| dc.description.abstract | SHIV infections in rhesus macaques have been used to extensively study accessory proteins of HIV-1 involved in pathogenesis as well as in vaccine development. All primate lentiviruses encode for a Vif (virion infectivity factor) protein, which is required for HIV-1 replication in primary CD4+ T cells and macrophages. The Vif protein interacts with APOBEC3 (A3) proteins promoting their accelerated degradation by the 26S proteosome. Sequence analysis of Vif proteins from different lentiviruses revealed that there are two highly conserved domains in the carboxyl terminus that are required for recruitment of the of the Vif-CBF-β-Cul5/Elongin B/C/Rbx-1 E3 ubiquitin ligase complex. These domains are the viral BC box, SLQ(Y/F)LAL, and Zn2+ binding (H-X5-C-X17-18-C-X3-5-H; HCCH) motifs. Previous cell culture studies have shown that the introduction of amino acid substitutions into the SLQ(Y/F)LA motif resulted in decreased binding of Vif to Elongin C, while substitutions into the HCCH motif prevented the interaction of Vif with Cullin 5. In this first study, we introduced two amino acid changes in the highly conserved SLQYLA domain (S147A, L148A; AAQYLA) of the SIV Vif protein in SHIV. In vitro, the resulting virus, SHIVVifAAQYLA, replicated in A3G negative CEM-SS cells but failed to replicate in A3G positive CEM cells. We also showed that hA3G was incorporated into the virion. Following these in vitro studies, SHIVVifAAQYLA was inoculated into three rhesus macaques, which were followed for over six months to assess various viral and immunological factors throughout the duration of infection. All three macaques did not develop significant CD4+ T cell loss over the course of infection, had plasma viral loads that were over 100-fold lower than macaques inoculated with parental SHIVKU-1bMC33, and developed no histological lesions in various lymphoid tissues, and developed immunoprecipitating antibodies. DNA and RT-PCR analysis revealed that only a select number of tissues were infected with virus, while sequence analysis of PBMC and select tissue DNA (at necropsy) showed that the site-directed changes were stable during the first three weeks after inoculation but thereafter the S147A amino acid substitution changed to a threonine in two of the three macaques. However, the L148A substitution remained stale in the vif amplified from the PBMC and select tissues at necropsy in all three macaques. Extensive sequence analysis of the vif, vpu, nef, and env genes revealed a increased number of G-to-A mutations in the genes amplified from macaques inoculated with SHIVVifAAQYLA. The majority of these mutations (>85%) were in the context of 5'-T85%) were in the context of 5'-TC (minus strand) and not 3'-CC, suggesting that one or more of the rhesus A3 proteins may be responsible for the observed mutational patterns. To determine if infectious virus was present in the plasma at necropsy, plasma from the three macaques inoculated with SHIVVifAAQYLA were pooled and intravenously inoculated into a naïve macaque. This macaque maintained its levels of circulating CD4+ T cells throughout the duration of infection, maintained viral loads below the limits of detection, and did not produce immunoprecipitating antibodies. However, gag was present in the DNA and RNA isolated from PBMC throughout infection and in select tissues at necropsy. The results from this first study showed for the first time the importance of the SLYQLA domain in vivo in viral pathogenesis. It also showed that mutations in vif could lead to a persistent infection in rhesus macaques resulting in the accumulation of G-to-A substitutions in the viral genome. In the second study, we used the SHIV/macaque model of infection to compare the replication and pathogenicity of SHIVs that express a Vif protein in which the entire SLQYLA (SHIVVif5A) or HCCH (SHIVVifHCCH(-)) domains were substituted with alanine residues. Each virus was inoculated into three rhesus macaques where various viral and immunological parameters were followed for six months. Our in vitro results indicate that in the presence of these mutant Vif viruses, rhA3G is incorporated into the virion, stably expressed, restricts, and accumulates G-to-A substitutions (plus strand) in