Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Mucociliary clearance is one of the lung's nonspecific host-defense mechanisms, carrying both trapped inhaled materials and locally produced biologic debris from the respiratory tract toward the oropharynx. This transport function depends on the amounts and rheologic properties of mucus, and on the beating action of cilia propelling the overlying secretions. The cellular mechanisms that regulate ciliary beating of airway epithelium may thus be considered as mechanisms regulating mucociliary clearance, and there is ample evidence that the ciliary beat frequency (CBF) can be influenced by a variety of neurotransmitters, hormones, and metabolic regulators (1).

Recent studies have shown that airway epithelial functions may be regulated through an autocrine mechanism. For example, airway epithelial cells synthesize and release eicosanoids (2, 3), platelet-activating factor (PAF) (4), endothelin (5), and nitric oxide (NO) (6), and these substances are capable of altering epithelial electrolyte transport and/or ciliary motility (7).

Arginine vasopressin (AVP) is a nonapeptide released from the posterior pituitary into the systemic circulation. This hormone not only acts as an antidiuretic hormone, but also produces potent vasoconstriction and cell proliferation (11). Rennick and colleagues (12, 13) have shown that AVP-like immunoreactivities exist in and are released from rabbit tracheal epithelium, but the physiologic significance of AVP in the regulation of airway epithelial function is unknown. Therefore, our present study with cultured rabbit tracheal epithelium was designed to: (1) determine whether AVP affects ciliary activity, as assessed from CBF, and if so, the mechanism by which it does this; and (2) to determine the AVP receptor subtype involved in the epithelial effect of this hormone.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Tissue Preparation

Japanese white rabbits weighing 2.0 to 2.4 kg were anesthetized with intravenous sodium pentobarbital (35 mg/ kg). The trachea was removed and its mucosa was dissected free from the underlying connective tissue, cut into small pieces (1 to 2 mm²), rinsed several times with Dulbecco's phosphate-buffered saline (D-PBS), and placed on a coverglass (18 × 24 mm) coated with human placental collagen (5.8 µg/cm²; Sigma Chemical Co., St. Louis, MO) in a Petri dish. Tissues were then incubated in medium-199 containing 5% fetal calf serum (FCS), 100 U/ml penicillin, 100 µg/ml streptomycin, and 50 µg/ml gentamicin at 37°C in a CO₂ incubator (5% CO₂-95% air). On the fifth day of incubation, the coverglass on which the cultured explant was adhered was mounted in a Rose chamber for the measurement of CBF.

Measurement of CBF

The photoelectric method for the measurement of CBF has been described in detail elsewhere (14). Briefly, the photometer that we used for this procedure (NFX-2; Hamamatsu Photonics, Hamamatsu, Japan), with a built-in periplanatic eyepiece, a limiting aperture, and lateral focusing telescope, was attached to the head of a microscope with a phase-contrast condenser and an on-base type of illuminator (Optiphoto-XF; Nikon, Tokyo, Japan). Because of the beating action of cilia, light from the illuminator passed through the preparation in varying intensities. These variations of light intensity were detected by the photometer and transduced to voltage impulses, which were amplified, displayed on an oscilloscope (LBO-522; Leader, Tokyo), and recorded by a pen recorder (VP-6213; Panasonic, Tokyo).

The same group of cilia in each preparation was studied throughout the experiment (i.e., the relative position of the field diaphragm of the photometer to the selected epithelial border was kept constant). The longitudinal axis of the photometer field was oriented perpendicularly to the cell border, and CBF measurements were taken from clumps of two or more cells with a free border devoid of debris. The choice of ciliated area was randomized by using an eyepiece graticule and selecting the group of cilia that lay closest to the crosspiece of the graticule on the equatorial line. After a 30-min period of equilibration at 37°C in Hanks' balanced salt solution (HBSS), baseline CBF values were obtained. Our preliminary studies showed that the variation of CBF among preparations was less than 60 beats/min (< 6%), and that there were no significant differences between experimental groups. The pH of all solutions tested was between 7.2 and 7.4, a range known not to affect CBF (15). All the CBF measurements were taken by the same observer, who was unaware of the nature of the solutions.

