HIV-1 competition experiments in humanized mice show that APOBEC3H imposes selective pressure and promotes virus adaptation

APOBEC3 (A3) family proteins are DNA cytosine deaminases recognized for contributing to HIV-1 restriction and mutation. Prior studies have demonstrated that A3D, A3F, and A3G enzymes elicit a robust anti-HIV-1 effect in cell cultures and in humanized mouse models. Human A3H is polymorphic and can be categorized into three phenotypes: stable, intermediate, and unstable. However, the anti-viral effect of endogenous A3H in vivo has yet to be examined. Here we utilize a hematopoietic stem cell-transplanted humanized mouse model and demonstrate that stable A3H robustly affects HIV-1 fitness in vivo. In contrast, the selection pressure mediated by intermediate A3H is relaxed. Intriguingly, viral genomic RNA sequencing reveled that HIV-1 frequently adapts to better counteract stable A3H during replication in humanized mice. Molecular phylogenetic analyses and mathematical modeling suggest that stable A3H may be a critical factor in human-to-human viral transmission. Taken together, this study provides evidence that stable variants of A3H impose selective pressure on HIV-1.

Small ruminant lentiviral Vif proteins commonly utilize cyclophilin A, an evolutionarily and structurally conserved protein, to degrade ovine and caprine APOBEC3 proteins

Mammals have co-evolved with retroviruses, including lentiviruses, over a long period. Evidence supporting this contention is that viral infectivity factor (Vif) encoded by lentiviruses antagonizes the anti-viral action of cellular apolipoprotein B mRNA editing enzyme catalytic polypeptide-like 3 (APOBEC3) of the host. To orchestrate E3 ubiquitin ligase complex for APOBEC3 degradation, Vifs utilize mammalian proteins such as core-binding factor beta (CBFB; for primate lentiviruses) or cyclophilin A (CYPA; for Maedi-Visna virus [MVV]). However, the co-evolutionary relationship between lentiviral Vif and the mammalian proteins associated with Vif-mediated APOBEC3 degradation is poorly understood. Moreover, it is unclear whether Vif proteins of small ruminant lentiviruses (SRLVs), including MVV and caprine arthritis encephalitis virus (CAEV), commonly utilize CYPA to degrade the APOBEC3 of their hosts. In this study, molecular phylogenetic and protein homology modeling revealed that Vif co-factors are evolutionarily and structurally conserved. It was also found that not only MVV but also CAEV Vifs degrade APOBEC3 of both sheep and goats and that CAEV Vifs interact with CYPA. These findings suggest that lentiviral Vifs chose evolutionarily and structurally stable proteins as their partners (e.g., CBFB or CYPA) for APOBEC3 degradation and, particularly, that SRLV Vifs evolved to utilize CYPA as their co-factor in degradation of ovine and caprine APOBEC3.

Functional mutations in spike glycoprotein of Zaire ebolavirus associated with an increase in infection efficiency

Ebola virus (EBOV) is extremely virulent, and its glycoprotein is necessary for viral entry. EBOV may adapt to its new host humans during outbreaks by acquiring mutations especially in glycoprotein, which allows EBOV to spread more efficiently. To identify these evolutionary selected mutations and examine their effects on viral infectivity, we used experimental-phylogenetic-structural interdisciplinary approaches. In evolutionary analysis of all available Zaire ebolavirus glycoprotein sequences, we detected two codon sites under positive selection, which are located near/within the region critical for the host-viral membrane fusion, namely alanine-to-valine and threonine-to-isoleucine mutations at 82 (A82V) and 544 (T544I), respectively. The fine-scale transmission dynamics of EBOV Makona variants that caused the 2014-2015 outbreak showed that A82V mutant was fixed in the population, whereas T544I was not. Furthermore, pseudotype assays for the Makona glycoprotein showed that the A82V mutation caused a small increase in viral infectivity compared with the T544I mutation. These findings suggest that mutation fixation in EBOV glycoprotein may be associated with their increased infectivity levels; the mutant with a moderate increase in infectivity will fix. Our findings showed that a driving force for Ebola virus evolution via glycoprotein may be a balance between costs and benefits of its virulence.

