Correlation of Receptor Sequestration with Sustained Diacylglycerol Accumulation in Angiotensin 11-stimulated Cultured Vascular Smooth Muscle Cells*

Angiotensin I1 stimulates sequential phospholipase C-mediated hydrolysis of initially the polyphosphoinositides and subsequently phosphatidylinositol (PI) in cultured rat aortic smooth muscle cells resulting in biphasic, sustained formation of diacylglycerol (DG). The mechanisms underlying this delayed induction of sustained DG accumulation are unknown but may be related to cellular events including processing of the angiotensin I1 receptor-ligand complex. In the present study, we characterized the kinetics of angiotensin I1 receptor sequestration and studied the effects of interventions which interfere with receptor processing on the pattern of angiotensin 11-induced DG formation and phosphoinositide hydrolysis. Conversion of the an- giotensin I1 receptor to an acid-resistant form was temperature-dependent, with half-times of 1.5 min at 37 “C and 7 min at 19 “C. Reducing the temperature to 25 or 19 “C caused a marked temporal separation be- tween the two phases of DG accumulation. There was a close temporal correlation between the effect of tem- perature on receptor sequestration and on sustained DG accumulation. Furthermore, phenylarsine oxide (5 min, 10

Angiotensin I1 stimulates sequential phospholipase C-mediated hydrolysis of initially the polyphosphoinositides and subsequently phosphatidylinositol (PI) in cultured rat aortic smooth muscle cells resulting in biphasic, sustained formation of diacylglycerol (DG). The mechanisms underlying this delayed induction of sustained DG accumulation are unknown but may be related to cellular events including processing of the angiotensin I1 receptor-ligand complex. In the present study, we characterized the kinetics of angiotensin I1 receptor sequestration and studied the effects of interventions which interfere with receptor processing on the pattern of angiotensin 11-induced DG formation and phosphoinositide hydrolysis. Conversion of the angiotensin I1 receptor to an acid-resistant form was temperature-dependent, with half-times of 1.5 min at 37 "C and 7 min at 19 "C. Reducing the temperature to 25 or 19 "C caused a marked temporal separation between the two phases of DG accumulation. There was a close temporal correlation between the effect of temperature on receptor sequestration and on sustained DG accumulation. Furthermore, phenylarsine oxide (5 min, 10 p~) , which inhibited angiotensin I1 receptor internalization, also selectively inhibited the sustained phase of DG accumulation (81 2 6% inhibition). Monensin and chloroquin, which interfere with receptor processing through the lysosomal-degradative pathway, had no effect on angiotensin 11-induced DG formation in these cells, suggesting that the processing event important to hormonally induced sustained DG accumulation occurs early in the internalization pathway, probably at the level of the plasma membrane. Moreover, the acid-resistant state of the angiotensin 11 receptor-ligand complex retained its ability to signal, since removal of the surface signal by competitive antagonism with Sarl-Iles-angiotensin I1 or acid-wash only slowly reversed accumulation of DG and depression of total cell calcium. These experiments support our previous observation that the initial and sustained phases of angiotensin 11-induced diacylglycerol formation in vascular smooth muscle are differentially * This work was supported by National Institutes of Health Grants HL38206, HL35013, HL36028, and HL07042. Funds were contributed by the Massachusetts Affiliate of the American Heart Association.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $$ To whom correspondence should be addressed: Cardiovascular Division, Dept. of Medicine, Brigham and Women's Hospital, 75 Francis St., Boston, MA 02115. controlled and suggest that an early event in the cellular processing of the angiotensin 11-receptor complex is essential to maintenance of DG accumulation.
Angiotensin I1 (ang 11)' has been shown to stimulate phosphoinositide metabolism in cultured vascular smooth muscle (VSMC) (1-5). Although ang 11-stimulated inositol trisphosphate formation is transient (l), diacylglycerol (DG) production is biphasic and sustained ( 3 ) . We have previously demonstrated that the initial, transient phase of DG formation results from polyphosphoinositide hydrolysis, while the sustained phase of DG accumulation results, at least in part, from delayed phospholipase C-mediated hydrolysis of phosphatidylinositol ( 3 ) . In VSMC, the two phases of ang IIinduced DG formation seem to be differentially controlled ( 3 ) . Early phosphatidylinositol 4-phosphate (PIP) and phosphatidylinositol 4,5-bisphosphate (PIP,) hydrolysis, DG and inositol trisphosphate formation, and calcium mobilization are attenuated by activation of the Ca'+/phospholipid-dependent enzyme protein kinase C with phorbol esters or 1oleolyl-2-acetylglycerol, suggesting that endogenous DG formation may attenuate these initial biochemical responses (2, 3 ) . However, the second, sustained phase of DG accumulation is not attenuated by phorbol esters, and the mechanisms underlying the delay in hormonal induction of this second phase remain unknown.
