3 Retinoic Acid-Regulated Target Genes During Development …
75
ChIP-Seq
To fully understand RA signaling, global gene expression (microarrays or RNA-seq)
can be combined with ChIP-seq (which combines chromatin immunoprecipitation
with NGS) to identify the binding sites of DNA-associated proteins (Furey, 2012;
Fig. 3.1). This approach can map global binding sites for any protein of interest (RARs
or other TFs), as well as histone modifications (Mendoza-Parra et al. 2016). Then,
appropriate computational analysis methods can be used to predict DNA-binding
sites from ChIP-seq read count data. Peak calling methods have been developed,
the most popular method being MACS, which empirically models the shift size of
ChIP-seq tags and uses it to improve the spatial resolution of predicted binding sites.
The limits of the approach is that it requires a large number of cells (several hundred
thousands) and relies on the availability of antibodies specific for an endogenous
target and compatible with the ChIP conditions.
Using highly specific antibodies, Mendoza-Parra et al. found that ~4000 RARγ
meta sites are co-occupied with RXRα at 5 five different time points during the
RA-induced endodermal differentiation of F9 cells, (Mendoza-Parra et al. 2011).
They also found that the binding of the heterodimers is highly dynamic. Indeed,
in the absence of RA, 50% of the meta binding sites was co-occupied by RARγ
and RXRα. However, despite significant amounts of de novo recruited heterodimers,
the overall number of binding sites decreased during the RA-induced differentiation
process with less than 1000 sites at 48 h. Integration with microarrays indicated
which heterodimers sites are associated with transcription regulation and thus which
genes are putative direct targets.
A shift in RAR/RXR heterodimers binding has been also observed by other groups
with F9 cells (Chatagnon et al. 2015) and mESCs (Mahony et al. 2011; Moutier et al.
2012), but in contrast to Mendoza-Parra they did not use subtype specific antibodies
and mapped the binding sites with pan RAR and pan RXR antibodies. Nevertheless
these studies revealed that RARs occupy a large repertoire of sites with not only
consensus DR2 and DR5 spacing but also novel ones with DR0, DR8 and IR0
(Chatagnon et al. 2015; Moutier et al. 2012). The interesting point was that DR2,
DR5 and DR8 elements were enriched in RA-regulated genes, while DR0 were
occupied in the absence of RA and were not able to activate transcription.
Other NGS-Based Technologies to Detect Regions Containing
RAR Binding Sites
The combination of NGS with a variety of other DNA enrichment technologies
has provided powerful and complementary genome-wide approaches to detect
nucleosome-depleted regions containing binding sites for TFs, without relying on
the availability and specificity of antibodies (Fig. 3.4; Furey 2012).
75
ChIP-Seq
To fully understand RA signaling, global gene expression (microarrays or RNA-seq)
can be combined with ChIP-seq (which combines chromatin immunoprecipitation
with NGS) to identify the binding sites of DNA-associated proteins (Furey, 2012;
Fig. 3.1). This approach can map global binding sites for any protein of interest (RARs
or other TFs), as well as histone modifications (Mendoza-Parra et al. 2016). Then,
appropriate computational analysis methods can be used to predict DNA-binding
sites from ChIP-seq read count data. Peak calling methods have been developed,
the most popular method being MACS, which empirically models the shift size of
ChIP-seq tags and uses it to improve the spatial resolution of predicted binding sites.
The limits of the approach is that it requires a large number of cells (several hundred
thousands) and relies on the availability of antibodies specific for an endogenous
target and compatible with the ChIP conditions.
Using highly specific antibodies, Mendoza-Parra et al. found that ~4000 RARγ
meta sites are co-occupied with RXRα at 5 five different time points during the
RA-induced endodermal differentiation of F9 cells, (Mendoza-Parra et al. 2011).
They also found that the binding of the heterodimers is highly dynamic. Indeed,
in the absence of RA, 50% of the meta binding sites was co-occupied by RARγ
and RXRα. However, despite significant amounts of de novo recruited heterodimers,
the overall number of binding sites decreased during the RA-induced differentiation
process with less than 1000 sites at 48 h. Integration with microarrays indicated
which heterodimers sites are associated with transcription regulation and thus which
genes are putative direct targets.
A shift in RAR/RXR heterodimers binding has been also observed by other groups
with F9 cells (Chatagnon et al. 2015) and mESCs (Mahony et al. 2011; Moutier et al.
2012), but in contrast to Mendoza-Parra they did not use subtype specific antibodies
and mapped the binding sites with pan RAR and pan RXR antibodies. Nevertheless
these studies revealed that RARs occupy a large repertoire of sites with not only
consensus DR2 and DR5 spacing but also novel ones with DR0, DR8 and IR0
(Chatagnon et al. 2015; Moutier et al. 2012). The interesting point was that DR2,
DR5 and DR8 elements were enriched in RA-regulated genes, while DR0 were
occupied in the absence of RA and were not able to activate transcription.
Other NGS-Based Technologies to Detect Regions Containing
RAR Binding Sites
The combination of NGS with a variety of other DNA enrichment technologies
has provided powerful and complementary genome-wide approaches to detect
nucleosome-depleted regions containing binding sites for TFs, without relying on
the availability and specificity of antibodies (Fig. 3.4; Furey 2012).