microRNAs (miRNAs) are a class of small non-coding RNAs that are key mediators of post-transcriptional gene regulation (Lee et al., 1993; Wightman et al., 1993). Mature miRNA sequences have an average of 22 nucleotides, many of which are conserved among metazoans (Wheeler et al., 2009; Fromm et al., 2015; Bartel, 2018). Evidence in the past three decades have demonstrated that they are highly evolutionarily conserved in performing critical regulatory roles in fine-tuning of gene expression to modulate cell proliferation, cell differentiation, and the physiological functions of cells and embryos (Brennecke et al., 2003; Xu et al., 2003; Chen et al., 2004; Wienholds et al., 2005; Robinson, 2009; Wahid et al., 2010). Studies have shown that miRNAs are essential for early embryogenesis, where depletion of global miRNAs with loss-of-function of key miRNA biogenesis enzymes, Drosha and Dicer, resulted in severe developmental defects and embryonic lethality (Bernstein et al., 2003; Giraldez et al., 2005; Song et al., 2012; Saurat et al., 2013). We previously found that gastrulation failure and embryonic lethality induced by Drosha and/or Dicer morpholino antisense oligonucleotide (MASO)-injection in sea urchin embryos were rescued by co-injection with four of the most abundantly expressed miRNAs (SpmiR-1, SpmiR-31, SpmiR-71 and SpmiR-2012) (Song et al., 2012). Interestingly, highly expressed miRNAs tend to be functionally important and evolutionarily conserved (Liang and Li, 2009). While extensive progress has been made in understanding the importance of miRNAs as post-transcriptional regulators, relatively little is known about transcriptional regulation of miRNAs.

To examine how miRNAs are transcriptionally regulated in early embryonic development, we use the purple sea urchin Strongylocentrotus purpuratus as our model organism. Sea urchins are closely related to chordates with about 70% of sea urchin genes having a human counterpart (Davidson et al., 2002; 2020; Davidson, 2006; Sodergren et al., 2006). We take advantage of their well-characterized signaling pathways and gene regulatory networks (GRNs), their high fecundity, and comparatively rapid and predictable early developmental life cycle (Sodergren et al., 2006; McClay, 2011). The sea urchin has only ∼50 annotated miRNAs, which is in contrast to humans that have 519 miRNAs (Hinman et al., 2003; Wheeler et al., 2009; Song et al., 2012; Bartel, 2018). About 80% of the miRNAs found in the sea urchin genome are also present in chordates, and many are found in protostomes as well (Wheeler et al., 2009; Song et al., 2012). Using a combination of pharmacological inhibitors, genetic constructs, and morpholino antisense oligonucleotides (MASOs), we examined the regulatory impact of key signaling pathways (Wnt, Nodal, MAPK, Sonic Hedgehog (SHH), Delta/Notch, VEGF, and BMP), in addition to transcription factors (Alx1, Ets1/2, and Tbr) on SpmiR-1, SpmiR-31, SpmiR-92, SpmiR-71, SpmiR-124, SpmiR-2002, SpmiR-2007, and SpmiR-2012. Of the set of miRNAs examined, miR-1, miR-31, miR-92 and miR-124 are highly evolutionarily conserved throughout metazoans (Wheeler et al., 2009; Takane et al., 2010; Concepcion et al., 2012; Song et al., 2012; Bartel, 2018). miR-71 is conserved across insects and invertebrates (Marco et al., 2010; de Souza Gomes et al., 2013; Pérez et al., 2019); miR-2012 has only been annotated in the sea urchin, sea star, acorn worm, and Xenoturbella (Philippe et al., 2011; Song et al., 2012); and miR-2002 and miR-2007 are sea urchin specific miRNAs originating in the clade Eleutherozoa (Wheeler et al., 2009). Additionally, computational predictions were performed to identify TF binding motifs upstream of miRNA genomic loci. We examined if these predicted TFs were downstream of the signaling pathways that we tested to impact the level of miRNAs.

