|Home | About | Journals | Submit | Contact Us | Français|
To examine the effects of asenapine on nitric oxide (NO) release and Ca2+ transients in H9C2 cell line, which were either subjected to peroxidation or not.
H9C2 were treated with asenapine alone or in presence of intracellular kinase blockers, serotoninergic and dopaminergic antagonists, and voltage Ca2+ channels inhibitors. Experiments were also performed in H9C2 treated with hydrogen peroxide. NO release and intracellular Ca2+ were measured through specific probes.
In H9C2, asenapine differently modulated NO release and Ca2+ movements depending on peroxidative condition. The Ca2+ pool mobilized by asenapine mainly originated from the extracellular space and was slightly affected by thapsigargin. Moreover, the effects of asenapine were reduced or prevented by kinases blockers, dopaminergic and serotoninergic receptors inhibitors, and voltage Ca2+ channels blockers.
On the basis of our findings, we can conclude that asenapine by interacting with its specific receptors, exerts dual effects on NO release and Ca2+ homeostasis in H9C2; this would be of particular clinical relevance when considering their role in cardiac function modulation.
Asenapine is currently approved by the Food and Drug Administration for the acute treatment of schizophrenia as well as for the acute treatment of manic or mixed episodes associated with bipolar 1 disorder with or without psychotic features. The effects of asenapine are mainly related to a high-affinity antagonism of dopamine D2 receptors and serotonin 5-HT2A receptors although the antagonism of serotonin 5-HT2B, 5-HT2C, 5-HT5A, 5-HT6, and 5-HT7 receptors is also involved. In addition, asenapine acts as a partial agonist at the 5-HT1A site.
Compared to other antipsychotics, asenapine has been reported to be well tolerated and to have minimal effects on metabolic and cardiovascular parameters. However, available information on this issue is scarce and mainly based on clinical studies and trials.
The results, recently obtained in coronary artery endothelial cells (CEC), have highlighted the role of asenapine on nitric oxide (NO) release and on different NO synthase (NOS) isoforms activation. In particular, in physiological conditions, asenapine was found to increase NO production through the endothelial NOS (eNOS) isoform activation, while it caused opposite effects in CEC that underwent peroxidation. Those effects were found to be related to cyclic adenosine monophosphate (cAMP)/protein kinase A (PKA) and phospholipase C (PLC) pathways. Furthermore, the involvement of 5-HT1A receptors was shown by the use of the selective antagonist, NAD-299.
It is widely accepted that changes of cytosolic Ca2+ ([Ca2+]c) levels are of primary importance in the regulation of NO production, considering that the constitutive isoform of NOS present in endothelial cells is a Ca2+ dependent enzyme.[5,6,7] Extracellular stimuli that initiate calcium signaling could either activate the voltage-gated Ca2+ channels in the plasma membrane or induce the activation of ligand-gated Ca2+ channels located on the intracellular stores of Ca2+.
NO could be involved in modulation of (Ca2+) c by changes of cAMP levels, as well. It is also to note that both alterations of NO and Ca2+ could exert dual effects on cardiomyoblasts depending on oxidative status of cells and their concentration. It is suggested the involvement of changes of free oxygen species production, of mitochondria function, and alterations of cardiac contractile myofilaments sensitivity to Ca2+.[10,11,12,13]
Thus, the aim of this study was to investigate in H9C2 about the effects of asenapine on NO release and (Ca2+) c handling in nonperoxidative and peroxidative conditions and to analyze the mechanisms involved.
Rat cardiac H9C2 cells were obtained from the American Type Culture Collection (Rockville, MD) and cultured in Dulbecco Modified Eagle Medium (DMEM; Sigma, Milan, Italy) supplemented with 10% heat-inactivated fetal bovine serum (fetal bovine serum [FBS]; sigma), 1% penicillin–streptomycin (sigma), and 2 mM L-glutamine (sigma) in a humidified incubator at 5% CO2, 95% air, and 37°C. Cells were subcultured when they reached about 90% confluence, and the experiments were performed with cells from passages 14–17. H9C2 (1.5 × 106 cells/ml) were plated into 0.1% gelatin-coated coverslips with DMEM and 10% FBS supplemented with L-glutamine, penicillin-streptomycin for 4 h. After this time, the cells were used for experiments.