In other membrane receptor systems, binding of an agonist to the receptor initiates movement of the receptor-agonist complex first within the plane of the membrane and subsequently into intracellular compartments (6). The half-time of receptor internalization at 37 "C ranges from 2.2 min for the asialoglycoprotein receptor in hepatoma cells ( 7 ) to 10 min for ang I1 in bovine adrenocortical cells (8). Movement or sequestration of the receptor-complex within the plane of the membrane occurs even more rapidly and is measureable within 30 s for insulin in hepatocytes (9) and within 2 min in ang 11-stimulated adrenal glomerulosa cells (10). These events occur well within the time frame observed for induction of sustained DG accumulation in VSMC ( 3 ) . In vascular smooth muscle, binding of angiotensin has been shown to initiate aggregation and subsequent internalization of the ang IIreceptor complex (ll), but no systematic studies of the kinet-The abbreviations used are: ang 11, angiotensin 11; VSMC, vascular smooth muscle cells; DG, diacylglycerol; PI, phosphatidylinositol; PIP, phosphatidylinositol 4-phosphate; PIP?, phosphatidylinositol 4,5-bisphosphate; HEPES, 4-(2-hydroxyethyl)-l-piperazineethane sulfonic acid PAO, phenylarsine oxide; DTT, dithiothreitol. ics of these events have been published. Furthermore, little consideration has been given to the possibility that these processing events are related to signal generation.
Based on these observations, we have postulated that the temporal delay in the onset of phospholipase C-mediated sustained DG accumulation (>2 min) observed in ang IIstimulated VSMC may be related to some obligatory cellular events possibly involving processing of the agonist-receptor complex. To test this hypothesis, we characterized the kinetics of ang 11-receptor internalization and studied the effects of interventions which interfere with receptor processing on the pattern of ang 11-induced DG formation and PI and polyphosphoinositide hydrolysis in VSMC. Our data indicate that the angiotensin I1 receptor is rapidly converted to a sequestered, acid-resistant form, and that agents which interfere with this process also preferentially inhibit sustained DG accumulation. Thus, there appears to be a close correlation between receptor sequestration and DG accumulation in ang 11-stimulated VSMC, suggesting that cellular processing of the ang I1 receptor-ligand complex is essential to initiating and sustaining accumulation of DG, and therefore possibly to tonic signal generation.

EXPERIMENTAL PROCEDURES
Cell Culture-VSMC were isolated from rat thoracic aorta by enzymatic dissociation as described previously (12, 13). Cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% calf serum, glutamine, and antibiotics and were passaged twice a week by harvesting with trypsin-versene and seeding at a 1:4 ratio in 80-cm2 flasks. For experiments, cells between passage levels 4-25 were seeded into 22-, 35-, or 100-mm dishes (2 X lo4 cells/cm2), fed every other day, and used at confluence (2-6 days).
Phospholipid Labeling and Extraction-Labeling, extraction, and separation of the phospholipids, neutral lipids, and inositol phosphates were performed as described previously (3). Briefly, VSMC cultures were incubated with either [3H]myoinositol (15 pCi/ml) for 24 h or [3H]arachidonic acid (1 pCi/ml) for 3 h. Cells were washed with a warm, balanced salt solution (130 mM NaCI, 5 mM KCl, 1 mM MgCl,, 1.5 mM CaC12, 20 mM HEPES, buffered to pH 7.4 with Tris base) and incubated for 20 min in this same solution at the indicated temperature. Cultures were then maximally stimulated with 100 nM ang I1 for various times. In acid-wash experiments, exposure to ang I1 was followed by a 10-min incubation at 4 "C with 0.05 M acetic acid in 150 mM NaCl (pH = 3.0), thorough washing, and rewarming to 37 "C for 5 min in balanced salt solution. In all cases the reaction was terminated by rapid aspiration of the buffer and addition of 1 ml of chloroform/methanol/HCl (1:2:0.05) for phospholipid extraction or chloroform/methanol (1:2) for neutral lipid extraction. Organic and aqueous phases were separated by addition of chloroform and distilled water, centrifugation, and two chloroform washes. The organic phase was immediately concentrated under nitrogen, and lipids were resolved by thin layer chromatography (3, 14). The aqueous phase was also concentrated under nitrogen, and the inositol phosphates were resolved by the column chromatography method of Downes and Michell (15) as described previously (1). All lipids and inositol phosphates were quantitated by liquid scintillation spectrophotometry.
