Effects of temperature and buffer composition on calcium sequestration by sarcoplasmic reticulum and plasma membrane of rabbit renal artery.код для вставкиСкачать
THE ANATOMICAL RECORD 228:288-296 (1990) Effects of Temperature and Buffer Composition on Calcium Sequestration by Sarcoplasmic Reticulum and Plasma Membrane of Rabbit Renal Artery LINDA J. McGUFFEE, SALLY A. LITTLE, JUANITA V. MERCURE, BETTY J. SKIPPER, AND ELLYN S. WHEELER-CLARK Departments of Pharmacology (L.J.M., S.A.L., J.V.M.) and Family, Community and Emergency Medicine (B.J.S.) School of Medicine, University of New Mexico, Albuquerque, New Mexico 87131; Department of Pharmacology, School of Pharmacy, University of Wisconsin, Madison Wisconsin 53706 (E.S.W.-C.) ABSTRACT 45Ca electron microscopic autoradiography was used to examine the effects of buffer composition and temperature on the distribution of calcium in rabbit renal artery smooth muscle cells. The results show that the relative distribution of calcium is dependent on both the buffer used (Tris or Krebs) and the temperature of the bathing solution (25°C or 34°C). Krebs buffer at 34°C gave the highest relative activity in the plasma membrane, sarcoplasmic reticulum, and mitochondria. Buffer and temperature had little effect on the relative activity of the nucleus or cytoplasm. Next, we identified the cellular sites of calcium accumulation after 5, 15, 30, or 60 min exposure to 45Ca in Krebs buffer at 34°C. The results show that sarcoplasmic reticulum and plasma membrane are the primary sites of calcium accumulation during influx into these cells. Although the amount of 45Ca in the cell continues to increase with longer exposure, the relative distribution of calcium is essentially the same after 5 or 60 min. The data also indicate that the relative activity of plasma membrane + sarcoplasmic reticulum (a combination site that includes sarcoplasmic reticulum within a mean distance of 275 nm of the plasma membrane) is similar to the membrane alone and is lower than the sarcoplasmic reticulum alone. Several laboratories have reported on the suitability of Tris buffer for studying contractile responses in smooth muscle. Although some investigators found that Tris has a significant depressant effect (Altura et al., 1980; Sakai et al., 1983; Turlapaty et al., 1978), others reported that contractile responses are not inhibited in Tris solutions (Hayashi and Hester 1987; Johansson et al., 1979; Karaki et al., 1981).In addition, temperature (25°C vs. 34°C) has been reported to affect smooth muscle contractility (Karaki et al., 1987). In 1986, Wheeler-Clark et al. (1986) reported significant 45Ca accumulation in the plasma membrane and sarcoplasmic reticulum of rabbit renal artery in the absence of pharmacological stimulation. This accumulation is two to five times greater than previously reported in guinea pig and rabbit vas deferens (McGuffee et al., 1985, 1987). The latter studies were conducted at room temperature (25°C) and in Tris buffered (23.8 mM) solution. The higher temperature (34°C) and lower concentration of Tris (6 mM) used by Wheeler-Clark et al. (1986)were suspected to be major contributors to the increase in plasma membrane and sarcoplasmic reticulum calcium observed in that study. However, tissue and species differences could not be ruled out. With this in mind, we designed experiments to define better the effect of buffer and temperature on the cellular distribution of calcium in rabbit renal ar0 1990 WILEY-LISS, INC. tery smooth muscle. In the present paper, we present data obtained at 25°C and 34°C in Krebs Ringer bicarbonate buffer and at 25°C in 23.8 mM Tris. We have examined these muscles using in vitro tension measurements and 45Ca electron microscope (EM) autoradiography (ARG), which we previously have shown has sufficient ultrastructural and spatial resolution to localize calcium within smooth muscle cells (McGuffee et al., 1985, 1987; Skipper and McGuffee 1985; WheelerClark et al., 1986). Calcium flux studies in intact smooth muscle tissue have shown that equilibrium between calcium influx and efflux is reached in about 60 min (Hester e t al., 1978). However, flux studies cannot identify sites of calcium accumulation within the cell. Moreover, they cannot be used to determine whether, over time, calcium redistributes among cellular sites. This information can only be obtained with a technique that combines ultrastructural resolution with calcium flux. In the present study, we used EM ARG to identify the cellular sites that sequester calcium after 5, 15, 30, or 60 min of exposure to 45Ca. Some of these results were reported previously (McGuffee et al., 1988). Received January 8, 1990; accepted March 2, 1990. CALCIUM SEQUESTRATION IN RENAL ARTERY MATERIALS AND METHODS Solutions Krebs Ringer bicarbonate solution (KRB) contained 125.1 mM NaC1, 4.7 mM KCl, 1.2 mM KH,PO,, 18.7 mM NaHCO,, 1.2 mM MgS0,.7H2O, 2 mM CaCl,, 5.6 mM glucose. A depolarizing KRB solution was made by replacing Na+ with K + , giving a final K + concentration of 129.1 mM. All KRB solutions were oxygenated with 95% 0, and 5% CO,. Solutions were adjusted to pH 7.4 a t the temperature at which they were to be used (25°C or 34°C). The Tris solution contained: 125 mM NaC1, 2.7 mM KC1, 2 mM CaCl,, 11 mM glucose, and 23.8 mM Tris (Trizma