Evolutionary Landscapes of Host-Virus Arms Races.

Vertebrate immune systems suppress viral infection using both innate restriction factors and adaptive immunity. Viruses mutate to escape these defenses, driving hosts to counterevolve to regain fitness. This cycle recurs repeatedly, resulting in an evolutionary arms race whose outcome depends on the pace and likelihood of adaptation by host and viral genes. Although viruses evolve faster than their vertebrate hosts, their proteins are subject to numerous functional constraints that impact the probability of adaptation. These constraints are globally defined by evolutionary landscapes, which describe the fitness and adaptive potential of all possible mutations. We review deep mutational scanning experiments mapping the evolutionary landscapes of both host and viral proteins engaged in arms races. For restriction factors and some broadly neutralizing antibodies, landscapes favor the host, which may help to level the evolutionary playing field against rapidly evolving viruses. We discuss the biophysical underpinnings of these landscapes and their therapeutic implications.

The structural basis for HIV-1 Vif antagonism of human APOBEC3G.

The APOBEC3 (A3) proteins are host antiviral cellular proteins that hypermutate the viral genome of diverse viral families. In retroviruses, this process requires A3 packaging into viral particles1-4. The lentiviruses encode a protein, Vif, that antagonizes A3 family members by targeting them for degradation. Diversification of A3 allows host escape from Vif whereas adaptations in Vif enable cross-species transmission of primate lentiviruses. How this 'molecular arms race' plays out at the structural level is unknown. Here, we report the cryogenic electron microscopy structure of human APOBEC3G (A3G) bound to HIV-1 Vif, and the hijacked cellular proteins that promote ubiquitin-mediated proteolysis. A small surface explains the molecular arms race, including a cross-species transmission event that led to the birth of HIV-1. Unexpectedly, we find that RNA is a molecular glue for the Vif-A3G interaction, enabling Vif to repress A3G by ubiquitin-dependent and -independent mechanisms. Our results suggest a model in which Vif antagonizes A3G by intercepting it in its most dangerous form for the virus-when bound to RNA and on the pathway to packaging-to prevent viral restriction. By engaging essential surfaces required for restriction, Vif exploits a vulnerability in A3G, suggesting a general mechanism by which RNA binding helps to position key residues necessary for viral antagonism of a host antiviral gene.

Capsid-dependent lentiviral restrictions.

Host antiviral proteins inhibit primate lentiviruses and other retroviruses by targeting many features of the viral life cycle. The lentiviral capsid protein and the assembled viral core are known to be inhibited through multiple, directly acting antiviral proteins. Several phenotypes, including those known as Lv1 through Lv5, have been described as cell type-specific blocks to infection against some but not all primate lentiviruses. Here we review important features of known capsid-targeting blocks to infection together with several blocks to infection for which the genes responsible for the inhibition still remain to be identified. We outline the features of these blocks as well as how current methodologies are now well suited to find these antiviral genes and solve these long-standing mysteries in the HIV and retrovirology fields.

Virology-the path forward.

In the United States (US), biosafety and biosecurity oversight of research on viruses is being reappraised. Safety in virology research is paramount and oversight frameworks should be reviewed periodically. Changes should be made with care, however, to avoid impeding science that is essential for rapidly reducing and responding to pandemic threats as well as addressing more common challenges caused by infectious diseases. Decades of research uniquely positioned the US to be able to respond to the COVID-19 crisis with astounding speed, delivering life-saving vaccines within a year of identifying the virus. We should embolden and empower this strength, which is a vital part of protecting the health, economy, and security of US citizens. Herein, we offer our perspectives on priorities for revised rules governing virology research in the US.

A modular CRISPR screen identifies individual and combination pathways contributing to HIV-1 latency.

