- Email: [email protected]
cGKI signaling in fundus smooth muscle
European Journal of Pharmacology 670 (2011) 266–271
Contents lists available at SciVerse ScienceDirect
European Journal of Pharmacology journal homepage: www.elsevier.com/locate/ejphar
Pulmonary, Gastrointestinal and Urogenital Pharmacology
Kinetics of relaxation by cGMP/cGKI signaling in fundus smooth muscle Claudia Ertl a, Robert Lukowski a, b, Katja Sigl a, Jens Schlossmann a, c, Franz Hofmann a, Jörg W. Wegener a,⁎ a b c
FOR923, Institut für Pharmakologie und Toxikologie, Technische Universität München, Germany Institut für Pharmazie, Pharmakologie, Toxikologie und Klinische Pharmazie, Universität Tübingen, Germany Institut für Pharmakologie und Toxikologie, Universität Regensburg, Germany
a r t i c l e
i n f o
Article history: Received 20 June 2011 Received in revised form 14 July 2011 Accepted 30 July 2011 Available online 19 August 2011 Keywords: Fundus Relaxation NANC Nitric oxide cGMP
a b s t r a c t cGMP-dependent kinase I (cGKI) is a major mediator of smooth muscle relaxation and exists in two isoforms, α and β. Both isoforms are supposed to mediate their effects via different intracellular signaling pathways. To verify this concept, the kinetics of relaxation mediated by either isoform was analyzed in gastric fundus smooth muscle from mice. Muscles from mice that express selectively the Iα or Iβ isoform of cGKI in smooth muscle (sm-cGKIα or sm-cGKIβ mice) were compared to muscles from conventional cGKI−/− mice. Fundus muscles were contracted by carbachol and then relaxed by 8-Br-cGMP or by electrical field stimulation (EFS). The time course of relaxation by 8-Br-cGMP was not different between muscles from sm-cGKIα and smcGKIβ mice. EFS induced a fast transient relaxation in muscles from sm-cGKIα and sm-cGKIβ mice that was blocked by the NO synthase inhibitor L-NAME. Recovery from this relaxation was about 4-times slower in muscles from sm-cGKIα mice than in muscles from sm-cGKIβ mice. The different kinetic of recovery from relaxation after EFS in sm-cGKIα and sm-cGKIβ mice suggests that different signaling pathways exist for each cGKI isoform in vivo in fundus muscles. © 2011 Published by Elsevier B.V.
1. Introduction Nitric oxide (NO) represents the major inhibitory neurotransmitter in the gastrointestinal tract (Shah et al., 2004; Toda and Herman, 2005). NO is released by non-adrenergic, non-cholinergic (NANC) neurons and causes relaxation of gastrointestinal smooth muscle by activating soluble guanylyl cyclase (sGC). sGC generates cGMP that activates cGMP-dependent protein kinase I (cGKI) to mediate relaxation by several intracellular mechanisms including the inhibition of IP3-dependent Ca 2+ release via the IP3-receptor associated cGMPkinase substrate (IRAG), the stimulation of sarco(endo)plasmatic reticulum ATPase, and the activation of myosin light chain phosphatase (Hofmann et al., 2006; Somlyo and Somlyo, 2003). The essential role of NO/cGMP signaling in the gut is supported by studies on mice that exhibit genetically-induced defects in this signaling pathway. For example, intestinal passage time and/or gastric emptying is impaired in mice lacking neuronal NO synthase (nNOS) (Mashimo et al., 2000), sGC (Friebe et al., 2007), cGKI (Ny et al., 2000; Pfeifer et al., 1998), or an intact IRAG (Geiselhoringer et al., 2004b). In addition, NANC-mediated relaxation of gastric muscle is hampered in nNOS−/− mice (Dick et al., 2002), in sGC−/− mice (Groneberg et al., 2009), and in cGKI−/− mice (Ny et al., 2000). ⁎ Corresponding author at: Institut für Pharmakologie und Toxikologie, TU München, Biedersteiner Str. 29, 80802 München, Germany. Tel.: +49 89 4140 3389; fax: +49 89 4140 3250. E-mail address: [email protected] (J.W. Wegener). 0014-2999/$ – see front matter © 2011 Published by Elsevier B.V. doi:10.1016/j.ejphar.2011.07.048
The main mediator of NO/cGMP signaling in smooth muscle is cGKI (Hofmann, 2005). cGKI exists in two isoforms, Iα and Iβ, that originate from one gene and differ only in their individual N-termini (Wernet et al., 1989). Both isoforms are supposed to relax smooth muscle through different molecular targets and mechanisms. For example, cGKIα interacts specifically with the myosin-interacting subunit of myosin phosphatase 1 (MYPT-1) (Surks et al., 1999) and/or the regulator of G-protein signaling 2 (RGS-2) (Tang et al., 2003), whereas cGKIβ is associated with IRAG and IP3-mediated Ca2+ release (Geiselhoringer et al., 2004b). However, recent evidence reconsidered the in vivo specificity of the isoforms using mice that selectively express cGKIα or cGKIβ in smooth muscle (Weber et al., 2007). In these mice, intestinal passage time as well as cGMP-mediated relaxation of vascular tone was completely restored (Weber et al., 2007) indicating that both isoforms can functionally compensate for each other. In the present study, we tested this new concept by analyzing the kinetics of relaxation by cGMP/cGKI signaling in murine fundus muscle from mice that selectively express cGKIα or cGKIβ. Relaxation was induced either by exogenously applied 8-Br-cGMP or by endogenously-derived cGMP which was expected to be produced upon activation of NANC neurons via electrical field stimulation (EFS). Onset of relaxation was analyzed during constant application of 8-Br-cGMP. Recovery from relaxation was analyzed after the transient EFS-induced relaxation. The results indicate that each isoform of cGKI mediates relaxation by a kinetically distinct pathway that, however, achieves a functionally equivalent level of relaxation in precontracted fundus muscle.
C. Ertl et al. / European Journal of Pharmacology 670 (2011) 266–271
2. Material and methods 2.1. Experimental preparation All experiments complied with the European guidelines for the use of experimental animals and were approved by the local animal ethics committee of TU München. Wild type (WT) mice, mice lacking cGMP-dependent protein kinase I (cGKI −/− mice) or IRAG (IRAG −/− mice), and mice expressing selectively cGKIα or cGKIβ isoforms in smooth muscle (sm-cGKIα or sm-cGKIβ mice) were generated as described previously (Desch et al., 2010; Geiselhoringer et al., 2004b; Weber et al., 2007). Mice of either sex were used at the age of 3 to 6 months. Mice were sacrificed by decapitation; the stomach was quickly transferred to buffer solution (in mM: NaCl 137, KCl 5.4, CaCl2 1.8, MgCl2 1, NaHCO3 12, NaH2PO4 0.42, glucose 5.6) bubbled with carbogen (95% O2, 5% CO2). Longitudinal strips were prepared from fundus muscle and cleaned from connective tissue. Muscle strips were mounted into organ baths (Myograph 601, www.dmt. dk). Tension was recorded isometrically at 37 ± 1 °C. All muscles used showed spontaneous contractile activity which, however, often disappeared during the time course of the experiments. Preload tension was individually applied to give the maximal amplitude of spontaneous contractions. 2.2. Tension recordings Electrical field stimulation (EFS) was performed as trans-mural stimulation of muscle strips with trains of pulses at 10 Hz (pulse strength 40 V, pulse width 1 ms, train duration 10 s) using a S48 Grass stimulator (www.grasstechnologies.com). This protocol produced maximal relaxation on carbachol-contracted fundus muscles in control experiments. Train interval was at least 10 min which allowed recovery of the response to EFS in control conditions as reported previously (Yano et al., 1995). Muscles used showed contractions in response to EFS; these contractions were blocked by atropine in control experiments. For analysis of EFS-induced relaxations, muscles were pre-contracted with carbachol (10 μM) to exclude the influence of cholinergic neurons on muscle tone. About 80% of all fundus muscles responded with relaxation to the EFS under these conditions. NO-mediated relaxations were characterized by their sensitivity to the inhibitor of NO synthase, NG-nitro-L-arginine methyl ester (L-NAME; 100 μM).
