INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE ARG-GLY-ASP (RGD) motif is a common motif found in many extracellular matrix proteins that bind to integrins. RGD-containing peptides induce vasodilatation in cremaster muscle arterioles by lowering the intracellular Ca²⁺ concentration ([Ca²⁺]_i; see Ref. 8). L-type Ca²⁺ channels (24), K⁺ channels (18), and _v3-integrins were suggested to be involved (15). However, ₅₁-integrins mediate an increase of the Ca²⁺ current in smooth muscle cells isolated from the same vascular bed (24). In rat renal afferent arterioles, RGD-containing peptides induce vasoconstriction rather than dilatation. The constriction is associated with a pronounced increase in the [Ca²⁺]_i in the smooth muscle cells as measured by confocal fluorescence microscopy (27). The [Ca²⁺]_i in renal vascular smooth muscle is the major determinant in the myogenic mechanism of renal autoregulation (19), in which renal arterioles constrict when the transmural pressure is increased. Extracellular mechanical stimuli can be transduced into the cytosol via interactions between the cytoskeleton and the cytoplasmic domain of integrins (23). It is hypothesized that variations of transmural pressure in renal arterioles will alter the interactions between integrins and the extracellular matrix, which contributes to the mechanotransduction in renal autoregulation (20, 27). It has been demonstrated that the myogenic component in renal autoregulation is oscillating at 0.1-0.2 Hz (4, 25). These oscillations can be driven by a temporally and spatially coordinated release of intracellular Ca²⁺ in the form of Ca²⁺ wave or [Ca²⁺]_i oscillation in afferent arteriolar vascular smooth muscle. [Ca²⁺]_i oscillation can be considered as a recurrence of fast-propagating intracellular Ca²⁺ waves. We hypothesized that occupancy of integrins would trigger spatial and temporal variations of [Ca²⁺]_i in afferent arteriolar vascular smooth muscle cells. The present study was performed to test these hypotheses in cultured renal vascular smooth muscle cells using confocal fluorescence microscopy.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Isolation and culture of rat renal preglomerular smooth muscle cells. Renal preglomerular smooth muscle cells were isolated from Sprague-Dawley rats by an iron oxide sieving method (10, 13, 29). Briefly, rats were anesthetized with halothane and laparotomized, and the kidneys were exposed. The abdominal aorta distal to the renal arteries was cannulated, and 1% iron oxide (Fe₃O₄) in calcium-free Hanks' balanced salts solution (HBSS) was perfused into the kidney. The kidneys were then removed, and the cortex was harvested. The tissue was minced and transferred to HBSS. Iron oxide-containing tissue was isolated from the suspension with a side-pull magnet (Perseptive Diagnostics) and was resuspended in HBSS. The suspension was passed through needles of decreasing size (18, 20, and 23 G) and was filtered on a 200-mesh sieve (Sigma). The iron oxide-containing tissue on the sieve was then digested with collagenase (type 1, 2 mg/ml; Worthington) in HBSS (with calcium chloride) at 37°C with gentle shaking for 30 min. The remaining iron oxide-containing tissue was removed by a magnetic method, and smooth muscle cells in the supernatant were centrifuged down and resuspended in 10 ml of DMEM with 10% FBS. The cells were seeded onto collagen I-coated T25 culture flasks (Becton-Dickinson) and incubated at 37°C in 5% CO₂ and 95% air at 98% humidity. The cells were allowed to seed for 2 days before the medium was changed. The medium was changed every 2-3 days until the cells had grown to confluence. The cells were passed approximately every 5 days and were passed onto collagen I-coated coverslips (Becton-Dickinson) during the sixth passage. Cells were grown close to confluence for the immunohistochemistry and Ca²⁺ measurement studies. Cells from at least three different preparations (rats) were studied in each experimental protocol.