Transcriptional silencing of latent HIV-1 proviruses entails complex and overlapping mechanisms that pose a major barrier to in vivo elimination of HIV-1. We developed a new latency CRISPR screening strategy, called Latency HIV-CRISPR which uses the packaging of guideRNA-encoding lentiviral vector genomes into the supernatant of budding virions as a direct readout of factors involved in the maintenance of HIV-1 latency. We developed a custom guideRNA library targeting epigenetic regulatory genes and paired the screen with and without a latency reversal agent-AZD5582, an activator of the non-canonical NFκB pathway-to examine a combination of mechanisms controlling HIV-1 latency. A component of the Nucleosome Acetyltransferase of H4 histone acetylation (NuA4 HAT) complex, ING3, acts in concert with AZD5582 to activate proviruses in J-Lat cell lines and in a primary CD4+ T cell model of HIV-1 latency. We found that the knockout of ING3 reduces acetylation of the H4 histone tail and BRD4 occupancy on the HIV-1 LTR. However, the combination of ING3 knockout accompanied with the activation of the non-canonical NFκB pathway via AZD5582 resulted in a dramatic increase in initiation and elongation of RNA Polymerase II on the HIV-1 provirus in a manner that is nearly unique among all cellular promoters.

Virology under the Microscope-a Call for Rational Discourse.

Viruses have brought humanity many challenges: respiratory infection, cancer, neurological impairment and immunosuppression to name a few. Virology research over the last 60+ years has responded to reduce this disease burden with vaccines and antivirals. Despite this long history, the COVID-19 pandemic has brought unprecedented attention to the field of virology. Some of this attention is focused on concern about the safe conduct of research with human pathogens. A small but vocal group of individuals has seized upon these concerns - conflating legitimate questions about safely conducting virus-related research with uncertainties over the origins of SARS-CoV-2. The result has fueled public confusion and, in many instances, ill-informed condemnation of virology. With this article, we seek to promote a return to rational discourse. We explain the use of gain-of-function approaches in science, discuss the possible origins of SARS-CoV-2 and outline current regulatory structures that provide oversight for virological research in the United States. By offering our expertise, we - a broad group of working virologists - seek to aid policy makers in navigating these controversial issues. Balanced, evidence-based discourse is essential to addressing public concern while maintaining and expanding much-needed research in virology.

Primate TRIM34 is a broadly-acting, TRIM5-dependent lentiviral restriction factor.

Human immunodeficiency virus (HIV) and other lentiviruses adapt to new hosts by evolving to evade host-specific innate immune proteins that differ in sequence and often viral recognition between host species. Understanding how these host antiviral proteins, called restriction factors, constrain lentivirus replication and transmission is key to understanding the emergence of pandemic viruses like HIV-1. Human TRIM34, a paralogue of the well-characterized lentiviral restriction factor TRIM5α, was previously identified by our lab via CRISPR-Cas9 screening as a restriction factor of certain HIV and SIV capsids. Here, we show that diverse primate TRIM34 orthologues from non-human primates can restrict a range of Simian Immunodeficiency Virus (SIV) capsids including SIVAGM-SAB, SIVAGM-TAN and SIVMAC capsids, which infect sabaeus monkeys, tantalus monkeys, and rhesus macaques, respectively. All primate TRIM34 orthologues tested, regardless of species of origin, were able to restrict this same subset of viral capsids. However, in all cases, this restriction also required the presence of TRIM5α. We demonstrate that TRIM5α is necessary, but not sufficient, for restriction of these capsids, and that human TRIM5α functionally interacts with TRIM34 from different species. Finally, we find that both the TRIM5α SPRY v1 loop and the TRIM34 SPRY domain are essential for TRIM34-mediated restriction. These data support a model in which TRIM34 is a broadly-conserved primate lentiviral restriction factor that acts in tandem with TRIM5α, such that together, these proteins can restrict capsids that neither can restrict alone.

HIV-1 Vif Gained Breadth in APOBEC3G Specificity after Cross-Species Transmission of Its Precursors.

