Oxytocin (OT), a hormone recently identified in the heart, induces embryonic and cardiac somatic stem cells to differentiate into cardiomyocytes (CM), possibly through nitric oxide (NO). We verified this hypothesis using P19 cells and P19 Clone 6 derivatives expressing a green fluorescent protein (GFP) reporter linked to cardiac myosin light chain-2v promoter. OT treatment of these cells induced beating cell colonies that were fully inhibited by N,G-nitro-l-arginine-methyl-ester (l-NAME), an inhibitor of NO synthases (NOS), partially reduced by 1400W, an inhibitor of inducible NOS, and ODQ, an inhibitor of NO-sensitive guanylyl cyclases. The NO generator S-nitroso-N-acetylpenicillamine (SNAP) reversed the l-NAME inhibition of cell beating and GFP expression. In OT-induced cells, l-NAME significantly decreased transcripts of the cardiac markers Nkx2.5, MEF2c, α-myosin heavy chain, and less, GATA4, endothelial NOS, and atrial natriuretic peptide, as well as the skeletal myocyte (SM) marker myogenin. Image analysis of OT-induced P19Cl6-GFP cells revealed ventricular CM coexpressing sarcomeric α-actinin and GFP, with some cells exclusively expressing α-actinin, most likely of the SM phenotype. The OT-mediated production of CM, but not SM, was diminished by l-NAME. In P19 cells, exogenously added OT stimulated the expression of its own transcript, which was reduced in the presence of l-NAME. Surprisingly, l-NAME alone decreased the expression of anti-stage specific embryonic antigen-1 marker of the undifferentiated state and induced some beating colonies as well as GFP in P19Cl6-GFP cells. Collectively, our data suggest that the pleiotropic action of NO is involved in the initiation of CM differentiation of P19 cells and maintenance of their undifferentiated state.
The in vitro differentiation of embryonic stem cells into cardiomyocytes (CM) is a poorly defined, inefficient, and relatively nonselective process . One of the most studied pluripotent stem cell models is the mouse P19 embryonal carcinoma cell line (P19EC), which is easy to culture, is amenable to genetic manipulation, and has a low frequency of spontaneous CM differentiation [2, 3]. Embryoid body (EB) formation and exposure to various inducers promote the differentiation of P19EC cells to mesodermal or ectodermal lineages. Among the mesodermal derivatives formed in EBs, the subtypes of cardiac cells (atrial CM, ventricular CM, and pacemaker cells) have been identified by histological, molecular, and electrophysiological criteria . Recently, a transgenic reporter cell line was derived from P19 Clone 6 (P19Cl6), which expresses green fluorescent protein (GFP) under the transcriptional control of the rat myosin light chain-2v (MLC-2v) promoter that is characterized by its ventricle-specific expression . P19Cl6 cells derived from P19EC cells by subculturing seem not to be committed to a mesodermal lineage, but represent a developmental stage closer to differentiated cardiac muscle than the parental cell line is . GFP reporter expression in P19Cl6 cells transduced with the MLC-2v promoter-GFP construct permits us to distinguish differentiation into CM or non-CM cells, facilitating CM quantification and, if necessary, the selection of CM generated under a variety of differentiation conditions . Identifying and optimizing these conditions are important as new approaches to probe the molecular mechanisms of CM differentiation and apply this knowledge to producing CM for cardiac therapy.
Oxytocin (OT), recently recognized as a cardiac hormone [7, 8], is one of the most potent CM differentiation factors for P19EC cells [9, 10]. It also has the unique ability to stimulate CM differentiation of cardiac somatic stem cells . This hormone, originally considered to be involved in uterine contraction and milk ejection, has broader physiological functions, including cardiovascular homeostasis and cell growth . In P19EC cells, OT induces beating CM formation earlier and more efficiently than does dimethylsulfoxide (DMSO), a frequently used, although not physiologically relevant, cardiomorphogen . Both OT- and DMSO-specific treatments upregulate OT receptors (OTRs) in P19EC cells during cell differentiation. Furthermore, an OT antagonist blocks not only OT-induced but also DMSO-induced CM differentiation, suggesting that DMSO acts via the OT pathway. Interestingly, OTRs have been implicated in retinoic acid-induced P19EC CM differentiation as well . The OT-activated signal transduction pathway leading to this differentiation is unknown. OTR activation in endothelial cells as well as in hypothalamic/hypophyseal and cardiac systems elicits nitric oxide (NO) [13, , , , –18]. Two studies, the first using N,G-nitro-l-arginine-methyl-ester (l-NAME) to inhibit nitric-oxide synthases (NOS), and the second employing S-nitroso-N-acetylpenicillamine (SNAP) as an exogenous source of NO, have demonstrated that NO is essential for optimal CM generation [19, 20]. We hypothesize that OTR signaling is linked to NO in the process of differentiation of stem cells into CM. Experiments were performed on pluripotent P19EC cells and, in addition, on the P19Cl6-GFP cell line to investigate differentiation in response to OT treatment.
