Genomic Evidence for the Nonpathogenic State in HIV-1–Infected Northern Pig-Tailed Macaques

Abstract HIV-1 is a highly host-specific retrovirus that infects humans but not most nonhuman primates. Thus, the lack of a suitable primate model that can be directly infected with HIV-1 hinders HIV-1/AIDS research. In the previous study, we have found that the northern pig-tailed macaques (NPMs) are susceptible to HIV-1 infection but show a nonpathogenic state. In this study, to understand this macaque–HIV-1 interaction, we assembled a de novo genome and longitudinal transcriptome for this species during the course of HIV-1 infection. Using comparative genomic analysis, a positively selected gene, Toll-like receptor 8, was identified with a weak ability to induce an inflammatory response in this macaque. In addition, an interferon-stimulated gene, interferon alpha inducible protein 27, was upregulated in acute HIV-1 infection and acquired an enhanced ability to inhibit HIV-1 replication compared with its human ortholog. These findings coincide with the observation of persistently downregulated immune activation and low viral replication and can partially explain the AIDS-free state in this macaque following HIV-1 infection. This study identified a number of unexplored host genes that may hamper HIV-1 replication and pathogenicity in NPMs and provided new insights into the host defense mechanisms in cross-species infection of HIV-1. This work will facilitate the adoption of NPM as a feasible animal model for HIV-1/AIDS research.


Introduction
A suitable nonhuman primate model that can be infected with HIV-1 and develop AIDS would facilitate the evaluation of anti-HIV-1 therapies and vaccines, but such a model is currently lacking. This is partly because HIV-1 is a highly host-specific retrovirus that replicates robustly and causes AIDS only in humans but does not infect other primates in general. Although Chimpanzees can be infected with HIV-1, it is greatly limited to use for the endangered status and extreme expense in animal experiment and hardly developing to AIDS after several years of infection (Nath et al. 2000). At present, the frequently used nonhuman primate models for HIV-1/AIDS research are simian immunodeficiency virus (SIV) or chimeric SIV encoding the HIV-1 envelope or reverse transcriptase (SHIV)-infected macaques. These models display similar viral replication and pathogenicity with those in HIV-1-infected humans (Hatziioannou and Evans 2012). However, because of the large genetic difference between HIV-1 and SIV (their genomes only share about 40-50% sequence identity at the nucleotide level) (Pollom et al. 2013), the anti-HIV-1 drug and vaccine evaluations in these models cannot directly applied to clinical studies.

MBE
Pig-tailed macaques (PTMs) are the only type of Old-World monkeys that were found to be susceptible to HIV-1 in the 1990s (Agy et al. 1992), and they were then found to be more amenable to infection by the HIV-1 derivative strain stHIV-1 in the 2000s. However, neither natural infection of these two strains in PTMs led to a pathogenic state (Hatziioannou et al. 2009(Hatziioannou et al. , 2014Thippeshappa et al. 2011;Schmidt et al. 2019). We have firstly posited that a fusion gene TRIM5-cyclophilin A replaced an important HIV-1 restrict factor TRIM5α in PTMs, resulting in the loss of anti-HIV-1 activity of this gene and thus rendering PTMs susceptible to HIV-1 infection (Liao et al. 2007). Thereafter, we found that one species of PTMs, northern pig-tailed macaque (Macaca leonine, NPM), which inhabits Southwest China and Southeast Asia, is amenable to HIV-1 infection (Kuang et al. 2009). HIV-1 could replicate at a low level and formed viral reservoirs in vivo during several years of infection and did not induce any diseases in this species . Whereas an HIV-1 derivative strain stHIV-1sv, in which HIV-1 vif gene is replaced by SIV mac vif to counterpart another important HIV-1 host restrict factor APOBEC3 in NPMs, presents a high level of replication in acute infection, but its replication was still impeded in chronic infection and also did not cause AIDS in NPMs . In stark contrast, SIV infection in this monkey caused a consistently high viremia with a slow AIDS progression . Therefore, it remains unclear why HIV-1, and in particular stHIV-1sv, maintain low levels of replications and do not induce AIDS in NPM compared with SIV.
To gain insight into the underlying molecular mechanism responsible for the distinct outcomes of viral infections, and ultimately, the nonpathogenic state of HIV-1 infection in NPMs, here, we conducted a comparative study between HIV-1 and stHIV-1sv infections in NPMs and SIV infection in NPMs or rhesus macaques (RMs). We assembled the genome of NPM (M. leonina) and profiled longitudinal transcriptomes after HIV-1 and SIV infections in NPMs by deep RNA-sequencing. Using comparative genomic and functional analyses, we identified several specific genes that are associated with the AIDS-free state in HIV-1-infected NPMs. In particular, Toll-like receptor 8 (TLR8) showed a weak ability to induce an inflammatory response, and interferon alpha inducible protein 27 (IFI27) obtained an enhanced capacity of inhibition HIV-1 replication in NPMs. This study provides a comprehensive view of the genomic landscape and transcriptomic changes during HIV-1 infection in nonhuman primates, and facilitates the understanding of the complexity of HIV-1 adaptation and pathogenicity in this monkey population.