Base, Sigma Chemical Co., St. Louis, MO) buffer. A depolarizing Tris solution was made by replacing Na+ with K + in the high Tris solution, giving a final K + concentration of 127.7 mM. All Tris solutions were oxygenated with 100% O,, and the pH was adjusted to 7.5 at either 34°C or 25°C. Tension Studies Renal arteries were removed from six male NZW rabbits, immersed in oxygenated KRB solution, cleaned of connective tissue, and cut into spiral strips. Each spiral strip was cut in half, giving a total of four tissues per animal. Strips were hung in muscle baths under 500 mg of tension, and equilibrated in oxygenated solution for 90 min. The equilibration solution was one of the following: KRB at 34”C, KRB at 25”C, 23.8 mM Tris at 34“C, or 23.8 mM Tris at 25°C. The solution was then drained from the bath and replaced with the similar K + substituted depolarizing solution for 10 min to induce contraction. The depolarizing solution was drained and replaced with one of the three remaining nondepolarizing solutions, and the sequence was repeated. I n this manner, muscles were exposed to all four experimental conditions in a randomized order. The magnitude of each contraction was measured and a two-way ANOVA using specimen and solution as the independent variables was used to test for significant differences among treatments. All possible pairwise comparisons among treatments were made using the Newman-Keuls procedure (SAS, 1988). 45Ca EM Autoradiography Studies Temperature and buffer solution studies Renal arteries were removed from three male NZW rabbits and placed in oxygenated KRB solution. Arteries were cleaned of connective tissue, cut into 2 mm ieces, and incubated in oxygenated KRB containing ‘%a (1mCi/ml, total Ca = 2.0 mM) for 60 min either a t room temperature (25°C) or at 34°C. Other tissues from the same renal arteries were incubated in Trisbuffered solution (23.8 mM) containing 45Ca at room temperature. After 60 min, tissues were quick frozen and freeze-dried as described below. 45Ca uptake vs. time studies Renal arteries from three male NZW rabbits were incubated in oxygenated KRB containing 45Ca (1mCi/ ml, total Ca = 2.0 mM) a t 34°C. Tissue from each rabbit was exposed to 45Ca for 5, 15, 30, or 60 min. After the appropriate exposure period, tissues were quick frozen and freeze-dried as described below. 289 Quick freezing, freeze-drying, and preparation of autoradiograms Tissues were quick frozen against a highly polished copper bar using the ‘“Gentleman Jim” freezing device (Ted Pella, Inc.). Frozen tissues were maintained below - 80°C throughout the drying procedure to prevent formation of liquid water within the specimens (Chiovetti et al., 1987). Tissues were dehydrated under vacuum a t low temperature in a glass freeze-drying apparatus (Chiovetti et al., 1987; McGuffee et al., 1981). The dried tissues were exposed to osmium tetroxide vapor in vacuo and embedded in Spurr resin. An example of the quality of tissue preservation with this technique is given in Figure 1. Tissues were osmicated in vacuo to prevent subsequent calcium loss from the freeze-dried, Spurr-embedded smooth muscle when the tissue is cut on water (McGuffee and Little, 1986). Thin (90-100 nm) sections were cut onto water using a Sorvall MT2B ultramicrotome, collected on EM grids, and carbon coated (Chiovetti e t al., 1987). Based on the ultrastructural preservation of the cells, eight blocks of tissue were selected from each treatment, and a minimum of six grids of tissue were collected from each block. Grids were coated with a monolayer of Ilford L4 emulsion according to the method described by Caro and Van Tubergen (1962). Briefly, this method involved melting the emulsion in a water bath then cooling the emulsion a t room temperature. To determine when the emulsion is a t the correct temperature to form a monolayer, a 4-cm-diameter wire loop was dipped into the emulsion and withdrawn at a uniform rate. The thin film in the loop was then touched to the surface of a microscope slide to deposit the film on the slide. When dry, a uniform monolayer of Ilford L4 emulsion has a purple interference color (Salpeter and Bachmann, 1972). Once a monolayer was obtained, slides containing the grids were coated. The emulsion was exposed in the dark a t 4°C for a n appropriate period (8-12 weeks). Autoradiograms were processed in Kodak D-19 developer and Fixer and in-cubated in Reynolds’ lead citrate for 2030 min to remove undeveloped emulsion. Grids were poststained in saturated uranyl acetate in 50% ethanol followed by Reynolds’ lead citrate to enhance contrast. Exposure to neither uranyl acetate nor Reynolds’ lead citrate causes a shift in the location of developed silver grains (McGuffee et al., 1981). Micrographs were taken only from grids that had a background grain density in tissue-free areas of plastic of 520% of the grain density over the tissue. Micrographs were taken at a n initial magnification of 10,000 and photographically enlarged to 24,000 for morphometric and grain distribution analyses. A typical autoradiogram is shown in Figure 2. Data collection and analysis Our goal is to associate the developed grains with their cellular origin. The sites of origin in which we are interested are the cytoplasmic matrix (C), which occupies most of the cell