We used genistein (Gen) and tyrphostin 23 (Tyr-23) to evaluate the importance of tyrosine phosphorylation in norepinephrine (NE)-induced changes in intracellular free calcium concentration ([Ca2+]i) in rat afferent arterioles. [Ca2+]i was measured in microdissected arterioles using ratiometric photometry of fura 2 fluorescence. The control [Ca2+]i response to NE (1 μM) consisted of a rapid initial peak followed by a plateau phase sustained above baseline. Pretreatment with the tyrosine kinase inhibitor Tyr-23 (50 μM, 10 min) caused a slow 40% increase in baseline [Ca2+]i. Tyr-23 attenuated peak and plateau responses to NE, both by ∼70%. In the absence of extracellular Ca2+ (0 Ca), Tyr-23 reduced the immediate [Ca2+]i response to NE by ∼60%, indicative of mobilization of internal stores, and abolished the plateau phase. In other arterioles, the [Ca2+]i response to depolarization induced by KCl (50 mM) was not attenuated by Tyr-23, indicating no direct effect on L-type Ca+ channels activated by depolarization. The Ca2+ channel blocker nifedipine (1 μM) inhibited the NE response by ∼50%; the effects of nifedipine and Tyr-23 were not additive. Nifedipine had no inhibitory effect after Tyr-23 pretreatment, indicating Tyr-23 inhibition of Ca2+ entry. Another tyrosine kinase inhibitor, Gen (5 and 50 μM), did not affect baseline [Ca2+]i. High-dose Gen inhibited the peak and plateau response to NE by 87 and 75%, respectively; low-dose Gen attenuated both responses by ∼20%. In 0 Ca, Gen (50 μM) abolished the immediate [Ca2+]i mobilization response. Combined nifedipine and Gen (50 μM) inhibited the rapid NE response by ∼90% in the presence of extracellular Ca2+. Gen (50 μM) also inhibited by 60% the [Ca2+]i response to 50 mM KCl, indicating a direct interaction with voltage-sensitive, L-type Ca2+ entry channels. These results indicate that tyrosine phosphorylation is an important link in the chain of events leading to α-adrenoceptor-induced Ca2+ recruitment (both entry and release) in afferent arteriolar smooth muscle cells. Furthermore, different blockers of tyrosine kinase appear to have different modes of action in renal microvessels.
- renal circulation
- vascular smooth muscle cells
- tyrophostin 23
- Ca2+ mobilization
- Ca2+ entry
- L-type Ca2+ channels
sympathetic nerves richly innervate the renal vasculature (3), releasing norepinephrine (NE) to exert its contractile effect by activation of cell-surface α1-adrenoceptors on vascular smooth muscle cells (VSMC) to increase cytosolic calcium concentration ([Ca2+]i) and constrict renal resistance vessels (6, 32, 34, 37). In this regard, the sympathetic autonomic system plays a crucial role in the regulation of renal hemodynamics and glomerular filtration, in addition to direct effects on tubular reabsorption, and thus in the short- and long-term regulation of extracellular fluid volume and arterial blood pressure (6).
The main vascular action of NE is exerted via G protein-coupled α1-adrenoceptors and their activation of phospholipase C with subsequent hydrolysis of membrane phosphoinositides generating inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (46). This is followed by an increase in [Ca2+]i mediated by at least one of several mechanisms. Ca2+ is recruited either as mobilization from intracellular stores via IP3 activation of release channels on the sarcoplasmic reticulum or as entry from the extracellular space through voltage-dependent and/or receptor-activated calcium channels located in the cell membrane. We previously demonstrated that the increase in [Ca2+]i induced by activation of cell-surface α1-adrenoceptors in rat preglomerular vessels derives from both intra- and extracellular sources (32, 33). The exact mechanisms and intermediate pathways responsible for Ca2+ signaling on adrenoceptor activation await further investigation.
Tyrosine kinases are widely recognized for their function as cell-surface receptors for growth factors such as epidermal growth factor (EGF) and platelet-derived growth factor (38). Activation of these mitogenic receptors leads to stimulation of gene expression and growth that requires multiple phosphorylation steps involving protein tyrosine kinases. In addition, there are examples of ligand activation of receptor tyrosine kinases that elevate [Ca2+]i (9, 40). Participation of tyrosine kinases in signaling events gains complexity as it is recognized that other families of tyrosine kinases are located in the cytosol and are devoid of extracellular binding sites for ligands (16). An example of cytosolic tyrosine kinases is the c-src family that has been detected in several tissues, including VSMC (6, 11). There is emerging evidence that cytosolic tyrosine kinase activity can be regulated by cell-surface G protein-coupled receptors during certain conditions (41). α-Adrenoceptor agonists such as NE and phenylephrine may stimulate protein tyrosine phosphorylation in VSMC (45, 47, 48). In addition to the well-characterized effects on growth and proliferation, some evidence suggests tyrosine kinases may play a role in the regulation of adrenoceptor-induced increase in [Ca2+]i and vascular tone (22).
