miR-1 as a Key Epigenetic Regulator in Early Differentiation of Cardiac Sinoatrial Region

A large diversity of epigenetic factors, such as microRNAs and histones modifications, are known to be capable of regulating gene expression without altering DNA sequence itself. In particular, miR-1 is considered the first essential microRNA in cardiac development. In this study, miR-1 potential role in early cardiac chamber differentiation was analyzed through specific signaling pathways. For this, we performed in chick embryos functional experiments by means of miR-1 microinjections into the posterior cardiac precursors—of both primitive endocardial tubes—committed to sinoatrial region fates. Subsequently, embryos were subjected to whole mount in situ hybridization, immunohistochemistry and RT-qPCR analysis. As a relevant novelty, our results revealed that miR-1 increased Amhc1, Tbx5 and Gata4, while this microRNA diminished Mef2c and Cripto expressions during early differentiation of the cardiac sinoatrial region. Furthermore, we observed in this developmental context that miR-1 upregulated CrabpII and Rarß and downregulated CrabpI, which are three crucial factors in the retinoic acid signaling pathway. Interestingly, we also noticed that miR-1 directly interacted with Hdac4 and Calm1/Calmodulin, as well as with Erk2/Mapk1, which are three key factors actively involved in Mef2c regulation. Our study shows, for the first time, a key role of miR-1 as an epigenetic regulator in the early differentiation of the cardiac sinoatrial region through orchestrating opposite actions between retinoic acid and Mef2c, fundamental to properly assign cardiac cells to their respective heart chambers. A better understanding of those molecular mechanisms modulated by miR-1 will definitely help in fields applied to therapy and cardiac regeneration and repair.


Introduction
During early embryogenesis, the cardiogenic mesoderm organizes-at both sides of the embryo-the primitive endocardial tubes, which fuse to form the primitive heart tube in the midline [1][2][3][4].Subsequently, the posterior cardiac tube segment differentiates into the cardiac sinoatrial region, including the atrium and inflow tract/sinus venosus [5,6].Concomitantly, the anterior cardiac tube segment contributes to ventricular and outflow tract differentiation [7,8].Previous studies have revealed that this segmentation process is regulated by several cardiac transcription factors, including Tbx5, Gata4 and Mef2c, among others, which activate the expression of early cardiac structural genes, such as myosin heavy-chain (MHC) and myosin light-chain (MLC) genes [9][10][11][12][13].It is also well known that the phenotype of cardiac sinoatrial region is influenced by retinoic acid (RA) signaling, mediating the expression pattern of early molecular markers such as Tbx5, Gata4 and AMHC1 (atrial myosin heavy chain), which together determine the posterior heart tube segment, leading to venous and atrial cell fate [14][15][16].In fact, exclusion of RA from ventricular precursors is essential for correct specification of the ventricles [17].Furthermore, RA synthesis is largely controlled by retinaldehyde dehydrogenase 2 (Raldh2/Aldh1a2) [18,19], in such a way that Raldh2 −/− knockout mice present reduced Tbx5 and Gata4 expression and severely impaired cardiac sinoatrial region development, while Mef2c expression, which induces ventricular cardiomyocyte differentiation [20], was unaltered [21].Along with the above, Mef2c −/− knockout mice displayed ectopic Tbx5 expression, suggesting that the mutant ventricle acquires atrial-specific characteristics and, therefore, Mef2c could act directly on cells and make them nonresponsive to RA inflow differentiation signaling [22].
Several authors have also reported that a large diversity of epigenetic processes, including the expression of non-coding RNAs (especially microRNAs), histone modifications and DNA methylation, are capable of regulating gene expression by influencing transcription or inhibiting translation through several functional overlapping and cross-talking mechanisms [23,24].In particular, microRNAs are considered as a post-transcriptional type of epigenetic regulatory processes, since the degree of complementarity microRNA/target mRNA determines its final outcome, either by degradation or translational inhibition [24,25].Specifically, miR-1 has been the first essential microRNA identified in cardiac development, with a specific expression pattern in cardiac muscle [26][27][28][29][30][31].Interestingly, experimental overexpression of miR-1 during heart development results in defective ventricular myocyte proliferation, accompanied by hypoplasia of the cardiac ventricular conduction system [32].To date, the specific effects of miR-1 during early differentiation on cardiac sinoatrial region are unknown.The previous study that analyzed the functional role of miR-1, which was performed in murine embryonic stem cells, was focused on the development and function of sinoatrial cells, corresponding specifically to the sinoatrial node, which determines the cardiac pacemaker [33].
Furthermore, it has been proposed that miR-1 modulates cardiomyocyte growth responses by negatively regulating Calm1/Calmodulin, a miR-1 target gene [34].In addition, miR-1 regulates human cardiomyocyte progenitor cell differentiation via the repression of histone deacetylase 4 (Hdac4), also identified as a miR-1 target gene, altering histone-DNA-binding activity [35].Nevertheless, Hdac4 can also interact with non-histones by covering the promoter region of a gene, thus inhibiting its expression [36][37][38].For instance, Hdac4 can interact with Mef2c by repressing its transcriptional activity, which is required for muscle cell differentiation [39].Additionally, as described in cardiac hypertrophy, it has been reported that the activation of Mef2c transcription is mediated both by CALM1 and ERK1/2 -extracellular signal regulated kinases 1 and 2, a signal transducer in the family of mitogen activated protein kinase, MAPK- [40][41][42][43][44].In this sense, it has been recently reported, in a zebrafish model, that ERK1/2 signaling is required to promote ventricular chamber differentiation [45].
Given the complexity of those molecular mechanisms involved in cardiac chamber formation, this experimental study aims to analyze miR-1 potential functions in early differentiation of the cardiac sinoatrial region.By means of gain-and loss-of-function experiments, we show that miR-1 plays a dual role by modulating both MEF2C activity and RA signaling pathway, placed as a key epigenetic regulator of appropriate cardiac cell specification, thus promoting the initial differentiation of the cardiac sinoatrial region.

miR-1 Regulates Specific Gene Expressions during Early Differentiation of the Cardiac Sinoatrial Region
In this research work, we carried out miR-1 gain-and loss-of-function experiments by means of respective microinjections of premiR-1 and antimiR-1 into the posterior cardiac precursors-in primitive endocardial tubes-committed to sinoatrial region fates (Figure 1).Noticeably, in experimental embryos, we observed morphological alterations based on an expansion of the posterior domain of the heart after miR-1 administration, concomitantly with a reduced ventricle region (Figure 1A,B).This phenotype is similar to that observed following RA treatment and after the deletion of Mef2c [22,46].Additionally, in an experimental group, cardiac asa were dissected and subsequently subjected to RNA extraction and RT-qPCR analysis under each experimental condition.Also, we performed RT-qPCR analyses, thus validating the miR-1 gain-and loss-of-function results with respect to the control on cardiac asa and fluorescence visualization of the dissected cardiac asa (Figure 1C).

