Propofol Stimulates Ciliary Motility via  the Nitric Oxide–Cyclic GMP Pathway in Cultured Rat Tracheal Epithelial Cells

Preparation of cultured rat tracheal epithelial cells was performed according to the methods previously described by Dirksen et al.  24The study protocol was approved by the animal research committee of The Johns Hopkins University School of Medicine. Twelve adult, male Sprague-Dawley rats (body weight, 300–350 g; Hilltop, Scottdale, PA) were anesthetized with use of halothane, and the thoracic cage was rapidly opened. The trachea was immediately removed and washed with Hank balanced salt solution (GibcoBRL, Grand Island, NY) with 25 mm N-(2-hydroxy-ethyl)-1-piperazineethanesulfonic acid (HEPES; GibcoBRL), pH 7.4. The excised tissues were cut into small pieces (0.5–1 mm²) and placed onto glass plates coated with rat tail collagen. The plates were incubated at 37°C in a humidified carbon dioxide incubator (95% air–5% carbon dioxide). Culture medium was a Dulbecco modified Eagle medium (DMEM; GibcoBRL) supplemented with 0.37% (wt/vol) NaHCO³, 10% fetal calf serum (GibcoBRL), 100 U/ml penicillin, 100 μg/ml streptomycin, 0.25 μg/ml amphotericin B (Sigma, St. Louis, MI), and 25 mm HEPES.

After 3–5 days of incubation, the culture plate was mounted on a phase-contrast microscope (IMT-2; Olympus, Tokyo, Japan). The observation medium was DMEM with 0.037% (wt/vol) NaHCO³supplemented with 10% fetal calf serum, 100 U/ml penicillin, 100 μg/ml streptomycin, 0.25 μg/ml amphotericin B, and 25 mm HEPES. The medium was adjusted to and maintained at pH 7.4 ± 0.1 at a constant temperature (24 ± 1°C) during the experimental period. If any bacterial–fungal contamination was found or if there were no more than three cells with actively beating cilia (defined as CBF greater than 6.0 beats/s) within a 400×-power observation field, the plate was discarded.

Propofol (RBI, Natick, MA), l-NMMA (Alexis, San Diego, CA), ODQ (Alexis) and KT5823 (Carbiochem, San Diego, CA) were dissolved in dimethyl sulfoxide (DMSO; Sigma) in phosphate-buffered saline and diluted appropriately; final DMSO concentration in the medium was 0.1%. Intralipid (KabiVitrum, Wilmington, DE), the propofol vehicle in clinical use, was diluted in the observation medium to a final concentration 0.05%.

The cells were viewed at 400× magnification. All observations were monitored and recorded for analysis using a 3CCD color videocamera (IK-TU40A; Toshiba, Tokyo, Japan), an S-VHS video cassette recorder (HR-S5400U; JVC, Tokyo, Japan), and a Trinitron color monitor (CVM-1271; Sony, Tokyo, Japan). After a 5-min control recording period, propofol, vehicle, or a blocker were added at random to the medium, and the image was recorded for the subsequent 25 min. To determine CBF, video images were later captured at 30 frames/s and digitized using a Macintosh computer (Apple Computer Inc., Santa Clara, CA) and Adobe Premiere software (Adobe, San Jose, CA). The images of cilia were viewed frame by frame by a researcher who was blind to the drug preparation administered. CBF was counted manually at least 3 times/cell and averaged. The average was regarded as the CBF value of the single cell. The CBF value of one plate was the average of values for at least three independent cells.

All values were expressed as mean ± SD. Values of n  represent the number of the plates. Comparisons of trends over time of two groups were performed with use of two-factor repeated-measures analysis of variance. Time-matched values in the groups were compared using the Bonferroni test after one-way analysis of variance. A P  value less than 0.05 was considered to be statistically significant.

Propofol at a dose of 100 μm (n = 10) increased CBF significantly (P < 0.0001), from 7.3 ± 0.7 beats/s at time 0 (baseline) to 8.4 ± 0.8 beats/s 25 min after the administration, compared with vehicle administration (0.1% DMSO, n = 11; 7.4 ± 0.6 beats/s at baseline to 7.4 ± 0.6 beats/s at 25 min;fig. 1A). The effect of propofol plateaued at 15–25 min postadministration.

Exposure to 0.05% intralipid (n = 5), the clinical vehicle of propofol, did not affect CBF (−0.2 ± 2.2% increase from baseline; baseline = 100%;fig. 1B). Observation medium (DMEM) alone also did not change CBF (−0.3 ± 3.3%).

Propofol increased CBF in a dose-dependent manner (0 mm or vehicle, 0.2 ± 1.7% increase from baseline, n = 11; 1 μm, 2.0 ± 1.9%, n = 6; 10 μm, 8.2 ± 6.7%, n = 9; 30 μm, 13.0 ± 7.1%, n = 8; 100 μm, 14.0 ± 4.7%, n = 10; 300 μm, 13.8 ± 6.4%, n = 9;fig. 2). The effect of propofol plateaued at doses of 30–300 μm.

Enhancement of CBF by 100 μm propofol was significantly and dose-dependently inhibited by coadministration of l-NMMA (0.1 mm, 11.0 ± 2.9%, n = 4; 1 mm, 5.6 ± 7.7%, n = 10; 10 mm, 2.4 ± 3.6%, n = 5;fig. 3). Ten millimoles of l-NMMA alone did not affect CBF (−0.3 ± 3.3%; n = 6).

The enhancement of CBF by 100 μm propofol was significantly and dose-dependently inhibited by coadministration of ODQ (1 μm, 14.0 ± 5.9%, n = 4; 10 μm, 8.0 ± 4.8%, n = 4; 100 μm, −0.3 ± 2.2%, n = 6;fig. 4). One hundred micromoles of ODQ completely abolished the propofol-induced CBF enhancement. ODQ alone did not affect CBF (−0.5 ± 2.2%; n = 6).

