|
Products Related to Zika, West Nile, Dengue, Malaria, T.B, Chikungunya, HIV, SARS |
|
Product# 1039 HIV-1 YU2 (M Tropic)Envelope Glycoprotein gp41 |
|
Product# HIV-1 M/T Tropic Envelope |
|
Product# 1081 HIV-1 gp120 (ADA) |
|
Product# 1011 HIV-1 gp120 (subtype C) |
|
Product# 1031 HIV-1 gp120 (YU2) |
ABSTRACT
The phase III RV144 human immunodeficiency virus (HIV) vaccine trial conducted in Thailand remains the only study to show efficacy in decreasing the HIV acquisition risk. In Thailand, circulating recombinant forms of HIV clade A/E (CRF01_AE) predominate; in such viruses, env originates from clade E (HIV-E). We constructed a simian-human immunodeficiency virus (SHIV) chimera carrying env isolated from an RV144 placebo recipient in the SHIV-1157ipd3N4 backbone. The latter contains long terminal repeats (LTRs) with duplicated NF-κB sites, thus resembling HIV LTRs. We devised a novel strategy to adapt the parental infectious molecular clone (IMC), R5 SHIV-E1, to rhesus macaques: the simultaneous depletion of B and CD8+ cells followed by the intramuscular inoculation of proviral DNA and repeated administrations of cell-free virus. High-level viremia and CD4+ T-cell depletion ensued. Passage 3 virus unexpectedly caused acute, irreversible CD4+ T-cell loss; the partially adapted SHIV had become dual tropic. Virus and IMCs with exclusive R5 tropism were reisolated from earlier passages, combined, and used to complete adaptation through additional macaques. The final isolate, SHIV-E1p5, remained solely R5 tropic. It had a tier 2 neutralization phenotype, was mucosally transmissible, and was pathogenic. Deep sequencing revealed 99% Env amino acid sequence conservation; X4-only and dual-tropic strains had evolved independently from an early branch of parental SHIV-E1. To conclude, our primate model data reveal that SHIV-E1p5 recapitulates important aspects of HIV transmission and pathobiology in humans.
IMPORTANCE Understanding the protective principles that lead to a safe, effective vaccine against HIV in nonhuman primate (NHP) models requires test viruses that allow the evaluation of anti-HIV envelope responses. Reduced HIV acquisition risk in RV144 has been linked to nonneutralizing IgG antibodies with a range of effector activities. Definitive experiments to decipher the mechanisms of the partial protection observed in RV144 require passive-immunization studies in NHPs with a relevant test virus. We have generated such a virus by inserting env from an RV144 placebo recipient into a SHIV backbone with HIV-like LTRs. The final SHIV-E1p5 isolate, grown in rhesus monkey peripheral blood mononuclear cells, was mucosally transmissible and pathogenic. Earlier SHIV-E passages showed a coreceptor switch, again mimicking HIV biology in humans. Thus, our series of SHIV-E strains mirrors HIV transmission and disease progression in humans. SHIV-E1p5 represents a biologically relevant tool to assess prevention strategies.
INTRODUCTION
Human immunodeficiency virus (HIV) continues to be a major global public health issue, and attempts to produce a safe and effective vaccine have not yet been successful. Although major research efforts have focused on AIDS vaccine development, a safe and effective vaccine is not yet available. The RV144 vaccine trial resulted in 31.2% protection, making it the first phase III study to show limited but significant efficacy in preventing HIV acquisition (1). Infection-risk analyses revealed that nonneutralizing antibodies (Abs) targeting the HIV Env V1/V2 region correlated inversely with infection risk (2, 3), and subsequent studies identified other nonneutralizing functional activities that also appear to be inversely correlated with risk (4, 5). A number of monoclonal antibodies (mAbs) have been isolated from RV144 vaccine recipients, including mAbs directed against the Env regions linked to lower risks of HIV acquisition (6). However, the potential of these mAbs to prevent acquisition in a relevant nonhuman primate (NHP) model has not yet been assessed.
A number of NHP studies have sought to replicate key findings of RV144 with corresponding simian immunodeficiency virus (SIV) immunogens. Of note, these challenge studies confirmed some of the key findings of RV144 (7, 8). However, anti-HIV Env-directed Ab responses cannot be evaluated in SIV models. A simian-human immunodeficiency virus (SHIV) displaying the HIV envelope circulating in Thailand will represent an important tool to assess active- and passive-immunization strategies in NHP models.
