Small molecules interfering with Rac1 activation are considered as potential drugs and are already studied in animal models. A widely used inhibitor without reported attenuation of RhoA activity is NSC23766 [(N6-[2-[[4-(diethylamino)-1-methylbutyl]amino]-6-methyl-4-pyrimidinyl]-2-methyl-4,6-quinolinediamine trihydrochloride]. We found that NSC23766 inhibits the M2 muscarinic acetylcholine receptor (M2 mAChR)-induced Rac1 activation in neonatal rat cardiac myocytes. Surprisingly, NSC27366 concomitantly suppressed the carbachol-induced RhoA activation and a M2 mAChR-induced inotropic response in isolated neonatal rat hearts requiring the activation of Rho-dependent kinases. We therefore aimed to identify the mechanisms by which NSC23766 interferes with the differentially mediated, M2 mAChR-induced responses. Interestingly, NSC23766 caused a rightward shift of the carbachol concentration response curve for the positive inotropic response without modifying carbachol efficacy. To analyze the specificity of NSC23766, we compared the carbachol and the similarly Giβγ-mediated, adenosine-induced activation of Gi protein–regulated potassium channel (GIRK) channels in human atrial myocytes. Application of NSC23766 blocked the carbachol-induced K+ current but had no effect on the adenosine-induced GIRK current. Similarly, an adenosine A1 receptor-induced positive inotropic response in neonatal rat hearts was not attenuated by NSC23766. To investigate its specificity toward the different mAChR types, we studied the carbachol-induced elevation of intracellular Ca2+ concentrations in human embryonic kidney 293 (HEK-293) cells expressing M1, M2, or M3 mAChRs. NSC23766 caused a concentration-dependent rightward shift of the carbachol concentration response curves at all mAChRs. Thus, NSC23766 is not only an inhibitor of Rac1 activation, but it is within the same concentration range a competitive antagonist at mAChRs. Molecular docking analysis at M2 and M3 mAChR crystal structures confirmed this interpretation.
Muscarinic acetylcholine receptors (mAChRs) mediate the parasympathetic regulation of the heart. Despite convincing evidence for the existence of all five mAChR in the heart (Wang et al., 2001), the quantitatively dominant cardiac mAChR is the M2 subtype (Gomeza et al., 1999). It mediates the negative chronotropic and dromotropic effects in the sinoatrial and atrioventricular nodes, respectively. Likewise, the negative inotropic effects of acetylcholine in the heart are attributed to this receptor subtype (Caulfield, 1993; Brodde and Michel, 1999). Mechanistically, these effects involve activation of pertussis toxin–sensitive G proteins of the Gi/o family, inhibition of cAMP production, and activation of the Gi protein–regulated potassium channel (GIRK) that generates the acetylcholine (Ach)-induced potassium current IK,Ach.
In addition, novel functions of cardiac ACh have emerged that are apparently unrelated to its function in the sinoatrial and atrioventricular nodes. Cardiac myocytes apparently can produce and release ACh, which in turn suppresses hypertrophy (Rocha-Resende et al., 2012). Exposure to ACh improves survival after both acute and chronic myocardial infarction (Katare et al., 2010). Moreover, cholinergic activation of M2 mAChR increases contractility in ventricular muscle from rats with chronic heart failure (Hussain et al., 2009). This positive inotropic effect is likely mediated by increased myosin light chain-2 (MLC-2) phosphorylation and enhanced Ca2+ sensitivity of the myofilaments. As the carbachol (2-[(aminocarbonyl)oxy]-N,N,N-trimethylethanaminium chloride) evoked positive inotropic response as well as myosin light chain-2 (MLC-2) phosphorylation were attenuated by pertussis toxin treatment or the inhibition of the Rho-dependent kinase (ROCK), the activation of Gi/o-proteins and the ROCK pathway are involved (Hussain et al., 2009).
G protein–coupled receptors such as M2 mAChR are important upstream activators of RhoGTPases (Kjoller and Hall, 1999). Prototypical Gi/o protein–coupled receptors (GiPCRs) such as M2 mAChR (Lechleiter et al., 1989) primarily induce the activation of Rac1 via Giβγ-dimers and guanine nucleotide exchange factors (GEFs) such as Tiam1 (Rossman et al., 2005; Brown et al., 2006; Vettel et al., 2012). Nevertheless, M2 mAchR is also able to activate RhoA and ROCK in neonatal rat cardiac myocyte-derived H10 cells under specific conditions (Vogt et al., 2007). Apparently, a high expression of the long isoform of regulator of G protein signaling 3 (RGS3L) is associated with a switch in the M2 mAchR-induced activation of RhoGTPases. Whereas at low expression of RGS3L, Rac1 activation via Giβγ-dimers prevails, at high expression levels this preference is switched to RhoA activation (Vogt et al., 2007).
