Estrogen increases Ca 2 efflux from female porcine coronary arterial smooth muscle

Transcription

1 Estrogen increases Ca 2 efflux from female porcine coronary arterial smooth muscle Y. S. PRAKASH, A. A. TOGAIBAYEVA, M. S. KANNAN, V. M. MILLER, L. A. FITZPATRICK, AND G. C. SIECK Departments of Anesthesiology, Physiology and Biophysics, Surgery, and Endocrinology, Mayo Foundation, Rochester 55905; and Departments of Veterinary PathoBiology and Pediatrics, University of Minnesota, St. Paul, Minnesota Prakash, Y. S., A. A. Togaibayeva, M. S. Kannan, V. M. Miller, L. A. Fitzpatrick, and G. C. Sieck. Estrogen increases Ca 2 efflux from female porcine coronary arterial smooth muscle. Am. J. Physiol. 276 (Heart Circ. Physiol. 45): H926 H934, Acute estrogen administration relaxes vascular smooth muscle by decreasing intracellular Ca 2 concentration ([Ca 2 ] i ). In the present study, we examined the hypothesis that this reduction in [Ca 2 ] i is mediated in part by enhanced Ca 2 efflux. Coronary artery smooth muscle cells were isolated from gonad-intact, sexually mature female pigs. The [Ca 2 ] i response to endothelin-1 was measured using fluo 3 and confocal microscopy. 17 -Estradiol (E 2 ), but not 17 estradiol or triamcinolone acetonide, caused a concentrationdependent (IC nm) decrease in the [Ca 2 ] i response to endothelin-1. This decrease was blocked by the specific estrogen receptor antagonist ICI Under conditions in which Ca 2 influx and sarcoplasmic reticulum Ca 2 reuptake were blocked, E 2 still decreased [Ca 2 ] i. The response was blocked by extracellular lanthanum. These data indicate that E 2 decreases [Ca 2 ] i in coronary artery smooth muscle by affecting Ca 2 efflux via a receptor-mediated mechanism. vasodilation; endothelin; receptor; calcium adenosine 5 triphosphatase ACUTE ADMINISTRATION of estrogens relaxes vascular smooth muscle, causing a reduction in vascular resistance and an increase in blood flow (3, 5, 8, 9, 16, 19, 21, 25, 30). The rapid relaxation of vasomotor tone induced by estrogens is likely mediated through nongenomic effects (for review, see Ref. 2) and probably involves estrogen receptors (2, 11, 20). However, despite considerable evidence for acute, nongenomic effects of estrogen on vascular tissue, the mechanisms underlying these effects are not completely understood. Estrogen-induced relaxation of vascular smooth muscle involves a reduction in intracellular Ca 2 concentration ([Ca 2 ] i ) (15, 19). One demonstrated mechanism for this effect is a decrease in Ca 2 influx (8, 9, 14). However, a reduction in [Ca 2 ] i induced by estrogens may also result from an increase in Ca 2 efflux, an inhibition of Ca 2 release from the sarcoplasmic reticulum (SR), or an increase in SR Ca 2 reuptake. Therefore, the purpose of the present study was to examine whether estrogens [17 -estradiol (E 2 )] affect Ca 2 The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. efflux and SR Ca 2 release in coronary artery smooth muscle. METHODS Animals. Ten sexually mature female Yorkshire pigs (at least 6 mo of age) were obtained from a local supplier. Animals were anesthetized with ketamine (8 mg/kg) and xylazine (12 mg/kg), and their hearts were removed. The three main coronary arteries (right and left circumflex and left anterior descending arteries) were removed and placed in oxygenated modified Krebs-Ringer bicarbonate solution at 4 C. Blood samples were also drawn to verify hormonal status using plasma concentrations of estrogen and progesterone. All animal surgical and care procedures were approved by the Institutional Animal Care and Use Committee of the Mayo Clinic and were in strict accordance with the American Physiological Society animal care guidelines. Immunocytochemical detection of estrogen receptors. Freshly dissociated coronary artery smooth muscle cells were fixed in 4% paraformaldehyde in 0.1 M phosphate buffer and stained for 2 3 h with a polyclonal antibody to the estrogen receptor (Research Biochemicals; 1:2,000 dilution in phosphate buffer). After staining with the primary antibody was completed, the sections were washed in phosphate buffer and reacted with a 1:1,000 Cy3-conjugated donkey anti-rabbit IgG secondary antibody. The fluorescently labeled coronary artery smooth muscle cells were visualized using a Bio-Rad MRC500 confocal microscope equipped with an argon-krypton laser. The 568-nm laser line was used to excite the Cy3, and the emissions were collected using a 590-nm long-pass filter. Cell preparation for Ca 2 imaging. The procedures for dissociation of single porcine coronary artery smooth muscle cells have been described previously (10). Briefly, the endothelium was removed and the tissue was minced thoroughly in Hanks balanced salt solution (HBSS) buffered with 10 mm HEPES (ph 7.4; Life Technologies, Gaithersburg, MD). The tissue was incubated first in 20 U/ml papain and 2,000 U/ml DNase (Worthington Biochemical, Freehold, NJ) and subsequently in 1 mg/ml type IV collagenase and 0.1 mg/ml elastase (Worthington). Single cells were released by trituration, centrifugation, and resuspension in minimum essential medium with 10% fetal calf serum. The cells were plated on glass coverslips coated with Cell-Tak (Collaborative Biomedical Products, Bedford, MA) and incubated for 1 2 h in a 5% CO 2 incubation chamber at 37 C. Cells were used within 4 h after dissociation. No time-dependent changes were observed in the gross morphology of the cells that were used for the experimental protocols. Exclusion of the dye trypan blue was used to assess cell viability just after dissociation ( 90% of all cells). An anti-smooth muscle myosin antibody (Sigma Immunochemicals, St. Louis, MO) was used to estimate the relative proportion of smooth muscle myocytes (immunoreactive) and fibroblasts (50:1). Each coverslip was washed with HBSS and incubated for min at 37 C in 5 µm fluo 3-AM (Molecular Probes, H /99 $5.00 Copyright 1999 the American Physiological Society

