INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

IT IS WELL ESTABLISHED that an increase in the intracellular free calcium concentration ([Ca²⁺]) initiates contraction of smooth muscle. One of the primary mechanisms by which this increase in Ca²⁺ initiates contraction involves the calmodulin-dependent, myosin light chain (MLC) kinase-catalyzed phosphorylation of the 20,000 relative molecular weight MLC (14). Secondary roles for Ca²⁺ have also been proposed in the regulation of cross-bridge behavior. Siegman et al. (36) demonstrated a Ca²⁺-dependent resistance to stretch in quiescent intestinal smooth muscle suggestive of attached noncycling cross bridges; these attached cross bridges were not dependent on MLC phosphorylation (35). Wagner and Rüegg (38) presented evidence suggesting a Ca²⁺- and calmodulin-dependent contraction of skinned smooth muscle fibers in the absence of an increase in MLC phosphorylation. Murphy and co-workers (1, 4, 6, 12) have presented a large amount of information concerning the relationship between MLC phosphorylation and force. These investigators coined the term "latch state," which was originally hypothesized to account for the maintenance of stress supported by slowly cycling, dephosphorylated cross bridges and was suggested to be regulated by an unidentified Ca²⁺-dependent mechanism with a higher sensitivity to Ca²⁺ than the MLC kinase. This was later revised to suggest that a single Ca²⁺-dependent step, that being MLC phosphorylation, was sufficient to account for all aspects of contraction (12), a revision not universally accepted (8, 11, 26, 38, 40, 41).

Although Ca²⁺ is the major cation involved in the regulation of smooth muscle, a second divalent cation, Mg²⁺, has been shown to significantly influence the activity of smooth muscle contractile proteins. Variations in the concentration of free Mg²⁺ have been demonstrated to modulate Ca²⁺-dependent actomyosin ATPase activity and the actomyosin superprecipitation response (16, 22, 29) and Ca²⁺-dependent actin-activated ATPase activity of phosphorylated smooth muscle myosin (15). In addition to modulation at the level of the isolated contractile proteins, alterations in [Mg²⁺] have been shown to affect the contractile behavior of permeabilized (skinned) fibers. Changes in the concentration of free Mg²⁺ affect the Ca²⁺ dependence of stress development (34) and unloaded shortening velocity but not MLC phosphorylation (3) in skinned smooth muscle fibers. Paul and Rüegg (33) suggested that the apparent requirement for relatively high [Mg²⁺] previously demonstrated in smooth muscle preparations (9, 22) is due to a Mg²⁺ dependence of the MLC kinase and not of actin-myosin interactions. However, these investigators (33) also suggested that any Ca²⁺-dependent regulatory system present in smooth muscle that does not involve MLC phosphorylation would require Mg²⁺ for expression. Ikebe et al. (15) and our laboratory (24) have demonstrated that in addition to modulation of a skinned fiber contraction, high [Mg²⁺] can elicit a contraction supported by active cycling cross bridges in the absence of both Ca²⁺ and MLC phosphorylation. We have suggested that this contraction in response to high [Mg²⁺] is due to the direct activation of latch bridges (24).

Another potentially important role for Mg²⁺ in the smooth muscle cell is the modulation of [MgADP]. Kerrick and Hoar (17) and, more recently, Somlyo and colleagues (10, 19, 20) have presented evidence to suggest that the high levels of sustained stress without proportional levels of Ca²⁺ or MLC phosphorylation in tonic smooth muscle cells are due to MgADP. More specifically, Khromov et al. (19, 20) have hypothesized that the affinity of the cross bridge for MgADP is a major determinant of the slower kinetics of tonic compared with phasic smooth muscle.