Relatively few studies have addressed the role of tyrosine kinases in agonist-induced renal vasoconstriction. In the rat juxtamedullary nephron preparation, afferent and efferent arteriolar contractile responses to ANG II are reported to be attenuated by tyrosine kinase inhibitors such as tyrphostin 23 (Tyr-23) and tyrphostin AG 1478 (4). It appears that ANG II transactivates a cell-surface growth factor receptor tyrosine kinase to increase [Ca2+]i in rat afferent arterioles (5). Also, deendothelialized rat juxtamedullary arterioles respond to ANG II with an increase in [Ca2+]i that can be blunted by genistein (Gen), a blocker of tyrosine kinase (29). NE-induced constriction of isolated rat interlobular arteries appears to rely, at least in part, on activation of tyrosine kinases (13). However, the importance of tyrosine kinases in mediating cathecholamine-induced Ca2+ recruitment and vasoconstriction of the glomerular arterioles is not clear.
The aim of the present study was to evaluate the role of tyrosine kinases in adrenoceptor-induced changes in [Ca2+]i in VSMC of rat renal resistance vessels isolated by microdissection. The pharmacological agents Gen and Tyr-23 were used to block tyrosine kinases on NE stimulation as [Ca2+]i in afferent arterioles was measured using ratiometric fluorescence of the indicator fura 2. Furthermore, we sought to identify tyrosine kinase effects on NE-induced Ca2+ mobilization and/or entry and on depolarization-induced Ca2+ entry through L-type channels.
Measurements of [Ca2+]i. Glomeruli with an attached isolated afferent arteriole were microdissected from 54 Wistar-Kyoto rats (weight 229 ± 5 g) of the Chapel Hill colony, Møllegaard (Ejby, Denmark) or Charles River (Sulzfeld, Germany). No differences were noted between strains. Thin slices (thickness 0.5–1 mm) were cut from the midregion of the kidney and transferred to a dissection dish containing ice-cold physiological salt solution (PSS) solution with bovine serum albumin (0.5 g/dl; Sigma). Sharpened forceps were used for the isolation procedure under microscopic visualization (magnification ×12–100) as described previously (32). A single afferent arteriole was cut using a sharp knife blade as close as possible to the bifurcation arising from an interlobular artery. A homogenous population of arterioles was obtained from the outer one-third of the cortex. If no preparation was obtained during the first 90 min of dissection, the kidney was discarded.
After dissection, an arteriole was loaded with fura 2-AM for 45–60 min in the dark at room temperature as previously described (18, 36). Fura 2-AM (Molecular Probes) was dissolved in DMSO as a stock solution (1 mM) and mixed with PSS to a final concentration of 2 μM plus Pluronic F127 (Molecular Probes; 0.01%) immediately before use. A fura 2-loaded arteriole was transferred to a chamber containing PSS on the stage of an inverted microscope (Olympus IX 70 or IX 50) using an Eppendorf micropipette. The proximal end of the arteriole and attached glomerulus were aspirated into concentric glass-holding pipettes using negative pressure generated by a syringe connected to the back of the pipette. For measurements of [Ca2+]i, the arteriole was centered in the optical field of ×40 quartz oil-immersion objective.
The first part of this study was performed using a photometer system. The preparation was visualized using a video camera (Sony) and monitor. Variable shutters were adjusted to center an arteriole in a sampling window. This made possible continuous control of the position of the preparation throughout an experiment. The arteriole was excited alternatively with UV light of 340- and 380-nm wavelengths from a dual-excitation wavelength DeltaScan equipped with dual monochromators and a light pathway chopper (Photon Technology International). Fluorescent light signals were directed through a 510-nm band pass filter and detected by a photometer. The fluorescence signal intensity was processed using Felix software (Photon Technology International) and stored on an IBM-compatible computer. For the second part, an intensified video camera and Image-Master software (Photon Technology International) were used for experiments conducted in Chapel Hill. Copenhagen studies used a digital video camera (SensiCam) and the Image Workbench software (Axon). The following experimental procedure was similar for the studies conducted in the two laboratories. A vessel was visualized on a computer monitor, and the region of interest for [Ca2+]i was encircled using a software-based routine. A monochromator produced alternative excitation of UV light of 340- and 380-nm wavelengths, and fluorescent emission at 510 nm was recorded. The [Ca2+]i was calculated using the Grynkiewicz equation (15): [Ca2+]i = Kd × [(R - Rmin)/(Rmax - R)] × (Sf/Sb), where Kd is the dissociation constant of fura 2 for calcium; Sf and Sb are the 380-nm fluorescence at zero and saturating calcium concentrations, respectively. Rmin and Rmax are values of R (fluorescence ratio 340/380) at zero and at saturating (39 μM) calcium concentration, respectively. Values for Kd, Rmin, Rmax, Sf, and Sb were determined in vitro for each experimental set-up as previously described (18, 35, 36).