HIV-1-Infected NPM Exhibited a Nonpathogenic State With Low Viramia and Immune Activation
We inoculated four NPMs with an HIV-1 strains, HIV-1 NL4-R3A (Nolan et al. 2009), and four NPMs with the HIV-1 derivative stHIV-1sv (Hatziioannou et al. 2009), respectively; we also infected four NPMs and three RMs with the SIV strain SIV mac239 as positive controls, respectively ( fig. 1A). The main viral and immunological features during infection were measured. The HIV-1 and SIV infections in NPMs and RMs were divided into acute infection (0-6 weeks post infection [wpi]) and chronic infection (7-69 wpi) according to the setpoints of plasma viral loads. In contrast to the efficient replication of SIV mac239 in NPMs and RMs, the replication of HIV-1 NL4-R3A was less efficient in NPMs, and stHIV-1sv showed a higher level of acute viremia than that of HIV-1 NL4-R3A , possibly due to the replacement of SIV vif; but its replication was still lower than that of SIV mac239 infection in NPMs. However, the replications of both HIV-1 strains were impeded during chronic stage, which kept a relatively low level, approximate 10-to 100-fold reductions than those of SIV-infected NPMs and RMs ( 1C). These results are in line with our previous studies Zhang et al. 2018) and suggest that beside TRIM5α and APOBEC3, some other restrict factors may inhibit HIV-1 replication in NPMs. An indirect indication was stHIV-1sv adaptation to another species of PTMs, southern PTMs, wherein the amino acid changes in viral Vpu, and Gag-capsid proteins have obtained enhanced antagonism to macaque restrict factors tetherin, and MX2, respectively, but the viral swarms or clones did not replicate well and induce the macaques to AIDS states (Hatziioannou et al. 2014;Schmidt et al. 2019), implying that some undiscovered host factors still hamper HIV-1 replication and pathogenicity in that species.
We also detected T-cell activation and proliferation in peripheral blood mononuclear cells (PBMCs) by the changes of the Ki67 marker and found that HIV-1 NL4-R3A -and stHIV-1sv-infected NPMs exhibited a substantially low level of Ki67 + expression on CD4 + T and CD8 + T cells during the acute and chronic stage, whereas SIV mac239 -infected NPMs showed a transient and moderate increase of Ki67 + expression on CD4 + T and CD8 + T cells in the acute infection and a decline in the chronic stage. In contrast, SIV mac239 infection in RMs presented a robust expression of Ki67 in the whole process ( fig. 1D and E). This feature was consistent with two other immune activation markers, CD38 and HLA-DR (supplementary fig. S2, Supplementary Material online). During the 69-week surveillance period of the infection, the HIV-1− and SIV-infected NPMs did not show any obviously AIDS-related diseases. In contrast, the SIV-infected RMs presented the typical AIDS-defining illnesses, such as the loss of weight and diarrhea (data not shown). Thus, these data demonstrated that: 1) Compared with SIV infections in NPMs and RMs, HIV-1 infections in NPMs presented a lower level of viral replication, especially in chronic infection stage. 2)

MBE
Compared with SIV infection in RMs, HIV-1 and SIV infections in NPMs led to a weaker level of immune activation. Together, the HIV-1-infected NPMs in particular exhibited a nonpathogenic state with two main features: low viramia and low immune activation.  Main virological and immunological characteristics of NPMs infected with HIV-1 NL4-R3A , stHIV-1sv, or SIV mac239 . (A) Twelve healthy NPMs were equally divided into three groups for HIV-1 NL4-R3A , stHIV-1sv, and SIV mac239 infection, and three healthy RMs were used for SIV mac239 infection. The infection stage was divided into acute infection (0-6 wpi) and chronic infection (7-69 wpi). Asterisks represent time points at which periphPBMCs were collected for transcriptome analysis. Changes in plasma viral load (B), CD4 + T cells (C ), CD4 + Ki67 + T cells (D), and CD8 + Ki67 + T cells (E) following HIV-1 NL4-R3A , stHIV-1sv, and SIV mac239 infection in NPMs, and SIV mac239 infection in RMs. In (B), dashed lines mean the limit of quantification, In (B) and (C), asterisks in RM-SIV mac239 panels indicate one RM died of AIDS at 41 wpi. In (B)-(E), each shape in different infections represents each macaque. Data in the last panel of (B)-(E) are presented as mean ± SEM, and the data in each infection during 1-6 wpi or 7-69/7-28 wpi are compared with each other by a two-way ANOVA test (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). MBE online), supporting the completeness of this assembly. These data demonstrated that the M. leonina reference genome was of sufficient quality for further population-scale genome and transcriptome analyses (table 1).