In H9C2, the oxidative stress was generated using 200 μM hydrogen peroxide for 20 min in DMEM without FBS and phenol red. Control cells were treated with DMEM + 10% FBS and phenol red only.
NO production was measured by Griess method (Promega, Milan, Italy) in 8 × 103 cells in 96-well plates in DMEM + 10% FBS without phenol red, as previously described.
Briefly, H9C2 were treated for 30 s, 120 s, and 300 s with asenapine (10 pM-100 μM; Sigma) in the dose-related and time-course studies. In control cells, DMEM and 10% FBS only were used. In addition, in other samples, asenapine was given alone or in presence of the adenylyl cyclase blocker 2'5'-dideoxyadenosine (1 μM; Sigma; 15 min), the selective cAMP-dependent PKA inhibitor, H89 (1 μM; Sigma; 15 min), the PLC γ inhibitor, U73122 (1 µM, sigma; 15 min), the Ca2+-calmodulin protein kinase (CaMKII) inhibitor, KN93 (1 µM; sigma; 15 min), the L-type Ca2+ channel blocker, amlodipine (1 μM, Santa Cruz Biotechnology; Dallas, USA; 15 min), the T-type Ca2+ channel blocker, ML218 (1 μM, Alomone Labs, Jerusalem, Israel; 15 min), the selective 5-HT1A antagonist, NAD-299 hydrochloride (1 μM, Tocris Bioscience, Bristol, United Kingdom; 15 min), the selective 5-HT2A antagonist, nefazodone hydrochloride (1 μM, Tocris Bioscience; 15 min), the selective D2 receptor antagonist, propionylpromazine hydrochloride (1 μM, Tocris Bioscience; 15 min), or the NOS blocker, Nω-nitro-L-arginine methyl ester (L-NAME; 10 mM; sigma; 15 min). The agonist-antagonists and their vehicle were also tested in the basal medium without agents. In some samples, the effects of 20 min hydrogen peroxide (200 µM) on NO release were also examined. In addition, the effects of 15 min prestimulation with asenapine (10 pM-100 μM) on NO release caused by hydrogen peroxide in H9C2 were analyzed. At the end of stimulations, NO production in the sample supernatants was examined by adding an equal volume of Griess reagent following the manufacturer's instruction. At the end of incubation, the absorbance at 570 nm was measured by a spectrometer (BS1000 Spectra Count, San Jose, CA, USA) and the NO production was quantified with respect to nitrate standard curve,[5,6,11] and expressed as a percentage in comparison with basal value.
The values obtained corresponded to the NO (μmol) produced, after each stimulation, by samples containing 1.5 μg of proteins each.
The coverslips were washed twice with sterile PBS 1X and incubated with Fura-2/acetoxymethyl ester (AM; 5 µM final concentration, sigma) for 30 min in the dark in DMEM 10% FBS and without phenol red supplemented with 1% penicillin-streptomycin and 2 mM L-glutamine. After further washing with DMEM, the coverslips in DMEM without Ca2+ were mounted in agitation at 37°C in thermostatic quartz cuvette in a Hitachi F-4500 fluorescence spectrometer, operating in continuous for 300 s at the wavelength pair 340 nm excitation/510 nm emissions.
Asenapine (10 pM-100 µM) was added to the suspension of Fura-2/AM loaded H9C2, in the presence or absence of Ca2+ in the incubation medium (obtained with 50 mM ethylene glycol tetraacetic acid [EGTA]). In some experiments, the effects of asenapine were compared with those of ATP (10 µM; sigma). Moreover, some experiments were performed by asenapine administration in the absence or presence of Ca2+ ionophore, A23187 (1 μM), H89 (1 μM), U73122 (1 µM), KN93 (1 µM), NAD-299 hydrochloride (1 µM), nefazodone hydrochloride (1 µM), propionyl promazine hydrochloride (1 µM), amlodipine (1 µM), ML218 (1 µM), and L-NAME (10 mM).
