INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE ROLE OF INTEGRINS in the physiology and pathophysiology of the cardiovascular system has become a field receiving increased interest. Recent work has revealed that in addition to serving as anchors to the cellular environment, integrins can function as transducers of a variety of cellular signals (7). Such signals result following interactions of integrins with extracellular matrix or cell surface ligands and include activation of tyrosine kinases, phospholipase C, protein kinase C, and modulation of membrane ion conductance (7). Many of these pathways directly overlap with known cell-signaling pathways involved in regulating vascular smooth muscle contractility. Thus we have hypothesized that integrins may represent important receptors that regulate vasomotor function.

Recently, Mogford et al. (16) have shown that synthetic Arg-Gly-Asp (RGD)-containing peptides that bind the _v3-integrins expressed by the vascular smooth muscle caused pronounced vasodilation in rat skeletal muscle arterioles. Similarly, proteolyzed fragments of collagen type I, which contains seven RGD sequences, produced vasodilations resembling those produced by RGD-containing peptides. This latter result coupled with previous work showing that extracellular matrix protein degradation can release soluble cryptic integrin-binding sites (11) has led to the concept that cryptic vasoactive signals exist within the extracellular matrix. The process of tissue injury may represent one relevant pathophysiological process leading to exposure and/or release of these vasoactive signals (11, 16).

Recently, D'Angelo et al. (9) found that _v3-integrin-mediated vasodilation was accompanied by a decrease in intracellular calcium concentration ([Ca²⁺]_i) of vascular smooth muscle cells (9). In this study we tested the hypothesis that integrins could in part reduce vascular smooth muscle [Ca²⁺]_i by activating K⁺ channels. This would lead to K⁺ efflux from the vascular smooth muscle cells causing hyperpolarization and subsequent inactivation of voltage-gated Ca²⁺ channels, leading to a decrease in vascular smooth muscle cell [Ca²⁺]_i and vasodilation. It is well documented that activation of K⁺ channels is a mechanism by which many vasodilators act (17), and integrins have been previously shown to be linked to K⁺ channel activity in other cell types (1-3). In this study we isolated rat cremaster arterioles and treated them with varying concentrations of synthetic cRGD peptide (i.e., GPenGRGDSPCA) with and without known K⁺ channel inhibitors.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Isolated vessel preparation. Male Sprague-Dawley rats (277 ± 13 g; n = 74) were anesthetized with an intraperitoneal injection of pentobarbital sodium (100 mg/kg). Anesthesia was confirmed by the loss of spinal reflexes. One cremaster muscle was carefully removed and placed in a cooled (4°C) Lexan chamber filled with physiological saline solution (PSS) containing (in mM) 145 NaCl, 4.7 KCl, 2.0 CaCl₂, 1.2 MgSO₄, 1.2 NaH₂PO₄, 0.02 EDTA, 5.0 glucose, 2.0 pyruvate, 3.0 3-(N-morpholino)propanesulfonic acid, and 0.15 albumin. A segment of first-order (1A) feed arteriole (175 ± 2 µm passive diameter) was surgically isolated and placed in a bath chamber. This chamber was fitted with pipette holders, and the entire apparatus was mounted on a microscope stage plate. The proximal end of the arteriole was tied onto a heat-polished micropipette (80-100 µm OD) using a single teased strand of 4-0 silk suture. Red blood cells were flushed from the lumen by application of gentle positive pressure (20 cmH₂O). The distal end was similarly tied onto a closed pipette. The vessel-mounting apparatus was transferred to the stage of an inverted microscope (Zeiss Axiovert 100). The arteriole was initially pressurized to 60 cmH₂O by adjusting the height of a water reservoir. After 1 h of being rewarmed to 34.5 ± 1°C, intraluminal pressure was raised in 10-cmH₂O steps every 10 min to a final pressure of 90 cmH₂O. Arterioles without leaks that gained spontaneous tone (75% of passive diameter) were used for these experiments. At the conclusion of each experiment, adenosine (1 mM) was added in a Ca²⁺-free PSS to obtain passive diameter. Diameters were recorded using video display equipment and a video caliper (13).