APOBEC3G (A3G) is a host-encoded cytidine deaminase that potently restricts retroviruses such as HIV-1 and depends on its ability to package into virions. As a consequence of this, HIV-1 protein Vif has evolved to antagonize human A3G by targeting it for ubiquitination and subsequent degradation. There is an ancient arms race between Vif and A3G highlighted by amino acids 128 and 130 in A3G that have evolved under positive selection due to Vif-mediated selective pressure in Old World primates. Nonetheless, not all possible amino acid combinations at these sites have been sampled by nature, and the evolutionary potential of species to resist Vif antagonism is not clear. To explore the evolutionary space of positively selected sites in the Vif-binding region of A3G, we designed a combinatorial mutagenesis screen to introduce all 20 amino acids at sites 128 and 130. Our screen uncovered mutants of A3G with several interesting phenotypes, including loss of antiviral activity and resistance of Vif antagonism. However, HIV-1 Vif exhibited remarkable flexibility in antagonizing A3G 128 and 130 mutants, which significantly reduces viable Vif resistance strategies for hominid primates. Importantly, we find that broadened Vif specificity was conferred through loop 5 adaptations that were required for cross-species adaptation from Old World monkey A3G to hominid A3G. Our evidence suggests that Vif adaptation to novel A3G interfaces during cross-species transmission may train Vif toward broadened specificity that can further facilitate cross-species transmissions and raise the barrier to host resistance. IMPORTANCE APOBEC3G (A3G) is an antiviral protein that potently restricts retroviruses like HIV. In turn, the HIV-1 protein Vif has evolved to antagonize A3G through degradation. Two rapidly evolving sites in A3G confer resistance to unadapted Vif and act as a barrier to cross-species transmission of retroviruses. We recently identified a single amino acid mutation in a simian immunodeficiency virus (SIV) Vif that contributed to the cross-species origins of SIV infecting chimpanzee and, ultimately, the HIV-1 pandemic. This mutation broadened specificity of this Vif to both antagonize the A3G of its host while simultaneously overcoming the A3G barrier in the great apes. In this work, we explore the evolutionary space of human A3G at these rapidly evolving sites to understand if the broadened Vif specificity gained during cross-species transmission confers an advantage to HIV-1 Vif in its host-virus arms race with A3G.

Highly-potent, synthetic APOBEC3s restrict HIV-1 through deamination-independent mechanisms.

The APOBEC3 (A3) genes encode cytidine deaminase proteins with potent antiviral and anti-retroelement activity. This locus is characterized by duplication, recombination, and deletion events that gave rise to the seven A3s found in primates. These include three single deaminase domain A3s (A3A, A3C, and A3H) and four double deaminase domain A3s (A3B, A3D, A3F, and A3G). The most potent of the A3 proteins against HIV-1 is A3G. However, it is not clear if double deaminase domain A3s have a generalized functional advantage to restrict HIV-1. In order to test whether superior restriction factors could be created by genetically linking single A3 domains into synthetic double domains, we linked A3C and A3H single domains in novel combinations. We found that A3C/A3H double domains acquired enhanced antiviral activity that is at least as potent, if not better than, A3G. Although these synthetic double domain A3s package into budding virions more efficiently than their respective single domains, this does not fully explain their gain of antiviral potency. The antiviral activity is conferred both by cytidine-deaminase dependent and independent mechanisms, with the latter correlating to an increase in RNA binding affinity. T cell lines expressing this A3C-A3H super restriction factor are able to control replicating HIV-1ΔVif infection to similar levels as A3G. Together, these data show that novel combinations of A3 domains are capable of gaining potent antiviral activity to levels similar to the most potent genome-encoded A3s, via a primarily non-catalytic mechanism.

TRIM34 restricts HIV-1 and SIV capsids in a TRIM5α-dependent manner.