2 H927 Eugene, OR). The coverslip was then washed briefly in HBSS and mounted on an open slide chamber (RC-25F, Warner Instruments, Hamden, CT) mounted on the stage of a Nikon Diaphot inverted microscope. The chamber was perfused at 2 3 ml/min at room temperature. Real-time Ca 2 imaging. An Odyssey XL real-time confocal system (Noran Instruments, Middleton, WI) attached to the Nikon Diaphot and equipped with an argon-krypton laser was used to visualize fluo 3-loaded coronary artery smooth muscle cells (18). An Olympus 40/1.3 oil-immersion objective lens was used to visualize the cells. Image size was set to pixels, and pixel area was calibrated using a stage micrometer (0.063 µm 2 /pixel). A fixed combination of laser intensity (20% of maximum) and photomultiplier gain (1,700 from a maximum of 4,096) was set a priori to ensure that pixel intensities within regions of interest ranged between 25 and 255 gray levels across different experimental protocols. Dye bleaching was kept to a minimum by maintaining laser intensity below 3 mw. Continued exposure to the laser did not exceed 3 min. With the use of these precautions, dye bleaching was estimated to be 5%overa3minperiod. The Odyssey confocal system is controlled by a Silicon Graphics workstation and is capable of acquiring 480 frames/s. In preliminary studies on fluo 3-loaded coronary artery smooth muscle cells, it was determined that an acquisition rate of 30 frames/s was sufficient to measure the changes in [Ca 2 ] i during various protocols without frequency aliasing. Therefore, image acquisition was limited to 30 frames/s. When necessary, image noise was reduced by acquiring at 60 or 120 frames/s with frame averaging. The sampling time for any pixel was 100 ns. A region of the coverslip containing at least coronary artery smooth muscle cells was selected, and a region-ofinterest software tool was used to define regions within the boundaries of individual cells. Each region of interest had a fixed dimension of pixels (6 µm 2 ). The optical section thickness for the 40 lens was set to 1 µm by controlling the confocal slit size. Therefore, Ca 2 measurements were obtained from a volume of 6 µm 3. The fluorescence intensity of fluo 3 was calibrated for Ca 2 concentrations as described previously (18). At fixed settings of laser intensity and photomultiplier gain, fluo 3-loaded coronary artery smooth muscle cells were exposed to A (Ca 2 ionophore) at varying levels of extracellular Ca 2 ranging from 0 (HBSS with EGTA) to 10 µm. Exposure to the ionophore at each extracellular Ca 2 concentration was limited to 1 min to ensure that other Ca 2 handling mechanisms such as mitochondria and SR did not compensate for the ionophore-induced Ca 2 influx and thus confound the calibration. Furthermore, fluo 3 is also capable of being incorporated into the mitochondria and nucleus with continued exposure. Accordingly, to ensure that the fluo 3 signal represented cytosolic Ca 2 only, the total duration of the calibration protocol was minimized. Based on a calibration curve constructed from the gray level values of fluorescence intensities at different Ca 2 concentrations, the average gray level within a region of interest was converted to nanomoles per liter of Ca 2. Assessment of cell viability. After each experiment, coronary artery smooth muscle cells on at least one coverslip per animal ( 80 cells) were evaluated for the exclusion of trypan blue. These cells were exposed to the confocal laser for varying periods of time ranging from 3 min to repeated exposures across a 30-min period. The exclusion of trypan blue in these cells confirmed that laser exposure did not injure the cells. In a subset of these cells (20 cells), the effect of prolonged laser exposure (three 3-min exposure periods separated by 5 min) on baseline [Ca 2 ] i level was evaluated. Baseline [Ca 2 ] i was found to vary 5% across this entire period. In another subset of cells (15 cells), the reproducibility of the [Ca 2 ] i response to 5 mm caffeine was evaluated over a 20-min period. The average coefficient of variation for the [Ca 2 ] i response to caffeine in these cells was 6.5%. [Ca 2 ] i response to endothelin-1. The [Ca 2 ] i response of coronary artery smooth muscle cells was measured to increasing concentrations of endothelin-1 (0.1, 1, 10, and 100 nm and 1 µm). Because tachyphylaxis is a well-known effect with endothelin-1, multiple-agonist exposures for the same cell were not possible. Instead, sets of coronary artery smooth muscle cells were exposed to only one endothelin-1 concentration, and the average response was determined. Effect of estrogens on endothelin-1-induced elevation of [Ca 2 ] i. Cells