Various plus unique: Viral protein U as a plurifunctional protein for HIV-1 replication

Human immunodeficiency virus type 1 (HIV-1), the causative agent of acquired immunodeficiency syndrome, encodes four accessory genes, one of which is viral protein U (Vpu). Recently, the study of Vpu has been of great interest. For instance, various cellular proteins are degraded (e.g. CD4) and down-modulated (e.g. tetherin) by Vpu. Vpu also antagonizes the function of tetherin and inhibits NF-κB. Moreover, Vpu is a viroporin forming ion channels and may represent a promising target for anti-HIV-1 drugs. In this review, we summarize the domains/residues that are responsible for Vpu's functions, describe the current understanding of the role of Vpu in HIV-1-infected cells, and review the effect of Vpu on HIV-1 in replication and pathogenesis. Future investigations that simultaneously assess a combination of Vpu functions are required to clearly delineate the most important functions for viral replication. Impact statement Viral protein U (Vpu) is a unique protein encoded by human immunodeficiency virus type 1 (HIV-1) and related lentiviruses, playing multiple roles in viral replication and pathogenesis. In this review, we briefly summarize the most up-to-date knowledge of HIV-1 Vpu.

A naturally occurring bovine APOBEC3 confers resistance to bovine lentiviruses: implication for the co-evolution of bovids and their lentiviruses

Mammals have co-evolved with lentiviruses for a long time. As evidence, viral infectivity factor (Vif), encoded by lentiviruses, antagonizes the anti-viral action of cellular APOBEC3 of their hosts. Here, we address the co-evolutionary dynamics of bovine APOBEC3 and the following two bovine lentiviruses: bovine immunodeficiency virus (BIV) and Jembrana disease virus (JDV). We determined the sequences of three APOBEC3 genes of bovids belonging to the genera Bos and Bison and showed that bovine APOBEC3Z3 is under a strong positive selection. We found that APOBEC3Z3 of gaur, a bovid in the genus Bos, acquired resistance to JDV Vif-mediated degradation after diverging from the other bovids through conversion of the structural composition of the loop 1 domain. Interestingly, the resistance of gaur APOBEC3Z3 can be attributed to the positive selection of residue 62. This study provides the first evidence, suggesting that a co-evolutionary arms race between bovids and lentiviruses occurred in Asia.

APOBEC3 proteins can copackage and comutate HIV-1 genomes

Although APOBEC3 cytidine deaminases A3G, A3F, A3D and A3H are packaged into virions and inhibit viral replication by inducing G-to-A hypermutation, it is not known whether they are copackaged and whether they can act additively or synergistically to inhibit HIV-1 replication. Here, we showed that APOBEC3 proteins can be copackaged by visualization of fluorescently-tagged APOBEC3 proteins using single-virion fluorescence microscopy. We further determined that viruses produced in the presence of A3G + A3F and A3G + A3H, exhibited extensive comutation of viral cDNA, as determined by the frequency of G-to-A mutations in the proviral genomes in the contexts of A3G (GG-to-AG) and A3D, A3F or A3H (GA-to-AA) edited sites. The copackaging of A3G + A3F and A3G + A3H resulted in an additive increase and a modest synergistic increase (1.8-fold) in the frequency of GA-to-AA mutations, respectively. We also identified distinct editing site trinucleotide sequence contexts for each APOBEC3 protein and used them to show that hypermutation of proviral DNAs from seven patients was induced by A3G, A3F (or A3H), A3D and A3G + A3F (or A3H). These results indicate that APOBEC3 proteins can be copackaged and can comutate the same genomes, and can cooperate to inhibit HIV replication.

Perspectives of Genome-Editing Technologies for HIV Therapy

BACKGROUND

Current HIV antiretroviral therapies potently suppress virus replication and prevent patients from progressing to AIDS but are unable to completely eliminate HIV due to the existence of dormant viral reservoirs which threaten to reemerge at anytime. Recently, genome-editing technologies that can recognize specific DNA sequences, including viral DNA, are being touted as promising tools for curing HIV, owing to their specificity, ease of use, and ability to be custom designed.

CONCLUSION

Here, we introduce several novel strategies aimed at eradicating HIV proviruses with state-of-the-art genome-editing technologies and discuss perspectives of these approaches for curing HIV.