The sea urchin utilizes highly conserved signaling pathways to regulate development, including Wnt, Nodal, MAPK, SHH, Delta/Notch, VEGF, BMP signaling pathways. The broader Wnt signaling pathway can be separated into three main branches: canonical Wnt (cWnt), non-canonical Wnt/Planar Cell Polarity (ncWnt/PCP) pathway and the ncWnt/Ca²⁺ pathway (Komiya and Habas, 2008). The cWnt branch, which uses β-catenin as the key transducer, is critical for anterior/posterior primary body axis formation, cell differentiation and germ layer specification (Wikramanayake et al., 2004; Kumburegama and Wikramanayake, 2008). While cWnt is involved in anterior-posterior body axis, Nodal signaling pathway is involved in ventral-dorsal body axis formation (Duboc et al., 2005). In fact, cWnt is involved in activating Nodal signaling, and together they function antagonistically to set up the body plan of the embryo (Wei et al., 2012). MAP kinases phosphorylate Yan/Tel transcriptional repressor, which restricts expression of Nodal (Molina et al., 2018). Yan/Tel morphants led to expanded expression of Nodal and a radialized embryo with disrupted dorsal-ventral axis. Nodal is involved in specifying the ventral ectoderm. Activated by Nodal, BMP signaling is required for specification of the dorsal/ventral and left/right (L/R) body axes and responsible for maintaining the dorsal gene expression (Duboc et al., 2004; 2005; Furtado et al., 2008). Nodal and BMP signaling pathways work together to set up the dorsal-ventral body axis, as well as repressing neural ciliary band gene fates (Saudemont et al., 2010). The Delta/Notch signaling pathway in echinoderms is involved in segregation of the endomesoderm and specification of the secondary mesenchyme cells (SMCs) (Yoon and Gaiano, 2005; Kopan, 2012; Yaguchi et al., 2012; Yankura et al., 2013; Burke et al., 2014; Jiao et al., 2017; Siebel and Lendahl, 2017; McClay et al., 2018). SHH is involved in muscle fiber organization and patterning of the mesoderm, as well as mediating Nodal’s patterning of the L/R axis (Johnson et al., 1994; Nelson et al., 1996; Walton et al., 2009; Low and Sauvage, 2010; Warner et al., 2016). VEGF signaling is involved in directed migration of skeletogenic cells (Duloquin et al., 2007; Adomako-Ankomah and Ettensohn, 2013; Adomako-Ankomah and Ettensohn, 2014).

We have previously found SpmiR-1 and SpmiR-31 to regulate sea urchin skeletogenesis (Stepicheva and Song, 2015; Sampilo and Song, 2024). The Wnt and VEGF signaling pathways we examined here have also been shown to play important roles in skeletal development (Croce et al., 2006; Adomako-Ankomah and Ettensohn, 2013; McIntyre et al., 2013; Morgulis et al., 2021). Sea urchin skeletogenesis may be analogous to vertebrate angiogenesis and vascularization (Morgulis et al., 2019; Gildor et al., 2021). Both processes use a common set of TFs (Ets1/2, Erg, Hex, Tel, and FoxO) and signaling pathways (VEGF Nodal, BMP, Delta/Notch, and Angiopoetin). Transcription factors and signaling pathways important for vascularization are expressed and utilized in the sea urchin skeletogenic cells (primary mesenchyme cells; PMCs) at the time of migration and patterning, and in skeletal formation (Duboc et al., 2004; Oliveri et al., 2008; Morgulis et al., 2019). Prior work has shown that Ets1/2, Tbr, and Alx1 are all key regulators of sea urchin skeletogenesis (Fuchikami et al., 2002; Ettensohn et al., 2003; Oliveri et al., 2008). Of these, we found SpmiR-1 to suppress reporters containing 3′UTRs of Ets1/2 and Tbr (Sampilo and Song, 2024) and SpmiR-31 to suppress reporters containing 3′UTR of Alx1 and VegfR7 (Stepicheva and Song, 2015). Moreover, among the deuterostomes, only echinoderms and vertebrates produce extensive skeletons, while other bilaterians such as hemichordates and tunicates do not form extensive skeletons (Murdock, 2020; Ben-Tabou de-Leon, 2022). Conserved TFs involved in skeletal development include Ets1 and Alx1. Ets1 is a TF which regulates skeletogenesis in vertebrates by promoting preosteoblast proliferation; in echinoderms, Ets1 is involved in skeletogenic cell specification (Kurokawa et al., 1999; Raouf and Seth, 2000; Consales and Arnone, 2002; Ettensohn et al., 2003; Oliveri et al., 2008; Sharma and Ettensohn, 2010; Damle and Davidson, 2011). Alx1 is involved in regulating craniofacial structures in vertebrates (Zhao et al., 1996; Lyons et al., 2016; Garg et al., 2017; Pini et al., 2020). In the sea urchin, it is the main driver of skeletogenic specification (Ettensohn et al., 2003; McCauley et al., 2012; Erkenbrack and Davidson, 2015; Koga et al., 2016; Khor and Ettensohn, 2017).