Comparative Genomic Analyses Identified Rapidly Evolving Genes in the NPM
Based on the annotated NPM protein sequences, we constructed a highly resolved phylogenetic tree to estimate the evolutionary status of NPMs among thirteen primate species, including human being, four apes, five Old-World monkeys, three New-World monkeys, and tree shrew (an experimental animal close affinity to primate). The evolutionary tree suggested that NPM diverged from the southern PTM (M. nemestrina) ∼3.9 Ma, and split from RM (M. mulatta) ∼6.2 Ma ( fig. 2). The phylogenetic characteristics of these primates are reflected in the diversity of evolutionary driving force, which may confer phenotypic and immunological differences between NPM and the other primates.
The continuous selective pressure from pathogens acts on host immune genes during primate evolution (Malfavon-Borja et al. 2013). Since the NPM is not a natural host of HIV-1 or SIV, we hypothesize that the history of some bacterial or viral diseases would have resulted in the rapid evolution of immune-related genes in the NPM, thereby shaping its present-day immune system to encounter specific pathogenic invasions, including HIV-1 and SIV. For instance, it is suspected that the TRIM5α and cyclophilin genes in PTMs  and African vervet monkey (Chlorocebus sabaeus); 3 New-World monkeys: white-tufted-ear marmoset (Callithrix jacchus), small-eared galago (Otolemur garnettii), gray mouse lemur (Microcebus murinus); and an experimental animal close affinity to primate: Chinese tree shrew (Tupaia belangeri). The MCMCTree algorithm in PAML4 (Yang 2007) was utilized to evaluate divergence time among 14 species. The nodal ages and the 95% confidence intervals (blue bar) were shown. Sequences of ∼5.59 Mbp single-copy orthologous coding genes from each of 14 species were concatenated as an input sequence of MCMCtree. Four fossil calibration points (A-D) were obtained from a previous report (Perelman et al. 2011), and the nodal divergence dating (E) between primates and tree shrew was obtained from a previous study (Fan et al. 2013). The species figures were obtained from the Ensembl database (https://asia.ensembl.org/).

MBE
had experienced the diversity selection from some unknown viruses, thus leading the generation of a new fusion gene TRIM5-cyclophilin A, which rendered the susceptibility of HIV-1 in PTMs (Newman et al. 2008;Virgen et al. 2008). To test this hypothesis, we compared the annotated NPM protein sequences with the other four closely related Old-World monkey species, including RMs (M. mulatta), cynomolgus macaque (M. fascicularis), southern PTMs (M. nemestrina), and African vervet monkey (Chlorocebus sabaeus). In total, we identified 11,536 homologous gene families and 10,620 single-copy one-to-one orthologous genes among the five Old-World monkey species ( fig. 3A). To characterize the potential adaptive changes in functional genes, we utilized the branch-site model implemented in the likelihood ratio test to detect positively selected genes (PSGs) and identified 197 PSGs in the NPM (M. leonine) lineage (supplementary table S5, Supplementary Material online). Using gene ontology (GO) enrichment analyses, we found that the 197 PSGs could retrieve 29 genes, which were enriched in 22 immune-related GO functional categories, mainly concentrating on regulations of innate immune response, defense response to virus, interferon-alpha (IFNα) production, and interleukin-6 (IL-6) secretion (supplementary table S6, Supplementary Material online). This result exhibits evidence of pathogenic selection on innate immune-related genes in NPMs. In these 29 PSGs, we found that 9 genes (TLR8, CR2, MAVS, CHRNA4, PGC, CLDN18, CLEC1B, ITGA4, and ENPP3) have potential positively selected amino acid sites using Bayes Empirical Bayes analyses (supplementary table S7, Supplementary Material online). These nine genes may play a role in shaping the unique immune response against certain specific pathogens and, concomitantly or coincidentally, have an impact on HIV-1 susceptibility in NPMs. According to the literatures in HIV-1 research, the genes among them most likely related to HIV-1 infection are TLR8, CR2, and MAVS ( fig. 3A). TLR8 is an endosomal pattern recognition receptor that senses single-stranded RNA (ssRNA) in viruses to initiate pro-inflammatory cytokine induction, maturation, and activation of T cells (Heil et al. 2004). CR2 encodes complement C3d receptor 2, which is the main receptor for complement protein C3d, plays an essential role in adaptive immune response, and is associated with susceptibility to HIV-1 infection in humans (Meza et al. 2020). The mitochondrial antiviral protein mitochondrial antiviral signaling protein (MAVS) is a key player in the induction of antiviral responses. HIV-1 can block MAVS-dependent signaling via the mitotic kinase polo-like kinase 1 (Gringhuis et al. 2017). Mutations in MAVS cause it to be insensitive to HIV-1-dependent suppression and a low viral load during disease progression (Stunnenberg et al. 2020). Therefore, we assumed that these genes may play an important role in HIV-1 infection, and examined TLR8 first to validate our hypothesis.

