cGMP-dependent protein kinase I (cGKI) induces relaxation of smooth muscle via several pathways that include inhibition of intracellular Ca2+ signaling and/or involve activation of myosin phosphatase. In the present study, we investigated these mechanisms comparatively in colon and jejunum longitudinal smooth muscle from mice. In simultaneous recordings from colon muscle, 8-bromo-cGMP (8-Br-cGMP) reduced both carbachol-induced tension and carbachol-induced increase in intracellular Ca2+ concentration ([Ca2+]i). These effects of 8-Br-cGMP were absent in colon from mice carrying a mutated inositol-1,4,5 trisphosphate receptor I-associated G kinase substrate (IRAG) gene or lacking cGKI. However, in jejunum, 8-Br-cGMP reduced carbachol-induced tension but did not change corresponding [Ca2+]i signals. This setting was also observed in jejunum from mice carrying a mutated IRAG gene, whereas no response to 8-Br-cGMP was observed in jejunum from mice lacking cGKI. After inhibition of phosphatase activity by calyculin A, 8-Br-cGMP did not relax jejunum but still relaxed colon muscle. In Western blot analysis, 8-Br-cGMP reduced the signal for phosphorylated MYPT-1 in carbachol-stimulated jejunum but not in colon. These results suggest that cGMP/cGKI signaling differentially inhibits contraction in the muscles investigated: in jejunum, inhibition is performed without changing [Ca2+]i and is dependent on phosphatase activity, whereas in colon, inhibition is mediated by inhibition of [Ca2+]i signals.
- inositol-1,4,5 trisphosphate receptor I-associated G kinase substrate
- myosin phosphatase targeting subunit 1
- guanosine 3′,5′-cyclic monophosphate-dependent protein kinase I
nonadrenergic, noncholinergic neurons of the gut use nitric oxide (NO) as a neurotransmitter to regulate smooth muscle contractility via cGMP/cGMP-dependent protein kinase I (cGKI) signaling (3, 13, 15, 25). Since NO is gaseous, it diffuses directly from the neurons to its receptor, the soluble guanylyl cyclase (sGC) located in smooth muscle cells. Stimulation of sGC increases intracellular levels of cGMP and induces relaxation via activation of cGKI (10).
Smooth muscle contraction is controlled by Ca2+-dependent and Ca2+-independent signaling pathways (6, 15, 22, 29). It is therefore likely that relaxation by NO/cGMP/cGKI interferes with both Ca2+-dependent and -independent signaling pathways. Indeed, several mechanisms have been reported by which cGKI mediates relaxation in smooth muscle. Ca2+-dependent mechanisms involve cGMP/cGKI inhibition of hormone-induced Ca2+ release from intracellular stores that is not observed after deletion of the cGKI gene in mice (18). Further studies have shown that the latter mechanism involves a protein complex that consists of at least three proteins, i.e., the inositol-1,4,5 trisphosphate receptor (IP3R), the cGKI, and, as a scaffold protein between both, the IP3R-associated G kinase substrate (IRAG) (21). Ca2+-independent mechanisms include cGMP/cGKI-dependent phosphorylation of myosin phosphatase targeting subunit 1 (MYPT-1) that increases or resumes the activity of myosin light chain phosphatase (MLCP) and leads to dephosphorylation of myosin regulatory light chain (14, 22). A further regulatory mechanism is the cGKI-induced increase in the open probability of Ca2+-activated K+ channels (BKCa) (1, 5, 20, 31), resulting in hyperpolarization of the membrane potential and closure of L-type Ca2+ channels.
To clarify whether cGMP/cGKI signaling involves multiple pathways for relaxation in intestinal smooth muscle, we investigated cGMP-mediated effects in longitudinal smooth muscle isolated from the mouse colon and jejunum by simultaneously monitoring carbachol-induced tension and Ca2+ signals. In addition, we used mutant mouse lines that lacked cGKI or exhibited a mutated IRAG to clarify in more detail the mechanisms of cGMP-mediated inhibition of smooth muscle tone.
