INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Bronchial hyperresponsiveness (BHR), the capacity of airways to constrict excessively in response to a variety of nonspecific stimuli, is a hallmark of asthma. Possible mechanisms for BHR include altered neural pathways, remodeling of the airway wall, and the presence of inflammatory mediators (1). Asthmatic individuals are frequently hyperresponsive to contractile agonists such as methacholine (MCh), even when not experiencing symptoms of asthma. The degree of hyperresponsiveness can be augmented in the context of airway inflammation, such as that following antigen exposure, and can be at least partly reduced after treatment with corticosteroids (2). Despite its complexity, BHR must be a function of excessive airway narrowing and thus of airway smooth-muscle (ASM) shortening.

An important limitation to investigating the mechanisms for BHR is the inability to replicate in vivo findings in in vitro systems. Previous attempts to do this, with human and animal materials, have yielded mixed results (3, 4). Most studies have failed to find a relationship between the force-generating capacity of tracheal or bronchial smooth muscle and measurements of tracheal or bronchial responsiveness in vivo. This has led to the notion that the contractile responses of ASM are normal in asthma, and that hyperresponsiveness must result from other factors, such as smooth-muscle quantity, airway remodeling, or changes in the load applied to the smooth muscle (5). Recently, there has been renewed interest in the possibility that the dynamic contractile properties of ASM are important in determining airway responsiveness. A number of investigators, using sensitized dogs (6), rats (7), and mice (8), have noted that in contrast to force generation, the velocity of shortening of ASM may correlate with in vivo measurements of responsiveness.

In the present study we used a mouse model to investigate possible relationships between airway responsiveness and in vitro measurements of smooth-muscle contractility. Mice are an attractive species with which to model asthma or asthma-related phenomena because of the large amount of genetic and immunologic information available for this species. As they have with many phenotypic characteristics, different highly inbred murine strains have been reported to exhibit different levels of bronchial responsiveness, and these differences appear to have a genetic basis. In particular, naïve A/J mice have been reported to be hyperresponsive to cholinergic stimulation (9). Different patterns of inheritance are found when A/J mice are bred with C57BL/6J mice (10) than when they are bred with C3H/HeJ mice (11), and linkage analysis has suggested a number of chromosomal regions as being responsible for these differences (10, 12). Nonetheless, the mechanisms accounting for these interstrain differences remain unclear.

Since strain-related differences in cholinergic responsiveness in mice are well established, we investigated which, if any, characteristics of ASM could account for measured values of cholinergic responsiveness in vivo. Using oscillatory mechanics to precisely measure changes in respiratory impedance following stimulation with MCh, we confirmed previous measurements of interstrain differences in responsiveness. Furthermore, we established that these differences were a function of the contractile properties of the whole airway tree. We then explored the possible contributions of various properties of ASM to responsiveness, giving particular attention to the possibility that the dynamic properties of the airways reflect strain-related differences in responsiveness. Our findings indicate that the rate of ASM shortening is a major determinant of airway narrowing and thus of bronchial responsiveness itself.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals

All procedures were reviewed and approved by the McGill University Animal Ethics Committee. Male mice (5 to 6 wk of age, weighing 24 to 26 g) were obtained from commercial suppliers (A/J and C57BL/6J strains, Charles River, St. Constant, PQ, Canada; C3H/HeJ strain, Jackson Laboratories, Bar Harbor, ME; Balb/C strain, Harlan Laboratories, Indianapolis, IN) and housed in a conventional animal facility at our laboratory. For all procedures, mice were anesthetized with sodium pentobarbital (65 mg/kg) (MTC Pharmaceuticals, Cambridge, ON, Canada). Initially, protocols were conducted only with A/J, C57BL/6J, and C3H/HeJ mice. Once in vitro interstrain differences in the apparent velocity of ASM shortening were noted (see RESULTS), measurements were also made of the in vivo responsiveness and rate of airway contraction of Balb/C mice.

Determination of Dose-Response Curve for Methacholine In Vivo

Respiratory mechanics. The assessment of respiratory mechanics was done with a computer-controlled small-animal ventilator (SAV) (Flexivent; SCIREQ, Montreal, Canada) as previously described in rats (13, 14). Briefly, the mice were sedated (xylazine, 8 mg/kg intraperitoneally), deeply anesthetized (pentobarbital sodium, 65 mg/kg intraperitoneally), and tracheostomized, and the trachea was cannulated with an 18-gauge metal needle. The mouse was then connected to the SAV and paralyzed (doxacurium, 0.5 mg/kg intraperitoneally). Regular mechanical ventilation was applied, with the mouse being ventilated quasisinusoidally at a frequency of 150 breaths/min and a tidal volume of 6 ml/kg. The expiratory valve of the SAV allowed the animal to expire passively through a water trap adjusted to maintain a positive end-expiratory pressure (PEEP) of 0.15 kPa. In preliminary experiments, this PEEP was shown to be optimal for the determination of MCh-induced changes in respiratory system resistance (RRS) and respiratory system elastance (ERS). A saline-filled venous line (5 µl in volume) was established in the tail vein for the administration of MCh (3.3 to 1,000 µg/kg).