Measurement of [Ca²+]_i in cultured renal smooth muscle cells. The [Ca²⁺]_i of cultured renal preglomerular smooth muscle cells grown on coverslips was measured with fluo 3 from the fluorescence images acquired with an MRC-1000 (Bio-Rad) confocal scanning unit, which was mounted on a Zeiss Axiovert 100TV inverted microscope. Renal preglomerular smooth muscle cells at sixth passage were grown on collagen I-coated coverslips for 3-4 days and used for measurement before they become fully confluent. The cells on the coverslips were loaded with 2 µM of fluo 3-AM (Molecular Probes) in DMEM for 20 min at 37°C. The excess dye was washed away by HBSS (with calcium) and incubated in the same buffer for 15 min. The coverslip was then inserted in the bottom of a perfusion chamber (Vestavia) and mounted on the inverted microscope. The changes in the smooth muscle cell [Ca²⁺]_i were measured at room temperature. Fluorescence was excited with the 488-nm line of the krypton-argon laser. Emission was collected through a band-pass filter 522/32 nm at 1 Hz and stored digitally. All fluorescence images were acquired with a Zeiss plan-apochromat objective [63 × at numeric aperture (NA) 1.4 or 40 × at NA 1.2]. Residence time of the laser on the coverslip is 0.29 s. The changes of [Ca²⁺]_i in each cell after exposure to 0.1 mM of either hexapeptide GRGDSP (Gly-Arg-Gly-Asp-Ser-Pro) or GRGESP (Gly-Arg-Gly-Glu-Ser-Pro) were obtained during the retrospective analysis of the stored fluorescence images with software (Time Course/Ratiometeric Software Module) supplied by Bio-Rad. A testing dose of 0.1 mM of the peptides was chosen so that the increase in fluorescence intensity does not saturate the detector during the image acquisition. A similar dosage was used in smooth muscle cells isolated from cremaster muscle arterioles (24). RGD peptides were administered in the perfusion chamber as a bolus in most studies. RGD peptides were present in the chamber throughout the recording period. When RGD was required to be applied locally on the coverslip, a micropipette (5-10 µm diameter) attached to a microperfusion pump was used (Hampel).

5

Immunofluorescence of -integrin subunits in cultured smooth muscle cells. To determine whether incubation of functional blocking anti-integrin antibodies induces any redistribution of integrin in the smooth muscle cells, the immunofluorescence of ₁ and ₃ was compared when cells were fixed by using two different fixation protocols. First, preglomerular smooth muscle cells grown on collagen I-coated coverslips were fixed in 2% paraformaldehyde for 10 min. The coverslips were blocked with 20% donkey serum in PBS for 60 min and then incubated with a mixture of anti-smooth muscle -actin antibodies and antibodies specific to ₁- or ₃-integrin subunits for double-labeling studies. After 2 h of incubation, the sections were washed three times with fresh PBS followed by incubation with the appropriate Cy3- or Cy5-conjugated secondary antibodies for 60 min. Cy3 (indocarbocyanine, absorption peak 550, emission peak 570) and Cy5 (indodicarbocyanine, absorption peak 650, emission peak 670) are cyanine-based fluorophores, which are brighter and more photostable than the widely used fluorescein-based fluorophores. All incubations were carried out in a moistened chamber at room temperature. The coverslips were then rinsed with fresh PBS three times, mounted in Gel/Mount (Fisher), and examined with a confocal microscope. Cultured cells incubated with anti-₂-integrin antibodies only were used as negative controls. All antibodies were diluted with PBS containing 20% donkey serum.