Materials and Methods
Cell Culture and Differentiation
P19EC cells were propagated and differentiated as described . For differentiation, 0.25 × 106 cells were grown as aggregates for 4 days in bacteriological-grade Petri dishes (6-cm diameter) containing 5 ml of complete α-minimum essential medium (α-MEM) in the absence (noninduced) or presence of inducers. Inducers were present over the 4 days of aggregation. On day 4, aggregates were transferred to tissue culture-grade vessels (10-cm diameter dishes or 24-well plates), and cultured in complete α-MEM in the absence of inducers until day 14 or 16 (with day 0 corresponding to the start of aggregation). P19Cl6-GFP cell  differentiation was carried out similarly to P19EC cells, except that cell aggregation was initiated in hanging drops from day 0 to day 2. The inducers, used alone or in combination, were OT (10−7 M; Peninsula Laboratories, San Carlos, CA, http://www.penlabs.com); DMSO (0.5% vol/vol; Sigma-Aldrich, Ontario, Canada, http://www.sigmaaldrich.com); l-NAME (10−4 M; Sigma-Aldrich); SNAP (200 μM; Sigma-Aldrich); 1400W (10 μM; Sigma-Aldrich); and ODQ (10 μM; Sigma-Aldrich). The compounds l-NAME, SNAP, 1400W and ODQ did not affect cell viability at the concentrations tested, as assessed by staining with acridine orange and propidium iodide .
Cellular morphology, MLC-2v-GFP fluorescence and some protein markers were examined with an inverted microscope (Carl Zeiss, Jena, Germany http://www.zeiss.com) equipped for epifluorescence analysis. Micrographs were taken with a Nikon Coolpix 5,000 camera (Nikon, Tokyo, Japan, http://www.nikon.com) and fluorescent areas analyzed with Image J software (National Institutes of Health, Bethesda, MD, http://www.nih.gov) employing the threshold function. Immunocytochemistry was performed as described elsewhere . The mouse monoclonal antibody immunoglobulin M (IgM) anti-stage specific embryonic antigen-1 (SSEA-1) was obtained from the Developmental Studies Hybridoma Bank under the auspices of the National Institute of Child Health & Human Development and maintained by the University of Iowa, Department of Biological Sciences (Iowa City, IA). Goat polyclonal antibody IgG anti-sarcomeric myosin heavy chain (MHC; antibody K-16) and IgG anti-dihydropyridine receptor-α 1 (DHPR; antibody N-19) were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, http://www.scbt.com/). Secondary antibodies were conjugated to fluorescein (Santa Cruz Biotechnology).
Confocal Microscopy Analysis
P19Cl6-GFP cells, subjected to different treatments and collected on day 14, were stained for α-actinin with indirect Alexa Fluor 568 antibody complex. Anti-sarcomeric α-actinin monoclonal antibody, which reacts with skeletal and cardiac muscle α-actinins (clone EA-53 from Sigma Chemical Co.) was used at 1:800 dilution. Goat antimouse IgG secondary antibody conjugated to red fluorophore Alexa Fluor 568 was from Invitrogen Life Technologies (Burlington, Ontario, Canada, http://www.invitrogen.com). Image analysis was performed in MRC1024 confocal microscope (Bio-Rad Microscience, Cambridge, MA, http://www.bio-rad.com) equipped with a Krypton argon laser (excitation at 488 nm, 568 nm) combined with an Eclipse Model TE 3000 inverted microscope (Nikon) with emission filters measuring green and red fluorescence at wavelengths of 488 and 568 nm. The image was registered by LaserSharp V3.2 software (Bio-Rad). The Image J program and intensity correlation analysis software were deployed to calculate green fluorescence, red fluorescence, and their colocalization areas, and expressed as a percentage of the total image area.