Experimental protocols. After equilibration (34.5°C, 90 cmH₂O intraluminal pressure), arterioles were treated with additive concentrations of cRGD peptide (2.1 × 10⁷ M to 3.0 × 10⁴ M) at 5- to 10-min intervals. Luminal diameter was measured at 1-min intervals during periods of peptide addition. These data were used to obtain a control concentration-response curve for each vessel. Arterioles were then rinsed three times over a 30-min period during which time they returned to their control diameter. Arterioles were then incubated with one of several K⁺ channel antagonists for 30 min followed by the addition of cRGD to obtain a second concentration-response curve. In experiments designed to test the involvement of K⁺ channels, antagonism of potassium channels was achieved with tetraethylammonium (TEA, 20 mM). Additionally, we used KCl (100 mM) to decrease the electrochemical driving force for K⁺ movement. To determine which specific K⁺ channel(s) might be involved, various antagonists were used. Iberiotoxin (100 nM) was used as a selective inhibitor of large conductance, calcium-activated K⁺ channels (K_Ca), and apamin (25 nM) was used to inhibit small conductance K_Ca channels. Voltage-gated K⁺ channels (K_V) were blocked using 4-aminopyridine (4-AP, 1 mM). Two doses of glibenclamide (1 µM) were used to inhibit ATP-sensitive K⁺ channels (K_ATP), and inward-rectifying K⁺ channels (K_IR) were inhibited with barium chloride (BaCl₂, 50 µM). Finally, further experiments were conducted using a combination of glibenclamide, barium, and 4-AP in the same concentrations as previously described to determine whether additive or interactive effects could be detected. Control experiments included the addition of the drug vehicle only and a time control series in which two consecutive cRGD concentration-response curves were obtained without a K⁺ channel inhibitor. To control for the possible increase in vasomotor tone that accompanied K⁺ channel antagonism, arterioles were preconstricted to the same degree with phenylephrine (5 × 10⁷ M) followed by the addition of cRGD to obtain a concentration-response curve.

Peptide/inhibitor preparation. Fresh cRGD was prepared by solubilizing the lyophilized peptide in albumin-free PSS. All drugs were prepared as concentrated stock solutions and were diluted to the appropriate experimental concentration immediately before each experiment. GPenGRGDSPCA and GRGESP were purchased from GIBCO Life Sciences (Gaithersburg, MD). All other drugs/inhibitors were purchased from Sigma (St. Louis, MO). For TEA experiments, TEA was substituted for 20 mM of NaCl in normal PSS in order to maintain a physiological osmolarity (290-310 mosM). For barium chloride experiments the MgSO₄ in normal PSS was substituted with MgCl₂ to prevent formation and precipitation of BaSO₄.

Data analysis. Arteriolar diameters were measured at 1-min intervals following addition of cRGD peptide. Maximal responses are expressed as means ± SE and are normalized to both basal tone (0%) and passive diameter (100%). Differences were tested using ANOVA, and the null hypothesis was rejected at P < 0.05.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

A concentration-dependent vasodilation was observed in response to cRGD (2.1 × 10⁷ M to 3.0 × 10⁴ M). All dilations were fully reversible by washing with fresh PSS. TEA (20 mM, n = 7) was used as a general K⁺ channel inhibitor to investigate whether K⁺ channels were involved in cRGD-mediated vasodilations (10). TEA treatment caused arterioles to constrict by 19.1 ± 5.0% and completely abolished cRGD-mediated vasodilation (Fig. 1A). As an additional test for the involvement of K⁺ channels, arterioles (n = 5) were treated with 100 mM KCl. This concentration of K⁺ reduces the driving force for K⁺ movement across the membrane and should prevent vasodilation due to K⁺ efflux from the vascular smooth muscle cells. After the addition of 100 mM KCl, arterioles constricted by 15.6 ± 8.5%. In addition, the vasodilation in response to cRGD was completely inhibited (Fig. 1B).

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Concentration-response curves for arterioles treated with cRGD (GPenGRGDSPCA) before and after treatment with 20 mM tetraethylammonium (TEA, A) or 100 mM KCl (B) as general K+ channel inhibitors. Concentrations of cRGD (x-axis) range from 0.21 to 300 µM. Percent dilation (y-axis) is a normalization for both basal tone and maximal diameter when treated with 1 mM adenosine in 0 calcium physiological saline solution.

To determine whether the constriction caused by 20 mM TEA or 100 mM KCl was involved in the inhibition of cRGD-mediated vasodilation, arterioles were preconstricted with phenylephrine (5 × 10⁷ M, n = 7). Phenylephrine constricted arterioles by 25.7 ± 2.6% and caused a slight inhibition of cRGD-mediated vasodilation at lower concentrations; however, there was no difference in the vasodilator response at higher concentrations (Fig. 2B).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Experiments were conducted to control for time and reduction in vascular diameter caused by K+ channel inhibitors. Time control experiments (A) show no difference in cRGD curves before and after treatment with vehicle. Preconstriction of arterioles with phenylephrine (5 × 107 M) to mimic the constriction caused by K+ channel antagonists had no effect on the maximal dilation to cRGD (B).

