Oliver TA et al. (2010), Whole-genome positive selection and habitat-dri...

ECB-ART-41789

Genome Biol Evol 2010 Jan 01;2:800-14. doi: 10.1093/gbe/evq063.

Show Gene links Show Anatomy links

Whole-genome positive selection and habitat-driven evolution in a shallow and a deep-sea urchin.

Oliver TA , Garfield DA , Manier MK , Haygood R , Wray GA , Palumbi SR .

???displayArticle.abstract???
Comparisons of genomic sequence between divergent species can provide insight into the action of natural selection across many distinct classes of proteins. Here, we examine the extent of positive selection as a function of tissue-specific and stage-specific gene expression in two closely-related sea urchins, the shallow-water Strongylocentrotus purpuratus and the deep-sea Allocentrotus fragilis, which have diverged greatly in their adult but not larval habitats. Genes that are expressed specifically in adult somatic tissue have significantly higher dN/dS ratios than the genome-wide average, whereas those in larvae are indistinguishable from the genome-wide average. Testis-specific genes have the highest dN/dS values, whereas ovary-specific have the lowest. Branch-site models involving the outgroup S. franciscanus indicate greater selection (ω(FG)) along the A. fragilis branch than along the S. purpuratus branch. The A. fragilis branch also shows a higher proportion of genes under positive selection, including those involved in skeletal development, endocytosis, and sulfur metabolism. Both lineages are approximately equal in enrichment for positive selection of genes involved in immunity, development, and cell-cell communication. The branch-site models further suggest that adult-specific genes have experienced greater positive selection than those expressed in larvae and that ovary-specific genes are more conserved (i.e., experienced greater negative selection) than those expressed specifically in adult somatic tissues and testis. Our results chart the patterns of protein change that have occurred after habitat divergence in these two species and show that the developmental or functional context in which a gene acts can play an important role in how divergent species adapt to new environments.

???displayArticle.pubmedLink??? 20935062
???displayArticle.pmcLink??? PMC2975446
???displayArticle.link??? Genome Biol Evol
???displayArticle.grants??? [+]

Genes referenced: LOC100887844 LOC115919910 LOC583082

???attribute.lit??? ???displayArticles.show???

	FIG. 1.—. Pairwise Allocentrotus fragilis: Strongylocentrotus purpuratus dN/dS histograms–(A) Per-read analysis, (B) whole-gene analysis. Note both distributions are plotted on a Log2 x axis scale.
	FIG. 2.—. Dendrogram of tissue-specific expression. The dendrogram represents similarities in expression profile of a given tissue using cluster analysis after binarizing expression in each tissue.
	FIG. 3.—. Venn diagram of four tissue expression categories using dN/dS calculated from per-read alignments. Shading represents median dN/dS ratios for a given expression category. Percentages within each section report the percentage of all expressed genes/alignments that fell into that category. Asterisks () denote a category whose dN/dS distribution is significantly different than the whole-genome dN/dS distribution via two-tailed Mann-Whitney test. A single asterisk () denote a category that is significant at the α = 0.05 level, while two asterisks (**) denote a category whose significance is robust to Bonferroni correction for 16 tests, giving α = 0.003.
	FIG. 4.—. Comparisons of the composition of the data set by expression categories for both all dN/dS values and a set of high dN/dS genes. Asterisks denote categories showing significant changes among sets, by χ2 with alpha corrected for four tests per data set (α = 0.012).
	FIG. 5.—. (A) All tissue mean branch-specific ωFG from branch-specific likelihood model. Error bars show the 95% CIs of 1,000 bootstrap replicates of the mean. Asterisks indicate a significant difference in distributions via Mann-Whitney test. (B) Branch-specific proportion of genes showing significant positive selection in the branch-specific likelihood model. Asterisk indicates a significant comparison by χ2.
	FIG. 6.—. (A) Mean branch-specific ωFG from branch-specific likelihood model across each tissue-specific expression category. Error bars show the 95% CIs of 1,000 bootstrap replicates of the mean. Asterisks indicate a significant difference between a given tissue-specific distribution and the all tissue distribution via Mann-Whitney test, Bonferroni corrected for eight comparisons. (B) Branch-specific proportion of genes showing significant positive selection in the branch-specific likelihood model across each tissue-specific expression category. Asterisk indicates a significant comparison by χ2, Bonferroni corrected for eight comparisons.
	FIG. 7.—. (A) Positive, (B) neutral, and (C) negative selection scores along each branch from the branch-specific likelihood model, across all tissue-specific expression categories. Asterisks indicate a significant difference between a given tissue-specific distribution and the all tissue distribution via Mann-Whitney test, Bonferroni corrected for eight comparisons.
	FIG. 8.—. Correlations between Strongylocentrotus purpuratus and Allocentrotus fragilis selection scores for each of 11.115 genes in branch-specific data set. (A) Correlations between negative selection scores (Pearson's r = 0.81); (B) Correlations between neutrality scores (r = 0.76); and (C) Correlations between positive selection scores (r = 0.08).

References [+] :

ALLISON, Protection afforded by sickle-cell trait against subtertian malareal infection. 1954, Pubmed