(Received for publication, July 10, 1996, and in revised form, December 16, 1996)
From the Department of Physiology and Biophysics, Georgetown University Medical Center, Washington, D. C. 20007
Using permeabilized, arterial smooth muscle strips where membrane-associated pathways remain intact but intracellular Ca2+ stores are depleted, we investigated mechanism(s) for the Ca2+ desensitization of contractile force by cGMP. The nonhydrolyzable analog 8-bromo-cGMP, when applied to these strips with submaximal Ca2+ levels clamped, dramatically and reversibly reduced the steady state levels of phosphorylation at 20-kDa myosin light chain and contractile force, with a nanomolar concentration required to obtain 50% reduction. Supramaximal concentrations of 8-bromo-cGMP (10 µM), however, did not change the steady state relationship between phosphorylation and force. When light chain phosphatase activity was blocked at pCa 6.7, 10 µM 8-bromo-cGMP did not affect the rates of rise of light chain phosphorylation and contractile force. When light chain kinase activity was blocked, 10 µM 8-bromo-cGMP significantly accelerated light chain dephosphorylation and force relaxation from the maximal contraction steady state. The light chain phosphorylation time course of a pCa 6.0-induced contraction in the presence of 8-bromo-cGMP exhibited kinetics that are predictable from a mathematical model in which only light chain phosphatase activity is increased. The results of this study strongly suggest that cGMP indirectly activates light chain phosphatase, the first proposed mechanism for cGMP-induced Ca2+ desensitization in vasodilatation.
The primary mechanisms that regulate smooth muscle contraction and relaxation are, respectively, phosphorylation of the regulatory 20-kDa myosin light chain (MLC20)1 at Ser-19 by MLC20 kinase (MLCK) and its dephosphorylation by MLC20 phosphatase (MLCP) (1, 2). Typically, intracellular Ca2+ levels modulate the MLCK to MLCP activity ratio and ultimately the degree of contractile force because MLCK activity depends on the amount of the Ca2+-calmodulin complex, which itself hinges on cytosolic Ca2+ levels. In many cases, however, the sensitivity of force to Ca2+ can be changed by physiological modulation of the Ca2+ dependence of MLC20 phosphorylation (3, 4) or by other mechanisms such as thin filament disinhibition (5, 6).
The nucleotide cyclic GMP has emerged as a potent, physiological second messenger involved in both vasodilator action and failure of vasoconstrictor activity. The neighboring endothelium, endogenous circulating hormones, as well as clinically administered nitrovasodilators all function to widen the lumen of vessels by stimulating guanylyl cyclase in vascular smooth muscle to produce cGMP (7). To date, cGMP has been implicated in lowering intracellular Ca2+ (8, 9) and in decreasing sensitivity of the contractile force to Ca2+ (10, 11). A myriad of Ca2+ lowering mechanisms have since been reportedly shown (see Ref. 9 for references): increased Ca2+ sequestration, increased Ca2+ efflux, decreased Ca2+ influx through decreased Ca2+ channel activity and through hyperpolarization via increased K+ channel activity, and decreased Ca2+ release through antagonism of second messenger formation and possibly through phosphorylation of its receptor. In stark contrast, there are currently no proposed mechanisms for the Ca2+ desensitization by cGMP because little research has been carried out on this effect. Still, no one has been able to belittle the significance of force desensitization in the vasodilating effect of cGMP (12). In this study, we not only demonstrate a very high affinity for the nonhydrolyzable cGMP analog, 8-bromo-cGMP (8Br-cGMP) to bring about Ca2+ desensitization, but we also provide convincing evidence that cGMP activates MLCP. A short report of our work has appeared previously in abstract form (13).
Smooth muscle strips (70 µm thick, 700-800 µm wide, and 3 mm long) were dissected from
rabbit femoral arteries and carefully freed of connective tissue; the
endothelia were removed by rubbing with a razor blade. The strips were
then tied with silk monofilaments to the fine tips of two tungsten
needles, one of which was connected to a force transducer (AM801,
SensoNor), and mounted over a well filled with solution on a Teflon
bubble plate to allow for moderately rapid (within a second) solution
exchange and freezing (3). Experiments were, with a single exception
(15 °C in Fig. 6), carried out at 20 °C.
