INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Although bronchial hyperresponsiveness (BHR) is recognized as a characteristic feature of asthma, the mechanisms underlying this excessive narrowing of the airways in response to a range of stimuli are unclear. An intrinsic abnormality of human airway smooth muscle (ASM) has been postulated as being responsible for BHR (1, 2), and evidence has recently been accumulating to suggest plausible mechanisms by which changes in the enzymatic pathways regulating contraction of ASM could result in the airway hyperresponsiveness (AHR) associated with asthma. In a canine model of allergic AHR (3), allergic sensitization alterated contractile properties of bronchial smooth muscle, producing a greater early shortening velocity (4), increased maximum shortening capacity (4), and prolonged isotonic relaxation (5). In this model, the content of 138-kD myosin light-chain kinase (MLCK) (6) was found to be increased in smooth muscle from sensitized airways (7). MLCK is a key regulator of contraction, since its activity is responsible for phosphorylation of the 20-kD regulatory light chains of myosin (MLC₂₀), allowing actomyosin adenosine triphosphatase (ATPase) to be activated by actin and thereby initiating smooth-muscle contraction via the sliding-filament process (8).

MLCK has a number of different isoforms. The two major classes that have been clearly identified by molecular cloning, and which have been shown to correspond to proteins in tissues and cells (9) are smooth-muscle MLCK (smMLCK; 130 to 150 kD) and nonmuscle MLCK (nmMLCK; 210 to 220 kD). However, this nomenclature is somewhat misleading, since smMLCK has been shown to be expressed in most if not all adult tissues and some cell lines (10), whereas nmMLCK has been shown to be expressed in smooth muscle (12). smMLCK and nmMLCK arise from the same gene (11), and the entire amino acid sequence of smMLCK is contained within the nmMLCK sequence. Therefore, immunoblotting with a single monoclonal antibody (clone K36) will detect both smMLCK and nmMLCK, since this antibody detects a common epitope at the NH₂-terminal (residues 29 to 80 [12]) of both isoforms.

As with the canine model, we have observed changes in contractile responsiveness in models of human allergic AHR. We have reported that isolated human bronchi passively sensitized by overnight incubation in serum from allergic asthmatic individuals exhibit an increased responsiveness to contraction induced by histamine (14) and a decreased relaxation response to verapamil and levcromakalim (15). Additionally, Mitchell and associates (16), after passive sensitization of normal human trachealis ASM, observed an increased maximal shortening capacity without alterations in the ability of the ASM to generate isometric force. In actively sensitized tissue, we have observed that human mast cell-derived tryptase potentiates the contractile response to histamine (17), as well as an increased responsiveness to electrical-field stimulation upon application of supernatants derived from activated neutrophils (18).

Since sensitization in a canine model results in BHR through an increased content of smMLCK (138 kD), and differences have been observed in the contractile reactivity of sensitized and nonsensitized human tissue (14), we conducted the present study to determine whether the content of smMLCK and activity in ASM differ in these two types of human tissue. Additionally, we examined whether the nmMLCK isoform exists in freshly isolated human ASM.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Human ASM Preparation

Macroscopically normal human lung tissue was obtained from 10 subjects. In two cases, tissue was obtained from donors whose lungs were not suitable for transplantation because of trauma; the remaining eight subjects were undergoing pulmonary transplantation. Details of the subjects are given in Table 1. Approval for all experiments with human lung was provided by the Human Ethics Committee of the University of Sydney and the Central Sydney Area Health Service.

On the day of retrieval, high-order bronchi (main to second-generation bronchi) distal to the carina were excised from the surrounding parenchyma and the bronchus was opened by cutting through the cartilaginous section (using sterile technique). Following manual removal of the epithelial layer, the underlying smooth-muscle bundles were collected with forceps and washed in ice-cold sterile saline before storage at 70° C for later biochemical analysis.

In addition, lower-order bronchi (third to sixth generation, from 3- to 5-mm internal diameter) were dissected free of surrounding parenchyma, cut into ~ 4-mm-long segments, and used for examination of sensitization status. These bronchi were stored at 4° C in Krebs-Henseleit solution (118.4 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl₂ · 2 H₂O, 1.2 mM MgSO₄ · 7 H₂O, 1.2 mM KH₂PO₄, 25 mM NaHCO₃ and 11.1 mM D-glucose; Sigma Chemical Company, St. Louis, MO) aerated with 5% CO₂ in O₂ (carbogen) until assessment of sensitization status was begun, which was within 24 h.