Artificial solutions. The PSS solution had the following composition (in mM): 135 NaCl, 5.0 KCl, 1.0 CaCl2, 1.0 MgCl2, 10 HEPES, and 5.0 d-glucose. A nominally calcium-free solution (0 Ca) was made by adding 2 mM EGTA (Sigma) to PSS and replacing CaCl2 with NaCl. In the 50 mM KCl solution (K50), 45 mM of NaCl was replaced with an equal concentration of KCl.
Drugs. Nifedipine (Sigma) was dissolved in DMSO and diluted in PSS to a final concentration of 1 μM, a concentration previously shown to completely inhibit L-type calcium channels (14, 32). Gen, Tyr-23, and tyrphostin 1 (Tyr-1; Sigma) were dissolved in DMSO and added to the PSS to produce the stated final concentrations.
Experimental protocol. All experimental solutions were added in a volume large enough to allow total exchange of the fluid in the experimental chamber several times. The fluid level in the experimental chamber was constantly maintained by means of a vacuum suction system. The viability of the preparation was assessed by [Ca2+]i and contractile stimulation with 1 μM NE or K50. Increases in [Ca2+]i typically accompanied visual contraction of the vessel. Discarded nonresponding preparations were almost always visibly damaged by physical dissection procedures, and/or the cells of the vessel were clearly swollen. Earlier studies revealed that 1 μM NE elicited roughly one-half of maximal response in [Ca2+]i; this concentration was used throughout the present studies (32, 33). We established that pretreatment for 10 min with the highest concentration of DMSO used to dissolve nifedipine, Gen, Tyr-23, and Tyr-1 did not significantly affect agonist-induced [Ca2+]i responses.
The responses to NE (1 μM) and K50 (50 mM) obtained during control conditions were compared in paired fashion with experimental responses recorded during pretreatment with blockers or 0 Ca. Arterioles were exposed to Tyr-23 (50 μM), Tyr-1 (50 μM), Gen (5 or 50 μM), or nifedipine (1 μM) for 2 or 10 min before and during NE stimulation. The mean prestimulation [Ca2+]i values were obtained between 10 and 15 s before addition of NE or K50. After stimulation, [Ca2+]i was measured as an immediate peak response (maximal peak value between 0 and 15 s) and a sustained plateau phase recorded between 30 and 35 s. In preliminary studies, we established that the response at 30–35 s was commonly maintained for several minutes.
The experiments were performed in random order to establish reversibility and to exclude possible effects of a prolonged action of a particular treatment. Also, to secure reversibility, a control response to NE or K50 was usually obtained both before and after the response in presence of blocker or 0 Ca. In poststimulation experiments, the responses to NE were recorded until a stable plateau level was reached (typically <30 s) and then during nifedipine inhibition during continued exposure to NE and pretreatment solution. For this series, we present the plateau values of [Ca2+]i before and after inhibition.
Statistical analysis. Data are presented as means ± SE. SigmaStat software (SPSS) was used for statistical analysis. Statistical significance was evaluated by analysis of variance or by analysis of variance for repeated measurements and the Newman-Keuls test. Student's t-test was used for paired and unpaired observations. When data did not exhibit normality, the data were transformed to natural logarithms before testing. A P value of <0.05 is considered statistically significant.
[Ca2+]i responses to NE and K50. Baseline [Ca2+]i averaged 74 ± 4 nM in a total of 76 afferent arterioles from 54 rats. Addition of NE (1 μM) to the bath caused an abrupt, sustained increase in [Ca2+]i. The initial peak was usually slightly greater than the sustained plateau. On average, NE caused [Ca2+]i to increase from a basal value of 78 ± 5 to an initial peak of 199 ± 10 nM, followed by a sustained plateau of 157 ± 7 nM at 30–35 s (n = 68). Stimulation with KCl (50 mM) solution caused an immediate increase in [Ca2+]i from 60 ± 7 to 118 ± 9 nM (n = 25), which was maintained at 112 ± 8 nM at 30–35 s. Contraction of arterioles visualized on a video monitor normally correlated temporally with the increases in [Ca2+]i.
Effect of Gen on NE responses. Administration of NE (1 μM) in the control period increased [Ca2+]i from 97 ± 8 to an immediate peak of 269 ± 31 nM, followed by a sustained plateau value of 187 ± 14 nM (n = 8). Subsequent 2-min exposure to Gen (50 μM) caused a small decrease in basal [Ca2+]i from 92 ± 8 to 81 ± 7 nM (P < 0.01). Gen pretreatment reduced peak and plateau responses to NE by ∼70%. NE challenge in the presence of Gen elicited a [Ca2+]i peak increase to 107 ± 10 nM and a sustained plateau of 106 ± 8 nM at 30–35 s (Figs. 1 and 2).
A lower concentration of Gen (5 μM) attenuated the [Ca2+]i responses to NE by ∼10–20%, which was significantly less than with 50 μM Gen. Under control conditions, NE increased [Ca2+]i from 86 ± 7 to 162 ± 17 nM, with a plateau phase of 132 ± 13 nM at 30–35 s (n = 10). Five micromolar Gen did not affect basal [Ca2+]i. NE stimulation in the presence of Gen increased [Ca2+]i from 80 ± 7 to a peak of 143 ± 12 nM; the sustained plateau was at 118 ± 11 nM.