Measurement of 125Z-Angiotensin II Binding and Internalizatwn-Measurement of surface-bound and internalized T -a n g I1 was performed on replicate-plated 22-mm cultures of attached cells. For experiments utilizing phenylarsine oxide (PAO), cells were preincubated with 0.1-100 pM PA0 prior to addition of Iz5I-ang 11. To determine internalization, samples were incubated for 90-120 min at 4 "C with 0.5 nM lZ5I-ang I1 (2200 Ci/mmol) in a binding buffer consisting of 50 mM Tris, 5 mM MgCl,, 100 mM NaC1, pH 7.4. Unbound radioactivity was removed by washing the cells 4 times with ice-cold binding buffer containing 0.25% bovine serum albumin. Cells were then rewarmed for various times at the indicated temperatures, rapidly cooled, washed four times with ice-cold buffer, and exposed to 0.05 M acetic acid in 150 NaCl for 10 min at 4 "C. This wash, plus two 0.5-ml rinses, was collected and counted as surface-bound radioactivity. In some experiments, the incubation medium was collected and analyzed for radioactivity. The cells were then solubilized with 1% sodium dodecyl sulfate, 0.03% NaOH to determine cell-associated, or internalized, radioactivity. Nonspecific binding of '"I-ang I1 was determined at each time point and was defined as that not displaced by 1 pM unlabeled ang 11. Such binding was routinely less than 5-10% of the total binding.
For binding experiments where surface and sequestered binding were not differentiated, measurement of bound lZ5I-ang I1 was performed using cell suspensions as described previously (12). Briefly, suspensions of cultured VSMC (final concentration IO6 cells/ml) were prepared from replicate-plated 100-mm culture dishes using collagenase, soybean trypsin inhibitor, and BSA (13) and incubated with increasing concentrations of lZ5I-ang I1 for 35 min at 25 "C. The reaction was terminated with cold buffer, and samples were filtered over a Whatman GF/C glass fiber filter and counted in a gamma counter. Nonspecific binding was determined as described above.
Degradation of 1Z5Z-Angiotensin ZI-Degradation of Iz5I-ang I1 was assayed by thin layer chromatography as described previously (12). Aliquots (50 pl) of surface-associated, medium or cell-associated radioactivity extracted as described above were spotted on thin layer cellulose plates (Eastman 12355), using authentic T -a n g I1 as a standard. Plates were developed with tert-butyl alcohol/3% NH3 (105:35) as solvent. Developed chromatograms were cut into 0.75-cm strips and counted in a gamma counter.
Measurement of '%a2+ Content-Cell cultures (35 mm) were equilibrated for 24 h at 37 "C in 2 ml of fresh culture medium containing "CaCl, (4 pCi). On the day of the experiment, ang I1 (10 nM) was added directly to the culture medium. After 5 min, Sarl-Iles-ang I1 (100 nM) was added to some dishes. The reaction was terminated after various intervals by washing the cultures five times with icecold calcium-free balanced salt solution containing 10 mM LaC13. Radioactivity was extracted with 1 ml of 0.1 N HNO,. Cell "Ca" content is expressed as nanomoles per milligram of protein and was calculated from the specific activity of 45Ca2+ in the medium. Protein was determined by the method of Lowry et al. Sari-Ilea-Angiotensin 11, Peninsula Laboratories, Belmont, CA; phenylarsine oxide, Aldrich; collagenase, soybean trypsin inhibitor, and bovine serum albumin, Cooper Biomedical, Malvern, PA.