Rapid Evolution of TLR8 Was Associated With Low
Immune Activation in HIV-1 Infected NPMs TLR8 can recognize ssRNA sequences originating from the HIV-1 long terminal repeat U5 region and concomitantly elicit proinflammatory signaling (Heil et al. 2004). TLR8 stimulation of host cells preferentially induces nuclear factor (NF)-κB-regulated cytokines, especially those associated with inflammation, such as TNF-α, IL-6, and IL-12, which contribute to HIV-induced immune activation (Gorden et al. 2005;Meås et al. 2020). The TLR8 variants with weak activity are related to slow decrease of CD4 + T cell and clinical progression of AIDS in HIV-1 infected individuals (Oh et al. 2008;Mackelprang et al. 2014). Based on the findings of genome sequencing, we confirmed the underlying difference in the TLR8 nucleotide sequences among NPMs, RMs, and humans using Sanger sequencing and identified 42 and 18 amino acid changes in TLR8 in NPMs compared with those in humans and RMs, respectively (supplementary fig. S3, Supplementary Material online), and in them the nine amino acid changes under positive selection were verified ( fig. 3B and C). Using western blotting, we observed that the amino acid changes in TLR8 did not disturb its expression in the PBMCs from humans, RMs, or NPMs ( fig. 3D). To validate the potential functional consequences, we overexpressed TLR8 genes from humans, RMs, and NPMs in HEK 293 T cells and treated them with TLR7/8 agonist R848 or TLR8 agonist VTX-2337 and found that both agonists induced dosedependent NF-κB release. Upon the R848 and VTX-2337 stimulation, NPM-TLR8 invoked a significantly lower NF-κB activity than human-or RM-TLR8, implying that NPM-TLR8 is weaker than human-or RM-TLR8 at inducing inflammatory response ( fig. 3E).
Then we checked the secretions of the representative NF-κB-regulated cytokines, TNF-α, and IL-6, in an in vitro coculture experiment. Because other nonviral TLRs, such as TLR2 and TLR4 can also elicit the inflammatory cytokine production, we cocultured whole bloods from healthy NPM, RM, and human donors with TLR8 agonist VTX-2337, and TLR2/4 agonist LPS and found that whole blood from NPMs induced lower TNF-α and IL-6 secretions than did RM or human blood following VTX-2337, but not LPS stimulation, suggesting that TLR8 is responsible for inducing a weaker inflammatory response in NPMs. Meanwhile, upon the inactivated stHIV-1sv or SIV mac239 stimulation, the whole bloods from NPMs also released a lower TNF-α and IL-6 than those from human and RM ( fig. 3F). Correspondingly, the reduced plasma TNF-α and IL-6 concentrations were detected in the NPMs infected with HIV-1 NL4-R3A , stHIV-1sv, and SIV mac239 ( fig. 3G). Thus, HIV-1 and SIV evoked a dampened activity of TLR8 in NPMs both in vitro and in vivo. The weaker NPM-TLR8 function was supported by the lower activation of CD4 + T and CD8 + T cells following TLR8 agonist stimulation in vitro

Upregulation of Interferon Stimulated Genes After stHIV-1sv Infection
We next assessed the biological function of genes that were upregulated following HIV-1 infections. Gene enrichment analysis showed that the tendencies of HIV-1 NL4-R3A and stHIV-1sv was similar in that they induced an innate immune response to acute viral infection; in particular, they activated the type I interferon (IFN-I) signaling pathway, which is a key innate control against microbial invasion (Ivashkiv and Donlin 2014). Whereas SIV mac239 infection exhibited few genes that were significantly upregulated ( fig. 4C-H 5B). However, some ISGs were upregulated following infection. Considering that stHIV-1sv is a more virulent strain than HIV-1 NL4-R3A in NPMs, and the ISGs are mostly upregulated at early time of infection in response to the viral invasion, we extracted the ISGs which were particularly upregulated at 1 wpi and 2 wpi during the stHIV-1 infection and retrieved a small cluster of them, including IFI27, IFI6, MX1, MX2, RSAD2, IFI44, OASL, HERC5, IFNG, and IFIT3 (fig. 5B). Among these genes, MX2 is known as HIV-1 restriction factors (Kane et al. 2013); RSAD2 (Nasr et al. 2012), HERC5 (Paparisto et al. 2018), and IFI44 (Power et al. 2015) have weak activities to inhibit HIV-1 replication; and IFNG is an anti-viral cytokine (Thobakgale et al. 2012). Their selective upregulations in stHIV-1sv infection, but not in SIV mac239 infection, could contribute to the observed suppression of stHIV-1sv replication in NPMs.
We then examined the anti-HIV-1 activity of the remaining genes, including IFI27, IFI6, OASL, and IFIT3, together with a gene ARRDC4 as a negative control, and an HIV-1 restriction factor Tetherin as a positive control. Interestingly, when these genes were over-expressed in 293 T cells, NPM IFI27 showed the strongest anti-pHIV-1 NL4.3-R3A and pstHIV-1sv packaging activity, which was in a similar level with that of Tetherin, implying that IFI27 may play an important role in controlling HIV-1 replication in NPMs (supplementary fig. S6, Supplementary Material online).