P19EC cultures induced by OT or DMSO were incubated with fresh α-MEM for 4 hours. At the end of incubation, the culture media were collected and directly analyzed for OT concentration by radioimmunoassay (RIA) . Culture medium not exposed to cells served as a control. Antibody, specific for OT nonapeptide (a gift from Dr. M. Morris, Wright State University, Dayton, OH), and synthetic OT standards (Peninsula Laboratories) were used to measure OT concentration.
Reverse Transcription Polymerase Chain Reaction
Total RNA was extracted from cells with TRIzol reagent (Invitrogen Life Technologies). For OT transcript analysis, poly(A)+ mRNA was affinity purified on Oligotex mRNA columns (Qiagen, Ontario, Canada, http://www.quiagen.com), and mRNAs for endothelial NOS (eNOS), GATA4 and atrial natriuretic peptide (ANP) were analyzed in DNase-treated samples of total RNA (Turbo DNase-free, Ambion Inc., Austin, TX, http://www.ambion.com). The samples were then reverse transcribed to cDNA, amplified by polymerase chain reaction (PCR; Robocycler Gradient 40 Thermocycler, Stratagene, La Jolla, CA, http://www.stratagene.com), and resolved on agarose gel. Bands stained by ethidium bromide were analyzed with the Storm 840 imaging system and ImageQuant software (Version 4.2, Molecular Dynamics, Inc., Sunnyvale, CA, http://www.ump.com/mdynamic.html). To validate this reverse transcriptase (RT)-PCR assay as a tool for the semiquantitative measurement of mRNA, dose-response curves were established for different amounts of total RNA extracted from P19EC cells, and the samples were quantified in the linear phase of PCR amplification. These data were normalized to the corresponding values of 18S RNA PCR products serving as the internal controls (Ambion). The PCR primers and conditions are shown in supplemental Table 1. Sequences of mouse ANP genes have been described . The conditions for RT-PCR analysis of mouse GATA4 were adapted from Nemer and Nemer . The sequence of OT-specific primers has been reported by Lefebvre et al.  OTA+ was a sense-strand primer corresponding to a sequence in exon A, starting three base pairs downstream of the initiation codon. OTB+ and OTB− were sense- and antisense-strand primers located at the 5′ and 3′ ends of exon B, respectively. OTC– was an antisense primer complementary to a sequence in exon C, terminating at the stop codon. 18 S (kit from Ambion, Canada) gene expression was an internal standard measure in this experiment.
Table Table 1.. Primers
Real-Time Quantitative PCR
Real-time PCR was performed according to standard protocols on a MyIQ Real-Time PCR detection system (Bio-Rad). Briefly, 1 μg of total RNA was DNase-treated and transcribed to cDNA. For amplification, 2 μl of diluted cDNA was added to a 20-μl reaction mixture containing 1× iQ SYBR Green Supermix (Bio-Rad) and 200 nM forward and reverse primers. The thermal cycling program was 95°C for 2 minutes, followed by 40 cycles of 95°C for 30 seconds, 60°C for 30 seconds, and 72°C for 30 seconds. The sequences of primers purchased from Invitrogen Life Technologies are shown in supplemental Table 1. All reactions were run in triplicate. As negative controls, PCR was performed on water and on RNA without reverse transcription. The cycle number at which the reaction crossed an arbitrarily placed threshold (CT) was determined for each gene. The relative amount of mRNA levels was quantified by 2 − CT. Relative gene expression was normalized to glyseraldehyde-3-phosphate dehydrogenase (GAPDH) expression.
Values are expressed as mean ± SEM. Multiple comparisons were made by analysis of variance, followed by Dunnett's modified t test or, in the case of comparison with only one group, by a two-tailed version of Student's t test.
l-NAME Inhibits the Appearance of Beating Colonies in OT-Induced P19EC Cells
Treatment of P19EC cell aggregates with 10−7 M OT induced beating cell colonies earlier and in higher numbers than treatment with 0.5% DMSO (Fig. 1A) by day 14 (20 ± 0.3 vs. 14 ± 0.7 positive wells, respectively; p < .05). No beating cells were seen in OT-induced cultures in combination with 10−4 M l-NAME, a nonspecific inhibitor of NOS. l-NAME supplementation significantly reduced the number of beating colonies but did not completely abolish the generation of contracting cells evoked by DMSO (to 7 ± 2.0; p < .05).