volume, and the plasma membrane (PM), nucleus (N), sarcoplasmic reticulum (SR), and mitochondria (MI, which occupy much smaller propor- 290 L.J. McGUFFEE ET AL. Fig. 1. Rabbit renal arteries were rapidly frozen, dehydrated under vacuum below -9O"C, exposed to osmium tetroxide vapor to enhance contrast and to preserve ionic distribution, embedded in Spurr resin, and thin sectioned onto water. The ultrastructure of the cells is similar to conventionally prepared tissue, with the exception of mitochondria (M), which are extremely electron dense in freeze-dried tissue. The nucleus (N) and sarcoplasmic reticulum (S) are well preserved. Cells are separated from one another by extracellular matrix (E). Bar = 1 p,m. tions of the cell. Due to the relatively high energy (mean energy 0.07 MeV; maximum energy = 0.25 MeV) of the p particle emitted during 45Cadecay, a developed grain will frequently be observed outside of its source. This is especially true when the source of radiation is small relative to the range of the beta emission, as in the case for PM, SR, and M. We account for radiation spread with the following analysis. Morphometric data are obtained from the autoradiograms using the point and circle technique described by Williams (1969,1977). Grain density information is obtained using the probability circle method of Salpeter and McHenry (1973). The statistical analysis that allows us to compare data from different treatments was developed in our laboratory (Skipper and McGuffee, 1985). To determine the grain distribution, a 275nm-radius circle is centered over each grain (McGuffee et al., 1985). Our choice of this size circle is based on both theoretical and experimental data showing that 275 nm approximates the half distance for a point source of calcium (Salpeter et al., 1987; Winegrad, 291 CALCIUM SEQUESTRATION IN RENAL ARTERY Fig. 2. A tangential section was cut through renal artery cells after incubation with 45Ca,freeze drying, and preparation for EM ARG. Thirteen 45Ca grains are shown on this autoradiogram. A 275 nm radius circle is drawn around one of the grains, which overlies the compound site “PM C + EM.” The method for partitioning grain density among PM, C, and E in this compound site is discussed in Data Collection and Analysis. Abbreviations as in Figure 1. Bar = 1 Km. + 1965). If a n organelle or PM is included within the circle, the grain is assigned to the respective compound site. For example, the circled grain in Figure 2 is assigned to the compound site “PM + C + E.” As can be seen in the figure, PM never fills a 275-nm-radius circle. It is always observed a s part of a compound site. The next step is to estimate the fractional volume of PM, E, and C in the compound site. This is done by placing a plastic overlay containing a n array of 275nm-radius circles, the center of each circle denoted by a point, on each autoradiogram (Wheeler-Clark et al., 1986). For circles overlying PM + C + E, the individual site (PM, C, or E) under the point is recorded. By repeating this process on all the autoradiograms, the fractional volume occupied by each component of PM + C E is calculated (Weibel and Bolender, 1973). The final step is to calculate the grain density of PM. From other data we have already calculated the grain + density of C more than 275 nm away from organelles and PM. We also have calculated the grain density for E. We must assume that the grain density of C included in PM + C + E is the same as the grain density of C observed alone a s a n individual site. A similar assumption is made for the E. We are unaware of any evidence in the literature t h a t would negate these assumptions. We then subtract the grain density contributed to PM + C + E by C and by E. The remaining grain density is allocated to PM. Autoradiographic data are expressed in units of relative activity: Relative activity = GilGT AiiAT ~ where Gi is number of grains observed in a cellular site, Ai is the fractional volume for that site, GT is the 292 L.J. McGUFFEE ET AL. total number of grains counted, and AT is the total volume. [Note: The method we used to calculate relative activity is based on the equation published by Williams (1977) and Kent and Williams (1974) to calculate what they called “relative crude specific activity.”] Estimates of relative activity can be obtained for PM, SR, M, and N a s long as they are surrounded by C and/or E. When two small sites, for example, PM and SR, are included in one compound site, additional equations would have to be used to attribute activity proportionately. Although we have not yet developed these equations, we predict that the variability estimates calculated with these new equations would be very large relative to the mean relative activity. Moreover, additional assumptions (i.e., homogeneity of activity among central and peripheral SR) would have to be made. However, if PM + SR is considered as one cellular site separate from PM or SR, the technique can be used to estimated the relative activity of the PM + SR combination site. The same is true for any other combination site. The following example illustrates what relative activity means in terms of our results. If the SR occupies 4% of the cell and contains 4% of the grains, the relative activity is 1. If 40% of the grains are in the SR, the