A monkey wrench in retroviral defense: loss of APOBEC3H activity in Old World monkeys

Inter-species transmission of viruses is critically dependent on the adaptation of viruses to their hosts. The evolution of lentiviral-host interactions within Old World monkeys has provided insights into the dynamics of the evolutionary arms-race between host antiviral proteins and their lentiviral targets. In particular, the African green monkey (AGM) is ideal for assessing the evolutionary forces responsible for interactions between lentiviruses and their hosts because this genus has four geographically distinct subspecies that are infected with subspecies-specific simian immunodeficiency virus (SIV). In Old World monkeys, the APOBEC3H (A3H) gene, which encodes a cytidine deaminase, has efficient antiviral activity against primate lentiviruses, is sensitive to inactivation by the viral antagonist Vif protein, and is capable of hypermutating retroviral genomes. In humans, A3H is highly polymorphic which influences its antiviral activity. As a result of the genetic polymorphisms in human populations, A3H is expressed at varying levels in different people. Fortuitously, there are seven members of the APOBEC3 family in primates, which makes an evolutionary approach to study this adaptation more feasible. Genes involved in antiviral immunity in AGMs have been reported to be polymorphic, with some changes at the protein level critical for interaction with Vif; however, it is unknown if A3H has a similar role in these primates.

Erin Garcia, a graduate student in Dr. Michael Emerman’s laboratory in the Human Biology Division, took on the challenge of investigating the host-virus “arms-race” between A3H and Vif proteins encoded by SIVs within AGMs. In their paper recently published in the Journal of Virology, Garcia and Dr. Emerman report that A3H antiviral activity was lost repeatedly throughout evolution, suggesting that A3H places a fitness cost on hosts.

They first sequenced A3H in 50 AGM DNA samples collected from all four subspecies infected with a subspecies-specific SIV, and found several polymorphisms, suggesting functional consequences for either antiviral activity or Vif antagonism. An excess of nonsynonymous mutations compared to synonymous mutations can be used to define positive selection of antiviral restriction factors. Evolutionary conflicts between host restriction factors and viral proteins to maintain or escape interactions usually result in an accumulation of nonsynonymous mutations at binding interfaces. To determine if the nonsynonymous A3H polymorphisms affect antiviral activity, the authors cloned haplotypes representative of all four subspecies into an expression vector for functional analysis in HEK293T cells. They found that A3H variants have poor antiviral activity against at least two separate lentiviruses and cannot be attributed to species-specificity. Similar to unstable haplotypes of human A3H, lower AGM A3H protein expression levels correlated with less potent antiviral activity. The authors determined that the difference in antiviral activity was not due to viral packaging or localization of the A3H protein. Instead, decreased protein expression levels, as well as amino acid mutations in the N-terminal region, resulted in lower antiviral activity. Additionally, the loss of A3H activity is possibly due to mutations in a putative RNA binding domain. This observation is supported by the finding that RNA binding has been previously shown to play an important role in the antiviral activity of A3H.

Next, by molecular reconstruction of ancestral A3H sequences leading to the modern AGM lineage, the authors found that the most recent common ancestor of AGMs likely encoded an active A3H, similar to other common ancestors throughout evolutionary history. To improve statistical power, the authors included a wider panel of Old World monkeys in sister clades. This suggests that the recurrent loss is a more recent event in primate evolution. Selective pressure by Vif does not appear to be a primary force behind the evolution of A3H in the AGM clade, but since a loss has occurred in both humans and other Old World monkeys, there may be a fitness cost to encoding this mutator protein over long evolutionary time periods.

Garcia explained: “Our work highlights the flexibility of the APOBEC3 locus in primates, which influences viral adaptation to new hosts during lentivirus transmission between species. Given how polymorphic A3H is in African green monkeys, we were surprised to find that all the variants we tested were less active. We expected to find a wider range of activity, but that was not the case. Since A3H proteins from other primate species are active, this hinted that something had happened to antiviral activity in the past and we decided to determine what that was. Publication of the A3H structure during the study was helpful because it allowed us to map where inactivating mutations were in the protein. It was interesting to us that these mutations were primarily in regions implicated in RNA binding because RNA binding has been shown to be important for the activity of other APOBEC3s. Although we haven’t directly shown that RNA binding defects are responsible for less potent antiviral activity, it makes sense based on current knowledge.”

Garcia EI and Emerman M. 2018. Recurrent loss of APOBEC3H activity during primate evolution. Journal of Virology Jun 20. pii: JVI.00971-18.

Funding was provided by the National Institutes of Health.