Unless otherwise specified, all chemicals used in the study were purchased from the Sigma Chemical Company.

Assessment of Sensitization Status of Human Lung Tissue

Bronchial rings were suspended in 5-ml organ baths containing Krebs- Henseleit solution bubbled with carbogen and maintained at 37° C (14, 15, 17, 18). A load of 1 to 2 g was placed on the tissues as determined by tissue size (19), since we have previously shown that such loads are optimal in tissues of the corresponding dimensions (20). Contractile responses were measured isometrically with Grass FTO3 transducers, and were recorded on a Grass Model 7P polygraph (Grass Instruments, Quincy, MA). After an equilibration period of ~ 1 h, during which time the tissue was washed and a stable baseline tone was achieved, 10 µl each of four common allergen extracts (timothy phloem pratense, 1:20 [wt/vol]; Alternaria tenuis, 1:10 [wt/vol]; Dermatophagoides pteronyssinus, 30,000 biologic allergen units [BAU]/ml; and cat pelt, 10,000 BAU/ml; Miles Laboratories, Elkhart, IN) were added sequentially to the bath. More than 97% of the atopic population of Australia exhibits a positive skin prick test reaction to these allergens (21). A contractile response to any of these allergens indicated that the tissue was sensitized. Tissues that did not contract in response to any of these allergens, but contracted to a subsequent maximal dose of acetylcholine (ACh) (50 µl of 0.1 M ACh) were classified as nonsensitized.

Measurement of Myosin Heavy Chain Content

Human ASM was thawed from 70° C, its wet weight was measured on a high-precision balance, and it was immediately placed into 1 ml of ice-cold myofibril homogenization buffer (20 mM imidazole, 60 mM KCl, 1 mM L-cysteine, 1 mM MgCl₂ · 6 H₂O, pH 6.9) containing 0.5% Triton X-100 (22), and was homogenized for 30 s at 22,500 rpm with a T25 Ultra-Turrax disperser (IKA, Staufen, Germany). After centrifugation at 10,000 × g for 10 min at 4° C with a fixed-angle (34-degrees) rotor (SS-34; Sorvall, Newton, CT), the supernatant was discarded and the resulting pellet was resuspended in 1 ml of myofibril homogenization buffer containing 0.3% Triton X-100 before recentrifugation. This procedure was repeated twice further in myofibril homogenization buffer without Triton X-100. After the final centrifugation, the pellet was resuspended in a sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer (50 mM Tris, 100 mM dithiothreitol [DTT], 2% SDS, 0.1% bromphenol blue, 10% glycerol, pH 6.8) and heated at ~ 95° C for 5 min.

SDS-PAGE was done with a 5% homogeneous separating gel (with 3% stacking gel). To semiquantify the amount of myosin heavy chain (MHC) in unknown samples, we applied aliquots (20 µl) of ASM extracts to the gels, along with a range of dilutions (from 15.6 to 125 µg MHC/ml) of a commercially available MHC standard (from rabbit muscle; Sigma). After electrophoresis (at 200 V constant), the gels were stained with Coomassie blue R-250 (Bio-Rad, Hercules, CA) and were subjected to scanning densitometry with a GS 690 laser densitometer with molecular analysis software (Bio-Rad). After densitometric analysis of known amounts of MHC standards (over a range of dilutions), a standard curve was established and the linear regression equation for the curve was used to semiquantify the MHC content in each ASM sample.

To confirm the identity of MHC in the ASM extracts, we performed MHC immunoblotting. Following 5% SDS-PAGE, sample proteins (along with broad-range prestained markers [33.5 to 230 kD]; Sigma) and biotinylated protein molecular-weight markers (6.5 to 165 kD; New England BioLabs, Beverly, MA) were transferred to nitrocellulose filters (0.45 µm; Schleicher & Schuell, Keene, NH), using transfer buffer (48 mM Tris, 39 mM glycine, 20% methanol, 1.3 mM SDS, pH 9.2) at 400 mA. After transfer, the nitrocellulose was blocked (blocking buffer: 5% skim milk powder, 0.1% Tween 20, 20 mM Tris, 137 mM NaCl, pH 7.6) and immunoblotted with a 1:500 dilution of antimyosin (smooth muscle) primary antibody (IgG₁ monoclonal, clone hSM-V; Sigma) and an antimouse IgG (heavy- and light-chain [H + L]) secondary antibody conjugated to horseradish peroxidase (1:3,000; Boehringer Mannheim, Indianapolis, IN). Both the primary and secondary antibodies were prepared in blocking buffer. Detection of MHC bands was done with an enhanced chemiluminescence (ECL) immunoblot detection system (Amersham International, Buckinghamshire, UK).

