Inhibition of RhoA and Cdc42 by miR-133a Modulates Retinoic Acid Signalling during Early Development of Posterior Cardiac Tube Segment

It is well known that multiple microRNAs play crucial roles in cardiovascular development, including miR-133a. Additionally, retinoic acid regulates atrial marker expression. In order to analyse the role of miR-133a as a modulator of retinoic acid signalling during the posterior segment of heart tube formation, we performed functional experiments with miR-133a and retinoic acid by means of microinjections into the posterior cardiac precursors of both primitive endocardial tubes in chick embryos. Subsequently, we subjected embryos to whole mount in situ hybridisation, immunohistochemistry and qPCR analysis. Our results demonstrate that miR-133a represses RhoA and Cdc42, as well as Raldh2/Aldh1a2, and the specific atrial markers Tbx5 and AMHC1, which play a key role during differentiation. Furthermore, we observed that miR-133a upregulates p21 and downregulates cyclin A by repressing RhoA and Cdc42, respectively, thus functioning as a cell proliferation inhibitor. Additionally, retinoic acid represses miR-133a, while it increases Raldh2, Tbx5 and AMHC1. Given that RhoA and Cdc42 are involved in Raldh2 expression and that they are modulated by miR-133a, which is influenced by retinoic acid signalling, our results suggest the presence of a negative feedback mechanism between miR-133a and retinoic acid during early development of the posterior cardiac tube segment. Despite additional unexplored factors being possible contributors to this negative feedback mechanism, miR-133a might also be considered as a potential therapeutic tool for the diagnosis, therapy and prognosis of cardiac diseases.


Introduction
Cardiac development is a complex morphogenetic process originating at the cardiac precursor cell level, located in the epiblast and primitive streak during early chick gastrulation stages. Precardiac cells invaginate through the primitive streak and migrate anterolaterally to both sides of the embryo, constituting the precardiac mesoderm between the ectoderm and the adjacent inductive endoderm [1][2][3][4][5][6][7]. In later stages, these fields form both primitive endocardial tubes, then fusing in the midline into a single heart tube and developing into the cardiac looping. Its anterior and posterior segments differentiate simultaneously, giving rise to the common ventricle and outflow tract, and the common atrium and inflow tract, respectively [8][9][10][11][12].
These morphogenetic processes are monitored by different molecular signalling pathways, including retinoic acid (RA), which has been demonstrated as necessary to define the

Results
In this research study, we obtained several results that altogether provide original data about RA pathway signalling modulatory factors during differentiation of the posterior cardiac tube segment. Our findings revealed that miR-133a plays a crucial role in RA signalling during the early genetic programme of the sino-atrial region. A novel contribution of this work also includes the analysis of miR-133a expression pattern during early cardiac looping. Moreover, in this process, we have identified the RhoA and Cdc42 distribution and Rho GTPases previously reported [32] as modulators of Raldh2, which is necessary for RA synthesis, during heart development. Through gain-of-function experiments, we observed that miR-133a downregulates RhoA and Cdc42 directly, decreasing Raldh2, and that the expression of specific atrial gene Tbx5 and AMHC1 are also diminished. Additionally, we found that miR-133a modulates p21 and cyclin A, previously reported in cell proliferation processes [34,35].
In this work, we further analysed the effect of RA administration on miR-133a expression, as well as on Raldh2, and atrial specific genes, Tbx5 and AMHC1, expression.

miR-133a Represses RA Synthesis by RhoA and Cdc42 Modulation
In order to understand miR-133a modulating mechanisms involved in RA signalling during posterior cardiac tube differentiation, we analysed the miR-133a targets. Bioinformatics analyses through Target-Scan software show that RhoA and Cdc42 3 UTRs are putative targets of miR-133a (Supplementary Figure S1), suggesting that they might exert post-transcriptional regulation of both Rho GTPases. Our luciferase assays show that luciferase signals from plasmids harbouring 3 UTR of each gene are reduced with respect to control, demonstrating that miR-133a is able to recognise and directly bind to RhoA and Cdc42 3 UTR, triggering their mRNA degradation (Figure 1).
In this work, we further analysed the effect of RA administration on miR-133a expression, as well as on Raldh2, and atrial specific genes, Tbx5 and AMHC1, expression.

miR-133a Represses RA Synthesis by RhoA and Cdc42 Modulation
In order to understand miR-133a modulating mechanisms involved in RA signalling during posterior cardiac tube differentiation, we analysed the miR-133a targets. Bioinformatics analyses through Target-Scan software show that RhoA and Cdc42 3′UTRs are putative targets of miR-133a (Supplementary Figure S1), suggesting that they might exert post-transcriptional regulation of both Rho GTPases. Our luciferase assays show that luciferase signals from plasmids harbouring 3′UTR of each gene are reduced with respect to control, demonstrating that miR-133a is able to recognise and directly bind to RhoA and Cdc42 3′UTR, triggering their mRNA degradation ( Figure 1).