Almost all newly acquired HIV infections involve R5-tropic strains; approximately 90% of new HIV infections occur through mucosal exposures. Most recently transmitted HIV strains are relatively difficult to neutralize and exhibit a tier 2 neutralization phenotype. Consequently, NHP models to evaluate strategies to prevent HIV acquisition should consider these key biological characteristics of HIV acquisition among humans.
SHIVs have been used as tools to evaluate anti-HIV Env Ab responses. Most SHIVs used to date have been built from the backbone of SIVmac239, with the exception of a recently reported strain (9). In general, SHIVs carry HIV vpu, tat, rev, and env. Previously, we generated a panel of non-clade B SHIVs, including several encoding HIV clade C envelopes. Furthermore, we engineered the long terminal repeat (LTR) of SIVmac239 to resemble that of HIV more closely (10). The HIV LTR generally contains at least two, but up to four, NF-κB sites, in contrast to the SIVmac239 LTR, which has only one such site (11). A larger number of NF-κB sites renders proviruses more responsive to stimulation by cytokines that act through the NF-κB pathway, including tumor necrosis factor alpha (TNF-α). In Thailand, where the RV144 trial was conducted, CRF01_AE strains circulate, and Env is predominantly derived from HIV clade E (1, 12).
Here we report the construction, novel in vivo adaptation, and pathogenicity of a SHIV encoding the env gene isolated from a placebo recipient of the RV144 vaccine efficacy trial in Thailand. This SHIV, termed SHIV-E1p5, is R5 tropic, has a tier 2 neutralization phenotype, is mucosally transmissible, and is pathogenic, as indicated by its ability to induce AIDS in NHPs. During adaptation, SHIV-E1 and progeny strains mimicked an important aspect of HIV CRF01_AE, namely, the ability to switch coreceptor usage and become dual tropic or solely X4 tropic. Deep-sequencing analysis of the various virus isolates during adaptation revealed env mutations uniquely associated with dual-tropic or X4-only phenotypes; such mutations were absent in the final R5-only SHIV-E1p5 isolate. Our newly created SHIV-E1 reflects key biological aspects of HIV clade E in humans, and the final isolate, SHIV-E1p5, can be used as a model to develop prevention strategies targeted against CRF01_AE.
RESULTS
Construction of SHIV carrying CRF01_AE env.
To determine the ability of RV144 vaccine-induced Abs to protect against HIV acquisition by passive immunization in an NHP model, we generated a SHIV carrying HIV CRF01_AE env. Six HIV CRF01_AE env clones of recently transmitted viruses isolated from placebo group RV144 participants were tested for infectivity as pseudotyped viruses generated by the cotransfection of HIV CRF01_AE env genes with an env-deleted provirus into 293T cells. All infectious env genes were used to generate SHIV clones according to the construction schema (Fig. 1). Overall, 30 infectious SHIV clones were obtained, as evidenced by the transfection of 293T cells and analysis of cell-free supernatants in TZM-bl cells (data not shown). One of them, SHIV harboring env clone 620345.2, was chosen for further development and renamed SHIV-E1 for the sake of simplicity. The backbone, SHIV-1157ipd3N4 (10), was chosen because it contains a 3′ engineered LTR with a duplication of the NF-κB site. As such, the engineered LTR resembles that of HIV more than that of SIVmac239, which contains only one NF-κB site. Of note, all HIV LTR elements contain at least two NF-κB sites, with different clades containing up to four such sites. The resulting SHIV-E1 was tested by DNA sequence analysis, coreceptor usage, and neutralization phenotype. SHIV-E1 was exclusively R5 tropic and relatively difficult to neutralize, corresponding to a tier 2 neutralization phenotype. Cell-free SHIV-E1, prepared by transfection of 293T cells, replicated in TZM-bl cells, U87.CD4.CCR5 cells, and human peripheral blood mononuclear cells (PBMC) depleted of CD8+ cells. PBMC from rhesus macaques (RMs) (25 donors) and pig-tailed macaques (5 donors) did not support the replication of the parental virus, even after the depletion of CD8+ cells. Clearly, a special approach needed to be designed to achieve successful adaptation to RMs.
Construction of SHIV-E1. (A) Several infectious HIV-E1 env clones were identified from newly infected individuals of the RV144 placebo group. env sequence divergence is shown. (B) To generate the initial parental SHIV-E1 provirus, env clone 620345.2 (0.0042) (red box) was inserted into the SHIV-1157ipd3N4 backbone (10). The latter infectious molecular clone had been engineered to contain duplicated NF-κB sites (NN) in the long terminal repeats (LTRs). As such, this SHIV LTR resembles the HIV LTR better than the original SIVmac239 LTR that contains only one NF-κB site. For the sake of simplicity, we use the term “SHIV-E1” for our initial infectious molecular SHIV clone. Of note, in the RV144 CRF01_AE strains used, the env sequences per se originated from HIV clade E strains rather than clade A strains. Thus, the env clone 620345.2 (0.0042) (shown in red boxes in panels A and B) used to generate the SHIV-E1 construct was renamed “HIV-E1 env.” The cloning sites KpnI (K) and BamHI (B) are shown. TM, transmembrane region.
