Relationship between α₁-Adrenergic Receptor-Induced Contraction and Extracellular Signal-Regulated Kinase Activation in the Bovine Inferior Alveolar Artery

Materials and Methods

Chemicals and Reagents.

Agents used were obtained from the following sources: desipramine, hydrocortisone, NE, propranolol, aprotonin, phenylmethylsulfonyl fluoride, and okadaic acid (Sigma-Aldrich, St. Louis, MO); 5-methylurapidil, BMY 7378, cyclazosin, rauwolscine, niguldipine, phenylephrine, and WB4101 (RBI/Sigma, Natick, MA); BE 2254 (Tocris Cookson Inc., Ballwin, MO); rabbit anti-phospho-ERK1/2 antibody and rabbit anti-total ERK1/2 antibody (Cell Signaling Technology Inc., Beverly, MA); anti-rabbit IgG-horseradish peroxidase-conjugated antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Phenylephrine and NE were dissolved in 0.9% saline containing 0.2% ascorbic acid. Phenylmethylsulfonyl fluoride was dissolved in 95% ethanol. All other drugs were dissolved in water.

Contraction Experiments.

Bovine lower jaws were obtained from a local slaughterhouse and immediately put into ice-cold phosphate-buffered saline (pH 7.6), containing 137 mM NaCl, 9.6 mM NaH₂PO₄, 2.7 mM KCl, 1.7 mM KH₂PO₄, for transportation to the laboratory (5 min). The BIAA was removed from the jaw and stripped of surrounding fat and connective tissue. To eliminate the release of nitric oxide or other factors from the endothelium, endothelium was removed by gentle rubbing with a stainless steel wire (Bockman et al., 1996). The absence of endothelium was confirmed by the loss of relaxation in response to 1 μM acetylcholine in arteries precontracted with 60 mM potassium chloride. Arteries were used immediately or were stored overnight in Krebs' solution (pH 7.4), containing 120 mM NaCl, 5.5 mM KCl, 2.5 mM CaCl₂, 1.4 mM NaH₂PO₄, 1.2 mM MgCl₂, 20 mM NaHCO₃, 11.1 mM dextrose, and 0.027 mM CaNa₂-EDTA, at 4°C for use the following day. No differences were observed in contractile responses between arteries used immediately and those stored overnight. The artery was cut into 3-mm-long ring segments, and the rings were attached to Grass FT 0.03 isometric force transducers (Grass Instruments, Quincy, MA) using stainless steel pins inserted through the lumen of the artery. The transducers were connected to a Grass polygraph for tension recordings. Rings were placed in Krebs' solution in glass muscle chambers, gassed with 95% O₂/5% CO₂, and maintained at 37°C. After 1 h equilibration at 2 g resting tension, tissues were contracted with 60 mM potassium chloride and then washed with Krebs' solution for 30 min. The potassium chloride contraction and washing procedure was then repeated twice more. After a final 30-min wash period, cumulative concentration-response curves for contraction were generated in each ring.

Measurement of Antagonist Affinity Values.

After initial NE concentration-response curves were obtained, the rings were washed with Krebs' solution and allowed to relax to the resting tension. Each ring was equilibrated for 1 h with a single concentration of antagonist, and NE concentration-response curves were then repeated. The concentrations of α₁-AR antagonists used were from 5 to 250 times the average literatureK_B value for an antagonist at its favored (highest affinity) α₁-AR (Table 1). In each experiment, NE concentration-response curves were also generated in a control ring that was not treated with antagonist, to determine whether there were changes in agonist potency or maximal contraction over time. Concentration-response curves were plotted and EC₅₀ values were obtained using nonlinear regression curve fitting (GraphPad Prism 3.0; GraphPad Software, San Diego, CA). Schild plots were constructed by the method of Arunlakshana and Schild (1959). Linear regression of the Schild plot gave pA₂ values from thex-intercept of the regression line. Slopes of the Schild plots were tested for difference from unity using a one-samplet test.

ERK1/2 Activation.