Figure 1.
Representative data of RhoA (A) and Cdc42 (B) 3′UTR luciferase assays after premiR-133a overexpression in 3T3 fibroblasts. Luciferase activity was compared to non-transfected controls. Each luciferase assay was carried out in triplicate. Student's t-test: ** p < 0.01. Figure 2, by using in situ hybridisation (ISH), we report the expression pattern of miR-133a, from primitive endocardial tubes to the early formation of cardiac asa, where the intensity is decreasing in the anterior-posterior sense, disappearing in the inflow tract region. On the other hand, by using immunohistochemistry (IMH), we observe the location of RhoA and Cdc42, at the same stages, being more intense in the inflow tract region and decreasing in the posterior-anterior sense. This pattern coincides with the cardiac distribution of Raldh2. Figure 1. Representative data of RhoA (A) and Cdc42 (B) 3 UTR luciferase assays after premiR-133a overexpression in 3T3 fibroblasts. Luciferase activity was compared to non-transfected controls. Each luciferase assay was carried out in triplicate. Student's t-test: ** p < 0.01.

As illustrated in
As illustrated in Figure 2, by using in situ hybridisation (ISH), we report the expression pattern of miR-133a, from primitive endocardial tubes to the early formation of cardiac asa, where the intensity is decreasing in the anterior-posterior sense, disappearing in the inflow tract region. On the other hand, by using immunohistochemistry (IMH), we observe the location of RhoA and Cdc42, at the same stages, being more intense in the inflow tract region and decreasing in the posterior-anterior sense. This pattern coincides with the cardiac distribution of Raldh2.
To analyse the functional role of these complementary anterior-posterior patterns, we performed miR-133a gain-and loss-of-functions experiments by injecting premiR-133a and antimiR-133a, respectively, into the posterior cardiac precursors of both primitive endocardial tubes ( Figure 3). The control and experimental embryos were analysed with anti-RhoA and anti-Cdc42 antibodies. Furthermore, a group of cardiac asa were dissected (Supplementary Figure S2) and subsequently subjected to RNA extraction and qPCR analysis. Our results show that miR-133a represses RNA and protein levels of RhoA and Cdc42 genes in the atrium and sinus venosus ( Figure 3B-C). Supporting these results, our loss-of-functions experiments by means of antimiR-133a demonstrate the opposite effects. In line with our gain-and loss-of-function experiments, we found that miR-133a inhibits RNA and protein levels of the Raldh2 gene, and antimiR-133a induces it ( Figure 3D). Since bioinformatics analyses do not show any evidence of Raldh2 as a putative target of miR-133a, our results suggest that Raldh2 would be indirectly modulated by miR-133a, by trigger mRNA degradation of those two pivotal genes, RhoA and Cdc42, required for Raldh2 expression and therefore for the adequate synthesis of RA. To analyse the functional role of these complementary anterior-posterior patterns, we performed miR-133a gain-and loss-of-functions experiments by injecting premiR-133a and antimiR-133a, respectively, into the posterior cardiac precursors of both primitive endocardial tubes ( Figure 3). The control and experimental embryos were analysed with anti-RhoA and anti-Cdc42 antibodies. Furthermore, a group of cardiac asa were dissected (Supplementary Figure S2) and subsequently subjected to RNA extraction and qPCR analysis. Our results show that miR-133a represses RNA and protein levels of RhoA and Cdc42 genes in the atrium and sinus venosus ( Figure 3B-C). Supporting these results, our loss-of-functions experiments by means of antimiR-133a demonstrate the opposite effects. In line with our gain-and loss-of-function experiments, we found that  , Tbx5 (E) and AMHC1 (F) are dramatically reduced, and an atrophic sino-atrial region in the heart tube after premiR-133a treatment is indicated by the red arrows, whereas they are markedly increased and expanded after miR-133a inhibition (blue arrows). RT-qPCR of RNA from dissected cardiac asa (left side) in embryos microinjected either with CFDA, premiR-133a or antimiR-133a. A high level of miR-133a leads to decreased RhoA, Cdc42, Raldh2, Tbx5 and AMHC1 transcripts, whereas miR-133a inhibition leads to increased transcripts. The standard deviations are from three independent experiments. Student's t-test: * p < 0.05, ** p < 0.01, *** p < 0.005 with respect to control (CFDA) embryos.

