Benzodiazepines Differentially Inhibit Phenylephrine-induced Calcium Oscillations in Pulmonary Artery Smooth Muscle Cells 

Pulmonary arteries were isolated from adult male mongrel dogs. The technique of euthanasia was approved by the Institutional Animal Care and Use Committee. All procedures were performed aseptically using general anesthesia (fentanyl, 15 micro gram/kg; pentobarbital sodium, 30 mg/kg, given intravenously), an endotracheal tube, and positive pressure ventilation. A catheter was placed in the right femoral artery, the mobilizable blood volume was removed, and 30 ml of saturated KCl was administered intravenously. A left thoracotomy was performed through the fifth intercostal space. The heart and lungs were removed en bloc, and the pulmonary arteries were isolated and dissected in a laminar flow hood under sterile conditions.

Primary cultures of PASMCs were obtained from segmental and subsegmental intralobar branches of pulmonary artery (the third and fourth generation of branches from the main pulmonary artery) having diameters < 4 mm. Explant cultures were prepared according to the method of Campbell and Campbell [13]with minor modifications. The endothelium was removed by gentle rubbing with a sterile cotton swab. The tunica adventitia was carefully removed together with the most superficial part of the tunica media. The remaining tunica media was cut into 2-mm²pieces that were explanted in 25-cm²culture flasks. The explants were nourished by D-MEM/F-12 medium (Gibco, Grand Island, NY) containing 10% fetal bovine serum and 1% antibiotic anti-mycotic mixture solution (10,000 U/ml penicillin, 10,000 micro gram/ml streptomycin, and 25 micro gram/ml of amphotericin B) and kept in a humidified atmosphere of 5% CO²:95% air at 37 [degree sign] Celsius. Pulmonary artery smooth muscle cells began to proliferate from explants after 7 days in culture. Cells were allowed to proliferate for an additional 7–10 days until subconfluence was achieved. Cells were then subcultured nonenzymatically to 35-mm culture dishes specially designed for fluorescence microscopy (Bioptechs Inc., Delta T system, Butler, PA). Pulmonary artery smooth muscle cells were used for experimentation within 72 h after subculture. Cells from the first and second passage were used for experiments. These cells were routinely identified as smooth muscle cells using a fluorescein-labeled antibody directed at smooth muscle alpha actin (Sigma Chemical Co., St. Louis, MO).

Twenty-four hours before experimentation, the culture medium containing 10% fetal bovine serum was replaced with serum-free medium to arrest cell growth, to allow for establishment of steady-state cellular events independent of cell division, and to prevent a false estimate of [Ca²+]ⁱbecause of binding of available dye to serum proteins in the medium. [14]Pulmonary artery smooth muscle cells were washed twice in loading buffer, which contained (in mm) 125 NaCl, 5 KCl, 1.2 MgSO⁴, 11 glucose, 1.8 CaCl², 25 HEPES, and 0.2% bovine serum albumin at pH 7.4 adjusted with NaOH. Pulmonary artery smooth muscle cells were then incubated in loading buffer containing 2 micro meter fura-2/AM, the acetoxymethyl derivative of fura-2 (Molecular Probes, Eugene, OR), at ambient temperature for 30 min. After the 30-min loading period, the cells were washed twice in loading buffer and incubated at ambient temperature for an additional 20 min before study. This provided sufficient time to wash away any extracellular fura-2/AM and for intracellular esterases to cleave fura-2/AM into the active fura-2. [15] 

Culture dishes containing fura-2-loaded PASMCs were placed in a temperature-regulated (37 [degree sign] Celsius) chamber (Bioptechs, Inc., Butler, PA) mounted on the stage of an Olympus IX-70 inverted fluorescence microscope (Olympus America Inc., Lake Success, NY). Fluorescence measurements were performed on individual PASMCs on a cultured monolayer using a dual wavelength spectrofluorometer (Deltascan RFK6002; Photon Technology International, So. Brunswick, NJ) at excitation wavelengths of 340 and 380 nm and an emission wavelength of 510 nm. The volume of the chamber was 1.5 ml. The cells were superfused continuously at 1 ml/min with Krebs-Ringer buffer, which contained (in mm) 125 NaCl, 5 KCl, 1.2 MgSO⁴, 11 glucose, 2.5 CaCl²and 25 HEPES at pH 7.4 adjusted with NaOH. The temperature of all solutions was maintained at 37 [degree sign] Celsius in a water bath. Just before data acquisition, background fluorescence (i.e., fluorescence between cells) was measured and subtracted automatically from the subsequent experimental measurements. Fura-2 fluorescence signals (340, 380, and 340/380 ratio) originating from single PASMCs were continuously monitored at a sampling frequency of 25 Hz and were collected using a software package from Photon Technology International (Felix[trademark symbol]).