the nef,Italics/> gene of the mutated viral genomes. gene of the mutated viral genomes. In vivo, all macaques maintained a stable level of circulating CD4+ T cells, developed low viral burdens, maintained engineered mutations, yielded no histological lesions, and developed immunoprecipitating antibodies by 12 weeks post-inoculation. However, the production of viral RNA only persisted in macaques inoculated with SHIVVifHCCH(-). The analysis of vif sequences amplified from PBMC DNA between weeks 0-16 during SHIVVifHCCH(-) infection revealed an increased number of G-to-A substitutions that increased with time in two of the three macaques. Sequence analysis of nef and vpu from the small intestine (ileum), thymus, and the spleen showed G-to-A substitutions in nef genes isolated from macaques inoculated with SHIVVifHCCH(-). Macaques inoculated with SHIVVif5A effectively controlled the virus three weeks post-inoculation and no viral sequences could be amplified from tissue DNA. These studies showed that the SLQYLA and HCCH domains are critical for viral pathogenesis in vivo and that there may exist APOBEC3 negative reservoirs in the rhesus macaque that allow for low levels of viral replication and persistence but not disease. Therefore, this study suggests that mutations targeted to one or more functional conserved domains within the Vif protein may limit viral replication and generate an effective immune response leading to the "self-inactivation" of the virus by the activities of various APOBEC3 proteins resulting in a possible live-attenuated vaccine candidate. The APOBEC3 family of restriction factors has been shown to inhibit certain retroviruses and retroelements. The APOBEC3 family in humans is comprised of seven cytidine deaminases (A3A, A3B, A3C, A3D, A3F, A3G, and A3H) that catalyze the deamination of cytidine to uracil on single-stranded DNA or RNA. While the human APOBEC3 repertoire has been extensively studied, the full complement of these proteins in the rhesus macaque remains unknown. Sequencing of the rhesus macaque genome has led to the identification of the rhesus homologues A3B, A3C, A3D, A3F, A3G, and A3H. Finally, we identified a human A3A (hA3A) homologue in the rhesus macaque (rhA3A) and presents evidence that both the human and rhesus Apobec3 genes are orthologous. We show that rhA3A is highly expressed in activated CD4+ T cells, widely expressed in both the visceral and central nervous system tissues of the rhesus macaque, and is degraded in the presence of the human immunodeficiency virus (HIV-1) and simian-human immunodeficiency virus (SHIV) genomes in a Vif-dependent manner. Our results also indicate that rhA3A reduced the level of infectious SHIVΔvif by approximately 20-fold and HIV-1Δvif by 3-fold. Human and monkey A3A amino acid sequences are 81% homologous and can be distinguished by a three amino acid indel located between residues 27-30. When these residues were deleted from rhA3A (rhA3AΔSVR), the antiviral activity of rhA3A was abolished suggesting that these residues are critical for lentivirus inhibition. Select APOBEC3 proteins are incorporated into the virion and can inhibit reverse transcription and/or induce G-to-A hypermutation in nascent reverse transcripts in the next target cell. Previous studies revealed that rhA3G is incorporated into SHIVΔvif virions and exerts its antiviral activity in target cells by an increase in cytidine deamination of newly synthesized minus-strand viral DNA from cytosines to uracils, leading to G-to-A substitutions (plus strand) in the viral genome. We were able to detect the incorporation of rhA3A into SHIVΔvif and to a lesser extent in SHIV virions; however, we were unable to detect the incorporation of rhA3A into either HIV-1 or HIV-1Δvif virions. Even though rhA3A is incorporated into SHIVΔvif virions and potently restricts SHIVΔvif similar to rhA3G, rhA3A produced an approximately 5-fold decrease in the number of G-to-A mutations compared to rhA3G. Unlike hA3A, rhA3A did not inhibit adeno-associated virus 2 (AAV-2) replication and L1 retrotransposition. This data suggests for the first time an evolutionary switch in primate A3A virus specificity and provides evidence that a primate A3A protein can inhibit lentiviral replication. | |