Inferior alveolar arteries were cut into 10-mm-long artery segments and placed in glass test tubes containing 10 ml of Krebs' solution maintained at 37°C and gassed with 95% O₂/5% CO₂. As in contraction studies, artery segments were equilibrated for 1 h, treated with 60 mM potassium chloride, and washed for 30 min. This process was then repeated once more. ERK1/2 was stimulated by the addition of adrenergic agonists or by TES. For TES experiments, artery segments were placed between two platinum electrodes and stimulated by applying square-wave pulses (10 Hz, 0.3 ms, 60 V) using a Grass S44 stimulator. In preliminary experiments, we have found that these stimulation parameters also cause α₁-AR-mediated contraction of the BIAA (data not shown). For each experiment, one artery segment was not treated with agonist or other drugs and was used for determination of basal ERK1/2 activity. For experiments involving antagonists, artery segments were incubated with antagonist for 1 h before TES or addition of agonist drugs. The concentrations of α₁-AR-selective antagonists used were approximately 13 (10 nM antagonist) or 130 nM (100 nM antagonist) times the average literature K_B value for an antagonist at its favored (highest affinity) α₁-AR subtype (Table 1; see Fig. 3 legend). Assays were terminated by placing artery segments in 0°C lysis buffer containing 137 mM NaCl, 20 mM Tris HCl, 1 mM CaCl₂, 1 mM MgCl₂, 1% Nonidet P-40, 1 mM EDTA, 1 μM aprotonin, 100 μM phenylmethylsulfonyl fluoride, and 10 nM okadaic acid. Tissues were minced with scissors and homogenized 3 times for 10-s intervals at 13,500 rpm with a Tissumizer using an SDT-080EN probe (Tekmar, Cincinnati, OH). Samples were centrifuged at 12,000g for 15 min to remove particulate, and the supernatants containing soluble cytosolic protein, including ERK, were aliquoted and stored at –20°C. Protein concentrations of the supernatant samples were determined using the Coomassie Plus Protein Assay (Pierce, Rockford, IL).

Immunoblotting.

Samples were thawed and boiled for 5 min in loading buffer (pH 6.8) containing 60 mM Tris-HCl, 25% glycerol, 2% SDS, 14.4 mM 2-mercaptoethanol, and 0.1% bromphenol blue. Protein samples (20 μg) were electrophoresed for 45 min on a 4 to 15% Tris-HCl/polyacrylamide gel (Bio-Rad, Hercules, CA) at a constant voltage of 200 V for 50 min on a Mini-PROTEAN electrophoresis apparatus (Bio-Rad). A colorimetric ladder (Bio-Rad) was included in each gel to identify protein size. Following electrophoresis, the contents of the gel were transferred to a nitrocellulose membrane (Osmonics, Westborough, MA) using a semidry transfer method (Bio-Rad). Nitrocellulose membranes were then washed three times for 10-min intervals in Tween 20 buffer containing 150 mM NaCl, 23 mM Tris-HCl, 2 mM Tris-Base, and 0.1% Tween 20. These membranes were blocked and probed with either anti-phospho (1:5000 diluted in 5% BSA/Tween 20 buffer) or anti-total (1:1000 diluted in 5% BSA/Tween 20 buffer) ERK1/2 antibodies overnight at 4°C. Nitrocellulose membranes were washed five times for 10-min intervals with Tween 20 buffer and incubated for 1 h at room temperature with a horseradish peroxidase-conjugated anti-rabbit IgG antibody (1:1000 diluted in 1% BSA-1% blotting grade nonfat dry milk/Tween 20 buffer). Nitrocellulose membranes were then washed five times for 10- min intervals with Tween 20 buffer, treated with West Pico enhancer and substrate solutions (Pierce, Rockford, IL) for 3 min to detect protein bands by chemiluminescence, and exposed to enhanced chemiluminescent Hyperfilm (Amersham Biosciences Inc., Piscataway, NJ). The film was developed and used for quantitation of the Western blots.

Western Blot Analysis.

A representative Western blot displaying the results from a single experiment and its analysis is shown in Fig. 1. Protein bands (Fig. 1B) were subjected to densitometry analysis (Molecular Analyst for Windows; Bio-Rad) and quantified as optical densitometry units. Each sample value was normalized by dividing the phospho-ERK1/2 densitometry volume by the total ERK1/2 densitometry volume. For each drug treatment, ERK1/2 stimulation over basal was calculated by dividing the treated normalized optical densitometry volume by the basal normalized optical densitometry volume. Each value was then expressed as percentage stimulation, with 100 μM agonist made equal to 100% stimulation of ERK1/2. Figure 1A shows a bar graph of the normalized percentage stimulation values calculated from the Western blot shown in Fig. 1B. Means ± S.E.M. of the data were calculated and were statistically compared using the one-sample t test. A p value less then 0.05 was used to indicate a significant difference between groups.