Antibodies. Mouse monoclonal IgM antibody to ₃ was purchased from Transduction Laboratories (Lexington, KY). Functional blocking antibodies to ₁ (hamster, monoclonal IgM, FITC conjugate or nonconjugated)-, ₂ (mouse, monoclonal IgG)-, and ₃-integrins (mouse, monoclonal IgG, nonconjugated) were purchased from PharMingen (San Diego, CA). Cy3-conjugated and nonconjugated mouse monoclonal IgG to smooth muscle -actin were purchased from Sigma Chemical (St. Louis, MO). Dilution for all primary antibodies was determined before the experiment. All secondary conjugated antibodies purchased have been solid-phase adsorbed to optimize the signal for the multiple labeling study. Secondary antibodies were diluted at 1:400 for all studies (Jackson Immunoresearch Laboratory).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effect of RGD-containing peptide on [Ca²+]_i. Exposing renal smooth muscle cells loaded with fluo 3 to 0.1 mM hexapeptide GRGDSP in HBSS elicited an immediate increase of fluorescence emission intensity, indicating an elevation of smooth muscle [Ca²⁺]_i. The increase of mean [Ca²⁺]_i reached a peak value of 287.0 ± 37.2 nM from a baseline of 91.4 ± 3.9 nM within 8 s and then returned to baseline within 20 s (P < 0.05, 34 cells/5 coverslips; Fig. 1). There were two types of responses in terms of the change in [Ca²⁺]_i profile in the individual cells. Eighty percent of the cells examined displayed a rapid increase of [Ca²⁺]_i and then simply returned to the baseline level similar to the profile shown in Fig. 1. In 20% of the cells studied, there were recurrences of smaller-amplitude Ca²⁺ waves after the initial response to the RGD peptide. The propagation of the Ca²⁺ wave could be visualized by the time delays between the [Ca²⁺]_i increases in the different regions of the cell (Fig. 2). The mean propagation velocity was 24.4 ± 1.7 µm/s (n = 6). As a result of these spontaneous recurrent Ca²⁺ waves, the mean [Ca²⁺]_i of the whole cell was set to oscillate. Hexapeptide GRGESP (0.1 mM) in HBSS or HBSS alone did not induce any change in [Ca²⁺]_i (Fig. 1). This confirmed our previous observation from isolated afferent arterioles that the increase of smooth muscle [Ca²⁺]_i is triggered by the RGD motif.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Time course of change in intracellular Ca2+ concentration ([Ca2+]i) when cells were exposed to either 104 M of GRGDSP or 104 M of GRGESP. The peptide was added to the perfusion chamber at time 0. Dotted lines are SE. , Increase of [Ca2+]i is significant (P < 0.05) compared with the baseline.

View larger version (26K): [in this window] [in a new window]   Fig. 2.   Recurrence of spontaneous Ca2+ waves induced by 104 M of GRGDSP. Fluorescence intensity of fluo 3 is measured simultaneously in 6 regions across a single smooth muscle cell. Each sampling region is a 12 µm × 12 µm square. The Ca2+ wave propagates from region 1 to region 6. The peptide is added to the perfusion chamber at time 0 as a bolus. The Ca2+ wave is propagating at 24 µm/s.

View larger version (77K): [in this window] [in a new window]   Fig. 3.   Propagation of Ca2+ wave in a single renal vascular smooth muscle cell. A: projection of 10 images collected before the appearance of Ca2+ wave. Arrow indicates the nucleus. B-L: consecutive images collected at 1 Hz to display the propagation of calcium wave. GRGDSP (0.1 mM) was applied locally on the coverslip with a micropipette.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Reversal of the nucleocytoplasmic Ca2+ gradient when a Ca2+ wave passes through the nucleus in the vascular smooth muscle cell shown in Fig. 3. Fluorescence intensity of fluo 3 is measured simultaneously in three regions, a cytoplasmic region anterior to the nucleus, a nuclear region, and a cytoplasmic region posterior to the nucleus. The area of each region is 12 µm × 12 µm. The Ca2+ wave is propagating at a velocity of 27 µm/s.