were exposed to 100 nm endothelin-1, and an elevation in [Ca 2 ] i was confirmed. When [Ca 2 ] i had reached a steady-state level (2 3 min), the cells were exposed to solvent (ethanol 10 nm), E 2 (0.1 nm to 10 µm), 17 -estradiol (E 2 ; 10 nm, the biologically inactive form of estradiol), or triamcinolone acetonide (10 nm; a synthetic steroid unrelated to estrogens). Role of estrogen receptors in [Ca 2 ] i response to estrogens. To determine whether estrogen receptors are necessary for mediation of the effect of E 2 on [Ca 2 ] i, coronary artery smooth muscle cells were exposed for 3 min to the estrogenreceptor antagonist ICI (1 µm; generously provided by Zeneca Pharmaceuticals, Cheshire, UK). Subsequently, the cells were exposed to endothelin-1, and when steady-state [Ca 2 ] i was reached, the cells were exposed to either E 2 or E 2 (10 nm). Effect of E 2 on Ca 2 efflux. To block Ca 2 influx, cells were exposed to zero extracellular Ca 2 or nifedipine (100 nm). The cells were then exposed to thapsigargin (1 mm) to block SR Ca 2 reuptake (23), which resulted in a gradual increase in [Ca 2 ] i, plateauing after 2 min. The cells were then exposed to E 2 in the presence or absence of endothelin-1. To determine whether estrogen receptors mediate the effect of E 2 on Ca 2 efflux under these conditions, cells were preexposed to ICI (1 µm) before endothelin-1 exposure. To confirm that Ca 2 efflux was actually involved, experiments were repeated in the presence of lanthanum (1 mm), which nonselectively blocks both Ca 2 influx and efflux (28). Effect of E 2 on SR Ca 2 release and reuptake. The response to caffeine (5 mm) was evaluated in the presence or absence of preexposure to E 2 in cells in which Ca 2 influx and SR Ca 2 reuptake were blocked by nifedipine (or zero extracellular Ca 2 ) and thapsigargin, respectively. Under these conditions, SR Ca 2 release was estimated from the peak [Ca 2 ] i response. To examine SR Ca 2 reuptake, cells were first exposed to nifedipine and lanthanum to block both Ca 2 influx and efflux. The [Ca 2 ] i response to endothelin-1 was then evaluated in the presence or absence of E 2.SRCa 2 reuptake was estimated from the initial rate of decline in [Ca 2 ] i. Effect of E 2 on Ca 2 influx. Cells were exposed to BAY K 8644 (100 nm, 1 and 10 µm) to induce Ca 2 influx. Nifedipine (100 nm) completely blocked the elevation of [Ca 2 ] i induced by 100 nm and 1 µm BAY K 8644, but it only partially blocked the [Ca 2 ] i response to 10 µm BAY K Accordingly, Ca 2 influx was induced by 1 µm BAY K 8644 in the presence of E 2. To explore whether the effects of estrogens on Ca 2 influx are mediated via Ca 2 -activated K channels (BK Ca ) (5, 13, 16, 29), cells were exposed to iberiotoxin (100 nm), a specific inhibitor of these channels, and then to endothelin-1. When

3 H928 [Ca 2 ] i had reached a steady state, the cells were exposed to E 2. Statistical analysis. In determining the statistical design for various experiments, the influence of interanimal variability in the [Ca 2 ] i response to 100 nm endothelin-1 was evaluated for 15 coronary artery smooth muscle cells from each of 10 animals using a one-way ANOVA. Interanimal variability was found to be nonsignificant at P Based on this result, the contribution of interanimal variability was determined to be insignificant, and the pooling of results from coronary artery smooth muscle cells obtained from different animals was justified. However, for each of the experimental protocols assessing the impact of estrogen on the [Ca 2 ] i response to 100 nm endothelin-1, cells from at least five animals were studied, and cells from any one animal did not represent 30% of the total. In addition, individual coronary artery smooth muscle cells were exposed to only one experimental protocol. The numbers of cells used for each protocol are reported in the results, and statistical analysis (one-way ANOVA) was based on this number. Statistical significance was tested at an 0.05 level. Data are reported as means SE. Reductions in [Ca 2 ] i are reported as percent changes from the maximum [Ca 2 ] i response. RESULTS Estrogen levels and estrogen-receptor status. Serum estrogen ranged from 37 to 103 pm and averaged pm in all 10 animals. Immunocytochemical staining confirmed the presence of estrogen receptors in coronary artery smooth muscle cells (Fig. 1). Although the immunocytochemical procedure could not distinguish between nuclear and cytosolic estrogen receptors, the higher intensity of fluorescence staining in the center of the cell suggested the presence of nuclear receptors. However, immunoreactivity for cytosolic estrogen receptors distributed throughout the cell was also clearly present. [Ca 2 ] i response to endothelin-1. Basal [Ca 2 ] i levels were not significantly different among cells and ranged from 120 to 160 nm (131 3 nm; n 426). At each concentration of endothelin-1, the [Ca 2 ] i response of 16 cells was determined. Endothelin-1 at 1 and 10 nm caused a slow and sustained increase in [Ca 2 ] i. In comparison, 100 nm and 1 µm endothelin-1 caused a more rapid increase in [Ca 2 ] i, reaching an initial peak after 4 30 s but then decreasing slowly over the next 2 3 min to 70% of the peak value. The [Ca 2 ] i responses to varying endothelin-1 concentrations at 30 s and 2 min are reported