Therefore, Mg²⁺ modulates the Ca²⁺-dependent contraction of smooth muscle, but the exact nature of this effect and whether Mg²⁺ acts at multiple sites is unknown. The present study was designed to determine the influence of free [Mg²⁺] on two phases of a smooth muscle contraction, development and maintenance of stress, that we have postulated are controlled by two separate Ca²⁺-dependent regulatory pathways. Specifically, we determined the effects of varying [Mg²⁺] from 0.1 to 6 mM on Ca²⁺-dependent MLC phosphorylation and concomitant stress development and stress maintenance in the absence of proportional levels of MLC phosphorylation. We also determined the effects of varying [Mg²⁺] on actin-activated myosin ATPase activity during the development and maintenance of stress. All studies were performed in the Triton X-100 detergent-skinned swine carotid medial fiber. These studies were undertaken to determine whether alterations in the concentration of free Mg²⁺ would differentially affect the MLC phosphorylation-dependent development of stress and the apparently MLC phosphorylation-independent maintenance of stress (41).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Tissue preparation. Swine carotid arteries were obtained at slaughter and transported to the laboratory in ice-cold physiological salt solution (PSS). Circumferential strips of the swine carotid media were dissected [100 µm × 2 mm × 8 mm at the optimal length for active force development (l_o)] and mounted between a micrometer for control of muscle length and a Grass FT.03 force transducer and Grass model 7 polygraph (Quincy, MA) for isometric force recording. The mounted strips were placed in water-jacketed muscle chambers at 37°C and aerated with 100% O₂ in PSS containing (in mM) 140 NaCl, 4.7 KCl, 1.2 NaH₂PO₄, 2.0 MOPS (pH 7.4), 5.6 D-glucose, and 0.02 Na₂-EDTA. The strips were stretched to ~100 mN force, allowed to stress-relax, and equilibrated for at least 90 min. After equilibration, a partial length-tension curve was constructed to determine l_o, using 110 mM KCl-PSS (equimolar substitution for NaCl) as the stimulus.

Permeabilization technique. The medial strips were made hyperpermeable by detergent skinning with Triton X-100 as described in detail previously (26). Briefly, the strips were depleted of intracellular and extracellular calcium by stimulation with 30 µM histamine in a Ca-free PSS containing 2 mM EGTA for 45 min before exposure to a skinning solution containing 0.5% Triton X-100, 5 mM EGTA, 20 mM imidazole (pH 7.0 at 22°C), 50 mM K-acetate, 1 mM dithiothreitol (DTT), and 150 mM sucrose for 60 min at 22°C. Exposure to the skinning solution was followed by two relaxing solutions containing MgCl₂, Na₂ATP, K-acetate, imidazole, DTT, and either 5 mM EGTA (first solution) or 0.1 mM EGTA (second solution).

The compositions of the contracting and relaxing solutions were calculated by a computer program that solves the appropriate multi-equilibrium association equations and has been described in detail previously (26). All solutions contained 20 mM imidazole (pH 7.0 at 22°C), 5 mM MgATP, 1 mM DTT, 0.1-6 mM Mg²⁺, and 10 nM-10 µM Ca²⁺. All experiments were performed at 22°C and pH 7.0. Solutions were adjusted to an ionic strength of 0.12 M with K-acetate and a pH of 7.0 with KOH. MgCl₂ was obtained from Sigma (St. Louis, MO) as a 4.92 M standard solution, and CaCl₂ was obtained from Orion Research (Boston, MA) as a 0.1 M atomic absorption standard solution.

Contraction protocols. The effect of Mg²⁺ on Ca²⁺-dependent stress development was studied by exposing the skinned fibers to increasing concentrations of Ca²⁺ in the presence of 0.1, 1.0, 3.0, or 6.0 mM Mg²⁺. The effect of Mg²⁺ on Ca²⁺-dependent stress maintenance was determined by contracting the fibers in 7 µM Ca²⁺ in the presence of 1.0, 3.0, or 6.0 mM Mg²⁺, or 3 µM Ca²⁺ in the presence of 0.1 mM Mg²⁺, and then reducing the Ca²⁺ concentration while maintaining the Mg²⁺ concentration constant. This protocol has been shown to demonstrate Ca²⁺-dependent stress maintenance without proportional levels of MLC phosphorylation, which is believed to be indicative of the latch state in skinned fibers (4, 25, 26).

MLC phosphorylation. MLC phosphorylation levels were measured in all tissues. The tissues were frozen in a dry ice-acetone slurry containing 6% trichloroacetic acid, slowly thawed, homogenized, and then subjected to two-dimensional gel electrophoresis as previously described (24, 27). Quantitation of the MLC phosphorylation levels was performed by scanning densitometry of the Coomassie blue-stained gels using a LKB laser densitometer equipped with a recording integrator.

