INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

ALTHOUGH HYPOXIC PULMONARY VASOCONSTRICTION has long been recognized as a unique contractile response of the pulmonary vasculature of humans and other species, the mechanism by which hypoxia stimulates smooth muscle contraction is poorly understood. One popular method by which hypoxia can be simulated experimentally is through the use of the oxygen scavenger sodium hydrosulfite (Na₂S₂O₄; also termed sodium dithionite) (3, 17, 18, 21, 27). However, sodium hydrosulfite is a potent reducing agent known to have several cellular actions (3), and the precise mechanism by which this compound evokes cellular responses is not known. We have used measurements of isometric tension and myosin light chain (MLC) phosphorylation to examine the mechanism of sodium hydrosulfite-induced tension development in smooth muscle.

The initiation of smooth muscle contraction is widely believed to require an increase in cytosolic calcium concentration, which activates calcium- and calmodulin-dependent MLC kinase, resulting in an increase in MLC phosphorylation. MLC phosphorylation, in turn, activates myosin, allowing actin-activated myosin ATPase activity and contraction (for review see Ref. 23). However, an increasing number of observations have clearly demonstrated that contractions can be initiated by both calcium- and MLC phosphorylation-independent pathways (for reviews see Refs. 9 and 28). The primary candidate for regulation of this alternate contractile pathway is the thin filament-based protein caldesmon (1). Caldesmon inhibits actin-activated myosin activity, and this inhibition can be reversed in vitro by either calcium and calmodulin or phosphorylation (22). The precise mechanism by which disinhibition of caldesmon occurs and the extent to which this pathway is physiologically important are currently unknown.

We report here that sodium hydrosulfite contracts smooth muscle in a calcium- and phosphorylation-independent manner. The induced contraction is not specific for pulmonary vascular smooth muscle and appears to bear little similarity to hypoxic vasoconstriction, which is calcium dependent (for review see Ref. 25). However, sodium hydrosulfite may be a useful agent in the investigation of the physiological relevance and mechanism underlying MLC phosphorylation-independent contractions.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Tension measurement. Adult Sprague-Dawley rats were euthanized by an intramuscular injection (0.1 ml/100 g body wt) of a mixture of xylazine (8.6 mg/ml) and ketamine (57.1 mg/ml). Smooth muscle segments (~2 × 4 mm) were prepared from the first branch of the right or left pulmonary artery, thoracic aorta, or trachea after connective tissues were dissected away under a microscope. Each segment was mounted in a cuvette of 200-µl volume, with one end fixed and the other attached to a force transducer (Scientific Instruments, Heidelberg, Germany). The output from the transducer was digitized at 2 Hz and recorded on computer disk (Axotape Software, Axon Instruments, Foster City, CA). HEPES-buffered Krebs-Henseleit (HKH) solution was continuously perfused at 37°C, and the tissue was equilibrated for ~40 min. The tissues were subjected to a passive tension of 0.2-0.3 mN, which had been previously determined to produce maximal active tension. During the equilibration period, tissues were stimulated with high KCl-HKH solution (60 mM KCl; equimolar substitution for NaCl) several times until a stable contractile response to KCl-HKH solution was achieved. Peak contractile responses were determined for each experimental condition and compared with a paired t-test. P < 0.05 was considered significant. All values are expressed as means ± SE.