IFI27 in NPM Selectively Inhibited HIV-1 Replication
IFI27 is a mitochondrial localized protein belonging to the ISG12 gene family that is commonly induced by IFNs (Parker and Porter 2004) and participates in multiple biological processes, including apoptosis and congenital immunity (Labrada et al. 2002;Rosebeck and Leaman 2008). IFI27 has the capacity to inhibit hepatitis C virus replication ), but whether it is related to HIV-1 replication remains unknown. In this study, IFI27 demonstrated the most considerable upregulations during acute stHIV-1sv infection ( fig. 5B). Furthermore, IFI27 expression at 2 wpi in the PBMCs of infected NPMs was validated by qRT-PCR ( fig. 5C). Genome sequences validated by Sanger sequencing showed that the protein sequences of IFI27 in NPM and RM were identical but showed multiple specific amino acid substitutions Insights into HIV-1-Infected Macaques · https://doi.org/10.1093/molbev/msad101 MBE compared with those of humans ( fig. 5D). Both humanand NPM IFI27 could be induced by exogenous IFN-α stimulation ( fig. 5E), implying that it might be a host-specific ISG that participates in HIV-1 replication.
To confirm this, we overexpressed human and NPM IFI27 in HOS CD4 + CCR5 + T cells and investigated their ability against HIV-1 and SIV replications. NPM IFI27 demonstrated selectively strong anti-HIV-1 but weak anti-SIV activities ( fig. 5F-G), supporting our finding of relatively low HIV-1 replication, while efficient SIV replication in NPMs ( fig. 1A). Furthermore, NPM IFI27 demonstrated a much stronger inhibition against HIV-1 replication than human IFI27 ( fig. 5F), as confirmed by a dose-dependent inhibition assay ( fig. 5H). These results suggest that the anti-HIV-1 activity of IFI27 is host-specific.
To further confirm it, using the siRNA targeting the IFI27, we observed that IFI27 mRNA expression was knocked down ∼50% in the PBMCs from healthy NPMs

MBE
or humans ( fig. 6A). Consequently, the IFI27 protein expressions were reduced 43% and 46% in the cells of NPMs and humans ( fig. 6B and C). After applying the same amount of infective stHIV-1sv and SIV mac239 (MOI = 0.1) to these cells, we observed that stHIV-1sv replication significantly increased ∼78% in the cells from NPMs, but only ∼25% in human cells ( fig. 6D). In turn, SIV mac239 replication was little changed in either cell types ( fig. 6E). These results indicate that IFI27 is a host-specific HIV-1 restriction factor. Together, the significant upregulation of IFI27 and its selective anti-HIV-1 activity provide an explanation for the low HIV-1 replication in NPMs. It also suggests a potential target for genetic intervention in the NPMs for the adoption of a more appropriate HIV-1/ AIDS model.