Both the Rho family GTPases, Rac1 and RhoA, are involved in the control of many different cellular processes, including actin cytoskeleton reorganization, adhesion, migration, polarity, gene expression, and cell cycle progression (Etienne-Manneville and Hall, 2002; Burridge and Wennerberg, 2004). We report herein that, as in failing rat hearts (Hussain et al., 2009), carbachol induces a positive inotropic response in neonatal rat heart ventricles via activation of M2 mAChR in a ROCK-dependent manner. Thus, to investigate the role of RhoGTPases, we used NSC23766, a small molecule that inhibits the activation of Rac1 by hindering its binding to the Rac-specific GEFs Tiam1 and Trio without altering the access of the related RhoA or Cdc42 to their specific GEFs (Gao et al., 2004). Interestingly, NSC23766 not only inhibited the carbachol-induced Rac1 activation but also the concomitantly induced RhoA activation as well as the M2 mAChR-induced positive inotropic response in the ventricles. We will provide evidence that these additional effects are due to a direct interaction of NSC23766 with M2 mAChR. It acts as a competitive antagonist in the same concentration range usually used for the inhibition of intracellular Rac1 activation. Based on the recently published crystal structures of M2 mAChR and M3 mAChR with bound inverse agonists (Haga et al., 2012; Kruse et al., 2012), we will additionally show the molecular docking of NSC23766 to these receptor subtypes.
Materials and Methods
We obtained pirenzepine [11-[(4-methylpiperazin-1-yl)acetyl]-5,11-dihydro-6H-pyrido[2,3-b][1,4]benzodiazepin-6-one], AF-DX 116 (11-[[2-[(diethylamino)methyl]-1-piperidinyl]acetyl]-5,11-dihydro-6H-pyrido[2,3-b][1,4]benzodiaze-pin-6-one), 4-DAMP (1,1-dimethyl-4-diphenylacetoxypiperidinium iodide), NSC23766 [(N6-[2-[[4-(diethylamino)-1-methylbutyl]amino]-6-methyl-4-pyrimidinyl]-2-methyl-4,6-quinolinediamine trihydrochloride], and H1152P [(S)-(+)-2-methyl-1-[(4-methyl-5-isoquinolinyl)sulfonyl]-hexahydro-1H-1,4-diazepine dihydrochloride] from Tocris Biosciences (Bristol, UK). We obtained carbachol (2-[(aminocarbonyl)oxy]-N,N,N-trimethylethanaminium chloride) and N6-cyclopentyladenosine (CPA) from Sigma-Aldrich (St. Louis, MO).
Contractility of Cardiac Muscles.
Our study conformed with the U.S. National Institutes of Health Guidelines for the Care and Use of Laboratory Animals (NIH Publication No. 85-23, revised 1996)and the Norwegian National Guidelines for Research Ethics in Science and Technology. The experimental protocol was approved by the Norwegian Animal Research Authority, and all procedures we describe were performed in accordance with their recommendations. Pregnant rats and mothers with litters were kept in separate cages; otherwise, two adult male rats were housed in each cage in a temperature-regulated room on a 12-hour light/dark cycle. The animals were given access to food and water ad libitum. Heart tissue was collected from neonatal rats 2 to 5 days after birth.
The left ventricle from neonatal rats was isolated (the right ventricle was removed) and cut open from the base to apex to form either one or two muscle strips and prepared for functional analyses as previously described elsewhere (Sjaastad et al., 2003) with minor modifications. In brief, the muscles were mounted in organ baths containing the relaxing solution and were allowed to adapt at 31°C for approximately 15 minutes before the solution was changed to one equilibrated with 95% O2/5% CO2 at 31°C and containing 119.2 mM NaCl, 3.0 mM KCl, 2.5 mM CaCl2, 1.2 mM MgSO4, 2.4 mM KH2PO4, 24.9 mM NaHCO3, 10 mM glucose, and 2.2 mM mannitol, pH 7.4. The muscles were field stimulated at 1.5 Hz with 5-millisecond impulses and current about 20% above the individual threshold (10–15 mA, determined in each experiment). The contraction-relaxation cycles (CRCs) were recorded and analyzed as previously described elsewhere (Sjaastad et al., 2003) with respect to the maximal developed force (Fmax) and the maximal rate of force development (dF/dt)max. The inotropic responses to agonists were expressed as the percentage increases in (dF/dt)max.
The descriptive parameters at the end of the equilibration period were used as basal (control) values. The values after agonist responses were expressed as a percentage of the control period (100%) before the addition of signaling inhibitors (NSC23766, H1152P) in the event an inhibitor altered the basal force. The experiments were performed in the presence of 1 μM prazosin (added ∼90 minutes before the agonist) to block α-adrenergic receptors. Prazosin did not influence the basal contraction-relaxation cycle characteristics or the electrical stimulation threshold. Carbachol was added cumulatively (concentration-response curves) or as a bolus (20 μM) until the maximal response was obtained (4–6 minutes after administration). We constructed concentration-response curves by estimating the centiles (EC10-EC100) and calculating the corresponding mean and horizontal positioning expressed as −logEC50 (Sjaastad et al., 2003).