MLC phosphorylation. MLC phosphorylation levels were determined in strips of rat pulmonary artery during the basal resting state and after stimulation with 10 mM sodium hydrosulfite and 110 mM KCl-HKH solution as previously described (16). Briefly, at appropriate times after stimulation, the tissues were rapidly frozen by immersion in a dry ice-acetone slurry containing 6% trichloroacetic acid. The frozen strips were slowly thawed, air-dried, and then homogenized in a solution containing 20 mM dithiothreitol, 10% glycerol, and 1% SDS. The ratio of homogenate solution to tissue was 20 mg wet weight tissue/ml solution. Homogenates were clarified by centrifugation and subjected to two-dimensional gel electrophoresis, and the electrophoresed proteins were transferred to 0.2-µm nitrocellulose membranes (1 A for 3 h at 4°C). The membranes were incubated in a phosphate-buffered saline solution (PBSS) containing 0.5% Tween 20 and 3% milk protein for 30 min and then overnight in PBSS containing 0.5% Tween 20, rabbit anti-chicken gizzard MLC antibody (1:250 dilution), and mouse anti-chicken gizzard tropomyosin antibody (1:800 dilution). The membranes were then washed extensively and incubated with the secondary antibodies (anti-rabbit immunoglobulin G and anti-mouse immunoglobulin G) conjugated to horseradish peroxidase. Antibody binding to MLC and tropomyosin was visualized by exposing nitrocellulose membrane to film to detect an enhanced chemiluminescent reaction (Amersham, Arlington Heights, IL). Quantitation of MLC phosphorylation levels was performed with a Molecular Dynamics personal laser densitometer. MLC phosphorylation levels were calculated by integration of the autoradiographic spot corresponding to the phosphorylated MLC as a percentage of the total of both the phosphorylated and unphosphorylated MLCs. The density of the satellite spots did not change during any experimental protocol and therefore was not used in the calculation (7, 14, 17). Results are expressed in moles of P_i per mole of MLC.

Drugs and chemicals. Sodium hydrosulfite, norepinephrine (NE), caffeine, carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone (FCCP), and mouse anti-tropomyosin antibody were obtained from Sigma (St. Louis, MO). Chelerythrine and cyclopiazonic acid (CPA) were obtained from Calbiochem (La Jolla, CA). All electrophoretic and immunoblot chemicals were obtained from Bio-Rad Laboratories (Richmond, CA). All other chemicals were analytic grade or better. Except for FCCP, all agents were dissolved to desired concentrations in either HKH solution or calcium-free HKH solution. The composition of HKH solution was (in mM) 126 NaCl, 10 HEPES (pH 7.4), 11 glucose, 6 KCl, 1 MgCl₂, and 2 CaCl₂. The calcium-free HKH solution was identical to regular HKH solution except that CaCl₂ was omitted and 2 mM EGTA was added. The pH of the sodium hydrosulfite solution was adjusted to 7.4 with NaOH. FCCP was first dissolved in DMSO and then diluted to 10 mM with HKH solution. The final concentration of DMSO in the FCCP solution was 0.05%, a concentration shown in preliminary studies not to influence smooth muscle contractions. All chemicals were applied by perfusion.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Sodium hydrosulfite contractions in rat pulmonary artery are calcium independent. The effects of sodium hydrosulfite on force development in rat pulmonary arterial segments were examined, and the results are shown in Fig. 1. Preliminary experiments indicated that when sodium hydrosulfite concentrations < 10 mM were used, contractile responses were variable. Only 56% of tissues examined (n = 9 segments) contracted in response to 5 mM sodium hydrosulfite. In contrast, 10 mM sodium hydrosulfite produced reversible and reproducible contractions in all vascular tissues. As shown in Fig. 1A, the contraction in response to 10 mM sodium hydrosulfite, although slower in rate than that in response to 60 mM KCl-PBSS, achieved significant levels of steady-state force. The maximal active force generated in response to 10 mM sodium hydrosulfite was 0.123 ± 0.013 mN (n = 9 segments), a value equivalent to 65.9 ± 12.8% of the contraction induced by 60 mM KCl-PBSS (Fig. 1B).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Sodium hydrosulfite (Na2S2O4)-induced contraction of rat pulmonary arterial smooth muscle. A: representative tracing showing contractile response of pulmonary arterial segments to KCl-HEPES-buffered Krebs-Henseleit (HKH) solution and Na2S2O4. Note reversibility of Na2S2O4 response. B: Na2S2O4-induced contraction is 65.9 ± 12.8% of maximal response to KCl-HKH solution. Values are means ± SE for 9 arterial segments.