Longer exposure to the low concentration of Gen (10 min) had no greater inhibitory effect than that at 2 min. Before exposure to Gen, NE stimulated afferent arteriolar [Ca2+]i from 58 ± 20 to a peak of 260 ± 40 nM and a plateau of 194 ± 19 nM (n = 7). Ten-minute Gen (5 μM) exposure had no discernable effect on baseline [Ca2+]i but inhibited responses to NE. NE increased [Ca2+]i immediately from 46 ± 18 to 205 ± 17 nM and then remained at 155 ± 17 nM at 30–35 s (both P < 0.001 vs. baseline; n = 7). Although Gen tended to reduce the peak, it was not statistically different from the control peak (P > 0.15). Gen attenuated the [Ca2+]i plateau to ∼80% of the control response to NE.
In other experiments, we evaluated whether Gen affected Ca2+ mobilization from intracellular stores, using the higher concentration of Gen (50 μM) and the 2-min pretreatment time. After the arterioles had exhibited a normal control response to NE, they were exposed to a nominally Ca2+-free EGTA-containing solution (0 Ca) and Gen. This bathing solution caused a significant reduction in resting [Ca2+]i from 87 ± 16 to 61 ± 8 nM (n = 5). NE had no stimulatory effect on arterioles pretreated with 0 Ca + Gen. In the presence of Gen, the average [Ca2+]i peak was 64 ± 10 nM and the plateau at 30–35 s was 63 ± 10 nM (Fig. 3). Control experiments without Gen when the bath was 0 Ca showed that the 0 Ca caused basal [Ca2+]i to decline from 95 ± 10 to 74 ± 7 nM (n = 11). Subsequent challenge with NE caused a rapid peak to 127 ± 16 nM, whereas the plateau value of 80 ± 6 nM at 30–35 s did not differ from baseline (P = 0.2). These findings indicate that Gen affects the peak phase of mobilization. The absence of the plateau phase when extracellular Ca2+ was absent indicates that the [Ca2+]i plateau response at 30–35 s represents Ca2+ entry. Thus it is reasonable to conclude that Gen influences Ca2+ entry as the plateau phase is attenuated by Gen when extracellular Ca2+ is normal.
This led us to investigate whether the effects of Gen might be exerted via blockade of L-type channels. We previously showed that ∼50% of the Ca2+ entry during the plateau response to NE is mediated via these dihydropyridine-sensitive channels and 50% by other entry pathways (32, 33). These observations were confirmed by our present results. We found that in the presence of 1 μM nifedipine, NE caused an increase in [Ca2+]i from 80 ± 6 nM to a peak value of 132 ± 21 nM and a plateau of 110 ± 13 nM (n = 5; Fig. 4). The combination of nifedipine (1 μM) and Gen (50 μM) inhibited the NE responses even more. The corresponding values were 69 ± 11 nM for baseline and 73 ± 10 nM for both peak and 73 ± 9 nM for plateau (n = 5). The small increases in [Ca2+]i above baseline were not statistically significant. The combination of Gen and nifedipine attenuated the plateau response significantly more than pretreatment with Gen only. The difference in peak response did not, however, reach the level of statistical significance (P = 0.13). Because the effects of the L-type Ca2+ channel antagonist nifedipine and Gen are additive, one might conclude that tyrosine kinases are involved in activating the ill-defined Ca2+ entry pathway distinct from voltage-gated channels.
Effect of Gen on K50 responses. From the data presented above, it is not clear whether Gen has a direct effect on voltage-gated L-type Ca2+ channels, as has been reported by others for nonrenal vessels (25, 49). We therefore performed experiments to more definitively investigate the issue of direct vs. indirect action. Administration of K50 in the control period increased [Ca2+]i from 88 ± 21 to 128 ± 22 nM in the initial 15 s and to 123 ± 22 nM at 30–35 s (n = 4; Fig. 2). After 2-min exposure to Gen (50 μM), K50-induced depolarization caused a smaller, but significant, increase in [Ca2+]i initially and during the plateau phase (from 75 ± 23 nM to 87 ± 23 and 89 ± 25 nM, for peak and plateau, respectively, n = 4).
A lower concentration of Gen (5 μM) had a smaller inhibitory effect. In the control period, K50 elevated [Ca2+]i from baseline 80 ± 8 to 118 ± 16 and 116 ± 14 nM for peak and plateau, respectively (n = 7). After pretreatment with Gen, the corresponding values were 78 ± 7, 108 ± 12, and 106 ± 12 nM. The plateau value was significantly attenuated, whereas the mean peaks did not differ from the control response (P > 0.1). Thus it is clear that Gen, at least at the higher concentration tested, exerts a direct inhibitory influence on the L-type channels.