Isolation and Culture of Neonatal Rat Cardiac Myocytes.
Neonatal rat cardiac myocytes (NRCM) were isolated from hearts of 1- to 3-day-old Wistar rats as described previously elsewhere (Will et al., 2010). In brief, rat hearts were minced and subjected to serial collagenase/pancreatin digestion to release single cells: 0.5 mg/ml collagenase type II (CellSystems Biotechnologie Vertrieb GmbH, Troisdorf, Germany) and 0.6 mg/ml pancreatin (Sigma-Aldrich, St. Louis, MO) were used. The obtained cell suspension was then placed on top of a Percoll gradient (GE Healthcare Bio-Sciences, Piscataway, NJ) to separate the cardiac myocytes from the other cell types. The cardiac myocyte fraction was seeded on collagen I–coated plates and cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% (w/v) fetal calf serum, 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin. We used 0.1 mM 5′-bromo-2′-deoxyuridine (BrdU) to prevent overgrowth of other cell types. The cells were used at the latest 3 to 4 days after isolation for the experiments.
Rac1 and RhoA GTPase Activation Assay.
The cellular levels of GTP-loaded RhoA and Rac1 were determined with effector pull-down assays using glutathione S-transferase (GST) fusion proteins containing the Rho-binding domain of rhotekin (GST-RBD) or Rac-binding domain of p21-activated kinase (PAK1) (GST-PBD). GST-RBD and GST-PBD were expressed in and purified from Escherichia coli. After activation of the NRCM with carbachol (1 mM, 90 seconds), cells were lysed in ice-cold GST-Fish buffer as described elsewhere (Vettel et al., 2012), and GTP-bound RhoGTPases were precipitated with either GST-RBD or GST-PBD. The amount of activated and total GTPases was then determined by immunoblot analysis.
For immunoblot analysis, the protein samples were separated by SDS-PAGE using 15% acrylamide and transferred onto nitrocellulose membranes. Membranes were blocked with Roti-Block (Carl Roth GmbH, Karlsruhe, Germany) for 1 hour at room temperature and were incubated with anti-RhoA (26 C4, 1:200; Santa Cruz Biotechnology, Santa Cruz, CA), or anti-Rac1 (610650, 1:1000; BD Biosciences, Franklin Lakes, NJ) overnight at 4°C. After incubation with appropriate secondary antibodies for 1 hour, the proteins were visualized by enhanced chemiluminescence and documented with an imaging system (Alpha Innotech/ProteinSimple, San Jose, CA).
Recording of Inward-Rectifier K+ Currents.
The study was approved by the local ethics committee of the Medical Faculty Mannheim, University of Heidelberg (no. 2011-216N-MA), and each patient gave written informed consent. Right atrial appendages were obtained from 12 patients with sinus rhythm who were undergoing open heart surgery for coronary bypass grafting.
The atrial myocytes were isolated using a standard protocol and were suspended in a storage solution (Voigt et al., 2010a). The membrane currents were measured with a standard voltage-clamp technique, and pClamp-Software (v10.2; Molecular Devices, Sunnyvale, CA) was used for data acquisition and analysis (Voigt et al., 2010a). The borosilicate glass microelectrodes had tip resistances of 2–5 MΩ when filled with pipette solution (80 mM potassium aspartate, 8 mM NaCl, 40 mM KCl, 5 mM Mg-ATP, 2 mM EGTA, 0.1 mM GTP-Tris, 10 mM HEPES, pH 7.4), and the seal resistances were 4–8 GΩ. Series resistance and cell capacitance were compensated for up to 70%. The current amplitudes are expressed as current densities (pA/pF). From a holding potential of −80 mV, the currents were activated using a standard ramp-pulse protocol from −100 to +40 mV (slope 0.112 mV/ms, 0.5 Hz), and the current amplitudes were analyzed at −100 mV.
Drugs were applied via a rapid solution exchange system (ALA Scientific Instruments, Long Island, NY). During the drug-free periods, the rapid solution exchange system supplied only a standard solution (120 mM NaCl, 20 mM KCl, 1 mM MgCl2, 2 mM CaCl2, 10 mM glucose, 10 mM HEPES, pH 7.4). The potassium current (IK,ACh and IK,Ado) was stimulated with carbachol (2 µM for 2 minutes each) and adenosine (0.01–1.0 mM for 0.5 minutes each) as described elsewhere (Voigt et al., 2010b). Ba2+ (1 mM) was applied to each myocyte, and the currents were analyzed after subtraction of the resulting leak current. Data were not corrected for the calculated liquid junction potential (−12 mV; JPCalc software, version 2.2; Axon Instruments, Inc., Union City, CA) (Barry, 1994).
Fluorometric Determination of the Cytosolic Ca2+ Concentration.
