Genomic regulatory blocks encompass multiple neighboring genes and maintain conserved synteny in vertebrates

Abstract

We report evidence for a mechanism for the maintenance of long-range conserved synteny across vertebrate genomes. We found the largest mammal-teleost conserved chromosomal segments to be spanned by highly conserved noncoding elements (HCNEs), their developmental regulatory target genes, and phylogenetically and functionally unrelated "bystander" genes. Bystander genes are not specifically under the control of the regulatory elements that drive the target genes and are expressed in patterns that are different from those of the target genes. Reporter insertions distal to zebrafish developmental regulatory genes pax6.1/2, rx3, id1, and fgf8 and miRNA genes mirn9-1 and mirn9-5 recapitulate the expression patterns of these genes even if located inside or beyond bystander genes, suggesting that the regulatory domain of a developmental regulatory gene can extend into and beyond adjacent transcriptional units. We termed these chromosomal segments genomic regulatory blocks (GRBs). After whole genome duplication in teleosts, GRBs, including HCNEs and target genes, were often maintained in both copies, while bystander genes were typically lost from one GRB, strongly suggesting that evolutionary pressure acts to keep the single-copy GRBs of higher vertebrates intact. We show that loss of bystander genes and other mutational events suffered by duplicated GRBs in teleost genomes permits target gene identification and HCNE/target gene assignment. These findings explain the absence of evolutionary breakpoints from large vertebrate chromosomal segments and will aid in the recognition of position effect mutations within human GRBs.

Genomic regulatory blocks underlie extensive microsynteny conservation in insects

“…The same trend has been observed in comparisons of mammalian genomes (Kent et al 2003;Pevzner and Tesler 2003;Murphy et al 2005). In vertebrates, conserved microsynteny can at least in part be explained by the occurrence of GRBs (Kikuta et al 2007).…”

“…We show that an equivalent organization can be observed in orthologous mosquito loci through comparisons of the genome sequences of Anopheles gambiae and Aëdes aegypti, and that the maintenance of HCNE clusters is likely to underlie preservation of microsynteny between flies and mosquitoes. The regions of HCNE arrays and microsynteny conservation also contain unrelated genes, probably in a similar way to bystander genes in vertebrate GRBs (Kikuta et al 2007). We provide genome-wide evidence that these genes generally differ from target genes in their type of core promoter, which might for the first time explain on a genomewide level why bystander genes do not specifically respond to long-range regulation in the region.…”

“…However, very often the regions spanned by HCNEs contain genes whose biological functions and expression patterns are unrelated to those of the presumptive target genes. These unrelated genes, which we refer to as "bystander genes," are independent of the regulatory input of HCNE arrays, but the pressure to maintain HCNE arrays have kept bystander and target genes together for hundreds of millions of years (Kikuta et al 2007). We termed the HCNE-spanned regions containing such genes "genomic regulatory blocks" (GRBs) and found GRBs to correspond to the longest regions of conserved gene order across vertebrate genomes.…”

BAC transgenic zebrafish reveal hypothalamic enhancer activity around obesity associated SNP rs9939609 within the human FTO gene

“…2), which is known to be regulated by an enhancer located 20 kb upstream of the SNP (Ragvin et al ., 2010). This is no proof, but from our enhancer detection screen and another transposon‐based enhancer detection approach we find that enhancer activity decreases with increasing distance from the gene (Kikuta et al ., 2007b; Kokubu et al ., 2009). Other support for this notion comes from p300 and H3K27ac ChIP sequencing data produced in embryonic and fetal mouse forebrain (Visel et al ., 2013).…”

“…It was suggested early on that the association of the LD block to FTO was misleading, and that categorically assigning risk SNPs to the closest open reading frame [termed the “nearest neighboring gene hypothesis” (Flint, 2013)], may generate false leads in regions of long‐range cis‐regulation (Ragvin et al ., 2010). Regions encompassing developmental regulatory genes with multiple cis‐regulatory inputs are termed genomic regulatory blocks (GRBs; Engstrom et al ., 2007; Kikuta et al ., 2007a, 2007b; Navratilova and Becker, 2009) and frequently span several genes, of which typically only one, identifiable by its transcriptional features, is the target of the regulatory elements in the region (Akalin et al ., 2009). Computational analysis showed that FTO is part of a GRB targeting IRX3 , encoding a developmental transcription factor, in all vertebrate genomes (Ragvin et al ., 2010).…”