Pulmonary mechanics were measured with an oscillation technique as previously described (13). After cessation of regular ventilation and expiration by the animal to the relaxation volume while PEEP was maintained, the SAV was used to apply a low-amplitude flow oscillation to the lung during a 16-s period of apnea. This perturbation consisted of a superposition of three mutually prime frequencies of 1.48 Hz, 5.45 Hz, and 19.69 Hz. Because measurements were made at a constant lung volume, the transient dynamic hyperinflation that invariably occurs with bronchoconstriction when an animal is mechanically ventilated in the conventional manner was avoided, and the measurements reflected the intrinsic responsiveness of the lungs rather than the effects of changes in lung volume. The volume history of the lungs was standardized to TLC at 2 min before every lung function measurement. After an initial baseline recording, sterile saline (1 µl/g) and increasing, noncumulative doses of MCh (3.3 to 1,000 µg/ kg intravenously) were administered at 5-min intervals. The saline or MCh were given by intravenous injection at the start of the 16-s oscillation period, for a period not exceeding 4 s. As soon as data collection was complete, normal ventilation was resumed and the tail cannula was flushed with 25 µl of saline. The SAV was calibrated daily, before lung function measurements were made. This enabled us to calculate tracheal pressure (Ptr) and volume (Vtr) from cylinder pressure (P) and volume displacement (V) (13). Values given were corrected for resistive and accelerative losses.

Signal analysis. The 16-s records of Ptr and Vtr were analyzed as previously described in detail (13). Briefly, the signals were bandpass filtered by calculating the fast Fourier transforms of the complete ventilatory records for each animal, with all frequency components set to zero except those in the band of interest, and by then calculating the inverse Fourier transforms. Three bands were separated in this way, spanning the frequency ranges from 1 to 2 Hz, 5 to 6 Hz, and 19.2 to 20.2 Hz (13). Each Ptr-Vtr signal pair was analyzed through recursive multiple linear regression, with exponential memory time constants of 0.67 s, 0.18 s, and 0.05 s, respectively. These values gave signals for RRS and ERS over time at each of the three oscillation frequencies. The peak response to MCh was taken as the mean of the lung function measurement occurring between 14 s and 15 s, since the last second of the recording period was affected by Fourier edge effects. The mean of this time window was also used to calculate baseline values.

Airway responsiveness to MCh was expressed as the mean effective dose of MCh producing a 200% increase in RRS (ED₂₀₀RRS) and a 150% increase in ERS (ED₁₅₀ERS) from the respective control saline baseline values. Values were determined by interpolation from each animal's dose-response curve. In addition, the maximal response to MCh and the percentage changes in RRS and ERS at the MCh dose of 330 µg/kg were determined. This dose was the highest at which measurements could be obtained at all frequencies in all animals. At MCh doses above this, nonlinearities prevented us from making valid observations at 19.7 Hz in some animals.

Quantity of Smooth Muscle

To investigate the possibility that structural factors contribute to interstrain differences in the airway responsiveness of mice, we performed a morphometric analysis of the trachea as follows. At the end of each measurement of isometric contraction, as subsequently described, tracheal rings were fixed in 10% buffered formalin for 48 h, and were then embedded in paraffin and sectioned parallel to the cartilaginous bands, using a microtome (Model 820; American Optical Corporation, New York, NY). The entire trachea was divided into cross-sections, and every 20th section was examined. Five-micron- thick sections were mounted on glass slides and stained with Masson's trichrome. An optical microscope with a drawing-tube attachment was used to measure airway dimensions with a computerized digitizing board (Jandel Scientific, Corte Madera, CA). The contours of the epithelial basement membrane, smooth muscle, and cartilage were traced. From these contours, the length of the epithelial basement membrane (PBM), area of tracheal smooth muscle, and area of tracheal cartilage were measured through the use of a commercial software package (Sigma Scan; Jandel Scientific).