relative activity is 10. If the N occupies 8% of the cell and contains 4% of the grains, the relative activity is 0.5. If 40% of the grains are in the N, the relative activity is 5. Thus relative activity estimates the distribution of cellular calcium among cellular components. It does not estimate the molar concentration of calcium. A relative activity of >1.0 indicates calcium accumulation by the site; a relative activity of <1.0 indicates calcium exclusion by the site. A relative activity = 1.0 indicates no accumulation or exclusion by the site. [Note: In previous publications (McGuffee et al., 1985, 1987; Wheeler-Clark et al., 19861, we have called this a random distribution.] As discussed in Skipper and McGuffee (19851, the calculation for relative activity assumes no variability in the area measurement (Ail for the cellular sites. Although this assumption does not appreciably affect the results for sites that occupy a relatively large portion of the cell such as C, PM, and N it is not valid for sites that occupy less than about 5% of the cell and have a high relative activity (Skipper and McGuffee, 1986). The net result is that the standard error of the mean (SEM) calculated from the equations in Skipper and McGuffee (1985) for a large site is a more reliable estimate of variability than is the SEM calculated for a small site. To obtain a better estimate for the SEM of all cellular sites, we have developed a simulation program that introduces variability in Ai (Skipper and McGuffee, 1986). We have used this computer program to determine the SEM for each experimentally determined value of relative activity reported in this paper. All possible pairwise comparisons of relative activities for each cellular site were made using a Z test (Skipper and McGuffee, 1985). Because four different treatments were compared for each site, the upper limit for P was set at 0.008 (a = 0.05/6 Bonferroni correction) for determining significance (Duncan e t al., 1983). TABLE 1. Contractile remonses generated in KRB or 23.8 mM-Tris buffer at 25°C or 34°C’ Treatment KRB 34°C KRB 25°C Tris 34°C Tris 25°C Grams of tension (mean t 2 SEMI 2.87 i 0.43* 2.02 t 0.39 2.54 i 0.39* 1.95 t 0.35 *Significantly different from all other treatments ( P < .01). *n = 22 tissues for each treatment; KRB = Krebs Ringer bicarbonate buffer. RESULTS In vitro tension studies were carried out to examine the effects of buffer type and temperature on the rabbit renal artery’s capacity to contract. The results are shown in Table 1. Temperature of the bathing solution had a greater effect on the amount of tension produced by the tissue than did buffer composition. Tissues generated the greatest contractile tension in KRB at 34”C, followed by Tris at 34”C, KRB at 25”C, and Tris at 25°C. Using the Newman-Keuls procedure, all possible pairwise comparisons were examined for significant differences. Both KRB at 34°C and Tris at 34°C were significantly different from each other and from KRB and Tris at 25°C (P < .01). KRB and Tris a t 25°C were not significantly different from each other. The fractional cell volume occupied by the cellular sites was determined in tissues from three animals that had been incubated in KRB at 34°C. On average, 76.5% of the cell is C, 8.7% PM, 7.9% N, 4.4% SR, 2.0% M, and 0.5%PM + SR. As is discussed below, sites that occupy less than about 5% of the cell will have the largest SEM relative to the magnitude of the relative activity. We used EM ARG to examine the effects of buffer solution and temperature on calcium distribution among the PM, SR, M, N, and C. Results are expressed in units of relative activity and are given in Figure 3. The relative distribution of calcium is dependent on both the buffer (Tris or KRB) and the temperature (25°C or 34°C). Use of KRB at 34°C gave the highest relative activity in the PM, SR, and M. The data in Figure 3 shows that there is calcium sequestration at the PM in KRB at 25°C and 34°C (relative activity significantly >1.0) but not in Tris at 25°C. Pairwise comparisons of the three different treatments were carried out using a Z test. The relative activities of PM in KRB at 25°C and 34°C (columns 2 and 3) are not significantly different from one other. This suggests that 45Cauptake at the PM is more sensitive to the type and concentration of buffer than to temperature. The relative activity in 23.8 mM Tris buffer at 25°C (column 1) is significantly less than with either of the other treatments. In contrast to the PM, sequestration of calcium a t the SR depends more on the temperature of the bathing medium than on the type of buffer. There is calcium sequestration in the SR in KRB a t 34°C (relative activity significantly >1.0) but not a t 25°C. In KRB a t 34°C (column 3 in Fig. 3),the relative activity of the SR is significantly greater than at 25°C in either Tris or 293 CALCIUM SEQUESTRATION IN RENAL ARTERY 16 14 1 2 3 .3 12 .-> 1 2 0TRIS, 25°C = KRB, 25°C KRB, 34°C 3 4 4 0 4 0 5 rnin = 15 0 = rnin 30 rnin 60 rnin lo a , E > .