The purpose of this study was to determine the mechanism by which hydrogen peroxide (H2O2), an important inflammatory mediator, relaxes canine tracheal smooth muscle (CTSM). H2O2caused concentration-dependent relaxations of CTSM strips contracted with ACh or isotonic KCl [EC50 of 0.24 ± 0.04 (SE) and 0.23 ± 0.04 mM, respectively]. Indomethacin (10 μM) decreased the sensitivity of both KCl- and ACh-contracted strips to H2O2. H2O2increased intracellular cAMP levels, an increase that was abolished by indomethacin. H2O2did not affect intracellular cGMP levels. In strips treated with indomethacin and contracted with ACh or isotonic KCl, H2O2-evoked relaxations were accompanied by increases in intracellular Ca2+ concentration and decreases in regulatory myosin light chain phosphorylation. We conclude that H2O2decreases Ca2+ sensitivity in CTSM by decreasing regulatory myosin light chain phosphorylation through inhibition of myosin light chain kinase and/or activation of smooth muscle protein phosphatases.
- canine tracheal smooth muscle
- guanosine 3′,5′-cyclic monophosphate
- adenosine 3′,5′-cyclic monophosphate
- regulatory myosin light chain
the appreciation of asthma as a chronic inflammatory disease of the airways and the recognition of the role of inflammation in the pathogenesis of acute lung injury has drawn attention to the effects of inflammation on airway smooth muscle (ASM) contractility. Hydrogen peroxide (H2O2), an important inflammatory mediator, has diverse effects on ASM in vitro. In unstimulated ASM, exposure to H2O2increases resting tone (2, 10, 14, 22-24, 27, 28). For example, in human ASM, H2O2causes a concentration-dependent increase in force that has an epithelial component and is inhibited by indomethacin (23). Conversely, in ASM contracted with ACh or a high-K+ solution, H2O2causes relaxation (9, 10, 14, 23) or a further increase in tone (2,28). This relaxation is produced, in part, by prostaglandins synthesized in the ASM and may be modulated by the airway epithelium (9, 10). However, even with a complete block of prostaglandin synthesis produced by indomethacin, H2O2still relaxes canine ASM contracted with acetylcholine (9). The mechanism of this indomethacin-insensitive component of relaxation, which has also been demonstrated in guinea pig (9) and equine (22) ASM, is unknown.
In general, bronchodilators may reduce the intracellular Ca2+ concentration ([Ca2+]i), reduce the amount of force produced by a given [Ca2+]i(the “Ca2+ sensitivity”), or both. H2O2increases [Ca2+]iin several cell types (21, 26) including vascular smooth muscle (16,25) and cat tracheal smooth muscle (2) by increasing Ca2+ influx and releasing Ca2+ from intracellular stores. If Ca2+ sensitivity was not affected by H2O2, these increases would be consistent with the observed contractile, but not relaxing, effects of H2O2. However, Iesaki et al. (16), studying endothelium-denuded smooth muscle from rabbit aorta contracted with phenylephrine, noted that H2O2simultaneously increased [Ca2+]iand decreased force, which suggests that H2O2decreased Ca2+ sensitivity. Whether such effects are also present in ASM is unknown. The purpose of this study was to determine the mechanism by which H2O2produces indomethacin-insensitive relaxation of canine ASM.
A 5- to 10-cm segment of cervical trachea was removed from mongrel dogs that had been anesthetized with pentobarbital sodium (30 mg/kg) and exsanguinated. The segment was immediately immersed in chilled physiological salt solution (PSS) with the following composition (in mM): 0.8 MgSO4, 1.2 KH2PO4, 3.4 KCl, 2.4 CaCl2, 110.5 NaCl, 25.7 NaHCO3, and 5.6 dextrose. After removal of fat, connective tissue, and epithelium, two sizes of smooth muscle strips were prepared: 3 mm wide × 10 mm long for organ bath experiments and 0.5 mm wide × 2–3 mm long for superfusion studies.