Myosin Content

To further investigate the possibility that differences in airway responsiveness among mouse strains are due to variation in the quantity of ASM, we estimated the amount of contractile protein in tracheas of A/J, C57BL/6J, and C3H/HeJ strains of mice by measuring the amount of myosin through Western blotting. Tracheas were homogenized with 200 µl of extraction buffer containing 0.5 M NaCl, 10 mM imidazole-HCl (pH 7.5), 5 mM -mercaptoethanol, 2 mM ethylene glycol-bis-(-aminoethyl ether)-N,N'-tetraacetic acid, 10 mM adenosine triphosphate, 10 mM MgCl₂, 1 mM benzamidine-HCl, 1 mM phenylmethylsulfonylfluoride, 10 µg/ml pepstatin A, 10 µg/ml leupeptin, and 10 µg/ ml aprotinin. Protein concentration was determined by using the protein dye-binding technique of Bradford, with bovine serum albumin (BSA) as a standard and emission recorded at 595 nm. Linear regression analysis was used to calculate unknown sample protein concentrations from the standard curve generated with BSA. Homogenate proteins (20 µg) were heated at 80° C for 5 min and were then loaded onto a 4 to 12% sodium dodecyl sulfate (SDS)-polyacrylamide gradient gel for 2 h at a constant voltage of 125 V. Proteins were then transferred in transfer buffer (pH 9.9) containing methanol (20%), NaHCO₃ (3.8 mM), and Na₂CO₃ (7.9 mM) onto nitrocellulose paper for 2 h at a constant current of 250 mA. Staining with Coomassie blue confirmed the electrotransfer efficiency, and was used as an indirect control for loading. The nitrocellulose membrane was blocked overnight at 4° C in 7% skim milk in Tris-buffered saline containing 0.1% Tween-20 (TTBS), and was then incubated for 2 h with a rabbit polyclonal antimyosin (smooth and skeletal) antibody (1:100; Sigma Chemical Co., St. Louis, MO) in TTBS. The blots were then incubated for 1 h with a horseradish peroxidase-linked antirabbit secondary antibody diluted 1:1,000 in TTBS. Between antibody incubations, nitrocellulose strips containing the blots were washed three times with TTBS. All incubation and washing steps were performed at room temperature. Heavy-chain myosin bands were detected with the ECL Plus chemiluminescence detection system (Amersham, Little Chalfont, UK), and were exposed on Amersham radiography film.

Measurement of Isometric Contraction of Tracheal Rings

We measured tracheal isometric tension with tracheal segments mounted in organ baths as follows. Mice were killed with an overdose of pentobarbital and the proximal trachea was rapidly removed and placed in gassed (95% O₂/5% CO₂) Krebs solution (NaCl 118.1 mM, KCl 4.7 mM, CaCl₂ 2.5 mM, MgSO₄ 1.2 mM, KH₂PO₄ 1.2 mM, NaHCO₃ 25 mM, glucose 11.1 mM). Tracheal rings were then mounted on two triangular supports, transferred to a 20-ml organ bath filled with Krebs solution heated to 37° C, and aerated with 95% O₂/5% CO₂ for an equilibration period of 60 min. The tissues were washed with fresh buffer every 15 min. Tension was measured isometrically with Grass force displacement transducers (FT-03C; Grass Instruments, Quincy, MA), and was recorded on a polygraph (Model 7414A; Hewlett-Packard Co., Cupertino, CA). During the equilibration period, tension was maintained at 0.5 g, which in preliminary experiments had been determined to be the tension that gave the maximal response to 10⁵ M MCh in all strains of mice. After stabilization of the tracheal tissues, a cumulative concentration-response curve was produced by adding increasing concentrations of MCh, from 10⁹ to 10⁴ M, to the tracheal tissue buffer. In each experiment, we determined the maximal tension generated (expressed in grams) and the concentration of MCh that produced 50% of the maximal tension (EC₅₀).

Lung Explants

To further characterize the in vitro contractile behavior of airways, we studied the responsiveness of intraparenchymal bronchi with the lung explant technique. Using two separate protocols, we constructed dose-response curves for MCh and apparent maximum velocity of shortening. Explants and culture media were prepared as previously described (7). Briefly, a midline incision from the neck to the lower abdomen was made in a mouse and the trachea was cannulated. The abdominal wall was incised and the animal was exsanguinated. The anterior chest wall was removed and the lungs were then inflated with warm (37° C) agrose-bicarbonate-buffered culture medium (BCM) through the endotracheal tube to a volume equal to the estimated TLC (0.05 ml/g body weight). The lungs were then dissected out, embedded in a syringe tube filled with 4% agarose, and refrigerated at 4° C for 30 min to gel the agarose. The resulting block was sliced into 0.5-mm-thick explant pieces that were placed in BCM and cultured at 37° C overnight. On the following day the explants were transferred to six-well plates with 1 ml of N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid-buffered culture medium (HCM) in each well, and were washed slowly. Airways were selected for study if they were cut in cross section and the epithelium was free of agarose. All studies were done at 37° C after a 30-min stabilization period. Images were recorded with an inverted microscope equipped with a video camera as previously described (7). For each airway, we measured the area (Ai) as the area enclosed by the epithelial luminal border.

Explant Protocol 1: Concentration-response curves. After baseline images were recorded, a 10-µl aliquot of stock MCh-HCM solution was added to the HCM outside the culture well inserts, resulting in a final MCh concentration of 10⁸ M. After incubation at 37° C for 10 min, all the airways were reimaged. This procedure was repeated for each concentration of MCh from 10⁸ to 10² M. We defined maximum contraction as the minimum area of each airway following exposure to MCh, expressed as the percentage of the baseline area. As a measure of bronchial responsiveness, we calculated EC₅₀ as the effective concentration of MCh causing a reduction in Ai equal to 50% of that with the maximum contraction.