Figure 3.
Effect of miR-133a gain-and loss-of-function on posterior cardiac segment. Whole-mount IMH for RhoA, Cdc42 and Raldh2. and ISH for Tbx5 and AMHC1. Embryos microinjected with CFDA (control), premiR-133a or antimiR-133a, at the level of the posterior cardiac precursors of both primitive endocardial tubes, and visualisation of CFDA (A). Note that, at the atrium and inflow tract levels, RhoA (B), Cdc42 (C), Raldh2 (D), Tbx5 (E) and AMHC1 (F) are dramatically reduced, and an atrophic sino-atrial region in the heart tube after premiR-133a treatment is indicated by the red arrows, whereas they are markedly increased and expanded after miR-133a inhibition (blue arrows). RT-qPCR of RNA from dissected cardiac asa (left side) in embryos microinjected either with CFDA, premiR-133a or antimiR-133a. A high level of miR-133a leads to decreased RhoA, Cdc42, Raldh2, Tbx5 and AMHC1 transcripts, whereas miR-133a inhibition leads to increased transcripts. The standard deviations are from three independent experiments. Student's t-test: * p < 0.05, ** p < 0.01, *** p < 0.005 with respect to control (CFDA) embryos.

miR-133a Suppresses Tbx5 and AMHC1 Expression
In this study, we also analyse the effect of miR-133a on Tbx5 and AMHC1 expressions during cardiac atrial differentiation ( Figure 3E,F). Our gain-of-function experiments show significant downregulation of both atrial markers, followed by a reduction in size of the corresponding area located in the posterior segment of the cardiac loop compared to the control embryos. In line with these data, loss-of-function experiments presented increased expression of these markers in most of the cardiac loop region. Additionally, our qPCR analyses supported these findings. Altogether, these data reveal a negative modulation of endogenous Tbx5 and AMHC1 expressions by miR-133a during cardiac tube formation.

miR-133a Modulates Cellular Proliferation during Posterior Differentiation of the Cardiac Tube
In order to determine the specific role of miR-133a in cellular proliferation during posterior cardiac segment differentiation, in this study, we analyse the expression of two key drivers of cellular proliferation: p21 and cyclin A ( Figure 4). A group of cardiac asa were dissected from experimental embryos (Supplementary Figure S2) and subsequently subjected to RNA extraction and qPCR analysis. Our gain-of-function experiments show that miR-133a upregulates p21, while represses cyclin A. In accordance with these data, our loss-of-function experiments show downregulation of p21 and upregulation of cyclin A. Since p21 and cyclin A are modulated by RhoA and Cdc42, respectively, our results suggest indirect modulation of miR-133a of those proliferative modulators.

miR-133a Suppresses Tbx5 and AMHC1 Expression
In this study, we also analyse the effect of miR-133a on Tbx5 and AMHC1 expressions during cardiac atrial differentiation ( Figure 3E,F). Our gain-of-function experiments show significant downregulation of both atrial markers, followed by a reduction in size of the corresponding area located in the posterior segment of the cardiac loop compared to the control embryos. In line with these data, loss-of-function experiments presented increased expression of these markers in most of the cardiac loop region. Additionally, our qPCR analyses supported these findings. Altogether, these data reveal a negative modulation of endogenous Tbx5 and AMHC1 expressions by miR-133a during cardiac tube formation.