in Table 1. Effect of estrogens on endothelin-1-induced elevation of [Ca 2 ] i. The effect of E 2 on the [Ca 2 ] i response to endothelin-1 was concentration dependent with an ED 50 of 10 nm (Fig. 2A). The effect of 10 nm E 2 on the endothelin-1 dose-response (1, 10, and 100 nm) was determined (Fig. 2B). Although the [Ca 2 ] i response varied with endothelin-1 concentration, the inhibitory effect of E 2 was proportionately similar across all endothelin-1 concentrations. Because the relative effect of 10 nm E 2 was comparable across endothelin-1 concentrations, fixed concentrations of E 2 (10 nm) and endothelin-1 (100 nm) were used in all subsequent protocols. Fig. 1. Immunocytochemical staining for estrogen receptors in porcine coronary artery smooth muscle cells. Dissociated cells were fixed in paraformaldehyde, stained with an antibody to estrogen receptors, and visualized using a Cy3-conjugated fluorescent secondary antibody. Nonspecific labeling of secondary antibody was determined by omitting primary antibody (A). Immunoreactivity confirmed presence of estrogen receptors in coronary artery smooth muscle cells. Note higher fluorescence intensity around nucleus (B). Exposure to 10 nm E 2 reduced the elevation of [Ca 2 ] i induced by 100 nm endothelin-1 to near basal levels within a 15- to 45-s period (Fig. 3; n 56). The endothelin-1-induced elevation in [Ca 2 ] i was also reduced by 10 nm E 2 (Fig. 3; n 53), but to a lesser extent compared with E 2. In contrast, the [Ca 2 ] i response to endothelin-1 was unaffected by triamcinolone acetonide (Fig. 3; n 14) or ethanol (n 15). Table 1. [Ca 2 ] i response of coronary artery smooth muscle cells to endothelin-1 Endothelin-1 Concentration [Ca 2 ] i at 30 s, nm [Ca 2 ] i at2min, nm 1nM nm nm µm Values are means SE. Time points for measurement of intracellular Ca 2 concentration ([Ca 2 ] i ) are based on average times to peak and steady-state responses at 100 nm and 1 µm endothelin-1.

4 H929 Fig. 2. Cumulative dose-response curve of 17 -estradiol (E 2 )- induced reduction of endothelin-1-induced intracellular Ca 2 concentration ([Ca 2 ] i ) in coronary artery smooth muscle cells from female pigs. In presence of a fixed concentration of 100 nm endothelin-1, cells were exposed to progressively larger concentrations of E 2 (A; n 15). Effect of IC 50 E 2 (10 nm) on [Ca 2 ] i response to different endothelin-1 concentrations was determined in separate sets of cells (B; n 12 per endothelin-1 concentration). [Ca 2 ] i response is reported as percentage of maximum response at a given endothelin-1 concentration. Role of estrogen receptors in [Ca 2 ] i response to estrogens. In the presence of ICI , E 2 had no effect on the [Ca 2 ] i response to endothelin-1 (Fig. 4; n 51). In contrast, ICI did not abolish the effect of E 2 on the [Ca 2 ] i response to endothelin-1 (Fig. 4; n 32). Effect of E 2 on Ca 2 efflux. When Ca 2 influx was blocked by nifedipine (n 22) or zero extracellular Ca 2 (n 13; data not shown), inhibition of SR Ca 2 reuptake by thapsigargin caused a gradual increase in [Ca 2 ] i (Fig. 5). Subsequent exposure to E 2 reduced [Ca 2 ] i by % (Fig. 5; P 0.05 compared with vehicle control). Under conditions of blocked Ca 2 influx (nifedipine: n 32; zero extracellular Ca 2 : n 21; data not shown) and SR Ca 2 reuptake (thapsigargin), endothelin-1 induced a large increase in [Ca 2 ] i (Fig. 6). Subsequent exposure to E 2 reduced [Ca 2 ] i (Fig. 6). This reduction in [Ca 2 ] i was blocked by ICI (n 22) and lanthanum (n 31) (Fig. 6). Effect of E 2 on SR Ca 2 release and reuptake. Preexposure to E 2 had no effect on the amplitude of the initial response to 100 nm and 1 µm endothelin-1. When Ca 2 influx and SR Ca 2 reuptake were blocked, the peak [Ca 2 ] i response to caffeine was comparable in the presence (n 24) and absence (n 25) of E 2 (Fig. 7). These results suggest that E 2 had no effect on SR Ca 2 release. In the presence of nifedipine and lanthanum, where Ca 2 influx and efflux were blocked, the rate of decline of the [Ca 2 ] i response to endothelin-1 was comparable in the presence (n 15) and absence (n 14) of E 2 (Fig. 7). These results suggest that E 2 had no effect on SR Ca 2 reuptake. Effect of E 2 on Ca 2 influx. Exposure to E 2 significantly reduced the Ca 2 influx induced by 1 µm BAY K 8644 (Fig. 8). When BK Ca channels were blocked by iberiotoxin, E 2 still reduced the [Ca 2 ] i response to endothelin-1 (Fig. 8). However, this reduction in [Ca 2 ] i was significantly less than that in the absence of iberiotoxin ( vs %, respectively). DISCUSSION In addition to confirming the inhibitory effect of estrogens on Ca 2 influx in coronary artery smooth muscle cells, the results of the present study also demonstrated that estrogens decrease [Ca 2 ] i by enhancing Ca 2 efflux via a receptor-mediated mechanism. Fig. 3. Effect of estrogens on endothelin-1-induced elevation of [Ca 2 ] i in porcine coronary artery smooth muscle cells. Exposure to 100 nm endothelin-1 (A) induced a large elevation of [Ca 2 ] i, which was reduced by both E 2 (B; n 56) and 17 -estradiol (E 2 ; C; n 53), albeit to different extents. Compared with control, solvent control (ethanol; n 15) and an unrelated steroid [triamcinolone acetonide (TAA); n 14] did not decrease [Ca 2 ] i (D). *Significant difference (P 0.05) from control.