Beating colonies had a different morphology after the various treatments. As seen in Figures 1C and 1F, OT and DMSO induced spherical beating cell colonies, displaying robust, synchronized contractions. Such colonies were absent in noninduced cultures (Fig. 1B) and in cultures treated with OT+l-NAME (Fig. 1D). Beating cell colonies in DMSO+ l-NAME treated cultures sometimes displayed large clusters of frequently beating cells (Fig. 1G), indicating that this treatment still promoted the CM phenotype. In contrast, a few very small, round clusters of weakly beating cells were seen in l-NAME treated cultures (Fig. 1E).
Changes of Gene Expression During OT-Mediated Cell Differentiation
Changes in GATA4 expression were an early indicator of OT-mediated cell differentiation as assessed by RT-PCR (Fig. 2). The results in Figure 2A demonstrate that, in P19EC cells exposed to OT, GATA4 mRNA was already increased in aggregates on the 4th day of differentiation (fivefold increment compared to day 1). A further elevation of GATA4 mRNA was seen on day 6 (10-fold), when the first beating cell colonies were detected. Maximally increased GATA4 mRNA (15-fold) was evident on days 10 and 14. Unexpectedly, in cells exposed to OT+ l-NAME or l-NAME alone, GATA4 mRNA expression was also detected, but at a relatively lower level. In l-NAME-treated cells, a sixfold GATA mRNA increment was noted on days 10 and 14.
As shown in Figure 2B, eNOS mRNA was barely detectable in cell aggregates (days 1 and 4) induced by OT in the presence or absence of l-NAME. Interestingly, in OT-induced cultures, the first appearance of beating CM (day 6) was associated with high eNOS mRNA expression (30 times greater than in cultures on day 1), which remained static through days 10–14. At the stage of cardiomyoblast generation on day 4 and with the onset of beating colonies on day 6, the eNOS mRNA level was very low in cells exposed to OT+ l-NAME or l-NAME alone. However, after CM formation on days 10 and 14, eNOS expression was significantly augmented in OT+ l-NAME induced cells and, to a lesser extent, in cells exposed to l-NAME alone; these augmentations were significantly smaller than in OT-treated cells.
ANP mRNA increased progressively until day 10 in cells exposed to OT (Fig. 2C). In contrast to effective beating cell inhibition, l-NAME had a less important impact on OT-induced ANP mRNA upregulation (respectively, 20% and 30% reductions at days 10 and 14). Interestingly, a high ANP mRNA level was found in l-NAME treated cultures displaying a few foci of beating cells, suggesting some cardiomyogenic process in NO-deficient cells.
The expression of other cardiac-specific genes was assessed in developing P19EC cells by real-time PCR. As shown in Figure 2, at a time corresponding to the cardiomyoblast stage (day 4), early CM (day 6), and beating CM (day 10), OT-treated cells expressed the cardiac transcription factors Nkx2.5 (Fig. 2D) and MEF2c (Fig. 2E) as well as the cardiac-specific α-MHC (Fig. 2F). At all such stages of cell differentiation, l-NAME largely reduced expression of these genes. Further analysis indicated that P19 cells responded to OT induction by expressing myogenin, the transcription factor participating in the generation of skeletal muscle in P19EC cells . This response was also attenuated in the presence of l-NAME (Fig. 2G).
