Differences in enhancer activity in mouse and zebrafish reporter assays are often associated with changes in gene expression

Abstract

Background: Phenotypic evolution in animals is thought to be driven in large part by differences in gene expression patterns, which can result from sequence changes in cis-regulatory elements (cis-changes) or from changes in the expression pattern or function of transcription factors (trans-changes). While isolated examples of trans-changes have been identified, the scale of their overall contribution to regulatory and phenotypic evolution remains unclear.

Results: Here, we attempt to examine the prevalence of trans-effects and their potential impact on gene expression patterns in vertebrate evolution by comparing the function of identical human tissue-specific enhancer sequences in two highly divergent vertebrate model systems, mouse and zebrafish. Among 47 human conserved non-coding elements (CNEs) tested in transgenic mouse embryos and in stable zebrafish lines, at least one species-specific expression domain was observed in the majority (83%) of cases, and 36% presented dramatically different expression patterns between the two species. Although some of these discrepancies may be due to the use of different transgenesis systems in mouse and zebrafish, in some instances we found an association between differences in enhancer activity and changes in the endogenous gene expression patterns between mouse and zebrafish, suggesting a potential role for trans-changes in the evolution of gene expression.

Conclusions: In total, our results: (i) serve as a cautionary tale for studies investigating the role of human enhancers in different model organisms, and (ii) suggest that changes in the trans environment may play a significant role in the evolution of gene expression in vertebrates.

Figure 1

Experimental workflow. A subset of 47 conserved non-coding elements (CNEs, [26]) were randomly selected (A), and tested for enhancer activity using transgenesis in zebrafish and mice (B). Transgenic expression was decomposed into major homologous anatomical terms, and systematically compared between mouse and zebrafish embryos to identify cases of differences in trans environments (C). Finally, 26 of these CNEs could be associated to putative target genes, for which endogenous gene expression data were gathered to detect changes in gene expression between zebrafish and mouse that were consistent with trans-changes between the two species (D).

Figure 2

Comparison of enhancer activity of different CNEs in mice and zebrafish. A, B) Expression driven by the Hs608 enhancer shows mouse-specific expression (A) in the dorsal root ganglia and spinal cord (arrow), and zebrafish-specific expression (B) in the forebrain (asterisk). C, D) The Hs278 enhancer drives expression in hindbrain (arrow) and spinal cord (asterisk) in mouse (C) but only in spinal cord in zebrafish embryos (D). E, F) The Hs123 enhancer drives similar expression in the forebrain of mouse (E) and zebrafish (F, arrow).

Figure 3

Examples of a change in endogenous gene expression associated to different enhancer activities in mouse and zebrafish.A) Expression driven by the Hs382 enhancer is detected in DRG (arrow) in mouse. B) This expression is coincident with the Hs382 target gene, Znf536 (arrow). C) In zebrafish, the Hs382 enhancer does not drive expression in DRGs (arrow). This contrasts with a positive control for DRG expression (inset, arrow; Tg(−3.1neurog1:GFP)sb2). D) This absence coincides with lack of expression of Znf536 in DRG (arrow).