ATPase activity. Actin-activated myosin ATPase activity was measured in the Triton X-100 detergent-skinned fibers using a Güth Muscle Research Station (Heidelberg, Germany) as previously described (40). The tissues were mounted between a force transducer and a motor in a 100-µl cuvette. The cuvette is equipped with a perfusion system that produces a complete solution change in <1 s. ATPase activity was monitored by the loss of fluorescence of NADH by an enzyme system coupled to ATP hydrolysis (ATP  ADP + P_i by myosin ATPase; ADP + phosphoenolpyruvate  ATP + pyruvate by pyruvate kinase; pyruvate + NADH  NAD⁺ + lactate by lactate dehydrogenase). The solution was excited at 340 nm, and the resultant fluorescence was detected at 470 nm by a Zeiss photometer objective and a Hamamatsu photomultiplier. The decay in fluorescence was measured over a 20-s time period, after which the cuvette solution was completely changed. The fluorescence data were analyzed by a microcomputer interfaced to the equipment. A linear regression of the data collected between 5 and 15 s after perfusion was used to calculate ATPase activity. A recorder hard copy of the change in fluorescence was also obtained to ensure linear responses and a return to similar peak fluorescence levels after each perfusion. ATPase activity was quantified as percent increase in stimulated ATPase activity using a 10-min perfusion with zero Ca²⁺ as baseline. We have previously published extensive controls demonstrating that this technique measures myosin ATPase activity and not other cellular ATPases in the Triton X-100-skinned carotid artery (40).

Calculations and statistics. Stress was calculated as previously described from strip length at l_o and tissue weight (25). The apparent K_m, defined as the Ca²⁺ concentration necessary for half-maximal response, was calculated with the use of a nonlinear curve-fitting program. Significance of differences between means was performed by Student's t-test for unpaired data. Significance of differences between apparent K_m values was performed by analysis of variance. P < 0.05 was taken as significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The experiments performed in this study were based on a protocol that mimics the normal stimulation-induced transient in intracellular [Ca²⁺] that occurs in the intact swine carotid media. Stylized results demonstrating both the protocol used and our definition of stress development, stress maintenance, and maintained stress as they relate to this present study are shown in Fig. 1. A stepwise increase in the [Ca²⁺] from nominally zero to 7 µM is used as an index of stress development. A stepwise decrease in the [Ca²⁺] from any suprabasal value is used as an index of stress maintenance. The difference in the magnitude of stress maintained and stress developed at any given [Ca²⁺] is defined as maintained stress and is used as an index of the latch state. We have previously used this protocol to propose that the development of stress in response to an increase in Ca²⁺ is the MLC phosphorylation-dependent component of a contraction, whereas the maintenance of stress that occurs without proportional levels of MLC phosphorylation after a decrease in Ca²⁺ is the MLC phosphorylation-independent component (24-27, 40, 41).

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Stylized results demonstrating protocols used and definition of stress development, stress maintenance, and maintained stress as they pertain to this study. Increase in stress after an increase in Ca2+ concentration ([Ca2+]) from nominally zero to maximal is used as an index of stress development. Magnitude of stress after a decrease in [Ca2+] from any suprabasal value is used an index of stress maintenance. Difference in stress between maintenance and development protocols is defined as maintained stress and is used as an index of latch state.

The data in Fig. 2A depict the results obtained using a constant [Mg²⁺] of 0.1 mM in the Triton X-100-skinned fiber. Strips were exposed to increasing [Ca²⁺] (stress development) or were first maximally contracted with 3 µM Ca²⁺ and then exposed to lower [Ca²⁺]. The data in Fig. 2B show the corresponding MLC phosphorylation levels during steady-state levels of Ca²⁺-dependent stress development and stress maintenance of the same fibers as shown in Fig. 2A. The levels of developed stress and MLC phosphorylation as well as the apparent K_m values for stress development, stress maintenance, and MLC phosphorylation during both the development and maintenance of stress are shown in Table 1. In the presence of 0.1 mM Mg²⁺, stress hysteresis is slight but significant, as indicated by the decrease in apparent K_m for stress maintenance compared with stress development. This stress hysteresis is suggestive of the presence of the latch state, i.e., force without proportional levels of MLC phosphorylation (1, 4, 6, 25, 26). One important feature is the almost complete relaxation of stress as [Ca²⁺] is decreased to 0.3 µM and lower.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Effect of 0.1 mM Mg2+ on Ca2+-dependent contractions of detergent-skinned fibers. A: detergent-skinned fibers of swine carotid media were exposed to increasing [Ca2+] (, stress development) or first contracted with 3 µM Ca2+ and then exposed to lower [Ca2+] (, stress maintenance) in presence of 0.1 mM Mg2+. Data were normalized as percentage of level of stress developed or as percentage of initial level of stress in response to 3 µM Ca2+ before reduction in [Ca2+]. Values shown are means ± SE for at least 6 determinations. B: skinned fibers used in generation of A were frozen for quantitation of myosin light chain (MLC) phosphorylation levels. MLC phosphorylation levels (shown as mol Pi/mol MLC) correspond to appropriate Ca2+-dependent stress development () and stress maintenance () curves in A. Values shown are means ± SE for at least 5 determinations.