In addition to the measurement of CBF, ciliary coordination was assessed from the image of ciliary beating action recorded on a video camera (C-1846-01; Hamamatsu Photonics) with a 0.75-inch videocassette recorder (VO-5800; Sony, Tokyo). Discoordination was defined as the loss of metachronal wave activity on the free border of the cell clumps (16, 17).

Effect of AVP on CBF

When we washed AVP out of the mucosal tissue preparations 15 min after its addition, and then reapplied AVP in the same dose, the response of CBF to the second dose was always less than that to the first dose. Because of this tachyphylaxis, only one dose of AVP was given per tissue preparation in all experiments. After determining the baseline CBF, medium was drained out of the chamber and replaced with HBSS containing AVP (10⁶ M), after which the CBF response was monitored for 20 min. In the control experiment, the response to HBSS alone was measured in an identical procedural sequence. To determine the dose-response relationship for AVP, various concentrations of AVP, ranging from 10⁹ to 10⁶ M, were added, and the maximal response of CBF to each concentration was measured. In addition, to determine the subtype of AVP receptors involved in the AVP action, the cells were incubated for 15 min with the V₁ receptor antagonist OPC-21268 (10⁶ M) (18) or the V₂ receptor antagonist OPC-31260 (10⁶ M) (19), and various concentrations of AVP were then added. Because the maximal response of CBF to AVP was observed in every tissue within 3 min after the addition of AVP, we measured the highest recorded value in the subsequent experiments.

It has been generally accepted that both intracellular cyclic adenosine monophosphate (cAMP) and Ca²⁺ are major determinants for airway epithelial ciliary motility (17, 20). We therefore assessed their contributions to the action of AVP. Tracheal mucosal epithelial tissues were incubated for 15 min with HBSS containing each of the following drugs: Rp-adenosine-3',5'-cyclic monophosphorothioate triethylamine (Rp-cAMPS, 5 × 10⁴ M), a specific cAMP-dependent protein kinase inhibitor (23); ethylene glycol-bis-(-aminoethyl ether) N,N,N',N'-tetraacetic acid (EGTA, 4 × 10³ M), a chelator of extracellular Ca²⁺; and thapsigargin (2 × 10⁶ M), an inhibitor of the endoplasmic reticulum Ca²⁺ pump (24). The response of CBF to the subsequent addition of AVP (10⁶ M) was then determined. In this separate experiment, Rp-cAMPS per se had no effect on the baseline value of CBF, whereas Ca²⁺-free medium containing EGTA decreased CBF by 7.6 ± 1.0% (P < 0.05, n = 9). Exposure to thapsigargin by itself caused a rapid increase in CBF by 23.9 ± 4.1% (P < 0.001, n = 9), but the CBF values returned to baseline levels within ~ 10 min.

Measurement of Intracellular Ca²⁺

Epithelial tissues dissected from the rabbit trachea were digested overnight at 4°C with PBS containing 0.05% protease (type XIV; Sigma). After terminating the digestion by adding 2.5% FCS, we pelleted the cells (200 × g for 10 min) and suspended the pellet in 50% Dulbecco's modified Eagle's medium (DMEM) and 50% Ham's nutrient F12 containing 5% FCS, nonessential amino acids, 100 U/ ml penicillin, 100 µg/ml streptomycin, and 50 µg/ml gentamicin. The isolated cells were plated at 2.5 × 10⁵ cells/ cm² on round coverslips (15 mm diameter; Matsunami Ltd., Tokyo) coated with human placental collagen (20 µg/ cm²). In this preparation of the cells, fibroblasts and other nonepithelial cells constituted < 5% of the total, and viability was > 95% as assessed by trypan blue exclusion. The cells were cultured in a CO₂ incubator for 5 d, when confluence was achieved. The coverslips were washed with HBSS containing N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (Hepes, 10² M, pH 7.4) and loaded with the acetoxymethyl ester of fura-2 (fura2-AM, 2 × 10⁶ M) at 37°C for 20 min (25). The coverslips were then washed again and held with a rigid holder in a continuously stirred cuvette containing Hepes-buffered HBSS maintained at 37°C, and the fluorescence intensity was measured spectrophotometrically (CAF-110; Japan Spectroscopic Co., Tokyo). Excitation at a frequency of 50 Hz was alternated automatically from 340 nm to 380 nm, and the light emitted from the cells was passed through a 500-nm filter and detected with a photomultiplier tube. Maximal (R_max) and minimal (R_min) values for the emission ratio were determined in the presence of ionomycin (10⁵ M) and EGTA (4 × 10³ M), respectively, and the intracellular Ca²⁺ concentration ([Ca²⁺]_i) was calculated with the formula described by Grynkiewicz and colleagues (26).