Organ Bath Studies
Tracheal strips were mounted in 25-ml organ baths filled with PSS (pH 7.4) and bubbled with 94% O2 and 6% CO2 (37°C). The strips were stretched between repeated contractions with 1 μM ACh until the optimal length was achieved.
Tracheal strips were incubated in 10 ml of PSS (bubbled with 94% O2 and 6% CO2; 22°C) containing the acetoxymethyl ester of fura 2 (fura 2-AM; 5 μM) for 3 h. Fura 2-AM was dissolved in dimethyl sulfoxide and 0.02% cremophor. After fura 2-AM loading, the strips were mounted in a 0.1-ml quartz cuvette and continuously superfused with PSS at 37°C (bubbled with 94% O2 and 6% CO2) for 30 min to remove excess fura 2-AM. One end of the strips was anchored via microforceps to a micrometer, and the other end was anchored via microforceps to an isometric force transducer. During a 2-h equilibration period, the length of the strips was increased after repeated contractions induced by 1 μM ACh until the optimal length was obtained.
Fura 2 fluorescence intensity was measured with a photometric system that measures optical and mechanical parameters of isolated tissue simultaneously (15). Light from a mercury high-pressure lamp was monochromatically filtered to restrict excitation light to 340- and 380-nm wavelengths. Fluorescence emitted from the strips was filtered at 500 ± 5 nm and detected by a photomultiplier assembly (Scientific Instruments, Heidelberg, Germany). The emission fluorescence intensities due to excitation at 340-nm (F340) and 380-nm (F380) wavelengths were measured, and the F340-to-F380ratio (F340/F380) was used as an index of [Ca2+]i. Absolute values of [Ca2+]iwere not calculated because the dissociation constant of fura 2 for Ca2+ within the smooth muscle cytosol is different from that measured in the absence of protein (18). In preliminary work, we found that H2O2alone did not affect F340 and F380 when added to solutions containing Ca2+ and fura 2 (data not shown).
Cyclic Nucleotide Measurements
Muscle strips were weighed and homogenized in 4 ml of cold (2°C) 95% ethanol with a ground-glass pestle and homogenizing tube. The precipitated pellet was separated from the soluble extract by centrifugation at 4,000 g for 10 min. The soluble extract was evaporated to dryness at ∼55°C under a stream of nitrogen and was then suspended in 0.3 ml of 4 mM EDTA (pH 7.5). cAMP or cGMP recovery determinations were made by adding [3H]cAMP (1.25 mCi) or [3H]cGMP (0.4 mCi), respectively, as tracers. Commercially available radioimmunoassay kits were used to determine the concentrations of cAMP and cGMP in the soluble extract (7). The protein content of the precipitated pellet was determined by the method described by Lowry et al. (19), with bovine serum albumin dissolved in 1 N NaOH as the standard. The intracellular concentrations of cAMP ([cAMP]i) and cGMP ([cGMP]i) are expressed as picomoles per milligram of protein.
Regulatory Myosin Light Chain Phosphorylation Measurements
Regulatory myosin light chain (rMLC) phosphorylation was measured in muscle strips by gel electrophoresis followed by Western blot analysis (13). After appropriate experimental interventions, the strips were flash-frozen by rapid immersion in a dry ice-acetone slurry containing 10% (wt/vol) trichloroacetic acid and 10 mM dithiothreitol (DTT; −80°C). Then the frozen strips were thawed to 25°C, washed in acetone containing 10 mM DTT to remove the trichloroacetic acid, and transferred to 200-ml Eppendorf tubes containing 65 ml of extraction buffer (7.3 M urea, 20 mM Tris, 21 mM glycine, and 10 mM DTT). Proteins were separated by glycerol-urea polyacrylamide gel [10% (wt/vol) acrylamide, 0.5% (wt/vol) bis-acrylamide, 40% (vol/vol) glycerol, 20 mM Tris, and 21 mM glycine] electrophoresis. The electrophoresis buffer contained 20 mM Tris, 21 mM glycine, 1 mM DTT, and 1 mM sodium thioglycolate. The gels were subjected to preelectrophoresis for 1 h at 400 V (10°C) to remove urea and to allow DTT and thioglycolate to enter the gels. Then 50 ml of sample were injected into the wells and initially subjected to electrophoresis for 1 h at 100 V and then for 17 h at 400 V (10°C).