Explant Protocol 2: Velocity of ASM shortening and maximum airway narrowing. We also used explants to investigate the dynamics of airway narrowing through a modification of Explant Protocol 1 (7). After baseline images had been recorded, 10 µl of 10² M MCh solution was dropped directly on the airway of interest. Images were then sequentially recorded at frequencies of 1 Hz for 8 s, 0.2 Hz for 10 s, 0.1 Hz for 40 s, 0.033 Hz for 120 s, and 0.016 Hz for the remainder of the 300-s recording period. Ai was measured and the internal perimeter (Pi) was calculated by assuming a circular lumen with Pi = 2r and r = (Ai/)^1/2. We estimated the velocity of shortening of ASM by numerically differentiating the values of Pi with respect to time. Because Pi values were normalized to baseline Pi, the results are reported in units of %.s¹. We chose the maximum value of the rate of change of Pi with respect to time ("peak velocity") as an index of apparent maximum velocity of shortening of the smooth muscle. We use the term peak velocity instead of "maximum velocity of shortening" because we measured shortening indirectly, from changes in airway lumen perimeter, without determining the exact load applied to the muscle. In addition to peak velocity, we took airway narrowing at 15 s after stimulation with MCh as an index of maximum narrowing.

Statistical Analysis

To compare different strains of mice, we first used a Kruskal-Wallis nonparametric test followed by Dunn's procedure to test for significance of differences among the different mouse strains used in the study. The Kolmogorov-Smirnov test was used to compare cumulative frequency distributions. Results are expressed as mean ± SE except where otherwise specified. A level of p < 0.05 was considered statistically significant. Statistical procedures were done with the Statview program (version 5.0; SAS Institute Inc., Cary, NC).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In Vivo Differences in Airway Responsiveness among Mouse Strains

RRS and ERS. As shown in Table 1, the baseline RRS following injection of saline was not statistically different among the different mouse strains studied at any of the frequencies examined. Baseline ERS was significantly greater in A/J and C57BL/ 6J than in C3H/HeJ mice at all frequencies. At 5.5 Hz and 19.7 Hz, ERS was significantly greater in C57BL/6J than in Balb/C mice. Figure 1 shows typical changes in RRS and ERS over time after injection of MCh (330 µg/kg) in a C3H/HeJ mouse. These changes were similar in all strains of mice. Increases in both RRS and ERS started approximately 5 s after MCh injection and continued progressively throughout the rest of the 16-s measurement period. The resistances at each of the oscillatory frequencies applied increased roughly in parallel (Figure 1, upper panel ), although the 1.5-Hz component of RRS was larger than the 5.5-Hz component, which was in turn larger than the 19.7-Hz component. In contrast, the rank order of the increases in ERS in response to MCh was reversed (Figure 1, lower panel ), in that the 19.7-Hz component of ERS was larger than the component at 5.5 Hz, which in turn was larger than the 1.5-Hz component.

View larger version (15K): [in this window] [in a new window]   Figure 1.   Representative curves of RRS and ERS versus time, obtained from a C3H/HeJ mouse (PEEP = 0.15 kPa) after injection of 330 µg/kg of MCh at t = 0. Curves correspond to frequencies of 1.5 Hz, 5.5 Hz, and 19.7 Hz, as indicated. There was a parallel increase in RRS at all frequencies. For ERS the largest increase was observed at 19.7 Hz.

RRS and ERS at 330 µg/kg. Figure 2 shows the RRS for the three oscillatory frequencies studied after the intravenous injection of 330 µg/kg of MCh. The rank order of in vivo airway responsiveness for the four strains of mice studied was A/J  Balb/C > C3H/HeJ  C57BL/6J. At 1.5 Hz, the RRS of A/J mice was significantly increased as compared with that of C57BL/6J or C3H/HeJ mice. Similarly, the RRS of Balb/C mice was significantly increased as compared with that of C3H/HeJ mice. At 5.5 Hz, RRS was significantly increased in A/J and Balb/C mice as compared with C3H/HeJ or C57BL/6J mice. At 19.7 Hz, no differences were found among the strains.

View larger version (20K): [in this window] [in a new window]   Figure 2.   RRS measured in vivo with the SAV at an MCh dose of 330 µg/kg given intravenously among four inbred strains of mice. Values given are means ± SE of n = 7 for A/J, n = 8 for Balb/C, n = 7 for C57BL/6J, and n = 8 for C3H/HeJ mice unless otherwise specified. Airways of A/J and Balb/C mice were more responsive than those of C57BL/6J or C3H/HeJ mice. The rank order of responsiveness was A/J  Balb/C > C3H/HeJ  C57BL/6J. * p < 0.05 compared with A/J mice; #p < 0.05 compared with Balb/C mice.