Sinoatrial Region
In this research work, we carried out miR-1 gain-and loss-of-function experiments by means of respective microinjections of premiR-1 and antimiR-1 into the posterior cardiac precursors-in primitive endocardial tubes-committed to sinoatrial region fates (Figure 1).Noticeably, in experimental embryos, we observed morphological alterations based on an expansion of the posterior domain of the heart after miR-1 administration, concomitantly with a reduced ventricle region (Figure 1A,B).This phenotype is similar to that observed following RA treatment and after the deletion of Mef2c [22,46].Additionally, in an experimental group, cardiac asa were dissected and subsequently subjected to RNA extraction and RT-qPCR analysis under each experimental condition.Also, we performed RT-qPCR analyses, thus validating the miR-1 gain-and loss-of-function results with respect to the control on cardiac asa and fluorescence visualization of the dissected cardiac asa (Figure 1C).Since, in our experimental study, the overexpression of miR-1 gave rise to an expansion of the posterior domain of the heart, we also analyzed the effects of miR-1 on RA target genes, including Amhc1, Tbx5 and Gata4, as well as Mef2c expression during early differentiation of the cardiac sinoatrial region.As illustrated in Figure 2, Supplementary Since, in our experimental study, the overexpression of miR-1 gave rise to an expansion of the posterior domain of the heart, we also analyzed the effects of miR-1 on RA target genes, including Amhc1, Tbx5 and Gata4, as well as Mef2c expression during early differentiation of the cardiac sinoatrial region.As illustrated in Figure 2, Supplementary Figures S1 and S2 and Supplementary Tables S1-S4, our results showed that miR-1 led to increased Amhc1, Tbx5 and Gata4 expressions.Also, RT-qPCR analyses support these findings (Figure 2A-C).Interestingly, miR-1 inhibited mRNA and protein levels of Mef2c gene in the cardiac sinoatrial region (Figure 2D).Supporting all the above data, our miR-1 loss-of-function experiments showed an opposite effect.of-function experiments showed an opposite effect.
Additionally, in this work, we observed that microinjections of RA-into the same areas mentioned above-positively mediated the expression of Gata4 (Supplementary Figure S3).This result coincides with the fact that RA increases Tbx5 and Amhc1 expression, as we previously reported [46], thus revealing that miR-1 mimics the effects observed after RA administration.Additionally, in this work, we observed that microinjections of RA-into the same areas mentioned above-positively mediated the expression of Gata4 (Supplementary Figure S3).This result coincides with the fact that RA increases Tbx5 and Amhc1 expression, as we previously reported [46], thus revealing that miR-1 mimics the effects observed after RA administration.

miR-1 Modulates RA Signaling Pathway during Early Differentiation of the Cardiac Sinoatrial Region
Since miR-1 modulates RA target gene expression during differentiation of cardiac sinoatrial region, through RT-qPCR, we analyzed RNAs from the dissected cardiac asa of embryos microinjected either with premiR-1 or antimiR-1, in order to explore the relationship between miR-1 and those factors involved in the RA signaling pathway (Figure 3). the endogenous expression levels of CrabpII (Figure 3B), a RA target gene that regulates RA delivery to its nuclear receptors (RARs) [48,49].In addition, our results showed that miR-1 ked to reduced levels of CrabpI (Figure 3C), which functions as a sequester of RA to facilitate its catabolism [50], thus supporting that miR-1 may upregulate RA signaling.In line with the above results, we obtained the opposite effects through antimiR-1 administration (Figure 3).With respect to the nuclear retinoic acid receptors, it is noteworthy that Rarß is a RA target gene essential for cardiovascular development [51][52][53][54].Using immunohistochemistry (IMH), in this study, we included a detailed analysis of this cardiac gene.In a control group of embryos, we identified Rarß distribution starting in both primitive endocardial tubes, followed by the primitive cardiac tube and continuing in the early cardiac looping formation, presenting a location in sinoatrial region (Supplementary Figure S4).Our miR-1 gain-of-function experiments led to an increased expression of Rarß and its protein levels (Figure 4, Supplementary Figure S1, Supplementary Table S1).Supporting these results, we obtained the opposite effect after our miR-1 loss-of-function experiments (Figure 4, right column, and RT-qPCRs; Supplementary Figure S1, Supplementary Table S2).On the other hand, expression modifications of nuclear retinoic acid receptors Rarα, Rxrα, Rarγ and Rxrγ were not detected after miR-1 experimental assays (Figure 5).Our findings revealed that miR-1 did not modify the expression levels of Raldh2 (Figure 3A), a RA synthesis modulator [47].We also observed that miR-1 was able to increase the endogenous expression levels of CrabpII (Figure 3B), a RA target gene that regulates RA delivery to its nuclear receptors (RARs) [48,49].In addition, our results showed that miR-1 ked to reduced levels of CrabpI (Figure 3C), which functions as a sequester of RA to facilitate its catabolism [50], thus supporting that miR-1 may upregulate RA signaling.In line with the above results, we obtained the opposite effects through antimiR-1 administration (Figure 3).
With respect to the nuclear retinoic acid receptors, it is noteworthy that Rarß is a RA target gene essential for cardiovascular development [51][52][53][54].Using immunohistochemistry (IMH), in this study, we included a detailed analysis of this cardiac gene.In a control group of embryos, we identified Rarß distribution starting in both primitive endocardial tubes, followed by the primitive cardiac tube and continuing in the early cardiac looping formation, presenting a location in sinoatrial region (Supplementary Figure S4).Our miR-1 gain-of-function experiments led to an increased expression of Rarß and its protein levels (Figure 4, Supplementary Figure S1, Supplementary Table S1).Supporting these results, we obtained the opposite effect after our miR-1 loss-of-function experiments (Figure 4, right column, and RT-qPCRs; Supplementary Figure S1, Supplementary Table S2).On the other hand, expression modifications of nuclear retinoic acid receptors Rarα, Rxrα, Rarγ and Rxrγ were not detected after miR-1 experimental assays (Figure 5).