For Western blot analysis, the proteins were transferred to nitrocellulose sheets (0.22 μm) for 4 h at 1.6 A (15°C) in a buffer of 25 mM Na2HPO4(pH 7.6). The nitrocellulose sheets were washed twice with 10 mM Tris-buffered saline containing 5% (wt/vol) bovine serum albumin for 1 h (25°C) before being labeled with polyclonal, affinity-purified rabbit anti-rMLC antibody (13). The anti-rMLC antibody was detected with125I-labeled protein A (DuPont, Boston, MA). The unphosphorylated and phosphorylated bands of rMLC were visualized by phosphorimage analysis (PhosphorImager, Molecular Dynamics, Sunnyvale, CA) and quantified by ImageQuant software (Molecular Dynamics). After local background subtraction, rMLC phosphorylation was calculated by integrating the bands corresponding to the mono- and diphosphorylated rMLC as a fraction of the total integration of both the phosphorylated and unphosphorylated rMLC.
Four protocols were conducted on strips prepared from separate sets of animals. For each protocol, the strips were initially contracted with either ACh or isotonic KCl before exposure to H2O2. This experimental design more closely approximates in vivo conditions where ASM tone is increased by constant stimulation with neuronally released ACh.
Effect of indomethacin on relaxation induced by H2O2.
Four pairs of strips from each of five dogs were mounted in organ baths containing PSS at 37°C (bubbled with 94% O2 and 6% CO2). The maximal response of each strip to 100 μM ACh was first determined. The strips then were contracted with either ACh (0.3 μM) or isotonic KCl solution (24 mM) in the absence and presence of indomethacin (10 μM, 30 min preincubation and throughout). These concentrations of ACh and KCl produce ∼50% of the maximal force developed in response to 100 μM ACh in the absence and presence of indomethacin. Fifteen minutes after the addition ACh or KCl, a cumulative concentration-response relationship to H2O2(0.1 μM to 1 mM) was generated in one strip from each pair. The other strip was not exposed to H2O2and served as a control for any effects of time. After generation of the concentration-response relationship, ACh, KCl, and H2O2were washed out, and the strips were allowed to relax. To demonstrate that H2O2did not cause irreversible cell damage, the strips were again exposed to 100 μM ACh, and the force response was measured.
Effect of H2O2 on cyclic nucleotide levels.
Four pairs of strips from each of six dogs were mounted in organ baths containing PSS at 37°C (bubbled with 94% O2 and 6% CO2). One strip from each pair was incubated with 10 μM indomethacin for 30 min before and throughout the experiment. Three pairs of strips were contracted with 0.3 μM ACh. One pair was flash-frozen with liquid nitrogen after 15 min of contraction. Two pairs of strips were then exposed to 1 mM H2O2and were frozen 2 and 10 min after the addition of H2O2. The final pair of strips was exposed to neither ACh nor H2O2and was frozen to provide baseline cyclic nucleotide values.
Effect of H2O2 on force and [Ca2+]i.
Strips were mounted in a superfusion cuvette after incubation in fura 2-AM. The strips were contracted with 0.3 μM ACh or 25 mM KCl. After stable ACh contractions were achieved, 100 μM (n = 5 dogs) or 1 mM (n = 6 dogs) H2O2was applied, and changes in F340/F380and force were recorded. The effect of 1 mM H2O2was also examined in strips treated with 10 μM indomethacin (n = 6 dogs). After stable KCl contractions (in the presence of 10 μM indomethacin) were achieved, 300 μM H2O2(n = 5 dogs) was applied, and changes in F340/F380and force were recorded.
Effect of H2O2 on rMLC phosphorylation during ACh-induced contractions.
