Recurrent Emergence of an Antiviral Defense through Repeated Retrotransposition and Truncation of CHMP3

Abstract

SummaryMost restriction factors directly recognize specific virus features to execute antiviral functions. In contrast, we recently discovered an antiviral protein, retroCHMP3, that instead impairs the host endosomal complexes required for transport (ESCRT) pathway to inhibit budding of diverse enveloped viruses, including HIV-1. The ESCRT pathway performs essential cellular functions, so ESCRT inhibition creates the potential for cytotoxicity. Here, we chart independent evolutionary courses of retroCHMP3 emergence and reduction of cytotoxicity in New World monkeys and Old World mice using ancestral reconstructions and functional analyses. Overexpression of full-length CHMP3 results in modest antiviral activity, which is enhanced by truncating mutations that eliminate an autoinhibitory domain but also increase cytotoxicity. We show that retroCHMP3 from squirrel monkeys acquired ancient missense mutations that mitigated cytotoxicity before gaining the activating truncation. In contrast, a truncating mutation arose soon after the independent appearance of murine retroCHMP3, but the variant also exhibits regulated expression by interferon signaling, illustrating distinct paths in the emergence of an antiviral function. Our identification of additional full-length, truncated, and degraded copies of mammalian retroCHMP3 genes reveals how retrogenes can repeatedly emerge in different species to independently create new immune functions.

Processed pseudogenes: A substrate for evolutionary innovation

“…The retrogene arose through independent retrotransposition events in New World monkeys and mice, producing variants that evolved the same antiviral function. [98,99] This demonstrates the ability for multiple evolutionary trajectories to lead to the repeated emergence of retrogenes that act as a genetic reservoir for the evolution of common immune defences across multiple species. The existence of functional truncated pseudogenes also highlights the lack of empirical basis for computationally distinguishing inert pseudogenes from functional genes.…”