Effects of nifedipine and external Ca²+ depletion on [Ca²⁺]_i response. To determine whether the increase of [Ca²⁺]_i triggered by RGD-containing peptides in renal vascular smooth muscle cells is due to an influx of extracellular Ca²⁺ through voltage-gated dihydropyridine-sensitive Ca²⁺ channels, the effects of nifedipine on the [Ca²⁺]_i were determined. Smooth muscle cells loaded with fluo 3 were first exposed to 0.1 mM of GRGDSP to confirm that these cells were responding to RGD motifs. The smooth muscle cells were then incubated with nifedipine (1 µM) for 5 min. There was no change in the fluo 3 emission intensity when the cells were exposed to 0.3 M KCl in HBSS, indicating that dihydropyridine-sensitive Ca²⁺ channels were successfully blocked by nifedipine (data not shown). Incubation with nifedipine did not inhibit the increase of [Ca²⁺]_i induced by GRGDSP or the induction and propagation of Ca²⁺ waves. The peaks of normalized fluo 3 emission intensity before and after the treatment with nifedipine were 2.04 ± 0.13 and 1.73 ± 0.23, respectively (29 cells/3 coverslips, Fig. 5A).

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Normalized time course of the change in fluo 3 fluorescence intensity when cells were exposed to 104 M GRGDSP after incubation with 1 µM nifedipine, 29 cells/3 coverslip (A) or in the absence of extracellular Ca2+, 24 cells/3 coverslips (B). The peptide was added to the perfusion chamber at time 0. Dotted lines are SE. , Fluorescence intensity is significantly different (P < 0.05) from the baseline.

Effects of IP₃ receptor blocker and ryanodine on [Ca²+]_i response in the absence of extracellular Ca²⁺. Smooth muscle cells loaded with fluo 3 were first exposed to 0.1 mM of GRGDSP to confirm that these cells were responding to RGD motifs. The smooth muscle cells were incubated with 50 µM xestospongin C for 30 min in the absence of extracellular Ca²⁺. There was a small transient increase of fluo 3 emission during the first few minutes of incubation, which was most likely due to the inhibiting effect of xestospongin C in endoplasmic reticulum Ca²⁺ pumps at this concentration (9). There was no change in fluo 3 emission intensity when the cells were exposed to 1 µM of phenylephrine after xestospongin C incubation, which indicated that the IP₃ receptors were blocked. However, 0.1 mM of GRGDSP could still induce mobilization of intracellular Ca²⁺ as reflected in an increase of fluo 3 emission (Fig. 6A). The peaks of normalized fluo 3 emission intensity before and after the incubation of xestospongin C were 1.76 ± 0.06 and 1.52 ± 0.08, respectively (58 cells/4 coverslips). On the contrary, preincubation of 20 µM of ryanodine for 20 min totally abolished the Ca²⁺ mobilization effect of 0.1 mM of GRGDSP (Fig. 6B). Phenylephrine (1 µM) could induce an increase of fluo 3 emission after the pretreatment of ryanodine. These observations suggest that the Ca²⁺ mobilization induced by RGD-containing peptides involves the ryanodine-sensitive Ca²⁺ stores but not the IP₃-sensitive Ca²⁺ stores.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Normalized time course of the change in fluo 3 fluorescence intensity when cells were exposed to 104 M GRGDSP after incubation with 50 µM xestospongin C, 58 cells/4 coverslip (A) or after the incubation of 20 µM ryanodine, 106 cells/6 coverslips (B). The peptide was added to the perfusion chamber at time 0. Dotted lines are SE.  and , Fluorescence intensity is significantly different (P < 0.05) from the baseline.