miR-1 Modulates Hdac4, Calm1/Calmodulin and Erk2/Mapk1 during Early Different of the Cardiac Sinoatrial Region
Given that our experiments revealed that miR-1 decreases both mRNA and p levels of Mef2c, which is not a miR-1 target gene, we searched for pivotal miR-1 genes as potential Mef2c regulators during differentiation of cardiac sinoatrial regio ing bioinformatics analyses through Target-Scan software, we identified Hdac4, Calm modulin and Erk2/Mapk1 as putative targets of miR-1 (Supplementary Figures S5-S7 gesting that it might exert control on their post-transcriptional regulation.Our luc assays showed that luciferase signals from plasmids harboring 3′UTRs of each gen reduced with respect to the control, proving that miR-1 recognizes and directly b Hdac4, Calm1 and Erk2/Mapk1 3′UTRs, thus triggering their mRNA degradation.Su ing these findings, and showing the specificity to the predicted sites, the reduced a  Given that our experiments revealed that miR-1 decreases both mRNA and protein levels of Mef2c, which is not a miR-1 target gene, we searched for pivotal miR-1 targets genes as potential Mef2c regulators during differentiation of cardiac sinoatrial region.Using bioinformatics analyses through Target-Scan software, we identified Hdac4, Calm1/Calmodulin and Erk2/Mapk1 as putative targets of miR-1 (Supplementary Figures S5-S7), suggesting that it might exert control on their post-transcriptional regulation.Our luciferase assays showed that luciferase signals from plasmids harboring 3 ′ UTRs of each gene were reduced with respect to the control, proving that miR-1 recognizes and directly binds to Hdac4, Calm1 and Erk2/Mapk1 3 ′ UTRs, thus triggering their mRNA degradation.Supporting these findings, and showing the specificity to the predicted sites, the reduced activity by miR-1 was not observed in Hdac4, Calm1 or Erk2 3 ′ UTR reporter vectors containing the mutated sequences of the predicted miR-1 binding sites, respectively (Figure 6).
In line with these results, by means of in situ hybridization (ISH), we analyzed and identified miR-1 expression patterns in early chick embryos, starting in both primitive endocardial tubes, followed by the primitive cardiac tube and continuing in the early cardiac looping formation.By means of IMH, we observed a similar distribution of HDAC4, CALM1 and phospho-ERK2 (pERK2) in the sinoatrial region during the stages under analyses.Interestingly, our comparative analysis demonstrated a complementary expression pattern among them, suggesting a plausible regulatory link during early posterior cardiac tube differentiation.Noticeably, this pattern coincided with the cardiac distribution of MEF2C (Figure 7).
As a clear demonstration of miR-1 functional role on Mef2c regulators, our gain-offunction experiments showed that miR-1 repressed both mRNA and protein levels of Hdca4, Calm1 and Erk2 in the sinoatrial region (Figure 8, Supplementary Figure S2 and Supplementary Tables S3 and S4).In agreement with these results, the opposite effects were observed after antimiR-1 administration (Figure 8, right column, and RT-qPCRs).Therefore, our findings indicate that miR-1 modulates Mef2c expression through the repression of HDAC4, Calm1 and Erk2 during early differentiation of the cardiac sinoatrial region.
FOR PEER REVIEW 7 of 20 by miR-1 was not observed in Hdac4, Calm1 or Erk2 3′UTR reporter vectors containing the mutated sequences of the predicted miR-1 binding sites, respectively (Figure 6).In line with these results, by means of in situ hybridization (ISH), we analyzed and identified miR-1 expression patterns in early chick embryos, starting in both primitive endocardial tubes, followed by the primitive cardiac tube and continuing in the early cardiac looping formation.By means of IMH, we observed a similar distribution of HDAC4, CALM1 and phospho-ERK2 (pERK2) in the sinoatrial region during the stages under analyses.Interestingly, our comparative analysis demonstrated a complementary expression pattern among them, suggesting a plausible regulatory link during early posterior cardiac tube differentiation.Noticeably, this pattern coincided with the cardiac distribution of MEF2C (Figure 7).Finally, our results showed that miR-1 led to reduced levels of Cripto, obtaining the opposite effect after our miR-1 loss-of-function experiments (Figure 9).Since Cripto/Tdgf1, a direct transcriptional target of MEF2C [55], shows a coincident expression pattern with Mef2c during early cardiac development restricted to sinoatrial region [56,57], our findings support miR-1 modulation activity on MEF2C signaling during the differentiation of the sinoatrial region.S3 and S4).In agreement with these results, the opposite were observed after antimiR-1 administration (Figure 8, right column, and RT-q Therefore, our findings indicate that miR-1 modulates Mef2c expression through pression of HDAC4, Calm1 and Erk2 during early differentiation of the cardiac sin  Finally, our results showed that miR-1 led to reduced levels of Cripto, obtaining the opposite effect after our miR-1 loss-of-function experiments (Figure 9).Since Cripto/Tdgf1, a direct transcriptional target of MEF2C [55], shows a coincident expression pattern with Mef2c during early cardiac development restricted to sinoatrial region [56,57], our findings support miR-1 modulation activity on MEF2C signaling during the differentiation of the sinoatrial region.