267
concentrations as indicated. 8-Br-cGMP (www.biolog.de) was used at 300 μM which produced maximal relaxation of carbacholcontracted intestinal muscle (Weber et al., 2007). 2.5. Analysis Results are presented as blots, original recordings or expressed as means ± S.E.M. Effects of substances were analyzed in steady-state conditions. Changes in tension were determined with respect to the maximum and the baseline. Effects of 8-Br-cGMP were calculated as the difference between tonic tension induced by carbachol (10 μM) and resting tension in the presence of 3-Isobutyl-1-methyl-xanthine (IBMX; 100 μM). Time courses were fitted with 1- or 2-exponential functions with R N 0.9 using Prism 4 software (www.graphpad.com). Statistical comparisons of data sets were performed by a Student's t-test or by ANOVA followed by Bonferroni post hoc test using Prism 4. Differences were considered significant at P b 0.05. 3. Results 3.1. Analysis of cGKI isoforms in fundus muscle Expression of cGKI and IRAG was analyzed by Western blot analysis in fundus protein lysates from sm-cGKIα, sm-cGKIβ, and wild type mice. Lysates from cGKI −/− and IRAG −/− mice served as controls to demonstrate the specificity of the used antibodies. Indeed, cGKI and IRAG were not detectable in the lysates from cGKI −/− and IRAG −/− mice, respectively (Fig. 1A). By using isoform specific antibodies, cGKIα was not detected in the lysates from cGKIβ, whereas cGKIβ was absent in the lysates from cGKIα, confirming that each mouse line expressed selectively the respective isoform (Fig. 1A). The expression of IRAG was unchanged in fundus from wild type (WT), sm-cGKIα and sm-cGKIβ mice, indicating that selective expression of the cGKI isoforms did not change expression levels of cGKI
2.3. Western blotting For western blot analysis, fundus muscles were frozen on liquid nitrogen. Homogenization was done in lysis buffer (2% SDS and 50 mM Tris–HCl; pH 7) using a Fastprep device (www.MPbioscience.com). The protein homogenates were boiled 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% non-fat milk powder in TRIS-buffered saline and labeled with specific primary antibodies followed by secondary antibodies conjugated to horseradish peroxidase or alkaline phosphatase. Primary antibodies against cGKI, cGKIα, cGKIβ, and IRAG were used as described previously (Geiselhoringer et al., 2004a). The signal obtained with an anti-β-actin antibody (1:5000; Novus Biologicals) was used as loading control. 2.4. Substances All salts and substances were used as pure as commercially available and purchased from Sigma (www.sigmaaldrich.com). Substances were applied as single dose or cumulatively to achieve the
Fig. 1. Biochemical and functional analyses of fundus muscle from wild type (WT), cGKI−/−, IRAG−/−, sm-cGKIα, and sm-cGKIβ mice. (A) Western blot analysis. Western blots of fundus muscle from WT, cGKI−/−, and IRAG−/−mice were decorated with anticGKI and anti-IRAG antibodies. Western blots from WT, sm-cGKIα (sm-Iα), and smcGKIβ (sm-Iβ) mice were decorated with anti-IRAG, anti-cGKIα and cGKIβ antibodies. β-Actin served as loading control. Similar results were obtained in 3 independent experiments. (B) Analysis of carbachol-induced contraction. Data represent magnitudes of peak tension induced by carbachol (10 μM) in fundus muscle from the indicated mouse strain. Columns represent means ± S.E.M. Numbers represent the number of experiments. Data were analyzed using ANOVA followed by Bonferroni post hoc test which indicated no statistically significant differences between WT and the respective mouse strain.
268
C. Ertl et al. / European Journal of Pharmacology 670 (2011) 266–271
Fig. 2. Relaxation of carbachol-contracted fundus muscle by 8-Br-cGMP. (A, B) Original recordings of tension in fundus muscle from WT (A) and cGKI−/− (B) mice. Bars indicate the presence of CCh (10 μM), 8-Br-cGMP (300 μM), and IBMX (100 μM). IBMX was used to determine the maximum of relaxation. (C) Magnitudes of carbachol-induced tension in the presence of 8-Br-cGMP in muscles from the indicated mouse strain. Columns represent means ± S.E.M. Numbers represent the number of experiments. Data were analyzed using ANOVA followed by Bonferroni post hoc test. Differences were calculated between WT mice and the respective mouse strain. ***, P b 0.001; n.s., non-significant. (D) Time course of cGMP-induced relaxation. Data points correspond to mean values of normalized time courses that were obtained in muscles from WT mice (n = 15). The white lines correspond to mono-exponential functions that were calculated from the data sets obtained in muscles from WT, sm-cGKIα, and sm-cGKIβ mice.
substrates. A quantitative analysis of the Western blots is presented in Supplemental Fig. 1. These results show that cGKIα and cGKIβ are selectively expressed in the absence of the respective other isoform in fundus from the sm-cGKIα and cGKIβ mouse lines.
that cGKIα or cGKIβ are similarly effective in relaxing fundus muscle if the enzymes are constantly activated by extracellularly applied 8Br-cGMP. Wash-out of 8-Br-cGMP did not re-establish contraction within 4 h prohibiting analysis of recovery from this type of relaxation.
3.2. Relaxation of fundus muscle by 8-Br-cGMP 3.3. Relaxation of fundus by electrical field stimulation (EFS) Stimulation of muscarinic receptors by carbachol similarly contracted fundus muscles from all mouse strains (Fig. 1B). 8-Br-cGMP at a maximal effective concentration (300 μM) relaxed carbacholinduced contractions in muscles from all mouse strains except from cGKI −/− mice (Fig. 2). The magnitude of relaxation was not different between muscles from sm-cGKIα, sm-cGKIβ and WT mice (Fig. 2C). 8-Br-cGMP induced a small relaxation in fundus from IRAG −/− mice, as reported for intestinal muscle from IRAGΔ12/Δ12 and IRAG−/− mice (Desch et al., 2010; Geiselhoringer et al., 2004b). Time courses of relaxation were calculated for sm-cGKIα, smcGKIβ and WT mice, since the magnitude of relaxation was not different between these mice strains. The analysis showed that the time courses were not different (Fig. 2D, Table 1). These findings indicate
Table 1 Time constants of relaxation by 8-Br-cGMP. t1/2 WT cGKI−/− sm-cGKIα sm-cGKIβ
2.1 ± 0.2 min i.d. 2.4 ± 0.2 min 1.9 ± 0.3 min
(n = 23) (n = 19) (n = 13)n.s. (n = 8)n.s.
All values are expressed as mean ± S.E.M. Numbers represent the number of experiments. Differences were analyzed between WT and the respective mouse strains. i.d. indeterminable; n.s. nonsignificant.
Short time application of EFS (10 s) induced transient relaxations in fundus muscles that were pre-contracted with carbachol (10 μM). As reported, the relaxations consisted of a fast and slow component (Lefebvre et al., 1992). The fast relaxation component was inhibited by L-NAME (100 μM, 10 min) and not observed in muscles from cGKI −/− mice. Therefore, this component was considered to be dependent on endogenous cGMP signaling resulting from NO that has been released by EFS-stimulated NANC neurons. For analysis, EFS-induced relaxations under control conditions were subtracted from the EFS-induced relaxations in the presence of L-NAME. The remaining relaxation was assumed to represent the relaxation by NO/cGMP signaling. The calculated magnitudes of L-NAME-inhibited relaxations were not different between smcGKIα, sm-cGKIβ, IRAG −/− and WT mice (Fig. 3D). Time course of the onset of relaxation was not analyzed since multiple steps are involved in this process. Instead, time courses of the recovery from relaxation were analyzed for sm-cGKIα, sm-cGKIβ, IRAG −/− and WT mice. Recovery from relaxation was not different between sm-cGKIβ and WT mice, whereas recovery from relaxation was about 4-times slower in muscles from sm-cGKIα than from WT mice (Fig. 3E; Table 2). The different time courses of recovery from relaxation in muscles from sm-cGKIα and sm-cGKIβ mice indicate that different signaling pathways are involved in the effects of either isoform. IRAG has been reported as a preferred substrate for cGKIβ that mediates cGMP-dependent inhibition of intracellular Ca 2+-release
C. Ertl et al. / European Journal of Pharmacology 670 (2011) 266–271
269
Fig. 3. EFS-induced relaxation of fundus muscle from WT, sm-cGKIβ, and sm-cGKIα mice. (A, B, C) Original recordings of EFS-induced relaxations in fundus muscles from WT (A), smcGKIβ (sm-Iβ, B), and sm-cGKIα (sm-Iα, C), mice. Muscles were pre-contracted by carbachol (CCh). Arrows indicate the EFS. Bars indicate the presence of L-NAME (100 μM). Carbachol (10 μM) was present throughout the experiment. Lines indicate fits of the data points with one- (right) or two- (left) exponential functions. (D) Magnitudes of L-NAME-inhibited, EFS-induced relaxations. Magnitudes were obtained from the exponential curves describing recovery from relaxation in each experiment. Columns represent means ± S.E.M. Numbers represent the number of experiments. No L-NAME-inhibited relaxations were observed in muscles from cGKI−/− mice. Differences were calculated between data from WT and the respective mouse strain using ANOVA followed by Bonferroni post hoc test. *, P b 0.05; **, P b 0.01; ***, P b 0.001; n.s., non-significant. (E) Recovery from L-NAME-inhibited relaxation. LNAME-resistant relaxations were subtracted from the values. Data points correspond to mean values of normalized time courses that were obtained in muscles from sm-cGKIα and sm-cGKIβ mice. White lines correspond to mono-exponential functions that describe the data sets obtained in muscles from sm-cGKIβ (solid line), sm-cGKIα (solid line), and WT (dashed line) mice. The line corresponding to the data set obtained in muscles from IRAG−/− mice was congruent to that obtained in muscles from sm-cGKIα mice.
(Schlossmann et al., 2000). Recovery from relaxation was about 3times slower in muscles from IRAG −/− mice than in WT or smcGKIβ mice (Fig. 3E; Table 2). This finding suggests that the recovery from relaxation in IRAG −/− mice represents a signaling pathway being not related to intracellular Ca 2+-release.