Association of lymph-node antigens with lower Gag-specific central-memory and higher Env-specific effector-memory CD8(+) T-cell frequencies in a macaque AIDS model

Virus-specific CD8⁺ T cells exert strong suppressive pressure on human/simian immunodeficiency virus (HIV/SIV) replication. These responses have been intensively examined in peripheral blood mononuclear cells (PBMCs) but not fully analyzed in lymph nodes (LNs), where interaction between CD8⁺ T cells and HIV/SIV-infected cells occurs. Here, we investigated target antigen specificity of CD8⁺ T cells in LNs in a macaque AIDS model. Analysis of virus antigen-specific CD8⁺ T-cell responses in the inguinal LNs obtained from twenty rhesus macaques in the chronic phase of SIV infection showed an inverse correlation between viral loads and frequencies of CD8⁺ T cells with CD28⁺ CD95⁺ central memory phenotype targeting the N-terminal half of SIV core antigen (Gag-N). In contrast, analysis of LNs but not PBMCs revealed a positive correlation between viral loads and frequencies of CD8⁺ T cells with CD28⁻CD95⁺ effector memory phenotype targeting the N-terminal half of SIV envelope (Env-N), soluble antigen. Indeed, LNs with detectable SIV capsid p27 antigen in the germinal center exhibited significantly lower Gag-N-specific CD28⁺ CD95⁺ CD8⁺ T-cell and higher Env-N-specific CD28⁻CD95⁺ CD8⁺ T-cell responses than those without detectable p27. These results imply that core and envelope antigen-specific CD8⁺ T cells show different patterns of interactions with HIV/SIV-infected cells.

Anti-HIV-1 potency of the CRISPR/Cas9 system insufficient to fully inhibit viral replication

The range of genome-editing tools has recently been expanded. In particular, an RNA-guided genome-editing tool, the clustered regularly interspaced short palindromic repeat (CRISPR)-associated 9 (Cas9) system, has many applications for human diseases. In this study, guide RNA (gRNA) to target gag, pol and a long terminal repeat of HIV-1 was designed and used to generate gRNA-expressing lentiviral vectors. An HIV-1-specific gRNA and Cas9 were stably dually transduced into a highly HIV-1-susceptible human T-cell line and the inhibitory ability of the anti-HIV-1 CRISPR/Cas9 lentiviral vector assessed. Although clear inhibition of the early phase of HIV-1 infection was observed, as evaluated by a VSV-G-pseudotyped HIV-1 reporter system, the anti-HIV-1 potency in multiple rounds of wild type (WT) viral replication was insufficient, either because of generation of resistant viruses or overcoming of the activity of the WT virus. Thus, there are potential difficulties that must be addressed when considering anti-HIV-1 treatment with the CRISPR/Cas9 system alone.

Quantifying the effect of Vpu on the promotion of HIV-1 replication in the humanized mouse model

BACKGROUND

Tetherin is an intrinsic anti-viral factor impairing the release of nascent HIV-1 particles from infected cells. Vpu, an HIV-1 accessory protein, antagonizes the anti-viral action of tetherin. Although previous studies using in vitro cell culture systems have revealed the molecular mechanisms of the anti-viral action of tetherin and the antagonizing action of Vpu against tetherin, it still remains unclear how Vpu affects the kinetics of HIV-1 replication in vivo.

RESULTS

To quantitatively assess the role of Vpu in viral replication in vivo, we analyzed time courses of experimental data with viral load and target cell levels in the peripheral blood of humanized mice infected with wild-type and vpu-deficient HIV-1. Our recently developed mathematical model describes the acute phase of this infection reasonably, and allowed us to estimate several parameters characterizing HIV-1 infection in mice. Using a technique of Bayesian parameter estimation, we estimate distributions of the basic reproduction number of wild-type and vpu-deficient HIV-1. This reveals that Vpu markedly increases the rate of viral replication in vivo.

CONCLUSIONS

Combining experiments with mathematical modeling, we provide an estimate for the contribution of Vpu to viral replication in humanized mice.

Species-specific differences in the ability of feline lentiviral Vif to degrade feline APOBEC3 proteins

How host-virus co-evolutionary relationships manifest is one of the most intriguing issues in virology. To address this topic, the mammal-lentivirus relationship can be considered as an interplay of cellular and viral proteins, particularly apolipoprotein B mRNA editing enzyme catalytic polypeptide-like 3 (APOBEC3) and viral infectivity factor (Vif). APOBEC3s enzymatically restrict lentivirus replication, whereas Vif antagonizes the host anti-viral action mediated by APOBEC3. In this study, the focus was on the interplay between feline APOBEC3 proteins and two feline immunodeficiency viruses in cats and pumas. To our knowledge, this study provides the first evidence of non-primate lentiviral Vif being incapable of counteracting a natural host's anti-viral activity mediated via APOBEC3 protein.