miRNAs in general are a necessary component of the developmental program (Bernstein et al., 2003; Giraldez et al., 2005; Song et al., 2012; Saurat et al., 2013). For example, miR-1, known as a myomiR, regulates heart formation in vertebrates (Mansfield et al., 2004; Sokol and Ambros, 2005; Wienholds et al., 2005; Zhao et al., 2005; 2007; McCarthy, 2011). In the sea urchin, we found SpmiR-1 to regulate circumpharygeal muscle fibers and skeletogenesis (Sampilo and Song, 2024). Additionally, previous work has shown that in vertebrates, miR-31 regulates osteoblast proliferation and myogenesis (Cacchiarelli et al., 2011; Crist et al., 2012; Baglìo et al., 2013; Deng et al., 2013; Stepicheva and Song, 2016). In the sea urchin embryo, inhibition of SpmiR-31 or blockage of SpmiR-31’s suppression of Alx1 results in skeletogenic cell patterning and spicule formation defects (Stepicheva and Song, 2015). While it is less well studied in echinoderms and vertebrates, work in C. elegans has shown miR-71 to be involved in L/R axis specification and aging, with additional work showing that it is necessary for survival of primary cells in the parasitic tapeworm Echinococcus multilocularis, and oogenesis in the migratory locust Locusta migratoria (Hsieh et al., 2012; Lucanic et al., 2013; Pérez et al., 2019; Song et al., 2019). In both the sea urchin and vertebrates, miR-124 has the conserved function of regulating neuronal development and neurogenesis; in addition, in the sea urchin, SpmiR-124 is involved in specification of immune cells by regulating Delta/Notch and Nodal signaling pathways (Liu et al., 2011; Jiao et al., 2017; Konrad and Song, 2022; Konrad et al., 2023). Based on the myriad roles of miRNAs in the developmental programs of various species, it is clear that investigation into the transcriptional regulation of miRNAs can provide valuable insights into the regulation of development.

Results from this study revealed that disruption of the signaling pathways, including Wnt, Nodal, MAPK, and SHH, resulted in expression level changes of several miRNAs, while perturbation of Delta/Notch, VEGF, and BMP signaling pathways did not yield significant changes of these miRNAs. Interestingly, one of the TFs identified as a potential regulator of miR-31 transcription, Alx1, has been previously identified as a target of SpmiR-31 (Stepicheva and Song, 2015). Similarly, we have previously identified SpmiR-1 to regulate Tbr; in this study, Tbr loss-of-function leads to significant decrease of SpmiR-1 (Figure 3) (Sampilo and Song, 2024). Our results identify specific signaling pathways and TFs that likely regulate the transcription of miRNAs. We also discover cross-regulation amongst miRNAs, TFs, and signaling pathways as important regulators of embryonic development.

We identified signaling pathways and transcription factors active during embryogenesis which may potentially regulate the transcript levels of several miRNAs. Signaling pathways were found to regulate SpmiR-1, SpmiR-31, SpmiR-71, SpmiR-92, SpmiR-124, and SpmiR-2012 (Figures 1, 2). We also found Tbr, Alx1, and Ets1/2 to regulate SpmiR-1, SpmiR-31, and SpmiR-71, respectively (Figure 3). With our experimental data, we used bioinformatic predictions, evolutionary conservation, and existing ATAC-Seq information (Arshinoff et al., 2022; Telmer et al., 2024) to identify TFs which may mediate the transcript levels of these miRNAs (Figure 4). A notable implication of several results was the possibility of negative feedback loops governing the regulation of miRNAs and their targets.