Table 2 Time constants of recovery from L-NAME-inhibited EFS-induced relaxation. t1/2 WT WT (+L-NAME) cGKI−/− sm-cGKIα sm-cGKIβ IRAG−/−
0.10 ± 0.03 min (n = 23) i.d. (n = 23) i.d. (n = 19) 0.44 ± 0.03 min (n = 13)a,b 0.08 ± 0.02 min (n = 15)n.s. 0.31 ± 0.12 min (n = 6)c,b
All values are expressed as mean ± S.E.M. Numbers represent the number of experiments. Differences were analyzed between WT and the respective mouse strains. i.d. indeterminable; n.s. non-significant; bno significant difference between sm-cGKIα and IRAG−/− mice; c, P b 0.05; a, P b 0.001.
4. Discussion It has been postulated that the Iα and Iβ isoforms of cGKI mediate relaxation of smooth muscle via different intracellular mechanisms including activation of myosin phosphatase activity and inhibition of intracellular Ca 2+ release (Schlossmann et al., 2000; Surks et al., 1999). Recent evidence indicates that both isoforms can functionally substitute for each other in vivo (Weber et al., 2007). The present study confirms that such a functional compensation exists in fundus muscle since a functionally equivalent level of relaxation was achieved by both, extracellularly-applied 8-Br-cGMP and by endogenously-derived cGMP via EFS-stimulation of NANC neurons. However, differences in the time course of recovery from relaxation were observed after EFS indicating that, indeed, different signaling pathways may account for the effects of each isoform. Two kinetic models of relaxation were studied using fundus muscle. In the first model, onset of relaxation was analyzed during constant activation of cGMP/cGKI signaling. Constantly applied 8-Br-cGMP similarly relaxed pre-contracted fundus muscles from WT, sm-cGKIα, and
270
C. Ertl et al. / European Journal of Pharmacology 670 (2011) 266–271
sm-cGKIβ mice. Muscles from cGKI−/− mice were not relaxed by 8-BrcGMP indicating that (i) cGKI is essential for the effects of 8-Br-cGMP, and (ii) cAMP-dependent protein kinase does not participate in the effects of 8-Br-cGMP in intact fundus muscle, in contrast to permeabilized aorta (Worner et al., 2007). It was expected that the time course of relaxation is faster in sm-cGKIα mice as compared to smcGKIβ mice since (i) the Ka-value for 8-Br-cGMP is about 10-fold lower for cGKIα than for cGKIβ in vitro (Ruth et al., 1991; Wolfe et al., 1989) and (ii) fundus from sm-cGKIα mice showed a pronounced level of cGKIα compared to WT mice. However, the time course of relaxation by 8-Br-cGMP was not different between the muscles from WT, sm-cGKIα, and sm-cGKIβ mice. These results suggest that the Kavalues as well as the concentration of cGKI are not rate-determining steps in the process of relaxation by 8-Br-cGMP which is mainly determined by changes in myosin light chain (MLC) phosphorylation (Ozaki et al., 1991). In the second model, kinetic of relaxation was analyzed after short-time activation of endogenous cGMP/cGKI signaling. Short time application of EFS induces the release of several neurotransmitters in the gut which produce relaxation (Costall et al., 1983; Li and Rand, 1990). Here, we focused on L-NAME-inhibited relaxations which are commonly mediated by the release of NO through NANC neurons (Sanders and Ward, 1992). The magnitudes of L-NAME-inhibited, EFS-mediated relaxations were not different in the muscles from the mouse strains investigated. This finding may explain why so far no obvious differences were detected in the phenotypes of sm-cGKIα and sm-cGKIβ mice (Weber et al., 2007). The maximum of EFS-induced relaxation was reached within 20 s which was not statistically different in muscles from sm-cGKIα and sm-cGKIβ mice. This “onset” of relaxation involves several distinct reactions (e.g. generation of NO, activation of sGC, binding of cGMP to cGKI, phosphorylation of target proteins like myosin phosphatase (MP), or dephosphorylation of MLC) which are, however, performed within this time scale. For example, short time application of NO transiently increased cGMP levels in intestinal smooth muscle (Boyer et al., 1998). The peak of the cGMP level was reached within 5 s after stimulation and returned to basal levels within 20 s. Further, association constants for cGMP to cGKI are very fast (b1 s) and cGMP dissociates rapidly from both cGKI isoforms after binding (Hofmann et al., 1985). Dissociation from the high affinity binding site was complete for both isoforms within 20 s at 4 °C (Ruth et al., 1991; Wolfe et al., 1989). In addition, NO-induced phosphorylation of cGKI target proteins, like the IP3 receptor or MP, is already detected within a time interval of 30 s during constant stimulation (Komalavilas and Lincoln, 1996; Lee et al., 1997). Moreover, NO donors induce fast dephosphorylation of MLC and CPI-17 within 30 s and a concomitant increase in MYPT1 phosphorylation (Johnson and Lincoln, 1985; Kitazawa et al., 2009). Since the “onset” of relaxation involves multiple reactions, as outlined above, we focused on the analysis of the recovery from relaxation. The recovery started at about 20 s after EFS at which cGMP levels and, probably, cGKI activity should be at rest (see above). Recovery was best-described by a one-exponential function indicating that only one process is involved, namely the re-phosphorylation of MLC which re-introduce contraction in fundus muscle. Indeed, changes in tension closely paralleled changes in MLC phosphorylation (Ozaki et al., 1991). Our results showed that recovery from relaxation was delayed in sm-cGKIα as compared to WT or sm-cGKIβ mice. This result can be best explained in a model that involves different signaling pathways for each cGKI isoform. In a simplified view, cGKIβ interferes with Ca 2+-dependent contractions, i.e. inhibition of IP3-dependent Ca 2+ release via IRAG (Geiselhoringer et al., 2004b), whereas cGKIα interferes with Ca 2+-independent contractions, i.e. stimulation of myosin phosphatase via MYPT-1 (Surks et al., 1999). Thus, the different time courses of recovery from relaxation may represent re-contraction via the Ca 2+-dependent pathway
in the sm-cGKIβ mice and re-contraction via the Ca 2+-independent pathway in the sm-cGKIα mice. This view is supported by the findings that the time course for Ca 2+-dependent MLC phosphorylation is faster than for Ca 2+-independent MLC phosphorylation. For example, Ca 2+-induced MLC phosphorylation was complete within 20 s in intact aorta (Kitajima et al., 1996) and within 1 min in permeabilized arteries (Moreland et al., 1992). In contrast, complete MLC phosphorylation in Ca 2+-independent conditions was about 3-fold prolonged, i.e. 3 min in permeabilized vascular smooth muscle (Niiro et al., 2003). Thus, the different kinetics of recovery from NANC-mediated relaxation in muscles from sm-cGKIα and sm-cGKIβ mice support the concept that both isoforms of cGKI acted on Ca 2+-dependent and Ca 2+-independent pathways, respectively, to produce relaxation. Additional evidence for this concept comes from the experiments performed on fundus muscle from IRAG −/− mice. IRAG has been identified as a substrate for cGKIβ and controls IP3-dependent Ca 2+ release (Schlossmann et al., 2000). Recovery from EFS-induced relaxation was 3-fold slower in fundus from IRAG −/− mice than in muscles from WT or sm-cGKIβ mice. This finding supports the view that this process reflects re-contraction via the Ca 2+-independent pathway. Since the recovery from relaxation was slow in sm-cGKIα mice, although IRAG was present, we conclude that cGKIα may not interfere with IRAG and Ca 2+-dependent processes, at least upon fast NANCmediated relaxation. However, we cannot exclude the possibility that cGKIα and cGKIβ may induce NO-mediated relaxation via still uncharacterized pathways. In summary, the results suggest that the Iα and Iβ isoforms of cGKI mediate relaxation by different cellular pathways which, however, are about equally effective in a physiological setting, i.e. NANCmediated relaxation. These findings may explain the absence of differences in the phenotypes of sm-cGKIα and sm-cGKIβ mice described previously (Weber et al., 2007). Supplementary materials related to this article can be found online at doi:10.1016/j.ejphar.2011.07.048.
Acknowledgment We thank Dr. Maria Huster for technical assistance. The work was supported by the Deutsche Forschungsgemeinschaft.