Estimates of [Ca²+]ⁱwere made by comparing the cellular fluorescence ratio with ratios acquired using fura-2 (free acid) in buffers containing known Ca²+ concentrations. [Ca²+] sub i was then calculated as described by Grynkiewicz et al. [16] 

Pulmonary artery smooth muscle cells were pretreated with the beta-adrenoreceptor antagonist propranolol (5 micro meter) to eliminate any beta-agonist effect of phenylephrine. The effects of increasing doses of phenylephrine on the amplitude and frequency of [Ca²+]ⁱwere first assessed. In subsequent experiments, after establishment of steady-state [Ca²+]ⁱoscillations induced by 10 micro Meter phenylephrine, lorazepam, diazepam, or midazolam was added to the superfusate in the continued presence of phenylephrine. Any given PASMC was exposed to only one agent. Lorazepam (2 mg/ml stock) was administered at concentrations ranging from 0.3–100.0 micro Meter, diazepam (5 mg/ml stock) at 3–300 micro Meter, and midazolam (5 mg/ml stock) at 1–100 micro Meter with 5-min intervals between doses. The changes in solution were accomplished by rapidly aspirating the buffer in the dish and transiently increasing the flow rate to 10 ml/min. In a separate series of experiments, the vehicle for diazepam (1 ml contains 0.4 ml propylene glycol, 0.1 ml alcohol, 0.015 ml benzyl alcohol and water) or lorazepam (1 ml contains 0.18 ml polyethylene glycol 400 in propylene glycol with 0.02 ml benzyl alcohol) was added to the superfusate as a vehicle control at a concentration corresponding to 300 micro Meter.

The amplitude and frequency of phenylephrine-induced [Ca²+]ⁱoscillations were measured in individual PASMCs. The amplitude was calculated by averaging the peak ratio obtained for 4–5 oscillations before and after each intervention. The change in the 340/380 fluorescence ratio was then calculated by subtracting the resting ratio value (baseline). The frequency of oscillations was calculated by averaging the time interval between the oscillation peaks and is reported as the number of oscillations observed per minute. Changes in the amplitude and frequency of the 340/380 fluorescence ratio in response to the benzodiazepines are expressed as percent of control, in which control equals the amplitude and frequency of oscillations induced by 10 micro Meter phenylephrine (without benzodiazepines). Data are expressed as means +/- SEM. Statistical analysis was performed using repeated-measures analysis of variance followed by Bonferroni/Dunn post hoc testing. Differences were considered statistically significant at P < 0.05.

Phenylephrine and propranolol were purchased from Sigma Chemical Co. Lorazepam (Ativan[registered sign]) was obtained from Wyeth Laboratories Inc. (Philadelphia, PA), diazepam (Valium[registered sign]) from ESI Pharmaceuticals (Cherry Hill, NJ), and midazolam (Versed[registered sign]) from Roche Laboratories (Nutley, NJ). Vehicle for diazepam and lorazepam was obtained from The Cleveland Clinic Pharmacy (Cleveland, OH).

A typical trace depicting the effect of phenylephrine (10 micro Meter) on [Ca²+]ⁱin a single PASMC is shown in Figure 1. Resting values of [Ca²+]ⁱwere 68 +/- 8 nM. Peak [Ca²+]ⁱachieved with 10 micro meter phenylephrine was 676 +/- 35 nM at a frequency of 1.08 +/- 0.1 transients/min. As previously noted, [11,12]tachyphylaxis to phenylephrine was never observed, and oscillations maintained a constant amplitude and frequency for > 30 min. Summarized data depicting the dose-dependent effects of phenylephrine on the amplitude and frequency of [Ca²+]ⁱoscillations are also shown in Figure 1(n = 19 cells). The lowest concentration of phenylephrine that induced [Ca²+]ⁱoscillations was 3 x 10 sup -8 M, achieving a peak [Ca²+]ⁱconcentration of 198 +/- 19 nM at a frequency of 0.31 +/- 0.1 transients/min. These results were defined as the control response, and the amplitude and frequency were normalized to 100%. The frequency and amplitude of phenylephrine-induced [Ca²+]ⁱoscillations were similar in first- and second-passage PASMCs.

