Vegetative insecticidal proteins （VIPs）, produced during the vegetative stage of their growth in Bacillus thuringiensis, are a group of insecticidal proteins and represent the second generation of insecticidal trans-genes that will complement the novel δendotoxins in future. Fewer structural and functional relationships of Vip proteins are known in comparison with those of δ-endotoxins. In this study, both the maximum-likelihood methods and the maximum parsimony based sliding window analysis were used to evaluate the molecular evolution of Vip proteins. As a result, strong evidence was found that Vip proteins are subject to the high rates of positive selection, and 16 sites are identified to be under positive selection using the Bayes Empirical Bayesian method. Interestingly, all these positively selected sites are located from site-705 to site-809 in the C-terminus of the Vip proteins. Most of these sites are exposed and clustered in the loop regions when mapped onto its computational predicted secondary tertiary and a part of the tertiary structure. It has been postulated that the high divergence in the C-terminal of Vip proteins may not result from the lack of functional constraints, but rather from the rapid mutation to adapt their targeted insects, driven by positive selection. The potential positive selection pressures may be an attempt to adapt for the ＂arm race＂ between Vip proteins and the targeted insects, or to enlarge their target＇s host range. Sites identified to be under positive selection may be related to the insect host range, which may shed a light on the investigation of the Vip proteins＇ structural and functional relationships.
