the nuclear pore complex, as shown by localization around the nucleus and transnuclear distribution (Fig. 6F, inset, arrowheads).

View larger version (30K): [in this window] [in a new window]   Fig. 7.   Western blot analysis of the SERCA2b antibody used to assess SERCA localization in coronary smooth muscle. The SERCA2b isoform was detected at the predicted molecular mass (MM; ~115 kDa) in canine cardiac muscle protein extracts [lane 1: positive control (PC)] and domestic pig right coronary artery (RCA) smooth muscle protein extracts (lanes 2 and 3: 5 and 20 µg of protein extract). The arrow denotes the molecular mass at 116 kDa.

Immunogold labeling of SERCA2b was performed to determine to which subcellular organelles SERCA was localized, e.g., the nuclear envelope. Tissue preparation for immunogold labeling enhances antigen detection but decreases the ability to resolve cell ultrastructure. Figure 6H depicts the subsarcolemmal distribution of SERCA (Fig. 6G, region of the sarcolemma outlined by a black box). SERCA distribution is in close proximity to the sarcolemma (~20-200 nm) with minimal distribution at distances >200 nm from the sarcolemma (Fig. 6, H and I, solid triangles). The nucleus shown in Fig. 6G is multilobular and contains many invaginations (arrowheads). The nuclear invagination outlined by a black box in Fig. 6G is depicted at higher magnification in Fig. 6I. Although the nuclear envelope is not as distinct as that shown in Fig. 6E, the immunogold labeling of SERCA is very near or on the ONM and INM. Our findings from these experiments demonstrated in living and fixed specimens that SERCA is localized near the sarcolemma and membranes in close proximity to the nucleus.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This study demonstrated that 1) phenotypic modulation of porcine coronary smooth muscle cells to endothelin-1 is attenuated after exercise training and 2) endothelin-1-induced Ca_c and Ca_n responses are attenuated after exercise training, unprecedented effects of exercise training in coronary smooth muscle. Furthermore, we also showed for the first time that noncultured/native coronary smooth muscle cells display an intricate nuclear design that is multilobular in shape and favorable for tight regulation of Ca_n and that SERCA is localized very near the sarcolemma and along the rim of the nucleus to regulate Ca²⁺.

Endothelin-1 is a potent vasoconstrictor and mitogen of vascular smooth muscle cells that has been implicated in the pathogenesis of coronary artery disease, particularly smooth muscle proliferation involved in neointimal thickening (29). Exercise training attenuated endothelin-1-induced smooth muscle phenotypic modulation compared with SED animals, as indicated by a 42% decrease in the relative DNA content. The organ culture model utilized to assess smooth muscle cell phenotypic modulation based on increased DNA content (and other factors) (19, 20) does not necessarily imply that these cells are undergoing proliferation but rather are phenotypically modulated compared with noncultured cells. Importantly, though, other investigators have shown that an increase in DNA content as measured by relative DAPI fluorescence positively correlates with cellular proliferation (26, 30). Furthermore, we show that vessels organ cultured for 2 days incorporate BrdU, an index of cellular proliferation. Thus we speculate that one mechanism for exercise-induced cardioprotection is decreased proliferative responses to potent agonists implicated in atherosclerosis.

The concept that subcellular Ca²⁺ events can regulate gene transcription is not new (2, 4) and has recently been addressed in cerebral smooth muscle (11). Cartin et al. (11) showed that membrane depolarization increases intracellular Ca²⁺ leading to increased nuclear expression of the transcription factors CREB (P-CREB) and levels of c-fos mRNA; evidence for Ca_c events in regulating gene transcription in smooth muscle. However, the outstanding issue regarding Ca_n in smooth muscle is whether levels of the ion can be regulated independent of the cytosol (23).

In coronary smooth muscle cells from SED pigs, we observed no difference in the time course and peak of the Ca_c and Ca_n responses to endothelin-1. These results suggest that in SED cells, Ca_n buffering is similar to the buffering of Ca²⁺ in the cytosolic compartment. Naik et al. (32) showed similar results in various strains of the rat aorta whereby the increase in Ca_n neither preceded nor followed the Ca_c response to endothelin-1 but occurred along the same time course. However, in various smooth muscle types, Ca_n responses were typically increased relative to Ca_c responses in response to endothelin-1 (32), histamine, ATP, carbachol (22), and bradykinin (5). Although it can be argued that the increased Ca_n response relative to the Ca_c response could be an artifact created by an uneven nucleocytoplasmic distribution of the Ca²⁺ indicator (23, 36), the differential rise in Ca_n over Ca_c in response to histamine, ATP, or carbachol did not occur with caffeine exposure (22). Also, the Ca²⁺ response to arginine vasopressin has been reported to be greater in the cytosol compared with the nucleus (21), further evidence that the changes in Ca_n do not always follow changes in Ca_c.