+ 0 - 6 z 4 2 0 PM SR M N C Cellular Sites Fig. 3. Relative activity is shown for each cellular site in the rabbit renal artery in the presence of either Tris or KRB a t 25°C or 34°C. Data for Tris buffer at 34°C are from Wheeler-Clark et al. (1986). (Note: in Figs. 3 and 4 and Table 1, mean values & 2 SEM are plotted, since 2 SEM represent the 95% confidence interval.) All possible pairwise comparisons were made for each tissue site. Relative activities that are significantly different ( P < .05) are PM: 1,2; 1,3; SR 1,3; 2,3; N: 1,3;C: 1,2; 1,3. KRB (columns 1 and 2). This suggests that a greater fraction of the total cellular calcium is in the SR at 34°C than at 25°C. There is no significant calcium sequestration in the M with any treatment (relative activity not significantly different from 1.0). The relative activity of the M is higher at 34°C (column 3) than at 25°C (column 1and 2). However, none of the three treatments are significantly different from one another due to large variability in relative activity estimates. As was discussed in a prior publication (Skipper and McGuffee, 19851, the SEM is inversely proportional to the fractional volume of the site. Morphometric results indicate that M occupy about 2% of the rabbit renal artery cell. Consequently, the SEM of the relative activity for M will be large. Thus it will be very difficult to detect statistically significant shifts in calcium associated with M unless unrealistically large sample sizes are analyzed. Type of buffer and temperature have very little effect on the relative activity of N and C. There is no significant sequestration of calcium in the N or C in any of the treatments (relative activity I1.0). However, N relative activity is significantly greater in Tris buffer at 25°C than in KRB at 34°C (columns 1 and 3 in Fig. 3). There is significantly more calcium in the C in Tris at 25°C than in KRB at 25°C (columns 1 and 2). These results suggest that more of the total cellular calcium is in the N and C in Tris buffer at 25°C than in KRB at 34°C. In addition, the C has more of the total cellular calcium in 23.8 mM Tris buffer at 25°C than in KRB at 25°C. Figure 4 shows how 45Ca is distributed among the cellular sites after 5, 15, 30, and 60 min exposure to 45Ca in KRB at 34°C. These data indicate only the fractional distribution of the total 45Cauptake at each time interval and the reader should be reminded that the total 45Ca accumulation has been shown to increase over time up to 60 min (Hester et al., 1978). Radioac- -2 J PM SR PM-SR M N C Cellular Sites Fig. 4. Relative activity 2 2 SEM for cellular sites after varying the time of exposure to 45Cain the bathing medium. All possible pairwise comparisons were made for each tissue site. Relative activities that are significantly different ( P < .05) are PM: 2,3; 3,4; SR: 3,4. The relative activity for M at 5 and 30 min is below, but not significantly different from, zero. Because of random variability and the small volume fraction occupied by M in this tissue, it is mathematically possible to calculate a relative activity value below zero. However, negative values are physiologically meaningless and should be considered zero. tive calcium is detected in all the cellular sites at every time period. However, the distribution of labeled calcium is not uniform across cellular sites, nor does the distribution remain constant with time. Within 5 min, calcium sequestration is observed at the PM. Sequestration is also observed after 15 and 60 min (relative activity significantly >1.0) but not after 30 min. The relative activity after 30 min (column 3) is significantly less than that after 15 or 60 min (columns 2 and 4) but not significantly different from that at 5 min (column 1). The SR sequesters a major proportion of the total cellular calcium within 5 min. The SR continues to sequester calcium after 15, 30, and 60 min (relative activity significantly >1.0). The only significant shift in SR calcium between 5 and 60 min is a decrease after 30 rnin compared to 60 min. In addition to examining PM and SR alone, we estimated the relative activity of the combination site PM + SR. PM + SR includes junctional and peripheral SR within 550 nm of the PM (mean distance = 275 nm). In this study, PM + SR does not sequester significant levels of calcium at any time period (relative activity not significantly different from 1.0) and relative activities after 5, 15, 30, and 60 min are not significantly different from each other. Calcium is not sequestered by the C a t any time period (relative activity <1.0), nor are there significant differences in relative activity among the four time periods. The same is true of the N and the M. DISCUSSION Contractile responses in vascular smooth muscle are temperature dependent. Significantly more tension is produced by strips of rabbit renal artery in KRB or in Tris at 34°C than at room temperature (25°C) (Table 1). A similar observation was made by Karaki et al. 294 L.J. McGUFFEE ET AL. (1987), who compared contractility of rabbit aorta at 37°C and 25°C. Using a bicarbonate-buffered system, they showed that a t 37°C both norepinephrine and high potassium elicited a larger increase in tension than at 25°C. Both norepinephrine (Bond et al., 1984; Kowarski e t al., 1985) and high potassium (Wheeler-Clark et al., 1986) can release calcium from SR. It may be that we observed a larger contraction at 34°C than a t 25°C because there is more releasable calcium present in the SR at the higher temperature. Our EM ARG results, discussed below, support this hypotheses, since proportionally more calcium is