Vasopressin-, angiotensin II-, and alpha 1-adrenergic-induced inhibition of Ca2+ transport by rat liver plasma membrane vesicles.

A rapid method for isolating highly purified rat liver plasma membrane vesicles using isotonic medium and Percoll self-forming gradient centrifugation is described. The vesicles were characterized by enzyme markers and electron microscopy. The method also yielded a fraction rich in nuclei. The vesicles transported Ca2+ in an ATP-dependent manner and this was enhanced by oxalate. The Vmax for Ca2+ uptake was 0.65 +/- 0.08 nmol/mg X min, which was approximately 18-fold higher than for other liver plasma membrane preparations, and the Km for Ca2+ was 5.2 +/- 0.4 nM. Calcium uptake was inhibited by 40-50% in vesicles isolated from rat livers perfused for 3 min with 10(-7)M vasopressin. The half-maximally effective concentration of vasopressin was 5 X 10(-10)M which correlates with that for raising cytosolic Ca2+ and phosphorylase a. Inhibition was not significant in vesicles from livers perfused with vasopressin for only 1 min, indicating that inhibition of the Ca2+ pump may not be involved in the rise in cytosolic Ca2+ observed at 1-2 s with this hormone. Epinephrine (10(-5)M) and angiotensin II (10(-7)M) inhibited Ca2+ uptake by 31 +/- 10 and 26 +/- 5%, respectively, at 3 min. Glucagon (10(-7)M) had no effect. It is proposed that the inhibitory action of the Ca2+-dependent hormones on the plasma membrane Ca2+ pump plays an important role in the actions of these hormones by prolonging the elevation in cytosolic Ca2+.

the Ca2+ indicator Quin-2 (7). However, since the intracellular Ca" stores are limited, other changes must occur to prolong the actions of the Caz+-dependent hormones. Consequently, it has been postulated that either the plasma membrane (Ca2+-M$+)-ATPase is inhibited and/or the influx of extracellular Ca2+ is stimulated (l).' Indeed, al-adrenergic stimulation has been shown to stimulate the influx of 45CaZ+ from the external medium (8)(9)(10).
In this communication, we describe a rapid method to prepare vesicles from rat liver plasma membranes utilizing Percoll gradient centrifugation. These vesicles are able to transport Ca2+ in an energy-dependent manner. When they are isolated from rat livers perfused with Ca2+-dependent hormones, the Ca2+ transport activity is inhibited up to 50%. It is postulated that this effect ensures that the Ca2+ mobilizing hormones maintain an elevated [CaZ+Ji and hence produce prolonged physiological responses.

EXPERIMENTAL PROCEDURES
Plasma Membrane Isolation-Male 200-250-g, body weight, Sprague-Dawley rats (Harlan Industries, Indianapolis, IN) fed ab libitum on Purina Lab Chow were killed by cervical dislocation and their livers rapidly removed and placed in ice-cold medium containing 250 mM sucrose, 5 mM HEPES2-KOH, and 1 mM EGTA, pH 7.4. They were minced with scissors and homogenized by 10 passes with a loosefitting Dounce homogenizer (Wheaton) followed by 3 passes with a tight-fitting homogenizer, then diluted to give a 6% (w/v) homogenate. The homogenate was then centrifuged at 1,464 X g for 10 min, and the resulting pellet was resuspended in the isolation medium and diluted to give a 6% (w/v) suspension. A volume (10.4 ml) of this was mixed with 1.4 ml of Percoll (Pharmacia) in 15-ml Corex tubes and centrifuged at 34,540 X g for 30 min. Two distinct layers close to the top of the tube were revealed. These were harvested and washed in 5 volumes of 250 mM sucrose, 50 mM Tris-HCI, pH 8.0, and the resulting pellets were resuspended in the same medium. The protein content of the membranes was measured according to Lowry et al. (11).
Rat liver plasma membranes were also prepared by the method of Song et al. (12) for comparison purposes.
Rat Liver Perfmion-Livers of rats were perfused as described previously with bicarbonate buffer containing 2.4 mM Ca" (13). Livers were initially perfused with hormone-free medium for 6 min. The median lobe was then tied off and rapidly removed for the preparation of "time zero" plasma membranes. Perfusion of the remaining liver lobes was continued with hormone or saline being continuously infused into the portal cannula line. At designated times, the left and right lobes were ligated and rapidly removed for preparation of plasma membranes.
Measurement of Ca2+ Transport-Caz+ uptake by liver plasma membrane vesicles was assayed essentially as described by Chan and Junger (161,unless stated otherwise, in 500 p1 of medium containing 180 mM sucrose, 50 mM Tris-C1, pH 8.0, 0.2 mM EGTA, 18 rCi/ml of "Ca2+ (0.18 mM total, 7.5 nM free), 10 mM MgC12, 20 mM sodium azide, 1 NM ruthenium red, 10 mM Tris-ATP, and 4 mM Tris oxalate. The free concentration of Ca2+ was varied in some experiments and was calculated using the COMICS program (17). stants were taken from Sillen and Martell (18). The reaction was started by adding 100 pg of plasma membrane protein and the mixture was incubated at 37 "C in a shaking water bath. At appropriate times, 450-p1 samples were removed and filtered under vacuum on 0.45-pm pore Millipore filters (type HA) that had been soaked in 0.25 M KC1 for 1 h. The filters were washed three times with 5 ml of 250 mM sucrose and 40 mM NaC1. The membrane vesicles trapped on the filter were dried and counted in 10 ml of ACS aqueous counting scintillant (Amersham).