Discussion
In this work, we have obtained original findings about numerous molecular factors involved in the posterior cardiac tube segment differentiation.In this sense, this is the first time there is clear evidence about the function of miR-1 modulating MEF2C activity and

Discussion
In this work, we have obtained original findings about numerous molecular factors involved in the posterior cardiac tube segment differentiation.In this sense, this is the first time there is clear evidence about the function of miR-1 modulating MEF2C activity and RA signaling pathway during early cardiac chamber differentiation.Our results revealed (Figure 10) that miR-1 administration into the posterior cardiac precursors in the primitive endocardial tubes gave rise to high expression levels of Amhc1, Tbx5 and Gata4, which have previously demonstrated to be RA target genes, essential for the early differentiation of the cardiac sinoatrial region [16,46,58].Significantly, the expression of Mef2c, a ventricular cardiomyocyte inductor [20,22], was diminished after miR-1 administration.Moreover, Cripto-a direct transcriptional target of MEF2C [55]-also was diminished after our miR-1 gain-of-function experiments.Additionally, morphological alterations were detected after miR-1 overexpression, characterized by an enlarged posterior domain of the heart and a reduced ventricle region.All the above results show that miR-1 mimics the effects previously associated to highlighted RA levels [14][15][16][17][18][19]21], indicating that miR-1 actively interferes with one or more factors involved in the RA signaling pathway.In this line, our results revealed that miR-1 upregulated CrabpII and Rarß, while it downregulated CrabpI, three crucial factors in RA signaling pathway.Interestingly, this microRNA also mimicked the results obtained after the deletion of Mef2c [20,22], indicating that miR-1 clearly modulates Mef2c transcription regulators.In addition to the above data, our results showed that: (i) miR-1 recognized and directly bonded to Erk2/Mapk1, Hdac4 and Calm1, the last two also previously reported as miR-1 target genes in culture cardiac cells [34,35]; (ii) this microRNA presented a complementary expression pattern with these three target genes during early posterior cardiac tube differentiation; and (iii) it repressed both mRNA and the protein levels of Hdac4, Calm1 and Erk2/Mapk1 in the sinoatrial region.Our findings support the fact that miR-1 actively interacts with these three potential molecular factors and that these factors are involved both in Mef2c regulation and RA signaling pathway during early differentiation of the cardiac sinoatrial region (Figure 10).

miR-1 Modulates MEF2C Activity
Our results showed that miR-1 diminished both mRNA and the protein Mef2c in the posterior domain of the developing heart.It has been proposed th promotes outflow-specific differentiation and directs cells in a nonresponsive m

miR-1 Modulates MEF2C Activity
Our results showed that miR-1 diminished both mRNA and the protein levels of Mef2c in the posterior domain of the developing heart.It has been proposed that Mef2c promotes outflow-specific differentiation and directs cells in a nonresponsive manner to the inflow-specifying actions of RA [22].Since Mef2c is not a miR-1 target gene, we analyzed the effects of miR-1 on Mef2c modulators and observed that this microRNA repressed both mRNA and the protein levels of cytoplasmatic Hdac4 (cHdac4), Calm1 and Erk2/Mapk1 (Figure 11).In this context, it is particularly known that nuclear HDAC4 (nHDAC4) directly binds to Mef2c and blocks its activity [36,39,59].Through pharmacological inhibition of Hdac4 activity by trichostatin A (TSA) [60], our experimental model (Supplementary Figure S8) revealed that Mef2c expression presented a dramatic increase, proving the importance of histone deacetylation in this transcription factor regulation.It is also known that Calm1-through CaMKII induction-leads to the phosphorylation of nHDAC4 and its translocation to the cytoplasm, thus promoting Mef2c activation [40][41][42]44,61].Therefore, in our model (Figure 11), given that low levels of Calm1--miR-1-induced-did not allow the export of nHDAC4 from the nucleus to the cytoplasm, miR-1 would increase Mef2c's interaction with nHDAC4, which would inhibit Mef2c expression.It has also been reported that nHDAC4, together with the nuclear corepressor NCoR1/SMRT, synergize to inhibit Mef2c activity [62,63].Furthermore, NCoR1/SMRT actively translocates cHDAC4 from the cytoplasm into the nucleus and prevents its nuclear export [63][64][65].The above data support the fact that there are opposite effects of CALM1 and NCoR1/SMRT on nHDAC4 modulated by miR-On the other hand, it is known that NCoR1/SMRT phosphorylation by ERK2 may destabilize the association between Mef2c and nHDAC4, thus increasing Mef2c expression [63,66,67].It has also been reported that Mef2c may be enhanced by ERK1/2 through the upregulation of coactivator P300/CBP-associated factor (PCAF), which inherently presents histone acetyltransferases (HAT) activity, thus increasing both the expression and function of Mef2c [43].Our results showed that miR-1 directly recognized and repressed Erk2/Mapk1 expression (Figure 9), and consequently, this microRNA diminished MEF2C activity; these results are supported by the above data.In this line, it has been observed during zebrafish development that excessive levels of ERK1/2 in atrial location cause the ectopic expression of ventricular specific genes [45], an effect that also supports our model, proposing miR-1 as a key regulator of atrial differentiation by repressing Erk1/2 expression.
Based on all the molecular mechanisms mentioned above, we propose in our model (Figure 11) that, in these early stages of cardiac development, miR-1 plays a crucial role to repress Mef2c through Erk2/Mapk1, Hdac4 and Calm1 modulation.
Noticeably, our results showed that miR-1 diminished the expression level of Cripto.It is known that MEF2C is required for Cripto enhancer activity and expression during cardiac development [55].Also, it has been reported in in vitro models during cardiac differentiation, that high levels of Cripto gene expression coincide with low levels of miR-1 and vice versa, establishing a crosstalk between miR-1 and Cripto during cardiomyogenesis [68].Furthermore, in mice, Cripto is a direct target of miR-1; thus, when miR-1 binds to the 3 ′ UTR Cripto, it suppresses its expression and promotes cardiomyocyte specification.Nevertheless, in silico analysis reveals that Cripto is not a direct target of miR-1 in chicken, thus suggesting that this interaction could be mediated through MEF2C.It has also been reported that Cripto is strongly expressed in undifferentiated cells, and it is rapidly downregulated during retinoic acid-induced differentiation [69].Therefore, in our model, miR-1 could be able to promote the RA signaling pathway to induce the early differentiation of the cardiac sinoatrial region.
tion.Nevertheless, in silico analysis reveals that Cripto is not a direct target of miR-1 in chicken, thus suggesting that this interaction could be mediated through MEF2C.It has also been reported that Cripto is strongly expressed in undifferentiated cells, and it is rapidly downregulated during retinoic acid-induced differentiation [69].Therefore, in our model, miR-1 could be able to promote the RA signaling pathway to induce the early differentiation of the cardiac sinoatrial region. .In these molecular mechanisms, miR-1 modulates opposite effects between Calm1 and NCoR1/SMRT on nuclear HDAC4 (nHDAC4), thus increasing Mef2c's interaction with nHDAC4, which inhibits Mef2c expression.Also, miR-1 suppresses Erk2/Mapk1 and, consequently, diminishes MEF2C activity.Additionally, our model indicates that RA function is modulated by miR-1, thus promoting the expression of RA target genes Tbx5, Gata4, Amhc1 and Rarβ.Moreover, miR-1 increases CrabpII and suppresses CrabpI, thus increasing RA activity.The presence of RA disrupts the interaction of nHDAC4-NCoR1/SMRT with RARs, thus allowing transcription of RARs target genes.Also, miR-1 represses Calm1's capability to inhibit RARs' activity, thus enhancing the expression of RA target genes.In our model, miR-1 modulates the opposite actions between RA and MEF2C, promoting the RA signaling pathway and suppressing MEF2C activity to allow properly assign cells to their cardiac chamber.OFT: outflow tract, V: ventricular region, SA: sinoatrial region, P: phosphorylate.Our model indicates that miR-1 plays a crucial role repressing Mef2c through modulation of Hdac4, Calm1 and Erk2/Mapk1.In these molecular mechanisms, miR-1 modulates opposite effects between Calm1 and NCoR1/SMRT on nuclear HDAC4 (nHDAC4), thus increasing Mef2c's interaction with nHDAC4, which inhibits Mef2c expression.Also, miR-1 suppresses Erk2/Mapk1 and, consequently, diminishes MEF2C activity.Additionally, our model indicates that RA function is modulated by miR-1, thus promoting the expression of RA target genes Tbx5, Gata4, Amhc1 and Rarβ.Moreover, miR-1 increases CrabpII and suppresses CrabpI, thus increasing RA activity.The presence of RA disrupts the interaction of nHDAC4-NCoR1/SMRT with RARs, thus allowing transcription of RARs target genes.Also, miR-1 represses Calm1's capability to inhibit RARs' activity, thus enhancing the expression of RA target genes.In our model, miR-1 modulates the opposite actions between RA and MEF2C, promoting the RA signaling pathway and suppressing MEF2C activity to allow properly assign cells to their cardiac chamber.OFT: outflow tract, V: ventricular region, SA: sinoatrial region, P: phosphorylate.

