Heterogeneous tempo and mode of conserved noncoding sequence evolution among four mammalian orders

Abstract

Conserved noncoding sequences (CNSs) of vertebrates are considered to be closely linked with protein-coding gene regulatory functions. We examined the abundance and genomic distribution of CNSs in four mammalian orders: primates, rodents, carnivores, and cetartiodactyls. We defined the two thresholds for CNS using conservation level of coding genes; using all the three coding positions and using only first and second codon positions. The abundance of CNSs varied among lineages, with primates and rodents having highest and lowest number of CNSs, respectively, whereas carnivores and cetartiodactyls had intermediate values. These CNSs cover 1.3-5.5% of the mammalian genomes and have signatures of selective constraints that are stronger in more ancestral than the recent ones. Evolution of new CNSs as well as retention of ancestral CNSs contribute to the differences in abundance. The genomic distribution of CNSs is dynamic with higher proportions of rodent and primate CNSs located in the introns compared with carnivores and cetartiodactyls. In fact, 19% of orthologous single-copy CNSs between human and dog are located in different genomic regions. If CNSs can be considered as candidates of gene expression regulatory sequences, heterogeneity of CNSs among the four mammalian orders may have played an important role in creating the order-specific phenotypes. Fewer CNSs in rodents suggest that rodent diversity is related to lower regulatory conservation. With CNSs shown to cluster around genes involved in nervous systems and the higher number of primate CNSs, our result suggests that CNSs may be involved in the higher complexity of the primate nervous system.

Keywords: carnivores; cetartiodactyls; conserved noncoding sequences; mammals; primates; rodents.

Fig. 1.—

Assessing the suitability of percent divergence thresholds. The two thresholds are comparable to the divergence of vista enhancer elements and higher than the average of all alignable noncoding sequences.

Fig. 2.—

The phylogenetic gain and loss of CNSs. The values on a branch (in black font) are the numbers of CNSs gained on the branch, whereas the values under a branch (in red font) are the number of CNSs lost on that branch with African elephant as the reference. For each point, the values are the numbers if whole coding thresholds were used, whereas the values in parentheses are the numbers if skip3 thresholds were used.

Fig. 3.—

The conservation levels in and around CNSs. (a) The divergence levels of CNSs (***t-test P-value < 10⁻²⁰). (b) The conservation levels of flanking regions of CNSs with whole coding thresholds. Point 0 is the average percent identity of 100 bp at the center of the CNSs, whereas other points are the average of 50-bp windows moved at 20-bp steps starting from 30 pb inside the CNSs. The bars are the standard error of the mean for each window.

Fig. 4.—

The SNP coverage of CNS. The average numbers of SNPs found in 100 bp of CNS are presented for each age category. Complete SNP data as well as SNV data were used. Random coordinates of number and lengths similar to whole coding primate unique CNSs were used (***t-test P-value < 0.001; **t-test P-value < 0.005).

Fig. 5.—

GC contents of CNS and flanking regions. (a) The GC contents of CNSs. The genomic average is from Karro et al. (2008) (***t-test P-value < 0.001). (b) Using sliding windows of 200-bp size and sliding steps of 10 bp, the percent GC contents of the mammalian unique CNSs and flanking regions were computed. Position 0 is the 100 bp in the center of the CNSs and the first window starts from 50 bp into the CNSs.

Fig. 6.—

The genomic distribution of the CNSs. Using whole coding thresholds, percentage of CNSs found in each genomic region for each lineage and category are presented. Eutherian CNSs are the 35,906 single-copy CNSs conserved in all the four lineages. For random coordinates, same number and lengths of eutherian CNSs were randomly selected from noncoding sequences of each reference species.