ED₂₀₀RRS and ED₁₅₀ERS. As shown in Table 2, changes in ED₂₀₀RRS and ED₁₅₀ERS in response to MCh were frequency dependent, with values at 19.7 Hz exhibiting the greatest departure from the saline baseline. Airways of C57BL/6J mice were consistently the least sensitive, followed by those of C3H/ HeJ mice. The ED₂₀₀RRS and ED₁₅₀ERS of A/J mice were significantly different from those of C3h/HeJ or C57BL/6J mice. Similarly, airway responses of Balb/C mice were also different from those of C3H/HeJ and C57BL/6J mice at all frequencies, except in the case of ED₂₀₀RRS at 1.5-Hz, for which interstrain differences did not reach statistical significance.

Isometric Contraction of Tracheal Rings

As shown in Table 3, no statistical differences were found among A/J, C57BL/6J, and C3H/HeJ mice with regard to EC₅₀. Although the maximal tracheal tension among C57BL/ 6J mice tended to be lower than that among mice of the other two strains, this did not reach statistical significance (p > 0.1 versus A/J or C3H/HeJ).

Smooth-Muscle Quantity and Myosin Content

We observed small differences in morphometric indices among the strains of mice examined in the study (Table 3). Whereas cartilage area was significantly greater in A/J than in C57BL/6J or C3H/HeJ mice (p = 0.005 and 0.002, respectively), basement membrane length was significantly less in the C3H/HeJ strain. After correction for airway size, the area of tracheal smooth muscle was significantly smaller among C57BL/6J mice (0.0016 ± 0.0003 mm²) than among either C3H/HeJ (0.0026 ± 0.0001 mm², p < 0.001) or A/J (0.0023 ± 0.0001 mm², p = 0.01) mice. Results of the myosin assay were slightly different. Smooth-muscle myosin extracted from trachealis appeared as a band of approximately 200 kD in Coomassie blue-stained 6% SDS- polyacrylamide gel electrophoresis gels. Densitometric scanning of Western blots indicated that trachealis from C3H/HeJ, A/J, and C57BL/6J mice contained similar quantities of myosin, although there was a tendency for A/J mice to have higher myosin values (Table 3).

Explant Protocol 1

The maximum narrowing of the airway in response to MCh did not differ significantly among the three strains of mice in which this was studied. In this protocol, in which explants were exposed to each dose of MCh for 10 min, mean values were over 90% for all strains, with most explanted airways exhibiting closure (Table 4). In contrast to airway narrowing, there were some interstrain differences in MCh sensitivity. The mean EC₅₀ was significantly lower among C3H/HeJ mouse airways (log₁₀ dose MCh = 5.53) than among A/J (4.5, p < 0.0001) or C57BL/6J mouse airways (4.3, p < 0.0001).

Explant Protocol 2

Rate of ASM shortening. Figure 3 shows the results of measurements of airway perimeter versus time (Figure 3A) and apparent velocity of ASM shortening versus time (Figure 3B) for a representative airway from an A/J mouse. Peak velocity was reached between 2 s and 3 s, and maximum narrowing occurred within the first 15 s. The peak velocity of shortening was significantly higher among A/J mice (34.0 ± 1.91%.s¹) than among C57BL/6J (26.0 ± 1.56%.s¹) or C3H/HeJ (25.1 ± 1.45%.s¹) mice. Given that the rank order of these results was similar to that observed in vivo, we chose to additionally study Balb/C mice, a strain previously reported to be relatively hyperresponsive to MCh (9). As with the airways of A/J mice, the apparent rate of shortening among Balb/C mice was relatively high (29.8 ± 1.92%.s¹).

View larger version (12K): [in this window] [in a new window]   Figure 3.   Example of a dynamic response of a representative airway from an A/J mouse (Protocol 2). (A) Change in bronchial perimeter, expressed as a percentage of baseline, versus time. (B) Corresponding velocity of shortening versus time. The velocity of shortening peaks at 2 s, and the maximum amount of contraction is achieved during the first 15 s.

Because there was variability of the peak velocity of ASM shortening in each mouse strain studied, we also analyzed the cumulative frequency distributions of these peak velocities (Figure 4). Significant differences were found between the distributions for A/J and both C3H/HeJ and C57BL/6J mouse airways (p < 0.001 and p = 0.04, respectively), and between those for Balb/C and both C3H/HeJ and C57BL/6J mouse airways (p < 0.01 and p = 0.04, respectively). The rank order of the peak velocity of shortening was A/J  Balb/C > C57BL/ 6J  C3H/HeJ.