The observation that, in response to OT treatment, P19EC cells generate both CM and skeletal muscle markers suggests multiple myogenic actions of the hormone. Therefore, we performed a series of experiments to visualize the population of P19 cells expressing CM markers upon stimulation with OT. For this purpose, we utilized P19Cl6-GFP cells induced with OT, l-NAME alone, and OT and l-NAME in combination. In cell cultures collected at day 14 of differentiation and displaying comparable and complete confluence (Fig. 3F), we analyzed the expression of sarcomeric (cardiac and skeletal) α-actinin with the antibody complex emitting red fluorescence (Alexa Fluor 568). Brilliant GFP-borne green fluorescence indicated the presence of MLC-2v ventricular marker. Analysis of fluorescence under confocal microscopy suggested colocalization of α-actinin and GFP fluorescence in 3.5% ± 0.4% of culture areas in OT-induced cell populations, but in only 0.6% ± 0.3% and 0.5% ± 0.2% of culture areas upon OT+ l-NAME and l-NAME treatments, respectively. The cultures stimulated with OT and OT+ l-NAME generated similar fractions of cells expressing only α-actinin (3.5% ± 0.5% and 3.2% ± 0.7% respectively). This fraction of cells was lower upon stimulation with l-NAME (1.7% ± 0.2%), and possibly contained some skeletal muscle cells (elongated shape) and/or nonventricular CM, indicated by round and triangular cell morphology. Noticeably, almost all GFP-positive cells displayed this latter shape and were induced by OT in a 12.3% ± 1.3% proportion. This proportion was reduced to 4.8% ± 0.8% by treatment with OT+ l-NAME and was 6.7% ± 0.3% in cultures induced with l-NAME alone (both p < .05 vs. OT), demonstrating l-NAME's inhibitory action on OT induction of cardiomyogenically driven GFP expression.
Expression of SSEA-1 and Contractile Proteins
In P19EC cells at an advanced stage of differentiation (day 14), we analyzed SSEA-1, the marker of the undifferentiated state. Positive fluorescence staining with monoclonal antibody against SSEA-1 has been demonstrated in noninduced cells (Fig. 4D). A smaller fraction of cells positive for SSEA-1 after induction by OT and/or l-NAME (Fig. 4A–4C) indicated that a differentiation process was initiated in these cultures. Secondly, we investigated whether the cells acquire two markers of the advanced contractile apparatus in CM, namely sarcomeric MHC and DHPR . As expected, noninduced cells were negative for DHPR and MHC (Fig. 4H, Fig. 4L), displaying only basal fluorescence. The two contractile proteins were clearly detected in OT-treated cells (Fig. 4E, 4I), whereas OT+ l-NAME and l-NAME treated cells exhibited only the low fluorescence of noninduced cells in the overall cultures (Fig. 4F, 4G, 4J, 4K). It is possible that, in addition to CM, some cells in OT- and/or l-NAME-treated cultures differentiated to other, nonmuscular cell phenotypes.
Endogenous OT Expression in P19EC Cells
The ability of P19 cells to undergo myogenesis depends in part on unknown factors in serum . Indeed, serum factors upregulate OTR , and we have already demonstrated that CM differentiation of P19EC cells is associated with OTR elevation . It is possible that differentiation depends on endogenous OT expression in P19 cells. Therefore, using RIA, we investigated whether P19EC differentiation is associated with OT production. On day 14 of OT-stimulated differentiation, OT was measured 4 hours after the addition of fresh medium. We detected 6.2 ± 0.2 pg of OT per milligram of protein in cell extracts and 14.2 ± 0.8 pg/ml in culture medium. Less OT was found in DMSO-stimulated P19EC cells: 3.7 ± 0.1 pg/mg of protein in cell extracts and 7.0 ± 0.1 pg/ml in medium. Because the contribution of serum-containing culture medium was only 1.1 ± 0.04 pg/ml, these results suggest that differentiated P19EC cells produce OT. The presence of OT transcript in OT-stimulated P19EC cells at a period preceding the appearance of beating colonies (day 6 of the differentiation protocol; Fig. 5) further indicates OT synthesis in these cells. In a series of experiments, RT-PCR was performed with OT gene-specific primers homologous to sequences on three exons of the OT gene, as represented schematically in Figure 5A. This strategy was chosen to reveal any structural differences in endogenous OT expression that might exist between the OT-coding regions in P19EC cells. The results reveal that OT-stimulated P19EC cells expressed all coding regions of the OT transcript. Secondly, we found that OT induction of P19EC stimulated OT mRNA expression in comparison to noninduced cells, but combined treatment with l-NAME abolished OT mRNA elevation (Fig. 5B).
SNAP Reverses the Inhibitory Effect of l-NAME on OT-Induced Cardiomyogenesis
In P19Cl6-GFP cultures, OT (22 wells) induced the generation of beating colonies more efficiently did than SNAP (18 wells), an NO generator (Fig. 6A). The number of wells with OT-induced contracting cells was reduced in the presence of l-NAME (4–5 wells) to levels seen in basal, spontaneous cardiomyogenic differentiation (noninduced, 3–4 wells). Compared with P19EC, where no beating cell colonies were observed in the absence of inducer (Fig. 1), some, although very rarely, contracting cells were spontaneously generated in P19Cl6-GFP cultures (Fig. 6A). When OT+ l-NAME treatment was given in the presence of SNAP, the number of contracting cultures increased from 4–5 to 17. Thus, NO supplementation partially reversed the inhibitory effect of l-NAME on OT-induced cardiomyogenesis.