Unlike SED smooth muscle cells, exercise training may have upregulated SERCA in the nuclear envelope to more strongly buffer rises in Ca_n in response to endothelin-1. This was observed by both a decrease in peak Ca_n response and a decrease in t_1/2 to minimum response compared with the Ca_c response. These data suggest that exercise training enhances or initiates a buffering mechanism to maintain a nucleocytoplasmic Ca²⁺ gradient in coronary smooth muscle. We observed a noticeable delay in the Ca_n response relative to the Ca_c response to caffeine, evidence that a selective barrier can be initiated for regulating Ca_n. Although not significantly different, there was a propensity for cells from EX arteries to display this delayed Ca_n response to caffeine. We propose that under more physiological conditions, i.e., caffeine doses <5 mM, submaximal release of Ca²⁺ from the SR presented to the perinuclear region would unmask this delay in a greater percentage of cells. Thus this would require closeness of the SR to the nucleus to morphologically support these observations.

In fact, there is increasing appreciation that many cell functions related to Ca²⁺ are regulated within microenvironments and small regions of the cytoplasm (15, 31). For example, Moore et al. (31) reported that the Na⁺/Ca²⁺ exchanger is distributed on unique regions of the plasma membrane in register with, and in close proximity to, calsequestrin-containing regions of the SR, in sites distinct from where contractile filaments come in contact with the membrane. This specific spatial localization of the Na⁺/Ca²⁺ exchanger in smooth muscle allows for the functional regulation of Ca²⁺ in the very small space (<100 nm) (13) between the superficial SR and the sarcolemma, the superficial buffer barrier (44). Therefore, to further understand the functional adaptations of Ca_n regulation associated with exercise training, an examination of coronary smooth muscle cell ultrastructure, particularly the nucleus, and spatial localization of SERCA was required.

In this study, we showed that coronary smooth muscle nuclei from domestic swine are mulitlobular in shape, containing many clefts and invaginations of its membrane, the nuclear envelope. These findings are very similar to other mammalian coronary smooth muscle cells as described extensively in a review by Forbes (13); however, the function of this intricate design was not addressed. Recently, our lab and others have described a unique distribution of the SR fluorescent probe DiOC₆ in living coronary smooth muscle (20, 47) and portal vein smooth muscle (15). DiOC₆ staining revealed that the SR lies in very close proximity to the sarcolemma, as previously described by electron microscopy (12), and labels the SR near the periphery of the nucleus and the nuclear envelope, as depicted by a weblike nuclear distribution. These observations, coupled with the functional data presented here, suggest that Ca²⁺ signals generated by release from the SR are either 1) rapidly buffered by the SR near the nucleus, 2) buffered by Ca_c binding proteins, 3) buffered by mitochondria near the nucleus, or 4) prevented from entering the nucleus by regulatory mechanisms operating at the nuclear envelope, i.e., SERCA and/or the nuclear pore complex.

To address this, we determined the subcellular localization of the SERCA2b isoform in living and fixed coronary smooth muscle cells. We demonstrated for the first time a distribution of SERCA that is arranged very similar to the SR network (15, 20) and multilobular shape of the nucleus (13). We were able to demonstrate this distribution in living smooth muscle cells using two fluorochromes, B-TSG and Di-8-ANEPPS, which stain SERCA and the sarcolemma, respectively. The spatial localization of SERCA2b near the sarcolemma and along the periphery of the nucleus was further confirmed by immunocytochemistry. With the use of immunofluorescence microscopy, the subsarcolemmal distribution of SERCA2b was very similar to the distribution of B-TSG in relation to Di-8-ANEPPS. A predominant perinuclear and transnuclear distribution SERCA2b was also observed, a distribution similar to B-TSG as well as the nuclear transport factor p97, a component of the nuclear pore complex. Immunogold labeling of SERCA2b revealed a distribution in close proximity to the sarcolemma (~20-200 nm) with minimal distribution at distances >200 nm from the sarcolemma and localization to the nuclear envelope. Because of the loss of intracellular membrane resolution associated with the fixation procedure before immunogold labeling, we cannot definitively confirm that SERCA lies on the INM or ONM. However, given the width of the perinuclear space, the distance between the ONM and INM (~30-50 nm) and the distribution pattern of SERCA on the nuclear envelope, it is possible that SERCA is localized to both the INM and ONM in coronary smooth muscle.

Other functional studies in smooth muscle also suggest that SERCA is localized near or on the nuclear envelope. The selective barrier maintaining the nucleocytoplasmic Ca²⁺ gradient in cultured smooth muscle cells at rest and during stimulation was blocked by the specific SERCA inhibitor thapsigargin (22) and maintained in Ca²⁺-free solutions (37), evidence for regulation of Ca_n independent of Ca_c regulation. In a biochemical study in which it was possible to separate the INM and ONM from rat liver nuclei, it was shown that SERCA was exclusively found on the ONM (24).