Ang ZI Receptor Binding and Temperature Dependence of
Internalization-We have previously demonstrated that ang I1 binds to a single class of high affinity binding sites in rat aortic VSMC (2). The binding is temperature-dependent and is saturable at 5 "C in 60 min (17). Using the acid-wash technique described above, we found that after incubation of VSMC cultures with lZ5I-ang at 4 "C for 90-120 min, 97.0 f 0.7% of specifically bound radioactivity was associated with the cell surface (acid-releasable). Rewarming the cells resulted in a rapid, temperature-dependent sequestration of the receptor-agonist complex to an acid-resistant form (Fig. l), with half-times of 1.5 min at 37 "C, 2.8 min at 25 "C, and 7 min at 19 "C. The ang I1 remaining on the cell surface comigrated with authentic ang I1 following thin layer chromatography, indicating an absence of degradation.
Effect of Temperature on Diglyceride Formation and Phosphoinositide Hydrolysis-To provide insight as to whether receptor sequestration is important in ang 11-induced DG formation in VSMC, we studied DG accumulation at three temperatures (37, 25, and 19 "C) which exhibited marked differences in rates of receptor internalization (Fig. l), at 30 "C, and at 4 "C where internalization did not occur. As we have previously demonstrated, DG accumulation at 37 "C is biphasic and sustained (3) (Fig. 2). Reducing the temperature to 30 "C did not alter the time course or amplitude of ang IIinduced DG production (data not shown). A further decrease in temperature to 25 "C caused a marked temporal separation between the early phase of DG formation and development of the second peak of DG accumulation. Reducing the tempera- for 20 min. Cells were then exposed to angiotensin I1 (100 nM) for various periods. Temperature had no effect on basal DG levels (average control, 2191 k 111 cpm). Each point represents the mean f S.E. of duplicate determinations from at least three experiments. ture to 19 "C delayed the early peak of DG formation from 15 s to 1 min and shifted formation of the second DG peak to 20-30 min (Fig. 2). A t 4 "C, DG accumulation was monophasic, peaking at 10 min (Fig. 2) and returning to base line by 80 min (data not shown).
The second or delayed phase of accumulation of DG most likely results from PLC-mediated hydrolysis of several phospholipids (18), of which PI is a major component (3). Preliminary evidence indicates that ang I1 also induces hydrolysis of phosphatidycholine in VSMC which is detectable after about 1 min.* Thus, the second phase of DG accumulation is probably the result of hydrolysis of a combination of phospholipids. However, since we have previously demonstrated that phosphoinositide hydrolysis makes a major contribution to ang 11-induced DG formation in VSMC ( 3 ) , we particularly examined the effect of altering temperature on inositol phospholipid hydrolysis and inositol phosphate formation following ang I1 stimulation. Reducing temperature to 4 "C completely abolished ang 11-induced PI breakdown and inositol monophosphate formation (data not shown). However, hy-K. K. Griendling and R. W. Alexander, unpublished observations. drolysis of the polyphosphoinositides and formation of inositol bisphosphate and inositol trisphosphate at this low temperature followed a time course identical to that seen for DG, suggesting that the DG formed at 4 "C resulted exclusively from PIPz and PIP breakdown. Decreasing temperature to 19 or 25 "C slightly reduced and slowed breakdown of PIP, (Fig.  3B) and PIP (data not shown), but polyphosphoinositide levels still rapidly began to return toward base line, especially at 25 "C. PI hydrolysis was significantly inhibited and delayed at both 19 and 25 "C (Fig. 3A). The PI levels measured here represent a combination of direct hydrolysis by PLC (3), phosphorylation by PI kinase to replenish the polyphosphoinositides (19), and reformation of PI from myoinositol and CDP-DG (19), all of which may have a temperature dependence. Thus, although the decrease in PI cannot be quantitatively related to increased DG formation at temperatures below 37 "C, it is likely significant that temperatures which inhibit DG accumulation also inhibit PI hydrolysis (Figs. 2  and 3A). Furthermore, formation of inositol monophosphate, the other product of PLC-mediated breakdown of PI, is also attenuated at 19 and 25 "C (data not shown).