Results

Time Course and Concentration-Response Relationship for ERK1/2 Activation.

To determine the conditions that would provide maximal ERK1/2 stimulation, time course and concentration-response relationships of adrenergic agonist activation of ERK1/2 were investigated. Maximal ERK1/2 activation was reached at 10 min with 100 μM NE or phenylephrine (data not shown). Therefore, in the following experiments, artery segments were stimulated with 100 μM agonist for 10 min to maximize the amount of ERK1/2 activation.

Identification of the AR Types Causing Phenylephrine-Mediated ERK1/2 Activation.

To isolate the α₁-AR component of ERK1/2 activation, we used the α₁-AR-selective agonist phenylephrine to stimulate ERK1/2. To confirm that α₁-ARs were mediating phenylephrine activation of ERK1/2, we used selective antagonists of α₁-ARs (BE 2254), α₂-ARs (rauwolscine), and β-ARs (propranolol). BIAA segments were incubated with receptor type-selective concentrations of each antagonist for 1 h and then stimulated with 100 μM phenylephrine for 10 min. As shown in Fig.2, the α₁-AR antagonist BE 2254 significantly inhibited ERK1/2 activation, with 100 nM BE 2254 reducing ERK1/2 activity to basal levels. Rauwolscine and propranolol were unable to significantly inhibit ERK1/2 activation. These data demonstrate that phenylephrine increases ERK1/2 activity in the BIAA through stimulation of α₁-ARs.

Characterization of the α₁-AR Subtype Causing ERK1/2 Activation.

To identify the α₁-AR subtypes that mediate ERK1/2 activation, artery segments were incubated with receptor subtype-selective concentrations of the antagonists niguldipine, 5-methylurapidil, WB4101 (α_1A-AR-selective), cyclazosin (α_1B-AR-selective), and BMY 7378 (α_1D-AR-selective). Except for BMY 7378, concentrations of these antagonists up to 100 nM were used to ensure selectivity for receptor subtypes. When used individually, none of the antagonists inhibited phenylephrine stimulation of ERK1/2, suggesting that more than one α₁-AR subtype may be mediating this response (data not shown). Therefore, we incubated BIAA segments with different combinations of two of the α₁-AR subtype-selective antagonists niguldipine, cyclazosin, and BMY 7378. Incubating BIAA segments with two of these antagonists (100 nM each) did not cause a significant decrease in phenylephrine-induced ERK1/2 activation (Fig.3). However, addition of all three antagonists simultaneously, at 10 nM concentrations of each, caused a moderate decrease in phenylephrine-mediated ERK1/2 activation, whereas 100 nM concentrations caused a marked inhibition (Fig. 3). These data suggest that all three α₁-AR subtypes can couple to ERK1/2 activation in the BIAA.

Identification of the AR Types Causing NE-Mediated ERK1/2 Activation.

To identify the adrenergic receptor types mediating activation of ERK1/2 by the endogenous agonist NE, artery segments were incubated for 1 h with the adrenergic receptor type-selective antagonists BE 2254, rauwolscine, and propranolol. Segments were then stimulated with 100 μM NE for 10 min. As shown in Fig.4, a 100 nM concentration of the α₁-AR antagonist BE 2254 significantly inhibited NE-mediated ERK1/2 activation, reducing activation levels by 94 ± 1.3%. Rauwolscine was unable to significantly inhibit ERK1/2 activation by NE, whereas 1 μM propranolol caused a small, nonsignificant decrease in NE-induced ERK1/2 activation. These data demonstrate that NE increases ERK1/2 activity in the BIAA through stimulation of α₁-ARs, and that α₂-ARs or β-ARs do not appear to be involved in mediating this response.

Characterization of the α₁-AR Subtype Causing Contraction.