View larger version (23K): [in this window] [in a new window]   Figure 4.   Cumulative frequency distribution of peak velocity of ASM shortening in response to MCh in explanted airways (Protocol 2). Each symbol represents a single airway: A/J (closed circles), Balb/C (open circles), C3H/HeJ (closed squares), and C57BL/6J (open squares) mice. ASM of A/J and Balb/C mice exhibited a higher velocity of shortening than that of C3H/HeJ and C57BL/6J mice. The rank order of the velocity of shortening in explants was similar to the phenotype of responsiveness observed in vivo.

Relation between airway narrowing and velocity of shortening. The magnitude of airway narrowing achieved at 15 s was greater in the hyperresponsive A/J and Balb/C strains of mice than in either the C3H/HeJ or C57BL/6J strains (minimal lumen area/initial lumen area: 19% and 26%, versus 30% and 30%, respectively). Airways from the hyperresponsive strains were also more likely to close completely (Figure 6, lower panel ). Further analysis revealed a strong correlation between the peak velocity of ASM shortening and the amount of shortening at 15 s in airways for the four strains of mice (Figure 5). In general, the higher the velocity of shortening, the smaller the resulting airway perimeter. Across strains, closure was observed among airways that shortened at a rate of 26%.s¹ or more. Below this value no closure was seen. The relationship between peak velocity of airway shortening and airway narrowing was also dependent on airway size. When airways were divided into groups according to the median of the baseline Pi, the peak velocity of shortening was higher in smaller airways (Figure 6). The percentage of closed airways was also higher for smaller airways (Figure 6). Furthermore, when bronchi that closed were excluded from the analysis, there was significant differences among mouse strains with regard to the slope of the relationship between the velocity of shortening and achieved airway narrowing. Among small airways, the slopes for the hyperresponsive strains were significantly lower than those for the less responsive strains (1.55 and 1.60 for A/J and Balb/C mice, versus 2.15 and 2.29 for C57BL/6J and C3H/HeJ mice, respectively).

View larger version (27K): [in this window] [in a new window]   Figure 6.   Explanted airways divided into groups according to median initial perimeter. In all mouse strains except C57BL/6J, the peak velocity of ASM shortening was significantly higher in smaller airways (p < 0.05, t test). In all strains, the percentage of closed airways was also significantly higher in small airways (p < 0.05, chi-square analysis).

View larger version (31K): [in this window] [in a new window]   Figure 5.   Relation between peak velocity of ASM shortening and airway narrowing in explants from the four strains of mice studied. Airway narrowing is expressed as percentage of baseline perimeter. Airway closure (perimeter = 0%) was not observed for velocities less than 26%.s1.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We used inbred strains of mice to investigate the relationship between measurements of bronchial responsiveness in intact animals and the characteristics of ASM in vitro. As in other studies, the strain-related differences in bronchial responsiveness that we observed were not paralleled by differences in maximal contractile force generation or smooth-muscle quantity. Similarly, although interstrain differences were found when responsiveness was measured in explanted bronchi through the use of dose-response curves, there was no relationship between rank order of responsiveness measured in vitro and that seen in vivo. In contrast, when the apparent velocity of shortening was measured in explanted intraparenchymal airways, the rank order of responsiveness among strains paralleled that seen in vivo, and was correlated with the magnitude of achieved airway narrowing. These findings underscore the importance of velocity of ASM shortening as a potential determinant of bronchial responsiveness.

Although the cholinergic responsiveness of the mouse strains used in our study has been previously measured (10, 11), one of the goals of our study was to accurately characterize the mechanical responses of these animals' airways to MCh. Using a computer-controlled SAV (13, 14), which enabled us to observe changes in respiratory mechanics with considerable precision while lung was volume kept constant, we investigated acute changes in respiratory impedance as a function of oscillatory frequency during MCh-induced bronchoconstriction. Because we measured RRS and ERS in the interval between 14 s and 15 s after MCh exposure, the maximal response plateau to MCh was in many cases not achieved before mechanical ventilation had to be resumed. Nevertheless, in that we effectively excluded the possibility that differences in lung volume contribute to the phenotype of airway responsiveness, we believe that our measurements represent a reasonable compromise.

These measurements provide clues about the physiologic events occurring within the lungs during bronchoconstriction (15, 16). Changes in impedance at high frequency primarily reflect events in more proximal airways, whereas resistance at low frequency includes a contribution from distal airways (16). Increases in RRS were proportionally similar at all three oscillatory frequencies used in our study (Figure 1, upper panel ), demonstrating that MCh-induced bronchoconstriction occurred across the entire airway tree. Thus, although the major site of increase in RRS is central, differences in the airway- responsiveness phenotype among inbred strains of mice appear to be a property of the airways at all levels of the bronchial tree, rather than of the larger airways only, for example. The various components of ERS changed in very different ways than did RRS (Figure 1, lower panel ). The greatest changes occurred at the highest frequency, indicating that as resistance of the peripheral airways increased, progressively more of the applied oscillations in Vtr were shunted to the central airways, with a resulting increase in elastance. These findings are also consistent with the concepts that after MCh infusion, airway narrowing involves both peripheral and central airways, and that to the extent that phenotypic differences in responsiveness are mirrored in vitro, they should be detectable at all levels of the airway tree.