For determination of the cytosolic Ca2+ concentration, the fluorometric Ca2+ quantification assay HitHunter Calcium Ca NW PLUS Assay (DiscoveRx, Fremont, CA) was used according to the manufacturer’s instructions. In brief, HEK-293 cells constitutively expressing M1, M2, or M3 mAChRs were grown in DMEM (Invitrogen, Carlsbad, CA) supplemented with 10% (w/v) fetal calf serum, 2 mM l-glutamine, 100 units/ml penicillin, and 100 μg/ml streptomycin in a humidified atmosphere of 5% CO2 at 37°C. Cells were seeded into 96-well plates at a density of 50,000 cells/well. Cells were transfected with a eukaryotic expression vector encoding Gαqi5 (Coward et al., 1999) (16 ng/µl) using Polyfect (Qiagen, Valencia, CA). The growth medium was aspirated from the cells and replaced with dye containing loading buffer. Probenecid was used to inhibit a potential efflux of the dye from the cells. After equilibration at 37°C for 30 minutes, NSC23766 or its solvent was added for an additional 30 minutes. Stimulation with carbachol and fluorescence measurements was performed at room temperature.
After agonist stimulation, the level of cytosolic Ca2+ was recorded for 20 seconds using an EnVision 2102 plate reader (PerkinElmer Life and Analytical Sciences, Waltham, MA). The results were expressed as ΔF values (maximum fluorescence intensity values after stimulation minus average fluorescence intensity values for unstimulated cytosolic Ca2+ concentration). All experiments were performed in 2–6 replicate measurements.
The structures of NSC23766 and methylene blue (3,7-bis (dimethylamino)-phenothiazin-5-ium chloride) were docked in both the M2 mAChRs crystal structure 3UON (Protein Data Bank [PDB]) (Haga et al., 2012) and in the M3 mAChRs crystal structure 4DAJ (Protein Data Bank) (Kruse et al., 2012) using the Cambridge Crystallographic Data Centre (CCDC, Cambridge, UK) software GOLD 5.1 with default settings and GOLDSCORE as the scoring function (Jones et al., 1995). The cocrystallized inverse agonists (3R)-1-azabicyclo[2.2.2]oct-3-yl-hydroxy(diphenyl)acetate (QNB) and tiotropium served as comparison for the general orientation of NSC23766 and methylene blue. The docking results and receptor interactions were analyzed with the software LigandScout (Inte:Ligand, Maria Enzersdorf, Austria) (Wolber and Langer, 2005; Wolber et al., 2006) by building three-dimensional (3D) pharmacophores.
Differences between group means for continuous data were compared by one-way analysis of variance (ANOVA) plus an appropriate post-test or unpaired Student’s t test. Data are expressed as mean ± S.E.M. P < 0.05 was considered statistically significant.
Carbachol Evokes a Positive Inotropic Response in Neonatal Rat Heart Ventricles in a M2 mAChR-Mediated and ROCK-Dependent Manner.
To test whether carbachol can evoke a positive inotropic response in the neonatal rat heart, we prepared left ventricles from freshly isolated hearts and measured the contraction-relaxation cycles of muscle strips in an organ bath. Any inotropic response was expressed as an increase in maximal development of force [(dF/dt)max]. As in failing adult rat hearts (Hussain et al., 2009), carbachol concentration-dependently increased contractility, reaching a maximum about 40% above basal (see Supplemental Fig. 1) with a half maximal effective concentration of about 1 µM. We found that 4-DAMP (9 nM), an antagonist of M2 and M3 mAChR, and AF-DX 116 (450 nM), a competitive antagonist preferentially of M2 mAChR, antagonized the carbachol-induced inotropic response. They shifted the carbachol concentration response curve to the right by 0.5 and 1.0 log units, respectively. In contrast, the M1 mAchR selective antagonist pirenzepine (63 nM) had no statistically significant effect (Fig. 1A). The calculated affinity constants (apparent pA2) for 4-DAMP, AF-DX 116, and pirenzepine were 8.55, 7.27, and 6.97, respectively.
This affinity pattern is in accordance with that described for the M2 mAChR type (Caulfield and Birdsall, 1998). Moreover, the carbachol-induced inotropic response was significantly suppressed by treatment with the ROCK inhibitor H1152P (1 µM, Fig. 1B). Therefore, the data demonstrate that carbachol induces a concentration-dependent positive inotropic response in the neonatal rat ventricle that is mediated by the M2 mAChR and is dependent on ROCK activity.
Carbachol-Induced Activation of Rac1 and RhoA in NRCM Is Suppressed by NSC23766.
To investigate the involvement of Rac1 and RhoA in M2 mAChR-mediated responses, we studied Rac1 and RhoA activation in NRCM by effector pull-down. Stimulation with 1 mM carbachol for 90 seconds induced a prominent activation of Rac1 (2-fold basal) and a weaker but significant activation of RhoA (1.4-fold basal) (Fig. 2). The carbachol-induced Rac1 activation was completely inhibited by treatment of NRCM with 50 µM NSC23766, which is in accordance with the recently reported role of the NSC23766-sensitive GEF Tiam1 in Gi-mediated Rac1 activation in NRCM (Vettel et al., 2012). Although no interference of NSC23766 with cellular RhoA activity has been reported (Gao et al., 2004), the carbachol-induced activation of RhoA in NRCM was surprisingly absent in NSC23766-treated cells (Fig. 2).