Effects of anti-integrin antibodies on the [Ca²+]_i of smooth muscle cells. To determine whether the Ca²⁺ mobilization induced by RGD peptides is mediated by integrins (15), cells loaded with fluo 3 were first exposed to 0.1 mM of GRGDSP to confirm that these cells were responsive to the RGD motifs. The cells were then incubated with functional blocking anti-₁-, anti-₂-, or anti-₃-integrin antibodies for 20 min. Preincubation of the cells with either anti-₁- or anti-₃-integrin antibodies completely inhibited the elevation of [Ca²⁺]_i induced by GRGDSP (Fig. 7). Anti-₂-integrin antibodies did not inhibit the increase of [Ca²⁺]_i induced by GRGDSP (Fig. 8). The latter observation was expected because ₂-integrins are exclusively expressed in leukocytes (21). These observations suggest that the increase of [Ca²⁺]_i triggered by RGD motifs is mediated by specific binding of RGD-containing peptides to functional -integrin and that the interactions are mediated by more than one class of integrin heterodimers. To test whether the ligation of ₁- and ₃-integrins by the functional blocking antibodies can trigger Ca²⁺ mobilization, the changes in emission intensity of fluo 3 were monitored when smooth muscle cells were exposed to these antibodies. Both anti-₁- and anti-₃-integrin antibodies could induce an increase in [Ca²⁺]_i in smooth muscle cells when applied separately (Fig. 9). During multiple exposure of the smooth muscle cells to the same antibody, an increase of [Ca²⁺]_i was only detected in the first exposure. During sequential application of these two antibodies into the smooth muscle cells, each antibody could induce a rise in [Ca²⁺]_i no matter which antibody was administered first. These observations indicate that these two antibodies ligate two different classes of integrin and that the ligation of either ₁- or ₃-integrin alone can mobilize intracellular Ca²⁺.

View larger version (22K): [in this window] [in a new window]   Fig. 7.   Normalized time course of change in fluo 3 fluorescence intensity when cells were exposed to 104 M GRGDSP after incubation with functional blocking anti-1 integrin antibody, 24 cells/4 coverslips (A) or after incubation of functional blocking anti-3-integrin antibody, 27 cells/4 coverslips (B). , Fluorescence intensity is significantly different (P < 0.05) from baseline.

View larger version (16K): [in this window] [in a new window]   Fig. 8.   Normalized time course of change in fluo 3 fluorescence intensity when cells were exposed to 104 M GRGDSP before (A) and after (B) incubation with functional blocking anti-2-integrin antibodies (41 cells/3 coverslips). The peptide was added to the perfusion chamber at time 0. Dotted lines are SE.  and , Fluorescence intensity is significantly different (P < 0.05) from the baseline.

View larger version (26K): [in this window] [in a new window]   Fig. 9.   Normalized time course of change in fluorescence intensity when cells were exposed to anti-1-integrin antibodies (30 cells/3 coverslips) and anti-3-integrin antibodies (27 cells/3 coverslips). The concentration of antibodies used was 50 µg/ml. The peptide was added to the perfusion chamber at time 0. Dotted lines are SE.  and , Fluorescence intensity is significantly different (P < 0.05) from the baseline.

Immunofluorescence of integrin -subunits. Figure 10 shows the immunofluorescence after double labeling for ₁- and -actin. ₁-Integrin was found mainly along the stress fibers and on the periphery of nuclei. The staining for -actin confirmed that the cells were smooth muscle cells. When functional-blocking anti-₁-integrin antibodies were incubated for 20 min with smooth muscle cells before fixation, the immunofluorescence signal of ₁-integrin along the stress fibres disappeared, and only the signal on the periphery of nuclei remained (Fig. 10, C and D). These observations suggest that incubation with functional-blocking anti-integrin antibodies in living cells induces a redistribution of integrins from the plasma membrane. ₃-Integrin showed a fibrillar distribution with more signal near the edge of the cells and the periphery of nuclei in controls (Fig. 11A). A redistribution of ₃-integrin was also found when functional blocking antibodies of ₃-integrin were incubated with the cultured cells before the cell fixation (Fig. 11C). Treatment of functional blocking anti-₁- or anti-₃-integrin antibodies before paraformaldehyde fixation did not result in cell detachment. In the controls, no immunofluorescence of ₂ was detected in the vascular smooth muscle cells as expected (image not shown).

View larger version (135K): [in this window] [in a new window]   Fig. 10.   Colocalization of 1-integrin (A and C) and smooth muscle -actin (B and D) in preglomerular smooth muscle cells. Cellular fixation was performed before (A and B) and after (C and D) incubation with functional-blocking antibodies. In C and D, cells were incubated with functional blocking anti-1-antibodies (hamster monoclonal IgG) for 20 min before fixation. Anti-1-integrin antibody was conjugated to FITC. Anti--actin antibody was conjugated to Cy3.

