Pathogenic infection of Rhesus macaques by an evolving SIV-HIV derived from CCR5-using envelope genes of acute HIV-1 infections

Abstract

For studies on vaccines and therapies for HIV disease, SIV-HIV chimeric viruses harboring the HIV-1 env gene (SHIVenv) remain the best virus in non-human primate models. However, there are still very few SHIVenv viruses that can cause AIDS in non-CD8-depleted animals. In the present study, a recently created CCR5-using SHIVenv_B3 virus with env gene derived from acute/early HIV-1 infections (AHI) successfully established pathogenic infection in macaques. Through a series of investigations on the evolution, mutational profile, and phenotype of the virus and the resultant humoral immune response in infected rhesus macaques, we found that the E32K mutation in the Env C1 domain was associated with macaque pathogenesis, and that the electrostatic interactions in Env may favor E32K at the gp120 N terminus and "lock" the binding to heptad repeat 1 of gp41 in the trimer and produce a SHIVenv with increased fitness and pathogenesis during macaque infections.

Keywords: Envelope; Pathogenesis; Rhesus macaque; SIV-HIV chimeric viruses.

Fig. 1. Viremia during infection and passage of SHIVenv B3 and cloned SHIVenv B3 in macaques

Twenty unique clade B AHI Envelopes were inserted into the SHIVenv_KB9 backbone using yeast homologous recombination as described. These AHI SHIVs (100 IU of each) were pooled and used intravenously inoculate two rhesus monkeys of which one, m328-08 was infected (A). Sequence analyses of the SHIVenv in the m328-08 plasma revealed infection by a single SHIVenv_B3 and the absence of any of the other SHIVenv’s. (B) Plasma and PBLs from animal 328-08 were extracted from day 19 and day 41 after inoculation, and then intravenously inoculated into two new rhesus monkeys, m165-05 and m349-08. Both macaques were infected and reached peak viremia at day 21. Since a prolonged infection was observed in the m165-05 and clearance by day 100 in the m349-08 animal, plasma and cells from the day 21 m165-05 animal were used to infect four animals (D). The day 21 virus in plasma was also PCR amplified and the env gene was cloned into an SIV_E660 and SHIVenv_KB9 backbone. These cloned viruses were then used to infect four macaques each (panels C and E, respectively). (F) Viral RNA levels in the macaques infected with the cloned SHIV_KB9-B3_m165-05d21 were compared to viral RNA levels previously described for SIVmac251 and SHIVenv_BaL infections.

Fig. 2. CD4 T cells in the blood of infected macaques

Total CD4 T cell counts were monitored in block in the m165-05 (A) and in the three sets of four macaque infected with the SHIV_E660-B3_m165-05d21 (B), with the m165-06 day21 plasma/cells (C), and the SHIV_KB9-B3_m165-05d21 clone (D). The percentage of central memory CD4+ T cells were also monitored for the first 40 days of infection in these macaques (F, G, and H, respectively). (E) Percentage of CM CD4+ T cell in the macaques infected with the cloned SHIV_KB9-B3_m165-05d21 were compared to those previously described for SIVmac251 and SHIVBaL infections.

Fig. 3. Genetic diversity in SHIVenv B3 during infections of m328-08, m349-09, and m165-05 animals

(A) Synonymous substitution above a 1% frequency in the SHIV populations infecting macaques m328-08, m349-09, and m165-05 were analyzed during the course of infection using Roche 454 Next Generation Sequencing. Genetic diversity in the m349-09 (B) and m165-05 (C) SHIVenv derived from approximately 2000 reads in each of the C1–C2 (set 1F), C2–C4 (set 2F), and C4–C5 domains (set 3F) as measured following Roche 454. PCR amplification and 454 sequencing of the SHIV env using the same primer sets revealed a background mutation frequency of ~0.001 substitutions/nt. See Materials and Methods for detailed protocols.