Measurement of smMLCK Content

Smooth muscle was thawed, its wet weight was measured, and it was immediately placed into 1 ml of ice-cold extraction buffer (22) (40 mM Na₄P₂O₇, 1 mM MgCl₂ · 6 H₂O, 1 mM ethylene glycol-bis-(-aminoethyl ether)-N,N,N',N'-tetraacetic acid [EGTA], 1 mM DTT, 250 µM phenylmethyl sulfonyl fluoride [PMSF], 20 µM leupeptin, pH 8.8). After homogenization for 30 s at 22,500 rpm, the homogenate was centrifuged at 15,000 × g for 10 min at 4° C. The samples were then separated into supernatant and pellet fractions. Protein extracts were prepared from the supernatant fraction by mixing the supernatants in a 1:4 (vol/vol) ratio with 5× concentrated SDS-PAGE sample buffer (250 mM Tris, 500 mM DTT, 10% SDS, 0.5% bromphenol blue, 50% glycerol, pH 6.8) before heating at ~ 95° C for 5 min. The pellet fraction was further extracted by resuspension in 500 µl SDS-PAGE sample buffer. The resuspended pellets were then heated at ~ 95° C for 12 min with frequent shaking. These latter two fractions were then pooled and concentrated by centrifugation at 3,000 × g for 2 h at room temperature (RT), using a 30-kD-cutoff microconcentrator (Centricon-30; Amicon, Bedford, MA). The concentrate (> 30 kD) was collected by inverting the retentate cup and centrifuging at 500 × g for 5 min at RT. The volume of the resulting concentrates was measured and the samples were then used for SDS-PAGE. To determine the molecular weight of MLCK extracted from human ASM, aliquots of broad-range prestained markers (33.5 to 230 kD; Sigma) and biotinylated protein molecular-weight markers (6.5 to 165 kD; New England BioLabs) were also applied to the gels.

Following electrophoresis, proteins were transferred to nitrocellulose filters (0.45 µm; Bio-Rad), using transfer buffer (25 mM Tris, 0.192 M glycine, 20% methanol, 0.5% SDS, pH 8.5). After transfer, the section of nitrocellulose filter containing the molecular-weight markers was stained with Coomassie blue and the R_F (electrophoretic mobility of each protein marker compared with the bromphenol blue solvent front) was measured after the nitrocellulose filter had been destained. The remainder of the nitrocellulose filter was subjected to immunoblotting with a 1:20,000 dilution of a mouse monoclonal anti-MLCK antibody (IgG_2b, clone K36; Sigma) followed by a 1:1,000 dilution of polyvalent alkaline phosphatase-conjugated antimouse immunoglobulin antibody. Detection of MLCK bands was done by incubating the nitrocellulose filter in a chromogen solution (0.1 M Tris, 5 mM MgCl₂ · 6 H₂O, 100 mM NaCl, pH 9.5) containing 0.5% 5-bromo-4-chloro-3-indolyl-phosphate (p-toluidine salt) and 0.5% nitroblue tetrazolium. MLCK content was semiquantified through scanning densitometry with a GS 690 laser densitometer.

Expression of MLCK Isoforms

In order to detect which MLCK isoforms were present in freshly isolated ASM, we compared the immunoblots obtained with the K36 mouse monoclonal anti-MLCK antibody from samples of ASM from nonsensitized and sensitized bronchi with lysates from A10 cells (rat embryonic thoracic aorta cell line), which express smMLCK, and with transfected rat embryo fibroblasts (REF) expressing recombinant nmMLCK as positive controls. REF cells have no detectable smMLCK or nmMLCK upon Northern or Western blotting (12), and we therefore used REF cells transfected with vector only as negative controls. Both A10 and transfected/mock-transfected REF cell lysates were generous gifts of Dr. P. J. Gallagher of Indiana University. We also examined lysates from human cultured ASM cells (passages, 3 to 5 from trachea obtained from lung transplant donors (n = 4) (a gift of Dr. R. A. Panettieri, Jr. of the University of Pennsylvania).