Discussion
HIV-1 and its precursors have a long-term and complex coevolutionary history with different hosts. SIVs, precursors of HIV-1, have overcome numerous specific defensive barriers in their hosts to cross-species transmissions from monkeys to chimpanzees and eventually to humans. Phylogenetic analyses suggested that the viruses had experienced two stages of evolutionary events: 1) several different SIV strains, at least including SIV gsn , SIV mon , and SIV mus in Cerconpithecus monkeys, had recombined to form a hybrid virus SIV cpz in chimpanzees; 2) the SIV cpz adapted to HIV-1 group M and transmitted in humans (Sauter and Kirchhoff 2019). In adaptations and crosstransmissions to new host species, the viral genes have evolved the ability to evade the defense mechanisms from the corresponding hosts, and in particular the viral accessory genes have obtained the antagonism to the host-specific restriction factors, facilitating the lowering of the species barriers. In turn, the high adaptability of the HIV-1/SIVs in each specific host has a negative fitness effect on a new host. As such, HIV-1 has evolved to be a highly human-specific retrovirus, hampering the development of a suitable nonhuman primate HIV-1/AIDS model. Nowadays, although RMs infected with SIV/SHIV have been widely used in HIV-1/AIDS research, these models have shown some shortcomings for anti-HIV-1 drug and vaccine applications. Besides RMs, the PTMs have been used in HIV-1/AIDS research due to their HIV-1 susceptibility (Agy et al. 1992;Liao et al. 2007). Lacking of functional TRIM5α gene in PTMs has been well documented as a molecular mechanism underlying these observations, which lowers the barrier for HIV-1 cross-species infection in these monkeys (Liao et al. 2007;Newman et al. 2008;Virgen et al. 2008). To overcome another barrier from macaque MBE restriction factor APOBEC3, an engineered HIV-1 derivative, stHIV-1sv, was constructed, and it presented a more robust replication in PTMs (Hatziioannou et al. 2009). However, neither HIV-1 (Agy et al. 1997) nor its derivative stHIV-1sv (Hatziioannou et al. 2009) are pathogenic in PTMs. Thereafter, the stHIV-1sv strain was adapted in vivo for several passages, and its descendants containing several mutations in structural and accessory genes such as vpu, gag-CA, or env have shown adaptations to other macaque-specific restriction factors tetherin, MX2, or the use of the viral coreceptor CCR5, respectively; they still cannot replicate well at chronic stage and induce AIDS in PTMs (Hatziioannou et al. 2014;Schmidt et al. 2019). Meanwhile, CD8 + T cells depletion in PTMs prior to HIV-1 infection can cause pathology but it is not a natural infection state (Hatziioannou et al. 2014;Schmidt et al. 2019). These results indicate that some unknown host-specific defense mechanisms still impede HIV-1 adaptation in this type of monkey. Therefore, revealing the unexplored barriers that inhibit HIV-1 fitness in PTMs will promote the development of HIV-1/AIDS animal model and extend the knowledge of HIV-1 evolution in primates. The interactions between host and pathogen are major drivers of immune-related gene evolution, and speciesspecific genetic changes can help to understand the substitutions differentially impacting on the diverse outcomes of infectious diseases (Chapman and Hill 2012). Here, we observed that the NPM infected with HIV-1 showed a typical feature of low level of immune activation, we then employed large-scale comparative genome analysis to reveal potential genetic underpinning of HIV-1 infection in NPM. We revealed that the TLR8 gene evolved to be functionally less efficient in NPM than in human and RM. As TLR8 has function to initiate pro-inflammatory cytokine induction, maturation and activation (Heil et al. 2004;Gorden et al. 2005;Meås et al. 2020), these changes on TLR8 may contribute to the observation that immunopathology is lacked in HIV-1-infected NPMs. This is similar to the natural SIV host species the sooty mangabey (Cercocebus atys), which does not develop AIDS following SIV infection (Chahroudi et al. 2012;Palesch et al. 2018). Specifically, functional genomics analysis in the Sooty mangabey has found that TLR4 special mutations could cause the attenuated production of pro-inflammatory cytokines, which is an important reason for the blunted immune activation in this monkey (Palesch et al. 2018). Considering that TLRs are an important class of pathogenrecognition receptors in hosts that recognized pathogenassociated molecular patterns in multiple exogenous viruses or bacteria (Wlasiuk and Nachman 2010), and they are subject to different evolutionary pressures from these pathogens, our study proposes another pattern of pathogen-recognition receptor in primates that may hinder HIV-1 or SIV pathogenicity. However, we should point out that in the present study, although TLR8 signaling was weaker in NPMs than in RMs or humans, it does not define a clear link between weak TLR8 signaling in NPMs and a lack of pathogenesis following HIV-1 or stHIVsv infection in this species and that further experiments directly evaluating this hypothesis would be required to make this connection.
Although NPM can be infected by HIV-1, viral replication is low, especially in chronic infection. To replicate well in a new host, HIV-1 must adapt and evade the known and unexplored species-specific antiviral factors. Understanding how the genetic variations and expressions that contribute to the suppression of HIV-1 adaptation in NPMs will provide us with prospects to develop this species as a suitable HIV-1/AIDS model and deepen our understanding of co-evolution between HIV-1 and the possible new hosts. In the present study, through longitudinal transcriptome analysis, we found that a number of ISGs were upregulated in early HIV-1 infections, and several of them have been demonstrated to have the capacities to impede HIV-1 replication. In particular, we identified IFI27, which exhibited more mutations in NPMs than in humans. This difference leads to an enhanced ability of NPM IFI27 to inhibit HIV-1 infection compared with human IFI27. Thus, we considered that NPM, as a nonnatural host of HIV-1 infection, could only have been made susceptible to HIV-1 in an evolutionary incident during which the TRIM5α gene was lost (Newman et al. 2008;Virgen et al. 2008). However, other intrinsic mechanisms of HIV-1 inhibition in ancestral macaques still remain, such as the IFN-I response and the subsequent multiple ISG productions, demonstrating a profound and combinatorial effect on the control of HIV-1 replication in NPM. Furthermore, the immune escape in HIV-1 evolution, well documented in humans, has not emerged in NPMs following several years of infection , which should be a short time for the adaptation of HIV-1 in this monkey. Thus, these factors act together to provide evidence as to why HIV-1 replicates at low levels and does not cause pathology in NPMs, and suggest that the barriers for successful adaptation and fitness of HIV-1 in NPMs are relatively high. However, the present study provides two potential hurdles to overcome for the future engineering or adaptation of HIV-1 derivatives in NPMs.
Taken together, our results provide a comprehensive resource for mechanistic and regulatory insight into HIV-1 infection in a new host, and these findings can be helpful in adapting NPM as a more appropriate HIV-1/AIDS model. Furthermore, the profile of molecular interactions between HIV-1 and NPMs will deepen the understanding of the innate and adaptive defense mechanisms from the host in the cross-species HIV-1 infection.