The effects of holding [Mg²⁺] at 1.0 mM, rather than 0.1 mM, on the Ca²⁺-dependent properties of contraction are shown in Fig. 3 and Table 1. The [Ca²⁺] that produced maximal stress in the presence of 1-6 mM Mg²⁺ was 7 µM. Increasing [Mg²⁺] to 1.0 mM had no significant effect on the levels of either developed stress or MLC phosphorylation compared with values obtained in the presence of 0.1 mM Mg²⁺. However, 1.0 mM Mg²⁺ significantly decreased the Ca²⁺ sensitivity of stress development as well as the Ca²⁺ sensitivity of MLC phosphorylation during both stress development and maintenance relative to the response in the presence of 0.1 mM Mg²⁺. In contrast to the data shown in Fig. 2, the results shown in Fig. 3 demonstrate a high degree of Ca²⁺-dependent stress maintenance without proportional levels of MLC phosphorylation, as previously reported using a similar [Mg²⁺] of 0.8 mM (4, 26). The results shown in Table 1 also demonstrate that the Ca²⁺ sensitivity for stress maintenance is significantly greater than that for stress development and MLC phosphorylation. The magnitude of the stress maintained in the absence of Ca²⁺ was increased to ~20% by the increase to 1.0 mM Mg²⁺.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Effect of 1 mM Mg2+ on Ca2+-dependent contractions of detergent-skinned fibers. A: detergent-skinned fibers of swine carotid media were exposed to increasing [Ca2+] (, stress development) or first contracted with 7 µM Ca2+ and then exposed to lower [Ca2+] (, stress maintenance) in presence of 1 mM Mg2+. Data were normalized as described for Fig. 2. Values shown are means ± SE for at least 6 determinations. B: MLC phosphorylation levels correspond to appropriate Ca2+-dependent stress development () and stress maintenance () curves in A. Values shown are means ± SE for at least 6 determinations.

The results obtained on increasing the level of free Mg²⁺ to 3.0 mM are shown in Fig. 4 and Table 1. Mg²⁺ at a concentration of 3 mM did not significantly alter the levels of developed stress or MLC phosphorylation relative to 1 mM Mg²⁺. Developed stress and MLC phosphorylation levels in the presence of 3 mM Mg²⁺ were significantly decreased compared with those obtained in the presence of 0.1 mM Mg²⁺. The Ca²⁺ sensitivity of stress development was not affected by the increase in the free [Mg²⁺], compared with that calculated in the presence of 1 mM Mg²⁺. However, the Ca²⁺ sensitivity of MLC phosphorylation in the presence of 3 mM Mg²⁺ was significantly decreased during both increasing and decreasing [Ca²⁺], relative to that obtained in the presence of 1 mM Mg²⁺. Therefore, the Ca²⁺ dependence of MLC phosphorylation was depressed even though the sensitivity of stress was not changed. The apparent K_m for stress maintenance was not determined because the significant elevation in the stress maintained in the absence of Ca²⁺ precluded a reliable calculation. The magnitude of the stress maintained after removal of Ca²⁺ in the presence of 3 mM Mg²⁺ was significantly greater than that observed in the presence of 1 mM Mg²⁺.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Effect of 3 mM Mg2+ on Ca2+-dependent contractions of detergent-skinned fibers. A: detergent-skinned fibers of swine carotid media were exposed to increasing [Ca2+] (, stress development) or first contracted with 7 µM Ca2+ and then exposed to lower [Ca2+] (, stress maintenance) in presence of 3 mM Mg2+. Data were normalized as described in Fig. 2. Values shown are means ± SE for at least 6 determinations. B: MLC phosphorylation levels correspond to appropriate Ca2+-dependent stress development () and stress maintenance () curves in A. Values shown are means ± SE for at least 5 determinations.