The interstrain differences we observed are quite similar to those previously described in studies using other techniques (9). At the dose of 330 µg/kg MCh, there were clear differences between the more responsive (A/J, Balb/C) and less responsive (C57BL/6J, C3H/HeJ) mouse strains. A/J mice developed much higher levels of impedance than did the other strains, suggesting a greater capacity for bronchoconstriction in response to a fixed stimulus. We also measured ED₂₀₀RRS and E₁₅₀ERS as alternate indicators of responsiveness. Although the differences were more subtle, both ED₂₀₀RRS and E₁₅₀ERS showed the C57BL/6J and C3H/HeJ strains to be generally less responsive than the A/J and Balb/C strains of mice (Table 2).

Given that our impedance measurements indicated that interstrain differences in responsiveness reflect airway phenomena, our attempt to find in vitro correlates of these differences focused on the airways. Investigations of tracheal smooth-muscle properties were not fruitful. As previously observed (17), we found no differences in tracheal smooth-muscle tension development among A/J, C57BL/6J, and C3H/HeJ mice. Although we did find an apparently reduced amount of smooth muscle in one hyporesponsive mouse strain (C56BL/6J), this was not the case for the other hyporesponsive strain (C3H/ HeJ). Because a limitation of morphometric measurements of smooth-muscle quantity is that tissue histologically identified as smooth muscle may include noncontractile elements, we also measured smooth-muscle myosin content in tracheas of these mice. Although we detected slight differences among strains (Table 3), these did not correspond with measurements of responsiveness in these strains. Therefore, the quantities of smooth muscle and of contractile protein do not appear to explain the interstrain differences in responsiveness among A/J, C57BL/6J, and C3H/HeJ mice.

A limitation of using tracheal smooth-muscle contractility as an index of in vitro responsiveness is that the possible contribution of more peripheral airways is ignored. We therefore measured responsiveness in intraparenchymal bronchi with the explant technique that we had previously developed and applied in the rat (18) and to a lesser extent in human airways (19). This approach allows determination of the contractile responses in a population of airways, and thus avoids some of the potential sampling bias inherent in measurements of only a few airways at a time. In contrast to measurements of tracheal isometric tension, we found significant differences among bronchial preparations from inbred strains of mice in their dose-response curves for MCh, although the rank order of responsiveness in vitro was quite different from that seen in vivo. In particular, the least responsive strain, C3H/HeJ, was the most responsive when this protocol was used. Thus, as with more traditional measurements, these in vitro results do not resemble those in intact animals, and do not appear to provide an explanation for interstrain differences in bronchial responsiveness.

In contrast, measurements of apparent maximal ASM shortening velocity yielded interstrain differences quite similar to those found in vivo. Indeed, the results for A/J, C57BL/6J, and C3H/HeJ mice were so compelling that we chose to measure responsiveness and shortening velocity in an additional strain of mice. As with the A/J strain, Balb/C mice also exhibited both relative hyperresponsiveness to MCh and high contraction rates. Although it was beyond the scope of the present study to establish whether ASM shortening velocity always corresponds to in vivo airway responsiveness, our results indicate that the rate of shortening of ASM can play a central part in the pathophysiology of BHR.

Interpretation of differences between mouse strains in the peak velocity of ASM shortening needs to be considered in the context of the load applied to the ASM. As previously observed in the rat (7), the peak velocity of shortening of murine airways was reached approximately 2 s after stimulation, and most of the reduction in airway lumen area was completed by about 15 s. These dynamics are remarkably similar to those measured under isotonic conditions with the load-release technique (20), and suggest that in airway explants the ASM is very lightly loaded, at least immediately after stimulation. Once the smooth muscle begins to shorten, however, load is expected to increase as shortening proceeds (23). It is this elastic load that appears to be the determinant of achieved airway narrowing for any particular rate of shortening (Figure 5). To examine this further, we performed an additional experiment to estimate the contribution of interdependence between the airway and parenchyma to the loading of ASM in explants. We reduced the amount of agarose-parenchymal matrix surrounding 11 airways from three A/J mice by cutting around the airways. When the surrounding parenchyma was removed from explants, there was a marked increase in the degree of airway narrowing without an apparent change in the rate of contraction (Figure 7). Although cutting back the agarose-parenchymal matrix did not change the velocity of shortening, there was a definite increase in the magnitude of the contraction at 15 s and in the number of closed airways. This suggests that changes in load predominantly affect the degree of airway narrowing achieved for a given shortening velocity. There was a close correlation between peak velocity and achieved airway narrowing in all four strains of mice studied (Figure 5). Therefore, under a given set of loading conditions, the degree of airway closure is determined by the velocity of shortening of ASM at the start of muscle contraction.