Fig. 4. Evolution of SHIVenv B3 during infections in macaques

The unique SHIVenv B3 clones found at a minimal of 5% in m165-05 over time of infection were aligned and presented in maximum likelihood phylogenetic trees (A). The synonymous amino acid substitutions, insertions and deletions in the V1 domain of Env are shown (B) over time for the SHIVenv_B3 infected macaque. (C) The phylogenetic trees for the SHIVenv viruses found at a minimal of 5% in the quasispecies of the three sets of four macaques infected with the SHIV_E660-B3_m165-05d21, with the m165-06 day21 plasma/cells, and the SHIV_KB9-B3_m165-05d21 clone.

Fig. 5. Entry efficiency of the env variants emerging in m165-05 and role of neutralizing antibodies in selection

HIV-1env_B3, the counterpart to SHIVenv_B3 has been propagated and tested for replicative fitness(Krebs, Tian et al., 2016). Mutations were generated in HIV-1env_B3 to generated the HIV-1env_B3_m165d98.1/d143 and HIV-1env_B3_m165d63.3/d98.3, variants of the SHIVenv virus identified at days 63, 98, and 143 in the m165-05 animal. (A) The entry efficiency of the HIV-1Env_B3, Env_B3_m165d98.1/d143, and Env_B3_m165d63.3/d98.3 was determined in 9 replicates in TZM-bl cells. Titers of viruses was determined by NL4-3 RT activity in cell-free supernatant. Following 2 hr incubation of TZM-bl with equal titers of virus, luciferase activity was quantified from cell lysates at 48 hours. (B) A bar graph showing the fraction of the three Env variants at different days post infection in m165-05. (C) Estimated production of SHIVB3₁₆₅ variants based on proportion at the different time points, the relative fitness of each variant, and inhibition by neutralizing antibodies (derived from Fig. 6). The line shows the actual viral loads during the course of this m165-05 infection.

Fig. 6. Neutralizing antibody activity of plasma from m165-05 and m349-08 on the HIVenvB3 variants representing the SHIVenv counterparts in m165-05

Plasma from m165-05 and m349-08 was obtained at days 28 (A, B), 42 (C, D), 63 (E, F), 98 (G, H), and 143 (I, J), diluted 1:5 up to 1:3125, and tested for inhibition of the HIV-1Env_B3, Env_B3_m165d98.1/d143, and Env_B3_m165d63.3/d98.3 variants in TZM-bl assays. All neutralizing antibody assays were performed in triplicate.

Fig. 7. Co-receptor usage and sensitivity to entry inhibitors

(A) The SHIV_KB9-B3_m165-05d21 clone was tested for ability to utilize human BOB, BONZO, CCR2b, CCR3, CCR5, CCR8 and CXCR4 with human CD4 for entry into GHOST cell lines. Virus in cell-free supernatant was measure by EIA for SIV p27. TZM-bl cells were incubated for 1 hour in varying concentrations of either the CXCR4-inhibitor AMD-3100 (B) or the CCR5-inhibitor TAK-779 (C), and subsequently infected with SHIV_KB9-B3_m165-05d21. TZM-bl luciferase activity was quantified 48 hours later.

Fig. 8. Models of SOSIP gp140 trimeric structure with the C1 E32K and V1 mutations

(A) Structural representation of the gp140 trimer (green;PDBID: 4TPV). Residue E32 from gp140 is represented in red and the relative position of gp140 trimer substituted residues are represented by the blue circles (I143, N148, S151). (B) Trimeric gp140 (PDBID: 4TPV) is depicted in complex with the 35O22 and PGT128 neutralizing antibodies. The relative position of E32 from the gp140 trimer is predicted to be in proximity to the 35O22 antibody. (C) gp120 (green) and gp41 (blue) are depicted in complex with the neutralizing 35O22 antibody. Residue E32 (red) from gp120 is predicted to be in proximity to gp41 and 35O22. (D) Spatial rotation of structural image depicted in (C). GP140 trimer representations are provided in green cartoon form in order to provide the correct orientation relative to the viral membrane (E; left). A Structural model of gp120 E32 (red circle) and surface electrostatics of gp41. The location of residues SE₆₁₉ of gp41 are circled in blue. (right) and both mutations E32K in gp120 and SE₆₁₉DD were modeled into the structure using in silico mutagenesis. Surface electrostatics were re-modeled to accommodate the new mutations. All structures were represented with Pymol (The PyMOL Molecular Graphics System, Version 1.8 Schrödinger, LLC.).