Seven strips from each of five dogs were mounted in organ baths containing PSS at 37°C (bubbled with 94% O2 and 6% CO2). All solutions contained 10 μM indomethacin. Six strips were contracted with 0.3 μM ACh; one strip did not receive ACh and was frozen for baseline measurements. Three of the six strips did not receive H2O2and were frozen 1, 15, and 25 min after contraction with ACh to obtain control values. H2O2(100 μM, 300 μM, or 1 mM) was added to the other three strips 15 min after contraction with ACh. These strips were frozen 10 min after the addition of H2O2(see Fig. 4 for schematic of protocol).
Effect of H2O2 on rMLC phosphorylation during KCl-induced contractions.
Four strips from each of four dogs were mounted in organ baths containing PSS at 37°C (bubbled with 94% O2 and 6% CO2). All solutions contained 10 μM indomethacin. Three strips were contracted with 25 mM KCl. One of the three strips was frozen 15 min after contraction with KCl. H2O2(300 μM) was added to one of the three strips 15 min after contraction with KCl and frozen 5 min later. The third strip was frozen 20 min after the addition of KCl and was used as control for the H2O2-treated strip. One strip did not receive KCl and was frozen for baseline measurements.
Fura 2-AM was purchased from Molecular Probes (Eugene, OR). The rMLC antibody was a gift from Dr. Susan Gunst (Department of Physiology and Biophysics, Indiana University, Indianapolis, IN). Protein A was purchased from DuPont. All other drugs and chemicals were purchased from Sigma (St. Louis, MO) and were prepared in distilled water. Na2CO3 was used to solubilize indomethacin. The osmolarity of high-K+ solutions was maintained by substituting equimolar amounts of KCl for NaCl.
Data are expressed as means ± SE;n is the number of dogs. The effect of indomethacin on the concentration-response curves to H2O2was compared by nonlinear regression analysis as described by Meddings et al. (20). In this method, force (F) at any concentration (C) of drug was given by the equation F = FmC/(EC50+ C), where Fm represents the maximal (or minimal) force and EC50 represents the concentration that produces half-maximal force for that drug. Nonlinear regression analysis was used to fit values of Fm and EC50 to data for F and C for each condition studied. This method allows for comparisons of curves to determine whether they are significantly different and whether this overall difference can be attributed to differences in Fm, EC50, or both. Paired comparisons were made with paired t-tests. Multiple comparisons were made with repeated-measures ANOVA, with Dunnett’s test for post hoc comparisons. In all cases,P < 0.05 was considered to indicate significance.
When canine tracheal smooth muscle (CTSM) strips were contracted with ACh or high-KCl solution in organ baths, H2O2added for 10 min induced comparable concentration-dependent relaxations (EC50 of 0.24 ± 0.04 and 0.23 ± 0.04 mM for strips contracted with ACh or high KCl, respectively; Fig. 1). Indomethacin attenuated relaxation induced by H2O2in strips contracted with either ACh or high KCl. Indomethacin significantly increased the EC50for H2O2from 0.24 ± 0.04 to 0.65 ± 0.18 mM in strips contracted with ACh. Although H2O2also relaxed strips contracted with high KCl in the presence of indomethacin (determined by repeated-measures ANOVA), the concentration at which this effect occurred could not be determined with post hoc analysis because of the variability in the data. In the strips contracted with 100 μM ACh before and after H2O2exposure, contractions after H2O2exposure were 100.2 ± 6.6% of the contractions obtained before exposure to H2O2. This indicates that H2O2did not affect the ability of the strips to produce maximal force (data not shown).
H2O2(1 mM) significantly increased [cAMP]i 2 and 10 min after addition to strips contracted with ACh. This increase was abolished by indomethacin. H2O2had no affect on [cGMP]i (Table1).
In fura 2-loaded strips, 0.3 μM ACh increased both force and F340/F380. The addition of 1 mM H2O2produced a sustained decrease in force, consistent with the results described above, and a sustained increase in F340/F380to a value greater than that before the addition of H2O2. Both the decrease in force and the sustained increase in F340/F380were significant, concentration dependent, and not affected by indomethacin (Fig. 2).