miR-1 Modulates the RA Signaling Pathway
With respect to miR-1 inductive role on the RA signaling pathway during the early differentiation of the cardiac sinoatrial region (Figure 10), we found that this microRNA increased the expression levels of RA target genes Tbx5, Gata4 and Amhc1.It is known that RA establishes an epigenetic switch for histone acetylation, mediated by p300/CBP, which inherently presents histone acetyltransferases (HAT) activity to allow transcription of RARs target genes [70][71][72][73][74][75].We also observed that miR-1 repressed epigenetic repressor Hdac4, which has been reported as a negative regulator of RA target genes [76].In line with the above, we showed in this work that Tbx5, Gata4 and Amhc1 expressions increased with HDAC4 inhibitor TSA (Supplementary Figure S8), confirming that RA regulates these cardiac genes through direct modulation of histone deacetylation.Furthermore, previous studies have reported that nHDAC4 is recruited and retained in the nucleus by NCoR1/SMRT, which is associated to RARs in the absence of RA [64,[77][78][79].As a matter of fact, it is known that the presence of RA disrupts the interaction of NCoR1/SMRT with RARs [67,80], thus allowing the transcription of RARs' target genes (Figure 11).Other studies have described that ERK2/MAPK1 phosphorylates nuclear corepressors N-CoR1/SMRT, reducing the interaction between N-CoR1/SMRT and RARα [66], which would allow us to suggest an increase of RARs target genes transcription.However, our results showed that miR-1 directly suppressed Erk2/Mapk1 and increased RA target genes.Based on these results, we propose that Erk2/Mapk1 is not sufficient to modulate atrial gene expressions, so an RA signal is necessary to disrupt the interaction between N-CoR1/SMRT and RARs [77].Keeping all the above in mind, our results showed that miR-1 upregulated RA function, thus promoting the expression of its target genes.
Interestingly, in our study, we found that miR-1 increased the expression levels of CrabpII, a RA target gene-with an RA response element (RARE)-involved in delivering RA to its nuclear receptors RARs [48,49,81,82].In addition, we observed that miR-1 increased both mRNA and the proteins of Rarβ, a previously defined RA target gene with RARE located in its promoter [51,54,73,80,82].Moreover, by means of IMH, we carried out a detailed dynamic analysis of RARβ distribution, which showed a specific location in the cardiac sinoatrial region.The above findings, together with the fact that RARβ repression is linked to cardiac abnormalities [52], support miR-1 active role on RA signaling pathway through Rarβ.Nevertheless, the levels of nuclear retinoic acid receptors Rarα, Rxrα, Rarγ and Rxrγ did not show significant changes after our miR-1 experimental assays, proving that these receptors are not targeted either by miR-1 or RA.It is noteworthy, based on our results, that miR-1 directly suppresses Caml1, a protein that also acts as an inhibitor of RARα activity by means of CaMKII induction, which mediates RARα phosphorylation and enhances the interaction between RARα and N-CoR1/SMRT, and subsequently supresses RARα target genes transcription [83].In our proposed model (Figure 11), miR-1 repressed CALM1's capability to inhibit RARs' activity, thus promoting the expression of RA target genes.Supporting these results, previous studies have shown that the blockage of RARα function diminishes Gata4 transcripts, specifically interfering with the inflow tract formation [84,85].
Additionally, our results show that miR-1 overexpression did not generate significant changes in the expression level of Raldh2, a RA synthesis modulator [47,79,86] involved in cardiac sinoatrial region development [18,19].Supporting our results, previous studies [87] have shown that Raldh2 expression is not altered in a nuclear corepressor mutant mouse model-SMRTmRID-which is characterized by enhanced transcription of RARs targets.
Noticeably, we observed that miR-1 diminished the expression levels of CrabpI, a protein described as a sequester of RA, facilitating its catabolism, limiting RA concentration and regulating the amount of RA that is accessible to nuclear receptors RARs [50,88].This supports the fact that miR-1 enhances RA activity.Given that CrabpI is not a miR-1 target gene, our model (Figure 11) proposes that miR-1 inhibits Med1/Trap220 (a miR-1 target gene identified through in silico analysis), which is a CrabpI coactivator associated with RA [89].Therefore, miR-1 would be able to promote RA signals.
Taking into account all the above, our novel experimental model integrates a network of molecular mechanisms modulated by miR-1, which plays a crucial role during early stages of cardiac chamber formation by promoting atrial differentiation together with complementary suppression of ventricular formation.In conclusion, our study reveals, for the first time, a key role of miR-1 as an epigenetic factor modulating RA and Mef2c in their opposite actions, which are required to properly assign cells to their respective cardiac chambers.Since miR-1 has also been identified in cardiomyopathy processes [90][91][92][93][94][95], further and deeper understanding of this microRNA as a modulator of molecular mechanisms governing specific signaling pathways could be helpful in therapy and cardiac regeneration and repair.

