Selection for translation efficiency on synonymous polymorphisms in recent human evolution

TitleSelection for translation efficiency on synonymous polymorphisms in recent human evolution
Publication TypeJournal Article
Year of Publication2011
AuthorsWaldman YY, Tuller T, Keinan A, Ruppin E
JournalGenome Biol Evol
Volume3
Pagination749–761
ISSN1759-6653
KeywordsCodon, Continental Population Groups, Evolution, Genetic, Genetics, Genome, Human, Humans, Medical, Models, Molecular, Polymorphism, Protein Biosynthesis, Proteins, Selection, Single Nucleotide
Abstract

Synonymous mutations are considered to be "silent" as they do not affect protein sequence. However, different silent codons have different translation efficiency ({TE)}, which raises the question to what extent such mutations are really neutral. We perform the first genome-wide study of natural selection operating on {TE} in recent human evolution, surveying 13,798 synonymous single nucleotide polymorphisms ({SNPs)} in 1,198 unrelated individuals from 11 populations. We find evidence for both negative and positive selection on {TE}, as measured based on differentiation in allele frequencies between populations. Notably, the likelihood of an {SNP} to be targeted by positive or negative selection is correlated with the magnitude of its effect on the {TE} of the corresponding protein. Furthermore, negative selection acting against changes in {TE} is more marked in highly expressed genes, highly interacting proteins, complex members, and regulatory genes. It is also more common in functional regions and in the initial segments of highly expressed genes. Positive selection targeting sites with a large effect on {TE} is stronger in lowly interacting proteins and in regulatory genes. Similarly, essential genes are enriched for negative {TE} selection while underrepresented for positive {TE} selection. Taken together, these results point to the significant role of {TE} as a selective force operating in humans and hence underscore the importance of considering silent {SNPs} in interpreting associations with complex human diseases. Testifying to this potential, we describe two synonymous {SNPs} that may have clinical implications in phenylketonuria and in Best's macular dystrophy due to {TE} differences between alleles.

URLhttp://keinanlab.cb.bscb.cornell.edu/sites/default/files/papers/waldman_etal_2011_selectionfortranslation_genbioevol.pdf
DOI10.1093/gbe/evr076