10 µM 8Br-cGMP on the time course of MLC20 phosphorylation (B) and contraction (A). After treatment in a 0.1 mM EGTA relaxing solution ± 8Br-cGMP for 10 or more min, strips were contracted by pCa 6.0 (10 mM EGTA) solution at 15 °C. The temperature was lowered slightly in order to slow down the time course so that the rate of rise could be more precisely analyzed. C, the following simple two-state model for the reaction of MLC20 phosphorylation is assumed for model calculation of its time course:
![]() |
(Eq. 1) |
![]() |
(Eq. 2) |
Solutions
The standard relaxing solution (3) used for
resting states of the permeabilized strips contained the following:
74.1 mM potassium methanesulfonate, 2 mM
Mg2+, 4.5 mM MgATP, 1 mM EGTA, 10 mM creatine phosphate, 30 mM
piperazine-N,N-bis(2-ethanesulfonic acid). At
times, we used slightly modified relaxing solutions in which the
concentration of EGTA was different. In the activating solution, 10 mM EGTA was used, and a calculated amount of calcium methanesulfonate was added to give the final desired concentration of
free Ca2+ ions (14). In the rigor solutions, the solute
composition was like that of the relaxing solutions except for the
absence of MgATP and creatine phosphate. All solutions were neutralized
to pH 7.1 with KOH at 20 °C, and an ionic strength of 0.2 M was achieved by appropriately using more or less
potassium methanesulfonate.
After measuring the force of
contractions induced by high K+ (154 mM) and by
phenylephrine (100 µM) in freshly dissected strips, these
strips were incubated in the standard relaxing solution for several
minutes. For plasma membrane permeabilization with -toxin, the
strips were then treated for 30 min at 30 °C with 5000 units/ml of
purified Staphylococcus aureus
-toxin (Life Technologies,
Inc.) at pCa 6.7 buffered with 10 mM EGTA (15). For plasma
membrane permeabilization with
-escin, the strips were then instead
treated with 40 µM
-escin in standard relaxing solution for 45 min at 5 °C and then for 15 min at 30 °C (15). The sarcoplasmic reticulum was also depleted of calcium by treating the
strips with 10 µM A23187, a Ca2+ ionophore,
for 15-20 min at 30 °C in standard relaxing solution. For the usage
of Triton X-100 to heavily permeabilize strips, we simply used 0.1% in
standard relaxing solution for 30 min at 4 °C and for 15 min at
30 °C
For measuring MLC20 phosphorylation, permeabilized preparations were rapidly frozen with liquid N2-cooled, liquid chlorodifluoromethane at the desired conditions, with force monitored up to the time of freezing. The various phosphorylated states of MLC20, unphosphorylated (U), monophosphorylated (P1), and diphosphorylated (P2), were separated by a two-dimensional isoelectric focusing SDS-polyacrylamide gel electrophoresis, blotted onto a nitrocellulose membrane, and stained with colloidal gold to produce separate spots for each phosphorylated state, and the amount at each spot was measured by its density (3). The percentage of MLC20 phosphorylation was calculated by dividing (P1 + P2) by (U + P1 + P2).
Fig.
1A graphically illustrates the marked effect
of 8Br-cGMP to cause Ca2+ desensitization of force in
-toxin-permeabilized rabbit femoral artery smooth muscle strips.
This trace demonstrates how 10 µM 8Br-cGMP potently and
reversibly elicits relaxation and how pretreatment with this compound
inhibits development of force.
A simple dose-response of the steady state, pCa 6.0-induced, submaximal
(near 60%) contractile force to 8Br-cGMP shown in Fig.
2 depicts the high efficacy of this cyclic nucleotide to exert its desensitizing effect on force. The IC50 was very
low at 8 ± 0.4 nM, whereas an 8Br-cGMP metabolite
8Br-5GMP had no significant effect on contraction, even in a
micromolar range (Fig. 2).
A more comprehensive and definitive example of the desensitization over
a range of buffered Ca2+ levels is shown in Fig.
3. The 8Br-cGMP effect on steady state levels of
contractile force (Fig. 3A) and MLC20
phosphorylation (Fig. 3B) is a parallel shift to the right,
with an average 2.5- and 2.9-fold increase in ED50 of
Ca2+ for force and phosphorylation, respectively. Of note,
force is reduced from 62 ± 2 to 8 ± 1%, and
phosphorylation is reduced from 62 ± 6 to 30 ± 6% at pCa
6.0, but the same concentration of the cyclic nucleotide affected
neither maximal phosphorylation steady state nor maximal force steady
state. With computer assistance, we extrapolated a line curve and its
equation from the data values in the absence of 8Br-cGMP at the
different Ca2+ levels of force and phosphorylation (Fig.
3C). The data points in the presence of the cyclic
nucleotide lie virtually on top of the control line with no significant
difference.