NSC23766 Inhibits the Carbachol-Induced but Not the CPA-Induced Positive Inotropic Response in Neonatal Rat Hearts.
We next investigated the effect of NSC23766 on the carbachol-induced inotropic response in neonatal rat hearts. As shown in Fig. 3A, the positive inotropic response evoked by 10 µM carbachol was largely suppressed when the muscle strips were treated with 50 µM NSC27366. It is surprising that no inhibition of the positive inotropic response by 50 µM NSC23766 was evident at 100 µM carbachol (Fig. 3B). Similar to the known antagonist AF-DX 116, 50 µM NSC27366 caused a rightward shift of the concentration-response curve for carbachol by 1 log unit without modifying the maximal inotropic response (Fig. 3 C). This behavior is typical for a competitive antagonist but not for an allosteric modulator. The calculated affinity constant (apparent pA2) of NSC27366 under these conditions was 4.56 (~2.7 µM).
Adenosine A1 receptors (A1R) are typical Gi-coupled receptors that initiate similar signaling cascades as M2 mAChR in cardiac myocytes and thus have overlapping functions in the heart (Headrick et al., 2011). In addition, A1R induces activation of Rac1 and RhoA in a similar manner as M2 mAChR (Vogt et al., 2007) and has recently been described as exerting positive inotropic effects in human trabeculae from the right atrium of patients with ischemic heart disease (Gergs et al., 2009). Therefore, we tested whether the selective A1R agonist CPA also increases the force of contraction in neonatal rat left ventricles.
As shown in Fig. 3D, 1 µM CPA induced a positive inotropic response that was approximately half as large as that evoked by M2 mAChR activation. In contrast with the carbachol-induced inotropic response, 50 µM NSC23766 did not inhibit the CPA-evoked increase in contractility.
NSC23766 Blocks Carbachol-Induced but Not Adenosine-Induced Activation of GIRK Currents in Human Atrial Myocytes.
GIRKs are a family of inward-rectifier potassium channels that are directly regulated by Giβγ-subunits. Thus, activation of M2 mAChR or A1R in the atria causes GIRK-mediated currents, contributing to the negative chronotropic effect of these GiPCRs (Yamada et al., 1998). Activation of Rac1 or RhoA is not required for this effect. To investigate whether NSC23766 also interferes with human M2 mAChR signaling upstream of the activation of RhoGEFs, we studied the effect of an acute application of 10 µM NSC23766 on carbachol-induced and adenosine-induced K+-inward currents, IK,ACh and IK,Ado, respectively, in human atrial myocytes at −100 mV.
In our experimental setting, we compared the first GIRK current activation (S1) by the respective agonist (2 µM carbachol or 1 mM adenosine) with the second stimulation (S2) in the presence or absence of 10 µM NSC23766 (Figs. 4 and 5). In time-matched controls, as shown elsewhere (Voigt et al., 2010a), the second peak current (S2) evoked by either agonist was smaller compared with S1 because of the well-known phenomenon of desensitization (Voigt et al., 2007) (Figs. 4, A and C, and 5, A and C).
A second process caused by desensitization is the rapid decline of the current amplitude from the initial peak to a quasi steady-state level despite the presence of the respective agonist (Voigt et al., 2007) (Figs. 4, A and C, and 5, A and C). Thus, the agonist-evoked GIRK currents IK,ACh and IK,Ado were evaluated by calculating the S2/S1 ratio of the respective peak and quasi steady-state currents (Figs. 4D and 5D) in the absence or presence of NSC23766, which largely suppressed the carbachol-induced IK,ACh current (Fig. 4, C and D).
As shown in Fig. 5B, maximal agonist-evoked IK,Ado activation occurred at 0.1–1 mM adenosine. NSC23766 did not inhibit A1R-mediated IK,Ado activation (Fig. 5, A, C, and D). On the contrary, the S2/S1 ratio of IK,Ado in the presence of NSC23766 appeared to be increased compared with IK,Ado in time-matched controls, indicating a possible interference of NSC27366 with the recovery from desensitization of IK,Ado. NSC27366 had no effect on the carbachol- or adenosine-independent basal current (IK1), largely mediated by inward-rectifier K+-current channels (Fig. 4B). Taken together, the data indicate that NSC23766 specifically inhibits the carbachol-induced IK,ACh activation by blockade of M2 mAChR.
NSC27366 Is an Unspecific, Competitive Inhibitor at M1, M2, and M3 Muscarinic Acetylcholine Receptors Recombinantly Expressed in HEK-293 Cells.