Characteristics of Plasm Membrane Vesicles Prepared by
Percoll Gradient Centrifugation-Preliminary differential centrifugation of rat liver homogenates showed that the bulk of the (Ca2+-Mg2+)-ATPase and 5'-nucleotidase activities sedimented at low speed (1,464 x g for 10 min) while most of the glucose-6-phosphatase and cytochrome c oxidase sedimented at higher g forces (data not shown). Therefore, the low speed pellet was resuspended and added to Percoll (11.9% final concentration) and centrifuged at 34,540 X g for 30 min to prepare plasma membranes. Two distinct layers appeared close to the top of the tube while all other material remained on the bottom. The two top layers were analyzed for enzyme markers and by electron microscopy. The very top layer was identified as rich in nuclei on the basis of electron microscope analysis. This fraction was low in 5'-nucleotidase, glucose-6phosphatase, (Ca2+-Mg2')-ATPase, and cytochrome c oxidase ( Table I). The second layer showed a large enrichment in two plasma membrane markers, namely 5'-nucleotidase and (Ca2+-Mg2')-ATPase. These were enriched 24-and 11-fold, respectively, over the total homogenate (Table I). Contamination of the fraction by microsomes and mitochondria was minimal since glucose-6-phosphate and cytochrome c oxidase showed 40 and 60% decreases in specific activity over the homogenate. For comparison, data from plasma membranes prepared by the Song et al. (12) procedure are shown in Table  I. It is evident that the membranes prepared using Percoll are more enriched in 5"nucleotidase and (Ca2+-Mg2')-ATPase. Electron microscopy revealed the presence of vesicles of different sizes in the Percoll-derived membranes and the virtual absence of mitochondria (data not shown). In general, these membranes had more vesicles and were less contaminated with other organelles than were Song et al. (12) membranes. ea2+ Transport by Plasma Membrane Vesicles-The time course of Ca2+ uptake by a preparation of Percoll-derived liver plasma membrane vesicles is shown in Fig. 1 (left). There was little uptake in the absence of ATP, whereas the ATP-dependent calcium uptake proceeded linearly for about 10-15 min. The ATP-dependent Ca2+ uptake at the end of 30 min for 3 separate preparations was 7.0 f 1.3 nmol/mg of protein in the presence of 4 mM oxalate and 4.5 f 0.7 nmol/mg of protein in its absence. The ATP-independent Ca2+ uptake or binding accounted for approximately 7% of the total Ca2+ accumulated at 30 min and oxalate had negligible effect on the basal activity.
For comparison, Caz+ uptake by Song et al. (12) membranes (Fig. 1, right) was measured under identical conditions to those in Fig. 1, left. It is very evident that these membranes showed much lower ATP-dependent Ca2+ transport, with rates similar to those reported by Chan and Junger (16).
Calcium uptake by Percoll-derived membrane vesicles was not altered by ruthenium red or NaN,, known inhibitors of mitochondrial Ca2+ transport. However, it was abolished by the ionophore A23187, indicating uptake into membrane bound vesicles (data not shown).
The ATP-dependent Ca2+ uptake of Percoll-derived vesicles was measured as a function of Mg2+ concentration at a constant ATP concentration of 10 mM and free Ca2+ concentration of 7.5 nM. No calcium uptake occurred when Mg2' was omitted. At 5 mM Mg2+, calcium uptake proceeded halfmaximally, and it reached a maximum at 10-15 mM M e . Ca2+ uptake was also studied as a function of ATP concentration. In these studies, approximately 20% of the initial ATP was hydrolyzed in 5 min as measured by [y3'P]ATP hydrolysis. The ATP concentrations for half-maximal and maximal activity were approximately 2 and 10-15 mM, respectively. The ATP-dependent calcium uptake by the vesicles was also measured as a function of free Ca2+ concentration (Fig. 2).