In the separate Triton X-100 experiments, 10 µM 8Br-cGMP was not capable of inducing significant relaxation of a submaximal steady state contraction at pCa 6.0 (Fig. 1). However, the addition of either calmodulin (Fig. 1, B1) or the selective phosphatase inhibitor microcystin-LR (MC-LR; Fig. 1, B2) did markedly enhance the level of force in the same preparations, thus demonstrating the integrity of contractile apparatus including MLCK and MLCP in the Triton X-100-permeabilized strips. Hence, it appears that cGMP is not able solely on its own to induce desensitization of contractile force to Ca2+ in the demembranated strips.
MLCK Activity Is UnaffectedAfter application of the
membrane-impermeable (PP1- and PP2A-selective) phosphatase inhibitor
MC-LR (16) at 10 µM in pCa 6.7 in -escin-permeabilized
strips, steady state levels of force and phosphorylation were attained
that did not increase further when the Ca2+ concentration
in the fibers was increased to a maximal pCa 4.5 (not shown). This
demonstrates the ability of the chosen inhibitor MC-LR at 10 µM to abolish in situ MLCP activity. In pCa
6.7 solution containing 10 µM MC-LR, pretreatment with
8Br-cGMP did not affect the steady state levels of or the rate of rise
for contraction (Fig. 4, A1) or
phosphorylation (Fig. 4, A2). To verify the efficacy of this
protocol, we used wortmannin (17) as a MLCK inhibitor that is known to
have inhibitory effects on other signaling pathways. Strips were
pretreated with 3 µM wortmannin instead of 8Br-cGMP, and
the rate of rise did indeed decrease (Fig. 4, A1).
In another set of experiments, the structurally distinct,
membrane-permeable (PP1- and PP2A-selective) phosphatase inhibitor calyculin A was used in combination with -toxin-permeabilized strips. Neither the rate of rise for contraction (Fig. 4,
B1) nor for phosphorylation (Fig. 4, B2) was
affected by the presence of 10 µM 8Br-cGMP, similar to
the results obtained using MC-LR in
-escin-permeabilized
preparations.
200 µM of the MLCK
inhibitor ML-9 (18) together with a Ca2+-free (10 mM EGTA) solution potently decreased maximum levels of contractile force and MLC20 phosphorylation to basal
levels, indicating practically complete inactivation of the MLCK under
these conditions. Under the same conditions, 8Br-cGMP significantly
increased the rates of both relaxation (Fig.
5A) and dephosphorylation (Fig. 5B); the average t1/2 for relaxation
decreased from 3.2 to 2.4 min, and the average t1/2
for dephosphorylation decreased from 53 to 39 s. ML-9, although
its effect is weak, was chosen for this protocol because of its
relatively rapid inhibition of Ca2+-activated contraction
and its relatively high selectivity for MLCK, whereas wortmannin was
not suitable for this protocol because of very slow kinetics of
inhibition of even isolated MLCK (17).
Time Course Kinetics Is Predictably Altered
In an unmodified pCa 6.0-induced contraction where MLCK and MLCP were not inhibited, 10 µM 8Br-cGMP decreased the peak levels of contractile force (Fig. 6A) and MLC20 phosphorylation (Fig. 6B) without affecting the initial rate of rise, thus decreasing the t1/2 as shown in the figure. According to a simple two-state MLC20 model (phosphorylated or unphosphorylated), we generated three predicted time courses of MLC20 phosphorylation (Fig. 6C). First, to simulate the steady state phosphorylation value obtained at pCa 6.0 (62% of total MLC20; see Fig. 3B), we chose an appropriate MLCK to MLCP activity ratio, 1.6 (control). To simulate possible 8Br-cGMP effects on the pCa 6.0-induced phosphorylation at steady state (30% of total MLC20), we either increased MLCP activity (MLCP × 3.8) or decreased MLCK activity (MLCK/3.8). The steady state levels were decreased to the same extent (30% of total MLC20) in both 8Br-cGMP simulations, but the actual experimental effect of 8Br-cGMP on initial time courses of both phosphorylation and contraction (Fig. 6, A and B) closely resembled the simulation with increased MLCP activity and deviated markedly from the simulation with decreased MLCK activity (Fig. 6C).
This study illustrates how nanomolar concentrations of 8Br-cGMP
but not 8Br-5GMP strongly and reversibly reduces the ability of
Ca2+ to cause the vascular smooth muscle contractile
machinery to generate force in situ. Moreover, it analyzes
this effect of 8Br-cGMP to identify which of the contractile or
regulatory apparatuses might be altered sufficiently to account for the
Ca2+ desensitizing effect.