References Boyer, J.C., Christen, M.O., Balmes, J.L., Bali, J.P., 1998. Calcitonin gene-related peptideinduced relaxation of isolated human colonic smooth muscle cells through different intracellular pathways. Biochem. Pharmacol. 56, 1097–1104. Costall, B., Naylor, R.J., Tan, C.C., 1983. Field stimulation-induced responses of circular smooth muscle from guinea-pig stomach. Characterisation of the receptor mechanisms involved. Naunyn Schmiedebergs Arch. Pharmacol. 323, 155–161. Desch, M., Sigl, K., Hieke, B., Salb, K., Kees, F., Bernhard, D., Jochim, A., Spiessberger, B., Hocherl, K., Feil, R., Feil, S., Lukowski, R., Wegener, J.W., Hofmann, F., Schlossmann, J., 2010. IRAG determines nitric oxide- and atrial natriuretic peptide-mediated smooth muscle relaxation. Cardiovasc. Res. 86, 496–505. Dick, J.M., Van Molle, W., Brouckaert, P., Lefebvre, R.A., 2002. Relaxation by vasoactive intestinal polypeptide in the gastric fundus of nitric oxide synthase-deficient mice. J. Physiol. 538, 133–143. Friebe, A., Mergia, E., Dangel, O., Lange, A., Koesling, D., 2007. Fatal gastrointestinal obstruction and hypertension in mice lacking nitric oxide-sensitive guanylyl cyclase. Proc. Natl. Acad. Sci. U. S. A. 104, 7699–7704. Geiselhoringer, A., Gaisa, M., Hofmann, F., Schlossmann, J., 2004a. Distribution of IRAG and cGKI-isoforms in murine tissues. FEBS Lett. 575, 19–22. Geiselhoringer, A., Werner, M., Sigl, K., Smital, P., Worner, R., Acheo, L., Stieber, J., Weinmeister, P., Feil, R., Feil, S., Wegener, J., Hofmann, F., Schlossmann, J., 2004b. IRAG is essential for relaxation of receptor-triggered smooth muscle contraction by cGMP kinase. EMBO J. 23, 4222–4231. Groneberg, D., König, P., Koesling, D., Friebe, A., 2009. The nitric oxide receptor guanylyl cyclase in gastrointestinal smooth muscle is dispensable for gastrointestinal motility and survival in mice. Naunyn Schmiedebergs Arch. Pharmacol. 379 (Suppl. 1), 20. Hofmann, F., 2005. The biology of cyclic GMP-dependent protein kinases. J. Biol. Chem. 280, 1–4.
C. Ertl et al. / European Journal of Pharmacology 670 (2011) 266–271 Hofmann, F., Gensheimer, H.P., Gobel, C., 1985. cGMP-dependent protein kinase. Autophosphorylation changes the characteristics of binding site 1. Eur. J. Biochem. 147, 361–365. Hofmann, F., Feil, R., Kleppisch, T., Schlossmann, J., 2006. Function of cGMP-dependent protein kinases as revealed by gene deletion. Physiol. Rev. 86, 1–23. Johnson, R.M., Lincoln, T.M., 1985. Effects of nitroprusside, glyceryl trinitrate, and 8bromo cyclic GMP on phosphorylase a formation and myosin light chain phosphorylation in rat aorta. Mol. Pharmacol. 27, 333–342. Kitajima, S., Harada, K., Hori, M., Ozaki, H., Karaki, H., 1996. Dissociation of P2 purinoceptor-mediated increase in intracellular Ca2+ level from myosin light chain phosphorylation and contraction in rat aorta. Br. J. Pharmacol. 118, 543–548. Kitazawa, T., Semba, S., Huh, Y.H., Kitazawa, K., Eto, M., 2009. Nitric oxide-induced biphasic mechanism of vascular relaxation via dephosphorylation of CPI-17 and MYPT1. J. Physiol. 587, 3587–3603. Komalavilas, P., Lincoln, T.M., 1996. Phosphorylation of the inositol 1,4,5-trisphosphate receptor. Cyclic GMP-dependent protein kinase mediates cAMP and cGMP dependent phosphorylation in the intact rat aorta. J. Biol. Chem. 271, 21933–21938. Lee, M.R., Li, L., Kitazawa, T., 1997. Cyclic GMP causes Ca2+ desensitization in vascular smooth muscle by activating the myosin light chain phosphatase. J. Biol. Chem. 272, 5063–5068. Lefebvre, R.A., Baert, E., Barbier, A.J., 1992. Influence of NG-nitro-L-arginine on nonadrenergic non-cholinergic relaxation in the guinea-pig gastric fundus. Br. J. Pharmacol. 106, 173–179. Li, C.G., Rand, M.J., 1990. Nitric oxide and vasoactive intestinal polypeptide mediate non-adrenergic, non-cholinergic inhibitory transmission to smooth muscle of the rat gastric fundus. Eur. J. Pharmacol. 191, 303–309. Mashimo, H., Kjellin, A., Goyal, R.K., 2000. Gastric stasis in neuronal nitric oxide synthase-deficient knockout mice. Gastroenterology 119, 766–773. Moreland, S., Nishimura, J., van Breemen, C., Ahn, H.Y., Moreland, R.S., 1992. Transient myosin phosphorylation at constant Ca2+ during agonist activation of permeabilized arteries. Am. J. Physiol. 263, C540–C544. Niiro, N., Koga, Y., Ikebe, M., 2003. Agonist-induced changes in the phosphorylation of the myosin- binding subunit of myosin light chain phosphatase and CPI17, two regulatory factors of myosin light chain phosphatase, in smooth muscle. Biochem. J. 369, 117–128. Ny, L., Pfeifer, A., Aszodi, A., Ahmad, M., Alm, P., Hedlund, P., Fassler, R., Andersson, K.E., 2000. Impaired relaxation of stomach smooth muscle in mice lacking cyclic GMPdependent protein kinase I. Br. J. Pharmacol. 129, 395–401. Ozaki, H., Gerthoffer, W.T., Publicover, N.G., Fusetani, N., Sanders, K.M., 1991. Timedependent changes in Ca2+ sensitivity during phasic contraction of canine antral smooth muscle. J. Physiol. 440, 207–224.
271
Pfeifer, A., Klatt, P., Massberg, S., Ny, L., Sausbier, M., Hirneiss, C., Wang, G.X., Korth, M., Aszodi, A., Andersson, K.E., Krombach, F., Mayerhofer, A., Ruth, P., Fassler, R., Hofmann, F., 1998. Defective smooth muscle regulation in cGMP kinase I-deficient mice. EMBO J. 17, 3045–3051. Ruth, P., Landgraf, W., Keilbach, A., May, B., Egleme, C., Hofmann, F., 1991. The activation of expressed cGMP-dependent protein kinase isozymes I alpha and I beta is determined by the different amino-termini. Eur. J. Biochem. 202, 1339–1344. Sanders, K.M., Ward, S.M., 1992. Nitric oxide as a mediator of nonadrenergic noncholinergic neurotransmission. Am. J. Physiol. 262, G379–G392. Schlossmann, J., Ammendola, A., Ashman, K., Zong, X., Huber, A., Neubauer, G., Wang, G.X., Allescher, H.D., Korth, M., Wilm, M., Hofmann, F., Ruth, P., 2000. Regulation of intracellular calcium by a signalling complex of IRAG, IP3 receptor and cGMP kinase Ibeta. Nature 404, 197–201. Shah, V., Lyford, G., Gores, G., Farrugia, G., 2004. Nitric oxide in gastrointestinal health and disease. Gastroenterology 126, 903–913. Somlyo, A.P., Somlyo, A.V., 2003. Ca2+ sensitivity of smooth muscle and nonmuscle myosin II: modulated by G proteins, kinases, and myosin phosphatase. Physiol. Rev. 83, 1325–1358. Surks, H.K., Mochizuki, N., Kasai, Y., Georgescu, S.P., Tang, K.M., Ito, M., Lincoln, T.M., Mendelsohn, M.E., 1999. Regulation of myosin phosphatase by a specific interaction with cGMP- dependent protein kinase Ialpha. Science 286, 1583–1587. Tang, K.M., Wang, G.R., Lu, P., Karas, R.H., Aronovitz, M., Heximer, S.P., Kaltenbronn, K.M., Blumer, K.J., Siderovski, D.P., Zhu, Y., Mendelsohn, M.E., 2003. Regulator of G-protein signaling-2 mediates vascular smooth muscle relaxation and blood pressure. Nat. Med. 9, 1506–1512. Toda, N., Herman, A.G., 2005. Gastrointestinal function regulation by nitrergic efferent nerves. Pharmacol. Rev. 57, 315–338. Weber, S., Bernhard, D., Lukowski, R., Weinmeister, P., Worner, R., Wegener, J.W., Valtcheva, N., Feil, S., Schlossmann, J., Hofmann, F., Feil, R., 2007. Rescue of cGMP kinase I knockout mice by smooth muscle specific expression of either isozyme. Circ. Res. 101, 1096–1103. Wernet, W., Flockerzi, V., Hofmann, F., 1989. The cDNA of the two isoforms of bovine cGMP-dependent protein kinase. FEBS Lett. 251, 191–196. Wolfe, L., Corbin, J.D., Francis, S.H., 1989. Characterization of a novel isozyme of cGMPdependent protein kinase from bovine aorta. J. Biol. Chem. 264, 7734–7741. Worner, R., Lukowski, R., Hofmann, F., Wegener, J.W., 2007. cGMP signals mainly through cAMP kinase in permeabilized murine aorta. Am. J. Physiol. Heart Circ. Physiol. 292, H237–H244. Yano, S., Kiyota, Y., Yamamoto, M., Watanabe, K., 1995. Pharmacological features of non-adrenergic non-cholinergic (NANC) relaxation induced by electrical vagal stimulation in isolated mouse stomach. Jpn. J. Pharmacol. 69, 9–15.
