Differences in snake venom composition occur across all taxonomic levels and it has been argued that this variation represents an adaptation that has evolved to facilitate the capture and digestion of prey and evasion of predators. Bothrops atrox is a terrestrial pitviper that is distributed across the Amazon region, where it occupies different habitats. Using statistical analyses and functional assays that incorporate individual variation, we analyzed the individual venom variability in B. atrox snakes from four different habitats (forest, pasture, degraded area, and floodplain) in and around the Amazon River in Brazil. We observed venom differentiation between spatially distinct B. atrox individuals from the different habitats, with venom variation due to both common (high abundance) and rare (low abundance) proteins. Moreover, differences in the composition of the venoms resulted in individual variability in functionality and heterogeneity in the lethality to mammals and birds, particularly among the floodplain snakes. Taken together, the data obtained from individual venoms of B. atrox snakes, captured in different habitats from the Brazilian Amazon, support the hypothesis that the differential distribution of protein isoforms results in functional distinctiveness and the ability of snakes with different venoms to have variable toxic effects on different prey.

The evolution of snake venoms resulted in multigene toxin families that code for structurally similar isoforms eventually harboring distinct functions. PLA₂s are dominant toxins in viper venoms, and little is known about the impact of their diversity on human envenomings and neutralization by antivenoms. Here, we show the isolation of three distinct PLA₂s from B. atrox venom. FA1 is a Lys-49 homologue, and FA3 and FA4 are catalytic Asp-49 PLA₂s. FA1 and FA3 are basic myotoxic proteins, while FA4 is an acid non-myotoxic PLA₂. FA3 was the most potent toxin, inducing higher levels of edema, inflammatory nociception, indirect hemolysis, and anticoagulant activity on human, rat, and chicken plasmas. FA4 presented lower anticoagulant activity, and FA1 had only a slight effect on human and rat plasmas. PLA₂s presented differential reactivities with antivenoms, with an emphasis on FA3, which was not recognized or neutralized by the antivenoms used in this study. Our findings reveal the functional and antigenic diversity among PLA₂s from B. atrox venom, highlighting the importance of assessing venom variability for understanding human envenomations and treatment with antivenoms, particularly evident here as the antivenom fails to recognize FA3, the most active multifunctional toxin described.