Animals Used in This Study
Twelve male NPMs and three male Chinese RMs were used in the present study. All subjects were aged 6-10 years and free from SIV, simian type-D retrovirus, and simian T-lymphotropic virus type 1 before HIV-1 or SIV infection. Animals were obtained from the Kunming Primate MBE Research Centre, Kunming Institute of Zoology, Chinese Academy of Sciences and were housed and cared for in accordance with the American Association for Accreditation of Laboratory Animal Care standards. All experimental procedures were approved by the Ethics Committee of Kunming Institute of Zoology, Chinese Academy of Sciences (IACUC18008).
Two HIV-1 strains, HIV-1 NL4-R3A and stHIV-1sv, and SIV strain SIV mac239 were cultured and titrated as described previously . Four NPMs infected with HIV-1 NL4-R3A and four NPMs infected with stHIV-1sv were established as models of low level of replications and nonpathogenic infections. Four NPMs infected with SIV mac239 were established as a model of high level of replication and slow AIDS progression. Three RMs infected with SIV mac239 were established as a model of high level of replication and rapid AIDS progression. A tissue culture infectious dose 50% (TCID 50 ) of 10 5 HIV-1 NL4-R3A , stHIV-1sv, or TCID 50 of 5,000 SIV mac239 was inoculated into each macaque by intravenous injection.
Whole blood from each macaque was collected by venipuncture into an EDTA vacutainer tube. Plasma was collected from the whole blood at 1000 g for 20 min, and PBMCs were separated by Ficoll density gradient centrifugation and stored in liquid nitrogen for further analysis.

Viral Load
The plasma viral loads from HIV-1 NL4-R3A -and stHIV-1svinfected NPMs were quantified using the COBAS TaqMan HIV-1 Test version 2.0 on the COBAS TaqMan 48 Analyzer (Roche Molecular Systems) with a detection limit of 20 copies/mL. The plasma viral loads from SIV mac239 -infected NPMs or RMs were determined by realtime PCR (ABI ViiATM 7 Real-Time PCR System) as previously described (Zhang et al. 2016), with a detection limit of 50 copies/mL. Each sample in each viral infection was tested in three repeated wells, and the mean value of each sample was imported into Graphpad Prism 8.0.1. The data in each infection during 1-6 wpi or 7-69 wpi were compared with each other by two-way analysis of variance (ANOVA) test. The P values were calculated using the "Grouped analyses-Two-way ANOVA-Multiple comparisons-Compare column means (main column effect)" model.

Genome Sequencing
Genomic DNA was extracted from blood samples of a male NPM (M. leonina) from Yunnan Province, China. The NPM genome was produced using an all-genome shotgun strategy on a next-generation sequencing platform. To reduce the risk of nonrandomness, 13 paired-end sequencing libraries with six insert sizes (250 bp, 450 bp, 2 kb, 5, 10, and 15 kb) were constructed and sequenced. In total, our sequencing data reached 425.38 Gbp with a high coverage of ∼141.79X.

RNA Purification and RNA-sequencing
Total RNA from 5 × 10 6 PBMCs from each infected macaque was extracted using TRIzol Reagent (Life Technologies, Carlsbad, CA, USA) and purified with RNeasy mini kits (Qiagen, Valencia, CA, USA) according to the manufacturers' protocols. RNA integrity was monitored on 1% agarose gel and assessed using the RNA Nano 6,000 assay kit of the Agilent Bioanalyzer 2,100 system (Agilent Technologies, Santa Clara, CA, USA). RNA purity was checked using a Nanophotometer spectrophotometer (IMPLEN, Los Angeles, CA, USA), and the concentration was measured using a Qubit RNA assay kit in a Qubit 2.0 MBE Fluorometer (Life Technologies). For sequencing, 1.5 μg RNA per sample was used by an RNA Library Prep kit for Illumina (NEB, USA). Following cluster generation, library preparations were sequenced on an Illumina HiSeq 4,000 platform, and 150-bp paired-end reads were generated.