The results obtained with the highest concentration of free Mg²⁺ used in this study, 6 mM, are shown in Fig. 5 and Table 1. In contrast to the effects of the lower [Mg²⁺], 6 mM Mg²⁺ significantly depressed the levels of Ca²⁺-dependent developed stress and MLC phosphorylation. Similar to the effects of increasing [Mg²⁺] from 1 to 3 mM, increasing [Mg²⁺] to 6 mM significantly decreased the Ca²⁺ sensitivity of MLC phosphorylation without affecting the Ca²⁺ sensitivity of stress development relative to those values calculated in the presence of lower [Mg²⁺]. The Ca²⁺ sensitivity of stress maintenance was not determined for reasons discussed above. The stress maintained in the absence of Ca²⁺ increased to ~70% of the maximal Ca²⁺-dependent stress in the presence of 6 mM Mg²⁺.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Effect of 6 mM Mg2+ on Ca2+-dependent contractions of detergent-skinned fibers. A: detergent-skinned fibers of swine carotid media were exposed to increasing [Ca2+] (, stress development) or first contracted with 7 µM Ca2+ and then exposed to lower [Ca2+] (, stress maintenance) in presence of 6 mM Mg2+. Data were normalized as described in Fig. 2. Values shown are means ± SE for at least 6 determinations. B: MLC phosphorylation levels correspond to appropriate Ca2+-dependent stress development () and stress maintenance () curves in A. Values shown are means ± SE for at least 6 determinations.

The magnitude of the stress maintained after removal of Ca²⁺, as shown in Figs. 2-5, appears to be Mg²⁺ dependent. To determine whether this is a correct assumption, skinned fibers were contracted with 7 µM Ca²⁺ in the presence of 6 mM Mg²⁺, then exposed to a Ca-free solution containing 5 mM EGTA. The [Mg²⁺] was then lowered from 6 to 3 to 1 to 0.1 mM. The results of these experiments are shown in Fig. 6. On the initial addition of 5 mM EGTA to a contracted fiber (7 µM Ca²⁺ and 6 mM Mg²⁺), the skinned fibers relaxed ~30%. This level of stress is shown in Fig. 6 along with the corresponding level of MLC phosphorylation. MLC phosphorylation levels were basal after the addition of the 5 mM EGTA-containing solution and were not affected by the subsequent changes in free [Mg²⁺]. As [Mg²⁺] was decreased, the steady-state level of maintained stress decreased in a concentration-dependent fashion until, in the presence of 0.1 mM Mg²⁺, the skinned fibers relaxed almost completely to baseline values. Reintroduction of Mg²⁺ up to 6 mM did not produce any increase in stress (data not shown). Therefore, once the medial fibers were completely relaxed, increases in both Ca²⁺ and Mg²⁺ are required to develop stress.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Mg2+-dependent relaxation of detergent-skinned fibers. Skinned carotid medial fibers were first contracted with 7 µM Ca2+ in presence of 6 mM Mg2+ and then exposed to a series of solutions containing no added CaCl2, 5 mM EGTA, and decreasing [Mg2+]. Level of maintained stress after each change in [Mg2+] () and corresponding MLC phosphorylation level () are shown. Values shown are means ± SE for at least 4 determinations.