Effect of Tyr-23 and Tyr-1 on NE responses. We tested the effect of a different blocker of tyrosine kinase, Tyr-23, that acts by binding to the substrate binding site of tyrosine kinase, in contrast to Gen which binds to the ATP site. Tyr-23 is reported to be a more specific blocker of tyrosine kinases than Gen (16). Although long-term exposure (up to 16 h) has been reported to be required for optimal effects of tyrphostins (26), more recent reports using different VSMC preparations, among them renal vasculature, indicate that 10 min of preincubation are sufficient to produce effective inhibition (4, 42, 44). Because tyrphostins may degrade to products with different potencies, we wanted to minimize the duration of preincubation (30).
In initial experiments, we tested the inhibitory effect of Tyr-23 (50 μM) on the [Ca2+]i response to NE after 3-min pretreatment. In the absence of the inhibitor, NE (1 μM) elicited a rapid peak rise in [Ca2+]i from 83 ± 8 to 156 ± 13 nM (n = 14). After 30–35 s, the [Ca2+]i plateau value was 130 ± 12 nM. After 3-min pretreatment with Tyr-23, the peak response to NE was significantly blunted (from a baseline of 101 ± 11 to 144 ± 18 nM). The plateau response, however, was not statistically lower (to 139 ± 16 nM). When we extended the treatment period to 10 min, the responses to NE were more markedly attenuated (P < 0.01). In the control period, NE stimulation elevated [Ca2+]i from 55 ± 6 to a peak of 169 ± 23 and a plateau of 132 ± 16 nM (n = 18; Fig. 5). After 10-min pretreatment with Tyr-23 (50 μM), baseline [Ca2+]i rose from 55 ± 5 to 84 ± 7 nM. NE stimulation caused an increase to 121 ± 12 nM within the initial 15 s and to 110 ± 10 nM at 30–35 s. Thus Tyr-23 reduced the NE-induced peak and plateau responses to 30 ± 6 and 30 ± 5% of those during control conditions, respectively. This degree of inhibition is similar to what we observed with 50 μM Gen. Preliminary experiments revealed that 250 μM Tyr-23 exerted no stronger inhibition than did the 50 μM dose. Thus, for the remaining studies involving Tyr-23, we elected to use a concentration of 50 μM and a pretreatment time of 10 min. We also tested the effect of an inactive analog Tyr-1 on the NE-induced response in paired experiments. We found that the responses after 10-min pretreatment with Tyr-1 (50 μM) were 112 ± 17 and 79 ± 8% (n = 6) of the control response for the peak and plateau, respectively. Both the peak and plateau responses are significantly greater than the corresponding values after Tyr-23 treatment (30 ± 6 and 30 ± 5%; see above).
We also evaluated the effect of Tyr-23 on NE-induced Ca2+ mobilization and entry. NE stimulation in the presence of the 0 Ca solution was compared with that during 0 Ca + Tyr-23 (50 μM). NE application in the presence of the 0 Ca solution caused the expected transient [Ca2+]i peak, from 38 ± 12 to 75 ± 14 nM (n = 7) that returned to baseline (Fig. 6). Pretreatment with 0 Ca + Tyr-23 further attenuated the peak [Ca2+]i response to NE, from the basal level of 42 ± 9 to 56 ± 13 nM (n = 7). In both 0 Ca experiments, [Ca2+]i returned to baseline by 30–35 s, indicating the absence of a plateau phase that normally represents Ca2+ entry.
Because we found that Gen and nifedipine had additive inhibitory effects on the [Ca2+]i response to NE, we investigated the combined effect of Tyr-23 (50 μM) and nifedipine (1 μM). Ten-minute pretreatment with nifedipine blocked the peak and plateau responses to NE to 69 ± 7 and 66 ± 9% of the control response, respectively. In the absence of nifedipine, NE increased the [Ca2+]i from 60 ± 11 to peak and plateau values of 201 ± 23 and 154 ± 16 nM (n = 9). Nifedipine reduced the NE responses as basal 59 ± 14 nM rose to a peak of 161 ± 27 and a plateau of 121 ± 18 nM (n = 9; Fig. 7). When the preparations were treated with the combination of Tyr-23 and nifedipine, greater inhibition was observed (P < 0.02). Peak and plateau responses were 35 ± 7 and 37 ± 10% (from 81 ± 9 to 129 ± 16 and 123 ± 14 nM; n = 9) of the control response (from 73 ± 11 to 212 ± 26 and 179 ± 16 nM; n = 9). On the other hand, the combined effect of Tyr-23 + nifedipine was no greater than the effect of Tyr-23 alone (30 ± 6 and 30 ± 5% of control response for peak and plateau, respectively; P > 0.5). This indicates that Tyr-23 blocks most of the Ca2+ entry via L-type Ca2+ channels.
We therefore determined the effect of addition of nifedipine on the NE-induced plateau phase in vessels exposed to Tyr-23. After 10-min Tyr-23 pretreatment, baseline [Ca2+]i was 65 ± 11 nM. NE caused a sustained [Ca2+]i plateau of 85 ± 11 nM (n = 3). Subsequent addition of 1 μM nifedipine had no effect on the plateau level (83 ± 12 nM) in the continued presence of NE and Tyr-23. This finding further strengthens the notion that Tyr-23 blocks Ca2+ entry via L-type channels.