Coevolutionary dynamics between tribe Cercopithecini tetherins and their lentiviruses

Human immunodeficiency virus, a primate lentivirus (PLV), causes AIDS in humans, whereas most PLVs are less or not pathogenic in monkeys. These notions suggest that the co-evolutionary process of PLVs and their hosts associates with viral pathogenicity, and therefore, that elucidating the history of virus-host co-evolution is one of the most intriguing topics in the field of virology. To address this, recent studies have focused on the interplay between intrinsic anti-viral proteins, such as tetherin, and viral antagonists. Through an experimental-phylogenetic approach, here we investigate the co-evolutionary interplay between tribe Cercopithecini tetherin and viral antagonists, Nef and Vpu. We reveal that tribe Cercopithecini tetherins are positively selected, possibly triggered by ancient Nef-like factor(s). We reconstruct the ancestral sequence of tribe Cercopithecini tetherin and demonstrate that all Nef proteins are capable of antagonizing ancestral Cercopithecini tetherin. Further, we consider the significance of evolutionary arms race between tribe Cercopithecini and their PLVs.

Cell-to-cell infection by HIV contributes over half of virus infection

Cell-to-cell viral infection, in which viruses spread through contact of infected cell with surrounding uninfected cells, has been considered as a critical mode of virus infection. However, since it is technically difficult to experimentally discriminate the two modes of viral infection, namely cell-free infection and cell-to-cell infection, the quantitative information that underlies cell-to-cell infection has yet to be elucidated, and its impact on virus spread remains unclear. To address this fundamental question in virology, we quantitatively analyzed the dynamics of cell-to-cell and cell-free human immunodeficiency virus type 1 (HIV-1) infections through experimental-mathematical investigation. Our analyses demonstrated that the cell-to-cell infection mode accounts for approximately 60% of viral infection, and this infection mode shortens the generation time of viruses by 0.9 times and increases the viral fitness by 3.9 times. Our results suggest that even a complete block of the cell-free infection would provide only a limited impact on HIV-1 spread.

DOI:http://dx.doi.org/10.7554/eLife.08150.001

Viruses such as HIV-1 replicate by invading and hijacking cells, forcing the cells to make new copies of the virus. These copies then leave the cell and continue the infection by invading and hijacking new cells. There are two ways that viruses may move between cells, which are known as ‘cell-free’ and ‘cell-to-cell’ infection. In cell-free infection, the virus is released into the fluid that surrounds cells and moves from there into the next cell. In cell-to-cell infection the virus instead moves directly between cells across regions where the two cells make contact.

Previous research has suggested that cell-to-cell infection is important for the spread of HIV-1. However, it is not known how much the virus relies on this process, as it is technically challenging to perform experiments that prevent cell-free infection without also stopping cell-to-cell infection.

Iwami, Takeuchi et al. have overcome this problem by combining experiments on laboratory-grown cells with a mathematical model that describes how the different infection methods affect the spread of HIV-1. This revealed that the viruses spread using cell-to-cell infection about 60% of the time, which agrees with results previously found by another group of researchers. Iwami, Takeuchi et al. also found that cell-to-cell infection increases how quickly viruses can infect new cells and replicate inside them, and improves the fitness of the viruses.

The environment around cells in humans and other animals is different to that found around laboratory-grown cells, and so more research will be needed to check whether this difference affects which method of infection the virus uses. If the virus does spread in a similar way in the body, then blocking the cell-free method of infection would not greatly affect how well HIV-1 is able to infect new cells. It may instead be more effective to develop HIV treatments that prevent cell-to-cell infection by the virus.

DOI:http://dx.doi.org/10.7554/eLife.08150.002

Experimental evaluation of the zoonotic infection potency of simian retrovirus type 4 using humanized mouse model

During 2001-2002 and 2008-2011, two epidemic outbreaks of infectious hemorrhagic disease have been found in Japanese macaques (Macaca fuscata) in Kyoto University Primate Research Institute, Japan. Following investigations revealed that the causative agent was simian retrovirus type 4 (SRV-4). SRV-4 was isolated by using human cell lines, which indicates that human cells are potently susceptible to SRV-4 infection. These raise a possibility of zoonotic infection of pathogenic SRV-4 from Japanese macaques into humans. To explore the possibility of zoonotic infection of SRV-4 to humans, here we use a human hematopoietic stem cell-transplanted humanized mouse model. Eight out of the twelve SRV-4-inoculated humanized mice were infected with SRV-4. Importantly, 3 out of the 8 infected mice exhibited anemia and hemophagocytosis, and an infected mouse died. To address the possibility that SRV-4 adapts humanized mouse and acquires higher pathogenicity, the virus was isolated from an infected mice exhibited severe anemia was further inoculated into another 6 humanized mice. However, no infected mice exhibited any illness. Taken together, our findings demonstrate that the zoonotic SRV-4 infection from Japanese macaques to humans is technically possible under experimental condition. However, such zoonotic infection may not occur in the real society.