miR-133a Modulates Cellular Proliferation during Posterior Differentiation of the Cardiac Tube
In order to determine the specific role of miR-133a in cellular proliferation during posterior cardiac segment differentiation, in this study, we analyse the expression of two key drivers of cellular proliferation: p21 and cyclin A (Figure 4). A group of cardiac asa were dissected from experimental embryos (Supplementary Figure S2) and subsequently subjected to RNA extraction and qPCR analysis. Our gain-of-function experiments show that miR-133a upregulates p21, while represses cyclin A. In accordance with these data, our loss-of-function experiments show downregulation of p21 and upregulation of cyclin A. Since p21 and cyclin A are modulated by RhoA and Cdc42, respectively, our results suggest indirect modulation of miR-133a of those proliferative modulators. RT-qPCR of RNA from dissected cardiac asa in embryos microinjected either with CFDA, premiR-133a or anti-miR-133a, at the level of the posterior cardiac precursors of both primitive endocardial tubes. Note that miR-133a treatment leads to increased p21 (A) and decreased cyclin A (B) transcripts, whereas miR-133a inhibition leads to decreased p21 (A) and increased cyclin A (B) transcripts. The standard deviations are from three independent experiments. Student's t-test: *p < 0.05, ** p < 0.01, *** p < 0.005 with respect to control (CFDA) embryos.
Furthermore, by means of in vitro miR-133a gain-of-function experiments on cardiomyocytes derived from undifferentiated H9c2 cells ( Figure 5), we observe that the Ki67 proliferative marker decreased, accompanied by a lower number of cells compared with the control, suggesting that miR-133a exerts as a negative microRNA proliferator. RT-qPCR of RNA from dissected cardiac asa in embryos microinjected either with CFDA, premiR-133a or anti-miR-133a, at the level of the posterior cardiac precursors of both primitive endocardial tubes. Note that miR-133a treatment leads to increased p21 (A) and decreased cyclin A (B) transcripts, whereas miR-133a inhibition leads to decreased p21 (A) and increased cyclin A (B) transcripts. The standard deviations are from three independent experiments. Student's t-test: * p < 0.05, ** p < 0.01, *** p < 0.005 with respect to control (CFDA) embryos.
Furthermore, by means of in vitro miR-133a gain-of-function experiments on cardiomyocytes derived from undifferentiated H9c2 cells ( Figure 5), we observe that the Ki67 proliferative marker decreased, accompanied by a lower number of cells compared with the control, suggesting that miR-133a exerts as a negative microRNA proliferator.

Retinoic Acid Represses miR-133a Expression
Since our results reveal that miR-133a represses RA synthesis, we analysed the relationship between these two factors ( Figure 6). We then performed gain-of-function experiments by means of RA microinjections into both primitive endocardial tubes ( Figure  6A), as well as loss-of-function experiments by using Citral injection, described as an in-

Retinoic Acid Represses miR-133a Expression
Since our results reveal that miR-133a represses RA synthesis, we analysed the relationship between these two factors ( Figure 6). We then performed gain-of-function experiments by means of RA microinjections into both primitive endocardial tubes ( Figure 6A), as well as loss-of-function experiments by using Citral injection, described as an inhibitor of RA synthesis [13,49,50]. Our results show that RA administration inhibits miR-133a expression at the cardiac asa level. In line with this, increased areas of miR-133a expression are observed after Citral injections. Our qPCR analysis supported the above findings ( Figure 6B). Furthermore, RA administration increased RNA and protein levels of the Raldh2 gene, being decreased by Citral administration ( Figure 6C). Moreover, we observed that Tbx5 and AMHC1 expressions in the cardiac loop stage were increased after RA injection as compared to control embryos. In this line, there was an inhibition of these atrial gene markers after Citral administration, also supported by qPCR analysis (Figure 6D,E). All the above results suggest a narrow interaction between RA and miR-133a during differentiation and early development of the posterior cardiac tube segment.