Fluorescence microscopy was employed to visualize both MLC-2v-GFP staining intensity and distribution in response to the differentiation inducers. Figures 6B–6G present fluoromicrographs taken on day 10 of differentiation. Only basal fluorescence was seen in noninduced cultures (Fig. 6B). OT-treated cultures produced intense GFP fluorescence areas (Fig. 6C). OT induction of MLC-2v-GFP expressing cells was reduced in the presence of l-NAME (Fig. 6D). SNAP partially reversed the inhibitory effect of l-NAME on OT-induced CM differentiation (Fig. 6E). SNAP exhibited large areas of weak fluorescence studded with intensively-emitting spots that enlarged upon combination with OT treatment (Fig. 6F, FG). High fluorescence-emitting areas, such as those seen in OT-differentiated cultures, overlapped the sites of contracting cell colonies (Fig. 6C). In contrast, areas of weak fluorescence were not always associated with beating (e.g., OT+ l-NAME; Fig. 6D). These results indicate OT effectiveness in CM differentiation to the ventricular phenotype and NO involvement in this process.
The formation of beating cell colonies expressing CM ventricular MLC-2v marker in response to both NO generation and inhibition of the NO pathway (albeit with different efficiencies) pointed to the pleiotropic action of NO in cardiomyogenesis. In this respect, we found that OT as well as l-NAME treatments of P19EC cell monolayers elicited proliferative actions, as assessed by crystal violet staining (1.6- ± 0.12-fold and 2.0- ± 0.11-fold higher cell numbers respectively vs. noninduced cells; p < .001), whereas the NO donor SNAP decreased proliferation by 60% compared to noninduced cultures (p < .001).
Inhibition of Inducible NOS and NO-Sensitive Soluble Guanylyl Cyclase Reduces OT-Evoked Cardiomyogenesis
To better understand the NO pathway involved in OT-induced cardiomyogenesis, we studied the effect of 1400W, a specific inhibitor of inducible NOS (iNOS) , and ODQ, a selective inhibitor of NO-sensitive soluble guanylyl cyclase (sGC) . Figure 7A shows that both inhibitors decreased the number of contracting colonies generated by OT treatment (OT, 16 wells on day 14; OT+ 1400W, 7 wells; and OT+ ODQ, 8 wells) and, in the absence of OT, they stimulated rare and undersized contracting cell clusters as well as MLC-2v mediated fluorescence (Fig. 7B).
This report follows up studies on the mechanism of OT-induced stem cell differentiation into CM. In previous investigations, we have provided evidence that OTR mRNA expression is enhanced during CM differentiation of pluripotent P19EC cells . Here, we demonstrated that NO contributes to OT-induced CM differentiation through a pathway involving eNOS mRNA upregulation as well as iNOS- and NO-dependent sGC activity. Image analysis of P19Cl6-GFP cells induced by OT revealed the presence of ventricular CM coexpressing sarcomeric α-actinin and MLC-2v specific GFP, with some cells exclusively expressing α-actinin, most likely skeletal myocytes (SMs). The OT-mediated production of CM and, less evidently, the SMs was reduced by l-NAME supplementation. We found that inhibition of the NO pathways in OT-mediated CM differentiation significantly reduced the generation of functional, contracting CM, indicating that NO is a transducing molecule in the cascade of intracellular events. This was further supported by reversal of the l-NAME effect with the NO donor SNAP.
The present work shows the functional involvement of iNOS/eNOS/sGC in OT-mediated CM differentiation. This was demonstrated by blockage of the NO pathways and by OT stimulation of eNOS mRNA (stimulation inhibited by l-NAME). The iNOS transcript was undetected under our conditions, but iNOS action was indicated by sensitivity to its specific inhibitor, 1400W. NO is a ubiquitous signaling molecule, characterized by high reactivity but with a self-limiting duration of action. NO acts through cGMP-dependent and -independent pathways to regulate gene expression by modulating transcription factors, translation, or the stability of mRNA and proteins [29, –31]. Its actions are compatible with differentiation potential, and the importance of NO signaling has been demonstrated in murine embryonic stem (ES) cells expressing iNOS, eNOS, and sGC during CM differentiation . The treatment of ES cells with SNAP or their transduction with iNOS gene increased the number of spontaneously contracting cell clusters and the expression of cardiac MLC protein, and these effects were decreased by treatment with l-NAME as well as an iNOS inhibitor .