(Figure 2) shows a typical trace demonstrating the dose-dependent inhibitory effect of lorazepam on peak [Ca²+]ⁱin a single PASMC. Lorazepam (1 micro Meter) exerted a significant inhibitory effect (P < 0.05) on the amplitude (87 +/- 3% of control) but not the frequency (96 +/- 4% of control) of [Ca²+]ⁱoscillations. The inhibitory effect of lorazepam on the amplitude of [Ca sup 2+]ⁱoscillations was dose-dependent. Figure 2summarizes the dose-dependent effects of lorazepam on both the amplitude and frequency of phenylephrine-induced [Ca²+]ⁱoscillations (n = 13 cells). Only the highest concentration of lorazepam (100 micro Meter) decreased (P < 0.05) the frequency (77 +/- 13% of control) of [Ca²+]ⁱoscillations. The vehicle for lorazepam had no effect on [Ca²+]ⁱoscillations induced by phenylephrine (n = 5 cells).

(Figure 3) shows a typical trace demonstrating the dose-dependent inhibitory effect of diazepam on the frequency of [Ca²+]ⁱoscillations in a single PASMC. Figure 3also summarizes the dose-dependent inhibitory effect of diazepam on the amplitude and frequency of [Ca²+]ⁱoscillations induced by phenylephrine. Diazepam (0.3 micro Meter) reduced (P < 0.05) the frequency (86 +/- 4% of control) but not the amplitude (98 +/- 2% of control) of [Ca²+] sub i oscillations (n = 14 cells). Higher concentrations of diazepam (100 and 300 micro Meter) resulted in a decrease (P < 0.05) in the amplitudes (87 +/- 5% and 45 +/- 9% of control, respectively) and frequencies (87 +/- 8% and 46 +/- 13% of control, respectively) of [Ca sup 2+]ⁱoscillations. The vehicle for diazepam had no effect on [Ca²+]ⁱoscillations induced by phenylephrine (n = 5 cells).

(Figure 4) summarizes the effects of midazolam on [Ca²+]ⁱoscillations induced by phenylephrine (n = 12 cells). Unlike lorazepam or diazepam, midazolam (30 micro Meter) had no significant effect on the amplitude (97 +/- 6% of control) or frequency (94 +/- 6% of control) of [Ca²+]ⁱoscillations. Only the highest concentration of midazolam (100 micro Meter) resulted in a decrease (P <0.05) in both amplitude (75 +/- 10% of control) and frequency (69 +/- 9% of control) of [Ca²+]ⁱoscillations.

The current study demonstrates that benzodiazepines exert dose-dependent inhibitory effects on the amplitude and frequency of phenylephrine-induced [Ca²+]ⁱoscillations in individual PASMCs. Both lorazepam and diazepam inhibited the phenylephrine-induced [Ca²+]ⁱoscillations in a differential manner. Lorazepam inhibited the amplitude of the [Ca²+]ⁱoscillations, whereas diazepam inhibited the frequency of the [Ca²+]ⁱoscillations. In contrast, the effects of midazolam were only apparent at very high concentrations.

Activation of sympathetic alpha-adrenoreceptors causes vascular smooth muscle contraction by increasing [Ca²+]ⁱ. [17]The alpha-adrenoreceptor is coupled via a G protein to phospholipase C, which hydrolyzes polyphosphatidylinositol 4,5-bisphosphate. The breakdown of polyphosphatidylinositol 4,5-bisphosphate leads to the generation of cytosolic inositol trisphosphate and diacylglycerol. [18]Inositol trisphosphate stimulates the release of Ca²+ from intracellular stores initiating contraction, whereas diacylglycerol, which stimulates protein kinase C, [19]is involved in the maintenance of sustained contractions. Several smooth muscle cell types [11,20,21]have been shown to exhibit agonist-induced oscillations in [Ca²+]ⁱwith dose-dependent amplitudes and frequencies. This suggests that [Ca²+]ⁱmay exert effects on smooth muscle tone through a frequency-dependent mechanism. We recently demonstrated that cultured and freshly dispersed individual PASMCs oscillate [Ca²+] sub i in response to alpha-adrenoreceptor activation with phenylephrine. [11]These [Ca²+]ⁱoscillations were dependent on the presence of extracellular Ca²+ but did not require activation of voltage-gated Ca²+ channels. In addition, the [Ca²+]ⁱoscillations required activation of phospholipase C and involved the release of Ca²+ from caffeine-sensitive intracellular stores. [11]The frequency of [Ca²+]ⁱoscillations may represent a digitization of the Ca²+ signal, allowing frequency-dependent control of the contractile response for “fine tuning” vessel tone. [22] 