Lysates (prepared in SDS-PAGE sample buffer and then heated at ~ 95° C for 3 min) were applied to 5% SDS-PAGE homogeneous separating gels along with broad-range prestained markers (33.5 to 230 kD; Sigma) and biotinylated protein molecular-weight markers (6.5 to 165 kD; New England BioLabs) to aid in the characterization of the MLCK isoforms. Following electrophoresis and transfer, the nitrocellulose filters were blocked and then incubated with a 1:20,000 dilution of K36 mouse monoclonal anti-MLCK antibody, followed by antimouse IgG (H + L) secondary antibody conjugated to horseradish peroxidase (1:3,000). An antibiotin antibody conjugated to horseradish peroxidase (1:1,000; New England BioLabs) was added to the secondary antibody solution to detect the biotinylated protein molecular-weight markers. Detection was done with the ECL immunoblot detection system.

Measurement of MLC₂₀ Phosphorylation

In an attempt to measure the activity of MLCK in human ASM from nonsensitized and sensitized bronchi, we measured the phosphorylation of MLC₂₀ by using a modified nondenaturing PAGE method (7, 23, 24) to separate monophosphorylated and unphosphorylated MLC₂₀, since the two species have different electrical charges. As this method was previously used in the canine model of AHR to measure phosphorylation of MLC₂₀ (7), we used canine tracheal ASM (a generous gift from Dr. K. X. F. Yang of the University of Sydney) as a positive control to examine the utility of the method for measuring MLC₂₀ phosphorylation in human ASM.

Human and canine ASM were thawed from 70° C, their wet weight was measured on a high-precision balance, and they were then immediately placed into 1 ml of ice-cold MLCK activity buffer (20 mM imidazole, 60 mM KCl, 1 mM cysteine, 1 mM MgCl₂ · 6 H₂O, 10 mM sodium azide, 1 mM ouabain, 0.25 mM PMSF, 0.001% leupeptin, and 1 mM DTT, pH 7.5) and homogenized (for 30 s at 22,500 rpm). Tissue homogenates, at a content of ~ 10 mg per assay tube, were equilibrated for 10 min at 37° C before the assay procedure was begun. The assay was initiated by adding 1 mM Mg²⁺-adenosine triphosphate and 1 mM CaCl₂ to the homogenates and incubating for 10 min (with constant mixing) before the reaction was stopped by precipitation with ice-cold 5% trichloroacetic acid in acetone on dry ice. The precipitated homogenates were then centrifuged (at 15,000 × g for 10 min at 4° C) before suspending the pellets in acetone containing 10 mM DTT and placing the suspensions on a rotator (2,400 rpm) for 1 h at RT. These preparations were then recentrifuged and the pellet was solubilized in a urea sample buffer (6.4 M urea, 17 mM Tris, 19.5 mM glycine, 10 mM DTT, 0.04% bromphenol blue, 10 mM EGTA, 1 mM ethylenediamine tetraacetic acid [EDTA], 5 mM NaF, 1 mM PMSF, pH 8.6) and rotated for 2 h at RT. After a further centrifugation step, the samples (100 µl) were loaded onto 10% native polyacrylamide gels (20 mM Tris, 22 mM glycine, 40% glycerol. 10% acrylamide, 0.5% bisacrylamide, pH 8.6) that had been preelectrophoresed for 1.5 h at 400 V at 15° C in a gel electrophoresis buffer containing 20 mM Tris, 22 mM glycine, 1 mM DTT, and 1 mM sodium thioglycollate, pH 8.6. After electrophoresis for 4 h at 400 V at 15° C, separated proteins were transferred to nitrocellulose filters at 1500 mA for 3 h at 15° C, using a transfer buffer containing 25 mM Tris, 0.192 M glycine, 20% methanol, and Na₂HPO₄ · 12 H₂O, pH 7.6. After transfer, the unphosphorylated and phosphorylated MLC₂₀ bands were detected by immunoblotting with a 1:500 dilution of a mouse monoclonal anti-MLC₂₀ antibody (IgM, clone MY-21; Sigma) followed by an antimouse IgM (H + L) secondary antibody conjugated to horseradish peroxidase (1: 3,000) and ECL.