We found that NO plays an important role in OT-mediated CM maturation of P19EC cells because inhibition of the NO pathway drastically reduced beating cell colonies, Nkx2.5 and MEF2c transcripts, and contractile protein expression, although there was a relatively less pronounced reduction of GATA4 and ANP transcripts. Interestingly, OT exhibited a more efficient cardiomyogenic action than did SNAP, as assessed by the number of beating cell cultures and their size and MLC-2v associated GFP expression. This could be due to fine regulation of NO at the OTR level and/or the contribution of additional NO-independent transduction pathways triggered by OT.
The optimal OT concentration for CM induction was in the range of pharmacological concentrations (10−7 M). Under these conditions, as with other G protein-coupled receptors, it is likely that OTRs internalize and/or desensitize upon continuous agonist exposure, which was indicated in other models by changes in intracellular calcium mobilization  and associated loss of cellular responsiveness [33, 34]. Interestingly, OTR-mediated mobilization of intracellular calcium was required to upregulate NO in endothelial cells . Therefore, we could speculate that under conditions of CM differentiation of P19EC cells, OTRs undergo a similarly extensive and rapid process of desensitization and internalization after OT exposure, with changes in intracellular calcium and NO mobilization. It is possible that this cycle of events is necessary to trigger OT-mediated cardiomyogenesis. Thus, NO deficits during the OTR desensitization period can provide a part of the signaling required to initiate cardiomyogenesis. Correspondingly, we found that l-NAME alone, as well as 1400W and ODQ, two other inhibitors of NO pathways, initiated some cardiomyogenesis in P19EC cells. l-NAME downregulated the embryonic cell marker SSEA-1, and increased GATA4 and ANP transcripts as well as MLC-2v associated GFP reporter gene. l-NAME, however, induced only a few small, round clusters of weakly beating cell colonies and hardly stimulated the synthesis of MHC and DHPR markers displayed in terminally differentiated P19EC. In P19Cl6-GFP cultures, l-NAME induced some cells expressing α-actinin and SMs marker myogenin. However, l-NAME treatment did not influence the OT-mediated production of cells expressing α-actinin, while being negative for the ventricular cell marker MLC-2v. These observations indicate that, in P19 cells induced by OT, l-NAME predominantly inhibits the production of ventricular CM and has a minimal effect on simultaneously-generated populations of SMs.
It is possible that the decrease of NO by l-NAME can remove a repressive barrier on differentiation and/or provide an inductive clue to initiate myogenesis but is insufficient to complete CM differentiation and maturation. Correspondingly, in cancer cells, high NO levels may be cytostatic or cytotoxic, whereas low-level activity can have the opposite effect and promote cell growth . In this respect, we observed that proliferation of P19EC cells was stimulated by both OT and l-NAME, but was decreased by the NO donor SNAP. The effect on proliferation can be one of the mechanisms of l-NAME's cardiomyogenic action, considering that the high density of cells can induce spontaneous differentiation . Therefore, the tuning of potentially positive and negative NO actions and the molecular recognition of these balances will be central to understanding the role of NO during CM differentiation.
The role of NO signaling in ES cell-derived CM remains unsolved. Recent studies of Krumenacker et al.  indicate that, upon differentiation, timely regulated expression of many NO signaling components is evident, suggesting complex mechanisms involved in early differentiation events. Confusion arose from the fact that, in our experiments, NOS inhibition both reduced and slightly induced cardiac and skeletal myogenesis. This observation of opposite actions is not unique to our study, as dual roles for NO have been reported in other developmental contexts [38, 39]. Dual roles have also been reported for other molecules in this respect. For instance, Wnt signaling, important in the maintenance of ES cell pluripotency, similarly has positive and negative influences on cardiogenesis . In mouse P19CL6 cells, DMSO (1%) induced Wnt3A and Wnt8 expression 2 days after treatment, and expression was quickly downregulated by day 4, but it was essential for subsequent cardiogenesis. It is interesting that the Wnt pathway has recently been shown to lead to a decrease of intracellular cGMP, indicating a putative role for Wnt in the modulation of NO signaling through alleviated cGMP levels .