Effect of AVP on [Ca²⁺]_i

After the cells had been equilibrated for 30 min, AVP (10⁶ M) was added to the cuvette in the absence or presence of EGTA (4 × 10³ M), and the response of [Ca²⁺]_i was continuously recorded. Additionally, to determine the role of intracellularly stored Ca²⁺, thapsigargin (2 × 10⁶ M) was added and, when the [Ca²⁺]_i values had returned to their baseline levels, AVP (10⁶ M) was applied.

To study the AVP receptor subtype, the responses of [Ca²⁺]_i induced by the following agonists were examined: AVP (10⁶ M); [Phe², Ile³, Orn⁸] vasopressin ([PO-VT], 10⁶ M), a V_1a receptor agonist; [deamino¹, D-3- (pyridyl) Ala², Arg⁸] vasopressin ([DP-VP], 10⁶ M), a V_1b receptor agonist; and 1-desamino-8-D-arginine vasopressin ([dDAVP], 10⁶ M), a V₂ receptor agonist (27). Furthermore, the effects of AVP receptor antagonists on the AVP-induced increase in [Ca²⁺]_i were tested. In this experiment, the cells were incubated for 15 min with OPC-21268 (10⁶ M), the specific V_1a receptor antagonist [d(CH₂)₅¹, O-Me-Tyr², Arg⁸] vasopressin ([CTM-VP], 10⁶ M) (27), or OPC-31260 (10⁶ M), and the response of [Ca²⁺]_i to AVP (10⁶ M) was measured.

Drugs

The following drugs were used: AVP ([Arg⁸] vasopressin), dDAVP (Peptide Institute Co., Osaka, Japan); PO-VT, CTM-VP (Peninsula Laboratories, Belmont, CA); DP-VP (Bachem, Torrance, CA); OPC-21268, OPC-31260 (Otsuka Pharmaceutical Co., Tokyo); Rp-cAMPS (Biolog Life Science Institute, Bremen, Germany); EGTA, penicillin, streptomycin, gentamicin (Sigma); thapsigargin, fura2-AM, and ionomycin (Dojin Laboratories, Kumamoto, Japan).

Statistics

All values are expressed as means ± SE, and n refers to the number of rabbits from which the tissues were taken. Statistical analysis was performed through analysis of variance (ANOVA), using Scheffe's F-test, and a value of P < 0.05 was considered statistically significant (28).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of AVP on CBF

The effects of incubation of rabbit tracheal epithelium with either AVP (10⁶ M) or HBSS alone (control) on CBF are shown in Figure 1. There was no significant difference in baseline CBF in the AVP group versus the control group (962 ± 32 beats/min, n = 12, versus 949 ± 43 beats/min, n = 11). Addition of AVP elicited a rapid and transient increase in CBF to 1,157 ± 58 beats/min (P < 0.001, n = 12) within 3 min. This was followed, after 20 min of incubation, by return to a stable value of 1,028 ± 32 beats/min, which was still significantly greater than the baseline CBF (P < 0.05). Neither ciliary discoordination nor mucus production was observed in any of the experiments. The effect of AVP on the initial peak response of CBF was dose dependent, the maximal increase from the baseline value and the AVP concentration required to produce a half-maximal effect (EC₅₀) being 20.6 ± 4.7% (P < 0.001, n = 8) and 3.9 ± 0.7 × 10⁸ M (n = 8), respectively (Figure 2). Incubation of cells with OPC-21268 reduced the AVP-induced maximal response to 5.9 ± 2.0% (P < 0.01, n = 8), whereas OPC-31260 was without effect.