Statistical Analysis

Relationships between two variables were analyzed through linear regression (StatView Version 4.51; Abacus Concepts Inc., Berkeley, CA), and the results were summarized as the regression equation and level of significance (p). Comparisons of two populations (summarized as mean ± SEM) were made with Student's unpaired t test (StatView) after the populations were shown be normally distributed through use of the Kolmogorov-Smirnov Test (StatView). Differences were considered statistically significant if the probability of the effect being due to chance alone was less than 5% (p < 0.05).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sensitization Status of Human Lung Tissue

Of the 10 subjects from whom tissue was obtained, five were classified as nonsensitized and five as sensitized (Table 1). There was no contraction upon the addition of allergens to nonsensitized tissue, and all such tissue specimens were viable (contracted in response to ACh). For tissue classified as sensitized, the addition of allergens elicited a contraction of 565.2 ± 132.3 mg (n = 5). There was no significant difference (p = 0.90) between the contraction elicited by a maximal dose of ACh in nonsensitized (2,612.5 ± 1,146.2 mg; n = 5) and sensitized tissue (2,449.0 ± 415.5 mg; n = 5). Further, the ages of sensitized and nonsensitized subjects from whom tissue was obtained were not significantly different (29.0 ± 7.6 yr [nonsensitized] versus 40.6 ± 6.8 yr [sensitized] [p = 0.29]), and the ratio of males to females in each group was similar (Table 1).

Measurement of MHC Content

To ensure that there was no difference in the amount of ASM collected through microscopic retrieval of tissue from main-order bronchi from nonsensitized and sensitized subjects, we compared their content of MHC per milligram of ASM (wet weight). After densitometric analysis of known amounts of MHC standards (over the range from 0.3 to 2.5 µg MHC/ml), the following regression equation was obtained: y = 0.056 + 0.411x (r² = 0.969). This regression equation was used to semiquantify the MHC content of nonsensitized and sensitized ASM. As shown in Figure 1A, there was no significant difference (p = 0.636) in the MHC content (µg) per milligram of ASM (wet weight) from the two study groups. Further, to confirm the identity of MHC in the ASM extracts, we performed immunoblotting for MHC. Figure 1B shows representative immunoblots of MHC in ASM from nonsensitized bronchi (lane 2) and sensitized bronchi (lane 3).

View larger version (29K): [in this window] [in a new window]   Figure 1.   Content of MHC in ASM from nonsensitized and sensitized bronchi is equal. (A) Content of MHC (µg) per milligram of ASM (wet weight) from nonsensitized (n = 5; open bar) and sensitized (n = 5; hatched bar) bronchi. Samples were prepared from freshly isolated ASM and electrophoresed with 5% SDS-PAGE before staining with Coomassie blue R-250. Statistical analysis was done with Student's unpaired t test. (B) To confirm the identity of MHC in ASM, MHC immunoblotting was performed with a monoclonal antimyosin (smooth muscle) antibody (clone hSM-V). Lane 1: biotinylated protein molecular-weight markers (relative kD are shown); lane 2: ASM from nonsensitized bronchi (0.56 µg MHC/ mg ASM); lane 3: ASM from sensitized bronchi (0.68 µg MHC/mg ASM). Immunoblots are representative results.

Measurement of smMLCK Content

Because this was the first study to examine smMLCK from human ASM, it was necessary to determine the molecular weight of smMLCK by comparison with prestained molecular-weight markers. The relationship between the log₁₀ of the molecular-weight markers (in kD) and R_F was described by the equation y = 2.430  0.739x (r² = 0.990). The average R_F values of the smMLCK bands from nonsensitized and sensitized ASM were 0.402 ± 0.005 (n = 5) and 0.399 ± 0.004 (n = 5), respectively. These values were not significantly different (p = 0.563). Therefore, by combining the R_F values for the two groups (i.e., R_F = 0.401 ± 0.003; n = 10), the molecular weight for smMLCK from human ASM was calculated to be 136.0 ± 0.7 kD.

Figure 2 shows the content of smMLCK per milligram of ASM (wet weight) from nonsensitized (n = 5) and sensitized (n = 5) bronchi. The average smMLCK content in ASM from sensitized bronchi was significantly higher (p = 0.049) than that in ASM from nonsensitized tissue (11.9 ± 3.3 versus 4.1 ± 0.7 smMLCK (a.u.)/mg ASM [wet weight], respectively).

View larger version (29K): [in this window] [in a new window]   Figure 2.   smMLCK content is increased in human sensitized ASM. Content of smMLCK (a.u.) per milligram of ASM (wet weight) from non-sensitized (n = 5; open bar) and sensitized (n = 5; hatched bar) bronchi. Samples were prepared from freshly isolated ASM and electrophoresed with 5% SDS-PAGE before immunoblotting with a monoclonal anti-MLCK antibody (clone K36). smMLCK content in nonsensitized versus sensitized ASM was compared semiquantitatively through scanning densitometry. Statistical analysis was done with Student's unpaired t test.