To determine whether calcium influx through voltage-dependent calcium channels or calcium release from intracellular stores was involved in sodium hydrosulfite-induced contractions, we examined the effect of nisoldipine, caffeine, and CPA on these contractions. After a 30-min incubation with 5 µM nisoldipine, a solution of 10 mM sodium hydrosulfite produced a contraction comparable in magnitude to the response obtained in the absence of the inhibitor (n = 3 segments). Similarly, incubation with 10 mM caffeine (n = 5) or 10 µM CPA (n = 2) elicited a transient muscle contraction due to the release of intracellular calcium stores, but neither intervention prevented the subsequent contractile response to sodium hydrosulfite.

Because neither inhibition of extracellular calcium influx through dihydropyridine-sensitive channels nor depletion of intracellular stores of calcium significantly affected the sodium hydrosulfite-induced contraction, we examined the effects of complete calcium removal. Extracellular calcium was removed by simple deletion of CaCl₂ from the HKH solution. Intracellular calcium was depleted by prolonged incubation in CaCl₂-free HKH solution and sequential exposure to calcium-releasing stimuli. Figure 2 shows a representative tracing of nine similar experiments and demonstrates that in normal 1.8 mM CaCl₂-containing HKH solution and calcium-replete tissue, 60 mM KCl-HKH solution, 10 µM NE, and 10 mM sodium hydrosulfite all produce contractions. After removal of extracellular and intracellular calcium, neither KCl-HKH solution nor NE induced a contraction; however, the contraction in response to sodium hydrosulfite was unchanged.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Na2S2O4-induced contractions occur in the absence of extracellular Ca2+ and after sarcoplasmic reticulum Ca2+ depletion. Contractile responses of rat pulmonary arterial segments were obtained with 60 mM KCl-HKH solution, 10 µM norepinephrine (NE), and 10 mM Na2S2O4 in presence of normal 1.8 mM CaCl2-HKH solution. After 1-h incubation in Ca2+-free HKH solution, pulmonary arterial segments were reexposed to same concentrations of KCl-HKH solution, NE, and Na2S2O4. Neither KCl-HKH solution nor NE elicited contractions after Ca2+-free incubation, whereas contraction in response to Na2S2O4 was quite similar to that obtained in 1.8 mM CaCl2-HKH solution. Results are representative of 9 experiments.

A similar experimental strategy was pursued with the use of agents that release or inhibit uptake of intracellular calcium (Fig. 3). Arterial segments were initially treated with 10 mM caffeine, 10 µM FCCP (a protonophore), or 10 µM CPA, followed by the sequential addition of all three agents. The segments were exposed to HKH solution containing 10 mM sodium hydrosulfite. As shown in Fig. 3, all three compounds (caffeine, FCCP, and CPA) produced a contraction when added to calcium-replete tissue. The complete absence of any increase in force development on the second addition of these agents clearly demonstrated complete calcium depletion of the arterial segments. Even after these harsh calcium-depletion protocols, the sodium hydrosulfite-induced contraction remained maximal and reversible. The active force generated in response to 10 mM sodium hydrosulfite in normal and calcium-free HKH solution was 0.123 ± 0.013 and 0.127 ± 0.015 mN, respectively (n = 9 segments).

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Na2S2O4-induced contractions of rat pulmonary arterial segments do not require release of Ca2+ from ryanodine-sensitive Ca2+ stores or release of mitochondrial Ca2+. A: in Ca2+-containing solutions, release of mitochondrial Ca2+ by carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone (FCCP; 10 µM) and inhibition of sarcoplasmic reticulum Ca2+ uptake by cyclopiazonic acid (CPA; 10 µM) induces transient contractions. Separate experiments are shown. B: release of Ca2+ by activation of ryanodine receptors with caffeine (Caff; 10 mM) produces transient contraction in Ca2+-containing HKH solution. However, after 1-h incubation in Ca2+-free HKH solution, Caff, FCCP, and CPA produced no contraction, whereas Na2S2O4 evoked a large contraction in same muscle segment. Results are representative of 5-9 experiments for each condition.