Discussion
In the last few years, microRNAs have been proven to play relevant functions during embryogenesis, including cardiovascular development. In previous studies, we have shown that miR-133a, in particular, is involved in the early stages of cardiogenesis [44]. We also observed that RA signalling participates in atrial chamber differentiation during early cardiac looping [13]. Furthermore, it has been reported that RA is necessary to define the anterior-posterior boundaries of the heart-forming mesoderm (even before Figure 6. Whole-mount ISH for miR-133a, Tbx5, AMHC1 and IMH for Raldh2. Embryos microinjected with CFDA (control), Retinoic acid (RA) or Citral, at the level of the posterior cardiac precursors into both primitive endocardial tubes, and visualisation of CFDA (A). The gain-of-function of RA leads to diminished miR-133a expression (red arrows) at the cardiac asa level (B), accompanied by increased protein levels of Raldh2 (C) and expanded expression of Tbx5 (D) and AMHC1 (E) in the heart tube and in the inflow tract (blue arrows). Note atrophic sino-atrial region with increased miR-133a expression (blue arrows) at cardiac asa level (B), and also (red arrows) diminished Raldh2 protein level (C), and decreased Tbx5 (D) and AMHC1 (E) expressions by RA synthesis inhibition. RT-qPCR of RNA from dissected cardiac asa (left side) in embryos microinjected either with CFDA, RA or Citral. The standard deviations are from three independent experiments. Student's t-test: * p < 0.05, ** p < 0.01, *** p < 0.005 with respect to control (CFDA) embryos.

Discussion
In the last few years, microRNAs have been proven to play relevant functions during embryogenesis, including cardiovascular development. In previous studies, we have shown that miR-133a, in particular, is involved in the early stages of cardiogenesis [44]. We also observed that RA signalling participates in atrial chamber differentiation during early cardiac looping [13]. Furthermore, it has been reported that RA is necessary to define the anterior-posterior boundaries of the heart-forming mesoderm (even before looping) and to form the atrium and sinus venosus [16]. It is well known that Raldh2, with an expression pattern mostly restricted to the caudal region and displaying a decreasing caudal-cranial trend, is largely involved in RA synthesis, which signal is essential to promote atrial cell identity within the cardiac progenitor fields [19][20][21][22][23][24].
In this study, we carried out a detailed analysis in order to establish a relationship between miR-133a and RA during posterior cardiac tube segment differentiation. Our results show that miR-133a downregulates Raldh2 expression. Nevertheless, we did not detect that miR-133a recognises Raldh2 3 UTR. Therefore, we explored the possible interaction between miR-133a and Rho GTPases, RhoA and Cdc42, described as relevant factors linked to Raldh2 expression [32]. Our findings show that miR-133a recognises RhoA and Cdc42 3 UTRs, interactions that are translated into repression of mRNA and codified protein levels of both genes, which have been observed in humans and mice [46]. Moreover, we observed that miR-133a downregulates RhoA and Cdc42. Consequently, it is evident that Raldh2 downregulation is induced by miR-133a via these Rho GTPases. Additionally, we observed miR-133a expression pattern mostly located in the cranial region, showing a decreasing cranial-caudal trend, while RhoA and Cdc42 are located-similar to Raldh2-in the caudal region, decreasing cranially. Taking into account all of the above, our proposed model illustrates that miR-133a modulates RA signalling via Raldh2 expression, a fundamental mechanism to differentiate the posterior cardiac tube segment (Figure 7). in transgenic hearts include upregulation of p21, consistent with inhibition of RhoA, and downregulation of cyclin A, consistent with RhoA and Cdc42 inhibition by Rho GDI [32].
In line with this, our cardiac tube experiments revealed that miR-133a modulates the cell cycle regulators p21 and cyclin A. Additionally, our H9c2 culture cells experiments showed that miR-133a diminishes proliferative marker Ki67. Considering all the above data, we hypothesise that miR-133a modulates differentiation and proliferation during posterior heart tube formation (Figure 7). In this line of research, several reports have pointed out the relevant roles of miR-133a in heart development and cardiac diseases, considering it to be a basis for adult heart regeneration and repair [58][59][60][61]. Noticeably, miR-133a has been described as a protective factor in myocardial infarction by regulating cardiac remodelling [62]. In addition, low circulating levels of RA have recently been shown to reduce the expression of critical RA-dependent gene programmes in heart failure [63]. Taking all these data into account, the relationship between miR-133a and RA described in our model may provide new insights into cardiac development and heart diseases as biomarkers and/or therapy. Model proposed for the interplay between miR-133a and RA during posterior heart tube formation. Our model indicates that miR-133a downregulates RhoA and Cdc42. Consequently, Raldh2 downregulation is induced by miR-133a, via these Rho GTPases. RA synthesis is Raldh2-dependent. Thus, miR-133a modulates RA signalling via Raldh2 expression. Also, RA negatively modulates miR-133a expression during the early genetic programme of the sinoatrial region. Additionally, our model indicates that miR-133a modulates cell proliferation by acting on the cell cycle regulators p21 (via RhoA) and cyclin A (via Cdc42). We hypothesise that there is a negative feedback mechanism between miR-133a and RA signalling during early development of the posterior cardiac tube segment. A: anterior segment, P: posterior segment.