In fura 2-loaded strips treated with 10 μM indomethacin, 25 mM KCl increased both force and F340/F380. The addition of 300 μM H2O2produced a significant decrease in force. With continued exposure, force increased from a nadir that occurred 2–3 min after exposure but remained significantly less at 10 min than the initial force (before H2O2exposure). These changes in force were accompanied by a progressive increase in F340/F380(Fig. 3). This increase in F340/F380occurred in all strips but did not reach significance because of the variability in responses between dogs.
In strips incubated with 10 μM indomethacin, 0.3 μM ACh produced a sustained increase in rMLC phosphorylation. H2O2(300 μM and 1 mM) decreased rMLC phosphorylation measured 10 min after its addition to ACh-induced contractions (Fig.4). In indomethacin-treated strips, 25 mM KCl produced a sustained increase in rMLC phosphorylation. H2O2at 300 μM decreased the rMLC phosphorylation measured 5 min after its addition to KCl-induced contractions (Fig.5).
In this study, H2O2produced epithelium-independent relaxations of CTSM contracted by receptor stimulation with ACh or membrane depolarization with KCl. The H2O2-induced relaxations were 1) significantly inhibited by indomethacin; 2) accompanied by an increase in cAMP that was abolished by indomethacin (H2O2did not affect cGMP levels); 3) accompanied by an increase in [Ca2+]i, indicating that H2O2decreased Ca2+ sensitivity; and4) accompanied by a decrease in rMLC phosphorylation.
The reported effects of H2O2on ASM are diverse, apparently depending on the concentration of H2O2, the stimulus used to contract the tissue, the presence of the epithelium, and the species examined (2, 9, 10, 14, 22-24, 27,28). Gao and Vanhoutte (10) found that H2O2caused relaxation of CTSM contracted with ACh that was associated with the increased release of prostaglandin E2 and 6-ketoprostaglandin F1α. In another study (9), these authors reported that H2O2-induced relaxation of canine third-order bronchi contracted with ACh was accompanied by an increase in [cAMP]i, whereas [cGMP]i did not change. The presence or absence of the epithelium did not affect these responses. Indomethacin abolished prostaglandin release and the increase in [cAMP]iand inhibited relaxation; however, significant relaxation remained. Our results in CTSM are consistent with these findings because indomethacin abolished increases in [cAMP]i produced by H2O2. H2O2-induced relaxations of ACh-induced contractions that are insensitive to cyclooxygenase inhibitors have also been observed in ASM from the guinea pig (9), rabbit (14), and horse (22).
To explore the mechanism(s) producing this indomethacin-insensitive relaxation, we examined the effects of H2O2during ACh- and KCl-induced contractions. KCl contracts ASM primarily by depolarizing the cell membrane and increasing [Ca2+]ivia increased Ca2+ influx through voltage-dependent Ca2+ channels. Ca2+ influx also occurs during ACh-mediated contractions. However, muscarinic-receptor stimulation also increases the force developed for a given [Ca2+]i, i.e., increases Ca2+ sensitivity. The finding that H2O2produced significant relaxation during both ACh- and KCl-induced contractions suggests that this effect is not dependent on receptor-mediated processes, although additional effects on these processes are possible.
To evaluate the possibility that H2O2decreases Ca2+ sensitivity during ACh- and KCl-mediated contractions, we studied fura 2-loaded strips. The application of H2O2produced a sustained increase in F340/F380during both types of smooth muscle activation. The finding that H2O2increases [Ca2+]iis consistent with previous reports in several cell types (21, 26) including vascular smooth muscle (16, 25) and cat tracheal smooth muscle (2). In cat tracheal smooth muscle, this increase depended on both the release of Ca2+ from intracellular stores and Ca2+influx (2). Proposed mechanisms responsible for this increase in various cell types include lipid peroxidation (21), direct stimulation of sarcoplasmic reticulum Ca2+channels (in skeletal muscle) (3, 8), direct activation of intracellular kinases (4, 29), and direct activation of membrane Ca2+ channels by a process related to thiol oxidation (25). In some of our strips, a transient decrease in F340/F380preceded the sustained increase in F340/F380. We are not aware of previous reports of a transient decrease in [Ca2+]iwith the acute application of H2O2and can offer no explanation for this effect. It was present in the presence and absence of indomethacin so that it would not seem to be related to products of cyclooxygenase.