Mean concentration-response curves for adrenergic agonist-induced contraction were generated in BIAA rings and are shown in Fig. 5. NE contracted BIAA rings with low potency (EC₅₀ = 11 ± 2.0 μM). To determine whether extraneuronal uptake, neuronal uptake, or stimulation of β-ARs contributed to the low potency of NE, concentration-response curves were generated in the presence of hydrocortisone, desipramine, and propranolol. These drugs had no significant effect on the potency of NE, suggesting that there are little or no effects of extraneuronal uptake, neuronal uptake, or β-AR stimulation on NE-mediated contraction. Concentration-response curves were also generated for the α₁-AR subtype-selective agonist phenylephrine and the α₂-AR subtype-selective agonist UK 14304. Phenylephrine contracted the BIAA with slightly higher potency (EC₅₀ = 5.12 ± 1.8 μM), but lower intrinsic activity (0.87 ± 0.09) compared with NE, whereas UK 14304 did not cause contraction. These data suggest that stimulation of α₁-ARs causes contraction of the BIAA and that α₂-AR stimulation does not.

To identify the α-AR type causing contraction, we determined affinity values for α-AR antagonists in blocking NE-induced contraction. The α₂-AR antagonist rauwolscine had a low affinity in blocking NE contraction (Table 1), whereas the α₁-AR antagonist BE 2254 inhibited NE-mediated contraction with high affinity (Fig. 6; Table 1). Schild plots for BE 2254 (Fig. 6B) and rauwolscine gave pA₂ values of –9.07 and –5.85, respectively (Table 1). The high affinity of the α₁-AR antagonist BE 2254 and the low affinity of the α₂-AR antagonist rauwolscine suggests that NE contraction of BIAA is mediated through stimulation of α₁-ARs.

To determine the α₁-AR subtype causing contraction, pA₂ values were generated using a variety of subtype-selective antagonists. Mean pA₂ values and Schild slopes for each antagonist are listed in Table 1. The α_1D-AR-selective antagonist BMY 7378 inhibited contraction with extremely low affinity, suggesting that α_1D-ARs did not mediate contraction. The affinity of the α_1B-AR-selective antagonist cyclazosin was 10-fold lower than affinity values reported for binding to α_1B-ARs, but similar to those reported for binding to α_1A-ARs. The affinities of 5-methylurapidil and WB4101 were consistent with their reported affinities at the α_1A-AR subtype. Taken together, these data demonstrate that the α_1A-AR is mediating NE-induced contraction of the BIAA.

Comparison of the Potency of NE Activation of ERK1/2 and Contraction.

NE mean concentration-response curves for contraction and ERK1/2 activation were generated and are shown in Fig.7. The EC₅₀ for NE in stimulating ERK1/2 activation was 21 ± 23 μM. Increases in ERK1/2 activation over basal were first observed at 1 μM NE with maximal ERK1/2 stimulation at 100 μM NE. NE also stimulated contraction with low potency (EC₅₀ = 11 ± 2 μM), with responses first observed at 1 μM NE. Thus, the potency of NE in causing contraction and ERK1/2 activation was nearly the same.

Effect of TES on ERK1/2.

As shown in Fig.8, TES of BIAA segments resulted in increases in ERK1/2 activation. TES for 2, 5, and 10 min caused progressive stimulation of ERK1/2 by 2.1-, 3.0-, and 4.3-fold over basal, respectively. We also compared ERK1/2 activation produced by TES to ERK1/2 activation caused by 100 μM phenylephrine for 10 min. TES for 5 min produced similar levels of ERK1/2 activation (93 ± 24.3%) relative to phenylephrine, whereas after 10 min of TES, ERK1/2 activation was slightly greater (133 ± 41.2%) compared with phenylephrine. To determine whether α₁-ARs were mediating ERK1/2 activation produced by TES, artery segments were incubated with 100 nM concentrations of the α₁-AR-selective antagonist BE 2254 1 h before TES. BE 2254 inhibited 10-min TES ERK1/2 activation to basal levels, suggesting that α₁-ARs are mediating this response. Taken together, these data suggest that in the BIAA, sympathetic nerves release norepinephrine to stimulate α₁-ARs, resulting in activation of ERK1/2.

