3 Retinoic Acid-Regulated Target Genes During Development …
77
direct RA target genes, there are the Homeobox genes, which are well known to
encode TFs regulating the expression of other downstream genes.
Taking advantage of in silico integrative approaches, Mendoza-Parra et al.
(Mendoza-Parra et al. 2016, 2011) attempted to reconstruct the cascade of TFs that
propagate the signal. Using the Dynamic Regulatory Events Miner (DREM) (Ernst
et al. 2007), they integrated TF-target gene annotations, including identified direct
putative RAR/RXRα targets, with the RA–induced gene response and FAIRE during
F9 and P19 induced differentiation. DREM analysis revealed the direct RA-target
genes and predicted six different gene co-expression paths for the endodermal differentiation of F9 cells and 10 paths for the neuronal cell fate of P19 cells. They
also reconstructed the RA-induced TFs-target gene networks involved in endodermal (F9 cells) and neuronal (P19 cells) differentiation by integrating the GRN (gene
regulatory networks) that constitute Cell Net (Morris et al. 2014) into the DREM
analysis. The reconstructed GRN reconstituted a scenario in which cascades of TFdriven common and specific regulatory programs are responsible for acquisition of
endodermal and neuronal fates.
Today, due to the increasing size of the omics datasets, a major challenge is to
develop novel algorithms reconstructing gene-regulatory networks from temporal
transcriptome data during cell fate transitions (Cholley et al. 2018) to describe the
dynamic regulation of RA-regulated genes.
Relevance
ECCs and ESCs have been extensively used as models to analyze early RA-regulated
gene programs because they differentiate into endodermal and neuronal phenotypes
upon RA addition and they provide large amounts of material that can be analyzed
by genome-wide and high-throughput NGS technologies. However, these in vitro
models do not provide insight into the developmental programs of all major organ
systems. Thus, the future challenge is to explore RA-regulated gene programs in specialized embryonic tissues from different species at different stages of development
using the same strategies that have been successful in ECCs and ESCs.
Different genome sequences are now available and microarrays have been developed to characterize early developmental gene expression patterns in several model
organisms, including mouse limb bud development (Taher et al. 2011), mouse somitogenesis (Buttitta et al. 2003), chick limb bud polarization (Pickering et al. 2017),
Drosophila life cycle (Arbeitman et al. 2002), Caenorhabditis elegans development
(Jiang et al. 2001), Xenopus gastrula embryos early development (Altmann et al.
2001; Sinner et al. 2006), endoderm and mesoderm specification in sea urchin (Davidson et al. 2002), and zebrafish development (Hermsen et al. 2013; Linney et al. 2004;
Lo et al. 2003; Ton et al. 2002). These studies have identified several regulated genes,
including TFs and signaling molecules.
Unfortunately, these approaches have not indicated whether the observed
gene expression patterns belong to RA-regulated paths. Moreover some of these
77
direct RA target genes, there are the Homeobox genes, which are well known to
encode TFs regulating the expression of other downstream genes.
Taking advantage of in silico integrative approaches, Mendoza-Parra et al.
(Mendoza-Parra et al. 2016, 2011) attempted to reconstruct the cascade of TFs that
propagate the signal. Using the Dynamic Regulatory Events Miner (DREM) (Ernst
et al. 2007), they integrated TF-target gene annotations, including identified direct
putative RAR/RXRα targets, with the RA–induced gene response and FAIRE during
F9 and P19 induced differentiation. DREM analysis revealed the direct RA-target
genes and predicted six different gene co-expression paths for the endodermal differentiation of F9 cells and 10 paths for the neuronal cell fate of P19 cells. They
also reconstructed the RA-induced TFs-target gene networks involved in endodermal (F9 cells) and neuronal (P19 cells) differentiation by integrating the GRN (gene
regulatory networks) that constitute Cell Net (Morris et al. 2014) into the DREM
analysis. The reconstructed GRN reconstituted a scenario in which cascades of TFdriven common and specific regulatory programs are responsible for acquisition of
endodermal and neuronal fates.
Today, due to the increasing size of the omics datasets, a major challenge is to
develop novel algorithms reconstructing gene-regulatory networks from temporal
transcriptome data during cell fate transitions (Cholley et al. 2018) to describe the
dynamic regulation of RA-regulated genes.
Relevance
ECCs and ESCs have been extensively used as models to analyze early RA-regulated
gene programs because they differentiate into endodermal and neuronal phenotypes
upon RA addition and they provide large amounts of material that can be analyzed
by genome-wide and high-throughput NGS technologies. However, these in vitro
models do not provide insight into the developmental programs of all major organ
systems. Thus, the future challenge is to explore RA-regulated gene programs in specialized embryonic tissues from different species at different stages of development
using the same strategies that have been successful in ECCs and ESCs.
Different genome sequences are now available and microarrays have been developed to characterize early developmental gene expression patterns in several model
organisms, including mouse limb bud development (Taher et al. 2011), mouse somitogenesis (Buttitta et al. 2003), chick limb bud polarization (Pickering et al. 2017),
Drosophila life cycle (Arbeitman et al. 2002), Caenorhabditis elegans development
(Jiang et al. 2001), Xenopus gastrula embryos early development (Altmann et al.
2001; Sinner et al. 2006), endoderm and mesoderm specification in sea urchin (Davidson et al. 2002), and zebrafish development (Hermsen et al. 2013; Linney et al. 2004;
Lo et al. 2003; Ton et al. 2002). These studies have identified several regulated genes,
including TFs and signaling molecules.
Unfortunately, these approaches have not indicated whether the observed
gene expression patterns belong to RA-regulated paths. Moreover some of these