To characterize the pharmacology of NSC23766 in more detail and to study the specificity at the different types of mAChR, we used HEK-293 cells constitutively expressing M1, M2, or M3 mAChR (Offermanns et al., 1994). To achieve an equal coupling of all three receptor subtypes to phospholipase C and subsequent Ca2+ release from intracellular stores, we transfected the cells transiently with Gαqi5 chimera containing the five carboxyl-terminal amino acids from Gi. This artificial G protein is able to divert also GiPCR (e.g., M2 AChR) to the activation of phospholipase C, thus generating increases in [Ca2+]i that can be quantified by a fluorometric assay (Kurko et al., 2009).
As shown in Fig. 6, stimulation of all three mAChR receptor subtypes with increasing concentrations of carbachol induced easily detectable [Ca2+]i elevations, but with the previously described difference pattern in carbachol potency (Offermanns et al., 1994). A half-maximal increase in [Ca2+]i was observed at 10 nM, 300 nM, and 3 µM concentrations of carbachol in cells expressing M3, M1, and M2 mAChR, respectively (Fig. 6). At all three receptor subtypes, 100 µM NSC23766 caused a rightward shift of the carbachol concentration response curves by 1.5 to 2 log units without altering the efficacy of carbachol. The affinities of NSC23766 (calculated from the apparent pA2 of this single concentration) for the M1, M2, and M3 mAChR were 1, 4.5, and 5 µM, respectively.
To confirm the competitive nature of the antagonism of NSC23766 and to get a more accurate affinity estimate, we studied the carbachol-induced M3 mAchR-mediated elevation of [Ca2+]i at increasing concentrations (1–100 µM) of NSC23766 (Fig. 7) and compared the resulting rightward shifts by Schild regression analysis (Kenakin and Beek, 1981; Kenakin, 1982). This analysis revealed a linear relation with a slope of 0.846 ± 0.137. From the pA2 of 5.36 (Kenakin and Beek, 1981; Kenakin, 1982), the affinity of NSC23766 for M3 AChR was calculated to be about 4 µM. Collectively, these data demonstrate that NSC23766 can act as a nonspecific competitive antagonist at all tested types of mAChR.
Modeling of the Binding of NSC23766 to M2 and M3 mAChR.
Recently, two crystal structures of muscarinic acetylcholine receptor/ligand complexes in their inactive conformation were solved: the human M2 mAChR (PDB: 3UON) (Haga et al., 2012) and M3 mAChR from rattus norvegicus (PDB: 4DAJ) (Kruse et al., 2012). A comparison of these crystal structures shows the high conservation of the orthosteric binding pocket (Haga et al., 2012; Kruse et al., 2012) and indicates the key residues for ligand binding (Fig. 8A). An important aspartic acid that can form an electrostatic interaction with a positive charge is present in all known ligands such as the cocrystallized QNB in the M2 mAChR structure or tiotropium in the M3 mAChR structure. This interaction strongly affects the general orientation of a ligand in the binding pocket. As QNB and tiotropium both show the same orientation and therefore form a reference for the interpretation of docking results (Fig. 8B), there is now an opportunity to rationalize the binding modes of the thus far uncharacterized ligands that bind to mAChRs. Thus, we calculated the docking results for NSC27366 and for methylene blue, an organic dye that has been shown to be an antagonist at mAChR (Abi-Gerges et al., 1997).
Although NSC23766 and methylene blue do not exhibit many structural similarities (Fig. 9A), they both dock well to the orthosteric binding pocket of M2 mAChR (Fig. 9, B and D) as well as M3 mAChR (Fig. 9, C and E). The orientation of NSC23766 in both receptors is very similar although they differ in a few interactions. The positively charged nitrogen of the tertiary amine shows electrostatic interaction with D103, Y104, Y403, and W400 in M2 mAChR (Fig. 9B) and with D147, Y148, Y506, and Y529 in M3 mAChR (Fig. 9C). The nitrogen in position 3 of the pyrimidine ring serves as a hydrogen bond acceptor for the hydroxyl groups of Y104 and Y403 of the M2 mAChR and for the hydroxyl groups of Y148 and Y506 of the M3 mAChR. The methyl group of the aliphatic chain is embedded in two differing hydrophobic pockets. For M2 mAChR, it consists of A191, F195, W400, and Y403, whereas Y148, W199, and W503 build that pocket for M3 mAChR. The quinoline ring is opposite T187 and W422 in M2 mAChR and opposite Y506 and W525 in M3 mAChR. The methyl group of the pyrimidine ring forms hydrophobic contacts with W155, F181, T187, T190, and A191 in M2 mAChR and with W199, T234, A235, and A238 in M3 mAChR.
In comparison, the binding mode of methylene blue to the M2 mAChR and the M3 mAChR is identical (Fig. 9, D and E; corresponding residue numbers are written in brackets). The positive ionizable nitrogen of methylene blue interacts with D103 (D147), Y403 (Y506), and W400 (W503). The nitrogen of the thiazine ring serves as hydrogen bond acceptor for N404 (N507). V111 (V155), A194 (A238), and W400 (W503) also are opposite to the benzyl ring with the dimethylamino group; V111 (V155), A194 (A238), and W155 (W199) stabilize the binding through hydrophobic contacts with the dimethylamino group.