Contents lists available at SciVerse ScienceDirect
European Journal of Pharmacology journal homepage: www.elsevier.com/locate/ejphar
Pulmonary, Gastrointestinal and Urogenital Pharmacology
Kinetics of relaxation by cGMP/cGKI signaling in fundus smooth muscle Claudia Ertl a, Robert Lukowski a, b, Katja Sigl a, Jens Schlossmann a, c, Franz Hofmann a, Jörg W. Wegener a,⁎ a b c
FOR923, Institut für Pharmakologie und Toxikologie, Technische Universität München, Germany Institut für Pharmazie, Pharmakologie, Toxikologie und Klinische Pharmazie, Universität Tübingen, Germany Institut für Pharmakologie und Toxikologie, Universität Regensburg, Germany
a r t i c l e
i n f o
Article history: Received 20 June 2011 Received in revised form 14 July 2011 Accepted 30 July 2011 Available online 19 August 2011 Keywords: Fundus Relaxation NANC Nitric oxide cGMP
a b s t r a c t cGMP-dependent kinase I (cGKI) is a major mediator of smooth muscle relaxation and exists in two isoforms, α and β. Both isoforms are supposed to mediate their effects via different intracellular signaling pathways. To verify this concept, the kinetics of relaxation mediated by either isoform was analyzed in gastric fundus smooth muscle from mice. Muscles from mice that express selectively the Iα or Iβ isoform of cGKI in smooth muscle (sm-cGKIα or sm-cGKIβ mice) were compared to muscles from conventional cGKI−/− mice. Fundus muscles were contracted by carbachol and then relaxed by 8-Br-cGMP or by electrical field stimulation (EFS). The time course of relaxation by 8-Br-cGMP was not different between muscles from sm-cGKIα and smcGKIβ mice. EFS induced a fast transient relaxation in muscles from sm-cGKIα and sm-cGKIβ mice that was blocked by the NO synthase inhibitor L-NAME. Recovery from this relaxation was about 4-times slower in muscles from sm-cGKIα mice than in muscles from sm-cGKIβ mice. The different kinetic of recovery from relaxation after EFS in sm-cGKIα and sm-cGKIβ mice suggests that different signaling pathways exist for each cGKI isoform in vivo in fundus muscles. © 2011 Published by Elsevier B.V.
1. Introduction Nitric oxide (NO) represents the major inhibitory neurotransmitter in the gastrointestinal tract (Shah et al., 2004; Toda and Herman, 2005). NO is released by non-adrenergic, non-cholinergic (NANC) neurons and causes relaxation of gastrointestinal smooth muscle by activating soluble guanylyl cyclase (sGC). sGC generates cGMP that activates cGMP-dependent protein kinase I (cGKI) to mediate relaxation by several intracellular mechanisms including the inhibition of IP3-dependent Ca 2+ release via the IP3-receptor associated cGMPkinase substrate (IRAG), the stimulation of sarco(endo)plasmatic reticulum ATPase, and the activation of myosin light chain phosphatase (Hofmann et al., 2006; Somlyo and Somlyo, 2003). The essential role of NO/cGMP signaling in the gut is supported by studies on mice that exhibit genetically-induced defects in this signaling pathway. For example, intestinal passage time and/or gastric emptying is impaired in mice lacking neuronal NO synthase (nNOS) (Mashimo et al., 2000), sGC (Friebe et al., 2007), cGKI (Ny et al., 2000; Pfeifer et al., 1998), or an intact IRAG (Geiselhoringer et al., 2004b). In addition, NANC-mediated relaxation of gastric muscle is hampered in nNOS−/− mice (Dick et al., 2002), in sGC−/− mice (Groneberg et al., 2009), and in cGKI−/− mice (Ny et al., 2000). ⁎ Corresponding author at: Institut für Pharmakologie und Toxikologie, TU München, Biedersteiner Str. 29, 80802 München, Germany. Tel.: +49 89 4140 3389; fax: +49 89 4140 3250. E-mail address: [email protected] (J.W. Wegener). 0014-2999/$ – see front matter © 2011 Published by Elsevier B.V. doi:10.1016/j.ejphar.2011.07.048
The main mediator of NO/cGMP signaling in smooth muscle is cGKI (Hofmann, 2005). cGKI exists in two isoforms, Iα and Iβ, that originate from one gene and differ only in their individual N-termini (Wernet et al., 1989). Both isoforms are supposed to relax smooth muscle through different molecular targets and mechanisms. For example, cGKIα interacts specifically with the myosin-interacting subunit of myosin phosphatase 1 (MYPT-1) (Surks et al., 1999) and/or the regulator of G-protein signaling 2 (RGS-2) (Tang et al., 2003), whereas cGKIβ is associated with IRAG and IP3-mediated Ca2+ release (Geiselhoringer et al., 2004b). However, recent evidence reconsidered the in vivo specificity of the isoforms using mice that selectively express cGKIα or cGKIβ in smooth muscle (Weber et al., 2007). In these mice, intestinal passage time as well as cGMP-mediated relaxation of vascular tone was completely restored (Weber et al., 2007) indicating that both isoforms can functionally compensate for each other. In the present study, we tested this new concept by analyzing the kinetics of relaxation by cGMP/cGKI signaling in murine fundus muscle from mice that selectively express cGKIα or cGKIβ. Relaxation was induced either by exogenously applied 8-Br-cGMP or by endogenously-derived cGMP which was expected to be produced upon activation of NANC neurons via electrical field stimulation (EFS). Onset of relaxation was analyzed during constant application of 8-Br-cGMP. Recovery from relaxation was analyzed after the transient EFS-induced relaxation. The results indicate that each isoform of cGKI mediates relaxation by a kinetically distinct pathway that, however, achieves a functionally equivalent level of relaxation in precontracted fundus muscle.
C. Ertl et al. / European Journal of Pharmacology 670 (2011) 266–271
2. Material and methods 2.1. Experimental preparation All experiments complied with the European guidelines for the use of experimental animals and were approved by the local animal ethics committee of TU München. Wild type (WT) mice, mice lacking cGMP-dependent protein kinase I (cGKI −/− mice) or IRAG (IRAG −/− mice), and mice expressing selectively cGKIα or cGKIβ isoforms in smooth muscle (sm-cGKIα or sm-cGKIβ mice) were generated as described previously (Desch et al., 2010; Geiselhoringer et al., 2004b; Weber et al., 2007). Mice of either sex were used at the age of 3 to 6 months. Mice were sacrificed by decapitation; the stomach was quickly transferred to buffer solution (in mM: NaCl 137, KCl 5.4, CaCl2 1.8, MgCl2 1, NaHCO3 12, NaH2PO4 0.42, glucose 5.6) bubbled with carbogen (95% O2, 5% CO2). Longitudinal strips were prepared from fundus muscle and cleaned from connective tissue. Muscle strips were mounted into organ baths (Myograph 601, www.dmt. dk). Tension was recorded isometrically at 37 ± 1 °C. All muscles used showed spontaneous contractile activity which, however, often disappeared during the time course of the experiments. Preload tension was individually applied to give the maximal amplitude of spontaneous contractions. 2.2. Tension recordings Electrical field stimulation (EFS) was performed as trans-mural stimulation of muscle strips with trains of pulses at 10 Hz (pulse strength 40 V, pulse width 1 ms, train duration 10 s) using a S48 Grass stimulator (www.grasstechnologies.com). This protocol produced maximal relaxation on carbachol-contracted fundus muscles in control experiments. Train interval was at least 10 min which allowed recovery of the response to EFS in control conditions as reported previously (Yano et al., 1995). Muscles used showed contractions in response to EFS; these contractions were blocked by atropine in control experiments. For analysis of EFS-induced relaxations, muscles were pre-contracted with carbachol (10 μM) to exclude the influence of cholinergic neurons on muscle tone. About 80% of all fundus muscles responded with relaxation to the EFS under these conditions. NO-mediated relaxations were characterized by their sensitivity to the inhibitor of NO synthase, NG-nitro-L-arginine methyl ester (L-NAME; 100 μM).