Effect of Tyr-23 on K50 responses. Additional studies assessed whether Tyr-23 inhibits L-type Ca2+ channels via a direct or indirect action. We examined effects of Tyr-23 on KCl-induced depolarization. The K50 solution caused a peak [Ca2+]i increase from 54 ± 7 to 109 ± 12 nM and a sustained plateau of 98 ± 4 nM (n = 6; Fig. 5). Pretreatment with Tyr-23 (50 μM) for 10 min did not inhibit the peak and plateau responses to K50 as in the presence of Tyr-23; K50 increased [Ca2+]i from 86 ± 8 to 152 ± 19 and 158 ± 22 nM. Thus we did not find any indication that Tyr-23 exerted any direct blocking effect on L-type channels.
Effect of Tyr-23 and Tyr-1 on resting baseline [Ca2+]i levels. As mentioned above, Tyr-23 for 10 min increased baseline [Ca2+]i. Tyr-1, on the other hand, did not increase the [Ca2+]i baseline. On the contrary, after 10-min pretreatment with 50 μM Tyr-1, there was a tendency for baseline [Ca2+]i to decrease (from 89 ± 23 to 70 ± 20 nM; n = 6). This difference was not, however, statistically significant (P > 0.08). To evaluate whether the Tyr-23-induced increase was due to Ca2+ entry via L-type channels, we pooled the paired experiments to compare Tyr-23 effects on basal [Ca2+]i in the absence and presence of nifedipine. Pretreatment with Tyr-23 (50 μM) for 10 min elevated [Ca2+]i from 66 ± 8 to 90 ± 10 nM (n = 9). In the presence of nifedipine, Tyr-23 increased the baseline [Ca2+]i from 63 ± 9 to 81 ± 9 nM. This increase was not significantly different from the increase obtained with Tyr-23 only (P > 0.5). Next, we pooled all paired pretreatment experiments using the 0 Ca solution and the 0 Ca + Tyr-23 combination. Treatment with 0 Ca for 10 min caused baseline [Ca2+]i to decrease by 38 ± 4 nM (n = 18). When treated with a combination of 0 Ca + Tyr-23, the decrease was 22 ± 4 nM, a smaller drop than that recorded using 0 Ca only (P < 0.01). These findings suggest that blocked extrusion of Ca2+ from the cytoplasm rather than increased entry might be the reason [Ca2+]i was increased by Tyr-23 treatment.
It is clear that an increase in [Ca2+]i is a major link in the chain of events leading to contraction of VSMC (46). Previous studies characterized NE effects on [Ca2+]i in isolated renal vessels (21, 28, 32–34, 51). We previously showed that the [Ca2+]i response to adrenoceptor stimulation is dependent on both entry from extracellular fluid and mobilization from intracellular Ca2+ stores (32, 33). There is a paucity of information concerning the possible role of tyrosine kinases in the intracellular signaling mechanisms that mediate adrenoceptor-induced action on renal resistance vessels. To gain insight into this issue, our aims were to evaluate the role of tyrosine phosphorylation in NE-induced Ca2+ mobilization, entry, or both.
Tyrosine kinases are thought to act at several levels of intracellular signaling leading to VSMC contraction (16). This includes modulation of Ca2+ and K+ channels, intracellular Ca2+ stores, Ca2+ sensitivity of the contractile apparatus, and interaction between the contractile apparatus and the cytoskeleton (16, 39). Several studies of nonrenal vascular beds report attenuation of adrenoceptor-induced vasoconstriction by inhibitors of tyrosine kinases (1, 8, 10, 13, 22, 43, 53). Although tyrosine kinases may affect contractility of VSMC unrelated to [Ca2+]i (12), other studies suggest that tyrosine kinases act, at least in part, via control of [Ca2+]i (22, 43). In addition, tyrosine kinase inhibitors impact on control of [Ca2+]i to attenuate ANG II- and vasopressin-induced vasoconstriction (20, 44). In several VSMC preparations, it has been shown that activation of adrenoceptors causes tyrosine phosphorylation (20, 45, 47, 48). The phosphorylation is inhibited by tyrosine kinase inhibitors such as Gen and Tyr-23 (20, 47, 48). It seems that many proteins can be phosphorylated by NE activation of adrenoceptors in VSMC. In cultured aortic smooth muscle cells, at least nine different proteins are phosphorylated on NE stimulation (45). Other observations include NE-induced phosphorylation of paxillin (47) and mitogen-activated protein kinase (20). The diversity of proteins phoshorylated raises the possibility of several different effector mechanisms.