View larger version (19K): [in this window] [in a new window]   Figure 1.   Time course of the effect of AVP on CBF in tracheal epithelial cells from rabbits. After the baseline CBF was determined, bathing solution was replaced at time zero (arrow) with HBSS either alone (control: dashed line) or containing 106 M AVP (solid line). Data are means ± SE; n = 11 for control, and n = 12 for AVP.

View larger version (19K): [in this window] [in a new window]   Figure 2.   Dose-dependent effect of AVP on CBF in rabbit tracheal epithelial cells. After the baseline CBF was determined, bathing solution was replaced with HBSS either alone (control: open circles) or containing various concentrations of AVP (closed circles), and the maximal response to each concentration was determined. In the experiment with AVP receptor antagonists, the cells were incubated for 15 min with OPC-21268 (106 M, V1 antagonist, closed squares) or OPC-31260 (106 M, V2 antagonist, closed triangles), and the solution was replaced with HBSS containing AVP and each antagonist. Values are expressed as percentage of the baseline CBF. Each point represents mean ± SE; n = 8. *P < 0.05, ***P < 0.001, significantly different from the corresponding control values. P < 0.05, P < 0.01, significantly different from the values for AVP alone.

As shown in Figure 3, pretreatment of cells for 15 min with the medium containing Rp-cAMPS or Ca²⁺-free medium containing EGTA did not significantly alter the response of CBF to the subsequent addition of AVP, but incubation with thapsigargin abolished the ciliary-stimulatory effect of AVP (P < 0.001, n = 9).

View larger version (13K): [in this window] [in a new window]   Figure 3.   Effects of pharmacologic blocking agents on the increase in CBF induced by AVP in rabbit tracheal epithelial cells. After incubation of cells with Rp-cAMPS (5 × 104 M), Ca2+-free solution containing EGTA (4 × 103 M), or thapsigargin (2 × 106 M) for 15 min, AVP (106 M) was added. Values are expressed as percentage of the baseline CBF obtained before the addition of AVP. Data are means ± SE; n = 9 for each column. ***P < 0.001, significantly different from the response to AVP alone.

Effect of AVP on [Ca²⁺]_i

As shown in Figure 4, AVP (10⁶ M) rapidly increased [Ca²⁺]_i in rabbit tracheal epithelium from the baseline level of 116 ± 12 nM to the peak value of 180 ± 17 nM (P < 0.001, n = 12). The response was generally in a spikelike pattern, and return to the baseline level occurred within 2.5 to 3.0 min. Incubation of cells with Ca²⁺-free solution containing EGTA did not affect the magnitude of the [Ca²⁺]_i increase induced by AVP or the pattern of the response. In contrast, exposure to thapsigargin caused a transient increase in [Ca²⁺]_i, from 124 ± 19 nM to 177 ± 26 nM (P < 0.01, n = 12), and abolished the subsequent response of [Ca²⁺]_i to AVP.

View larger version (14K): [in this window] [in a new window]   Figure 4.   Representative recordings of fura-2 fluorescence ratio in rabbit tracheal epithelial cells. (A) Responses to consecutively added vehicle (HBSS) and AVP (106 M). (B) Response to AVP (106 M) in Ca2+-free solution containing EGTA (4 × 103 M). (C) Response to AVP (106 M) after treatment with thapsigargin (2 × 106 M).

The effects of AVP-receptor agonists on [Ca²⁺]_i are shown in Figure 5. Addition of DP-VP increased [Ca²⁺]_i by 58 ± 8 nM (P < 0.001, n = 7), an effect that was equipotent to that of AVP and with a response pattern similar to that with AVP, whereas PO-VT caused only a small increase in [Ca²⁺]_i, and dDAVP had no effect. The rank order of potency was thus AVP = DP-VP > PO-VT > dDAVP.