267
concentrations as indicated. 8-Br-cGMP (www.biolog.de) was used at 300 μM which produced maximal relaxation of carbacholcontracted intestinal muscle (Weber et al., 2007). 2.5. Analysis Results are presented as blots, original recordings or expressed as means ± S.E.M. Effects of substances were analyzed in steady-state conditions. Changes in tension were determined with respect to the maximum and the baseline. Effects of 8-Br-cGMP were calculated as the difference between tonic tension induced by carbachol (10 μM) and resting tension in the presence of 3-Isobutyl-1-methyl-xanthine (IBMX; 100 μM). Time courses were fitted with 1- or 2-exponential functions with R N 0.9 using Prism 4 software (www.graphpad.com). Statistical comparisons of data sets were performed by a Student's t-test or by ANOVA followed by Bonferroni post hoc test using Prism 4. Differences were considered significant at P b 0.05. 3. Results 3.1. Analysis of cGKI isoforms in fundus muscle Expression of cGKI and IRAG was analyzed by Western blot analysis in fundus protein lysates from sm-cGKIα, sm-cGKIβ, and wild type mice. Lysates from cGKI −/− and IRAG −/− mice served as controls to demonstrate the specificity of the used antibodies. Indeed, cGKI and IRAG were not detectable in the lysates from cGKI −/− and IRAG −/− mice, respectively (Fig. 1A). By using isoform specific antibodies, cGKIα was not detected in the lysates from cGKIβ, whereas cGKIβ was absent in the lysates from cGKIα, confirming that each mouse line expressed selectively the respective isoform (Fig. 1A). The expression of IRAG was unchanged in fundus from wild type (WT), sm-cGKIα and sm-cGKIβ mice, indicating that selective expression of the cGKI isoforms did not change expression levels of cGKI
2.3. Western blotting For western blot analysis, fundus muscles were frozen on liquid nitrogen. Homogenization was done in lysis buffer (2% SDS and 50 mM Tris–HCl; pH 7) using a Fastprep device (www.MPbioscience.com). The protein homogenates were boiled 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% non-fat milk powder in TRIS-buffered saline and labeled with specific primary antibodies followed by secondary antibodies conjugated to horseradish peroxidase or alkaline phosphatase. Primary antibodies against cGKI, cGKIα, cGKIβ, and IRAG were used as described previously (Geiselhoringer et al., 2004a). The signal obtained with an anti-β-actin antibody (1:5000; Novus Biologicals) was used as loading control. 2.4. Substances All salts and substances were used as pure as commercially available and purchased from Sigma (www.sigmaaldrich.com). Substances were applied as single dose or cumulatively to achieve the
Fig. 1. Biochemical and functional analyses of fundus muscle from wild type (WT), cGKI−/−, IRAG−/−, sm-cGKIα, and sm-cGKIβ mice. (A) Western blot analysis. Western blots of fundus muscle from WT, cGKI−/−, and IRAG−/−mice were decorated with anticGKI and anti-IRAG antibodies. Western blots from WT, sm-cGKIα (sm-Iα), and smcGKIβ (sm-Iβ) mice were decorated with anti-IRAG, anti-cGKIα and cGKIβ antibodies. β-Actin served as loading control. Similar results were obtained in 3 independent experiments. (B) Analysis of carbachol-induced contraction. Data represent magnitudes of peak tension induced by carbachol (10 μM) in fundus muscle from the indicated mouse strain. Columns represent means ± S.E.M. Numbers represent the number of experiments. Data were analyzed using ANOVA followed by Bonferroni post hoc test which indicated no statistically significant differences between WT and the respective mouse strain.
268
C. Ertl et al. / European Journal of Pharmacology 670 (2011) 266–271
Fig. 2. Relaxation of carbachol-contracted fundus muscle by 8-Br-cGMP. (A, B) Original recordings of tension in fundus muscle from WT (A) and cGKI−/− (B) mice. Bars indicate the presence of CCh (10 μM), 8-Br-cGMP (300 μM), and IBMX (100 μM). IBMX was used to determine the maximum of relaxation. (C) Magnitudes of carbachol-induced tension in the presence of 8-Br-cGMP in muscles from the indicated mouse strain. Columns represent means ± S.E.M. Numbers represent the number of experiments. Data were analyzed using ANOVA followed by Bonferroni post hoc test. Differences were calculated between WT mice and the respective mouse strain. ***, P b 0.001; n.s., non-significant. (D) Time course of cGMP-induced relaxation. Data points correspond to mean values of normalized time courses that were obtained in muscles from WT mice (n = 15). The white lines correspond to mono-exponential functions that were calculated from the data sets obtained in muscles from WT, sm-cGKIα, and sm-cGKIβ mice.
substrates. A quantitative analysis of the Western blots is presented in Supplemental Fig. 1. These results show that cGKIα and cGKIβ are selectively expressed in the absence of the respective other isoform in fundus from the sm-cGKIα and cGKIβ mouse lines.
that cGKIα or cGKIβ are similarly effective in relaxing fundus muscle if the enzymes are constantly activated by extracellularly applied 8Br-cGMP. Wash-out of 8-Br-cGMP did not re-establish contraction within 4 h prohibiting analysis of recovery from this type of relaxation.
3.2. Relaxation of fundus muscle by 8-Br-cGMP 3.3. Relaxation of fundus by electrical field stimulation (EFS) Stimulation of muscarinic receptors by carbachol similarly contracted fundus muscles from all mouse strains (Fig. 1B). 8-Br-cGMP at a maximal effective concentration (300 μM) relaxed carbacholinduced contractions in muscles from all mouse strains except from cGKI −/− mice (Fig. 2). The magnitude of relaxation was not different between muscles from sm-cGKIα, sm-cGKIβ and WT mice (Fig. 2C). 8-Br-cGMP induced a small relaxation in fundus from IRAG −/− mice, as reported for intestinal muscle from IRAGΔ12/Δ12 and IRAG−/− mice (Desch et al., 2010; Geiselhoringer et al., 2004b). Time courses of relaxation were calculated for sm-cGKIα, smcGKIβ and WT mice, since the magnitude of relaxation was not different between these mice strains. The analysis showed that the time courses were not different (Fig. 2D, Table 1). These findings indicate
Table 1 Time constants of relaxation by 8-Br-cGMP. t1/2 WT cGKI−/− sm-cGKIα sm-cGKIβ
2.1 ± 0.2 min i.d. 2.4 ± 0.2 min 1.9 ± 0.3 min
(n = 23) (n = 19) (n = 13)n.s. (n = 8)n.s.
All values are expressed as mean ± S.E.M. Numbers represent the number of experiments. Differences were analyzed between WT and the respective mouse strains. i.d. indeterminable; n.s. nonsignificant.
Short time application of EFS (10 s) induced transient relaxations in fundus muscles that were pre-contracted with carbachol (10 μM). As reported, the relaxations consisted of a fast and slow component (Lefebvre et al., 1992). The fast relaxation component was inhibited by L-NAME (100 μM, 10 min) and not observed in muscles from cGKI −/− mice. Therefore, this component was considered to be dependent on endogenous cGMP signaling resulting from NO that has been released by EFS-stimulated NANC neurons. For analysis, EFS-induced relaxations under control conditions were subtracted from the EFS-induced relaxations in the presence of L-NAME. The remaining relaxation was assumed to represent the relaxation by NO/cGMP signaling. The calculated magnitudes of L-NAME-inhibited relaxations were not different between smcGKIα, sm-cGKIβ, IRAG −/− and WT mice (Fig. 3D). Time course of the onset of relaxation was not analyzed since multiple steps are involved in this process. Instead, time courses of the recovery from relaxation were analyzed for sm-cGKIα, sm-cGKIβ, IRAG −/− and WT mice. Recovery from relaxation was not different between sm-cGKIβ and WT mice, whereas recovery from relaxation was about 4-times slower in muscles from sm-cGKIα than from WT mice (Fig. 3E; Table 2). The different time courses of recovery from relaxation in muscles from sm-cGKIα and sm-cGKIβ mice indicate that different signaling pathways are involved in the effects of either isoform. IRAG has been reported as a preferred substrate for cGKIβ that mediates cGMP-dependent inhibition of intracellular Ca 2+-release
C. Ertl et al. / European Journal of Pharmacology 670 (2011) 266–271
269
Fig. 3. EFS-induced relaxation of fundus muscle from WT, sm-cGKIβ, and sm-cGKIα mice. (A, B, C) Original recordings of EFS-induced relaxations in fundus muscles from WT (A), smcGKIβ (sm-Iβ, B), and sm-cGKIα (sm-Iα, C), mice. Muscles were pre-contracted by carbachol (CCh). Arrows indicate the EFS. Bars indicate the presence of L-NAME (100 μM). Carbachol (10 μM) was present throughout the experiment. Lines indicate fits of the data points with one- (right) or two- (left) exponential functions. (D) Magnitudes of L-NAME-inhibited, EFS-induced relaxations. Magnitudes were obtained from the exponential curves describing recovery from relaxation in each experiment. Columns represent means ± S.E.M. Numbers represent the number of experiments. No L-NAME-inhibited relaxations were observed in muscles from cGKI−/− mice. Differences were calculated between data from WT and the respective mouse strain using ANOVA followed by Bonferroni post hoc test. *, P b 0.05; **, P b 0.01; ***, P b 0.001; n.s., non-significant. (E) Recovery from L-NAME-inhibited relaxation. LNAME-resistant relaxations were subtracted from the values. Data points correspond to mean values of normalized time courses that were obtained in muscles from sm-cGKIα and sm-cGKIβ mice. White lines correspond to mono-exponential functions that describe the data sets obtained in muscles from sm-cGKIβ (solid line), sm-cGKIα (solid line), and WT (dashed line) mice. The line corresponding to the data set obtained in muscles from IRAG−/− mice was congruent to that obtained in muscles from sm-cGKIα mice.
(Schlossmann et al., 2000). Recovery from relaxation was about 3times slower in muscles from IRAG −/− mice than in WT or smcGKIβ mice (Fig. 3E; Table 2). This finding suggests that the recovery from relaxation in IRAG −/− mice represents a signaling pathway being not related to intracellular Ca 2+-release.