Pandemic HIV-1 Vpu overcomes intrinsic herd immunity mediated by tetherin

Among the four groups of HIV-1 (M, N, O, and P), HIV-1M alone is pandemic and has rapidly expanded across the world. However, why HIV-1M has caused a devastating pandemic while the other groups remain contained is unclear. Interestingly, only HIV-1M Vpu, a viral protein, can robustly counteract human tetherin, which tethers budding virions. Therefore, we hypothesize that this property of HIV-1M Vpu facilitates human-to-human viral transmission. Adopting a multilayered experimental-mathematical approach, we demonstrate that HIV-1M Vpu confers a 2.38-fold increase in the prevalence of HIV-1 transmission. When Vpu activity is lost, protected human populations emerge (i.e., intrinsic herd immunity develops) through the anti-viral effect of tetherin. We also reveal that all Vpus of transmitted/founder HIV-1M viruses maintain anti-tetherin activity. These findings indicate that tetherin plays the role of a host restriction factor, providing ‘intrinsic herd immunity’, whereas Vpu has evolved in HIV-1M as a tetherin antagonist.

Down-modulation of primate lentiviral receptors by Nef proteins of simian immunodeficiency virus (SIV) of chimpanzees (SIVcpz) and related SIVs: implication for the evolutionary event at the emergence of SIVcpz

It has been estimated that human immunodeficiency virus type 1 originated from the zoonotic transmission of simian immunodeficiency virus (SIV) of chimpanzees, SIVcpz, and that SIVcpz emerged by the recombination of two lineages of SIVs in Old World monkeys (SIVgsn/mon/mus in guenons and SIVrcm in red-capped mangabeys) and SIVcpz Nef is most closely related to SIVrcm Nef. These observations suggest that SIVrcm Nef had an advantage over SIVgsn/mon/mus during the evolution of SIVcpz in chimpanzees, although this advantage remains uncertain. Nef is a multifunctional protein which downregulates CD4 and coreceptor proteins from the surface of infected cells, presumably to limit superinfection. To assess the possibility that SIVrcm Nef was selected by its superior ability to downregulate viral entry receptors in chimpanzees, we compared its ability to down-modulate viral receptor proteins from humans, chimpanzees and red-capped mangabeys with Nef proteins from eight other different strains of SIVs. Surprisingly, the ability of SIVrcm Nef to downregulate CCR5, CCR2B and CXCR6 was comparable to or lower than SIVgsn/mon/mus Nef, indicating that ability to down-modulate chemokine receptors was not the selective pressure. However, SIVrcm Nef significantly downregulates chimpanzee CD4 over SIVgsn/mon/mus Nefs. Our findings suggest the possibility that the selection of SIVrcm Nef by ancestral SIVcpz is due to its superior capacity to down-modulate chimpanzees CD4 rather than coreceptor proteins.

Vif determines the requirement for CBF-β in APOBEC3 degradation

APOBEC3 (apolipoprotein B mRNA editing enzyme catalytic polypeptide-like 3) proteins are cellular DNA deaminases that restrict a broad spectrum of lentiviruses. This process is counteracted by Vif (viral infectivity factor) of lentiviruses, which binds APOBEC3s and promotes their degradation. CBF-β (core binding factor subunit β) is an essential co-factor for the function of human immunodeficiency virus type 1 Vif to degrade human APOBEC3s. However, the requirement for CBF-β in Vif-mediated degradation of other mammalian APOBEC3 proteins is less clear. Here, we determined the sequence of feline CBFB and performed phylogenetic analyses. These analyses revealed that mammalian CBFB is under purifying selection. Moreover, we demonstrated that CBF-β is dispensable for feline immunodeficiency virus Vif-mediated degradation of APOBEC3s of its host. These findings suggested that primate lentiviruses have adapted to use CBF-β, an evolutionary stable protein, to counteract APOBEC3 proteins of their hosts after diverging from other lentiviruses.