MATERIAL AND METHODS
All experiments complied with the European guidelines for the use of experimental animals and were approved by the local animal ethics committee. Wild-type mice, mice lacking cGKI (cGKI−/− mice), or mice exhibiting a mutated IRAG (IRAGΔ12/Δ12 mice) were bred as described previously (8). All mice used were euthanized by decapitation; intestinal segments were quickly transferred to buffer solution (in mM: 137 NaCl, 5.4 KCl, 1.8 CaCl2, 1 MgCl2, 12 NaHCO3, 0.42 NaH2PO4, and 5.6 glucose) bubbled with carbogen (95% O2-5% CO2). The segments were washed and cleaned from connective tissue. Rectangular pieces of longitudinal smooth muscle were obtained from intestinal segments as described previously (9). Briefly, the segments were pulled onto a glass pipette, and sheets of the longitudinal muscle were smoothly teased free from the underlying circular muscle layer after the outer muscle layer was scored longitudinally with a razor blade. Muscle segments were mounted longitudinally into organ baths (Myograph 601, www.dmt.dk, or Fibermic, www.si-heidelberg.com). Tension was recorded isometrically at 37 ± 1°C. All muscles used for contraction experiments showed spontaneous contractile activity that, however, often declined within 30 min of equilibration in buffer solution at 37°C. Preload tension was individually applied to give the maximum of spontaneous contractions.
Ca2+ signals in longitudinal smooth muscle were recorded at 1 Hz using a RatioMaster system (Photon Technology International) with a photomultiplier as described previously (9). Briefly, muscle strips were loaded with 10 μM fura-2 AM in the presence of 0.01% Pluronic acid for 1 h at 37°C and then mounted into the organ bath. Extracellular dye and Pluronic acid were washed out by superfusion with buffer solution at 1 ml/min for 30 min. The fluorescence signal at 510 nm was recorded at 1 Hz during repeated alternating excitation at 340 and 380 nm.
For Western blot analysis, longitudinal muscles from colon and jejunum were stimulated with carbachol (10 μM, 1 min) with or without 8-bromoguanosine 3′,5′-cyclic monophosphate (8-Br-cGMP; 300 μM, 10 min) in reaction tubes containing buffer solution at 37°C. After this procedure, muscles were frozen in liquid N2. Homogenization of frozen preparations was done in preheated lysis buffer (2% SDS and 50 mM Tris·HCl, pH 7) using a Fastprep device (MP Bio Science). The homogenates were heated at 95°C for 10 min and then centrifuged twice for 5 min at 18,000 g. Protein concentration was determined using bicinchoninic acid assay. The supernatant fractions were separated on a 10% SDS-polyacrylamide gel and electrotransferred to polyvinylidene difluoride membranes. Blots were treated with 5% nonfat milk powder in Tris-buffered saline and labeled with anti-MYPT-1 antibody (1:200; Upstate), anti-P-Thr-696MYPT-1 antibody (1:200; Upstate), or anti-β-actin antibody (1:5,000; Novus Biologicals).The β-actin signal was used as loading control. Density of bands was analyzed with EZQuant (EZQuant Biology Software Solutions).
All salts and substances were used as pure as commercially available and purchased from Sigma unless otherwise indicated. Substances were applied as single dose or cumulatively to achieve the concentrations as indicated.
Results are presented as blots, original recordings, or means ± SE. Effects of substances were analyzed in steady-state conditions. Ca2+ signals were expressed as emission ratios at 510 nm during alternate excitation at 340 and 380 nm. Changes in tension or Ca2+ signals were determined with respect to the maximum of the signals and the baseline disregarding spontaneous activity. Statistical comparisons of data sets were performed by Student's t-test using Prism 4 software (GraphPad). Differences were considered significant at P < 0.05.
Effects of 8-Br-cGMP in smooth muscle from wild-type mice.
The muscarinic agonist carbachol induced contractions and simultaneously increased Ca2+ signals in both longitudinal colon and jejunum smooth muscle from wild-type mice (Fig. 1, A–F). Preincubation with a membrane-permeable derivative of cGMP (8-Br-cGMP; 300 μM) inhibited carbachol-induced contractions in both muscles. However, 8-Br-cGMP prevented carbachol-induced Ca2+ signals only in colon and not in jejunum smooth muscle (Fig. 1, A–F). These results imply that cGMP reduces smooth muscle tone by different mechanisms in the respective smooth muscle: in colon smooth muscle, 8-Br-cGMP inhibits contraction by reducing the increase in intracellular Ca2+ concentration ([Ca2+]i), whereas in jejunum muscle, 8-Br-cGMP inhibits contraction without changing [Ca2+]i.
Effects of 8-Br-cGMP in smooth muscle from cGKI−/− mice.
8-Br-cGMP represents a potent activator of cGKI but also is able, at least at high concentrations, to interact with cAMP-dependent protein kinase (19, 28). As a consequence, we investigated the effects of 8-Br-cGMP on muscles from cGKI−/− mice to clarify to which extent the effects of 8-Br-cGMP are mediated by cGKI. In muscles from cGKI−/− mice, carbachol induced contractions that were not different in amplitude from those recorded in muscles from wild-type mice. 8-Br-cGMP did not affect the carbachol-induced contractions and Ca2+ signals in both colon and jejunum muscles from cGKI−/− mice, indicating that the effects of 8-Br-cGMP observed in wild-type mice are exclusively mediated by cGKI (Fig. 1, G–J).