The HIV-1 capsid protein makes up the core of the virion and plays a critical role in early steps of HIV replication. Due to its exposure in the cytoplasm after entry, HIV capsid is a target for host cell factors that act directly to block infection such as TRIM5α and MxB. Several host proteins also play a role in facilitating infection, including in the protection of HIV-1 capsid from recognition by host cell restriction factors. Through an unbiased screening approach, called HIV-CRISPR, we show that the CPSF6-binding deficient, N74D HIV-1 capsid mutant is sensitive to restriction mediated by human TRIM34, a close paralog of the well-characterized HIV restriction factor TRIM5α. This restriction occurs at the step of reverse transcription, is independent of interferon stimulation, and limits HIV-1 infection in key target cells of HIV infection including CD4+ T cells and monocyte-derived dendritic cells. TRIM34 can also restrict some SIV capsids. TRIM34 restriction requires TRIM5α as knockout or knockdown of TRIM5α results in a loss of antiviral activity. Through immunofluorescence studies, we show that TRIM34 and TRIM5α colocalize to cytoplasmic bodies and are more frequently observed to be associated with infecting N74D capsids than with WT HIV-1 capsids. Our results identify TRIM34 as an HIV-1 CA-targeting restriction factor and highlight the potential role for heteromultimeric TRIM interactions in contributing to restriction of HIV-1 infection in human cells.

Combinatorial mutagenesis of rapidly evolving residues yields super-restrictor antiviral proteins.

Antagonistic interactions drive host-virus evolutionary arms races, which often manifest as recurrent amino acid changes (i.e., positive selection) at their protein-protein interaction interfaces. Here, we investigated whether combinatorial mutagenesis of positions under positive selection in a host antiviral protein could enhance its restrictive properties. We tested approximately 700 variants of human MxA, generated by combinatorial mutagenesis, for their ability to restrict Thogotovirus (THOV). We identified MxA super-restrictors with increased binding to the THOV nucleoprotein (NP) target protein and 10-fold higher anti-THOV restriction relative to wild-type human MxA, the most potent naturally occurring anti-THOV restrictor identified. Our findings reveal a means to elicit super-restrictor antiviral proteins by leveraging signatures of positive selection. Although some MxA super-restrictors of THOV were impaired in their restriction of H5N1 influenza A virus (IAV), other super-restrictor variants increased THOV restriction without impairment of IAV restriction. Thus, broadly acting antiviral proteins such as MxA mitigate breadth-versus-specificity trade-offs that could otherwise constrain their adaptive landscape.

Macaque interferon-induced transmembrane proteins limit replication of SHIV strains in an Envelope-dependent manner.

HIV-1 does not persistently infect macaques due in part to restriction by several macaque host factors. This has been partially circumvented by generating chimeric SIV/HIV-1 viruses (SHIVs) that encode SIV antagonist of known restriction factors. However, most SHIVs replicate poorly in macaques unless they are further adapted in culture and/or macaques (adapted SHIVs). Therefore, development of SHIVs encoding HIV-1 sequences derived directly from infected humans without adaptation (unadapted SHIVs) has been challenging. In contrast to the adapted SHIVs, the unadapted SHIVs have lower replication kinetics in macaque lymphocytes and are sensitive to type-1 interferon (IFN). The HIV-1 Envelope (Env) in the chimeric virus determines both the reduced replication and the IFN-sensitivity differences. There is limited information on macaque restriction factors that specifically limit replication of the more biologically relevant, unadapted SHIV variants. In order to identify the IFN-induced host factor(s) that could contribute to the inhibition of SHIVs in macaque lymphocytes, we measured IFN-induced gene expression in immortalized pig-tailed macaque (Ptm) lymphocytes using RNA-Seq. We found 147 genes that were significantly upregulated upon IFN treatment in Ptm lymphocytes and 31/147 were identified as genes that encode transmembrane helices and thus are likely present in membranes where interaction with viral Env is plausible. Within this group of upregulated genes with putative membrane-localized proteins, we identified several interferon-induced transmembrane protein (IFITM) genes, including several previously uncharacterized Ptm IFITM3-related genes. An evolutionary genomic analysis of these genes suggests the genes are IFITM3 duplications not found in humans that are both within the IFITM locus and also dispersed elsewhere in the Ptm genome. We observed that Ptm IFITMs are generally packaged at higher levels in unadapted SHIVs when compared to adapted SHIVs. CRISPR/Cas9-mediated knockout of Ptm IFITMs showed that depletion of IFITMs partially rescues the IFN sensitivity of unadapted SHIV. Moreover, we found that the depletion of IFITMs also increased replication of unadapted SHIV in the absence of IFN treatment, suggesting that Ptm IFITMs are likely important host factors that limit replication of unadapted SHIVs. In conclusion, this study shows that Ptm IFITMs selectively restrict replication of unadapted SHIVs. These findings suggest that restriction factors including IFITMs vary in their potency against different SHIV variants and may play a role in selecting for viruses that adapt to species-specific restriction factors.

