Potential arms race in the coevolution of primates and angiosperms: brazzein sweet proteins and gorilla taste receptors

Abstract

We explored whether variation in the sweet taste receptor protein T1R3 in primates could contribute to differences in sweet taste repertoire among species, potentially reflecting coevolution with local plants. Specifically, we examined which primates are likely to be sweet "tasters" of brazzein, a protein found in the fruit of the African plant Pentadiplandra brazzeana that tastes intensely sweet to humans, but provides little energy. Sweet proteins like brazzein are thought to mimic the taste of sugars to entice seed dispersers. We examined the evolution of T1R3 and assessed whether primates are likely "deceived" by such biochemical mimicry. Using published and new sequence data for TAS1R3, we characterized 57 primates and other mammals at the two amino acid sites necessary to taste brazzein to determine which species are tasters. We further used dN/dS-based methods to look for statistical evidence of accelerated evolution in this protein across primate lineages. The taster genotype is shared across most catarrhines, suggesting that most African primates can be "tricked" into eating and dispersing P. brazzeana's seeds for little caloric gain. Western gorillas (Gorilla gorilla), however, exhibit derived mutations at the two brazzein-critical positions, and although fruit is a substantial portion of the western gorilla diet, they have not been observed to eat P. brazzeana. Our analyses of protein evolution found no signature of positive selection on TAS1R3 along the gorilla lineage. We propose that the gorilla-specific mutations at the TAS1R3 locus encoding T1R3 could be a counter-adaptation to the false sweet signal of brazzein.