Table 2 Time constants of recovery from L-NAME-inhibited EFS-induced relaxation. t1/2 WT WT (+L-NAME) cGKI−/− sm-cGKIα sm-cGKIβ IRAG−/−
0.10 ± 0.03 min (n = 23) i.d. (n = 23) i.d. (n = 19) 0.44 ± 0.03 min (n = 13)a,b 0.08 ± 0.02 min (n = 15)n.s. 0.31 ± 0.12 min (n = 6)c,b
All values are expressed as mean ± S.E.M. Numbers represent the number of experiments. Differences were analyzed between WT and the respective mouse strains. i.d. indeterminable; n.s. non-significant; bno significant difference between sm-cGKIα and IRAG−/− mice; c, P b 0.05; a, P b 0.001.
4. Discussion It has been postulated that the Iα and Iβ isoforms of cGKI mediate relaxation of smooth muscle via different intracellular mechanisms including activation of myosin phosphatase activity and inhibition of intracellular Ca 2+ release (Schlossmann et al., 2000; Surks et al., 1999). Recent evidence indicates that both isoforms can functionally substitute for each other in vivo (Weber et al., 2007). The present study confirms that such a functional compensation exists in fundus muscle since a functionally equivalent level of relaxation was achieved by both, extracellularly-applied 8-Br-cGMP and by endogenously-derived cGMP via EFS-stimulation of NANC neurons. However, differences in the time course of recovery from relaxation were observed after EFS indicating that, indeed, different signaling pathways may account for the effects of each isoform. Two kinetic models of relaxation were studied using fundus muscle. In the first model, onset of relaxation was analyzed during constant activation of cGMP/cGKI signaling. Constantly applied 8-Br-cGMP similarly relaxed pre-contracted fundus muscles from WT, sm-cGKIα, and
270
C. Ertl et al. / European Journal of Pharmacology 670 (2011) 266–271
sm-cGKIβ mice. Muscles from cGKI−/− mice were not relaxed by 8-BrcGMP indicating that (i) cGKI is essential for the effects of 8-Br-cGMP, and (ii) cAMP-dependent protein kinase does not participate in the effects of 8-Br-cGMP in intact fundus muscle, in contrast to permeabilized aorta (Worner et al., 2007). It was expected that the time course of relaxation is faster in sm-cGKIα mice as compared to smcGKIβ mice since (i) the Ka-value for 8-Br-cGMP is about 10-fold lower for cGKIα than for cGKIβ in vitro (Ruth et al., 1991; Wolfe et al., 1989) and (ii) fundus from sm-cGKIα mice showed a pronounced level of cGKIα compared to WT mice. However, the time course of relaxation by 8-Br-cGMP was not different between the muscles from WT, sm-cGKIα, and sm-cGKIβ mice. These results suggest that the Kavalues as well as the concentration of cGKI are not rate-determining steps in the process of relaxation by 8-Br-cGMP which is mainly determined by changes in myosin light chain (MLC) phosphorylation (Ozaki et al., 1991). In the second model, kinetic of relaxation was analyzed after short-time activation of endogenous cGMP/cGKI signaling. Short time application of EFS induces the release of several neurotransmitters in the gut which produce relaxation (Costall et al., 1983; Li and Rand, 1990). Here, we focused on L-NAME-inhibited relaxations which are commonly mediated by the release of NO through NANC neurons (Sanders and Ward, 1992). The magnitudes of L-NAME-inhibited, EFS-mediated relaxations were not different in the muscles from the mouse strains investigated. This finding may explain why so far no obvious differences were detected in the phenotypes of sm-cGKIα and sm-cGKIβ mice (Weber et al., 2007). The maximum of EFS-induced relaxation was reached within 20 s which was not statistically different in muscles from sm-cGKIα and sm-cGKIβ mice. This “onset” of relaxation involves several distinct reactions (e.g. generation of NO, activation of sGC, binding of cGMP to cGKI, phosphorylation of target proteins like myosin phosphatase (MP), or dephosphorylation of MLC) which are, however, performed within this time scale. For example, short time application of NO transiently increased cGMP levels in intestinal smooth muscle (Boyer et al., 1998). The peak of the cGMP level was reached within 5 s after stimulation and returned to basal levels within 20 s. Further, association constants for cGMP to cGKI are very fast (b1 s) and cGMP dissociates rapidly from both cGKI isoforms after binding (Hofmann et al., 1985). Dissociation from the high affinity binding site was complete for both isoforms within 20 s at 4 °C (Ruth et al., 1991; Wolfe et al., 1989). In addition, NO-induced phosphorylation of cGKI target proteins, like the IP3 receptor or MP, is already detected within a time interval of 30 s during constant stimulation (Komalavilas and Lincoln, 1996; Lee et al., 1997). Moreover, NO donors induce fast dephosphorylation of MLC and CPI-17 within 30 s and a concomitant increase in MYPT1 phosphorylation (Johnson and Lincoln, 1985; Kitazawa et al., 2009). Since the “onset” of relaxation involves multiple reactions, as outlined above, we focused on the analysis of the recovery from relaxation. The recovery started at about 20 s after EFS at which cGMP levels and, probably, cGKI activity should be at rest (see above). Recovery was best-described by a one-exponential function indicating that only one process is involved, namely the re-phosphorylation of MLC which re-introduce contraction in fundus muscle. Indeed, changes in tension closely paralleled changes in MLC phosphorylation (Ozaki et al., 1991). Our results showed that recovery from relaxation was delayed in sm-cGKIα as compared to WT or sm-cGKIβ mice. This result can be best explained in a model that involves different signaling pathways for each cGKI isoform. In a simplified view, cGKIβ interferes with Ca 2+-dependent contractions, i.e. inhibition of IP3-dependent Ca 2+ release via IRAG (Geiselhoringer et al., 2004b), whereas cGKIα interferes with Ca 2+-independent contractions, i.e. stimulation of myosin phosphatase via MYPT-1 (Surks et al., 1999). Thus, the different time courses of recovery from relaxation may represent re-contraction via the Ca 2+-dependent pathway
in the sm-cGKIβ mice and re-contraction via the Ca 2+-independent pathway in the sm-cGKIα mice. This view is supported by the findings that the time course for Ca 2+-dependent MLC phosphorylation is faster than for Ca 2+-independent MLC phosphorylation. For example, Ca 2+-induced MLC phosphorylation was complete within 20 s in intact aorta (Kitajima et al., 1996) and within 1 min in permeabilized arteries (Moreland et al., 1992). In contrast, complete MLC phosphorylation in Ca 2+-independent conditions was about 3-fold prolonged, i.e. 3 min in permeabilized vascular smooth muscle (Niiro et al., 2003). Thus, the different kinetics of recovery from NANC-mediated relaxation in muscles from sm-cGKIα and sm-cGKIβ mice support the concept that both isoforms of cGKI acted on Ca 2+-dependent and Ca 2+-independent pathways, respectively, to produce relaxation. Additional evidence for this concept comes from the experiments performed on fundus muscle from IRAG −/− mice. IRAG has been identified as a substrate for cGKIβ and controls IP3-dependent Ca 2+ release (Schlossmann et al., 2000). Recovery from EFS-induced relaxation was 3-fold slower in fundus from IRAG −/− mice than in muscles from WT or sm-cGKIβ mice. This finding supports the view that this process reflects re-contraction via the Ca 2+-independent pathway. Since the recovery from relaxation was slow in sm-cGKIα mice, although IRAG was present, we conclude that cGKIα may not interfere with IRAG and Ca 2+-dependent processes, at least upon fast NANCmediated relaxation. However, we cannot exclude the possibility that cGKIα and cGKIβ may induce NO-mediated relaxation via still uncharacterized pathways. In summary, the results suggest that the Iα and Iβ isoforms of cGKI mediate relaxation by different cellular pathways which, however, are about equally effective in a physiological setting, i.e. NANCmediated relaxation. These findings may explain the absence of differences in the phenotypes of sm-cGKIα and sm-cGKIβ mice described previously (Weber et al., 2007). Supplementary materials related to this article can be found online at doi:10.1016/j.ejphar.2011.07.048.
Acknowledgment We thank Dr. Maria Huster for technical assistance. The work was supported by the Deutsche Forschungsgemeinschaft.