Of note is that we did not find SpmiR-2002 and SpmiR-2007 to be responsive to any perturbations, indicating that these miRNAs are not regulated by these pathways and TFs and/or they may be functional later in development. Of the factors we have discovered to regulate SpmiR-1 levels, Tbr is previously characterized as a target of SpmiR-1 (Sampilo and Song, 2024). The echinoderm Tbr proteins are orthologous to vertebrate Eomesodermin (Eomes) (also known as Tbr2), Tbr1, and Tbx21 (also called T-bet) (Papaioannou and Silver, 1998; Croce et al., 2001). In the sea urchin, Tbr is zygotically expressed in the skeletogenic mesoderm of the cleavage and blastula stage embryo (Croce et al., 2001; Oliveri et al., 2002). Tbr is involved in skeletogenic mesoderm specification, as well as skeletogenesis in the larva, with Tbr loss-of-function resulting in complete loss of the larval skeleton (Croce et al., 2001; Fuchikami et al., 2002; Hinman et al., 2007; Gao and Davidson, 2008; Oliveri et al., 2008). We have previously found that SpmiR-1 overexpression results in mispatterning of the skeletogenic cells and duplicated branching of the larval skeleton (Sampilo and Song, 2024). We showed that SpmiR-1 inhibited blastulae have significantly increased Tbr mRNA levels compared to the control (Sampilo and Song, 2024). Here we found that Tbr knockdown results in a significant decrease in SpmiR-1 (Figure 3). It is interesting to note that expression of SpmiR-1 decreases in the early and mesenchyme blastulae and increases in the gastrula stage with enrichment in the mesenchymal cells (Sampilo and Song, 2024). Based on SpmiR-1 and Tbr expression data in the purple sea urchin, we speculate that while Tbr is not likely to be involved in inhibiting miR-1 expression in the blastula stage, it may be partly involved in activating miR-1 in the gastrula stage. Additionally, bioinformatic searches for Tbr binding sites identified one site at ∼9500 bp upstream of SpmiR-1, and 8,660 bp upstream in LvmiR-1 (Table 1). This set of data indicate that SpmiR-1 and Tbr are in a regulatory feedback loop where SpmiR-1 inhibits Tbr and Tbr activates SpmiR-1, suggesting that cross-regulation of miR-1 and Tbr may be important for proper skeletogenesis (Figure 4A).

Additionally, investigation of the SpmiR-1 genomic locus suggested that it was intronic, raising the possibility of co-transcription with its host gene, Mib1 (Supplementary Figure S4). While it is possible for SpmiR-1 level to be entirely dependent on its host gene, previous research has shown that at least a third of miRNAs located within introns in C. elegans retained independent promoter regions, and as such potential mechanisms for independent regulation of miR-1 in the sea urchin should not be discounted (Isik et al., 2010). Existing literature on regulation of Mib1 transcription is scarce, but does not suggest that it is regulated by Tbr or cWnt signaling (Kenny et al., 2001; Chen et al., 2021; 2023). This is inconsistent with our finding that SpmiR-1 level is significantly decreased upon Wnt perturbation with ΔLv-Cadherin injection, suggesting that inhibition of cWnt/β-catenin promotes the transcription or stabilization of miR-1 (Figure 1). Interestingly, Mib1 has been shown to activate cWnt signaling, suggesting a possible indirect mechanism of miR-1 regulation (Berndt et al., 2011).

In addition, previous work has shown miR-1 to inhibit components of the cWnt/β-catenin pathway. For example, miR-1 has been shown to inhibit FZD7 in breast cancer cells and Wnt1 ligand in the sea urchin (Liu et al., 2015; Sampilo and Song, 2024). In Drosophila, miR-1 directly suppresses Prickle, an essential ncWnt/PCP signaling component (King et al., 2011). Interestingly, miR-1 was found to suppress vertebrate oncogenic factor, TCF7 of the cWnt/β-catenin pathway during prostate cancer (Siu et al., 2017). Thus, our prior work and other studies indicate that miR-1 regulates components of the Wnt signaling pathway, and the current work indicates that miR-1 itself is also regulated by the cWnt signaling pathway (Figures 1, 4A).

The C59 drug treatment did not result in significant changes of miRNA levels, whereas, injection of truncated cadherin leading to sequestering of β-catenin, resulted in significant changes in levels of SpmiR-1, SpmiR-71, and SpmiR-2012 (Figure 1B). The reason for this could be that injection of the truncated cadherin provides an immediate perturbation directly targeting the cWnt pathway. There may be maternal sources of Wnt ligands present that were not immediately affected by the C59 treatment in abrogating palmitoylation, secretion, and the biological activity of Wnt ligands. Treatment with Bisindolylmaleimide-I resulted in significant increase of SpmiR-92 and significant decrease of SpmiR-124 (Figure 1C). Also, treatment with ROCK inhibitor, Y-27632, resulted in significant decrease of SpmiR-2012. We do not understand why C59 treatment did not result in any significant changes compared to drugs against ncWnt/Ca²⁺ (Bisindolylmaleimide-I) and ncWnt/PCP (Y-27632).