The K, and Vmx values derived from 4 separate experiments
using Eadie-Hofstee plots were 5.2 f 0.4 nM and 0.65 k 0.08 nmol/mg of protein. min, respectively. Effect of Vasopressin on ea2+ Transport by Plasma Membrane Vesicles-Direct addition of vasopressin ( 10-9-10-7 M) to plasma membrane vesicles 10 min before or simultaneously with ATP did not alter Ca2+ transport (data not shown). Therefore vesicles were prepared from the lobes of rat livers that had been perfused with lo-? M vasopressin. As shown in Fig. 3 (left), little difference was observed in the Ca2+ transport activity of lobes perfused with saline (control). However, the infusion of M vasopressin markedly inhibited Ca2+ uptake by the vesicles as shown in Fig. 3 (left). At 2,3, and 6 min, there was a 32, 39, and 50% inhibition, respectively, of Ca2+ uptake compared with control at each time. The effect of vasopressin at 1 min was not statistically significant.
Rat liver lobes were also perfused for 3 min with concentrations of vasopressin ranging from 10"' to lo-? M (Fig. 3, right). Maximal effects were seen at M vasopressin and the concentration for half-maximal effects was 5 X 10"O M (Fig. 3, right). These concentrations are very similar to those for vasopressin effects on [Ca2+Iir Ca2+ efflux, and phosphorylase in hepatocytes (e.g. see Ref. 3).l

TABLE I Distribution of enzyme markers
Plasma membrane and nuclei fractions were isolated on 11.9% Percoll from the low speed pellet (1464 X g for 10 min) of liver homogenates prepared in isotonic sucrose and were assayed for enzyme markers as described under "Experimental Procedures." Values are means & S.E. for 4-8 preparations. ND, not determined. Other hormones were tested to see if they also caused inhibition at 3 min. Epinephrine M) produced a 31 k 10% ( n = 4, p < 0.01) inhibition of Ca2+ uptake, which was reduced by the a,-adrenergic antagonist prazosin ( M) to produce a slight 2 f 12% ( n = 3) increase in Ca" uptake, thus indicating that the epinephrine effect was mediated by

DISCUSSION
This report describes a method for isolating rat liver plasma membrane vesicles capable of high ATP-dependent Ca" transport activity. The entire procedure can be performed within 1 h and involves centrifugation in a Percoll selfforming gradient of the low speed pellet obtained by homogenization of liver in isotonic medium. Plasma membranes prepared in this way are purer with respect to enzyme markers and electron microscopic appearance compared with the other well characterized liver plasma membrane preparations (Table I and Ref. 12). Another interesting feature of the method is that another fraction enriched in nuclei can be separated at the same time.
The plasma membrane vesicles prepared using Percoll transport approximately 5-fold more Ca2+ than other membrane preparations ( Fig. 1 and Ref. 16). One explanation may be that the Percoll-derived membranes have higher proportion of inside-out polarity. Another reason may be that the method involves isotonic medium and requires a relatively short time, whereas other methods use hypotonic media and/ or take longer (4-16 h). Under the latter conditions, some peripheral proteins may be lost from the membranes and there may be less formation of well sealed vesicles.
The characterization of Ca2+ transport in hepatic membranes with a K,,, of 14-17 nM for Ca2+ has been reported previously (16, 19). Our measurements on the K,,, for Ca2+ uptake, namely 5.2 nM (Fig. 2), agree with these studies. However, all of these K,,, values are below the concentration of cytosolic Ca" in unstimulated hepatocytes, which is approximately 200 nM (7). This suggests that factors in the intact cell may decrease the affinity of the Ca2+ pump.
In this work, we report for the first time the marked effects of vasopressin, angiotensin 11, and epinephrine on liver Ca2+ transport activity in subsequently isolated plasma membrane vesicles. Although the dose response for vasopressin inhibition of Ca2+ uptake by plasma membrane vesicles correlates well with that for the other effects of vasopressin in liver, it is not likely that this inhibition is involved in the primary signaI(ing) by which Ca2+-mobilizing hormones elevate cytosolic Ca". This elevation takes place 1-2 s after hormone addition to the liver (5, 71,' whereas the inhibition of the plasma membrane Ca2+ pump is not significant at 1 min. On the other hand, this inhibition may be related to the maintenance of elevated cytosolic Ca2+. This is possible because during hormone stimulation cytosolic Ca2+ remains elevated for some time before it returns to its resting level i.e. well beyond the time at which Ca2+ efflux from the liver cell ceases (3, 7).' It is interesting to note that the time frame of vasopressin inhibition of Ca2+ uptake by plasma membrane vesicles correlates with that far phosphatidylinositol hydrolysis in liver (20). Phosphoinositides are known to influence Ca2+-ATPase activity in some tissues (21-23). The nature of the alterations induced in the liver (Ca2+-Mg2+)-ATPase pump by the Ca2+-dependent hormones is under further study.