View larger version (141K): [in this window] [in a new window]   Fig. 11.   Colocalization of 3-integrin (A and C) and smooth muscle -actin (B and D) in preglomerular smooth muscle cells. Cellular fixation was performed before (A and B) and after (C and D) incubation with functional blocking antibodies. In C and D, cells were incubated with functional blocking anti-3-antibodies (mouse monoclonal IgG) for 20 min before fixation. A mouse monoclonal IgM anti-3-integrin antibody was used to label 3-integrin after fixation. Anti--actin-antibody (mouse monoclonal IgG) was conjugated to Cy3. 3-Integrin was visualized with Cy5-conjugated secondary antibody.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The utility of isolated afferent arterioles for studying the subcellular variations of [Ca²⁺]_i is limited by the geometry of blood vessels, where the individual smooth muscle cell wraps around the vessel (26). Refined methodology is now available for isolating vascular smooth muscle cells from preglomerular resistance vessels for acute studies (10, 29). Because of the enzymatic digestion process used in the isolation of the smooth muscle cells, the integrity of integrins on freshly isolated smooth muscle cells might be compromised. We took an alternative approach by culturing preglomerular smooth muscle cells on coverslips so that the subcellular spatial and temporal variations in intracellular Ca²⁺ could be monitored by confocal fluorescence microscopy. We first demonstrated that exposing the cells to the hexapeptide GRGDSP resulted in an increase of [Ca²⁺]_i. In contrast, no response was observed when the hexapeptide GRGESP was used. These observations are in agreement with previous studies performed in isolated afferent arterioles (27). The baseline [Ca²⁺]_i in the cultured smooth muscle cells is 91 ± 4 nM, which is consistent with reports from another laboratory (29). GRGDSP (0.1 mM) increased the [Ca²⁺]_i to 287 ± 37 nM within 8 s, and the values were restored to the baseline within 20 s. The time-course profile of the [Ca²⁺]_i increase is similar to that in isolated afferent arterioles (27). These observations suggest that the machinery required for RGD-dependent mobilization of intracellular Ca²⁺ in intact arteriolar smooth muscle is also present in the primary cultures.

Local application of RGD peptides on the coverslip triggered propagation of Ca²⁺ waves traveling across the smooth muscle cell (Figs. 3 and 4). Because wave propagation did not depend on extracellular Ca²⁺, it suggests that the propagation is due to a regenerative release of Ca²⁺ from intracellular stores (22, 28). The propagation velocity of intracellular Ca²⁺ waves in the present study is in the range of 24-27 µm/s, which is faster than those observed in human vein smooth muscle cells (7.5 µm/s, induced by histamine; see Ref. 16), in the smooth muscle cell line A7r5 (16 µm/s, induced by vasopressin; see Ref. 1), and in human bronchial smooth muscle cells (18 µm/s, induced by arachidonic acid; see Ref. 6). This is the first report of integrin-mediated Ca²⁺ waves in smooth muscle cells. No other comparable data are available in the literature. The baseline fluorescence signal in the nucleus is in general lower than that of the cytoplasm, which indicates the existence of a nucleocytoplasmic Ca²⁺ gradient. The gradient was reversed when the cells were stimulated with RGD peptides. Similar observations of reversing the nucleocytoplasmic Ca²⁺ gradient were also reported in a smooth muscle cell line when stimulated by pharmacological agonists (12). Because the increase of nuclear [Ca²⁺]_i was initiated when the cytoplasmic Ca²⁺ wave reached the nucleus, it is likely that the increase of nuclear [Ca²⁺]_i is also the result of Ca²⁺-induced Ca²⁺ release. It is known that a mechanical signal can be transduced from the plasma membrane to the nucleus via the integrins and the cytoskeleton (14). If ligand binding on integrins had triggered a nuclear [Ca²⁺]_i increase directly, the rise of nuclear [Ca²⁺]_i would most likely have preceded the arrival of the cytoplasmic Ca²⁺ wave. The propagation of integrin-initiated Ca²⁺ waves was not limited to an individual cell. In some cases, it was observed that the wave propagated into the adjacent smooth muscle cells. The recurrent Ca²⁺ waves resulted in local oscillation of [Ca²⁺]_i. The coupling of smooth muscles cells by Ca²⁺ oscillations might allow a population of cells to operate in unison to regulate myogenic tone (16). Interestingly, we have found that there are myogenic oscillations (vasomotion) in afferent arterioles, and an increase of transmural pressure induced by acute hypertension enhances the power of the vasomotion (25). These observations are consistent with the notion that there are oscillations of [Ca²⁺]_i in renal vascular smooth muscle in vivo. Together with our previous observation that RGD-containing peptides can induce vasoconstriction in isolated afferent arterioles by increasing the vascular smooth muscle [Ca²⁺]_i (27), the present study strongly suggests that integrins might be part of the signaling mechanism of the myogenic response in renal autoregulation.