Materials and Methods
Experimental protocols with animals were performed in agreement with the Spanish law in application of the EU Guidelines for animal research and conformed to the Guide for the Care and Use of Laboratory Animals, published by the US National Institutes of Health (NIH Publication no.85-23).Approval by the University of Extremadura bioethics board was obtained prior to the initiation of the study.

Embryo Microinjections into the Posterior Cardiac Precursors of Both Primitive Endocardial Tubes
Stage HH 8 (3-4 somites) cultured embryos were microinjected (using an Inject + Matic micro injector system) in both primitive endocardial tubes into the posterior cardiac precursors committed to sinoatrial region fates [19,46].
For gain-of-function experiments, two different groups of embryos were microinjected with premiR-1 (Thermo Fisher AM17150-Vacaville, CA, USA) and all-trans retinoic acid (RA; Sigma R2625-St.Louis, MO, USA), respectively.premiR is a precursor miRNA that contains single stem-looped structure to be further processed into mature miRNA, which takes place in the miRNA-induced silencing complex.This, based on the sequence complementary between target mRNA and miRNA, gives rise to translation repression or target RNA degradation [99,100].
After 10-12h of additional incubation, embryos were photographed under bright and fluorescent light (Nikon digital, SIGHT DS-U1) and were selected according to the location and extent of the injection.The selected embryos were either fixed in 4% PFA-processed for gene expression (in situ hybridization, ISH) or immunochemistry (IMH) analysis-or cardiac loops were collected for RNA isolation as previously described [46].

RNA Isolation and RT-qPCR
Cardiac loops isolated from experimental and control (CFDA) embryos (Figure 1) were subjected to RT-qPCR analysis following MIQE guidelines [105][106][107].RNA was extracted and purified using ReliaPrep RNA Cell Miniprep System Kit (Promega) according to the manufacturer's instructions.For mRNA expression measurements, 0.5 µg of total RNA was used for retro-transcription with Maxima First Strand cDNA Synthesis Kit for RT-qPCR (Thermo Fisher Vicaville USA).Real-time PCR experiments were performed with 2 µL of cDNA, GoTaq qPCR Master Mix (Promega) and corresponding primer sets (Supplementary Table S5).For miR-1 expression analyses, 20 ng of total RNA was used for retro-transcription with Universal cDNA Synthesis Kit II (Exiqon), and the resulting cDNA was diluted 1/80.Real-time PCR experiments were performed with 1 µL of diluted cDNA, Go Taq qPCR Master Mix (Promega) as well.All RT-qPCRs were performed using a CFX384TM thermocycler (Bio-Rad) following the manufacture's recommendations.The relative expression of each gene was calculated using Gusb and Gapdh as internal controls for mRNA expression analyses and 5S and 6U for miR-1 expression analyses, respectively [108].Each PCR reaction was carried out in triplicate and repeated in at least three distinct biological samples to obtain representative means.

Analysis In Silico
Target-Scan 8 software was used to perform bioinformatics analysis of the binding sites of miR-1 at 3 ′ UTRs of predicted targets as described before [109,110].

Luciferase Assays and 3T3 Transfection
For luciferase assays, Hdac4, Calm1/Calmodulin and Erk2/Mapk1 3 ′ untranslated regions (UTRs) were amplified from mouse genomic DNA and cloned into the pGLuc-Basic vector (New England BioLabs Ipswich England).Hdac4, Calm1 and Erk2/Mapk1 gene 3 ′ UTRs were amplified from chicken genomic DNA with specific primers bearing HindIII/BamHI restriction sites and cloned into the pGLuc-Basic vector (New England BioLabs).PCR-based site-directed mutagenesis was performed using the Stratagene QuikChange site-directed mutagenesis kit but with the enzymes and buffers from the Bio-Rad iPROOF PCR kit.Primers used for site-directed mutagenesis (Supplementary Table S6) introduced mutations into miR-1 seed sequence present in the Hdac4 3 ′ UTR, Calm1 3 ′ UTR and Erk2/Mapk1 3 ′ UTR.Independent cotransfection experiments with pre-miRNAs were carried out simultaneously in 3T3 cells with 20 µL of culture medium; luciferase activity was measured 24 h after transfection using a Pierce™ Gaussia Luciferase Flash Assay Kit or a Pierce™ Cypridine Luciferase Flash Assay Kit (Thermo Fisher Vicaville USA).In all cases, transfections were carried out in triplicate.

Statistical Analyses
For statistical analyses of datasets, unpaired Student's t-tests were used as previously reported [111,112].Significance levels or p-values are stated in each corresponding figure legend.A p < 0.05 was considered statistically significant.

Image Analysis
Quantitative analysis of images has been performed using the Image J ® software (v.1342).

Figure 1 .
Figure 1.(A) Embryo microinjected with CFDA (control), premiR-1 or antimiR-1 at the level of the prospective atrial cells of both primitive endocardial tubes (arrowheads) and light microscopy of control and experimental embryos (HH11) subjected to miR-1 gain-and loss-of-function.Note, in the experimental embryos and dissected cardiac asa (HH11), the morphological alterations based on an expansion of the posterior domain of the heart after miR-1 administration, concomitantly with a reduced ventricle region.The opposite effects were observed after antimiR-1 administration.(B) Area percentage of structural SA and V after miR-1 gain-and loss-of-function experiments.(C) RT-qPCR validating the miR-1 gain-and loss-of-function results respect to control on cardiac asa and fluorescence visualization of dissected cardiac asa.OFT: outflow tract; V: ventricle; SA: sinoatrial region.Student's t-test: *** p < 0.005, **** p < 0.0001 with respect to control (CFDA) embryos.