As a control for time, experiments were performed (n = 5) in which two consecutive concentration-response curves to cRGD were generated. No significant differences were observed between the two curves (Fig. 2A).

To elucidate the specific K⁺ channel(s) involved in the vasodilator response, arterioles were treated with a series of specific K⁺ channel inhibitors. Iberiotoxin (100 nM), a selective peptide derived from scorpion venom that is specific for large conductance, K_ca channels (12) had no effect on cRGD-mediated vasodilation (Fig. 3A). Similarly, apamin (25 nM), which selectively inhibits small conductance K_Ca channels, was unable to affect cRGD-mediated vasodilation (Fig. 2B) (8). By comparison, treatment of arterioles with 4-AP (1 mM) to inhibit voltage-gated K⁺ channels produced 31.7 ± 12.4% inhibition at 3 × 10⁴ M cRGD. Overall, 4-AP produced a rightward shift in the concentration-response curve that was significant at higher concentrations of cRGD (Fig. 4A). Glibenclamide (1 µM) was used to assess a role for ATP-sensitive K⁺ channels. Significant inhibition (13.8 ± 6.7%) was only apparent at higher concentrations of cRGD (Fig. 5). The contribution of the inward rectifying K⁺ channel was tested with BaCl₂ (50 µM, n = 7). Barium caused significant inhibition of cRGD-mediated vasodilation at concentrations from 8 × 10⁶ M to 3 × 10⁴ M, which reached a maximum of 54.0 ± 7.6% at 3 × 10⁵ M (Fig. 4B). Finally, experiments were conducted in which all three antagonists that produced significant inhibition (glibenclamide, barium, and 4-AP) were added collectively to determine whether the inhibition was additive. The three-drug combination caused highly significant inhibition of cRGD-mediated vasodilation but was unable to completely inhibit this response (Fig. 6A). At the highest concentration of cRGD (3 × 10⁴ M) arteriolar vasodilation was inhibited 67.6 ± 6.4%. This is approximately equal to the calculated vasodilation from the three inhibitors added individually.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Inhibition of large conductance Ca2+-activated K+ (KCa) channels (A) and small conductance KCa channels (B). Neither iberiotoxin (IBTX) nor apamin had any effect on cRGD-mediated vasodilation, leading to the conclusion that KCa channels are not involved in this response.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   A: voltage-gated K+ (KV) channels were blocked using 1 mM 4-aminopyridine (4-AP). A rightward shift in concentration-response curve when compared with control is shown. This shift is significant at higher concentrations of cRGD. B: barium chloride (50 µM) was used as an inhibitor of inward rectifying K+ (KIR) channel. A significant inhibition of cRGD-mediated vasodilation at higher doses of cRGD is shown.

View larger version (14K): [in this window] [in a new window]   Fig. 5.   ATP-sensitive K+ (KATP) channels were inhibited using glibenclamide (1 µM). In these experiments there is a small statistically significant inhibition of cRGD-mediated vasodilation that is significant only at highest concentrations of cRGD.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   A: concentration-response curve generated using glibenclamide, barium, and 4-AP in combination. cRGD-mediated vasodilation was inhibited by 67.6 ± 6.4%. B: results from individual inhibitor experiments are compared with combination experiments in this graph. Percent inhibition was calculated using the difference in control vs. inhibitor curves at maximal cRGD concentration (3 × 104 M). Individual inhibitor experiments are stacked in left bar and are compared with combination experiments on right. Graph shows that these are essentially equal, leading to the conclusion that all 3 channels (KV, KATP, and KIR) are involved in this response.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The purpose of our study was to test the hypothesis that efflux of K⁺ through membrane K⁺ channel(s) is involved in cRGD-mediated vasodilation. We observed that a synthetic peptide containing the RGD sequence caused concentration-dependent vasodilation, which has been previously reported to act through vascular smooth muscle _v3-integrin. In another recent study we observed that vascular smooth muscle [Ca²⁺]_i decreased in response to _v3 ligation in response to RGD-containing peptide (9). The magnitude of the reduction in [Ca²⁺]_i could be mimicked by the addition of acetylcholine (1 µM) or adenosine (10 µM), which produced dilation similar to the RGD peptide. These observations suggested that the integrin-mediated vasodilator response was linked to modulation of [Ca²⁺]_i perhaps through Ca²⁺ entry. Our data would suggest that activation of K⁺ channels through a link to integrins could alter membrane Ca²⁺ conductance. A negative feedback link between K⁺ channel-induced hyperpolarization and Ca²⁺ entry has been previously reported in vascular smooth muscle by Brayden and Nelson (5).