The fact that early DG formation and phospholipid hydrolysis remain relatively intact during drastic reductions in temperature suggests that the major effect of lowering temperature on the late response is not to decrease the activity of PLC. Rather, the differential effect of temperature on the late DG response correlates well temporally with the effect of temperature on receptor sequestration (Fig. 4). At all temperatures, receptor sequestration precedes or occurs simultaneously with delayed DG formation, and peak DG accumulation occurs only after about 70% of the receptors have become resistant to acid wash. Although the relationship between receptor internalization and sustained DG accumulation could not clearly be analyzed at 37 "C because of overlap between the two phases of DG formation, the clear separation between the early and late peaks of DG accumulation at 25 and 19 "C permitted regression analysis of this relationship. The corre- arachidonic acid (1 pCi/ml, 3 h, 37 "C) or "'I-ang I1 (0.5 nM, 90 min, 4 "C). For measurements of DG accumulation, cells were washed and exposed to ang I1 (100 nM) at the indicated temperatures. For measurement of internalization, cells were washed and rewarmed to the indicated temperatures, and internalized radioactivity was measured as that remaining associated with the cell following acid wash. A, 37 "C; B, 25 "C; C, 19 "C. lation coefficients at 25 and 19 "C were 0.95 and 0.84; respectively, indicating a linear relation between receptor internalization and sustained DG accumulation a t both temperatures.
Effect of Phenylarsine Oxide on Angiotensin II Receptor Internalization and DG Formation-PA0 has been shown to inhibit internalization of a variety of receptors including padrenergic, epidermal growth factor, and insulin receptors (20,21). Sequestration of the ang I1 receptor in VSMC was also effectively inhibited by P A 0 in a dose-dependent manner. The threshold for this inhibition was about 1 p~ PAO, the ICso was about 10 p~ PAO, and maximal inhibition occurred at 100 p~ PAO.
To test further the correlation between receptor sequestration and DG accumulation, we preincubated cells with P A 0 to block internalization and measured ang 11-induced phosphoinositide metabolism. P A 0 selectively inhibited the late (5 min) sustained phase of ang 11-stimulated DG formation in VSMC (Fig. 5 ) . Inhibition was dose-dependent with a threshold of about 0.1 p~ PAO, and a half-maximal effect occurring at about 6 p~ P A 0 (data not shown). The correlation between P A 0 inhibition of receptor sequestration and inhibition of sustained DG formation is linear (Fig. 6). How-

. Relationship of P A 0 inhibition of angiotensin IIinduced DG accumulation and receptor sequestration.
Cultured VSMC were labeled with either [3H]arachidonic acid (1 pCi/ml, 3 h, 37 "C) or lZ5I-ang 11 (0.5 nM, 90 min, 4 "C). For measurements of DG accumulation, cells were washed, exposed to P A 0 (0.1-100 p~) for 5 min, and stimulated with ang I1 (100 nM) for 5 min in the continued presence of PAO. For measurements of internalization, cells were exposed to P A 0 at 37 "C for 5 min prior to lZ5I-ang I1 labeling, and then rewarmed to 37 "C for 5 min in the presence of P A 0 after Iz5Iang I1 labeling. Results are expressed as a percentage of the ang IIinduced DG or internalization response in the absence of PAO. Each point corresponds to a different dose of PAO. ever, DG formation is more sensitive to inhibition by P A 0 than is internalization. To determine whether the ability of P A 0 to inhibit sustained DG accumulation is specifically related to its ability to inhibit receptor sequestration, we incubated VSMC with ang 11 (100 nM) for 5 min a t 37 "C to allow internalization to occur, and then exposed cells to P A 0 (10 p~) in the continued presence of ang 11. As shown in Fig.  7, preincubating cells with P A 0 resulted in an 81 & 6% inhibition of subsequent ang 11-induced DG accumulation (Fig. 7A), but the same duration of exposure to P A 0 inhibited DG formation by only 29 * 3% after the receptor was internalized (Fig. 7B). Thus, the 35% rightward shift of the relationship between P A 0 inhibition of internalization and DG accumulation depicted in Fig. 6 most likely reflects a nonspecific effect of P A 0 on DG accumulation. PA0 has also been shown to have other effects on cell function, such as reducing binding to certain receptors (20). However, in VSMC, P A 0 had no effect on either ang I1 receptor number or affinity as determined by Scatchard analysis (data not shown). Nonspecific and specific effects of P A 0 with respect to internalization can be experimentally separated since the former can be abolished by thorough washing, while the specific effects of P A 0 on internalization can be reversed using bifunctional sulfhydryl compounds such as dithiothreitol (DTT) (20). In ang 11-stimulated VSMC, the inhibitory effect of PA0 on DG was not reversed by extensive washing and removal of PAO. D T T (1 mM, 5 min), however, partially reversed the PA0 inhibition of DG accumulation (PAO: 86% inhibition; P A 0 + DTT: 36% inhibition). This reversal was likely incomplete because DTT (5 mM) alone causes a 45% decrease in lZ5I-ang I1 binding to VSMC (19). These observations suggest that the major effect of P A 0 on sustained DG accumulation is specifically related to its ability to inhibit receptor internalization.