The last set of experiments performed was to determine the effect of altering free [Mg²⁺] on actin-activated myosin ATPase activity during a Ca²⁺-dependent contraction. Measurement of actin-activated myosin ATPase activity would allow us to determine whether Mg²⁺ effected actin and myosin interactions directly compared with an indirect action through effects on MLC phosphorylation. The skinned strips were maximally contracted by the addition of 7 µM Ca²⁺ in the presence of either 1 or 6 mM Mg²⁺. After a stable force and ATPase recording was attained, the [Mg²⁺] was either increased from 1 to 3 to 6 mM or decreased from 6 to 3 to 1 mM. Steady-state force and actin-activated myosin ATPase activities were measured after each change in [Mg²⁺]. The results, presented as a percentage of the maximal response, are shown in Table 2. Increasing [Mg²⁺] significantly reduced both stress and ATPase activity, and decreasing [Mg²⁺] significantly increased both stress and ATPase activity. Similar changes in [Mg²⁺] had no effect on MLC phosphorylation levels (Table 1). These results suggest that the effect of Mg²⁺ on stress is due to direct effects on the actin-activated myosin ATPase and not indirect effects via changes in MLC phosphorylation.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Studies using intact smooth muscle have demonstrated that contractile activation increases cytosolic [Ca²⁺], MLC phosphorylation levels, and stress. With continued stimulation, stress is maintained while the levels of cytosolic Ca²⁺ and MLC phosphorylation fall to suprabasal values (1, 6, 13, 28). We have used the Triton X-100 detergent-skinned swine carotid medial tissue and a protocol originally designed by Chatterjee and Murphy (4) to intentionally mimic this transient increase in cytosolic [Ca²⁺] (25, 26). As [Ca²⁺] is raised, stress and MLC phosphorylation increase concomitantly. In contrast, if the tissues are first contracted with a high [Ca²⁺] and then exposed to a lower [Ca²⁺], MLC phosphorylation levels fall in proportion to [Ca²⁺], but stress is maintained at higher than expected levels. This results in a leftward shift in the apparent K_m for the Ca²⁺ dependence of stress from 1.6 to 0.3 µM. We have proposed that this "stress hysteresis" is indicative of the latch state in the detergent-skinned smooth muscle tissues.

In this study we have shown that alterations in the free [Mg²⁺] differentially affect the magnitude and Ca²⁺ dependence of stress development and stress maintenance, as defined in Fig. 1. An increase in [Mg²⁺] above 0.1 mM significantly decreased the Ca²⁺ sensitivity of stress development; however, further increases in [Mg²⁺] (i.e., 1.0-6.0 mM) had no effect on the Ca²⁺ sensitivity of stress development. A major decrease in the magnitude of developed stress was not noted until [Mg²⁺] reached 6 mM. This decrease in the magnitude of developed stress in the presence of 6 mM Mg²⁺ was not accompanied by a decrease in Ca²⁺ sensitivity. In contrast, sequential increases in [Mg²⁺] (from 0.1 to 6 mM) significantly decreased the Ca²⁺ sensitivity of MLC phosphorylation at each step change in [Mg²⁺]. The Mg²⁺-dependent decrease in the Ca²⁺ sensitivity of MLC phosphorylation without a concomitant decrease in the Ca²⁺ sensitivity of stress development resulted in a separation of stress development and MLC phosphorylation at each [Ca²⁺]. All Mg²⁺-dependent alterations in actin-activated myosin ATPase activity correlated with changes in stress and not changes in MLC phosphorylation.

Increasing [Mg²⁺] from 0.1 mM did not significantly affect the Ca²⁺ sensitivity of stress maintenance but dramatically increased the magnitude of the maintained stress as defined in Fig. 1. In agreement with previous findings (4, 26), the Ca²⁺ sensitivity of stress maintenance was significantly higher than that for stress development at all [Mg²⁺] examined. In addition, the Ca²⁺ sensitivity of MLC phosphorylation was not different in the stress development and stress maintenance protocols at any [Mg²⁺] examined. Therefore, an increase in [Mg²⁺] decreased stress development and MLC phosphorylation as well as the Ca²⁺ sensitivity of MLC phosphorylation but augmented the magnitude of stress maintenance. Low [Mg²⁺] (0.1 mM) attenuated stress hysteresis, suggesting that the latch state requires Mg²⁺, directly or indirectly, for expression.

The simplest explanation for the Mg²⁺-dependent decrease in magnitude and Ca²⁺ sensitivity of MLC phosphorylation is Ca²⁺/Mg²⁺ competition at calmodulin cationic binding sites (39). An increase in [Mg²⁺] would effectively decrease the level of activation of the MLC kinase and therefore reduce the apparent Ca²⁺ sensitivity and magnitude of MLC phosphorylation. However, at an ionic strength similar to that used in this study (0.12 M), Dedman et al. (5) showed that high [Mg²⁺] does not compete with Ca²⁺ for calmodulin binding sites. This would then suggest that calmodulin-dependent activation of MLC kinase may not be the site of Mg²⁺ action.