Sodium hydrosulfite-induced contractions are not unique to pulmonary arterial smooth muscle and do not involve hypoxia or protein kinase C. To determine whether sodium hydrosulfite-induced contractions were unique to pulmonary arterial segments, we performed experiments similar to those shown in Fig. 2 in rat aortic and tracheal smooth muscle strips. The aortic and tracheal tissues were incubated in a calcium-free HKH solution for 60 min, during which time they were repeatedly stimulated with either 10 µM NE (aortic strips) or 50 µM methacholine (tracheal strips) to deplete intracellular stores of calcium. Similar to the results obtained with pulmonary arterial segments, the addition of 10 mM sodium hydrosulfite induced significant levels of force that were unaffected by calcium depletion (Table 1).

Because sodium hydrosulfite is an oxygen scavenger, we sought to determine whether hypoxia was a necessary condition for the observed calcium-independent contractions described above. As shown in Fig. 4, addition of 10 mM sodium hydrosulfite to the perfusion chamber contracted rat pulmonary arterial segments and lowered the PO₂ of HKH solution from ~140 to <2 mmHg. However, 10 mM hydrosulfite contracted the pulmonary arterial segments (n = 4) to a similar degree, even though the solution was continually aerated with 95% O₂-5% CO₂ for 30 min so that the PO₂ levels increased to between 200 and 414 mmHg.

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Na2S2O4 contractions are independent of PO2. Contractile response of rat pulmonary arterial segments to 10 mM Na2S2O4 was examined under conditions of low (left) and high (right) PO2. Elevated PO2 (200-414 mmHg) conditions were produced by aerating HKH solution with 95% O2-5% CO2 for ~30 min. Na2S2O4-induced contractions were similar at both ranges of PO2. Tracing is representative of 3 experiments.

To determine whether the contraction to sodium hydrosulfite results from the formation of H₂O₂, we tested the effects of H₂O₂ on rat pulmonary arterial segments. Consistent with a previous study (19), the pulmonary segments contracted in response to 1 and 10 mM H₂O₂ in normal, calcium-containing HKH solution (Fig. 5). However, neither concentration of H₂O₂ contracted segments (n = 3) in calcium-free HKH solution, although contractions to sodium hydrosulfite were not affected. These data suggest that generation of H₂O₂ is not the mechanism by which sodium hydrosulfite initiates force development.

View larger version (28K): [in this window] [in a new window]   Fig. 5.   Na2S2O4 contractions do not occur by formation of H2O2. A: H2O2 contracted pulmonary arterial segments slightly at 1 mM and more forcefully at 10 mM in Ca2+-containing solution. B: in Ca2+-free HKH solution, however, 1 and 10 mM H2O2 failed to produce contractions. In contrast, 10 mM Na2S2O4 contracted segments in Ca2+-free solution (see Fig. 2 and Table 1).

In addition to the widely studied and accepted role of MLC kinase, protein kinase C-catalyzed protein phosphorylation has been implicated in contractile regulation and especially in calcium-independent pathways (2, 6). To determine whether protein kinase C is important in the sodium hydrosulfite-induced contraction, we utilized the protein kinase C inhibitor chelerythrine (7). We first determined that rat pulmonary arterial segments contracted in response to the addition of an activator of protein kinase C, phorbol 12,13-dibutyrate, and that chelerythrine was capable of inhibiting this contraction. As shown in Fig. 6A, exposure of muscle segments to 0.1 µM phorbol 12,13-dibutyrate produced a large, sustained contraction and this contraction was completely abolished by pretreatment with 10 µM chelerythrine. For examination of the role of protein kinase C in sodium hydrosulfite contractions, calcium-depleted pulmonary arterial segments were contracted by the addition of 10 mM sodium hydrosulfite, relaxed, and then incubated in calcium-free HKH solution containing 10 µM chelerythrine. The calcium-depleted tissues were challenged a second time with 10 mM sodium hydrosulfite, this time in the presence of chelerythrine. As shown in the tracing in Fig. 6B (representative of 4 experiments), inhibition of protein kinase C by chelerythrine had no effect on the sodium hydrosulfite-induced contraction.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Inhibition of protein kinase C does not alter Na2S2O4-induced contraction. A: phorbol 12,13-dibutyrate (PDBu) produced a large, sustained contraction of rat pulmonary arterial segments, which was blocked by preincubation with chelerythrine (10 µM; right). B: conversely, contraction of rat pulmonary arterial segments to 10 mM Na2S2O4 was not affected by incubation with 10 µM chelerythrine.