associated with the SR a t 34°C than at 25°C. A similar observation has been made using skinned rat caudal artery. In that preparation, there is over 75% more calcium in the SR at 38°C than a t 24°C (M. Stout, personal communication). We examined the effect of buffer composition and temperature on the distribution of calcium in rabbit renal artery cells. Results show that the highest relative activities in the PM and SR are obtained when tissues are exposed to KRB buffer a t 34°C (Fig. 3). The data also show that 45Ca uptake a t the SR is more sensitive to the temperature of the bathing medium than to the type of buffer. Proportionally more calcium is associated with the SR at 34°C than a t 25°C. This finding is in good agreement with the results of Wheeler-Clark et al. (1986). They calculated that at 34°C in 6 mM Tris buffer, a n SR relative activity of 10.5 t 2.0 in this tissue. This is not significantly different from our SR relative activity of 11.6 2 0.9 in KRB at 34°C (Fig. 3). On the other hand, it is significantly more than the relative activity a t 25°C in 23.8 mM Tris (1.8 -t 0.3) or KRB (2.1 2 0.4) shown in Figure 3. In other words, both 6 mM Tris and KRB at 34°C support sequestration of calcium in the SR. Neither 23.8 mM Tris at 25°C nor KRB at 25°C support sequestration in the SR. On the other hand, we found t h a t calcium sequestration at the PM is more sensitive to the type of buffer than to temperature of the bathing solution. This finding is also in agreement with the results of WheelerClark e t al. (19861, who reported a PM relative activity of 4.1 t 0.8 in 6 mM Tris at 34°C. This is not significantly different from the relative activity of 4.7 2 0.4 in KRB at 34°C or 4.4 2 0.4 in KRB at 23°C (Fig. 3). However, it is significantly more than the relative activity of 0.1 2 0.5 in Tris at 23°C (Fig. 3). Thus, in 23.8 mM Tris buffer the PM excludes calcium but in 6 mM Tris and in KRB i t sequesters calcium. We conclude from the data in Figure 3 and from the previous results of Wheeler-Clark et al. (1986) that both type of buffer and temperature can significantly alter the relative distribution of calcium in the renal artery smooth muscle cell. Both are important parameters to consider when designing in vitro experiments on smooth muscle. The effect of Tris buffer on isolated smooth muscle function has been studied in several different laboratories with conflicting results. Turlapaty and coworkers (1978) compared the effects of Tris and KRB buffers on contractile responses of rat portal vein and aorta. Tris buffer in a range of 5-30 mM decreased spontaneous mechanical activity in a concentration dependent manner. Tris (5 mM) also attenuated contractile responses to KC1, epinephrine, and angiotensin. In a later paper, this same group (Turlapaty et al., 1979) used calcium flux to study the effect of Tris and KRB on tissue calcium content and found that total calcium uptake was reduced in the presence of 30 mM Tris. In other laboratories, Tris in a range of 5-25 mM had little or no effect on elicited contractions in rabbit aorta particularly when the sodium concentration was kept constant (Karaki et al., 1981). Conflicting results such a s these have raised questions about the advisability of using Tris-buffered solutions, particularly in high concentrations, in in vitro studies of contractile tension. Kwan and Daniel (1981) investigated the effect of Tris on calcium accumulation by membrane fractions isolated from rat mesenteric arteries and veins. They observed that 50 and 100 mM Tris reduced both ATPdependent and -independent calcium accumulation in a purified PM fraction and that the inhibition was concentration dependent. This suggests that Tris may interfere with both ATP-dependent calcium efflux from the cells and also with non-ATP-dependent binding to PM vesicles. Although previous results suggested that exposure to Tris might physiologically or biochemically alter calcium accumulation in smooth muscle, they did not identify unequivocally which calcium pools are affected by Tris buffer in situ. However, these studies do provide important information on the concentration dependence of a Tris effect and did suggest that a t least part of the effect is at the PM. The EM ARG data reported here confirm the PM as the primary locus of the Tris effect in intact tissue. In addition, the effect appears to be concentration dependent, since the relative activity of the membrane in 23.8 mM Tris is two orders of magnitude less than the activity in 6 mM Tris (Fig. 3). The exact site at which Tris alters calcium binding at the PM is not clear. As a cationized molecule, Tris might be acting on the external surface of the membrane to compete with calcium for binding sites. Alternatively, Tris also could interfere with internally bound calcium, because a significant portion of Tris (30%) is unionized at pH 7.4 and can penetrate the cell membrane (Holmdahl and Nahas, 1962). Therefore, the reduction of PM calcium sequestration by Tris in intact cells does not necessarily clarify on which side of the membrane bilayer the PM calcium pool is located. The mechanism by which Tris inhibits PM calcium binding is also unclear. Cationic Tris might compete with calcium for binding to negatively charged proteins, lipids, andlor carbohydrates, such