Moreover, the effects of asenapine on the “capacitive” Ca2+ entry through the plasma membrane Ca2+ channels were examined by the evaluation of the rate of Ca2+ overshoot in H9C2. The cells on coverslips were pretreated with EGTA (50 mM) and were subsequently exposed to thapsigargin (10 µM) and asenapine alone or in co-stimulation for 5 min.
Finally, 60 mM CaCl2 was added to the samples and the effects on Ca2+ overshoot were analyzed.
Fura-2/AM loaded H9C2 were treated for 20 min with 200 μM hydrogen peroxide. In some samples, asenapine (10 pM-100 µM) was added to the suspension of H9C2, which underwent peroxidative stress, after 20 min hydrogen peroxide.
Some experiments were also performed in H9C2 that had undergone 20 min peroxidation by administering asenapine after NAD-299 hydrochloride (1 μM), nefazodone hydrochloride (1 μM), propionyl promazine hydrochloride (1 μM), amlodipine (1 µM), ML218 (1 µM) and L-NAME (10 mM). All blockers were given for 15 min before asenapine.
Quantification of (Ca2+) c was conventionally obtained by measuring the Fura-2/AM fluorescence in Ca2+-free (0.1 M EGTA) and Ca2+-saturated conditions by the equation (Ca2+) c = Kd ([R − Rmin]/[Rmax − R]).[14,15,16] The fluorescence intensities obtained were corrected for cell autofluorescence at the wavelengths employed.
The results obtained were examined through one-way ANOVA followed by Newman–Keuls post hoc test. A simple regression analysis was performed to examine the correlation between the dose of asenapine administrated and the observed (Ca2+) c effects in the dose-response study. All data are presented as mean ± standard deviation of five different experiments for each experimental protocol. A P < 0.05 was considered statistically significant.
As shown in Figure 1a, in nonperoxidative (physiologic) condition, asenapine increased NO release in H9C2 in a dose-dependent and time-related way (P < 0.05). Those results were linearly related to the dose of asenapine administered (at 30 s, R: 0.74; at 120 s, R: 0.64; at 300 s, R = 0.61). A plateau was nearly reached at 10 μM asenapine 120 s, which was used for all subsequent experiments.
Different results were obtained in H9C2 which had undergone peroxidation. The 20 min treatment with hydrogen peroxide increased NO release by about 100% of control values [P < 0.05; Figure 1b], an effect which was dose-dependently counteracted by asenapine [P < 0.05; Figure 1b].
In H9C2 pretreated with 2'5' dideoxyadenosine, H89, U73122, ML218, amlodipine, nefazodone, propionyl promazine, and L-NAME, the effects of asenapine were abolished. NAD-299 and KN93 reduced the response of H9C2 to asenapine on NO release in comparison with what was observed with asenapine alone [P < 0.05; Figure 1c].
As shown in Figure 2, asenapine (10 pM-100 µM), caused a dose-dependent and stable increase of (Ca2+) c (P < 0.05). Those results were linearly correlated to the dose of asenapine administered (at 30 s, R: 0.51; at 60 s, R: 0.52; at 180 s, R = 0.54; at 300s, R: 0.52).
The plateau was nearly obtained at 10 µM asenapine and amounted to 123.6 ± 1.3 nM (P < 0.05) from control values of 107.8 ± 1.9 nM; this concentration was maintained for all subsequent experiments.
As depicted in Figures Figures3a,3a, ,bb and and4a,4a, ,b,b, the effects of asenapine on (Ca2+) c were almost abolished in H9C2 cultured in Ca2+-free medium (P > 0.05) and potentiated by Ca2+ ionophore, A23187. Moreover, the effects of asenapine on (Ca2+) c were abolished by H89, U73122, amlodipine, ML218, propionyl promazine, nefazodone, and L-NAME and reduced by KN93 [Figures [Figures3,3, ,4c4c–f and and5b5b–e]. NAD-299 increased the response of H9C2 to asenapine [Figures [Figures3c,3c, ,dd and and5a5a].
As shown in Figures Figures6a6a and and7a,7a, ,b,b, the pool mobilized by asenapine was different from the one affected by ATP. Hence, the increase of (Ca2+) c caused by either asenapine or ATP did not significantly differ, irrespective of the sequence of addition (P > 0.05).