Effects of 8-Br-cGMP in smooth muscle from IRAGΔ12/Δ12 mice.
Inhibition of hormone-induced increases in [Ca2+]i by 8-Br-cGMP has been ascribed to cGKI-mediated phosphorylation of IRAG and subsequent inhibition of IP3R-mediated Ca2+ release from intracellular stores (21). IRAG has been detected in both colon and small intestinal muscle segments (7). In this study, we investigated the effects of 8-Br-cGMP on colon and jejunum longitudinal smooth muscle from mutant mice that express a mutated IRAG (IRAGΔ12/Δ12 mice) (8). Carbachol induced contractions in both muscles from IRAGΔ12/Δ12 mice that were not different in amplitude from those recorded in muscles from wild-type mice. 8-Br-cGMP did not affect carbachol-induced tension or Ca2+ signals in colon smooth muscle from IRAGΔ12/Δ12 mice. However, the effects of 8-Br-cGMP in jejunum smooth muscle from IRAGΔ12/Δ12 mice were not different from those observed in muscles from wild-type mice; i.e., 8-Br-cGMP reduced carbachol-induced tension but did not influence the carbachol-induced Ca2+ signals (Fig. 2, A–F). These results suggest that the cGMP-mediated inhibition of muscle contraction in colon muscle relies on IRAG-controlled inhibition of the increase in [Ca2+]i, whereas in jejunum muscle, 8-Br-cGMP relaxes without affecting IRAG-dependent changes in [Ca2+]i.
Relaxation by 8-Br-cGMP of carbachol-induced tension after phosphatase inhibition.
So far, the results indicate that 8-Br-cGMP inhibits carbachol-induced contractions in jejunum muscle independently of changes in [Ca2+]i. Ca2+-independent relaxations are mainly mediated by an increase or restoration in the phosphatase activity of the MLCP (22). Thus we studied the relaxing effects of 8-Br-cGMP on carbachol-induced contraction after inhibition of phosphatase activity by the unspecific phosphatase inhibitor calyculin A (11). Treatment with 100 nM calyculin A increased the carbachol-induced tone (Fig. 3A) and induced a slow rundown of the contraction with a decrease of 0.05 ± 0.1 mN/mg per 20 min. In the presence of calyculin A, 8-Br-cGMP induced relaxation in colon muscle but not in jejunum muscle from wild-type mice (Fig. 3A). This finding indicates that phosphatase activity is essential for 8-Br-cGMP-mediated relaxation in jejunum but not in colon muscle.
Effects of Y27632 in smooth muscle from wild-type mice.
Inhibition of Rho kinase reduced carbachol-mediated contraction but did not change intracellular Ca2+ signals in longitudinal ileum muscle from guinea pig (24). Likewise, the Rho kinase inhibitor Y27632 (10 μM) reduced carbachol-mediated contractions in longitudinal jejunum muscle but, however, not in colon muscle from mice (Fig. 3, B and C). Carbachol-mediated Ca2+ signals were not changed by Y27632 in both tissues (Fig. 3, B and C). These results indicate that carbachol-mediated contraction involves Rho kinase activity in jejunum muscle but not in colon muscle.
Effect of 8-Br-cGMP on phosphorylation of MYPT-1.
Next, we investigated the effects of 8-Br-cGMP on the phosphorylation levels of MYPT-1. MYPT-1 represents a regulatory subunit of MLCP. Agonist-activated RhoA/Rho kinase signaling has been shown to increase phosphorylation of MYPT-1 at Thr696, which is assumed to inhibit MLCP activity and thus maintain or induce [Ca2+]i-independent muscle contraction (16, 27). MYPT-1 is supposed to also be regulated by cGMP/cGKI-mediated phosphorylation to counteract agonist-induced MLCP inhibition (22, 23). Indeed, cGMP signaling inhibited the signal for MYPT-1 phosphorylation at Thr696 (16). To prove whether 8-Br-cGMP influences the phosphorylation of MYPT-1 in jejunum smooth muscle, we performed Western blot analysis using a P-Thr-696MYPT-1-specific antibody (Fig. 4). Carbachol increased the signal for P-Thr-696MYPT-1 in jejunum but only barely in colon muscle. In jejunum, preincubation with 8-Br-cGMP (300 μM) decreased P-Thr-696MYPT-1 signals after stimulation with carbachol. In contrast, 8-Br-cGMP did not significantly change P-Thr-696MYPT-1 signals in colon. No effects of 8-Br-cGMP on P-Thr-696MYPT-1 signals were observed in muscles from cGKI−/− mice (not shown). Together, these findings suggest that inhibition of phosphorylation of MYPT-1 at Thr696 is a prominent mechanism involved in relaxation of jejunum but not of colon longitudinal smooth muscle.