Discussion

In this study we compared adrenergic agonist stimulation of ERK activation and contraction in the BIAA. We identified the adrenergic receptor types that are stimulated by NE to cause contraction and ERK activation. Both the α₂-AR antagonist rauwolscine and the β-AR antagonist propranolol had no effect on NE or phenylephrine activation of ERK- or NE-induced contraction, whereas the α₁-AR antagonist BE 2254 significantly antagonized both responses, suggesting that α₁-ARs mediate both contraction and ERK activation in the BIAA. We also determined the α₁-AR subtypes mediating these responses in the BIAA, by using α₁-AR subtype-selective antagonists that can distinguish between the α_1A-, α_1B-, and α_1D-AR subtypes. We used a variety of α₁-AR subtype-selective antagonists to inhibit α₁-AR-mediated ERK activation and contraction. Using concentrations of antagonists reported to be selective for each subtype, we found that all three α₁-ARs appear to participate in the stimulation of ERK activation. However, affinity values generated for α₁-AR subtype-selective antagonists in blocking NE-induced contraction showed that only the α_1A-AR mediates this response. We also compared the relative potencies of NE in activating ERK and causing contraction, and found that NE activates each response with almost equal potency. In addition, TES of the BIAA caused activation of ERK through stimulation of α₁-ARs. These data suggest that in some vascular beds in vivo, the sympathetic nervous system activates ERK while also regulating vascular tone.

To understand the functional roles of α₁-ARs in the BIAA, we first identified the α₁-AR subtype mediating contraction. To do this, affinity values of α₁-AR subtype-selective antagonists in blocking contraction were determined and compared with their reported affinities at α₁-AR subtypes reported in previous studies. The α_1A-AR-selective antagonists 5-methylurapidil and WB4101 inhibited NE-induced contraction with high affinity. WB4101 had a slightly higher affinity then 5-methylurapidil, which is consistent with previous reports of the relative affinities of these compounds at the α_1A-AR (Hanft and Gross, 1989). The α_1B-AR-selective antagonist cyclazosin antagonized NE contractions with relatively low affinity (5 nM), similar to its affinity (11 nM) in inhibiting α_1A-AR-mediated contractions of rat small mesenteric artery (Stam et al., 1998). This result, combined with the high affinity of α_1A-AR-selective antagonists, suggests that the α_1B-AR does not mediate contraction of the BIAA. The α_1D-AR-selective antagonist BMY 7378 did not inhibit NE-induced contraction, indicating that the α_1D-AR does not contract the BIAA. This result is not surprising, as the α_1D-AR mediates contraction in only a few vascular tissues, including the rat (Piascik et al., 1995) and mouse (Yamamoto and Koike, 2001) aorta. The high affinity of WB4101 and 5-methylurapidil, and the low affinities of cyclazosin and BMY 7378 suggest that only α_1A-ARs mediate NE-induced contraction of the BIAA. This is consistent with other reports that α_1A-ARs are the primary receptor subtype mediating contraction of many blood vessels (Docherty, 1998).

We also identified the α₁-AR subtype mediating ERK activation in the BIAA, to determine whether the α_1A-AR receptor subtype is mediating both contraction and ERK activation or whether other α₁-AR subtypes are also involved in activation of ERK. We found that concentrations of individual α₁-AR subtype-selective antagonists reported to be selective for each subtype were unable to inhibit ERK activation. Combinations of two α₁-AR subtype-selective antagonists also failed to inhibit ERK activation. However, when using α_1A-, α_1B-, and α_1D-AR-selective antagonists together, ERK activation was inhibited significantly. These data suggest that more than one α₁-AR subtype is coupled to ERK activation in the BIAA. Previous studies have reported that all three α₁-AR subtypes can activate ERK when they are transfected separately into rat PC12 cells (Zhong and Minneman, 1999). However, few reports have attempted to determine the α₁-AR subtype that activates ERK in native tissues that endogenously express α₁-ARs. In contrast to our findings, Xin et al. (1997) found that only the α_1D-AR stimulated ERK in rat aorta vascular smooth muscle cells. However, that study used a primary cell culture of aortic smooth muscle cells, which did not contain the various cell types found in an intact blood vessel (i.e., smooth muscle cells, fibroblasts, macrophages, monocytes, nerves). α₁-ARs are present in both the medial (primarily smooth muscle cells) and adventitial (primarily fibroblasts and other cell types) layers of arteries (Faber et al., 2001). Thus, α_1B- and α_1D-ARs may mediate ERK activation in the fibroblast-containing adventitial layer, whereas the α_1A-AR may mediate ERK activation in the smooth muscle cell-containing medial layer of the BIAA. These data may explain why mRNA (Scofield et al., 1995; Faber et al., 2001) and protein (Piascik et al., 1997; Hrometz et al., 1999; Faber et al., 2001) for each α₁-AR subtype can be detected in most blood vessels, but only one α₁-AR subtype is typically responsible for mediating contraction. It may be possible that α₁-ARs that are expressed but not linked to contraction can activate ERK or other signaling pathways in the supporting adventitial layer of blood vessels.