It is possible that the link between K⁺ channels and Ca²⁺ channel conductance involves more than membrane hyperpolarization. In endothelial cells it has been shown that integrin signaling can alter [Ca²⁺]_i (19). Schwartz et al. (18) postulated the involvement of an integrin-linked protein, possibly a channel (18) called integrin-associated protein, which may interact with _v-integrins. The signaling mechanisms that link integrins with intracellular Ca²⁺ levels in vascular smooth muscle may include a physical link between the integrin and channel proteins or second messenger-linked Ca²⁺ channel regulation. In recent studies we have found that Ba²⁺ current through L-type Ca²⁺ channels is reduced by soluble RGD-containing peptide (20), supporting the possibility that Ca²⁺ channel modulation may occur through a non-K⁺ channel-linked pathway(s).

In the present study our data suggest that integrin-mediated vasodilation may be linked to several K⁺ channel types. Numerous vasodilators have been shown to act through activation of K⁺ channels (4, 6, 17). Examples include adenosine, which acts through K_ATP channels, and nitric oxide, which acts through K_Ca channels (4). Evidence also suggests it is possible for a single stimulus to cause vasodilation through the combined involvement of several types of K⁺ channels. Bruch et al. (6) have shown that pituitary adenylate-cyclase-activating peptides activate both K_ATP and K_Ca channels in coronary vascular smooth muscle cells. Similarly, time-averaged shear stress has been shown to activate these same two channels in the vascular endothelium (14). Other investigators have shown that the second messengers protein kinase A and protein kinase G can modulate multiple K⁺ channels, further supporting the notion that more than one type of K⁺ channel can be activated (17). Given the importance of K⁺ channel activity and vasomotor tone, an association of K⁺ channels with integrins may provide a partial explanation for RGD-induced vasodilation.

Inhibition of RGD-induced vasodilation by TEA (20 mM) and KCl (100 mM) strongly suggest K⁺ channel involvement (Fig. 1). Using specific K⁺ channel antagonists, we found evidence for the involvement of K_V, K_ATP, and K_IR channels. Quantitatively, it appears that K_V channels account for 31.7 ± 12.4%, K_ATP 13.8 ± 6.7%, and K_IR 24.1 ± 14.0% of the dilation as estimated by the maximal response to cRGD. To determine whether the effects were additive, which would be expected if all three K⁺ channels participated, experiments using a combination of all three inhibitors (glibenclamide, barium, and 4-AP) were conducted, and it was found that the inhibition of vasodilation was equivalent to the sum of the inhibitions observed with the individual inhibitors (Fig. 6B). However, compensatory interactions between various K⁺ channel types in the regulation of membrane potential may make these estimates problematic.

The inability of individual antagonists to completely abolish the cRGD-mediated vasodilation despite complete inhibition with 20 mM TEA and 100 mM K⁺ suggests several possibilities. First, a unique type of K⁺ channel may be involved that is not blocked by the specific inhibitors utilized in our experiments. Second, the inhibitors may not completely inhibit their respective classes of K⁺ channels and may antagonize more than one K⁺ channel type. For example, it has been reported that there is a subset of K_V channels that are not inhibited by 4-AP (17). Also, a portion of the K_Ca channel current in erythrobalstoma cells that responded to integrin ligation was reportedly insensitive to -charybdotoxin and apamin (3). The involvement of several K⁺ channel types may reflect activation of diverging signaling cascades by integrin ligation (7), which in turn affect different K⁺ channels.

From our observations we propose that arteriolar vasodilation in response to RGD-integrin ligation is in part linked to K⁺ channels. Induction of hyperpolarization by K⁺ eflux would result in reduced Ca²⁺ entry and thus could provide a partial explanation for our previous observation of reduced vascular smooth muscle [Ca²⁺]_i in response to RGD-containing peptide. Further electrophysiological and biochemical studies will be required to determine the nature of the link between integrins and K⁺ channel activity.