Expression of MLCK Isoforms

In order to detect which MLCK isoforms were present in freshly isolated ASM, we compared immunoblots of ASM samples from nonsensitized and sensitized bronchi with lysates from A10 cells and transfected REF cells (expressing recombinant nmMLCK only) as positive controls. Negative controls were REF cells transfected with vector only. Comparision with lysates from human cultured ASM cell lines (n = 4) was also done. The immunoblots shown in Figure 3 are composites of representative results. Interestingly, the only MLCK isoform expressed in ASM from both nonsensitized and sensitized tissue (lane 2 and lane 3, respectively) was apparently smMLCK (136 kD). There was no nmMLCK (210 to 220 kD) in these samples. However, cultured ASM cells from different cell lines (lanes 7 to 10) expressed either smMLCK alone, or both the smMLCK and nmMLCK isoforms.

View larger version (40K): [in this window] [in a new window]   Figure 3.   Freshly isolated ASM expresses only smMLCK, whereas cultured ASM cells can express either smMLCK alone or both smMLCK and nmMLCK isoforms. ASM samples from nonsensitized and sensitized bronchi were compared with those from A10 cell lysates (which express smMLCK) and transfected REF cells (expressing recombinant nmMLCK only) as positive controls. Negative controls were REF cells transfected with vector only. Comparison was also done with lysates from human cultured ASM cell lines (n = 4). Lysates were electrophoresed with 5% SDS-PAGE and immunoblotted with an anti-MLCK monoclonal antibody (clone K36), which detects both smMLCK and nmMLCK. Lane 1: biotinylated protein molecular weight markers (relative kD are shown); lane 2: ASM from nonsensitized bronchi; lane 3: ASM from sensitized bronchi; lane 4: A10 cells; lane 5: REF cells expressing recombinant nmMLCK; lane 6: REF cells transfected with vector only; lanes 7 to 10: lysates from different human cultured ASM cell lines (n = 4). Immunoblots shown are composites of representative results.

Measurement of MLC₂₀ Phosphorylation

Figure 4 shows the results of an attempt to measure the activity of MLCK in human ASM by detecting the phosphorylation of MLC₂₀ with a nondenaturing PAGE method for separating monophosphorylated from unphosphorylated MLC₂₀. As previously reported (7), this method can be used to measure phosphorylation of MLC₂₀ in canine tracheal ASM. Lane 3 shows unphosphorylated canine MLC₂₀ (before stimulation of MLCK activity), and lane 4 shows the electrophoretic pattern obtained as a result of MLCK activity (with both unphosphorylated MLC₂₀ and monophosphorylated MLC₂₀ [MLC₂₀-P] present). However, as shown in lanes 1 and 2, it was not possible to use this method to measure the activity of MLCK in human ASM, since it was not possible to detect unphosphorylated MLC₂₀ in human ASM.

View larger version (24K): [in this window] [in a new window]   Figure 4.   Activity of smMLCK in human ASM cannot be measured by detecting MLC20 phosphorylation with nondenaturing PAGE. In an attempt to measure the activity of MLCK in human ASM (as compared with canine ASM as a positive control), phosphorylation of MLC20 was measured by using a 10% nondenaturing PAGE method to separate unphosphorylated MLC20 and monophosphorylated MLC20 (MLC20-P). MLCK activity in homogenates of human and canine ASM (10 mg) was initiated by adding 1 mM Mg2+-ATP and 1 mM CaCl2 for 10 min. The reactions were stopped by precipitation, and the samples were prepared in a urea sample buffer, electrophoresed, and immunoblotted with a monoclonal anti-MLC20 antibody (clone MY-21). Lanes 1 and 2: human ASM before and after addition of 1 mM Mg2+-ATP and 1 mM CaCl2 for 10 min; lanes 3 and 4: canine ASM before and after addition of 1 mM Mg2+-ATP and 1 mM CaCl2 for 10 min. Immunoblots are representative results.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We found that the content of smMLCK in smooth muscle from human sensitized airways was significantly higher than that in ASM from nonsensitized airways. This was the first study to demonstrate such a difference in human tissue. Further, smMLCK was the only MLCK isoform present in freshly isolated ASM. However, using lysates from cultured ASM, we found that in contrast, there was considerable heterogeneity in the MLCK isoforms expressed. Some cell lines expressed smMLCK only, whereas others expressed both smMLCK and nmMLCK. This expression pattern was independent of passage number.