Because all concentrations of P A 0 tested had some effect on basal levels of DG, 10 p~ P A 0 was routinely used for phospholipid experiments to insure near-maximal inhibition of DG formation with minimal base-line changes. A 5-min incubation with P A 0 (10 p~) caused a 24 f 5% increase in basal DG and decreased base-line PIP to 74 f 9% control. Other basal phospholipid levels were not significantly altered. As expected from its effects on ang 11-stimulated DG accumulation, P A 0 selectively inhibited late, but not early, changes in phospholipid metabolism (Table I). Ang 11-induced PI breakdown (5 min) was significantly inhibited, as was phosphatidic acid formation, suggesting that the reduction in DG accumulation by P A 0 occurred by inhibition of PLCmediated PI breakdown, rather than by acceleration of DG phosphorylation.
Effect of Other Inhibitors of Endocytosis-The acid-wpsh technique for measuring internalization does not distinguish among the various steps in the endocytotic pathway; that is, the receptor-agonist complex may become inaccessible to acid-wash by some conformational change, movement within the plane of the membrane, or movement into some intracellular compartment. T o further pinpoint the site of the inter- for 3 h or [3H]myoinositol (15 pCi/ml) for 24 h, and were then either exposed to ang I1 (100 nM) alone or preincubated with PA0 (10 p~, 5 min) and exposed to ang I1 in the continued presence of PAO.
Values are expressed as the mean f S.E. of duplicate determinations from at least four experiments. Each condition was calculated as a percentage of its own control. PA, phosphatidic acid. p~ chloroquin had no effect on subsequent ang 11-induced DG accumulation. These observations suggest that the degradative portion of the endocytotic pathway is not involved in sustained hormonally induced phosphoinositide metabolism.
Contribution of Retroendocytosis-In other systems, particularly insulin recycling in adipocytes, it has been proposed that a portion (25%) of the agonist-receptor complex recycles back to the plasma membrane without dissociating by a process which is chloroquin-insensitive (22). This retroendocytotic movement of the agonist-receptor complex to a different domain on the cell surface could conceivably be important to subsequent signal generation. To determine whether such a pathway is present for the ang I1 receptor in VSMC, we incubated cells with '2'I-ang I1 for 90 min at 4 "C, washed off unbound ang 11, rewarmed the cells to 37 "C for 5 min to allow internalization, and removed surface-bound ang I1 by acid wash. Samples were subsequently rewarmed and the distribution of radioactivity between surface, medium, and cell was monitored (Fig. 8). Over a period of 30 min, 60% of the radioactivity originally associated with the cell remained inaccessible to acid wash, 34% appeared in the medium, and 6% became surface-associated. Both the cell-associated radioactivity and the radioactivity in the medium were degraded ang I1 as assessed by thin layer chromatography, suggesting that this radioactivity resulted from cycling through the lysosomal degradative pathway. Thus, it appears that retroendocytosis does not represent a major pathway contributing to signal generation in these cells.
Functional Significance of Sequestered Receptor Compkx-To determine whether the ang I1 receptor-ligand complex remains capable of generating an intracellular signal following its conversion to an acid-resistant form, we measured DG accumulation and 45CaZ' content after removal of the readily dissociable surface signal.
Two types of experiments were performed. In the first experiment, we exposed cells to ang I1 for 5 min at 37 "C to permit development of the acid-resistant form of the ang 11-receptor complex, and surface-bound ang I1 was removed by acid wash. Subsequent rewarming of the samples for 5 min at 37 "C resulted in only a 32 f 14% ( n = 5) decrease in DG accumulation compared with cells not exposed to acid. In a second set of experiments, we used Sari-

Receptor Processing and
Diacylglycerol Formation in VSMC  FIG. 8. Fate of angiotensin 11-receptor complex. Cultured VSMC were incubated with '"I-ang I1 (0.5 nM) for 90-120 min at 4 "C, washed, and rewarmed to 37 "C for 5 min to permit sequestration of the receptor. Radioactivity was then stripped from the surface by acid wash at 4 "C, and samples were rewarmed to 37 "C for the indicated times. Incubation medium was then counted, and surfacebound and cell-associated radioactivity were separated by acid wash and counted.