In vivo transfection and double-immune depletion to adapt SHIV-E1.
To achieve replication of SHIV-E1 in RMs, we did not rely solely on the inoculation of cell-free virus. Instead, we opted for intramuscular (i.m.) inoculation of plasmid DNA encoding infectious SHIV-E1 (Fig. 2A and andBB and and3,3, top left). The rationale for this approach was to generate virus particles displaying RM host proteins on their surfaces after budding from transfected muscle cells. In addition, we inoculated the first animal, animal R547, intravenously (i.v.) with three doses of cell-free SHIV-E1 on alternate days (Fig. 2A and andBB and and3,3, top left). To give the virus the best chance to replicate, we eliminated adaptive immunity by depleting CD8+ cells with a cytotoxic mAb and by depleting B cells with the anti-CD20 mAb rituximab (Rituxan; Genentech Inc.) (Fig. 2A and and3,3, top left). Of note, the anti-CD8 mAb affects both CD8+ T cells as well as the majority of natural killer (NK) cells in RMs, thereby also significantly interfering with host innate immune responses. This approach was successful: peak viremia reached >109 viral RNA (vRNA) copies in the first animal (Fig. 3, top left).
Early passage history of SHIV-E1 in rhesus macaques (RMs). (A) Schema of serial SHIV-E1 passage through different RMs. The first animal, R547, was inoculated intravenously (i.v.) with the cell-free supernatant from 293T cells transfected with the infectious molecular clone SHIV-E1 as well as with SHIV-E1 proviral DNA by the intramuscular (i.m.) route. The second RM, R551, received blood i.v. from animal R547 on day 53. Both animals R547 and R551 were immunodepleted for CD8+ cells and B cells prior to virus exposure (black box). Animal R547 is denoted in red, indicating that the reisolated virus was exclusively R5 tropic. Virus isolated from animal R551 (day 57; day of necropsy) was found to be dual tropic (gradient from red to blue); initial red shading for animal R551 indicates that virus isolated early (day 14) was R5 tropic. Both animals R555 and R561 received dual-tropic virus (blue shading). Solid red arrows indicate the passage direction; olive arrows denote dead-end passages with filtered plasma transfer with dual-tropic virus. Dotted black boxes indicate RMs that were PCR positive for T. cruzi. Circles with P1, P2, P3a, and P3b indicate the passage numbers. (B) Summary of inocula, route of inoculation, and source of the inoculum for each animal.
Replication kinetics of SHIV-E1 in RMs during serial passage. RMs R547 and R551 were immunodepleted by the administration of anti-B cell (rituximab) (green arrows) and anti-CD8+ cell (cM-T807) (black arrows) mAbs on the days indicated. Animal R547 was inoculated with SHIV-E1 IMC DNA i.m. at two sites on day 0 (red asterisk with down arrow) and cell-free virus i.v. on three different days (7 ml each dose) (red down arrows). Blood was transferred to animal R551 on day 53 after inoculation of animal R547. Animal R551 developed signs of AIDS and was euthanized on day 57. Red up arrows indicate the isolation of virus from animals R547 (day 203) and R551 (days 14 and 57); olive green arrows indicate simultaneous, dead-end passages with filtered plasma containing dual-tropic virus from donor R551 to animals R555 and R561 (neither recipient was pretreated with rituximab or cM-T807); circles with P1, P2, P3a, and P3b indicate the passage numbers; red boxes indicate R5-tropic virus; blue boxes indicate dual-tropic virus; dotted black boxes indicate RMs that were PCR positive for T. cruzi; red lines with red circles indicate viral RNA copies per milliliter on a log scale; and axes on the right indicate absolute cell numbers (103).
As expected, the depletion of B cells and CD8+ cells resulted in profound lymphopenia, to which the host responded initially with the expected compensatory rise in absolute CD4+ T-cell numbers (Fig. 3, top left), reaching a peak of >3,000 cells/mm3 concomitant with a high peak vRNA level, thus providing abundant target cells for SHIV-E1. The latter replicated to very high levels, with the simultaneous destruction of target cells, as evidenced by the precipitous drop of absolute CD4+ T-cell numbers to <200 cells/mm3 on day 53, a clear indication of viral pathogenicity. In vivo transfection with infectious proviral DNA combined with the repeated administration of cell-free virus and double-immune depletion of the host was successful.