We also compared the potencies of NE in causing contraction and ERK activation. We found that NE had nearly the same potency for causing contraction of the BIAA as it does for activating ERK, suggesting that in vivo, the degree of ERK activation is directly proportional to the level of vasoconstriction. Furthermore, we found that TES resulted in activation of ERK through the stimulation of α₁-ARs. These data suggest that during moment-to-moment control of vas-cular tone, sympathetic nerves release NE to cause contraction and to activate ERK. However, the physiological role of α₁-AR-mediated ERK activation in vascular tissues is unknown. Numerous in vivo and in vitro studies have suggested that the sympathetic nervous system regulates mitogenic responses in vascular tissues. For example, increased sympathetic innervation (Head, 1991) or exogenous elevation of plasma catecholamines (Stewart et al., 1992) has been associated with vascular smooth muscle hypertrophy in vivo. In addition, sympathetic denervation of blood vessels through mechanical (Bevan, 1975), chemical (Fronek et al., 1978), or immunological (Lee et al., 1987) methods results in decreases in DNA synthesis and cell number. The addition of catecholamines to vascular smooth muscle cell cultures increases DNA (Nakaki et al., 1990) and protein synthesis (Chen et al., 1995; Hu et al., 1999), as well as cell growth (Blase and Boissel, 1993). Previous reports have also suggested that ERK is involved in regulating agonist-induced vasoconstriction. For example, inhibition of MAPK activation has been shown to attenuate not only 5-hydroxytryptamine_2A (Florian and Watts, 1998; Banes et al., 1999) but also angiotensin (Epstein et al., 1997)- and α₁-AR (DiSalvo et al., 1993) mediated vasoconstriction. Possible mechanisms by which ERK activation has been proposed to affect contractile responses include phosphorylation of caldesmon (Adam et al., 1989; Hedges et al., 2000) or myosin (Jin et al., 1996). Although evidence exists for ERK-regulating mitogenic responses and contractile function in blood vessels, the role of α₁-AR-stimulated ERK activation may be dependent on cell type. For example, α₁-AR activation of ERK in smooth muscle cells within the medial layer may regulate contraction, whereas α₁-AR activation of ERK in fibroblasts within the adventitial layer may be involved in vascular remodeling.

In conclusion, we found that only α₁-ARs mediate contraction and ERK activation in the BIAA. Activation of the α_1A-AR subtype causes contraction, whereas α_1A-, α_1B-, and α_1D-ARs can all activate ERK. The endogenous adrenergic agonist NE was equipotent in stimulating both contraction and ERK activation. Our data also suggest that ERK is activated following sympathetic stimulation of α₁-ARs in intact blood vessels. The activation of ERK in the BIAA may play a role in modifying α₁-AR-induced contraction, or ERK may be responsible for regulating growth or other mitogenic pathways in vascular smooth muscle cells or in other cell types found in intact blood vessels. The human inferior alveolar artery is the primary source of blood supply to the mandible and mandibular teeth. During aging, this artery slowly degenerates until it no longer contributes the major source of blood to the mandible, teeth, and oral tissues (Zoud and Doran, 1993). The human inferior alveolar artery also develops luminal narrowing much earlier in life than other cranial arteries (Bradley, 1975). One explanation for these changes is that vascular remodeling or arteriosclerotic narrowing of the inferior alveolar artery develops due to excessive growth factor pathway activation. Thus, sympathetic nervous system activation of ERK via α₁-ARs may be important for regulation of normal blood vessel growth and/or contraction as well as abnormal growth and remodeling associated with aging or other conditions.