Sodium hydrosulfite-induced contractions are independent of MLC phosphorylation. The potential mechanism of sodium hydrosulfite-induced contractions was investigated by quantifying MLC phosphorylation levels. MLC phosphorylation levels were measured at rest, at various time points during a 10 mM sodium hydrosulfite-induced contraction, and at the peak of a KCl-HKH solution-induced contraction, with both contractions in the presence of calcium. The results of these experiments are shown in Fig. 7. MLC phosphorylation levels did not increase above the basal level at any time point measured during the sodium hydrosulfite-induced contraction. In contrast, after 3 min of membrane depolarization by KCl-HKH solution, MLC phosphorylation levels increased to >0.5 mol P_i/mol MLC. Therefore, these data clearly demonstrate that the contraction of pulmonary arterial segments in response to sodium hydrosulfite is independent of an increase in MLC phosphorylation and must be initiated by an alternate pathway.

View larger version (10K): [in this window] [in a new window]   Fig. 7.   Na2S2O4 contractions occur in absence of myosin light chain (MLC) phosphorylation. MLC phosphorylation levels were measured in rat pulmonary arterial segments during contractions to 10 mM Na2S2O4 (normal HKH solution). Although Na2S2O4 produced a significant increase in force, MLC phosphorylation levels were not elevated above basal value at any time point measured. For comparison, MLC phosphorylation after 3-min stimulation with 60 mM KCl is shown. KCl depolarization produced significant increase in MLC phosphorylation levels. Values are means ± SE; n = 3 segments. * P < 0.001 compared with basal value.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

We report in this study that the oxygen scavenger and reducing agent sodium hydrosulfite contracts arterial and tracheal smooth muscle preparations by a mechanism that is not dependent on an increase in either calcium or MLC phosphorylation. Contractions of the smooth muscles in response to sodium hydrosulfite were consistently obtained during conditions in which the muscle segments were repeatedly exposed to calcium-mobilizing agents until no contraction in response to standard receptor- or membrane-mediated pathways could be elicited. Reversible, maximal contractions of the pulmonary artery were obtained in response to sodium hydrosulfite after calcium depletion, whereas contractions in response to the addition of NE, caffeine, FCCP, or CPA were abolished. Moreover, contractions in response to sodium hydrosulfite were not associated with an increase in MLC phosphorylation levels, whereas membrane depolarization resulted in a concomitant increase in force and MLC phosphorylation. The fact that these calcium- and MLC phosphorylation-independent contractions are maximal and reproducible suggests that sodium hydrosulfite may be activating a physiologically relevant regulatory pathway rather than simply inducing an artifactual conformational change in one or more contractile proteins.

The two most likely candidates for coupling sodium hydrosulfite stimulation with a calcium- and MLC phosphorylation-independent contractile event are protein kinase C and hypoxia. The addition of sodium hydrosulfite to physiological salt solutions has been shown to result in the generation of superoxide radicals (3). Such radicals have been shown to activate protein kinase C (4), which has been implicated in the regulation of smooth muscle contractions (2, 6, 8), including calcium-independent contractions (6, 11, 12). Moreover, direct measurements of protein kinase C isoforms in smooth muscle have provided evidence of a role for calcium-independent isoforms in contractile regulation (8). We therefore examined the hypothesis that sodium hydrosulfite contractions resulted from the activation of protein kinase C. However, the protein kinase C inhibitor chelerythrine had no effect on sodium hydrosulfite-induced contraction but completely inhibited phorbol ester-induced contractions (Fig. 6).