Passage of SHIV-E1-infected blood to the second donor and virus isolation.
On day 53 postinoculation, blood from animal R547 was transferred i.v. to the second recipient, animal R551. On day 56 postinoculation, this RM developed neck and facial swelling, which prompted an emergency diagnostic workup. Peripheral blood smears showed severe parasitemia. The animal was euthanized immediately on day 57 (Fig. 3, top right); the parasite was identified subsequently as Trypanosoma cruzi. Animal R551's profound peripheral blood CD4+ T cell loss toward the end of its course may have facilitated severe parasitemia by a known opportunistic pathogen of humans with end-stage HIV disease (13).
In order to continue virus passage, plasma collected at the time of necropsy was filtered to remove T. cruzi and given i.v. to the next two recipients, animals R555 and R561 (passage 3a [P3a] and P3b) (Fig. 2A and andBB and and3,3, bottom). Of note, these two animals were not immunodepleted because the virus in the previous donor, R551, had replicated robustly. At the time of necropsy, virus was also isolated from animal R551 by cocultivation of PBMC, resulting in SHIV-E1p2d57. We conclude that after the first two passages, virus adaptation had succeeded, resulting in a highly replication-competent virus that eliminated CD4+ T cells in the first two doubly immunodepleted recipients.
Unexpected rapid decline of CD4+ T cells in P3a and P3b animals.
Since SHIV-E1 replicated to very high levels in immunodepleted RMs, we sought to determine if this passaged virus could replicate in immunocompetent hosts. Both P3a and P3b RMs had rapidly rising vRNA levels, with peaks of >108 vRNA copies/ml on day 14. Surprisingly, we noted a precipitous drop of CD4+ T cells to <200 cells/mm3, a pattern reminiscent of the acute pathogenicity of dual-tropic SHIV89.6P (14, 15). Indeed, in animal R555 (P3a), the CD4+ T cells never recovered. In the other recipient, R561 (P3b), there was a partial recovery of CD4+ T cells. Prompted by these unanticipated data, we examined the coreceptor usage of the virus isolated from passage 2, SHIV-E1p2d57. Indeed, this isolate turned CEMx174-GFP cells green, consistent with CXCR4 coreceptor usage (Fig. 4A). Next, we performed extended coreceptor assays in U87.CD4 and GHOST(3) cells expressing various individual human coreceptors. SHIV-E1p2d57 replicated in cells expressing either CCR5 or CXCR4 coreceptors (Fig. 4B), while the parental clone, SHIV-E1, infected only CCR5-expressing cells (Fig. 4B). Clearly, virus isolated at necropsy from the passage 2 animal had become dual tropic, in contrast to the parental SHIV-E1 infectious molecular clone (IMC), which used only CCR5.
Coreceptor usage of SHIV-E1p2d57 reisolated from animal R551. (Ai and iv) Infectivity assays for SHIV-E1p2d57 isolated from T. cruzi-infected animal R551 on day 57 using CEMx174-GFP cells. (ii and v) X4-tropic HIV-1NL4-3 was used as a positive control for X4 tropism. (iii and vi) No-virus controls. Top panels represent images acquired in bright field; bottom panels represent the corresponding GFP fluorescence. (B) Virus production measured by a p27 or p24 ELISA of culture supernatants harvested on different days after exposure of U87 or GHOST cells expressing different human coreceptors to cell-free virus preparations. (Left) p27 levels of SHIV-E1p2d57; (middle) p27 levels of the SHIV-E1 parental IMC from transfected 293T cells; (right) p24 levels of HIV-196USSN20, known to have extended coreceptor usage.
The search for partially adapted SHIV-E1 with exclusive R5 tropism.
Given the unexpected coreceptor switch, we aimed to isolate R5-only virus from the latest time point possible during serial adaptation. To do this, we chose a two-pronged approach: (i) the generation of infectious molecular clones and (ii) the isolation of infectious, partially adapted virus by cocultivation with uninfected RM PBMC. The strategy to generate infectious env clones is depicted in Fig. 5A. Initially, env regions from the day 14 or day 57 isolate were cloned into the backbone of a proviral HIV clone, NL-LucR.T2A (16), to generate NL-LucR.E1p2d14 or NL-LucR.E1p2d57 IMCs, respectively.