Impacts of humanized mouse models on the investigation of HIV-1 infection: illuminating the roles of viral accessory proteins in vivo

Human immunodeficiency virus type 1 (HIV-1) encodes four accessory genes: vif, vpu, vpr, and nef. Recent investigations using in vitro cell culture systems have shed light on the roles of these HIV-1 accessory proteins, Vif, Vpr, Vpu, and Nef, in counteracting, modulating, and evading various cellular factors that are responsible for anti-HIV-1 intrinsic immunity. However, since humans are the exclusive target for HIV-1 infection, conventional animal models are incapable of mimicking the dynamics of HIV-1 infection in vivo. Moreover, the effects of HIV-1 accessory proteins on viral infection in vivo remain unclear. To elucidate the roles of HIV-1 accessory proteins in the dynamics of viral infection in vivo, humanized mouse models, in which the mice are xenotransplanted with human hematopoietic stem cells, has been utilized. This review describes the current knowledge of the roles of HIV-1 accessory proteins in viral infection, replication, and pathogenicity in vivo, which are revealed by the studies using humanized mouse models.

A high excision potential of TALENs for integrated DNA of HIV-based lentiviral vector

DNA-editing technology has made it possible to rewrite genetic information in living cells. Human immunodeficiency virus (HIV) provirus, an integrated form of viral complementary DNA in host chromosomes, could be a potential target for this technology. We recently reported that HIV proviral DNA could be excised from the chromosomal DNA of HIV-based lentiviral DNA-transduced T cells after multiple introductions of a clustered regularly interspaced short palindromic repeat (CRISPR)/Cas9 endonuclease system targeting HIV long terminal repeats (LTR). Here, we generated a more efficient strategy that enables the excision of HIV proviral DNA using customized transcription activator-like effector nucleases (TALENs) targeting the same HIV LTR site. A single transfection of TALEN-encoding mRNA, prepared from in vitro transcription, resulted in more than 80% of lentiviral vector DNA being successfully removed from the T cell lines. Furthermore, we developed a lentiviral vector system that takes advantage of the efficient proviral excision with TALENs and permits the simple selection of gene-transduced and excised cells in T cell lines.

APOBEC3D and APOBEC3F potently promote HIV-1 diversification and evolution in humanized mouse model

Several APOBEC3 proteins, particularly APOBEC3D, APOBEC3F, and APOBEC3G, induce G-to-A hypermutations in HIV-1 genome, and abrogate viral replication in experimental systems, but their relative contributions to controlling viral replication and viral genetic variation in vivo have not been elucidated. On the other hand, an HIV-1-encoded protein, Vif, can degrade these APOBEC3 proteins via a ubiquitin/proteasome pathway. Although APOBEC3 proteins have been widely considered as potent restriction factors against HIV-1, it remains unclear which endogenous APOBEC3 protein(s) affect HIV-1 propagation in vivo. Here we use a humanized mouse model and HIV-1 with mutations in Vif motifs that are responsible for specific APOBEC3 interactions, DRMR/AAAA (4A) or YRHHY/AAAAA (5A), and demonstrate that endogenous APOBEC3D/F and APOBEC3G exert strong anti-HIV-1 activity in vivo. We also show that the growth kinetics of 4A HIV-1 negatively correlated with the expression level of APOBEC3F. Moreover, single genome sequencing analyses of viral RNA in plasma of infected mice reveal that 4A HIV-1 is specifically and significantly diversified. Furthermore, a mutated virus that is capable of using both CCR5 and CXCR4 as entry coreceptor is specifically detected in 4A HIV-1-infected mice. Taken together, our results demonstrate that APOBEC3D/F and APOBEC3G fundamentally work as restriction factors against HIV-1 in vivo, but at the same time, that APOBEC3D and APOBEC3F are capable of promoting viral diversification and evolution in vivo.

Mutation can produce three outcomes in viruses: detrimental, neutral, or beneficial. The first one leads to abrogation of virus replication because of error catastrophe, while the last one lets the virus escape from anti-viral immune system or adapt to the host. Human APOBEC3D, APOBEC3F, and APOBEC3G are cellular cytidine deaminases which cause G-to-A mutations in HIV-1 genome. Here we use a humanized mouse model and demonstrate that endogenous APOBEC3F and APOBEC3G induce G-to-A hypermutation in viral genomes and exert strong anti-HIV-1 activity in vivo. We also reveal that endogenous APOBEC3D and/or APOBEC3F induce viral diversification, which can lead to the emergence of a mutated virus that converts its coreceptor usage. Our results suggest that APOBEC3D and APOBEC3F are capable of promoting viral diversification and functional evolution in vivo.