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 from the University of Extremadura bioethics board was obtained prior to the initiation of the study. Model proposed for the interplay between miR-133a and RA during posterior heart tube formation. Our model indicates that miR-133a downregulates RhoA and Cdc42. Consequently, Raldh2 downregulation is induced by miR-133a, via these Rho GTPases. RA synthesis is Raldh2-dependent. Thus, miR-133a modulates RA signalling via Raldh2 expression. Also, RA negatively modulates miR-133a expression during the early genetic programme of the sinoatrial region. Additionally, our model indicates that miR-133a modulates cell proliferation by acting on the cell cycle regulators p21 (via RhoA) and cyclin A (via Cdc42). We hypothesise that there is a negative feedback mechanism between miR-133a and RA signalling during early development of the posterior cardiac tube segment. A: anterior segment, P: posterior segment.
On the other hand, our results show that RA administration downregulates miR-133a expression. In fact, RA modulates many microRNA expressions to exert their biological functions, establishing a large number of regulatory networks currently under study [48]. Furthermore, we observed in this work that RA administration upregulates atrial genes Tbx5 and AMHC1, as we reported in previous studies [13], in agreement with other authors [25,28,51,52]. As a novelty, our experiments reveal that miR-133a downregulates both atrial gene expressions. In addition, our findings obtained by means of Citral administration indicate that the blockage of endogenous RA synthesis gives rise to Raldh2 inhibition, in parallel to that of Tbx5 and AMHC1. In accordance with previous works [26,53], in our study we established a correlation between Raldh2, an RA synthetic enzyme, and Tbx5 and AMHC1, both RA inducible genes, coincident with Tbx5 and AMHC1 upregulation cranially after RA injections in the prospective atrial cells. Consequently, in our experiments, exogenous RA leads to atrialisation of the heart, whereas inhibition of RA synthesis causes an ablated atrial chamber, in line with previous authors [25]. Significantly, our miR-133a gain-of-function experiments led to a reduction in the size of the corresponding sinoatrial domain together with lower expression of Tbx5 and AMHC1. Therefore, we hypothesise that, through this network, miR-133a modulates the posterior cardiac tube differentiation. Thus, our model (see Figure 7) supports the fact that there is a negative feedback mechanism between miR-133a and RA signalling during sino-atrium formation.
It is well established that growth and differentiation processes require a modulatory cardiomyocyte proliferation factor [41,46,54]. In the absence of miR-133a expression there is an ectopic expression of smooth muscle genes in the heart, as well as aberrant cardiomyocyte proliferation. These abnormalities have been attributed, at least in part, to elevated expression of serum response factor (SRF) and cyclin D2, which are targets of miR-133a. SRF regulates the genes responsible for cardiac muscle and smooth muscle differentiation and growth, and cyclin D2 promotes cardiomyocyte cell cycle progression [41]. Moreover, it has been proposed in humans and mice, that miR-133a plays a crucial role in determining cardiomyocyte hypertrophy by means of RhoA and Cdc42 modulation [46]. RhoA has been implicated in the regulation of hypertrophic cardiac muscle cell growth. Transgene-positive mice expressing high levels of activated RhoA showed pronounced atrial enlargement and manifested a lethal phenotype [55]. Cdc42 functions as a cell cycle progression determinant [56]. Thus, inactivation of Cdc42 causes embryonic lethality with heart development defects [57]. Increased cardiac expression of Rho GDI reduces the proliferation of cardiomyocytes. The mechanisms affecting cell proliferation in transgenic hearts include upregulation of p21, consistent with inhibition of RhoA, and downregulation of cyclin A, consistent with RhoA and Cdc42 inhibition by Rho GDI [32]. In line with this, our cardiac tube experiments revealed that miR-133a modulates the cell cycle regulators p21 and cyclin A. Additionally, our H9c2 culture cells experiments showed that miR-133a diminishes proliferative marker Ki67.
Considering all the above data, we hypothesise that miR-133a modulates differentiation and proliferation during posterior heart tube formation (Figure 7). In this line of research, several reports have pointed out the relevant roles of miR-133a in heart development and cardiac diseases, considering it to be a basis for adult heart regeneration and repair [58][59][60][61]. Noticeably, miR-133a has been described as a protective factor in myocardial infarction by regulating cardiac remodelling [62]. In addition, low circulating levels of RA have recently been shown to reduce the expression of critical RA-dependent gene programmes in heart failure [63]. Taking all these data into account, the relationship between miR-133a and RA described in our model may provide new insights into cardiac development and heart diseases as biomarkers and/or therapy.