Endogenous OT produced in P19EC cells can act in paracrine and autocrine ways to provide the secondary stimulus triggering further steps of CM differentiation. In accordance with this hypothesis, our experiments showed that OT gene expression in P19EC cells was enhanced after OT induction and was lowered in the presence of l-NAME. OT RIA in P19EC cells revealed that, although most of the peptide was targeted for secretion, consistent amounts of OT were also found in cells, raising the additional possibility of an “intracrine” action of the hormone in cells coexpressing OT and its receptor. This OT upregulation occurring early in induced cells (day 6) could provide a means to sustain NO production (for instance, through eNOS upregulation), even after removal of the exogenously supplied hormone. In addition, the stimulatory effect of OT on P19EC cell proliferation could facilitate NO-mediated cell selection and differentiation.
It has been shown that, in DMSO-induced P19EC cells, the expression of GATA4 transcript and protein is restricted to cells committed to the cardiac lineage, and GATA4 induction precedes the expression of cardiac marker genes as well as the appearance of beating cells . GATA4 transcript was delayed in samples treated with OT+ l-NAME compared to OT alone. This observation indicates that l-NAME treatment modulates GATA4 expression and possibly functions at early stages of CM differentiation. Detailed mechanisms are still to be defined, but it has been shown that NO can alter GATA4 function through post-translational modifications .
In our studies, whereas GATA4 expression was only moderately reduced when l-NAME was administered together with OT, the factors MEF2c and Nkx2.5 were extensively downregulated. This lack of balance of transcription factors can severely impair the cardiomyogenic program, which requires physical interaction and synergistic modulation of target gene expressions [1, –3]. The imbalance may explain ANP, MHC, and DHPR expression responses to OT and NO modulators since these genes are controlled to varying degrees by GATA4, MEF2c, and Nkx2.5 regulatory elements [43, , , –47]. For example, the lower Nkx2.5-to-GATA4 ratio observed in l-NAME treated cells compared to OT-treated cells (Fig. 2) could explain the otherwise unexpected spontaneous ANP upregulation resulting from l-NAME. In line with this possibility, there is a report of ANP activation as a result of decreased binding of Nkx2.5 to the ANP promoter . OTR is another potential target of the imbalance since the presence of GATA4 and Nkx2.5 responsive elements was recently uncovered in its gene . Besides transcription factor deficits, specific conditions of cell culture, such as cell aggregation and hypoxic conditions inside EBs can play a role in gene expression . NO can elicit both pro-oxidant and antioxidant effects, and, therefore, in our experimental conditions, l-NAME supplementation and hypoxia may stimulate the formation of free radical scavengers . It was reported recently that free radical scavengers promoted GATA4 expression, although this was not sufficient to enhance the cardiomyogenesis of ES cells and MEF2c expression . Most likely, free radical generation is not required for OT-induced GATA4 expression, indicating that a finely tuned interplay between OT-dependent and -independent transcription factors may be required for proper cardiomyogenic differentiation.
In summary, we found that OT upregulated eNOS expression during differentiation of mouse P19EC cells into CM, and that inhibition of NO signaling during this process reduced CM yield. The NO generated by SNAP treatment, although cardiomyogenic, did not enhance CM differentiation induced by OT. Moreover, beating colonies were larger in OT- than in SNAP-induced cultures. This suggests that additional signaling pathway(s) besides NO are involved in OT-stimulated cardiomyogenesis. Because NO inhibition, in addition, initiates CM differentiation, we postulate that NO influences cardiac differentiation in P19EC cells through a pleiotropic mechanism.
The authors indicate no potential conflicts of interest.
We acknowledge the editorial work of Ovid Da Silva and the secretarial assistance of Antoinette Paolitto. B.D. is the recipient of a studentship from Fonds Québécois de la Recherche sur la Nature et les Technologies and a studentship from Université du Québec à Montréal. This work was supported by Canadian Institutes of Health Research and Canadian Heart and Stroke Foundation Grants MOP-53217 (J.G. and M.J.), and NET SRD-63193 (J.G., M.J., and J.P.).