as sialic acid, associated with the PM-surface coat. In addition, Tris has been shown to have a variety of effects on animal and plant proteins (Ljungbery et al., 1984; Murphy and Sastre, 1983). For example, Tris exposure can alter membrane phospholipid content of certain bacteria (Donohue et al., 1982), release outer membrane components (Irvin et al., 1981a), and increase membrane permeability (Irvin et al., 1981b). Because of the variety and complexity of interactions between membranes and Tris, it seems equally plausible that the effects described in this study could result from direct competitive antagonism with calcium for PM binding sites or from indirect effects on membrane composition and permeability. It also seems possible that subtle differences in membrane composition could result in spe- CALCIUM SEQUESTRATION IN RENAL ARTERY cies-andlor tissue-dependent differences in susceptibility to the effects of Tris. Such differences could account, in part, for the conflicting results reported for Tris in the literature. Therefore, additional studies would be helpful to clarify the mechanism of T r i s membrane interactions. Regardless of the manner in which Tris affects membranes, we agree with previous investigators who have warned against using Tris in high concentrations in studies of calcium accumulation by isolated membrane fractions (Kwan and Daniel, 1981; Kwan et al., 1984) and in intact muscle preparations (Altura et al., 1980; Sakai et al., 1983; Turlapaty et al., 1978, 1979). In addition, we suggest that, to obtain maximum uptake by SR in vascular smooth muscle, experiments should be carried out at 34°C rather than a t room temperature. In a separate series of experiments, we examined calcium sequestration in the PM, SR, PM + SR M, N, and C after 5, 15, 30, and 60 min exposure to 45Ca in KRB at 34°C. These results are the first to show that SR and PM are the primary sites of calcium accumulation during influx into rabbit renal artery. The PM and SR accumulated significant proportions of the total cellular calcium after only 5 min; the M, N, and C did not. Although the amount of 45Ca in the cell continues to increase with longer exposure (not shown on graph), the relative distribution is essentially the same after 5 or 60 min. At the present time, we do not have a n explanation for the significant decrease in the relative activity at both the PM and SR after 30 min (Fig. 4). It has been suggested that “peripheral” SR in smooth muscle can sequester and release large amounts of calcium for excitation-contraction coupling as do the terminal cisernae and junctional SR of skeletal and cardiac muscle. To investigate this possibility, other investigators have used X-ray microprobe analysis to identify “hot spots” within 50 nm of the PM in smooth muscle (Bond et al., 1984). In guinea pig portal vein, these “hot spots” are presumed to be SR, although SR has not yet been visually identified as such in these preparations. Using ARG, we have shown that the relative calcium activity of the PM + SR region is very close to that of the PM alone and much lower than that of the “central” SR (Fig. 4). A direct comparison of these results with those previously published by Bond et al. (1984) is not possible due to the differences in the technique. However, our findings show that on average, PM + SR sites in rabbit renal artery do not sequester greater amounts of calcium than the “central” SR or PM alone. Therefore, the importance of calcium in the PM + SR region relative to “nonjunctional” PM or SR in excitation-contraction coupling of smooth muscle requires further investigation. ACKNOWLEDGMENTS This work was supported by NIH grant GM30003 to L.J.M. LITERATURE CITED Altura, B.M., B.T. Altura, A. Carella, and P.D.M.V. Turlapaty 1980 Adverse effects of artifical buffers on contractile responses of arterial and venous smooth muscle. Br. J. Pharmacol., 69t207-214. Bond, M., T. Kitazawa, A.P. Somlyo, and A.V. Somlyo 1984 Release and recycling of calcium by the sarcoplasmic reticulum in guineapig portal vein smooth muscle. J. Physiol., 355t677-695. 295 Caro, L.G., and R.P. Van Tubergen 1962 High-resolution autoradiography. I. methods. J . Cell Biol., 15t173-188. Chiovetti, R., L.J. McGuffee, S.A. Little, E. Wheeler-Clark, and J. Brass-Dale 1987 Combined quick freezing, freeze-drying, and embedding tissue a t low temperature and in low viscosity resins. J. Electron Microsc. Techn., 5tl-15. Donohue, T.J., B.D. Cain, and S. Kaplan 1982 Alterations in the phospholipid composition of Rhodopseudomonas sphaeroides and other bacteria induced by tris. J . Bacteriol., 152t595-606. Duncan, R.C., R.G. Knapp, and M.M. Clinton 1983 Introductory Biostatistics for the Health Sciences. John Wiley and Sons, New York, pp. 151-152. Hayashi, S., and R.K. Hester 1987 Endothelium-dependent relaxations in rabbit aorta are depressed in artificial buffered solutions. J . Pharmacol. Exp. Ther., 242r523-530. Hester, R.K., G.B. Weiss, and W.J. Fry 1978 Effects of norepinephrine, dopamine and potassium on tension and Ca-45 fluxes in canine and rabbit renal arteries. J . Pharmacol. Exp. Ther., 207: 364-371. Holmdahl, M.H., and G.G. Nahas 1962 Volume of distribution of C14labeled tris(hydroxymethy1)aminomethane. Am J. Physiol., 202: 1011-1014. Irvin, R.T., T.J. MacAlister, R. Chan, and