Vpx overcomes a SAMHD1-independent block to HIV reverse transcription that is specific to resting CD4 T cells.

Early after entry into monocytes, macrophages, dendritic cells, and resting CD4 T cells, HIV encounters a block, limiting reverse transcription (RT) of the incoming viral RNA genome. In this context, dNTP triphosphohydrolase SAM domain and HD domain-containing protein 1 (SAMHD1) has been identified as a restriction factor, lowering the concentration of dNTP substrates to limit RT. The accessory lentiviral protein X (Vpx) proteins from the major simian immunodeficiency virus of rhesus macaque, sooty mangabey, and HIV-2 (SIVsmm/SIVmac/HIV-2) lineage packaged into virions target SAMHD1 for proteasomal degradation, increase intracellular dNTP pools, and facilitate HIV cDNA synthesis. We find that virion-packaged Vpx proteins from a second SIV lineage, SIV of red-capped mangabeys or mandrills (SIVrcm/mnd-2), increased HIV infection in resting CD4 T cells, but not in macrophages, and, unexpectedly, acted in the absence of SAMHD1 degradation, dNTP pool elevation, or changes in SAMHD1 phosphorylation. Vpx rcm/mnd-2 virion incorporation resulted in a dramatic increase of HIV-1 RT intermediates and viral cDNA in infected resting CD4 T cells. These analyses also revealed a barrier limiting HIV-1 infection of resting CD4 T cells at the level of nuclear import. Single amino acid changes in the SAMHD1-degrading Vpx mac239 allowed it to enhance early postentry steps in a Vpx rcm/mnd-2-like fashion. Moreover, Vpx enhanced HIV-1 infection of SAMHD1-deficient resting CD4 T cells of a patient with Aicardi-Goutières syndrome. These results indicate that Vpx, in addition to SAMHD1, overcomes a previously unappreciated restriction for lentiviruses at the level of RT that acts independently of dNTP concentrations and is specific to resting CD4 T cells.

Recurrent Loss of APOBEC3H Activity during Primate Evolution.

Genes in the APOBEC3 family encode cytidine deaminases that provide a barrier against viral infection and retrotransposition. Of all the APOBEC3 genes in humans, APOBEC3H (A3H) is the most polymorphic: some genes encode stable and active A3H proteins, while others are unstable and poorly antiviral. Such variation in human A3H affects interactions with the lentiviral antagonist Vif, which counteracts A3H via proteasomal degradation. In order to broaden our understanding of A3H-Vif interactions, as well as its evolution in Old World monkeys, we characterized A3H variation within four African green monkey (AGM) subspecies. We found that A3H is highly polymorphic in AGMs and has lost antiviral activity in multiple Old World monkeys. This loss of function was partially related to protein expression levels but was also influenced by amino acid mutations in the N terminus. Moreover, we demonstrate that the evolution of A3H in the primate lineages leading to AGMs was not driven by Vif. Our work suggests that the activity of A3H is evolutionarily dynamic and may have a negative effect on host fitness, resulting in its recurrent loss in primates.IMPORTANCE Adaptation of viruses to their hosts is critical for viral transmission between different species. Previous studies had identified changes in a protein from the APOBEC3 family that influenced the species specificity of simian immunodeficiency viruses (SIVs) in African green monkeys. We studied the evolution of a related protein in the same system, APOBEC3H, which has experienced a loss of function in humans. This evolutionary approach revealed that recurrent loss of APOBEC3H activity has taken place during primate evolution, suggesting that APOBEC3H places a fitness cost on hosts. The variability of APOBEC3H activity between different primates highlights the differential selective pressures on the APOBEC3 gene family.