View larger version (27K): [in this window] [in a new window]   Figure 5.   Effects of AVP and AVP-receptor agonists at 106 M on intracellular [Ca2+]i in rabbit tracheal epithelial cells. Values are expressed as absolute [Ca2+]i variations (peak  baseline [Ca2+]i levels). Data are means ± SE; n = 7 for each column. *P < 0.05, ***P < 0.001, significantly different from the response to vehicle (HBSS). PO-VT, [Phe2, Ile3, Orn8] vasopressin (V1a agonist); DP-VP, [deamino1, D-3-(pyridyl) Ala2, Arg8] vasopressin (V1b agonist); dDAVP, 1-desamino-8-D-arginine vasopressin (V2 agonist).

The effects of AVP receptor antagonists, including CTM-VP, OPC-21268, and OPC-31260, on the AVP (10⁶ M)- induced increase in [Ca²⁺]_i are shown in Figure 6. Addition of each antagonist by itself did not significantly alter the baseline value of [Ca²⁺]_i. Incubation of cells for 15 min with OPC-21268 reduced the effect of AVP by 76 ± 11% (P < 0.01, n = 7), whereas CTM-VP and OPC-31260 had no effect.

View larger version (14K): [in this window] [in a new window]   Figure 6.   Effects of AVP receptor antagonists on the increase in [Ca2+]i induced by AVP. After incubation of cells for 15 min with CTM-VP ( [d(CH2)51, O-Me-Tyr2, Arg8] vasopressin, 106 M, V1a antagonist), OPC-21268 (106 M, V1 antagonist), or OPC-31260 (106 M, V2 antagonist), the baseline [Ca2+]i levels were determined and AVP (106 M) was added. Values are expressed as absolute [Ca2+]i variations (peak  baseline [Ca2+]i levels). Data are means ± SE; n = 7 for each column. **P < 0.01, significantly different from response to AVP alone.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

It has been generally accepted that mucociliary transport in the respiratory tract is governed by ciliary activity of airway epithelium and by the depth and rheologic properties of periciliary fluid (1, 29). Our in vitro studies demonstrate for the first time that the antidiuretic hormone AVP stimulates ciliary motility of rabbit tracheal epithelium. We found that addition of AVP to preparations of tracheal epithelium caused a rapid increase in CBF in a dose-dependent manner. However, the effectiveness of ciliary action depends on several characteristics of ciliary beating, of which CBF is but one. The coordination of the beating pattern, for example, also plays a role in ciliary performance (29). In the present experiments, no ciliary discoordination or apparent secretion of mucus was noted among adjacent cilia on the same cell or on several bordering cells in association with the increased CBF in response to AVP. We therefore speculate that the observed ciliary stimulatory effect of AVP might be translated into enhanced mucociliary transport, as predicted by theoretical models of mucociliary pumping (30).

Ciliary motility of airway epithelium is regulated mainly by cAMP and Ca²⁺ (17, 20). Intracellular cAMP activates glycogenolysis, which in turn stimulates the production of ATP, an energy source for ciliary beating, via the Krebs cycle (17). On the other hand, mobilization of intracellular Ca²⁺ apparently acts on the ciliary axoneme via formation of Ca²⁺-calmodulin complexes to increase the frequency of dynein-microtubule interactions (21). In our experiments, the increase in CBF produced by AVP was not altered by pretreatment of cells with Rp-cAMPS, a specific inhibitor of cAMP-dependent protein kinase, at a concentration sufficient to block cAMP-mediated ciliary stimulation under the same experimental conditions (30), indicating that the action of AVP is not related to cAMP. The ciliary-stimulatory effect of AVP remained unchanged with the addition of EGTA to eliminate Ca²⁺ influx from the extracellular solution, but was greatly attenuated by addition of the endoplasmic reticulum Ca²⁺-pump inhibitor thapsigargin to deplete internal Ca²⁺ stores (24). These results suggest that the release of intracellularly stored Ca²⁺, but not Ca²⁺ influx through membrane channels, may account for the stimulation of ciliary motility by AVP.