Figure 8 illustrates a model that provides a rational basis to explain the results obtained in this study. Smooth muscle cell Ca²⁺ regulation after exercise training is denoted by solid lines, and dashed lines represent Ca²⁺ regulation in SED pigs. Previously, we have shown that exercise training increases coronary smooth muscle Ca²⁺ regulation within the confined functional region between the superficial SR and the sarcolemma (Fig. 8). We hypothesized that Ca²⁺ entering the cell through L-type voltage-gated Ca²⁺ channels is sequestered by SERCA on the superficial SR and slowly released from the SR toward sarcolemmal Ca²⁺-ATPase and the Na⁺/Ca²⁺-exchanger for extrusion (8, 17, 41, 42). This adaptation to exercise has been described as a mechanism to attenuate vasoconstriction to potent Ca²⁺-mobilizing agonists, i.e., endothelin-1 (8) and norepinephrine (33), by depleting the SR Ca²⁺ store. Thus SERCA localization <200 nm from the sarcolemma is consistent with Ca²⁺ regulation in the superficial buffer barrier (Fig. 8) (44).

View larger version (30K): [in this window] [in a new window]   Fig. 8.   Hypothetical model of Can regulation and exercise-induced cardioprotection. Exercise training results in increased subsarcolemmal Ca2+ (Cas) cycling, decreased Cac responses to ET-1, and attenuated vasoconstriction (10, 43). Deeper within the cell, exercise training increases Can buffering, which may in part attenuate ET-1-induced smooth muscle proliferation. The spatial localization of SERCA (S) on the superficial sarcoplasmic reticulum (SR), near the nucleus, and on the nuclear envelope serves in regulating Cas and Can. The inositol 1,4,5-trisphosphate (IP3) resulting from ET stimulation rapidly diffuses through the cytosol to activate IP3 receptors localized on the nuclear envelope and induce Ca2+ release. Without strong buffering of Ca2+ by SERCA, large increases in Can would occur. We propose that enhanced Can buffering associated with exercise training may be one mechanism by which coronary smooth muscle cells are less responsive to mitogen-induced growth and proliferation. Solid lines, smooth muscle cell Ca2+ regulation after exercise training; dashed lines, Ca2+ regulation in SED pigs; CP, Ca2+-ATPase pump; NX, Na+/Ca2+ exchanger; CR, calcium release channel; NPC, nuclear pore complex; INM, inner nuclear membrane; ONM, outer nuclear membrane; SBB, superficial buffer barrier; NBB, nuclear buffer barrier; Ca:Cam, Ca2+-calmodulin complex; MLCK, myosin light chain kinase; G, G protein; PL, phospholipase.

We propose here that decreased endothelin-1-induced DNA content may be associated with enhanced Ca_n regulation and that SERCA may be one site for enhanced regulation after exercise training. Therefore, Ca²⁺ that passively diffuses from the cytosol into the nucleus via the nuclear pore complex is actively sequestered by SERCA on 1) the SR deep within the cell and superficial to the nucleus, and/or 2) SERCA on the nuclear envelope (Fig. 8). This spatial localization of SERCA allows for shielding of the nucleoplasm from changes in Ca_c, a mechanism that, when functioning, could attenuate Ca²⁺-activated gene transcription in the nucleus. Possibly more important than acting as a buffer for Ca²⁺ passively diffusing through the cytosol is the role of SERCA on the nuclear envelope in buffering localized increases in Ca_n resulting from Ca²⁺ release from the nuclear envelope itself. Indeed, inositol 1,4,5-trisphosphate receptors on the nuclear envelope (Fig. 8, Ca²⁺ release channels and arrows on the nuclear envelope) could be activated with minimal delay compared with inositol 1,4,5-trisphosphate receptors on the superficial SR because of the very rapid diffusion of inositol 1,4,5-trisphosphate in the cytosol (1). Thus enhancing Ca_n buffering may be one mechanism by which coronary smooth muscle cells are less responsive to mitogen-induced phenotypic modulation. We have defined this region of Ca_n regulation as the nuclear buffer barrier (Fig. 8).

Because atherosclerosis involves the phenotypic modulation of vascular smooth muscle cells, our data provide a mechanistic basis for future studies on the cardioprotective effects of exercise training and atherosclerosis. In fact, we have shown that compared with healthy animals, coronary smooth muscle from swine with atherosclerosis have increased Ca_n responses to endothelin-1, which were positively correlated with the extent of coronary atherosclerosis (47). The dysregulation of Ca_n was also directly associated with the structural breakdown of the nuclear buffer barrier, including retraction of transnuclear Ca²⁺ stores (47). Thus if exercise training has similar effects on Ca_n regulation in smooth muscle from arteries with coronary artery disease, such adaptations might have the potential to attenuate Ca²⁺-mediated responses to local increases in potent mitogens associated with atherosclerosis. In conclusion, this is the first report of altered Ca_n regulation and an intrinsic resistance to coronary smooth muscle phenotypic modulation produced by exercise training.