Iles-ang 11, a potent ang 11 antagonist, to compete off surfacebound ang 11, and measured the temporal decline in DG levels and increase in total cell calcium content (Fig. 9). Addition of Sar'-Ilea-ang I1 (IO p~) prior to ang II(100 nM) completely abolished subsequent DG formation (data not shown). Addition of Sar'-Ile*-ang I1 to VSMC after a 5-min incubation with ang 11 (100 nM) only slowly reversed the accumulation of DG, requiring at least 15 min for DG levels to return to baseline (Fig. 9, top). In both acid-wash and antagonist experiments, following removal of the surface signal, there is significant residual DG long after the time necessary for metabolism of previously formed DG (see time course of early DG peak, Fig. 2) Another manifestation of the cell response to ang 11, a decrease in 45Ca2+ content, which results from initial release of intracellular calcium and stimulation of 45Ca2+ efflux, is depressed for up to 2 h in the continued presence of the hormone (Fig. 9, bottom). This persistent decrease in 45Ca2+ content reflects an inability to refill intracellular calcium stores in the presence of ang 11. Upon addition of antagonist, there is a lag time of 15 min in the return of 45Ca2+ levels to base line (Fig. 9, bottom). The gradual reversal of changes in DG and calcium reflect a gradual decrease in signal generation by a receptor that, by virtue of its sequestered state, is only slowly accessible to the antagonist.

DISCUSSION
The results of the present study confirm and extend our previous observations which suggested that the two phases of DG formation measured after stimulation of VSMC with ang I1 are biochemically distinct, with different sources and mechanisms of control (3). The two phases can be clearly separated by step-wise reduction in temperature. Furthermore, P A 0 markedly inhibits sustained DG accumulation while only minimally affecting the early DG peak. The effects of these two interventions on the sustained phase of DG accumulation correlate closely with their effects on conversion of the ang I1 receptor-ligand complex to an acid-resistant form, suggesting that cellular processing of the ang I1 receptor is essential to sustained DG accumulation.
There have been relatively few studies concerning ang I1 receptor internalization. A recent report has described movement of the ang I1 receptor in adrenal glomerulosa cells from the cell surface to the lysosome within 20 min of lZ5I-ang I1 binding (10). By 5 min, the majority of bound ang I1 was located in coated pits and coated vesicles (10). In our cells, [3H]arachidonic acid (1 pCi/ml) for 3 h or "Ca" for 24 h, and then exposed to ang 11. After 5 min, Sar'-Ilea-ang I1 was added to half the samples. the half-time for development of an acid-resistant form of the agonist-receptor complex a t 37 "C was 1.5 min, suggesting that development of acid resistance occurs during movement of the receptor on the cell surface but prior to removal of the receptor-agonist complex from the plasma membrane.
The close correlation between the rate of development of acid resistance of the ang I1 receptor-ligand complex and the development of the late phase of DG accumulation at various temperatures suggests that the two phenomena are related. This concept embraces two possibilities: either DG accumulation plays a role in the initiation of internalization or receptor sequestration/movement is essential to the development of sustained DG accumulation. In other systems, phor-bo1 ester has been shown to stimulate receptor internalization (23). In VSMC, however, 4P-phorbol 12-myristate 13-acetate (100 nM), an exogenous activator of protein kinase C, had no effect on internalization at 37 " C 3 This suggests that DG does not initiate internalization via a protein kinase C-dependent mechanism, and makes the second possibility considered above more likely.
The actual relationship of receptor sequestration to development of sustained DG formation encompasses several possible mechanisms: different signaling domains on the cell K. K. Griendling, L. S. Ekstein, and R. W. Alexander, unpublished observations. surface, movement of the receptor to an intracellular compartment from which signal generation then occurs, or movement of the receptor or substrate to the cell surface following internalization, resulting in secondary signaling. Signaling from an intracellular compartment has been suggested by the work of Richards et al. (24), who demonstrated direct breakdown of PI by PLC and accumulation of DG in rat liver lysosomes. Additionally, based on his observation that 80% of acetylcholine-stimulated 32P incorporation into PI in pigeon pancreatic slices occurs in the endoplasmic reticulum, Hokin (25) has suggested that the agonist-receptor complex may move through various endocytotic vesicles to a PI-rich source, or alternatively, initiate transfer of PI from endoplasmic reticulum to the plasma membrane. However, in VSMC, the inability of chloroquin and monensin, two wellcharacterized lysosomotrophic amines, to inhibit DG formation suggests that the lysosomal degradative pathway has no role in the coupling of receptor activation to the phosphoinositide response. Furthermore, there appears to be very little retroendocytosis in VSMC stimulated with angiotensin I1 (Fig. 8). Thus, in VSMC, the receptor-processing event integral to sustained DG formation does not appear to be related to either the lysosomal degradative or the retroendocytotic pathway.