To confirm that AVP actually affects intracellular Ca²⁺ in rabbit tracheal epithelium, we measured [Ca²⁺]_i, using the calcium-sensitive fluorescent probe fura-2. Application of AVP elicited a rapid increase in [Ca²⁺]_i, and this response was transient, with [Ca²⁺]_i returning to baseline levels within 2.5 to 3.0 min. As is well known in most cell types, such a transient pattern of [Ca²⁺]_i elevation is largely attributable to Ca²⁺ release from intracellular stores produced by phospholipase C (PLC) activation, and to concomitant phosphatidylinositol hydrolysis, and that conversely, a sustained increase in [Ca²⁺]_i following the transient response is derived from Ca²⁺ influx (25, 31). We found that the AVP-induced increase in [Ca²⁺]_i was not changed by EGTA, but was abolished by thapsigargin. This finding is consistent with the notion that AVP increases [Ca²⁺]_i through mobilization from thapsigargin-sensitive internal stores, and the resulting increase in cytosolic Ca²⁺ may constitute the message that initiates stimulation of airway ciliary motility. In contrast to the response of [Ca²⁺]_i to AVP, which was only transient, the CBF values remained increased for at least 20 min after the addition of AVP. The mechanism for this discrepancy is uncertain. Similar observations have been reported for the CBF of rabbit oviductal ciliated cells in response to other putative agonists such as adenosine triphosphate (ATP) or prostaglandin F₂ (32), and the effect of AVP on ciliary motility might therefore not be fully explained by the release of Ca²⁺ from thapsigargin-sensitive stores.

To date, three subtypes of AVP receptor have been identified: V_1a, V_1b, and V₂ receptors. V_1a receptors are expressed predominantly in hepatocytes and vascular smooth muscle, and V_1b receptors are expressed in adenohypophysis and renal-collecting-duct epithelial cells (27). Both receptor subtypes are linked to PLC, which consequently induces intracellular Ca²⁺ release (27, 33). V₂ receptors, on the other hand, are expressed in renal medullary tubular cells and skeletal myogenic cells (34, 35), which are coupled to the adenylate-cyclase pathway (36). To determine the AVP receptor subtype responsible for the observed actions of AVP in rabbit tracheal epithelium, we conducted a series of experiments with specific AVP receptor agonists and antagonists. The increase in CBF produced by AVP was attenuated by the V₁-receptor antagonist OPC-21268 (18) but not by the V₂-receptor antagonist OPC-31260 (19), implying that the ciliary stimulatory effect of AVP may be mediated by V₁ receptors. In the experiment with intracellular Ca²⁺, addition of the V_1b- receptor agonist DP-VP increased [Ca²⁺]_i to the same degree as did AVP, whereas the V_1a-receptor agonist PO-VT produced only a small effect and the V₂-receptor agonist dDAVP failed to alter [Ca²⁺]_i, suggesting that V_1b receptors may be involved. However, because we did not study concentration-dependent effects of these agonists, no definitive conclusion can be drawn about the presence of V_1b receptors. Instead, we examined the effects of AVP receptor antagonists on the Ca²⁺-mobilization response to AVP. Pretreatment of cells with CTM-VP, a V_1a-receptor antagonist whose affinity for V_1b receptors is extremely low (37), or with OPC-31260, did not alter the effect of AVP. By contrast, OPC-21268, which may antagonize both V_1a and V_1b receptors (18), substantially inhibited the AVP-induced increase in [Ca²⁺]_i. Although a V_1b-receptor antagonist with strong selectivity over the V_1a receptors is not now available, our findings indicate a possible role of V_1b receptors in ciliary motility and intracellular Ca²⁺ mobilization in rabbit tracheal epithelium.

The pituitary-derived antidiuretic hormone AVP belongs to the family of vasoactive and mitogenic peptides. Previous study with scattered immunolabeling showed that AVP-like immunoreactivities were present in the cytoplasmic matrix of cultured and intact epithelial cells from rabbit trachea (12), and it has been reported more recently that AVP can be released from these cells (13). Although the local concentration of this hormone in the airway is unknown, it is possible that epithelium-derived AVP could affect ciliary activity through an autocrine mechanism.

In conclusion, AVP increases ciliary motility of rabbit tracheal epithelium through mobilization of intracellular Ca²⁺ from thapsigargin-sensitive stores. This Ca²⁺-mobilizing action of AVP may be mediated by stimulation of epithelial V_1b receptors.