Figure 1 .
Figure 1.(A) Embryo microinjected with CFDA (control), premiR-1 or antimiR-1 at the level of the prospective atrial cells of both primitive endocardial tubes (arrowheads) and light microscopy of control and experimental embryos (HH11) subjected to miR-1 gain-and loss-of-function.Note, in the experimental embryos and dissected cardiac asa (HH11), the morphological alterations based on an expansion of the posterior domain of the heart after miR-1 administration, concomitantly with a reduced ventricle region.The opposite effects were observed after antimiR-1 administration.(B) Area percentage of structural SA and V after miR-1 gain-and loss-of-function experiments.(C) RT-qPCR validating the miR-1 gain-and loss-of-function results respect to control on cardiac asa and fluorescence visualization of dissected cardiac asa.OFT: outflow tract; V: ventricle; SA: sinoatrial region.Student's t-test: *** p < 0.005, **** p < 0.0001 with respect to control (CFDA) embryos.

Figure 2 .
Figure 2. Effects of miR-1 gain-and loss-of-function experiments during the early differentiation of the cardiac sinoatrial region.Whole-mount in situ hybridization (ISH) for Amhc1 (A) Tbx5 and (B) Gata4 (C) and immunohistochemistry (IMH) for MEF2C (D).Note the increased and expanded expression of AMHC1, Tbx5 and Gata4 after premiR-1 treatment (green arrows), accompanied by diminished protein levels of MEF2C (orange arrow).The posterior diminished expressions of Amhc1, Tbx5 and Gata4 in the heart tube after antimiR-1 treatment are indicated by orange arrows, whereas Mef2c was increased (green arrows).The left side illustrates the RT-qPCR of RNA from the dissected cardiac asa of embryos microinjected either with CFDA, premiR-1 or antimiR-1.A high level of miR-1 led to increased Amhc1, Tbx5 and Gata4 transcripts, whereas miR-1 inhibition led to Mef2c increased transcripts.V: ventricle; SA: sinoatrial region.Standard deviations are from three independent experiments.Student's t-test: * p < 0.05, *** p < 0.005, **** p < 0.001 with respect to control (CFDA) embryos.

Figure 2 .
Figure 2. Effects of miR-1 gain-and loss-of-function experiments during the early differentiation of the cardiac sinoatrial region.Whole-mount in situ hybridization (ISH) for Amhc1 (A) Tbx5 and (B) Gata4 (C) and immunohistochemistry (IMH) for MEF2C (D).Note the increased and expanded expression of AMHC1, Tbx5 and Gata4 after premiR-1 treatment (green arrows), accompanied by diminished protein levels of MEF2C (orange arrow).The posterior diminished expressions of Amhc1, Tbx5 and Gata4 in the heart tube after antimiR-1 treatment are indicated by orange arrows, whereas Mef2c was increased (green arrows).The left side illustrates the RT-qPCR of RNA from the dissected cardiac asa of embryos microinjected either with CFDA, premiR-1 or antimiR-1.A high level of miR-1 led to increased Amhc1, Tbx5 and Gata4 transcripts, whereas miR-1 inhibition led to Mef2c increased transcripts.V: ventricle; SA: sinoatrial region.Standard deviations are from three independent experiments.Student's t-test: * p < 0.05, *** p < 0.005, **** p < 0.001 with respect to control (CFDA) embryos.

Figure 3 .
Figure 3. Effects of miR-1 gain-and loss-of-function experiments on RA pathway during posterior differentiation of cardiac tube.RT-qPCR of RNA from the dissected cardiac asa of embryos microinjected either with CFDA, premiR-1 or antimiR-1, illustrating Raldh2 (A), CrabpII (B) and CrabpI (C).A high level of miR-1 led to an increased expression of CrabpII (B) and a decreased expression of CrabpI (C).The opposite effect was observed with anti-miR1 treatment.Note that nonsignificant expression of Raldh2 (A) was observed after miR-1 overexpression or miR-1 inhibitor treatment, as compared to the control (CFDA).Standard deviations are from three independent experiments.Student's t-test: * p < 0.05, ** p < 0.01, **** p < 0.001 with respect to control (CFDA) embryos.

Figure 3 .
Figure 3. Effects of miR-1 gain-and loss-of-function experiments on RA pathway during posterior differentiation of cardiac tube.RT-qPCR of RNA from the dissected cardiac asa of embryos microinjected either with CFDA, premiR-1 or antimiR-1, illustrating Raldh2 (A), CrabpII (B) and CrabpI (C).A high level of miR-1 led to an increased expression of CrabpII (B) and a decreased expression of CrabpI (C).The opposite effect was observed with anti-miR1 treatment.Note that nonsignificant expression of Raldh2 (A) was observed after miR-1 overexpression or miR-1 inhibitor treatment, as compared to the control (CFDA).Standard deviations are from three independent experiments.Student's t-test: * p < 0.05, ** p < 0.01, **** p < 0.001 with respect to control (CFDA) embryos.

Figure 4 .
Figure 4. Embryos subjected to Rarß study.Whole-mount IMH illustrate RARß markedly increased in the sinoatrial (SA) region after miR-1 administration (green arrows).Note RARß diminished after antimiR-1 administration (orange arrows).RT-qPCR of RNA from the dissected cardiac asa (left side) show the levels of Rarß transcripts.V: ventricle.Standard deviations are from three independent experiments.Student's t-test: ** p < 0.01 with respect to control (CFDA) embryos.

Figure 4 .
Figure 4. Embryos subjected to Rarß study.Whole-mount IMH illustrate RARß markedly increased in the sinoatrial (SA) region after miR-1 administration (green arrows).Note RARß diminished after antimiR-1 administration (orange arrows).RT-qPCR of RNA from the dissected cardiac asa (left side) show the levels of Rarß transcripts.V: ventricle.Standard deviations are from three independent experiments.Student's t-test: ** p < 0.01 with respect to control (CFDA) embryos.

Figure 4 .
Figure 4. Embryos subjected to Rarß study.Whole-mount IMH illustrate RARß markedly in in the sinoatrial (SA) region after miR-1 administration (green arrows).Note RARß diminish antimiR-1 administration (orange arrows).RT-qPCR of RNA from the dissected cardiac side) show the levels of Rarß transcripts.V: ventricle.Standard deviations are from three ind ent experiments.Student's t-test: ** p < 0.01 with respect to control (CFDA) embryos.