5 H930 independent of the level of endothelin-1 stimulation, as indicated by the comparable reduction in [Ca 2 ] i at different endothelin-1 concentrations, relative to the peak response at each concentration (e.g., 50% decrease from the peak response with 10 nm E 2 ). Confirmation of the effect of E 2 on Ca 2 efflux was derived primarily from the fact that the response was blocked by lanthanum. However, lanthanum also blocks Ca 2 influx, which has been previously demonstrated to be inhibited by E 2. Therefore, we designed studies to block Ca 2 influx independent of any inhibition on Ca 2 efflux. In experimental protocols designed to examine the underlying mechanisms of E 2 action, a 100 nm endothelin-1 concentration was selected based on studies by other investigators (24, 28), who found that the initial [Ca 2 ] i response at similar endothelin-1 concentrations represents SR Ca 2 release and that the subsequent response represents a balance between Ca 2 influx and efflux. In this regard, the initial peak [Ca 2 ] i response at 100 nm endothelin-1 observed in our study most likely reflects SR Ca 2 release, whereas the lower steady state reflects a balance between Ca 2 influx and efflux. In contrast, endothelin-1 concentrations of 10 nm or less elicit only slow, monotonic elevations in [Ca 2 ] i (also observed in the present study), which most likely reflect Ca 2 influx (24). Accordingly, the selection of 100 nm allowed the examination of both SR Ca 2 release and Ca 2 flux across the cell membrane. In contrast to these measurements, studies using multicellular coronary artery smooth muscle strips as well as single cells have reported a decline in [Ca 2 ] i to baseline after the initial peak [Ca 2 ] i response to endothelin-1. However, it must be noted that this decline in multicellular preparations occurred over a considerably longer Fig. 4. Involvement of estrogen receptors in estrogen-induced reduction of [Ca 2 ] i in porcine coronary artery smooth muscle cells. In presence of estrogen-receptor antagonist ICI (ICI), E 2 did not reduce endothelin-1-induced elevation of [Ca 2 ] i (A; n 51), suggesting that estrogen receptors mediate this effect of E 2. In contrast, even in presence of ICI , E 2 reduced endothelin-1- induced elevation of [Ca 2 ] i (B; n 32 cells), suggesting a nonestrogen receptor-mediated effect. C: summary of [Ca 2 ] i responses. *Significant difference (P 0.05) from response to E 2. The conclusion that E 2 enhances Ca 2 efflux was supported by three observations: 1) the reduction in [Ca 2 ] i by E 2 was blocked by the estrogen-receptor antagonist ICI ; 2)theE 2 -induced reduction in [Ca 2 ] i was observed even when Ca 2 influx and reuptake were blocked; and 3) thee 2 -induced reduction in [Ca 2 ] i was absent when both Ca 2 influx and efflux were blocked by nifedipine (and/or zero extracellular Ca 2 ) and lanthanum. The enhancement of Ca 2 efflux by E 2 represents a novel mechanism by which estrogens regulate [Ca 2 ] i in response to agonist stimulation. As would be expected, the relative enhancement of Ca 2 efflux was Fig. 5. Effect of E 2 on Ca 2 efflux at basal [Ca 2 ] i levels. In absence of agonist-induced sarcoplasmic reticulum (SR) Ca 2 release, when Ca 2 influx was blocked by nifedipine (or zero extracellular Ca 2 ; not shown) and SR Ca 2 reuptake was blocked by thapsigargin, basal [Ca 2 ] i levels slowly increased (A; n 22). This trend in [Ca 2 ] i was reversed by E 2 (B; n 22), suggesting enhancement of Ca 2 efflux.