Smooth-muscle contraction occurs via a sliding-filament process that involves actin and myosin (25). Actin filaments are composed of two linear polymers of a 42-kD globular protein, wrapped together in a helical configuration. Myosin filaments are thick, bipolar, and arranged asymmetrically in a hexameric structure comprising one pair of heavy chains (each ~ 200 kD) and two pairs of light chains (one pair of 17 kD each [MLC₁₇] and the other of 20 kD each [MLC₂₀]). The globular head section of myosin contains both the binding sites for attachment to actin and the enzymatic (actomyosin ATPase) sites that cleave ATP and so provide the energy necessary to fuel the binding reactions. The sliding of actin and myosin filaments past each other is achieved by attachment of the globular heads of the myosin molecule to actin (so-called crossbridge formation), a flexing change in the configuration of the myosin head with respect to actin, and detachment of the myosin head followed by its subsequent, rapid reattachment at another site further down the actin filament. This rapid cyclic attachment and detachment of crossbridges (crossbridge cycling) is driven by the energy derived from the breakdown of ATP by actomyosin ATPase. The phosphorylation of MLC₂₀ is a key step in smooth-muscle contraction, since it allows stimulation of actomyosin ATPase (8). The enzyme responsible is a Ca²⁺/calmodulin-dependent MLCK (26), which phosphorylates MLC₂₀ at the serine-19 position (27) and results in conformational changes that increase ATP hydrolysis and actin binding (28).

The molecular mechanism responsible for the increased maximal shortening velocity that accompanies antigen sensitization in a canine model of allergic AHR (3) has been partly elucidated. Immunoblot analysis of tracheal and bronchial smooth muscle from ragweed pollen-sensitized dogs showed that sensitization increases the content and total activity of MLCK (although the enzyme-specific activity of this MLCK remains constant) over that of littermate controls (7). Reflecting the increased MLCK activity was increased actomyosin ATPase activity (29) and phosphorylation of MLC₂₀ (30) in ASM from sensitized animals. Further, there was no difference in the activity of calmodulin in ASM from nonsensitized as opposed to sensitized animals (31, 32). Accordingly, it follows that an increased content (and therefore activity) of MLCK in ASM would result in a greater velocity of contraction due to an increase in the actin-myosin crossbridge cycling rate. These results led the authors (7, 31) to suggest that an increased MLCK content in ASM from sensitized animals may account for the observed hyperresponsiveness and increased contractile reactivity of this muscle. The isoform of MLCK responsible for this appears to be smMLCK (138 kD) (6).

We wished to extend these findings to humans in order to improve our understanding of the mechanisms underlying the BHR characteristic of asthma. Ideally, we wanted to measure the content of smMLCK in human ASM from asthmatic subjects. However, tissue from asthmatic subjects rarely becomes available for study. We therefore used a model of human allergic asthma. On the basis of the finding by ourselves (14, 15, 17, 18) and others (16) of an increased reactivity in vitro in airway preparations from both passively and actively sensitized patients, we hypothesized that differences in smMLCK could be apparent in tissue from these patients. We used human lung tissue that we classified as sensitized or nonsensitized, and separated smMLCK from other smooth-muscle proteins by using SDS-PAGE and immunoblotting. smMLCKs of smooth muscle from a number of species have been characterized through immunoblotting (9), and for some species complete amino acid sequences of these enzymes have been obtained (10, 11, 13). The present study, however, was the first to examine smMLCK in human ASM. The resultant molecular weight (136 kD) of this isoform is similar to that obtained for smMLCK from human myometrium (137 kD) (33), although the degree of sequence homology is currently unknown.

We also wished to examine whether the other major isoform of MLCK, nmMLCK (210 to 220 kD) was also present in freshly isolated ASM, since it has been reported than nmMLCK is present in cultured canine ASM cells (34, 35). In these latter reports (34, 35), the polyclonal antibody (clone D119) used to detect nmMLCK was described as having been raised in rabbits, against a peptide representing the 12-amino-acid repeat sequence (KPVGNAKPAETL) between amino acid residues 102 and 293 in rabbit and bovine smMLCK (11). The D119 antibody was originally used (12) to detect a protein of higher molecular mass (208 kD) in cultures of primary cells, cell lines, whole embryos, and embryonic tissues. However, this MLCK isoform (called embryonic MLCK) is unique and is not nmMLCK (12). Although it has been reported that D119 preferentially detects nmMLCK as compared with smMLCK (36), the use of D119 as an antibody for nmMLCK is controversial, since it does not react very well with smMLCK or nmMLCK (both recombinant and endogenous proteins) (personal communication, Dr. P. J. Gallagher, Indiana University). Therefore, we chose to use the K36 antibody in our study. This antibody detects both the smMLCK and nmMLCK isoforms, since smMLCK and nmMLCK arise from the same gene (11), and the entire amino acid sequence of smMLCK is therefore contained within the nmMLCK sequence. Accordingly, immunoblotting with the same monoclonal antibody (clone K36) will detect both smMLCK and nmMLCK, since this antibody detects a common epitope at the NH₂-terminal of these two isoforms (residues 29 to 80) (12). We found that nmMLCK was not present in freshly isolated ASM, but could be detected in half of the lysates from human cultured ASM cells. However, all human cell lines expressed smMLCK, and this expression pattern was independent of passage number. The mechanistic basis for the heterogeneity of MLCK isoforms in cultured ASM cells, and the mechanism by which control of isoform expression is regulated, are unclear, but warrant further investigation.