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 . Approval from the University of Extremadura bioethics board was obtained prior to the initiation of the study.
After 16-18 h 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 (ISH) or immunochemistry (IMH) analysis, or cardiac loops were collected for RNA isolation.

RNA Isolation and qRT-PCR
Cardiac loops isolated from experimental and control (CFDA) embryos (Supplementary Figure S2) were subjected to qRT-PCR analysis following MIQE guidelines [72][73][74]. RNA was extracted and purified using the ReliaPrep RNA Cell Miniprep System Kit (Promega) according to the manufacturer's instructions. For mRNA expression measurements, 1 µg of total RNA was used for retro-transcription with a Maxima First Strand cDNA Synthesis Kit for qRT-PCR (Thermo Scientific). Real-time PCR experiments were performed with 2 µL cDNA, Go Taq qPCR Master Mix (Promega) and corresponding primer sets (Supplementary Table S1). For microRNA 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 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 Gadph as internal controls for mRNA expression analyses and 5S and 6U for microRNA expression analyses, respectively [75]. 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 software was used to perform bioinformatics analysis of binding site by miR-133a at 3 UTRs of predicted targets as described before [76].

Luciferase Assays and 3T3 Transfection
RhoA and Cdc42 3 UTR constructs (Supplementary Table S2) were PCR amplified and cloned into the pMIR_REPORT vector. 3t3 fibroblasts (ATCC) were co-transfected with 100 ng of different constructs luciferase vector and 300 ng of pcLux vector control for internal normalisation. Luciferase activity (Pierce™ Gaussia Luciferase Flash Assay Kit) was normalised to the pcLux vector control (Pierce™ Cypridina Luciferase Flash Assay Kit) and compared to non-transfected controls. Each luciferase assay was carried out in triplicate and repeated in at least three distinct biological samples to obtain representative assays.

Immunohistochemistry in H9c2 Cell Cultures
Control and experimentally premiR-133a treated cells were rinsed in PBS for 10 min at room temperature and fixed with 1% PFA for 2 h at 4 • C. After fixation, the samples were rinsed three times (10 min each) in PBS at room temperature and then permeabilysed with 1% Triton X-100 in PBS for 30 min at room temperature. To block nonspecific binding sites, PBS containing 5% goat serum and 1% bovine serum albumin (Sigma, St. Louis, MO, USA) was applied to the cell cultures overnight at 4 • C. Subsequently, they were incubated with DAPI (1:1000; Sigma) for 7 min at room temperature and rinsed three times in PBS for 5 min each. Alternatively, for DAPI staining, control and experimentally premiR-133a treated cells were immunofluorescently labelled to detect Ki67 expression. Primary antibodies against Ki67 (ab16667) were used, diluted (1:200) in PBS, and applied to each culture overnight at 4 • C. Subsequently, the samples were rinsed three times (for 1 h each) in PBS to remove excess primary antibody and incubated 2 h at room temperature with Alexa-Fluor 546 anti-rabbit (1:100; Invitrogen) as the secondary antibody. Cell cultures were stored in PBS in darkness at 4 • C until analysed using a Leica TCS SP5 II confocal scanning laser microscope.

Statistical Analyses
For statistical analyses of datasets, unpaired Student's t-tests were used, as previously reported [73,78]. Significance levels or p-values are stated in each corresponding figure legend. A p-value < 0.05 was considered statistically significant.

Author 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. and V.G.-L. contributed to the conception and design of the study, performed the experiments, and analysed the data. All authors have contributed to the experimental work and analysed and discussed the results. All authors have read and agreed to the published version of the manuscript.
Funding: This work has been financed with research grants IB18123 (to CL-S) and GR21174 (to VG-M, CTS005) from the Junta de Extremadura, with FEDER co-financing, and CTS-446 (to DF and AA) from the Junta de Andalucía Regional Council.
Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.