6 H931 Fig. 6. Effect of E 2 on Ca 2 efflux after endothelin-1 (ET) activation. When Ca 2 influx and SR Ca 2 reuptake were blocked by nifedipine (Nif; or zero extracellular Ca 2 ; not shown) and thapsigargin (Tg), respectively, endothelin-1 induced a large, sustained elevation of [Ca 2 ] i (A; n 32). Subsequent exposure to E 2 reduced [Ca 2 ] i (B; n 32). This effect of E 2 on [Ca 2 ] i was inhibited by lanthanum (La; C; n 31), indicating that E 2 enhances Ca 2 efflux. D: summary of [Ca 2 ] i responses. *Significant difference (P 0.05) from response to endothelin-1 alone. time period ( 10 min) and most likely reflects the ensemble of individual cellular responses. The serum estrogen levels of 50 pm are consistent with previous reports in swine of pm (7, 26). However, because estrous stage was not controlled in the present study, the 50 pm value most likely reflects an average across different stages. Given the picomolar serum estrogen levels, the 10 nm concentration of E 2 used to examine [Ca 2 ] i regulation is supraphysiological. It is difficult to directly extrapolate the reduction in [Ca 2 ] i observed with 10 nm E 2 in single coronary artery smooth muscle cells to the extent of reduction in vascular tone in vivo because, if anything, the estrogen concentration at the tissue is likely to be smaller. Fig. 7. Effect of E 2 on SR Ca 2 release and reuptake. In absence of Ca 2 influx and SR reuptake, the large, transient [Ca 2 ] i response to caffeine (A; n 25) was not affected by preexposure to E 2 (B; n 24). In absence of Ca 2 influx and efflux (blocked by nifedipine and lanthanum), endothelin-1 produced a large, sustained elevation of [Ca 2 ] i (C; n 14). Rate of decline of [Ca 2 ] i response to endothelin-1 was not affected by preexposure to E 2 (D; n 15). These data suggest that SR Ca 2 release and reuptake are not substantially affected by estrogens.

7 H932 Fig. 8. Effect of E 2 on Ca 2 influx. Exposure of coronary artery smooth muscle cells to BAY K 8644 (BayK), an agonist of voltage-gated Ca 2 influx channels, resulted in a slow elevation of [Ca 2 ] i (A; n 14). Preexposure to E 2 inhibited BAY K 8644-induced elevation of [Ca 2 ] i (B; n 14). Exposure to E 2 following BAY K 8644 also decreased [Ca 2 ] i (tracing not shown). Preexposure to iberiotoxin (IBTX), a specific inhibitor of Ca 2 -activated K (BK Ca ) channels, only partially blocked the reduction in [Ca 2 ] i levels induced by E 2 under conditions of inhibited SR reuptake and endothelin-1 stimulation (C; n 16). In the statistical summary (D), control responses are [Ca 2 ] i elevations by BAY K 8644 alone, or decrease in [Ca 2 ] i by E 2 in the absence of IBTX. *Significant difference (P 0.05) from the control response. However, it must be noted that we observed decreases of 20 30% in [Ca 2 ] i even at E 2 concentrations between 100 pm and 1 nm. These observations would suggest that circulating estrogen levels do decrease [Ca 2 ] i to a somewhat smaller extent, which would be consistent with a role of estrogen as a modulator of vascular tone. We opted for a concentration of 10 nm E 2 because it approximates an IC 50 and thus provided a greater sensitivity for assessing the effects of various inhibitors and activators and distinguishing between the effects of Ca 2 efflux versus other mechanisms of estrogen action. It is difficult to compare the results of the present study, in which 10 nm E 2 was used, with those of previous reports because a wide range of estrogen concentrations have been used in these studies (4, 14, 16, 19, 21a, 29) that are often greater than circulating estrogen levels and considerably higher than that used in the present study. Regardless, it appears that the mechanisms by which estrogen modulates vascular tone are similar, and it is likely that Ca 2 efflux plays a role in estrogen-induced reduction of [Ca 2 ] i even in vivo. The specificity of the E 2 effect on [Ca 2 ] i regulation is supported by the fact that the estrogen-receptor antagonist ICI blocked the [Ca 2 ] i response to E 2. In addition, compared with E 2, the biologically inactive isomer E 2 at a similar concentration was significantly less potent in reducing agonist-induced elevation of [Ca 2 ] i, consistent with previous studies (25, 31). The lack of inhibition of the [Ca 2 ] i response to E 2 by ICI suggests a nonspecific non-receptormediated effect of E 2. However, the mechanisms underlying the effects of E 2 are not clear. Finally, triamcinolone acetonide, a steroid not related to estrogens, did not reduce the endothelin-1-induced elevation of [Ca 2 ] i in coronary artery smooth muscle cells. This also suggests that the nonspecific effect of E 2 is unlikely to be an experimental artifact. The relatively rapid changes in [Ca 2 ] i observed in the present study, occurring over seconds