After measuring the content of smMLCK from freshly isolated human ASM by densitometry, we found approximately three times more smMLCK in ASM from sensitized tissue than in ASM from nonsensitized controls. The content of smMLCK was normalized for wet weight of ASM tissue according to the method established by Jiang and colleagues (7). This greater content of smMLCK in ASM from sensitized tissue was unaccompanied by any concomitant increase in MHC. However, since our study did not specifically examine the MHC isoform distribution in ASM (37, 38), we cannot exclude the possibility that differences in the MHC isoform pattern also exist in human ASM, although Stephens and coworkers (35) found no significant differences in the content of MHC isoforms in canine ASM.

Although it could be presumed that an increased content of MLCK in human sensitized ASM may result in greater MLCK activity, proof of this would require measurement of MLC₂₀ phosphorylation. To this end, we attempted to measure MLC₂₀ phosphorylation in human ASM, using methods previously established for canine ASM (7, 23) and a sample of canine tracheal ASM. Although activity of MLCK in canine ASM could be measured by detecting MLC₂₀ phosphorylation with nondenaturing PAGE (Figure 4), it was not possible with this method to detect activity of MLCK in human ASM. This may have been due either to the relatively low content of MLC₂₀ in human ASM (as compared with canine ASM), or to low cross-reactivity with human MLC₂₀ by the mouse monoclonal anti-MLC₂₀ antibody (IgM, clone MY-21) raised against chicken lens membrane and used in our assay. Perhaps a greater amount of human ASM would have increased the sensitivity of the assay, but, it is unlikely that we could have obtained more than 10 mg of ASM from pulmonary transplant recipients. Therefore, it was not possible to measure human MLC₂₀ phosphorylation in nonsensitized and sensitized human ASM through nondenaturing PAGE (7, 23).

Because sensitization is associated with increased smMLCK content, the amount of smMLCK in smooth muscle may be able to be modulated in vivo. However, this modulation does not appear to be exclusively restricted to ASM. Recently, Stephens and colleagues (39) extended their original work with canine sensitized ASM to examine the smMLCK content of vascular smooth muscle of the canine saphenous vein, and found that ragweed sensitization also increased the content of smMLCK in vascular smooth muscle. In a rat model of cardiac hypertrophy (40) and in rats with diabetic cardiomyopathy (24), duration- and region-specific changes were found in the levels of smMLCK and MLC₂₀ phosphorylation that were correlated with cardiac failure. Further, in human myometrial tissue, the amount of smMLCK per myocyte was found to be increased during pregnancy, suggesting that the myosin-phosphorylation-stress relationship is adapted to result in a pregnancy-associated enhancement in myometrial stress-generating capacity without appreciable increases in the myometrial content of actin or myosin (33).

The mechanisms underlying this in vivo modulation of smMLCK content are unclear at present, but increasing evidence suggests that expression of ASM contractile proteins is not static but instead undergoes regulation in response to the physiologic and pathologic environment, such as in phenotypic modulation of ASM cells in culture (41). This is further supported by our evidence of heterogeneity of smMLCK isoforms in human cultured ASM cells. The transcriptional regulation of gene expression in ASM is now a subject of intense investigation (42), and although the role of inflammatory mediators in ASM contractile-protein gene expression is less well understood, it is interesting to speculate that the proximity of increased numbers of mast cells in ASM of sensitized airways (43) may play a role in regulating smMLCK gene expression.

In summary, we examined the smMLCK content in actively sensitized ASM from humans, and achieved similar results to those obtained in a canine model (7), finding that sensitization increases smMLCK content. Our study was the first to examine smMLCK in human ASM, and suggests that increased content of smMLCK may be one of the processes responsible for the enhanced contractile reactivity observed in sensitized tissue.