Differentially, as reported in Figures Figures6b6b and and7c,7c, the addition of 10 µM asenapine in costimulation with thapsigargin markedly changed the kinetics of (Ca2+) c fluctuations promoted by thapsigargin.
Finally, the effects of asenapine on (Ca2+) c were markedly reduced in H9C2 which had undergone peroxidation [Figure [Figure8a8a and andb].b]. As shown in Figure Figure8a8a and andb,b, in 20 min hydrogen peroxide treated cells, the increase of (Ca2+)c by asenapine reached a plateau at 100 nM asenapine. This concentration was used for the experiments performed with blockers. In this respect, it is notable that in H9C2, which were treated with hydrogen peroxide, the effects of asenapine were abolished by nefazodone, propionyl promazine, amlodipine, ML218, and L-NAME and were increased by NAD-299 [Figure 8c].
The results of this study have shown for the 1st time that in cardiomyoblasts asenapine exerts dual effects on both NO release and changes of (Ca2+)c, depending on the oxidative status of cells. Mechanisms related to D2 receptors and 5-HT receptors-dependent signaling and L- and T-type voltage Ca2+ channels opening would be involved in the response of H9C2 to asenapine.
Changes of (Ca2+) c levels are of primary importance in the regulation of NO production, eNOS being a Ca2+ dependent enzyme. In the cytosol, Ca2+ is maintained at a very low level and is concentrated in intracellular calcium stores such as the endoplasmic reticulum (ER). The dynamic steady state of Ca2+ in the cytosol is the result of the balance between active and passive fluxes through the cell membranes of various stores and is strictly regulated through mechanisms involving inositol-1,4,5-triphosphate-phosphate (IP3) generation, “capacitive Ca2+ entry” and the activation of voltage-gated Ca2+ channels.[9,18,19]
The results obtained in the present study have shown that in nonperoxidative conditions, asenapine can cause a persistent increase of (Ca2+) c mainly from extracellular origin by 5HT2A and through D2 receptors involvement, and L- and T-type Ca2+ channels opening. In addition, by the same way, asenapine would increase NO release in H9C2. Mechanisms related to PKA and PLC-dependent signaling would be involved in such effects. Hence, D2-like receptors have been reported to inhibit cAMP/PKA pathway and L-type voltage-gated Ca2+ channels, and finally decrease (Ca2+)c. By this way, the antagonistic effect of asenapine on D2 receptors could explain the results obtained about (Ca2+) c in H9C2.
Observations about the involvement of 5-HT2A receptors in the effects of asenapine are not in agreement with previous reports. Hence, 5-HT2A receptors have been found to be responsible for Ca2+ increase in response to serotonin by modulation of 5-HT4 subtype activity and by increasing the magnitude of the L-type Ca2+ current. For this reason, the finding of the abolishing effects of asenapine on Ca2+ fluctuations in H9C2 by nefazodone, would not confirm those data. That discrepancy could be related to the different cellular model and to the rather complex mechanism of action of asenapine, which acts as a high affinity antagonist of D2 receptors and 5-HT2A receptors, but also as antagonist of 5-HT2B, 5-HT2C, 5-HT5A, 5-HT6, and 5-HT7 receptors. Moreover, asenapine acts only as a partial agonist at the 5-HT1A site. Regarding this issue, it is notable that the 5-HT1A receptors were found to play an inhibitory role on the effects of asenapine on Ca2+ movements, being the response of H9C2 to asenapine increased by the selective 5-HT1A antagonist, NAD-299.[5,23] Hence, since 5-HT1A have been reported to negatively modulate the Ca2+/calmodulin pathway in neurons, our results could be assumed to be in agreement with previous observations, although obtained in a different cellular model.
In addition, the Ca2+ pool mobilized by asenapine was found to be independent from that mobilized by an agent acting through IP3 generation, such as ATP, but partly similar to the one affected by thapsigargin, the Ca2+-ATPase inhibitor which is able to deplete the ER Ca2+ pool. These results would confirm the involvement of 'capacitive Ca2+ entry' in the response of H9C2 to asenapine.