References Boyer, J.C., Christen, M.O., Balmes, J.L., Bali, J.P., 1998. Calcitonin gene-related peptideinduced relaxation of isolated human colonic smooth muscle cells through different intracellular pathways. Biochem. Pharmacol. 56, 1097–1104. Costall, B., Naylor, R.J., Tan, C.C., 1983. Field stimulation-induced responses of circular smooth muscle from guinea-pig stomach. Characterisation of the receptor mechanisms involved. Naunyn Schmiedebergs Arch. Pharmacol. 323, 155–161. Desch, M., Sigl, K., Hieke, B., Salb, K., Kees, F., Bernhard, D., Jochim, A., Spiessberger, B., Hocherl, K., Feil, R., Feil, S., Lukowski, R., Wegener, J.W., Hofmann, F., Schlossmann, J., 2010. IRAG determines nitric oxide- and atrial natriuretic peptide-mediated smooth muscle relaxation. Cardiovasc. Res. 86, 496–505. Dick, J.M., Van Molle, W., Brouckaert, P., Lefebvre, R.A., 2002. Relaxation by vasoactive intestinal polypeptide in the gastric fundus of nitric oxide synthase-deficient mice. J. Physiol. 538, 133–143. Friebe, A., Mergia, E., Dangel, O., Lange, A., Koesling, D., 2007. Fatal gastrointestinal obstruction and hypertension in mice lacking nitric oxide-sensitive guanylyl cyclase. Proc. Natl. Acad. Sci. U. S. A. 104, 7699–7704. Geiselhoringer, A., Gaisa, M., Hofmann, F., Schlossmann, J., 2004a. Distribution of IRAG and cGKI-isoforms in murine tissues. FEBS Lett. 575, 19–22. Geiselhoringer, A., Werner, M., Sigl, K., Smital, P., Worner, R., Acheo, L., Stieber, J., Weinmeister, P., Feil, R., Feil, S., Wegener, J., Hofmann, F., Schlossmann, J., 2004b. IRAG is essential for relaxation of receptor-triggered smooth muscle contraction by cGMP kinase. EMBO J. 23, 4222–4231. Groneberg, D., König, P., Koesling, D., Friebe, A., 2009. The nitric oxide receptor guanylyl cyclase in gastrointestinal smooth muscle is dispensable for gastrointestinal motility and survival in mice. Naunyn Schmiedebergs Arch. Pharmacol. 379 (Suppl. 1), 20. Hofmann, F., 2005. The biology of cyclic GMP-dependent protein kinases. J. Biol. Chem. 280, 1–4.
C. Ertl et al. / European Journal of Pharmacology 670 (2011) 266–271 Hofmann, F., Gensheimer, H.P., Gobel, C., 1985. cGMP-dependent protein kinase. Autophosphorylation changes the characteristics of binding site 1. Eur. J. Biochem. 147, 361–365. Hofmann, F., Feil, R., Kleppisch, T., Schlossmann, J., 2006. Function of cGMP-dependent protein kinases as revealed by gene deletion. Physiol. Rev. 86, 1–23. Johnson, R.M., Lincoln, T.M., 1985. Effects of nitroprusside, glyceryl trinitrate, and 8bromo cyclic GMP on phosphorylase a formation and myosin light chain phosphorylation in rat aorta. Mol. Pharmacol. 27, 333–342. Kitajima, S., Harada, K., Hori, M., Ozaki, H., Karaki, H., 1996. Dissociation of P2 purinoceptor-mediated increase in intracellular Ca2+ level from myosin light chain phosphorylation and contraction in rat aorta. Br. J. Pharmacol. 118, 543–548. Kitazawa, T., Semba, S., Huh, Y.H., Kitazawa, K., Eto, M., 2009. Nitric oxide-induced biphasic mechanism of vascular relaxation via dephosphorylation of CPI-17 and MYPT1. J. Physiol. 587, 3587–3603. Komalavilas, P., Lincoln, T.M., 1996. Phosphorylation of the inositol 1,4,5-trisphosphate receptor. Cyclic GMP-dependent protein kinase mediates cAMP and cGMP dependent phosphorylation in the intact rat aorta. J. Biol. Chem. 271, 21933–21938. Lee, M.R., Li, L., Kitazawa, T., 1997. Cyclic GMP causes Ca2+ desensitization in vascular smooth muscle by activating the myosin light chain phosphatase. J. Biol. Chem. 272, 5063–5068. Lefebvre, R.A., Baert, E., Barbier, A.J., 1992. Influence of NG-nitro-L-arginine on nonadrenergic non-cholinergic relaxation in the guinea-pig gastric fundus. Br. J. Pharmacol. 106, 173–179. Li, C.G., Rand, M.J., 1990. Nitric oxide and vasoactive intestinal polypeptide mediate non-adrenergic, non-cholinergic inhibitory transmission to smooth muscle of the rat gastric fundus. Eur. J. Pharmacol. 191, 303–309. Mashimo, H., Kjellin, A., Goyal, R.K., 2000. Gastric stasis in neuronal nitric oxide synthase-deficient knockout mice. Gastroenterology 119, 766–773. Moreland, S., Nishimura, J., van Breemen, C., Ahn, H.Y., Moreland, R.S., 1992. Transient myosin phosphorylation at constant Ca2+ during agonist activation of permeabilized arteries. Am. J. Physiol. 263, C540–C544. Niiro, N., Koga, Y., Ikebe, M., 2003. Agonist-induced changes in the phosphorylation of the myosin- binding subunit of myosin light chain phosphatase and CPI17, two regulatory factors of myosin light chain phosphatase, in smooth muscle. Biochem. J. 369, 117–128. Ny, L., Pfeifer, A., Aszodi, A., Ahmad, M., Alm, P., Hedlund, P., Fassler, R., Andersson, K.E., 2000. Impaired relaxation of stomach smooth muscle in mice lacking cyclic GMPdependent protein kinase I. Br. J. Pharmacol. 129, 395–401. Ozaki, H., Gerthoffer, W.T., Publicover, N.G., Fusetani, N., Sanders, K.M., 1991. Timedependent changes in Ca2+ sensitivity during phasic contraction of canine antral smooth muscle. J. Physiol. 440, 207–224.
271
Pfeifer, A., Klatt, P., Massberg, S., Ny, L., Sausbier, M., Hirneiss, C., Wang, G.X., Korth, M., Aszodi, A., Andersson, K.E., Krombach, F., Mayerhofer, A., Ruth, P., Fassler, R., Hofmann, F., 1998. Defective smooth muscle regulation in cGMP kinase I-deficient mice. EMBO J. 17, 3045–3051. Ruth, P., Landgraf, W., Keilbach, A., May, B., Egleme, C., Hofmann, F., 1991. The activation of expressed cGMP-dependent protein kinase isozymes I alpha and I beta is determined by the different amino-termini. Eur. J. Biochem. 202, 1339–1344. Sanders, K.M., Ward, S.M., 1992. Nitric oxide as a mediator of nonadrenergic noncholinergic neurotransmission. Am. J. Physiol. 262, G379–G392. Schlossmann, J., Ammendola, A., Ashman, K., Zong, X., Huber, A., Neubauer, G., Wang, G.X., Allescher, H.D., Korth, M., Wilm, M., Hofmann, F., Ruth, P., 2000. Regulation of intracellular calcium by a signalling complex of IRAG, IP3 receptor and cGMP kinase Ibeta. Nature 404, 197–201. Shah, V., Lyford, G., Gores, G., Farrugia, G., 2004. Nitric oxide in gastrointestinal health and disease. Gastroenterology 126, 903–913. Somlyo, A.P., Somlyo, A.V., 2003. Ca2+ sensitivity of smooth muscle and nonmuscle myosin II: modulated by G proteins, kinases, and myosin phosphatase. Physiol. Rev. 83, 1325–1358. Surks, H.K., Mochizuki, N., Kasai, Y., Georgescu, S.P., Tang, K.M., Ito, M., Lincoln, T.M., Mendelsohn, M.E., 1999. Regulation of myosin phosphatase by a specific interaction with cGMP- dependent protein kinase Ialpha. Science 286, 1583–1587. Tang, K.M., Wang, G.R., Lu, P., Karas, R.H., Aronovitz, M., Heximer, S.P., Kaltenbronn, K.M., Blumer, K.J., Siderovski, D.P., Zhu, Y., Mendelsohn, M.E., 2003. Regulator of G-protein signaling-2 mediates vascular smooth muscle relaxation and blood pressure. Nat. Med. 9, 1506–1512. Toda, N., Herman, A.G., 2005. Gastrointestinal function regulation by nitrergic efferent nerves. Pharmacol. Rev. 57, 315–338. Weber, S., Bernhard, D., Lukowski, R., Weinmeister, P., Worner, R., Wegener, J.W., Valtcheva, N., Feil, S., Schlossmann, J., Hofmann, F., Feil, R., 2007. Rescue of cGMP kinase I knockout mice by smooth muscle specific expression of either isozyme. Circ. Res. 101, 1096–1103. Wernet, W., Flockerzi, V., Hofmann, F., 1989. The cDNA of the two isoforms of bovine cGMP-dependent protein kinase. FEBS Lett. 251, 191–196. Wolfe, L., Corbin, J.D., Francis, S.H., 1989. Characterization of a novel isozyme of cGMPdependent protein kinase from bovine aorta. J. Biol. Chem. 264, 7734–7741. Worner, R., Lukowski, R., Hofmann, F., Wegener, J.W., 2007. cGMP signals mainly through cAMP kinase in permeabilized murine aorta. Am. J. Physiol. Heart Circ. Physiol. 292, H237–H244. Yano, S., Kiyota, Y., Yamamoto, M., Watanabe, K., 1995. Pharmacological features of non-adrenergic non-cholinergic (NANC) relaxation induced by electrical vagal stimulation in isolated mouse stomach. Jpn. J. Pharmacol. 69, 9–15.