A direct inhibitory affect of Mg²⁺ on MLC kinase activity is a possible explanation for the decrease in magnitude and Ca²⁺ sensitivity of MLC phosphorylation. It should be noted that Paul and Rüegg (33) demonstrated that contracted skinned fibers of the guinea pig taenia coli rapidly relax with concomitant MLC dephosphorylation if transferred to a Mg-free but Ca-containing solution. These investigators further demonstrated that the Ca²⁺-dependent increase in MLC phosphorylation and force apparently required 2 mM Mg²⁺. This supports the idea that an increase in Mg²⁺ does not inhibit MLC kinase, but rather that MLC kinase activity is actually dependent on a finite [Mg²⁺]. Thus their results are not consistent with the Mg²⁺-dependent decrease in net MLC phosphorylation levels seen in the present study if the MLC kinase is the primary site of Mg²⁺ action. DiSalvo et al. (7) presented evidence demonstrating no Mg²⁺ dependence of a MLC phosphatase isolated from aortic smooth muscle, whereas several studies (23, 31, 32) have shown that increasing the [Mg²⁺] increases the activity of a MLC phosphatase. Activation of a MLC phosphatase could account for the decrease in Ca²⁺ sensitivity of MLC phosphorylation.

A more difficult finding to explain is the rightward shift in the Ca²⁺ sensitivity of MLC phosphorylation without any significant change in the Ca²⁺ sensitivity of stress development. In the presence of high [Mg²⁺], developed stress was supported by lower levels of MLC phosphorylation than that in the presence of low [Mg²⁺]. Consistent with previous work by Barsotti et al. (3), more stress develops at high [Mg²⁺] than at low [Mg²⁺], when measured at equivalent levels of MLC phosphorylation. We have previously demonstrated in the skinned swine carotid media that [Mg²⁺]  6 mM induced stress development in the absence of both Ca²⁺ and MLC phosphorylation (24). Qualitatively similar results were shown by Ikebe et al. (15), using skinned gizzard fibers. We have suggested that this Mg²⁺-induced contraction may be the result of direct activation of a MLC phosphorylation-independent contractile regulatory system (24). In the present study, [Mg²⁺] below the threshold for a Ca²⁺-independent contraction produced a separation in stress and MLC phosphorylation. It is possible that a synergistic or cooperative interaction between Mg²⁺-dependent and Ca²⁺-dependent activation increased stress in the absence of proportional levels of MLC phosphorylation. On the other hand, Ikebe et al. (15) have shown that high [Mg²⁺] induces a conformational change in myosin similar to that caused by phosphorylation of the MLC and therefore both Mg²⁺ and Ca²⁺ initiate contraction through a similar mechanism. Our previous results demonstrating that calmodulin antagonists inhibit Mg²⁺-dependent, MLC phosphorylation-independent stress development (24) and that increasing tissue length inhibits Mg²⁺-dependent but not Ca²⁺-dependent stress (24, 25) would suggest this is not the case and that a true dissociation of stress and MLC phosphorylation results from an increase in the cellular [Mg²⁺].

As discussed above, the [Mg²⁺] that induced stress development are substantially greater than those used in this study. It is evident from the results of this study, however, that increasing [Mg²⁺] from 0.1 to 6 mM augments the magnitude of stress maintenance. The Mg²⁺ effect may be due to an increase in the cellular [MgADP], as would be suggested by the work of Kerrick and Hoar (17). Although the addition of an ATP regenerating system (10 mM creatine phosphate; 100 U/ml creatine kinase) did not significantly affect our results (data not shown), we cannot rule out this potentially important mechanism. This is supported by work from Somlyo's laboratory (20) demonstrating that an increase in [MgADP] significantly slowed relaxation and enhanced force maintenance. Moreover, they have directly demonstrated that an increase in the [MgADP]-to-[MgATP] ratio increased stress hysteresis in skinned tonic smooth muscle (19). The enhancement of stress hysteresis by an increase in [MgADP] is similar to that documented in our present study by an increase in [Mg²⁺]. Moreover, if the increase in [Mg²⁺] increased [MgADP], then one would expect the decrease in actin-activated myosin ATPase activity as was seen in Table 2. The increase in [MgADP] would increase the population of cross bridges in the force-generating state, while slowing ATPase activity.