The present study shows that NO/cGMP/cGKI signaling inhibits intestinal smooth muscle tone by a Ca2+-dependent mechanism in colon smooth muscle and by a Ca2+-independent mechanism in jejunum smooth muscle. No effects of cGMP were observed in muscles from cGKI−/− mice, indicating that both mechanisms are mediated exclusively by cGKI without any cross-activation of cAMP-dependent protein kinase.
Ca2+-dependent inhibition of smooth muscle tone by cGMP/cGKI was directly observed in colon longitudinal smooth muscle. Stimulation with carbachol elicited contractions and intracellular Ca2+ signals that were blocked by cGMP. None of these effects of cGMP were observed in colon muscle from IRAGΔ12/Δ12 mice. Similarly, cGMP did not reduce bradykinin-induced Ca2+ transients after inactivation of IRAG in cultured human colon cells (4). These results confirm the view that cGMP/cGKI-induced inhibition of contraction is mediated by IRAG-controlled reduction of intracellular Ca2+ signals in colon smooth muscle, most probably via interaction with the IP3R and, consequently, inhibition of IP3-induced Ca2+ release from intracellular Ca2+ stores (8). In line with this view, inhibition of phosphatase activity did not interfere with the relaxing effects of cGMP in colon longitudinal muscle.
Ca2+-independent relaxation of smooth muscle tone has been mainly observed after pharmacological inhibition of Rho kinase by Y27632 and, in addition, by cGMP-mediated activation of MLCP in permeabilized preparations from ileum smooth muscle (24, 30). In the present study, Ca2+-independent inhibition of smooth muscle tone by cGMP/cGKI was directly monitored in jejunum longitudinal smooth muscle. Stimulation with carbachol induced contractions and intracellular Ca2+ signals. cGMP blocked the contraction but did not change the intracellular Ca2+ signals in jejunum from both wild-type and IRAGΔ12/Δ12 mice. Thus cGMP/cGKI seem to inhibit jejunum smooth muscle tone via increasing/restoring MLCP activity as described also for ileum (22). This view is supported by the findings that 1) inhibition of phosphatase activity abolished relaxation by cGMP and 2) phosphorylation of MYPT-1 at Thr696, which has been reported to be involved in Rho kinase-induced contraction (27), was reduced by cGMP.
The exact mechanism by which cGMP/cGKI signaling reduces MYPT-1 phosphorylation is still under debate. Evidence has been presented that cGMP/cGKI-mediated phosphorylation of Ser695 prevented Rho-kinase induced phosphorylation of Thr696 and, thus, inhibited contraction (27). However, relaxation by cGMP was not prevented by Ala substitution of Ser695 using an active MYPT-1 fragment in permeabilized ileum smooth muscle (12). In addition, in jejunum smooth muscle, no increase in Ser695 phosphorylation by cGMP could be detected using a commercially available antibody (data not shown).
Phosphorylation of MYPT-1 after muscarinic stimulation has been observed in several intestinal tissues, e.g., in guinea pig ileum (24) or in rabbit colon smooth muscle cells (17). MYPT-1 has been detected in both longitudinal colon and jejunum smooth muscle from mice, but a decreased signal for phosphorylated MYPT-1 at Thr696 after stimulation with carbachol and incubation with 8-Br-cGMP was only observed in jejunum and not in colon smooth muscle (Fig. 4). In addition, inhibition of Rho kinase by Y27632 reduced carbachol-induced contractions in jejunum but not in colon smooth muscle. Thus Ca2+ sensitization by Rho kinase signaling seems to be only of minor importance for carbachol-induced contraction in colon muscle. Carbachol-induced contraction in colon muscle may rather depend on L-type Ca2+ channel activity. In support of this view, muscarinic-induced contraction was completely inhibited by the Ca2+ channel blocker nifedipine (2) and clearly reduced in Cav1.2 channel-deficient colon muscle preparations (26).
In conclusion, colon and jejunum longitudinal smooth muscle from mice represent two prototypes of smooth muscle with respect to cGMP/cGKI signaling: in colon muscle, inhibition of muscle tone depends on [Ca2+]i with cGMP/cGKI acting primarily via IRAG-dependent inhibition of [Ca2+]i increase, whereas in jejunum muscle, inhibition of muscle tone is independent on [Ca2+]i with cGMP/cGKI acting mainly via increasing or restoring phosphatase activity.
This study was supported by the Deutsche Forschungsgemeinschaft.
We thank Lucia Koblitz for excellent technical assistance.
- Copyright © 2009 the American Physiological Society