Transcriptome Analyses
Read alignment and quality control: We obtained 99 transcriptomes spanning multiple infection stages from NPMs infected with different viruses. First, quality control was performed using fastp ), any reads with N content exceeding 10% or low-quality (Q ≤ 5) bases exceeded 50% of the base total, and the corresponding paired reads were removed. Second, Tophat 2 (Kim et al. 2013) was used to align paired-end reads to the M. leonina reference genome, which was assembled in our lab. Lastly, further quality control was performed using Picard Tools (http:// broadinstitute.github.io/picard/) with the command Collect RnaSeqMetrics. The sequencing metrics %mRNA Bases, % Intergenic Bases, and Median 5′ to 3′ Bias were selected to measure the sequence quality. To detect outliers, a quality z-score was calculated for each metric, and samples with low quality (z-score > 2 for % intergenic bases; z-score < −2 for %mRNA Bases, Median 5′ to 3′ Bias) were identified, and any sample with greater than one outlier value was removed due to sequencing quality concerns. Four samples were removed, and the remaining 95 samples were used for further analysis (supplementary table S10, Supplementary Material online).
Expression quantification: Cufflinks (Trapnell et al. 2010) was used to measure the expression of genes; cuffquant calculated the read count of each gene in each sample and generated an intermediate file that was used for further analyses with cuffnorm and cuffdiff; cuffnorm normalized the read count from cuffquant to FPKM (fragments per kilobase of exon model per million mapped reads). Gene expression levels were quantified using read count and FPKM. Differential expression analyses: Cuffdiff ) was used to perform pairwise differential expression analyses with default parameters between samples at each infection and noninfection stage. P values were corrected for multiple comparisons using the Benjamini-Hochberg procedure to estimate the false discovery rate (FDR). FDR < 0.05 and | log 2 (fold change) | > 2 were required to define DEGs. We used g: Profiler (https://biit.cs.ut.ee/ gprofiler/) (Reimand et al. 2016) to perform functional enrichment analyses of DEGs, and P-values were corrected for multiple comparisons using the Benjamini-Hochberg procedure to estimate the FDR.
To verify whether IFI27 can be induced by IFN-α, PBMCs (1 × 10 6 /well in 24-well plates) from three healthy human or NPM donors were co-cultured with 200 IU IFNα (Sigma). Cells were collected at different time points and total RNA was extracted using TRIzol reagent (Life Technologies), and cDNA was generated with a PrimeScript RT reagent Kit with gDNA Eraser (Takara). Expression levels of IFI27 were determined by the real-time qPCR assays described above.
After the anti-HIV-1 packaging activities of IFI27 were screened out, HOS CD4 + CCR5 + cells were transfected with 200 ng of IFI27 or IFI6-expressing vector from NPMs or humans (pcDNA3.1-NPM-IFI27, pcDNA3.1-human-IFI27) in 24-well plates using jetPRIME transfection reagent. Tetherin and PSGL-1 from humans cloned into the pcDNA3.1 vector were used as positive controls, and the pcDNA3.1 vector was used as a negative control. 24 h post-transfection, the cells were infected with HIV-1 NL4-R3A or stHIV-1sv strains at a multiplicity of infection (MOI) of 0.1 (infectious titers were determined based on TZM cells). At 6 h post-infection, the cells were washed three times with PBS at 1000 g/8 min to remove the uninfected viruses, and a new culture medium was added. After 48 h of infection, the HIV-1 particles in the culture supernatant were determined using the real-time qPCR assay described above, and the inhibition of viral replication by each gene of interest was calculated.
To determine the species-specific anti-HIV-1 activity of IFI27, we overexpressed NPM IFI27 and human IFI27 in HOS CD4 + CCR5 + cells, and determined the reduction in viral replication. Briefly, a series of pcDNA3.1-NPM IFI27 or pcDNA3.1-human IFI27 (0 ng, 100 ng, 300, 500, and 700 ng) was transfected into HOS CD4 + CCR5 + cells in 24-well plates. At 24 h post-transfection, the cells were infected with stHIV-1sv or SIV mac239 strains at an MOI of 0.1 and washed at 6 h later. At 48 h post-infection, the different antiviral activities of human IFI27 and NPM IFI27 were evaluated as described above.
In contrary, we knocked down NPM IFI27 and human IFI27 in PBMCs from NPMs and humans, and examined the increase in viral replication. Briefly, PBMCs (2 × 10 6 / mL) from 3 healthy NPMs and human donors were cultured in RPMI 1,640 medium with 10% FBS in 6-well plates, after stimulated with ConA (5 μg/mL) and IL-2 (50 IU/mL) for 2 days, the cells were transfected with siRNA targeting NPM IFI27, human IFI27 and siRNA-NC. After 48 h, the transfected cells from one plate were collected for