Dihydropyridine-sensitive Ca²⁺ channels are abundant in renal arterioles (11) and are known to mediate Ca²⁺ influx in renal vascular smooth muscle. A patch-clamp study in smooth muscle cells isolated from cremaster muscle arterioles has suggested that exogenous RGD peptides modulate the activity of Ca²⁺ channels (24). Inhibition or stimulation of the Ca²⁺ current (measured by Ba²⁺ current) depends on which integrin heterodimers are ligated. However, the inability of nifedipine and depletion of extracellular Ca²⁺ to inhibit the increase of [Ca²⁺]_i induced by GRGDSP in the present study suggests that the increase of [Ca²⁺]_i is not the result of an extracellular Ca²⁺ influx but is due to the release of Ca²⁺ from intracellular stores. It is not surprising that smooth muscle cells from cremaster muscle arterioles and renal arterioles could employ very different Ca²⁺ signaling mechanism when ligands bind to integrins. RGD-containing peptides induce constriction in afferent arterioles but dilatation in cremaster skeletal arterioles (15, 27). Blocking of the IP₃ receptor with xestospongin C had no effects on Ca²⁺ mobilization induced by RGD-containing peptides, but ryanodine completely abolished it. These observations strongly suggest that the increase of [Ca²⁺]_i is due to the release of Ca²⁺ from ryanodine-sensitive Ca²⁺ stores and that the Ca²⁺ wave observed is due to Ca²⁺-induced Ca²⁺ release through ryanodine receptors. In the vascular smooth muscle cell line A7r5, the vasopressin-induced Ca²⁺ waves are mediated by IP₃-sensitive Ca²⁺ stores (1).

The next hypothesis tested was whether specific interactions between RGD peptides and integrins are required for the elevation of [Ca²⁺]_i. Antibodies against ₁- and ₃-integrins were chosen to test this hypothesis because these are the two most widely expressed integrin -subunits through which the extracellular matrix is connected to the cytoskeleton (5). Furthermore, a study in smooth muscle cells isolated from cremaster muscle arterioles has shown that exogenous RGD motifs increase the Ca²⁺ current via ₅₁-integrins and decrease the Ca²⁺ current via _v3-integrins (24). Immunofluorescence evidence from the present study indicates that these two -integrin subunits are also abundant in the primary cultures of renal vascular smooth muscle cells. Pretreatment with anti-₁- or anti-₃-integrin functional blocking antibodies totally abolished the elevation of [Ca²⁺]_i triggered by the hexapeptide GRGDSP (Fig. 7). The inhibition could be due to the blocking of specific interactions between integrins and RGD motifs or deterioration of cellular physiological conditions. However, 1 µM phenylephrine still triggered an increase of [Ca²⁺]_i even though the cells no longer responded to RGD peptides (data not shown), which indicates that the cells could still respond to a pharmacological agonist with mobilization of Ca²⁺. There was a transient decrease in the fluo 3 emission intensity when RGD peptides were introduced into the perfusion chamber (Fig. 7). This transient decrease (4-5 s) in emission intensity was due to the temporary dislocation of the coverslip from the focal plane during solution switching.