The increased (Ca2+) c caused by asenapine could be the basis for the observed increase of NO release. Hence, like in CEC, asenapine was able to augment NO production in a dose-dependent and time-related way in H9C2. NO release was measured by the Griess system, which has been previously used for NO detection in endothelial cells and H9C2, as well.[5,6]
Moreover, the effects of asenapine were abolished by cAMP/PKA, PLC, 5HT2A, and D2 receptors inhibitors, and L- and T-type Ca2+ channel blockers. The various antagonists were used at the same concentrations which were able to prevent the effects of asenapine in CEC or isolated arteries and to abolish the nitrite release in vascular smooth muscle cells.
As previously observed in CEC, NO release caused by asenapine was only reduced by NAD-299. Although not examined, it could be speculated that 5-HT1A receptors could play a dual role on NO release. Hence, the potentiating effects on Ca2+ movements elicited by asenapine through 5-HT1A-related mechanisms could counteract their inhibitory effects on NO production.
Overall, the results obtained in presence of various antagonists seem to suggest common mechanisms at the basis of increased NO release and Ca2+ influx in H9C2. As mentioned above, the interaction of asenapine with 5HT2A and D2 receptors would activate an intracellular PKA and PLC-dependent signaling that could increase NO release either directly through eNOS phosphorylation or indirectly through augmented Ca2+ influx from extracellular space by L- and T-type Ca2+ channels. Meanwhile, the activation of 5-HT1A receptors by asenapine could also contribute to Ca2+-dependent NO release in H9C2.
It is to note that in peroxidative conditions, the effects of asenapine on Ca2+ movements were reduced both in terms of maximum increase and as duration. As observed in nonperoxidative conditions, also in this case the effects were abolished by L- and T-type Ca2+ channels inhibitors, 5HT2A and D2 receptors blockers and slightly increased by NAD-299.
Moreover, asenapine was able to counteract the effects of hydrogen peroxide on NO release, an effect which was also observed in CEC.
Thus, although not specifically examined, the reduction of NO release in peroxidative conditions could be linked to the lower Ca2+ influx, eNOS being a Ca2+ dependent enzyme, as reported above.
It should be noted that changes in NO could also be involved in modulation of Ca2+ transients. Hence, cAMP levels are regulated by NO/cGMP-dependent mechanisms related to phosphodiesterase II (PDE) inhibition. By this way, changes of NO release could influence cAMP levels which would interfere with the activity of Ca2+ channels, serotonin and dopaminergic receptors, and Ca2+ pump.
Interestingly, regarding this issue, in presence of L-NAME, which was able to prevent the increased NO release in H9C2, the effects of asenapine on Ca2+ movements were nearly abolished in both nonperoxidative and peroxidative conditions.
Overall, the dual effects elicited by asenapine on Ca2+ in H9C2 would be of particular clinical relevance. On the one hand, in nonperoxidative conditions, asenapine would exert a beneficial role on cardiac contraction through sustained increase of Ca2+; on the whole, in peroxidation, asenapine would limit massive intracellular Ca2+ accumulation, which could be detrimental for myocytes through increased free radical generation, damage of mitochondria, and activation of caspase cascade.
Although not specifically examined, NO as well could be involved in mediating the protective effects elicited by asenapine in H9C2. Hence, NO has been shown to modulate several aspects of “physiological” myocardial function. The effects of NO are influenced by its cellular and enzymatic source, the amount generated, the presence of reactive oxygen species, and the activation of cGMP-dependent and independent signal transduction pathways.
Regarding this issue, it is to note that on the one hand, at low concentration, NO could contribute to increasing myocardial contractility by cGMP/PDE II activation and intracellular Ca2+ increase; on the other hand, at high concentration, NO could exert negative effects on cardiac performance through desensitization of cardiac contractile myofilaments to Ca2+ and peroxynitrite release.
Although the results obtained in the present study evidence a role for NO in the modulation of Ca2+ transients caused by asenapine in H9C2, further studies could help clarify the relationship between NO release and Ca2+ and the implications in terms of cardiac function.
Asenapine was found for the 1st time to affect NO release and Ca2+ transients in H9C2 through its specific receptors and L- and T-type Ca2+ channels opening. Intracellular signaling involving PKA and PLC-related pathways was also shown to play a role.
There are no conflicts of interest.
We thank the Azienda Ospedaliera Maggiore Della Carità di Novara for its help.