In the present study, we observed that two inhibitors of tyrosine kinase, Gen and Tyr-23, differing structurally and in mechanism of action, attenuated the [Ca2+]i response to NE in isolated afferent arterioles. We found that Gen (50 μM) and Tyr-23 (50 μM) elicited similar buffering effects on the NE-induced [Ca2+]i response, each blocking ∼70 to 80%. It appears that IC50 doses for these blockers vary among different preparations and subtypes of tyrosine kinases (16). In one study on large renal arteries (diameter ∼150–300 μm), Tyr-23 (50 μM) is reported to have no effect on NE potency or maximal contraction (13). In the same study, Gen (50 μM) had no effect on NE potency and only a minor effect on maximal contraction. Tyr-23 (100 μM), however, attenuated maximal contraction to an extent similar to that found in the present study. Thus it is possible that the smaller renal resistance vessels are more sensitive or reactive than larger arteries. Tyr-23 is considered to be a more specific inhibitor of tyrosine phosphorylation than Gen (16). We also found that the inactive tyrphostin analog Tyr-1 had no attenuating effect on the peak response to NE and tended to have a small (∼20%) attenuating effect on the plateau (at 30 s). Although Tyr-1 may have had a minor effect on the sustained phase, it should be appreciated that it was considerably smaller than the ∼70% inhibition of both peak and plateau phases elicited by Tyr-23.
The action of another agonist of G protein-coupled receptors, ANG II, is thought to be partially dependent on tyrosine kinases in renal resistance vessels (4). Tyr-23 (100 μM) was found to block ∼35% of ANG II-induced contraction of afferent arterioles in the rat juxtamedullary nephron preparation. A subsequent study indicated that the EGF receptor tyrosine kinase mediates the response to ANG II by inhibiting Ca2+ influx but not mobilization (5). When the preparation was bathed in a solution containing a low concentration of Ca2+ (100 nM), the response to 100 nM ANG II consisted of a transient peak that was not affected by tyrosine kinase blockade with Tyr-AG 1748.
Our results indicate that tyrosine kinase inhibition attenuates the Ca2+ entry response to NE in renal arterioles. As we previously reported and as indicated by our present results, there is no elevation of sustained [Ca2+]i during NE stimulation when Ca2+ entry is prevented by the absence of extracellular Ca2+ (32, 33). This finding indicates that Ca2+ recruitment during the sustained plateau phase occurs solely via Ca2+ entry from the extracellular fluid. Thus the attenuation of the [Ca2+]i plateau by tyrosine kinase inhibition reflects blockade of Ca2+ entry. We have in this and other studies shown that 40–70% of the Ca2+ entry in response to NE is mediated via dihydropyridine-sensitive L-type Ca2+ channels (32, 33). We therefore sought to identify the role of tyrosine kinases in the NE-induced activation of L-type Ca2+ channels.
We found that 2 min of combined treatment with Gen (50 μM) and nifedipine (1 μM) caused an additive attenuation of the plateau response to NE. Under these conditions, there was no significant response to NE. This contrasts with the 35–40% of control responses observed when Gen or nifedipine was a single treatment. Thus it is clear that a substantial fraction of the non-L-type Ca2+ entry component is blocked by Gen.
Less certain, however, is at what level this inhibition occurs. Gen blocks a larger part of the response than can be accounted for by the non-L-type pathway, implicating an effect on inhibition of entry via the L-type channels. As other investigators reported that Gen may inhibit these channels (25, 49), we tested whether this occurred in afferent arterioles. It has also been suggested that pp60c-src, a cytosolic tyrosine kinase, stimulates L-type Ca2+ channel currents in VSMC (50). Indeed, we found that Gen blocks ∼60% of the Ca2+ entry through voltage-gated Ca2+ channels in response to K50. This observation supports the notion that part of the Gen effect on the total NE response is due to direct inhibition of L-type channels. This, however, does not exclude the possibility that Gen in addition exerts this effect by an action on upstream events that subsequently trigger activation of L-type channels.
Because Gen is thought to exert an indirect effect on L-type channels, we also used a more specific tyrosine kinase inhibitor Tyr-23 (52). As mentioned above, Tyr-23 (50 μM) attenuates the [Ca2+]i response to NE to a similar extent as Gen (50 μM). When preatreatment with nifedipine and Tyr-23 was combined, there was no additive effect, in contrast to the effect observed with Gen. The combination of nifedipine and Tyr-23 produced no stronger inhibition of responses to NE than did Tyr-23 alone. On the other hand, the response was attenuated to a larger degree than after nifedipine pretreatment alone. Thus Tyr-23 appears to abolish NE-induced [Ca2+]i entry via L-type channels and possibly inhibits some entry via an alternative pathway. This notion is further strengthened by results from posttreatment experiments in which addition of nifedipine failed to further inhibit the plateau response to NE after 10-min pretreatment with Tyr-23. We previously showed that the response to K50 is completely blocked by nifedipine (1 μM) (32). It has been reported that Tyr-23 directly inhibits Ca2+ channels in VSMC (43, 49). We found, however, in accord with findings from a recent study, that the response to depolarization with high extracellular [K+] was not attenuated by Tyr-23 (4).
