Lineage-specific accelerated sequences underlying primate evolution

Abstract

Understanding the mechanisms underlying phenotypic innovation is a key goal of comparative genomic studies. Here, we investigated the evolutionary landscape of lineage-specific accelerated regions (LinARs) across 49 primate species. Genomic comparison with dense taxa sampling of primate species significantly improved LinAR detection accuracy and revealed many novel human LinARs associated with brain development or disease. Our study also yielded detailed maps of LinARs in other primate lineages that may have influenced lineage-specific phenotypic innovation and adaptation. Functional experimentation identified gibbon LinARs, which could have participated in the developmental regulation of their unique limb structures, whereas some LinARs in the Colobinae were associated with metabolite detoxification which may have been adaptive in relation to their leaf-eating diet. Overall, our study broadens knowledge of the functional roles of LinARs in primate evolution.

Fig. 1.. Landscape and characteristics of lineage-specific accelerated regions (LinARs) in primates.

The number distribution of LinARs in 18 major primate evolutionary nodes (suborders and families) and 49 extant species. More than 70% of LinARs in each lineage overlapped with long noncoding RNAs (lncRNAs) or human candidate cis-regulatory elements (cCREs). The size of the circles represents the number scale of the LinARs, whereas the gray area in the pie chart represents the overlap ratio with the lncRNA or cCREs. Species pictures are copyrighted by S. D. Nash/IUCN/SSC Primate Specialist Group and are used with their permission in this study.

Fig. 2.. Comparison between our identified human-accelerated regions and those of previous studies.

(A) The Venn diagram depicts the overlap between the human LinARs identified in our study and previously detected HARs. The dot plot shows the distribution of conservation score P values across the primates and human-accelerated signal P values for human LinARs detected specifically in our study [orange corresponds to (B), pink to (C), green to previously identified HARs, and blue to shared regions]. The red lines represent the significance threshold of the P values (conserved: P = 7.24 × 10⁻⁵, accelerated: P = 1.45 × 10⁻⁸). (B) A case of our newly identified human LinAR (HLinAR-Rank114). The top panel shows the branch length of the neighbor-joining tree across primate species with humans having a significantly longer branch than other lineages. The bottom panel shows the branch length of the neighbor-joining tree across the mammals, the branch length in the human lineage being no different from those in other mammalian lineages (20). n represents the number of species. The scale bar denotes the mean number of nucleotide substitutions per site. (C) An example of our newly identified human LinAR (HLinAR-Rank18), which was not detected by previous studies because, although this region was specifically conserved in primates, it was not conserved across the mammals. (D) Results of in situ hybridization for the expression of five lncRNAs in the adult human cerebral cortex. The tissue was dissected from the motor cortex. The expression signals of all genes were detected and marked (red). The blue dot is 4′,6-diamidino-2-phenylindole staining, showing the location of the nucleus.

Fig. 3.. Analysis of enhancer activity for two gibbon-specific accelerated regions in a lacZ reporter transgenic mouse assay.

(A and B) Gibbon-specific accelerated regions downstream of two genes (DLX5 and EMX2) associated with limb development. Black rectangles represent exons, orange rectangles represent rapidly evolving regions, and arrows represent the direction of transcription. The red lines in the phylogenetic tree represent the gibbon lineage. (C) Representative E11.5 transgenic embryos obtained for two gibbon LinAR (Rank148 and Rank181) reporters. Each construct shows three embryos resulting from independent transgene integration events. The images in the left-hand panels show close-up views of forelimb and hindlimb expressions in a representative embryo for each construct, with arrows indicating the positions where limb expression is present. Right-hand panels show zoom-in figures of the shoulder. See also fig. S10.

Fig. 4.. Dual-luciferase reporter gene assay for the assessment of LinAR promoter activity.

(A) Sequence alignment shows a Colobinae LinAR (Rank88) located within a SOD1 intron. Black rectangles denote the exon-intron structure of SOD1, the orange rectangle represents the LinAR, and the arrow denotes the direction of transcription. (B) RNA sequencing data indicated that SOD1 gene expression was decreased in many tissues from the black-and-white snub-nosed monkey (R. bieti) as compared with humans and macaques. The expression level was normalized with quantile normalization. (C) Relative luciferase activity of reporter vectors containing orthologous sequences of this LinAR in humans, Macaca, and Rhinopithecus was measured in human embryonic kidney cells. Eight replications were performed for each experimental group. ***P < 0.001; ****P < 0.0001 (t test, calculated by GraphPad Prism software).