Although reducing temperature has effects on cell function other than receptor processing, our experiments with low temperature provide evidence suggestive of distinct signaling domains on the cell surface. One important effect of reducing temperature on cell metabolism is to decrease enzyme activity. However, this does not appear to be the explanation for alteration in sustained DG formation. PLC activity, as evidenced by the early DG response at the various temperatures (19-37 "C, Fig. 2), remains intact. The apparent enhancement of the early DG peak at 19 "C may be related to removal of the inhibitory effect of protein kinase C stimulation normally resulting from sustained DG formation (2, 3). Phosphokinase activity probably does decrease, and this may explain the inability of the cell to replenish PIP2 at the lower temperatures (Fig. 3). However, since DG formation at 30 "C, a temperature which should cause a major change in enzyme activity, is virtually identical to DG formation at 37 "C, it is unlikely that changes in enzyme activity are solely responsible for alterations in the late DG response. It seems more likely that low temperature is altering membrane fluidity or fusion. In other cell systems, a reduction in temperature to 19 "C has been shown to inhibit lysosomal-endosomal fusion (6,26). However, as noted above, the lysosomal degradative pathway does not appear to be involved in sustained DG signaling, suggesting that the pertinent effects of temperature may be related to membrane fluidity rather than fusion. In reconstituted erythrocyte membranes, a sharp decrease in the velocity of concanavalin A-conjugated gold microsphere movement within the bilayer occurred in the temperature range 17-28 "C (27). This is the same range of temperature which markedly delayed late DG accumulation and development of an acidresistant form of the agonist-receptor complex in the present study. Taken together, these observations suggest that the events in the endocytotic pathway important to sustained DG accumulation most likely are localized to the plasma membrane.
Additional evidence for distinct signaling domains at the level of the plasma membrane comes from our experiments with P A 0 and potassium depletion (28). The mechanism of action of PA0 is unknown, but PA0 has been shown to prevent the appearance of receptors in a light vesicular fraction (20), suggesting that its site of action is early in the internalization pathway. Our own experiments with cell fractionation have shown that PA0 (100 PM) completely abolishes the ang 11-induced fusion of different vesicular fractions of VSMC and prevents the appearance of lZ5I-ang I1 in the light vesicular fra~tion.~ This suggests that PA0 exerts its effect at the level of the plasma membrane. Moreover, in related studies, we have shown that potassium depletion selectively inhibits sustained DG accumulation and concomitantly inhibits receptor internalization (28). The effects of PA0 and potassium depletion on DG accumulation are strikingly similar. In human fibroblasts, potassium depletion has been shown to decrease endocytosis of receptor-bound low density lipoprotein (29), and in particular, to decrease coated-pit formation (29) and assembly of the clathrin lattice (30). These observations suggest the intriguing possibility that the coated pit plays an important role in DG formation. Campbell et al. (31) have demonstrated that coated vesicles isolated from bovine brain are enriched in active PI kinase and concluded that coated vesicles may be involved in cellular PI metabolism. Our data suggest the additional possibility that the coated pit/vesicle, or some other distinct domain on the cell surface, may be important in hormonal signaling.
In summary, we have provided evidence that upon ang I1 stimulation of VSMC, the angiotensin I1 receptor is rapidly converted to a sequestered acid-resistant form, and that development of this state is essential to sustained accumulation of DG. Low temperature and phenylarsine oxide, both of which inhibit ang I1 receptor internalization, preferentially interfere with the late phase of DG accumulation, while having little effect on the initial transient phase of DG formation from polyphosphoinositide breakdown. The processing event important to hormonally induced sustained DG accumulation appears to occur very early in the internalization pathway, almost certainly at the level of the plasma membrane.