Ligation of either ₁- or ₃-integrin seems to be sufficient to induce Ca²⁺ mobilization. It is expected that blocking of both ₁- and ₃-integrin subunits is required to completely inhibit the effects of RGD peptides in [Ca²⁺]_i. Our observations indicate that functional blocking of either ₁ or ₃ is sufficient to inhibit that. It might be because soluble RGD peptide is less efficient to ligate integrins compared with anti-integrin antibodies and/or the binding of functional blocking antibodies to ₁-integrin reduces the binding affinity of ₃ to soluble RGD peptides and vice versa. The observation that sequential application of these two antibodies both can trigger Ca²⁺ signaling is consistent with this notion. Recognition of integrin by anti-integrin antibodies does not depend on the RGD binding motifs.

Incubation of the smooth muscle cells with anti-₁- or anti-₃-integrin antibodies before cellular fixation resulted in very different immunofluorescence patterns of integrins when compared with preparations with cellular fixation before incubation with antibodies. The difference in immunofluorescence patterns suggests that there is a redistribution/clustering of integrins on the plasma membrane as a result of the occupancy by the functional-blocking antibodies. The redistribution and clustering of integrins in the plasma membrane might result in conceding the binding sites for RGD or changing the microenvironment of the RGD binding sites, which prevents the interactions of RGD with the -integrins. All of the above observations indicate that specific interactions between RGD peptides and integrins are required for the increase of [Ca²⁺]_i in renal vascular smooth muscle cells. It is not certain whether the changes in [Ca²⁺]_i are the result of the RGD-containing peptides interacting with free integrin on the cell surface or the result of the RGD peptides acting through inhibition of the attachment of the vascular smooth muscle cells to the collagen matrix (7). It can be a combination of both. These two events are not mutually exclusive. The end result of either mechanism is the formation of new ligand-integrin interactions. The anti-₁- or anti-₃-integrin antibodies can trigger Ca²⁺ mobilization in the presence of 0.1 mM GRGDSP (data not shown). By assuming that all free RGD binding sites have been occupied by GRGDSP at this condition, this observation indicates that inhibition of smooth muscle cell attachment is capable of mobilizing intracellular Ca²⁺.

Integrins are -heterodimers. The integrin -subunits associated with the ₁- and ₃-subunits to mediate the intracellular Ca²⁺ mobilization were not identified in this study. Wu et al. (24) have reported that Ca²⁺ current (measured by Ba²⁺ current) was enhanced by ₅₁-agonists and was reduced by _v3-agonists in smooth muscle cells isolated from cremaster muscle arterioles. However, our observations indicate that the Ca²⁺ mobilization in renal vascular smooth muscle does not depend on an influx of extracellular Ca²⁺. ₁-Integrin can pair up a wide range of -subunits, ₁ to ₉ and _v (3). ₃-Integrin is most commonly paired with _v and _IIb (3). Identification of which heterodimer of integrins is involved in this signaling process will be the next step to elucidate the possible role of integrins in the mechanotransduction process of renal autoregulation (15, 27).

In summary, we have demonstrated that the hexapeptide GRGDSP induces an increase of [Ca²⁺]_i in cultured renal vascular smooth muscle cells similar to that observed in smooth muscle cells of isolated afferent arterioles. Local application of GRGDSP could trigger a regenerative Ca²⁺ wave that propagates across the cell. The increase of intracellular [Ca²⁺]_i depends on the Ca²⁺ released through ryanodine receptors and requires interactions of the RGD motif and integrins. Finally, the study is the first to show that RGD-containing peptides can trigger subcellular spatial and temporal variations in intracellular Ca²⁺.