Results indicate that loss-of-function of Alx1 leads to significant decreases in the level of miR-31 (Figure 3). In addition, previous results have shown that miR-31 is a post-transcriptional regulator of Alx1, which is a primary driver of skeletogenic specification (Stepicheva and Song, 2015; Shashikant et al., 2018). Previously we have shown that miR-31 inhibition of specific block of miR-31’s suppression of Alx1 leads to significant shortening of skeletal spicules and mispatterning of skeletogenic cells (Stepicheva and Song, 2015). SpAlx1 mRNA is expressed specifically by skeletogenic cells throughout gastrulation (Ettensohn et al., 2003), and miR-31 is ubiquitously expressed in all cells throughout development (Stepicheva and Song, 2015). Knockdown of SpAlx1 and LvAlx1 revealed that Alx1 is essential for skeletogenic cell differentiation and skeletogenesis (Ettensohn et al., 2003). Potentially, Alx1 may be involved in activating SpmiR-31 to impact skeletogenesis, but the exact mechanism is not known. In vertebrates, Alx1 has known roles in neural crest development, and craniofacial structure specifically (Kayserili et al., 2009). Thus, prior and current work in the sea urchin indicate that miR-31 and Alx1 are in a possible feedback loop where Alx1 activates the transcription of miR-31, and miR-31 suppresses Alx1 (Figure 4B). However, we were not able to bioinformatically identify potential Alx1 binding site upstream of SpmiR-31 and LvmiR-31 (Table 1). This may be due to indirect regulation of miR-31 by Alx1, or may represent a predictive failure where the inferred binding specificity based on available data from human and murine Alx1 does not match the binding specificity of SpAlx1, as a result of evolutionary changes to the protein acquired in echinoderms (Lynch and Wagner, 2008; Khor and Ettensohn, 2017; Shashikant et al., 2018). Overall, our results strongly suggest that Alx1 and miR-31 participate in a possible regulatory loop that impacts sea urchin skeletogenesis (Figures 3, 4B) (Stepicheva and Song, 2015).

Worth noting is that our results indicate that knockdown of Ets1 results in a significant increase in miR-71 levels (Figures 3, 4C). miR-71 is an invertebrate-specific miRNA with known roles in L/R axis specification, olfactory neuron function, and aging in C. elegans (Boulias and Horvitz, 2012; Hsieh et al., 2012). miR-71 is also necessary for survival of primary cells in E. multilocularis, and oogenesis in L. migratoria (Lucanic et al., 2013; Finger et al., 2019; Pérez et al., 2019; Song et al., 2019). It is additionally present in the roundworm Brugia malayi, and is involved in helminth parasitism (Liu C. et al., 2015; Rojas-Pirela et al., 2022). Ets1 is a highly evolutionarily conserved transcription factor, which in vertebrates is involved in development of melanocytes and the coronary vascular endothelium, as well as organ formation from mesodermal cells (Kola et al., 1993; Saldana-Caboverde et al., 2015; Wang et al., 2022). In echinoderms, Ets1 is involved in skeletogenesis alongside Alx1 (Kurokawa et al., 1999; 1999; Ettensohn et al., 2003; Rizzo et al., 2006; Oliveri et al., 2008; Yajima et al., 2010). We do not currently know the function of SpmiR-71. Further research into miR-71 may lead to discovery of conserved roles in echinoderm neurogenesis and skeletogenesis.

Our prior research has shown that SpmiR-124 directly suppresses Notch to regulate neural development and SMC differentiation (Konrad and Song, 2022; Konrad et al., 2023). Bioinformatic results from this study indicate that SpmiR-124 may be regulated by transcription factors downstream of Delta/Notch signaling (Table 1), suggesting a possible cross-regulatory relationship during these processes (Figure 4E).