An important point to discuss is the possible mechanism(s) by which Mg²⁺ or MgADP may dissociate stress from MLC phosphorylation levels. One potential mechanism may simply be that the increase in Mg²⁺ produces a rightward shift in the Ca²⁺ sensitivity of MLC phosphorylation as discussed above coupled with the concomitant increase in [MgADP] producing a leftward shift in the Ca²⁺ sensitivity of stress (17). Therefore, a dissociation of stress and MLC phosphorylation would result without invoking a true regulatory protein pathway. A second possibility is that dephosphorylated, strained cross bridges have a high affinity for MgADP, similar to that measured in rigor. Thus this population of cross bridges could maintain high levels of stress without proportional levels of MLC phosphorylation (10, 19, 20). The last possibility arises from a recent study from our laboratory demonstrating that loss of the thin filament protein caldesmon results in active cross-bridge cycling in unstimulated swine carotid arterial strips (8). This cross-bridge cycling was not associated with an increase in MLC phosphorylation levels above basal. It is interesting to speculate that the effect of Mg²⁺ shown in this present study to induce high levels of maintained stress may be due to disinhibition of caldesmon. For example, as [Ca²⁺] is decreased in the presence of 0.1 mM Mg²⁺, the inhibitory effects of caldesmon may increase as [Ca²⁺] is decreased, producing complete and relatively rapid relaxation. However, if Mg²⁺ depresses the inhibitory effect of caldesmon, then as [Ca²⁺] is lowered in our stress hysteresis protocol, the magnitude of the maintained stress would be increased. In reality, all three mechanisms described above are most likely involved in mediating the effect of Mg²⁺.

One possible explanation for the high maintained stress in the presence of Mg²⁺ but absence of Ca²⁺ that can be ruled out is cellular deterioration. Relaxation of the maintained stress by a decrease in [Mg²⁺] suggests that exposure of the tissues to 6 mM Mg²⁺ did not have deleterious effects on the smooth muscle cell. This suggestion is further supported by work from our group demonstrating that a prior contraction in response to high Mg²⁺ had no effect on either Ca²⁺-dependent stress development or Ca²⁺-dependent MLC phosphorylation (24). Moreover, the fact that the high levels of maintained stress are relaxed by a decrease in [Mg²⁺] and that exposure to millimolar [Mg²⁺] does not adversely affect Ca²⁺-dependent responses argues against structural alterations as a possible explanation for our results.

Estimates of the intracellular free [Mg²⁺] in the swine carotid artery suggest that the range of Mg²⁺ in this study that elicited stress hysteresis may be within physiological limits. Kushmerick et al. (21) calculated a [Mg²⁺] of 0.40 and 0.46 mM from NMR measurements of rabbit uterus and bladder, respectively. Of particular interest is the work of Kopp et al. (18), who used NMR measurements to calculate a [Mg²⁺] of 0.54 mM in flaccid swine carotid arteries and 0.99 mM in arteries pressurized to 100 mmHg. Similarly, the [Mg²⁺] has been estimated to be 0.98 mM in guinea pig taenia cecum (37). It has also been shown that agonist activation of vascular smooth muscle can increase free Mg²⁺ levels by >2 mM (30). This information would suggest that the intracellular [Mg²⁺] is in the range necessary for stress maintenance without proportional levels of MLC phosphorylation. It has long been known that smooth muscle has a relatively high [Mg²⁺] requirement for activation (9, 16, 23, 29, 34); our results would suggest that the latch state or stress maintenance may be the step involved in this requirement.

In summary, we interpret the results of this study to suggest that Mg²⁺ differentially affects the Ca²⁺-dependent regulatory system responsible for the rapid development of stress (MLC phosphorylation dependent) and the Ca²⁺-dependent regulatory system responsible for the slow development and/or maintenance of stress (MLC phosphorylation independent; the latch state). Mg²⁺ or MgADP is apparently a requirement for expression of stress hysteresis indicative of the latch state. We believe the results of this study provide additional evidence demonstrating that MLC phosphorylation is not a simple switch for the initiation of a smooth muscle contraction and that additional events must be involved. The results of this study also demonstrate that Mg²⁺ is an important tool in experiments designed to characterize and discern MLC phosphorylation-independent regulatory systems.