Cytidine deaminase efficiency of the lentiviral viral restriction factor APOBEC3C correlates with dimerization.

The seven APOBEC3 (A3) enzymes in primates restrict HIV/SIV replication to differing degrees by deaminating cytosine in viral (-)DNA, which forms promutagenic uracils that inactivate the virus. A polymorphism in human APOBEC3C (A3C) that encodes an S188I mutation increases the enzymatic activity of the protein and its ability to restrict HIV-1, and correlates with increased propensity to form dimers. However, other hominid A3C proteins only have an S188, suggesting they should be less active like the common form of human A3C. Nonetheless, here we demonstrate that chimpanzee and gorilla A3C have approximately equivalent activity to human A3C I188 and that chimpanzee and gorilla A3C form dimers at the same interface as human A3C S188I, but through different amino acids. For each of these hominid A3C enzymes, dimerization enables processivity on single-stranded DNA and results in higher levels of mutagenesis during reverse transcription in vitro and in cells. For increased mutagenic activity, formation of a dimer was more important than specific amino acids and the dimer interface is unique from other A3 enzymes. We propose that dimerization is a predictor of A3C enzyme activity.

A CRISPR screen for factors regulating SAMHD1 degradation identifies IFITMs as potent inhibitors of lentiviral particle delivery.

The InterFeron Induced TransMembrane (IFITM) proteins are interferon stimulated genes that restrict many viruses, including HIV-1. SAMHD1 is another restriction factor blocking replication of HIV-1 and other viruses. Some lentiviruses evolved Vpx/Vpr proteins to degrade SAMHD1. However, this viral antagonism can be perturbed by host mechanisms: a recent study showed that in interferon (IFN) treated THP1 cells, Vpx is unable to degrade SAMHD1. In the present work, we designed an Interferon Stimulated Genes (ISGs)-targeted CRISPR knockout screen in order to identify ISGs regulating this phenotype. We found that IFITM proteins contribute to the IFNα-mediated protection of SAMHD1 by blocking VSV-G-mediated entry of the lentiviral particles delivering Vpx. Consistent with this, IFNα treatment and IFITM expression had no effect when the A-MLV envelope was used for pseudotyping. Using an assay measuring viral entry, we show that IFNα and IFITMs directly block the delivery of Vpx into cells by inhibiting VSV-G viral fusion. Strikingly, the VSV-G envelope was significantly more sensitive to this IFNα entry block and to IFITMs than HIV-1's natural envelope. This highlights important differences between VSV-G pseudotyped and wild-type HIV-1, in particular relative to the pathways they use for viral entry, suggesting that HIV-1 may have evolved to escape restriction factors blocking entry.

A Single Nucleotide Polymorphism in Human APOBEC3C Enhances Restriction of Lentiviruses.