Computational predictions were one of the lines of evidence we used to identify possible regulatory relationships governing level of miRNAs, and analysis of those predictions requires context in order to evaluate results thoroughly and draw reasonable conclusions. Even with a strong p-value for a given TF binding prediction, we have to take other factors into account. For example, inherent biological variability exists in both the binding sites recognized by a given TF, and conversely, the variety of TFs that can bind to a given short DNA sequence (Todeschini et al., 2014; Kribelbauer et al., 2019). CIS-BP, the data base we used for TF binding information, creates inferred sequence-binding motifs based on available evidence, sometimes in organisms that have large evolutionary distance from the sea urchin. In the case of Alx1 and Tbr, they may have acquired changes to their coding sequences over evolutionary time resulting in different binding specificities in echinoderms compared to vertebrates (Lynch and Wagner, 2008; Cheatle Jarvela et al., 2014; Cary et al., 2017; Shashikant et al., 2018). While TFs can retain DNA-binding specificity over evolutionary distance and through significant changes in sequence, orthologous transcription factors can obtain new activities and lose others in different species (Hanks et al., 1998; Siegal and Baker, 2005; Lynch and Wagner, 2008). An additional caveat to consider is that the statistical significance of individual motif occurrences alone is not directly comparable across all transcription factors. The length and complexity of the binding motif affects the “baseline” p-value for a given occurrence, and these qualifiers vary widely among different transcription factors. For example, the consensus binding motifs for FoxC1 and other Forkhead TFs tend to be shorter (∼10nt) and A-rich, and the binding motif for Klf15 is highly C-rich, resulting in numerous hits for these factors scattered across lower-complexity stretches of sequence in the scanned region (Weirauch et al., 2014). Furthermore, sequence-based predictions do not necessarily correlate with other lines of evidence, such as in SpmiR-1, where 10 Kruppel-like factor 15 binding sites are predicted, but only four overlap with ATAC-seq (Table 1). Given that sequence specificity alone may not prove conclusive, we used various additional lines of evidence to strengthen the confidence derived from a predicted TF binding sites. This includes agreement with our own experimental data, where we demonstrate that levels of some miRNAs were modulated by predicted TFs downstream of a particular targeted signaling pathway. Also, the genome resource site for echinoderms, the Echinobase, provides ATAC-seq data across multiple timepoints of embryonic development (Arshinoff et al., 2022; Telmer et al., 2024). ATAC-seq data yields alignments to the genome wherever DNA is not closed off by chromatin, and is thus open to TF binding and transcription (Buenrostro et al., 2013; Buenrostro et al., 2015; Shashikant et al., 2018; Shashikant and Ettensohn, 2019). Any region of DNA which is accessible to ATAC-seq is potentially accessible for TF binding, and vice versa. Thus, we use the ATAC-Seq data to evaluate the possible functionality of a TF at a predicted binding site at any given stage in development.

Another criterion that we use to analyze our results is conservation of sequence in the proposed regulatory region across species (Brown et al., 2005; Rebeiz et al., 2015). For example, we identified a shared occurrence of caudal type homeobox 1-like binding sites upstream of SpmiR-2012 and LvmiR-2012 (Table 1). In general, we find the majority of TFs predicted in the purple sea urchin to be corroborated in the green sea urchin. In SpmiR-1, SpmiR-31, and SpmiR-124, 100% of predictions were corroborated with LvmiR-1, LvmiR-31, and LvmiR-124, respectively (Table 1, Supplementary Table S3). In general, using a cross-species evolutionary analysis approach, our results indicate that 73% of TF sequences found in S. purpuratus miRNA loci are also predicted in the corresponding L. variegatus miRNA loci (Table 1, Supplementary Table S3).

Proximity of the predicted binding site to the miRNA is an additional consideration. It has been shown that regions proximal to a given transcription start site (TSS) are more likely to contain TF binding sites (Lin et al., 2010; Whitfield et al., 2012). Considering this, the distance of a predicted binding motif from the miRNA itself can also serve as evidence for or against the validity of that prediction. For example, in the case of the Kruppel like factor 15, SNAI1, and HMG protein Tcf/Lef, each is predicted to bind within 500 bp upstream of SpmiR-1, with Tcf/Lef binding within 1000 bp in LvmiR-1 (Table 1).

In summary, significant changes in levels of multiple miRNAs were observed upon disruption of signaling pathways and transcription factors. We have identified several instances where miRNA level is dependent on developmental signaling pathways. The fundamental goal of the computational analysis performed was to identify the specific TFs which mediate the regulation of miRNA expression demonstrated in the experimental data presented. Our results provide multiple lines of evidence to propose reasonable TFs downstream of specific signaling pathways that may regulate miRNA levels (Figure 4). However, to definitively assess direct regulation of these TFs of specific miRNA, experimental testing will be required.

Overall, this study provides a deeper insight and understanding of how miRNAs are transcriptionally regulated by signaling pathways and transcription factors during embryogenesis. Since post-transcriptional regulation mediated by miRNAs is a key regulator of development, alongside signaling pathways and transcription factors, a greater understanding of how they regulate and cross-regulate contributes to our overall understanding of development.