Hypoxia produced by the oxygen-scavenging properties of sodium hydrosulfite also does not appear to be the mechanism underlying the calcium-independent contraction because hyperoxygenating the sodium hydrosulfite-containing solution did not abolish the contractions. Moreover, hypoxic pulmonary vasoconstriction is associated with an increase in cellular calcium (20, 24, 26) and myosin phosphorylation (14, 29), and hypoxia relaxes systemic vessels such as the aorta, whereas sodium hydrosulfite-induced, calcium-independent contractions were obtained in both pulmonary artery and aorta. The conclusion that hypoxia is not important in the sodium hydrosulfite-induced contraction raises concerns with respect to the use of this compound as a convenient agent to generate hypoxic solutions, as previously discussed by Archer and co-workers (3). The wide variety of chemicals derived from sodium hydrosulfite, such as superoxide radicals, may confuse or even mislead the interpretation of the results of experiments based on this chemical. The generation of H₂O₂ by sodium hydrosulfite could not explain the observed calcium-independent contractions, however, because such contractions were blocked by removal of extracellular calcium (Fig. 5).

Our determination that sodium hydrosulfite contractions are calcium independent relies on the adequacy of the calcium-depletion protocol as assessed by abolition of receptor-mediated contraction. This appears to be a reasonable assumption because the depletion protocol effectively eliminated contractions to NE, caffeine, FCCP, and CPA but had no effect on sodium hydrosulfite-induced contractions. Moreover, the absence of increased MLC phosphorylation levels further suggests that these contractions occur by a fundamentally different mechanism. MLC phosphorylation values were found to be within the linear range for densitometric analysis of an autoradiographic spot with enhanced chemiluminescence, and immunoblot analysis verified the location of the MLC on the nitrocellulose membrane.

On the basis of the results presented in this study, sodium hydrosulfite induces smooth muscle contraction by a novel mechanism that is independent of MLC phosphorylation. Although we have no evidence to indicate a specific signaling mechanism, we speculate that exposure to sodium hydrosulfite either induces a conformational change in MLC that mimics the effect of phosphorylation or results in a disinhibition of regulatory thin filaments such as caldesmon. Such an effect could occur because of a direct action of sodium hydrosulfite, an action of a metabolite such as bisulfite or bisulfate, or the generation of a reactive oxygen species other than H₂O₂. With respect to the former mechanism, Ikebe et al. (10) suggested that specific changes in ionic constituents, such as a high Mg²⁺ concentration, may contract smooth muscle by mimicking the phosphorylation-induced conformational change in MLC. However, such contractions are typically submaximal and are characterized by a substantially slower rate of force development (15). With respect to the latter mechanism, Katsuyama et al. (13) have shown that a peptide that competes for the caldesmon-binding site on actin but is not inhibitory to myosin ATPase activity induces a calcium-independent contraction in permeabilized smooth muscle. Disinhibition of caldesmon may allow expression of an inherent level of activated myosin that is normally under tonic inhibition by the caldesmon protein (6, 28). Such disinhibition could occur by a conformational change in caldesmon produced by sodium hydrosulfite or its metabolites. Whether the sodium hydrosulfite-induced contraction is initiated by this mechanism is unknown; however, we are not aware of any other reported calcium-independent contractile mechanism that could explain these findings.

In summary, we have demonstrated that stimulation of several smooth muscles with sodium hydrosulfite produces a reproducible, maximal contraction that is independent of calcium and MLC phosphorylation levels. The contraction is not dependent on protein kinase C and does not result from hypoxic conditions. Although no information is presently available to suggest a mechanism for the contraction, our results do provide a simple and inexpensive means of initiating calcium- and MLC phosphorylation-independent contractions in smooth muscle, which may facilitate mechanistic studies.