J.W. Costerton 1981a Citrate-tris(hydroxymethy1)aminomethane-mediated release of outer membrane sections from the cell envelope of a deep-rough (heptose-deficient lipopolysaccharide) strain of Escherichia coli 08. J. Bacteriol., 145t1386-1396. Irvin, R.T., T.J. MacAlister, and J.W. Costerton 1981b Tris(hydroxymethy1)aminomethane buffer modification of Escherichia coli outer membrane permeability. J. Bacteriol., 145t1397-1403. Johansson, B., B. Ljung, P. Hellstrand, B. Uvelius, and S.B. Sigurdsson 1979 Effects of tris on vascular smooth muscle (Letter to the editor). Am. J. Physiol., 237:H409-H411. Karaki, H., H.Y. Ahn, and N. Urakawa 1987 Caffeine-induced contraction in vascular smooth muscle. Arch. Int. Pharmacodynam., 285t60-71. Karaki, H., T. Suzuki, and N. Urakawa 1981 Tris does not inhibit isolated vascular or intestinal smooth muscle contraction. Am. J . Physiol., 241:H337-H341. Kent, C., and M.A. Williams 1974 The nature of hypothalamo-neurohypophyseal neurosecretion in the rat. J. Cell Biol., 60t554-570. Kowarski, D., H. Shuman, A.P. Somlyo, and A.V. Somlyo 1985 Calcium release by noradrenaline from central sarcoplasmic reticulum in rabbit main pulmonary artery smooth muscle. J. Physiol., 366r153-175. Kwan, C.Y., and E.E. Daniel 1981 Tris inhibits calcium accumulation by plasma membrane fractions isolated from vascular smooth muscle. Biochem. Int., 2t429-436. Kwan, C.Y., Y. Sakai, and E.E. Daniel 1984 Tris inhibits binding and transport of calcium in microsomal fraction isolated from rat vas deferens. Arch. Int. Pharacodynam., 269t252-262. Ljungberg, U., H.-E., Akerlund, and B. Anderson 1984 The release of a 10-kDa polypeptide from everted photosystem I1 thylakoid membranes by alkaline tris. FEBS Lett., 175r255-258. McGuffee, L.J., L. Hurwitz, S.A. Little, and B.E. Skipper 1981 A 45Ca autoradiographic and stereological study of freeze-dried smooth muscle of the guinea pig vas deferens. J . Cell Biol., 90t201-210. McGuffee, L.J., and S.A. Little 1986 Effect of osmium vapor on calcium loss during sectioning of freeze-dried, embedded tissue. Proc. Annu. Meet. EM SOC.Am., 44t258-259. McGuffee, L.J., S.A. Little, and B.J. Skipper 1987 Localization of calcium in vas deferens using 45Ca EM autoradiography: Relationship to species and the effect of 45Ca removal. Anat. Rec., 21 7: 321-327. McGuffee, L.J., J . Mercure, and S.A. Little 1988 Cellular sites of 45Ca accumulation during influx into smooth muscle cells. FASEB J., 2tA755 (abstract). McGuffee, L.J., E.S. Wheeler-Clark, B.J. Skipper, and S.A. Little 1985 The cellular distribution of calcium in freeze-dried rabbit vas deferens using EM autoradiography. Anat. Rec., 211t117125. Murphy, K.M.M., and A. Sastre 1983 Obligatory role of a trisicholine allosteric site in guanine nucleotide regulation of t3H]-L-QNB binding to muscarinic acetylcholine receptors. Biochem. Biophys. Res. Commun., 1 13t280-285. Sakai, Y., C.Y. Kwan, and E.E. Daniel 1983 Effect of tris buffer on the contractile responses of rat vas deferens. Eur. J. Pharmacol., 95: 161-169. Salpeter, M.M., and L. Bachmann 1972 Autoradiography. In: Principles and Techniques of Electron Microscopy. Vol. 2. M.A. Hayat, ed. Van Nostrand Reinhold Co., New York, pp. 221-278. 296 L.J. McGUFFEE ET AL. Salpeter, M.M., and F.A. McHenry 1973 Electron microscope autoradiography. In: Advanced Techniques in Biological Electron Microscopy. J.K. Koehler, ed. Springer-Verlag, Berlin, pp. 113-152. Salpeter, M.M., R. Nelson, R. Harris, and E.E. Salreter 1987 Resolution in EM autoradiography V: 45Ca (and PI. J . Histochem. Cytochem., 35t1047-1051. SAS Institute Inc. 1988 SAS/STATT'MUser's Guide, Release 6.03 Edition. SAS Institute Inc., Cary, NC, pp. 1028. Skipper, B.J., and L.J. McGuffee 1985 Statistical significance tests for autoradiographic data. Anat. Rec., 211t126-132. Skipper, B.J., and L.J. McGuffee 1986 Sample size estimates for autoradiographic experiments. Proc. R. Microsc. SOC., 21r268 (abstract). Turlapaty, P.D.M.V., B.T. Altura, and B.M. Altura 1978 Influence of tris on contractile responses of isolated rat aorta and portal vein. Am. J . Physiol., 235rH208-H213. Turlapaty, P.D.M.V., B.T. Altura, and B.M. Altura 1979 Tris(hydroxymethy1)aminomethane inhibits calcium uptake in vascular smooth muscle. Biochim. Biophys. Acta, 551t459-462. Weibel, E.R., and R.P. Bolender 1973 Stereological techniques for electron microscopic morphometry. In: Principles and Techniques of Electron Microscopy. M.A. Hayat ed. Van Nostrand Reinhold Co., New York, Vol. 3, pp. 237-296. Wheeler-Clark, E.S., G.B. Weiss, L.J. McGuffee, and L.M. Buja 1986 Cellular redistribution of 45Ca in rabbit renal artery determined by electron microscopic autoradiography: Changes in response to K+-induced depolarization and nitrendipine. J. Pharmacol. Exp. Ther., 239t286-295. Williams, M.A. 1969 The assessment of electron microscopic autoradiographs. In: Advances in Optical and Electron Microscopy. R. Barer and V.E. Cosslett, Academic Press, New York, Vol. 3, pp. 219-277. Williams, M.A. 1977 Quantitative methods in biology. In: Practical Methods in Electron Microscopy. A.M. Glauert. North Holland Publishing Co., Amsterdam, Vol 6, Part 1, pp. 85-169. Winegrad, S. 1965 Autoradiographic studies of intracellular calcium in frog skeletal muscle. J . Gen. Physiol., 48t455-479.