We next compared the binding of NSC23766 and methylene blue to that of QNB. As shown in Fig. 7B, the aromatic rings in QNB and tiotropium fill the identical two hydrophobic pockets in the orthosteric binding site. One of them (near N404; see Fig. 8B) overlaps with NSC23766 (Fig. 9F), and the other sterically matches methylene blue (Fig. 9G) in their surmised binding conformations, respectively. Whereas the positive ionizable center shows the same position for NSC23766, QNB, and tiotropium, it differs in the binding of methylene blue. Surprisingly, N404 in the M2 mAChR (corresponding to N507 in M3 mAChR), which forms hydrogen bonds to the inverse agonist in both crystal structures, seems to be involved in the interaction with methylene blue but not with NSC23766.
The RhoGTPases RhoA and Rac1 have pivotal functions in the heart and are involved in the development of several cardiovascular diseases (Brown et al., 2006). Especially, an overactivation of Rac1 has been shown to be deleterious. Mice expressing constitutively active Rac1 developed cardiomyopathy (Sussman et al., 2000). Cardiac myocyte-specific overexpression of an active form of Rac predisposes the heart to postischemic contractile dysfunction (Talukder et al., 2013). Through activation of NADPH oxidase, Rac1 activation contributes to oxidative stress, which promotes cardiac injury and dysfunction (Maack et al., 2003; Endou et al., 2004; Elnakish et al., 2012; Nagase et al., 2012; Zhang et al., 2012; Ma et al., 2013). Moreover, Rac1 is essential for the hypertrophic response in the heart (Sussman et al., 2000; Satoh et al., 2006; Vettel et al., 2012). Therefore, Rac1 inhibitors are potential cardioprotective agents for treatment of cardiac hypertrophy, atrial as well as ventricular fibrosis, and thus of atrial fibrillation and heart failure (Shen et al., 2009; Vettel et al., 2012; Ma et al., 2013). In addition to cardiac diseases, Rac1 inhibitors are thought to be beneficial and thus have been tested in vivo for a variety of other disorders (Müller et al., 2008; Shibata et al., 2008; Tan et al., 2011, 2012).
NSC23766 inhibits the binding and activation of Rac1 GTPase via the Rac-specific GEFs Trio or Tiam1 without altering the access of RhoA or Cdc42 to their GEFs (see Fig. 10). In cells, it blocks serum or platelet-derived growth factor-induced Rac1 activation and lamellipodia formation without affecting the activity of Cdc42 or RhoA (Gao et al., 2004). Therefore, NSC23766 is the most widely used small molecule to inhibit Rac1 activity in different cell types and, most importantly, in in vivo mouse models by injection, intrathecal catheter, or even implanted Alzet osmotic pump (Binker et al., 2008; Müller et al., 2008; Shibata et al., 2008; Shen et al., 2009; Colomba et al., 2011; Tan et al., 2011, 2012; Ma et al., 2013). From different studies in which NSC23766 was used to inhibit Rac1-dependent responses, IC50 values of 50–100 µM were reported (Montalvo-Ortiz et al., 2012).
In NRCM, we used 30–100 µM NSC27366 to suppress α1A-adrenoceptor/Giβγ-induced Rac1 activation by Tiam1 (Vettel et al., 2012). Our data clearly demonstrate that NSC23766 is not only an inhibitor of the activation of Rac1 GTPase by the GEFs Tiam and Trio but also is a nonselective, competitive antagonist at mAChRs (see Fig. 10) in the generally used concentration range for the inhibition of Rac1 activation, with a calculated affinity constant of 3–5 µM. Therefore, NSC23766 application rapidly and reversibly blocked all carbachol-induced processes in cardiac myocytes such as activation of GIRK, Rac1, RhoA, and the positive inotropic response (see Fig. 9). Accordingly, NSC23766 shifted the concentration-response curve for carbachol to the right. The extent of the rightward shift was increased by increasing NSC23766 concentrations, and a Schild analysis revealed a linear correlation with a slope almost equal to 1, consistent with a competitive nature of the antagonism. In addition, docking of NSC23766 to the orthosteric binding pocket of M2 mAChR and M3 mAChR (Haga et al., 2012; Kruse et al., 2012) revealed that NSC23766 can be accommodated into the binding pocket and has the possibility of establishing electrostatic as well as hydrophobic interactions with the receptor protein (see Fig. 9).
