NEW YORK (GenomeWeb) – In a new analysis, researchers were unable to find any evidence of recent, human-specific selection at FOXP2, a gene with ties to language.
The Max Planck Institute for Evolutionary Anthropology's Svante Pääbo and his colleagues reported in 2002 that FOXP2 — particularly two missense substitutions — appeared to be under recent, positive selection in humans.
As FOXP2 encodes a transcription factor that is expressed in the brain during fetal development — and mutations in it have been linked to language dysfunction — this recent selection suggested that the gene could have a key role in the evolution of modern language among humans. However, these substitutions have since been found among ancient hominins that split from the human lineage before that selective sweep was thought to have occurred.
Researchers from Stony Brook University have now revisited that initial conclusion, conducting a new analysis using a larger, more diverse dataset, and that appears to have made a difference.
"We wanted to test whether their hypothesis stood up against a larger, more diverse dataset that more explicitly controlled for human demography," senior author Brenna Henn from Stony Brook and the University of California, Davis, said in a statement.
As they reported today in Cell, the researchers uncovered no evidence of positive selection on FOXP2. They attributed the previous finding to the ancestral makeup of the initial samples.
The initial analysis from 2002 relied on a small set of 20 people on whom the researchers conducted Sanger sequencing of three FOXP2 introns. That cohort included seven individuals from Africa, four from Europe, one from South America, five from Asia, and three from Australia and Papua New Guinea.
For this new analysis, Henn and her colleagues drew upon data from the Human Genome Diversity Panel and 1000 Genomes Project. In addition, they had sequencing data from across the full FOXP2 gene region.
Applying the same Tajima's D statistic approach to the HGDP dataset that Pääbo and his colleagues had previously used, Henn and her colleagues initially did find an indication of positive selection on FOXP2. However, when they separated out African individuals and individuals whose ancestors underwent the out-of-Africa expansion, that signal disappeared. They reported a similar effect when they analyzed the 1000 Genomes Project dataset.
Henn and her team also applied other statistical approaches to search for signals of selection — including signals of balancing selection — but were unable to tease any out.
This suggested to the researchers that the signal the 2002 team uncovered was instead due to the ancestral composition of their sample set. This, Henn said, underscores the importance of collecting and using diverse human samples. "There's a severe Eurocentric bias in a lot of medical and other scientific studies, but we've found a scientific impetus for emphasizing diversity and inclusivity in data collection because it clearly yields more accurate results," she added.
Meanwhile, she and her colleagues also scanned the FOXP2 region using gene-level genomic evolutionary rate profiling (GERP) scores to uncover an intronic region containing SNPs that appeared to be evolutionarily conserved and many of which appeared unique to humans.
Through in vitro and in vivo studies, the researchers found that this intronic region was not a novel exon, but instead suggested that it could encode a long non-coding RNA or act as a lowly transcribed enhancer RNA.
"[W]hile we're not questioning the functional work of FOXP2 or its role in language production, we're finding that the story of FOXP2 is really more complex than we'd ever imagined," coauthor Sohini Ramachandran from Brown University said in a statement.