The cGMP-induced desensitization of contractile force to Ca2+ must involve a Ca2+-independent reduction in the MLC20 phosphorylation by cGMP at a given Ca2+ level and/or a change to some MLC20 phosphorylation-independent regulatory mechanism(s). An example of the latter could be a cGMP-induced increase in inhibition of actomyosin ATPase by the thin filament-associated proteins caldesmon and calponin (5, 6, 19). However, because 8Br-cGMP does not appear to change the steady state relationship between phosphorylation and force (Fig. 3C), it is reasonable to exclude a significant involvement of such phosphorylation-independent mechanisms. In this tissue, it appears that the Ca2+ desensitization of force by 8Br-cGMP is mainly mediated through Ca2+-independent decreases in phosphorylation at MLC20. However, we cannot rule out the possibility of more complicated mechanisms of Ca2+ desensitization of force by cGMP during the nonsteady state phases of contraction and relaxation.
With that in mind, we set out to establish whether cGMP causes inactivation of MLCK, activation of MLCP, or some combination thereof. It is well known that in vitro phosphorylation of MLCK at its site A by protein kinase A, protein kinase C, and Ca2+-calmodulin-dependent kinase II inactivates MLCK by decreasing the affinity of MLCK for Ca2+-calmodulin (20). Hence phosphorylation at site A by any of the above mentioned kinases would cause a lower than normal MLC20 phosphorylation at a given Ca2+ level. Not surprisingly, it had been proposed that the cGMP-activated kinase (PKG) phosphorylates MLCK and accounts for the Ca2+ desensitizing effect of cGMP. Direct phosphorylation of MLCK by PKG does indeed occur in vitro but not at site A; furthermore, this phosphorylation does not modify the activity of MLCK (21, 22). Thus from the biochemical studies, it seems that cGMP is not capable of affecting MLCK activity through its protein kinase PKG. Still, the biochemical research on isolated MLCK could not provide any useful information to evaluate whether an indirect inhibition of MLCK by cGMP occurred. Our experiment, however, involving virtually complete elimination of MLCP activity (Fig. 4, A and B) lends physiological support to the idea that MLCK is unaffected by cGMP. 8Br-cGMP did not change the activity of MLCK, as evidenced by its ineffectiveness on the phosphorylation rate and rate of development of force. This result was confirmed in two different types of permeabilized preparations using two structurally different selective MLCP inhibitors. We realize that it is conceivable, although unlikely, that our chosen two phosphatase inhibitors MC-LR and calyculin A somehow interfere with the putative ability of PKG to inhibit MLCK, thus hiding such an effect.
At this point it now seemed by default that cGMP was exerting Ca2+ desensitization through activation of MLCP. The experiment summarized in Fig. 5 indeed indicates such a mechanism; 8Br-cGMP significantly increased the rates of dephosphorylation and relaxation. The t1/2 for dephosphorylation was 1.4 times greater in the absence of 8Br-cGMP and the t1/2 for relaxation 1.3 times greater. One concern we have, however, is over the inability of the Ca2+ chelator EGTA and the MLCK inhibitor ML-9 to diffuse instantaneously into the permeabilized strips and to abolish the MLCK activity immediately, creating some curiosity as to whether the effect might be even greater under ideal, diffusion-independent conditions.
An alternative approach to investigating the cGMP effect is to compare the kinetics of our experimental time courses with what can be predicted from a theoretical model (23, 24) that is based on the simple two-state phosphorylation/dephosphorylation reaction (Fig. 6; also see below). Our results mentioned above indicated that MLCP activation predominantly accounted for Ca2+ desensitization of the contractile force, but we suspected that diffusion limitations with our Fig. 5 protocol had been decreasing the visibility of such MLCP activation. Additionally, we could not rule out completely the possibility that our Fig. 4 protocol was somehow masking a putative MLCK inhibition by 8Br-cGMP. A comparison of the experimental and theoretical time course kinetics in Fig. 6 addresses those concerns in an ideological manner and suggests that cGMP elicits Ca2+ desensitization of force through a substantial activation of MLCP. A 3.8-fold increase in MLCP activity changed the kinetics of the simulated pCa 6.0-induced phosphorylation in a manner very similar and comparable to how 8Br-cGMP changed the phosphorylation kinetics of actual pCa 6.0-induced contractions. Thus we conclude from experimental results that cGMP creates marked Ca2+ desensitization of contractile force in vascular smooth muscle by activating MLCP. That finding would also imply that MLCP activity in situ under normal physiological conditions is not maximal. It has been interestingly suggested that PKG in pituitary tumor cells stimulates Ca2+- and voltage-activated potassium channels through activation of an okadaic acid-sensitive protein phosphatase (25). We suggest that activation of protein phosphatase by cGMP may be a general function in biological systems.