Collectively, our results suggest different modes of action of Gen and Tyr-23. Both compounds appear to inhibit Ca2+ entry via L-type channels. Tyr-23 seems to effectively block this component of the NE-induced [Ca2+]i response without interacting with the channel directly. Instead, Tyr-23 probably exerts a primary action on an upstream mechanism that secondarily activates L-type channels. Gen appears to directly inhibit L-type channel activity, but, paradoxically, the blockade of the sustained phase of the NE-induced response is less complete than with Tyr-23. Not clear is whether the effect of Gen on the Ca2+ entry component is solely due to direct action on the L-type channel or an indirect action on an upstream mechanism similar to that affected by Tyr-23. However, in the case of Gen, the direct action may predominate over the indirect. Further studies are required to resolve these differences. It is noteworthy that Tyr-23 and Gen inhibit tyrosine kinases by different mechanisms. Gen acts on ATP binding, whereas tyrphostins bind to the substrate-sensitive site and are accordingly considered to be more specific (2, 23).
In other experiments, we examined the effect of tyrosine kinase blockade on the transient [Ca2+]i response to NE in the absence of extracellular Ca2+. As previously established by us, this response, in afferent arterioles, consists of a clearly demonstrable immediate peak (32, 33). In the absence of Ca2+ entry, this transient is limited to release of Ca2+ from intracellular stores. We found that after 2-min pretreatment with 0 Ca, the transient peak to NE was reduced to ∼45% of the peak response when Ca2+ entry is allowed. When afferent arterioles were pretreated with 0 Ca + Gen, this transient peak was totally abolished. Our findings are in agreement with earlier results for the rat aorta (1). We also noted an attenuation of the peak response, although not as pronounced, after 10-min pretreatment with Tyr-23 (50 μM) and 0 Ca. As previously mentioned, one study reported that the intracellular release component of the Ca2+ response to ANG II was not affected by the tyrphostin-AG 1478, an agent considered to be specific for EGF receptor tyrosine kinase (5). This might indicate that the release component is not affected by plasma membrane EGF receptor kinase but by cytosolic tyrosine kinases as Tyr-23 does not distinguish between cytosolic and receptor tyrosine kinases. Another explanation is that release from intracellular stores plays a larger role in the NE response than that to ANG II, as has been indicated by other studies (17, 31, 32). Evidence implicates that tyrosine kinases act directly on IP3 receptors of the sarcoplasmic reticulum (19) or upstream between the cell-surface G protein-coupled receptor and mobilization of intracellular Ca2+ (24, 27). The latter reports suggest that stimulation of rat VSMC with endothelin and ANG II activates tyrosine kinase-dependent production of IP3. It is noteworthy that Gen does not appear to affect [Ca2+]i release activated directly by IP3 or caffeine (24). One view is that activation of G protein-coupled receptors commonly tightly linked to phospholipase C (PLC)-β stimulates receptor or cytosolic tyrosine kinases that, in turn, activate PLC-γ, which leads to phosphoinositide hydrolysis and IP3 production (27).
In contrast to Tyr-23, Gen does not affect baseline [Ca2+]i. The inactive analog Tyr-1 was also without effect in this regard. At present, we do not know whether the rise in baseline [Ca2+]i with Tyr-23 reflects accelerated influx or retarded efflux. However, our results indicate that this increase in baseline [Ca2+]i is at least partially due to impaired Ca2+ extrusion as the fall in baseline [Ca2+]i after 10-min pretreatment with 0 Ca was reduced by concomitant Tyr-23 treatment. Furthermore, blockade of Ca2+ entry through nifedipine-sensitive channels does not blunt the increased baseline [Ca2+]i due to Tyr-23.
In summary, our results indicate that two structurally different effective tyrosine kinase inhibitors, Gen and Tyr-23, but not inactive Tyr-1, buffer the [Ca2+]i response to NE in rat afferent arteriolar VSMC. We noted similarities and differences that may reflect preferential actions of the inhibitors on different tyrosine kinases. Both Gen and Tyr-23 blunted the intracellular release of Ca2+ in response to NE. Both inhibitors impaired the Ca2+ entry response to NE; however, the mechanisms appear to differ. Gen affected a non-L-type Ca2+ entry pathway as well as voltage-gated L-type Ca2+ channels. Tyr-23 affected Ca2+ entry via L-type channels and the effect seems to be indirect, secondary to upstream events involving mobilization. Tyr-23 had no direct effect on high KCl-induced depolarization activation of Ca2+ entry, whereas Gen exerted partial inhibition on this entry pathway, suggesting a direct interaction with voltage-sensitive L-type Ca2+ entry channels. Taken together, our findings suggest that tyrosine phosphorylation is an important event in α-adrenoceptor-mediated control of renal vascular resistance. Tyrosine kinases may act at a proximal signal transduction site in α-adrenoceptor-induced Ca2+ recruitment via mobilization and entry pathways.
The skillful technical assistance of A. Salomonsson is acknowledged.
These studies were supported by a National Institutes of Health Research Grant (HL-02334) from the National Heart, Lung, and Blood Institute, the Novo-Nordisk Foundation, the König-Petersen Foundation, the Berth von Kantzow's Foundation, and the Crafoord Foundation.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society