rather than minutes or hours, are clearly not compatible with the classic genomic mechanism for the action of E 2, which involves translocation of receptors to the nucleus and protein synthesis (1, 12). Therefore, the acute effects observed in the present study are nongenomic in nature and are also consistent with previous studies both in vivo (13, 17, 30) and in vitro (5, 8, 9, 19, 21a, 25) (also see Ref. 2 for a review). The existence of cytosolic estrogen receptors and the recent interest in the role of plasma membrane estrogen receptors are also indicative of nongenomic effects. It is possible that these plasma membrane receptors are also involved in [Ca 2 ] i regulation by estrogens and mediate some of the effects on [Ca 2 ] i observed in the present study. However, with the use of light microscopy, it was not possible to localize the estrogen receptors to the plasma membrane. In addition to an enhancing effect on Ca 2 efflux, the results of the present study also confirmed previous observations that E 2 inhibits Ca 2 influx (14, 19, 22, 29, 31). For example, E 2 inhibited the BAY K induced elevation of [Ca 2 ] i. These results of the present study extend previous reports by establishing that Ca 2 influx, at least through L-type Ca 2 channels, is blocked even by nanomolar concentrations of estrogen, considerably smaller than those used in previous studies. In porcine coronary artery smooth muscle from castrated males, estrogens were shown to activate BK Ca channels through a cgmp-dependent mechanism, thus indirectly reducing Ca 2 influx (29). In the present study, the observation that iberiotoxin (a potent inhibitor of BK Ca channels) decreased the extent to which E 2 reduced the endothelin-1-induced elevation of [Ca 2 ] i supports this mechanism. Also in general agreement with previous studies (14, 31) is the observation that E 2 had no effect on SR Ca 2 release. In the present study, E 2 did not affect the [Ca 2 ] i response to caffeine, an agonist for ryanodine-

8 H933 receptor channels. Endothelin-1 is also known to activate inositol 1,4,5-trisphosphate (IP 3 ) production and release Ca 2 through IP 3 -receptor channels (10, 27). In the present study, preexposure to E 2 did not affect the initial [Ca 2 ] i response to endothelin-1, suggesting that SR Ca 2 release through IP 3 -receptor channels is also not affected. Indeed, in permeabilized vascular smooth muscle, Kitazawa et al. (14) found that the force response to exogenous IP 3 was unaffected by E 2. Therefore, it is unlikely that SR Ca 2 release through either IP 3 - or ryanodine-receptor pathways is affected by E 2. However, these observations do not rule out the possibility that estrogens inhibit SR Ca 2 release via other agonists and/or signal transduction pathways. For example, Han et al. (4) found that E 2 does inhibit SR Ca 2 release induced by thromboxane A 2. Further experimentation is required to distinguish between these potentially differential mechanisms of estrogen action on the SR. In the present study, the lack of an E 2 effect of the rate of [Ca 2 ] i decline under conditions of blocked Ca 2 influx and efflux suggests that SR Ca 2 reuptake is also unaffected by estrogens. Thus the predominant site of action for estrogens appears to be the plasma membrane. In summary, the results of the present study demonstrate that E 2 enhances Ca 2 efflux, thereby providing a novel mechanism by which estrogens decrease the [Ca 2 ] i response to endothelin-1 in coronary artery smooth muscle cells from gonad-intact female pigs. This effect of E 2 requires estrogen receptors. We thank Thomas Keller and Kevin Rud for expert technical assistance. This research was supported by National Heart, Lung, and Blood Institute Grant HL-51736, the Mayo Foundation, and the University of Minnesota Graduate School. Y. S. Prakash was supported by a fellowship from Abbott Laboratories. Address for reprint requests: G. C. Sieck, Anesthesia Research, Mayo Clinic, Rochester, MN ( sieck.gary@mayo.edu). Received 4 June 1998; accepted in final form 16 November REFERENCES 1. Evans, R. The steroid and thyroid hormone receptor superfamily. Science 240: , Farhat, M. Y., S. Abi-Younes, and P. W. Ramwell. Nongenomic effects of estrogen and the vessel wall. Biochem. Pharmacol. 51: , Gilligan, D. M., D. M. Badar, J. A. Panza, A. A. Quyyumi, and R. O. I. Cannon. Acute vascular effects of estrogen in postmenopausal women. Circulation 90: , Han, S. Z., H. Karaki, Y. Ouchi, M. Akishita, and H. Orimo. 17 -Estradiol inhibits Ca 2 influx and Ca 2 release induced by thromboxane A 2 in porcine coronary artery. Circulation 91: , Harder, D., and P. Coulson. Estrogen receptors and effects of estrogen on membrane electrical properties of coronary vascular smooth muscle. J. Cell. Physiol. 