Humans express seven human APOBEC3 proteins, which can inhibit viruses and endogenous retroelements through cytidine deaminase activity. The seven paralogs differ in the potency of their antiviral effects, as well as in their antiviral targets. One APOBEC3, APOBEC3C, is exceptional as it has been found to only weakly block viruses and endogenous retroelements compared to other APOBEC3s. However, our positive selection analyses suggest that APOBEC3C has played a role in pathogen defense during primate evolution. Here, we describe a single nucleotide polymorphism in human APOBEC3C, a change from serine to isoleucine at position 188 (I188) that confers potent antiviral activity against HIV-1. The gain-of-function APOBEC3C SNP results in increased enzymatic activity and hypermutation of target sequences when tested in vitro, and correlates with increased dimerization of the protein. The I188 is widely distributed in human African populations, and is the ancestral primate allele, but is not found in chimpanzees or gorillas. Thus, while other hominids have lost activity of this antiviral gene, it has been maintained, or re-acquired, as a more active antiviral gene in a subset of humans. Taken together, our results suggest that APOBEC3C is in fact involved in protecting hosts from lentiviruses.

Conservation and Innovation of APOBEC3A Restriction Functions during Primate Evolution.

LINE-1 (long interspersed element-1) retroelements are the only active autonomous endogenous retroelements in human genomes. Their retrotransposition activity has created close to 50% of the current human genome. Due to the apparent costs of this proliferation, host genomes have evolved multiple mechanisms to curb LINE-1 retrotransposition. Here, we investigate the evolution and function of the LINE-1 restriction factor APOBEC3A, a member of the APOBEC3 cytidine deaminase gene family. We find that APOBEC3A genes have evolved rapidly under diversifying selection in primates, suggesting changes in APOBEC3A have been recurrently selected in a host-pathogen "arms race." Nonetheless, in contrast to previous reports, we find that the LINE-1 restriction activity of APOBEC3A proteins has been strictly conserved throughout simian primate evolution in spite of its pervasive diversifying selection. Based on these results, we conclude that LINE-1s have not driven the rapid evolution of APOBEC3A in primates. In contrast to this conserved LINE-1 restriction, we find that a subset of primate APOBEC3A genes have enhanced antiviral restriction. We trace this gain of antiviral restriction in APOBEC3A to the common ancestor of a subset of Old World monkeys. Thus, APOBEC3A has not only maintained its LINE-1 restriction ability, but also evolved a gain of antiviral specificity against other pathogens. Our findings suggest that while APOBEC3A has evolved to restrict additional pathogens, only those adaptive amino acid changes that leave LINE-1 restriction unperturbed have been tolerated.

Evolutionary Analyses Suggest a Function of MxB Immunity Proteins Beyond Lentivirus Restriction.

Viruses impose diverse and dynamic challenges on host defenses. Diversifying selection of codons and gene copy number variation are two hallmarks of genetic innovation in antiviral genes engaged in host-virus genetic conflicts. The myxovirus resistance (Mx) genes encode interferon-inducible GTPases that constitute a major arm of the cell-autonomous defense against viral infection. Unlike the broad antiviral activity of MxA, primate MxB was recently shown to specifically inhibit lentiviruses including HIV-1. We carried out detailed evolutionary analyses to investigate whether genetic conflict with lentiviruses has shaped MxB evolution in primates. We found strong evidence for diversifying selection in the MxB N-terminal tail, which contains molecular determinants of MxB anti-lentivirus specificity. However, we found no overlap between previously-mapped residues that dictate lentiviral restriction and those that have evolved under diversifying selection. Instead, our findings are consistent with MxB having a long-standing and important role in the interferon response to viral infection against a broader range of pathogens than is currently appreciated. Despite its critical role in host innate immunity, we also uncovered multiple functional losses of MxB during mammalian evolution, either by pseudogenization or by gene conversion from MxA genes. Thus, although the majority of mammalian genomes encode two Mx genes, this apparent stasis masks the dramatic effects that recombination and diversifying selection have played in shaping the evolutionary history of Mx genes. Discrepancies between our study and previous publications highlight the need to account for recombination in analyses of positive selection, as well as the importance of using sequence datasets with appropriate depth of divergence. Our study also illustrates that evolutionary analyses of antiviral gene families are critical towards understanding molecular principles that govern host-virus interactions and species-specific susceptibility to viral infection.