H2O2decreased force while increasing [Ca2+]i, a finding indicating a decrease in Ca2+ sensitivity. A similar finding has been reported in rabbit aorta strips without endothelium (16, 17). Unlike our findings, relaxation was observed during agonist-induced contractions but not during KCl-mediated contractions (17). The effects of H2O2on force and [Ca2+]iin ASM has not been previously reported. However, in the cat trachea, increases in [Ca2+]iproduced by H2O2are associated with a slow increase in baseline muscle tension and augmentation of contractions evoked by electrical field stimulation (2), effects that would not seem consistent with the decreases in Ca2+ sensitivity that we observed. We observed decreases in Ca2+sensitivity in both the presence and absence of indomethacin. This decrease was not caused by reduced viability of the contractile machinery because H2O2did not impair the ability of the muscle to produce maximal force after H2O2 washout.
We explored the mechanism responsible for this decrease in Ca2+ sensitivity by measuring rMLC phosphorylation. In smooth muscle, the binding of Ca2+ to calmodulin increases myosin light chain kinase activity and phosphorylation of the 20-kDa rMLC. rMLC phosphorylation allows the binding of myosin to actin, which increases actomyosin ATPase activity. ACh increases Ca2+ sensitivity in CTSM by increasing the level of rMLC phosphorylation maintained for a given [Ca2+]i, probably by inhibiting smooth muscle protein phosphatase activity by a G protein-linked mechanism (1, 5, 6). We found that H2O2decreased rMLC phosphorylation during contraction with both ACh and KCl. When combined with the finding that H2O2did not decrease [Ca2+]iunder either experimental condition, this demonstrates that H2O2decreased Ca2+ sensitivity, at least in part, by decreasing the level of rMLC phosphorylation maintained for a given [Ca2+]i. The fact that inhibition was observed in both the presence and absence of muscarinic-receptor stimulation implies that this effect is not specific to the G protein-linked systems that mediate agonist-induced increases in Ca2+ sensitivity. However, we cannot rule out such additional effects of H2O2during receptor activation. The fact that the inhibitory effect of H2O2on rMLC phosphorylation was present during KCl-induced contraction, which is produced solely by Ca2+activation, implies that H2O2inhibited myosin light chain kinase or activated smooth muscle protein phosphatases. Furthermore, it is possible that mechanisms other than rMLC phosphorylation may also regulate Ca2+ sensitivity (11, 12) and could be inhibited by H2O2. However, these effects were not examined in our experiments.
In summary, we confirm that there are two components of H2O2-induced relaxation of CTSM without epithelium: one component that is associated with an increase in [cAMP]i and is abolished by indomethacin and another that is indomethacin insensitive. The mechanism responsible for the latter component involves a decrease in Ca2+ sensitivity caused, at least in part, by an inhibitory effect of H2O2on rMLC phosphorylation.
We thank Dr. Susan J. Gunst (Department of Physiology and Biophysics, Indiana University School of Medicine, Indianapolis, IN) for generously providing us with polyclonal affinity-purified rabbit anti-regulatory myosin light chain antibody and Drs. Cheryl A. Conover and James B. Lawrence (Endocrine Research Unit, Mayo Clinic, Rochester, MN) for use of the PhosphorImager. Our special thanks to Kathy Street for expert technical assistance and Janet Beckman for her usual superb secretarial support.
Address for reprint requests and other correspondence: R. R. Lorenz, Mayo Clinic, 200 First St. SW, Rochester, MN 55905 (E-mail:).
This work was supported by National Heart, Lung, and Blood Institute Grants HL-45532 and HL-54757 and the Mayo Foundation.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 1999 the American Physiological Society