100: , Hendricks, D. M. Estrogen concentrations in bovine and porcine tissues. J. Toxicol. Environ. Health 4: , Jiang, C., P. A. Poole-Wilson, P. M. Sarrel, S. Mochizuki, P. Collins, and K. T. MacLeod. Effect of 17 beta-oestradiol on contraction, Ca 2 current and intracellular free Ca 2 in guinea pig isolated cardiac myocytes. Br. J. Pharmacol. 106: , Jiang, C., P. M. Sarrel, P. A. Poole-Wilson, and P. Collins. Acute effect of 17 beta-estradiol on rabbit coronary artery contractile response to endothelin-1. Am. J. Physiol. 263 (Heart Circ. Physiol. 32): H271 H275, Kannan, M. S., A. M. Fenton, Y. S. Prakash, and G. C. Sieck. Cyclic ADP-ribose stimulates sarcoplasmic reticulum calcium release in porcine coronary artery smooth muscle. Am. J. Physiol. 270 (Heart Circ. Physiol. 39): H801 H806, Kannel, W. B., and T. J. Thom. Incidence, prevalence and mortality of cardiovascular disease. In: The Heart,editedbyR.C. Schlant and R. W. Alexander. New York: McGraw-Hill, 1994, p Katzenellenbogen, B. Dynamics of steroid hormone receptor action. Annu. Rev. Physiol. 42: 17 35, Kishi, Y., and F. Numano. A study of the mechanism of estrogen as an antiatherosclerotic: the inhibitory effect of estrogen on A23187-induced contraction of the aortic wall. Mech. Ageing Dev. 18: , Kitazawa, T., E. Hamada, K. Kitazawa, and A. K. Gaznabi. Non-genomic mechanism of 17 beta-oestradiol-induced inhibition of contraction in mammalian vascular smooth muscle. J. Physiol. (Lond.) 499: , Morley, P., J. F. Whitfield, B. C. Vanderhyden, B. K. Tsang, and J. L. Schwartz. A new nongenomic estrogen action: the rapid release of intracellular calcium. Endocrinology 131: , Mugge, A., M. Riedel, M. Barton, M. Kuhn, and P. R. Lichtlen. Endothelium independent relaxation of human coronary arteries by 17 beta-oestradiol in vitro. Cardiovasc. Res. 27: , Penney, L., R. Frederick, and G. Parker. 17 Beta estradiol stimulation of uterine blood flow in oophorectomized rabbits with complete inhibition of uterine ribonucleic acid synthesis. Endocrinology 109: , Prakash, Y. S., M. S. Kannan, and G. C. Sieck. Regulation of intracellular calcium oscillations in porcine tracheal smooth muscle cells. Am. J. Physiol. 272 (Cell Physiol. 41): C966 C975, Raddino, R., C. Manca, E. Poli, R. Bolognesi, and O. Visioli. Effects of 17 beta-estradiol on the isolated rabbit heart. Arch. Int. Pharmacodyn. Ther. 281: 57 65, Rodriguez, J., M. J. Garcia de Boto, and A. Hidalgo. Mechanisms involved in the relaxant effect of estrogens on rat aorta strips. Life Sci. 58: , Salas, E., M. G. Lopez, M. Villarroya, P. Saanchez-Garcia, R. De Pascual, W. R. Dixon, and A. G. Garcia. Endotheliumindependent relaxation by 17 alpha estradiol of pig coronary arteries. Eur. J. Pharmacol. 258: 47 55, a. Silva De Sa, M. F, and R. Meirelles. Vasodilating effect of estrogen on the human umbilical artery. Gynecol. Invest. 8: , Stice, S. L., S. P. Ford, J. P. Rosazza, and D. E. Van Orden. Interaction of 4-hydroxylated estradiol and potential-sensitive Ca 2 channels in altering uterine blood flow during the estrous cycle and early pregnancy in gilts. Biol. Reprod. 36: , Thastrup, O., A. P. Dawson, O. Scharff, B. Foder, P. J. Cullen, B. K. Drobak, P. J. Bjerrum, S. B. Christensen, and M. R. Hanley. Thapsigargin, a novel molecular probe for studying intracellular calcium release and storage Agents Actions 43: , Ushio-Fukai, M., J. Nishimura, S. Kobayashi, and H. Kanaide. Endothelin-1 and endothelin-3 regulate differently vasoconstrictor responses of smooth muscle of the porcine coronary artery. Br. J. Pharmacol. 114: , Vargas, R., G. Thomas, W. B., and P. W. Ramwell. Differential effects of 17 alpha and 17 beta estradiol on PGF 2 mediated contraction of the porcine coronary artery. Adv. Prostaglandin Thromboxane Leukot. Res. 19: , Velle, W. Endogenous anabolic agents in farm animals. Environ. Qual. Saf. Suppl. 5: , 1976.

9 H Wagner-Mann, C., Q. Hu, and M. Sturek. Multiple effects of ryanodine on intracellular free Ca 2 in smooth muscle cells from bovine and porcine coronary artery: modulation of sarcoplasmic reticulum function. Br. J. Pharmacol. 105: , Wagner-Mann, C., and M. Sturek. Endothelin mediates Ca influx and release in porcine coronary smooth muscle cells. Am. J. Physiol. 260 (Cell Physiol. 29): C771 C777, White, R. E., D. J. Darkow, and J. L. Falvo-Lang. Estrogen relaxes coronary arteries by opening BK Ca channels through a cgmp-dependent mechanism. Circ. Res. 77: , Williams, J. K., M. R. Adams, D. M. Herrington, and T. B. Clarkson. Short-term administration of estrogen and vascular responses of atherosclerotic coronary arteries. J. Am. Coll. Cardiol. 20: , Zhang, F., J. L. Ram, P. R. Standley, and J. R. Sowers. 17 -Estradiol attenuates voltage-dependent Ca 2 currents in A7r5 vascular smooth muscle cell line. Am. J. Physiol. 266 (Cell Physiol. 35): C975 C980, 1994.