View larger version (33K): [in this window] [in a new window]   Figure 7.   Peak velocity of shortening (left panel ) and achieved narrowing (right panel ) in intact explants (Intact) and explants with parenchyma removed (Cut). *p < 0.05.

Although the relationship between velocity of ASM shortening and achieved airway narrowing was similar in all strains of mice studied, there was some variation with airway size and mouse strain. At a given velocity of shortening, smaller airways appeared to close more easily than larger airways (Figure 6). The slope of the relationship between velocity and achieved contraction differed between the hyperresponsive (A/J, Balb/C) and the less responsive strains (C3H/HeJ, C57BL/6J) of mice, particularly in smaller airways, suggesting that differences in load may also contribute to the airway hyperresponsiveness of the A/J and Balb/C strains (see RESULTS).

Our findings provide insight into the role of ASM shortening velocity as a determinant of bronchial responsiveness. Traditional analyses of the forces acting on the airways during bronchoconstriction have emphasized the steady-state balance of forces across the airway wall (24). In that type of analysis, maximal force generation is predicted to be the main determinant of the magnitude of airway narrowing achieved. In such a model, the quantity of ASM per unit area is the variable most likely to explain BHR (25). Our data suggest that in the mouse at least, this is unlikely. On the other hand, there has been interest in the possible importance of the dynamics of ASM shortening as possibly contributing to bronchial responsiveness. This interest is based on observations indicating the importance of lung-volume history as a determinant of bronchial responsiveness. Small decreases in lung volume can markedly increase bronchial responsiveness (26), and prevention of deep breaths during induced bronchoconstruction in normal subjects can lead to the development of airway obstruction similar to that seen in asthma (27). Furthermore, smooth-muscle contractility appears to be quite sensitive to changes in muscle length (28). Cyclical changes in muscle length, on the order of those occurring during tidal breathing, can substantially reduce smooth-muscle contractility (29).

On the basis of these observations, it has been hypothesized (30, 31) that the relationship between ASM shortening velocity and cyclical changes in ASM length during tidal breathing may be a critical determinant of bronchial responsiveness. If smooth muscle contracts fast enough, it would enter the "latch state" (32) before tidal oscillations could intervene to relax the muscle. Slower muscle would be more susceptible to the influence of periodic tidal stretches that would relax the muscle and prevent airway narrowing. In the present study we observed a close relationship between the velocity of ASM shortening and the achieved degree of constriction even in the absence of tidal oscillations. Although our findings do not contradict the hypothesis of a relationship between ASM shortening velocity and cyclical changes in ASM length as a determinant of bronchial responsiveness, they suggest that the velocity of shortening may be an important determinant of responsiveness regardless of the relative rates of tidal breathing and ASM shortening. As long as ASM can contract quasiisotonically at the beginning of contraction, the duration of shortening of the muscle would be expected to be similar to that observed in our study and in others (21). Under such circumstances, the amount of airway narrowing or closure achieved would be a function of the rate of muscle shortening for any given external load applied to the muscle. In this regard, it is interesting to note that the increase in RRS after MCh injection followed a time course similar to that of contraction in explants (Figure 1, upper panel ). Given that RRS reflects the integrated behavior of many airways, its rapid increase indicates that at least some airways in intact mice are capable of contracting as rapidly as in the explant preparation.

Other studies have found a similar correspondence between muscle shortening velocity and responsiveness. The hyperresponsive Fisher rat exhibits an increased rate of ASM shortening as well as increased force generation and smooth-muscle quantity (33). However, the only commonality between the Fisher rat and hyperresponsive mice appears to be the increased rate of ASM shortening seen in both models. Correlations between responsiveness and the rate of smooth-shortening are, moreover, not limited to genetic models of responsiveness. The Stephens group has found that the BHR associated with sensitization in both dogs (20) and mice (8) is better reflected by measurements of shortening velocity than by measurements of isometric tension. There is also at least one report of increased shortening of asthmatic smooth muscle (34). Furthermore, passive sensitization of human bronchial smooth muscle has been associated with increased shortening velocity (35). Taken together, these observations strongly suggest that the shortening velocity of ASM may be a crucial determinant of bronchial responsiveness.

In summary, we found that isometric tension of tracheal rings, morphometric analysis of trachea, evaluation of tracheal myosin content, and the dose-response curve for MCh in explanted airways could not explain the phenotype of bronchial responsiveness observed in highly inbred strains of mice in vivo. On the other hand, the rank order of the velocity of shortening of the airways among the mouse strains that we studied corresponded to that of airway responsiveness in vivo. Furthermore, shortening velocity was closely related to achieved narrowing and airway closure in all of the strains of mice used in our study. Our findings strongly support the notion that the shortening velocity of ASM is important in the pathophysiology of BHR.