Figure 5 .
Figure 5. Effects of miR-1 gain-and loss-of-function experiments on RA pathway during p differentiation of cardiac tube.RT-qPCR of RNA from the dissected cardiac asa of embr croinjected either with CFDA, premiR-1 or antimiR-1.Note that nonsignificant expression (A), Rxrα (B), Rarγ (C) and Rxrγ (D) were observed after miR-1 overexpression or miR-1 in treatment as compared to the control (CFDA).Standard deviations are from three independ periments.

Figure 5 .
Figure 5. Effects of miR-1 gain-and loss-of-function experiments on RA pathway during posterior differentiation of cardiac tube.RT-qPCR of RNA from the dissected cardiac asa of embryos microinjected either with CFDA, premiR-1 or antimiR-1.Note that nonsignificant expression of Rarα (A), Rxrα (B), Rarγ (C) and Rxrγ (D) were observed after miR-1 overexpression or miR-1 inhibitor treatment as compared to the control (CFDA).Standard deviations are from three independent experiments.

Figure 7 .
Figure 7. Whole-mount ISH for miR-1 and whole-mount IMH for HDAC4, CALM1, phosph (pERK2) and MEF2C during early chick cardiac development, from HH8 through HH11, in embryos.Note the miR-1 expression pattern in both primitive endocardial tubes (PET), b servable in the sinoatrial (SA) region and ventricle (V) at later stages.Note the location of H CALM1, pERK2 and MEF2C in the PET and, subsequently, in the SA region.Arrows: positiv

Figure 7 .
Figure 7. Whole-mount ISH for miR-1 and whole-mount IMH for HDAC4, CALM1, phospho-ERK2 (pERK2) and MEF2C during early chick cardiac development, from HH8 through HH11, in control embryos.Note the miR-1 expression pattern in both primitive endocardial tubes (PET), being observable in the sinoatrial (SA) region and ventricle (V) at later stages.Note the location of HDAC4, CALM1, pERK2 and MEF2C in the PET and, subsequently, in the SA region.Arrows: positive stains.

Figure 9 .
Figure 9. Effects of miR-1 gain-and loss-of-function experiments on Cripto during posterior differentiation of cardiac tube.RT-qPCR of RNA from dissected cardiac asa of embryos microinjected either with CFDA, premiR-1 or antimiR-1.A high level of miR-1 led to the decreased expression of Cripto.The opposite effect was observed with anti-miR1 treatment.Standard deviations are from three independent experiments.Student's t-test: * p < 0.05, *** p < 0.005 with respect to contro (CFDA) embryos.

Figure 9 .
Figure 9. Effects of miR-1 gain-and loss-of-function experiments on Cripto during posterior differentiation of cardiac tube.RT-qPCR of RNA from dissected cardiac asa of embryos microinjected either with CFDA, premiR-1 or antimiR-1.A high level of miR-1 led to the decreased expression of Cripto.The opposite effect was observed with anti-miR1 treatment.Standard deviations are from three independent experiments.Student's t-test: * p < 0.05, *** p < 0.005 with respect to control (CFDA) embryos.

Figure 11 .
Figure 11.Proposed model illustrating the network of molecular mechanisms governed by miR-1 during early differentiation of the cardiac sinoatrial region.Solid lines indicate our results from this work.Dashed lines correspond to previous studies in different research fields referred in the "Discussion" section.Our model indicates that miR-1 plays a crucial role repressing Mef2c through modulation of Hdac4, Calm1 and Erk2/Mapk1.In these molecular mechanisms, miR-1 modulates opposite effects between Calm1 and NCoR1/SMRT on nuclear HDAC4 (nHDAC4), thus increasing Mef2c's interaction with nHDAC4, which inhibits Mef2c expression.Also, miR-1 suppresses Erk2/Mapk1 and, consequently, diminishes MEF2C activity.Additionally, our model indicates that RA function is modulated by miR-1, thus promoting the expression of RA target genes Tbx5, Gata4, Amhc1 and Rarβ.Moreover, miR-1 increases CrabpII and suppresses CrabpI, thus increasing RA activity.The presence of RA disrupts the interaction of nHDAC4-NCoR1/SMRT with RARs, thus allowing transcription of RARs target genes.Also, miR-1 represses Calm1's capability to inhibit RARs' activity, thus enhancing the expression of RA target genes.In our model, miR-1 modulates the opposite actions between RA and MEF2C, promoting the RA signaling pathway and suppressing MEF2C activity to allow properly assign cells to their cardiac chamber.OFT: outflow tract, V: ventricular region, SA: sinoatrial region, P: phosphorylate.

Figure 11 .
Figure 11.Proposed model illustrating the network of molecular mechanisms governed by miR-1 during early differentiation of the cardiac sinoatrial region.Solid lines indicate our results from this work.Dashed lines correspond to previous studies in different research fields referred in the Section 3.Our model indicates that miR-1 plays a crucial role repressing Mef2c through modulation of Hdac4, Calm1 and Erk2/Mapk1.In these molecular mechanisms, miR-1 modulates opposite effects between Calm1 and NCoR1/SMRT on nuclear HDAC4 (nHDAC4), thus increasing Mef2c's interaction with nHDAC4, which inhibits Mef2c expression.Also, miR-1 suppresses Erk2/Mapk1 and, consequently, diminishes MEF2C activity.Additionally, our model indicates that RA function is modulated by miR-1, thus promoting the expression of RA target genes Tbx5, Gata4, Amhc1 and Rarβ.Moreover, miR-1 increases CrabpII and suppresses CrabpI, thus increasing RA activity.The presence of RA disrupts the interaction of nHDAC4-NCoR1/SMRT with RARs, thus allowing transcription of RARs target genes.Also, miR-1 represses Calm1's capability to inhibit RARs' activity, thus enhancing the expression of RA target genes.In our model, miR-1 modulates the opposite actions between RA and MEF2C, promoting the RA signaling pathway and suppressing MEF2C activity to allow properly assign cells to their cardiac chamber.OFT: outflow tract, V: ventricular region, SA: sinoatrial region, P: phosphorylate.
Contributions: The experimental study was designed and supervised by C.L.-S., A.A., D.F. and V.G.-M., who also wrote the manuscript.C.G.-P., E.L.-V.and V.G.-L.contributed to the conception and design of the study, performed the experiments, and analyzed the data.All authors have contributed to the experimental work and analyzed and discussed the results.All authors have read and agreed to the published version of the manuscript.