Unexpected inhibition of mAChRs by organic small molecules is not a novel concept. Structurally diverse molecules such as tacrine derivates (Svejdova et al., 1990), substituted naphthofurans (Monte et al., 1998), or the dye methylene blue (Abi-Gerges et al., 1997) have been previously described as exhibiting mAChR antagonistic properties. In addition, many drugs used today for treatment, such oral antipsychotics (for recent review see Ozbilen et al., 2012), are antagonists at mAChRs and cause parasympatholytic side effects such as vision blurry, dry mouth, and urinary retention, which are often a cause for a discontinuation of treatment (Leucht et al., 2012). Therefore, the unexpected antagonistic effect of NSC23766 at mAChRs in the concentration range required to substantially suppress Tiam1- and Trio-mediated Rac1 activation will clearly limit its potential as novel therapeutic agent. Our data also imply a need for consequent testing of newly found “Rac1 inhibitors” (Ferri et al., 2009; Montalvo-Ortiz et al., 2012) for mAChR antagonistic properties before pursuing them further in drug development.
As our docking and 3D pharmacophore analyses show, the binding of NSC27366 to mAChRs can be rationally explained; our studies provide a 3D interaction model that allows prioritizing the selection of compounds for further experimental testing through virtual screening.
In addition to the finding that NSC27366 is a competitive antagonist at mAChR, we show that in neonatal rat hearts the stimulation of M2 mAChR or A1R induced ROCK-dependent positive inotropic responses, which are usually absent in adult myocardium (Hussain et al., 2009). Because such effects reoccur in failing rat myocardium and also have been reported in diseased human myocardium (Gergs et al., 2009; Hussain et al., 2009), it might be speculated that such coupling is part of the well-known fetal gene program being reinitiated in the failing heart (Katz, 1990). Further research to investigate the molecular mechanisms underlying this positive inotropic effect of primarily GiPCRs and its putative potential for therapy is ongoing in our laboratories.
Our data demonstrate that NSC23766 is a bona fide competitive antagonist at mAChRs in the concentration ranges used for the inhibition of the activation of the monomeric GTPase Rac1 by GEFs such as Tiam1 and Trio. This property must be considered in any future experimental and therapeutic use of NSC23766.
The authors thank Dr. Martina Schmidt (Groningen, The Netherlands) for providing HEK-293 cells constitutively expressing M1, M2, and M3 mAChRs; Dr. Barbara Möpps (Ulm, Germany) for the generous gift of the expression vector for the Gαqi5 chimera; Dr. Klaus Mohr (Bonn, Germany) for critically reading the manuscript and scientific advice; and Doris Baltus, Heike Rauscher, and Katrin Kupser for expert technical assistance.
Participated in research design: Levay, Krobert, Wittig, Bermudez, Wolber, Dobrev, Levy, Wieland.
Conducted experiments: Levay, Krobert, Wittig, Voigt, Bermudez.
Performed data analysis: Levay, Krobert, Voigt, Wolber, Dobrev, Levy, Wieland.
Wrote or contributed to the writing of the manuscript: Levay, Krobert, Voigt, Bermudez, Wolber, Dobrev, Levy, Wieland.
- Received June 14, 2013.
- Accepted July 22, 2013.
M.L. and K.A.K. contributed equally to this work.
The part of the study performed in Germany was supported by the DZHK (German Centre for Cardiovascular Research) and the BMBF (German Ministry of Education and Research). The work performed in Oslo was supported by the Norwegian Council on Cardiovascular Disease, the Research Council of Norway, Stiftelsen Kristian Gerhard Jebsen, South-Eastern Norway Regional Health Authority, Anders Jahre’s Foundation for the Promotion of Science, the Novo Nordisk Foundation, the Family Blix Foundation, the Simon Fougner Hartmann Family Foundation, and grants from the University of Oslo.
- adenosine A1 receptors
- AF-DX 116
- AF-DX 384
- N-[2-[2-[(dipropylamino)methyl]-1-piperidinyl]ethyl]-5,6-dihydro-6-oxo-11H-pyrido[2,3-b][1,4] benzodiazepine-11-carboxamide
- 2-[(aminocarbonyl)oxy]-N,N,N-trimethylethanaminium chloride
- 1,1-dimethyl-4-diphenylacetoxypiperidinium iodide
- Dulbecco’s modified Eagle’s medium
- guanine nucleotide exchange factor
- Gi protein–regulated potassium channel
- Gi/o protein–coupled receptor
- glutathione S-transferase fusion proteins containing Rac-binding domain of p21-activated kinase
- glutathione S-transferase fusion proteins containing the Rho-binding domain of rhotekin
- (S)-(+)-2-methyl-1-[(4-methyl-5-isoquinolinyl)sulfonyl]-hexahydro-1H-1,4-diazepine dihydrochloride
- human embryonic kidney
- muscarinic acetylcholine receptor
- methylene blue
- 3,7-bis (dimethylamino)-phenothiazin-5-ium chloride
- neonatal rat cardiac myocytes
- (N6-[2-[[4-(diethylamino)-1-methylbutyl]amino]-6-methyl-4-pyrimidinyl]-2-methyl-4,6-quinolinediamine trihydrochloride
- Protein Data Bank
- regulator of G-protein signaling 3
- Rho-dependent kinase
- Copyright © 2013 by The American Society for Pharmacology and Experimental Therapeutics