In our theoretical model, we made predictions regarding
MLC20 phosphorylation based on a single phosphorylation
site for MLCK. Although in vitro MLCK can phosphorylate 2 mol/mol of MLC20 (Ser-19 and Thr-18), the rate of
phosphorylation at the latter site is much lower than at the former
site and their dephosphorylations by MLCP occur at similar rate (26),
suggesting that only a very small portion of MLC20 is
diphosphorylated under physiological conditions. In fact, the extent of
phosphorylation is 1 mol/mol or lower in intact smooth muscle (27) and
1.1-1.2 mol/mol in the permeabilized muscle maximally stimulated by
Ca2+ ions alone or with GTPS (15). Additional Thr-18
phosphorylation increases the actin-activated myosin ATPase (26) but
does not increase the rate of movement of actin filaments in the
in vitro motility assay (28). Therefore, only one
phosphorylation site at MLC20 is enough and appropriate to
simulate the time course of MLC20 phosphorylation in our
conditions (23, 24).
The time course of force development was similar to that of MLC phosphorylation (Fig. 4), but the relaxation occurred much more slowly than the dephosphorylation (Fig. 5). The relaxation and dephosphorylation rates were somewhat affected by diffusion of EGTA and ML-9, but MLC phosphorylation at 1 min in the Ca2+-free (EGTA 10 mM), ML-9 solution was decreased by 60% in control and by 73% in the presence of 8Br-cGMP, even though force levels in both conditions were reduced only by 3-4% (Fig. 5). These observations are consistent with previous results (29) and can be explained by a slow onset detachment of dephosphorylated cross-bridges (30). Recently, Khromov et al. (31) precisely analyzed the biphasic time course of relaxation in permeabilized arterial smooth muscle by photolysis of a caged Ca2+ chelator. They suggested that the initial plateau phase of relaxation followed by an exponential decay may be due to continued cycling of remaining phosphorylated cross-bridges and cooperative cycling of dephosphorylated cross-bridges. It is therefore the length of the plateau phase of relaxation in tonic smooth muscle that may depend inversely upon the MLC dephosphorylation rate. The second exponential decay phase was suggested to be representative of the detachment rate of dephosphorylated cross-bridges, which is independent of the dephosphorylation rate. The time course of relaxation initiated by EGTA and ML-9 in this study was also biphasic (Fig. 5A). 8Br-cGMP reduced the duration of the plateau phase of relaxation but did not significantly affect the subsequent exponential decay. Using the work of Khromov et al. (31) to interpret our results shown in Fig. 5, one again concludes that 8Br-cGMP increases the rate of MLC dephosphorylation but does not affect detachment rate of dephosphorylated cross-bridge.
The loss of 8Br-cGMP's desensitizing effect in heavily permeabilized, demembranated samples implies that some soluble cytosolic target(s), such as PKG and/or diffusible cofactor(s), mediates that effect in intact and selectively permeabilized fibers, confirming previous work of Pfitzer et al. (10). This idea is in agreement with other studies involving cGMP because no known action of cGMP in smooth muscle relaxation has been shown to function independently of PKG (9). One can then reasonably assume from the low nanomolar IC50 of 8Br-cGMP shown in this study that the unknown anticipatory target involved in coupling cGMP to the activation of MLCP is likely PKG. However, because cGMP has also been shown to activate cAMP-dependent protein kinase (32), the possibility that high micromolar concentrations of 8Br-cGMP stimulate both PKG and protein kinase A signaling pathways cannot be excluded.
The Ca2+ level in smooth muscle is the primary determinant in the phosphorylation of MLC20 and hence in the contraction. The known mechanisms in this tissue for handling cytoplasmic Ca2+ and for modulating Ca2+ sensitivity of the contraction are synergistic and cooperative pathways. Excitatory agonists produce contraction by elevating Ca2+ levels and increasing Ca2+ sensitivity (33, 34), whereas vasodilating substances cause relaxation by reducing Ca2+ levels and decreasing Ca2+ sensitivity (35, 36). Furthermore, the excitatory agonists cause Ca2+ sensitization through inhibition of MLCP (23, 37), whereas we have shown in this study that cGMP cause Ca2+ desensitization through activation of MLCP. Thus, nitrovasodilators and the endothelium via production of NO appear to antagonize excitatory agonists in almost every reported mechanism, including a now apparent activation of MLCP versus agonist-induced inhibition of MLCP.