In vivo studies suggest that concentrations of benzodiazepines used to induce surgical anesthesia can produce significant hemodynamic alterations in the systemic and pulmonary circulations. [10,23]Our results are consistent with other in vitro studies in which benzodiazepines attenuated alpha-adrenoreceptor-mediated increases in vessel tone [24]and caused relaxation of airway smooth muscle. [6]The vasodilator effects of benzodiazepines are thought to be mediated, at least in part, via inhibition of voltage-dependent Ca²+ currents in vascular smooth muscle cells. [4,24,25]It is noteworthy that midazolam had no effect on pulmonary vasomotor tone in an isolated rat lung model [9]or in patients undergoing cardiac surgery. [26]Those results are consistent with our findings in PASMCs that midazolam inhibited phenylephrine-induced [Ca²+]ⁱoscillations only at the highest concentration tested. Therefore, use of midazolam in the clinical setting could be an advantage compared with other benzodiazepines in that it may not alter pulmonary vascular reactivity.

In the intact blood vessel, the vasodilating effects of benzodiazepines could be due to inhibition of voltage-gated Ca²+ channels preventing Ca²+ entry in the smooth muscle cells. In the current study, however, it is unlikely that the inhibitory mechanism of action of lorazepam or diazepam involves an interaction with voltage-gated Ca²+ channels, because we previously demonstrated that the [Ca²+]ⁱoscillations induced by phenylephrine were not dependent on activation of voltage-gated Ca²+ currents. [11]Lorazepam, however, may attenuate the amplitude of the [Ca²+]ⁱ, oscillations either by inhibiting influx of extracellular Ca²+ via capacitance-gated Ca²+ channels that open when intracellular Ca sup 2+ stores are depleted or by reducing the amount of Ca²+ released from intracellular stores. The inhibitory effect of diazepam on the frequency of the [Ca²+]ⁱoscillations is likely to be more complex and may involve alterations in the binding of phenylephrine to alpha-adrenoreceptors via competitive, noncompetitive, or allosteric interactions. Alternatively, diazepam may alter more distal points in the signal transduction pathway for phenylephrine-induced [Ca²+] sub i oscillations, such as the coupling between alpha-adrenoreceptor activation and the rate of inositol trisphosphate production [27]or binding to its receptor. It is also possible that some actions could be due to nonspecific effects of the benzodiazepine on the cell surface membrane. Additional experiments are required to address these possibilities.

The clinically relevant free serum concentrations of benzodiazepines have been estimated at [nearly =] 1 micro meter for diazepam, midazolam, and lorazepam. [28–31]Clinically relevant concentrations of anesthetic agents in vivo, however, are difficult to ascertain and can vary over time depending on the speed of injection, volume of distribution, plasma protein concentration, and pH. All of the benzodiazepines used in this study avidly bind to serum proteins (> 90%), which significantly reduces the free level of drug in the serum capable of interacting with the tissue. Protein binding in vivo, however, is unlikely to be instantaneous, so free drug concentrations with a bolus induction dose are likely higher than those measured at equilibrium. In addition, first-pass concentrations in the pulmonary circulation are much higher than steady-state levels in the systemic circulation, [9]and the microkinetic behavior within the vascular space (drug transfer rate from serum to protein-bound state and cellular constituents) has not been defined. Despite the difficulty and uncertainty in estimating the in vivo concentrations and the likelihood that these estimates are different in different pathologic conditions (e.g., hemodilution, liver disease, hypoproteinemia, hypovolemia), it seems unlikely that the concentrations of lorazepam and diazepam that had significant effects on [Ca²+]ⁱoscillations would be routinely encountered in clinical practice.

Lorazepam and diazepam but not midazolam have differential effects on phenylephrine-induced [Ca²+]ⁱoscillations in PASMCs. The precise cellular mechanisms that mediate these effects of benzodiazepines on [Ca²+]ⁱsignaling in PASMCs remain to be elucidated.

The authors thank Ronnie Sanders for outstanding work in preparing the manuscript.