Using reporter gene assays to identify cis regulatory differences between humans and chimpanzees

Abstract

Most phenotypic differences between human and chimpanzee are likely to result from differences in gene regulation, rather than changes to protein-coding regions. To date, however, only a handful of human-chimpanzee nucleotide differences leading to changes in gene regulation have been identified. To hone in on differences in regulatory elements between human and chimpanzee, we focused on 10 genes that were previously found to be differentially expressed between the two species. We then designed reporter gene assays for the putative human and chimpanzee promoters of the 10 genes. Of seven promoters that we found to be active in human liver cell lines, human and chimpanzee promoters had significantly different activity in four cases, three of which recapitulated the gene expression difference seen in the microarray experiment. For these three genes, we were therefore able to demonstrate that a change in cis influences expression differences between humans and chimpanzees. Moreover, using site-directed mutagenesis on one construct, the promoter for the DDA3 gene, we were able to identify three nucleotides that together lead to a cis regulatory difference between the species. High-throughput application of this approach can provide a map of regulatory element differences between humans and our close evolutionary relatives.

Figure 1.—

Quantitative RT–PCR results. Mean fold differences (y-axis) and standard errors for three replicates are given for either the human (open bars) or the chimpanzee (shaded bars) RNA templates. For each gene (x-axis), results were standardized on the basis of the species with the lower expression level (set to one). All gene expression differences between human and chimpanzee are significant at P < 0.05.

Figure 2.—

Differences in promoter activity between human and chimpanzee. Mean fold differences (y-axis) and standard errors for 5 replicates (or 15 in the case of DDA3) are given for either the human (open bars) or the chimpanzee (shaded bars) promoters. For each gene (x-axis), results were standardized on the basis of the species with the lower promoter activity level (set to one). P-values are given below the gene names for a one-tailed t-test of the difference in activity between the human and the chimpanzee promoters (see materials and methods).

Figure 3.—

Reporter gene assays with DDA3 constructs. (A) The promoter constructs of the human, the chimpanzee, and the different combinations are shown. (B) Mean fold differences (y-axis) and standard errors for 15 replicates are given for the human (open bar), the chimpanzee (shaded bar), or the different combo promoters. The results were standardized on the basis of the combination with the lower promoter activity level (i.e., combo 4 was set to one). All pairwise comparisons between the chimpanzee or combination 1 mean activity levels on the one hand and the human or combinations 2–6 mean activity levels on the other are significant at P < 0.01. The raw data for all 15 replicates of all the DDA3 constructs are available in supplemental Table 2 at http://www.genetics.org/supplemental/.

Figure 3.—

Reporter gene assays with DDA3 constructs. (A) The promoter constructs of the human, the chimpanzee, and the different combinations are shown. (B) Mean fold differences (y-axis) and standard errors for 15 replicates are given for the human (open bar), the chimpanzee (shaded bar), or the different combo promoters. The results were standardized on the basis of the combination with the lower promoter activity level (i.e., combo 4 was set to one). All pairwise comparisons between the chimpanzee or combination 1 mean activity levels on the one hand and the human or combinations 2–6 mean activity levels on the other are significant at P < 0.01. The raw data for all 15 replicates of all the DDA3 constructs are available in supplemental Table 2 at http://www.genetics.org/supplemental/.