Enhancement of CBF by 100 μm propofol was significantly and dose-dependently inhibited by coadministration of KT5823 (0.03 μm, 13.8 ± 3.9%, n = 5; 0.3 μm, 9.0 ± 4.2%, n = 5; 3 μm, 5.0 ± 3.5%, n = 5; 30 μm, −0.1 ± 4.1%, n = 8;fig. 5). Thirty micromoles of KT5823 totally inhibited the propofol-induced CBF enhancement. KT5823 alone did not affect CBF (−1.3 ± 4.1%; n = 7).

In this study, we demonstrated that propofol stimulates CBF in a dose-dependent fashion and suggest that the NO–cGMP signaling pathway is involved in its stimulation in cultured rat tracheal epithelial cells (fig. 6).

The plasma concentrations of propofol reported in humans and rats during sedation or anesthesia are approximately 2–10 μg/ml, or 10–50 μm. 25–27These doses of propofol stimulate CBF in our in vitro  setting. Because propofol in blood is more than 95% protein-bound in rats and humans, 28–30actual concentrations of the drug in tissues are uncertain, but probably are much less than the apparent plasma concentrations. Although we performed our experiment using the serum protein–contained medium, we cannot eliminate the possibility that the concentrations that stimulate CBF in vitro  are different from those that can stimulate CBF in vivo .

Raphael and Butt 31demonstrated that CBF of nasal tissues in patients anesthetized using propofol and alfentanil did not change, whereas isoflurane anesthesia decreased CBF. 31His group also reported that propofol at a dose of 70 μm did not change CBF in human nasal turbinate explants. 32The discrepancy between their results and ours may be, at least in part, the result of different experimental conditions and use of tissues from different species. We used rat tracheal tissues after 3–5 days of culture, whereas they used human nasal tissues on the same day of harvest.

Inhalation anesthetics such as halothane and isoflurane have been shown to depress ciliary function, 2–5,31although short exposure (< 2 min) stimulates CBF. 33,34Other anesthesia-related drugs, such as local anesthetics, barbiturates, and benzodiazepines, also inhibit CBF. 35–38Patients with airway diseases who have impaired ciliary function are at greater risk of postoperative pulmonary complications. Anesthetic or sedative drugs that have an accelerating effect on CBF are advantageous in patients with respiratory risk. Considering the limitations of our model, additional studies, including a series of experiments for in vivo  measurements, are needed to verify our observations.

We and others have reported that endothelial and inducible NOS (eNOS and iNOS), sGC, and PKG-I are localized in ciliated airway epithelia. 8–10Nasal NO con- centration correlates with ciliary functions. 7NO or cGMP production from airway epithelium has also been reported. 8,12NOS inhibitors decrease CBF after prestimulation with isoproterenol, substance P, bradykinin, tumor necrosis factor α, interleukin-1β, l-arginine, and ethanol. 11–188-Bromo-cGMP activates PKG and increases CBF. 8,18PKG inhibitors significantly inhibit l-arginine–, nitroprusside-, or isoproterenol-induced enhancement of CBF. 8,18Thus, these studies indicate that ciliary motility is regulated by the NO–cGMP pathway in an autocrine–paracrine fashion. Petros et al.  20reported that propofol stimulated cGMP formation in cocultured porcine endothelial and smooth muscle cells, and this stimulation was inhibited by a NOS inhibitor. Liu et al.  21demonstrated that propofol increased cGMP content in cultured bovine vascular smooth muscle cells, and this increase was inhibited by sGC inhibitors. These results raise the possibility that propofol stimulates NO and cGMP production through activation of NOS and sGC in the airway. Presently, we cannot say which type of cell produces NO by propofol stimulation because, in addition to ciliated epithelial cells, there are other kinds of cells, including nonciliated epithelial cells, smooth muscle cells, and fibroblasts, in our cultured tracheal tissues. 24Miyawaki et al.  39reported that propofol inhibited acetylcholine-stimulated cGMP formation in rat aortic strips, indicating a need for further studies to clarify the propofol actions on NO–cGMP formation not only in airways, but also in the vasculature.

The intracellular free-calcium concentration ([Ca²⁺]i) is crucially involved in the regulation of CBF. 2Elevation of [Ca²⁺]i increases CBF, whereas the decrease of [Ca²⁺]i causes CBF to slow. Recently, Uzlaner and Priel 17demonstrated that dibutylyl–cGMP alone did not stimulate CBF in cultured rabbit tracheal epithelial cells when [Ca²⁺]i was not increased, whereas it could elevate CBF strongly when [Ca²⁺]i was increased by ionomycin or adenosine triphosphate (ATP), suggesting that elevated [Ca²⁺]i is necessitated for the NO–cGMP system-induced enhancement of CBF. Elevated [Ca²⁺]i also activates endothelial NOS and the NO–cGMP pathway. 6,40There is a possibility that propofol increases [Ca²⁺]i in airway epithelial cells. Propofol is reported to increase [Ca²⁺]i in cultured rat embryonic brain cells and human glial cells, 41whereas it reduced [Ca²⁺]i in cardiomyocytes 42and vascular smooth muscle cells. 43The discrepancy between results obtained by Raphael and Butt 31and Hann et al.  32and ours are possibly attributable to a difference in [Ca²⁺]i condition. Further studies are needed to investigate the effect of propofol on [Ca²⁺]i in airway epithelial cells.

The current study clearly shows that propofol stimulates CBF via  the NO–cGMP pathway in cultured rat tracheal ciliated epithelial cells. Although this study suggests a possible advantage for using propofol in patients with respiratory risk, further studies, including clinical trials, are necessary.