Generation of the SHIV-E1p2 infectious molecular clone. (A) Steps involved in the generation of an exclusively R5 SHIV-E1p2 infectious molecular clone. (B) Virus production measured by a p24 ELISA of culture supernatants harvested on day 7 after exposure of U87.CD4.CCR5 or U87.CD4.CXCR4 cells to cell-free supernatants of 293T cells transfected with various NL-LucR.E1p2 IMCs. HIV-1NL4-3 and HIV-11084i were used as X4- and R5-tropic controls, respectively. (C) Infectivity on TZM-bl cells of serially diluted supernatants of 293T cells transfected with various IMCs of SHIV-E1p2. SHIV-KNH1144p4 (SHIV-A) (our unpublished data) was used as a positive control. (D) Ability of PBMC from 7 random naive rhesus macaque donors to support the replication of the SHIV-E1p2c183 IMC. The cell-free filtered supernatant of transfected 293T cells was assayed. Supernatants of IMC-exposed PBMC were collected on the days indicated.
Next, we examined the coreceptor usage of the different NL-LucR.E1p2 IMCs. Cells expressing either CCR5 or CXCR4 were incubated with virus, and p24 levels were measured from cell supernatants on day 7. All of the NL-LucR.E1p2d14 clones (clone 3 [c3], c7, c18, c20, and c21) replicated in CCR5-expressing cells but not in cells expressing CXCR4 (Fig. 5B). However, three of the four NL-LucR.E1p2d57 IMCs (c9, c23, and c26) replicated in both CCR5- and CXCR4-expressing cells (Fig. 5B). A single clone, c32, grew only in CXCR4-expressing cells (Fig. 5B). These data are consistent with the SHIV-E1p2d57 isolate containing strains that had switched coreceptor usage to become either dual tropic or X4-only tropic (c32). In contrast, NL-LucR.E1p2d14 IMCs were exclusively R5 tropic. Consequently, we used NL-LucR.E1p2d14 IMCs for further development. env regions of the latter were cloned into the backbone of SHIV-1157ipd3N4 (10) to generate SHIV-E1p2 IMCs (Fig. 5A). These proviral plasmids were transfected into 293T cells, followed by infectivity tests of filtered supernatants in TZM-bl cells; all SHIV-E1p2 IMCs were infectious, with SHIV-E1p2c183 showing the highest infectivity (Fig. 5C). SHIV-E1p2c183 replicated in PBMC from three of seven randomly selected naive RM donors (Fig. 5D). Compared to the parental clone, SHIV-E1, which had been unable to replicate in any RM PBMC previously, the partially adapted IMC SHIV-E1p2c183 replicated in 43% of randomly selected RM donor PBMC. These data indicate progress, albeit not yet full success, in adapting SHIV-E1 to the new host species.
P4 and P5 in immunocompetent animals.
Having successfully rescued partially adapted, cell-free virus as well as IMCs, we decided to combine all of these virus forms and use the successful “in vivo transfection” method again. Thus, the P4 animal, R798, received an i.m. inoculation of the IMC SHIV-E1p2c183 plus two i.v. inoculations of cell-free virus stocks of SHIV-E1p1d203 and SHIV-E1p2d14 on day 0 (Fig. 6A and andB).B). Two additional inoculations of the cell-free virus were given on subsequent days (Fig. 6B and andC,C, left). These combined inoculations resulted in high-level peak viremia of ∼108 vRNA copies/ml. Of note, this animal was not immunodepleted and thus had normal peripheral blood lymphocyte counts at the time of virus exposure. Over the course of the infection, the absolute number of CD4+ T cells dropped significantly. On day 28 after inoculation of P4 animal R798, infected blood was transferred to a new naive recipient, animal R974, again without immunodepletion. Viral titers approached 108 vRNA copies/ml, and viremia persisted (Fig. 6C, right). During acute viremia, the number of CD4+ T cells again dropped, reaching levels of <200 cells/mm3, indicating the pathogenicity of the passaged virus. Cell-free virus was reisolated from animal R974 on day 42 postinoculation and termed SHIV-E1p5 (Fig. 6A and andC).C). On day 332, this animal was euthanized due to reaching the study endpoint.
Serial passages P4 and P5 followed by isolation of SHIV-E1p5. (A) Animal R798 was inoculated with cell-free SHIV-E1p1d203 and SHIV-E1p2d14 i.v. and with SHIV-E1p2c183 IMC DNA by the i.m. route. Blood was transferred to animal R974 on day 28 after inoculation of animal R798. SHIV-E1p5 was isolated on day 42 after exposure of animal R974. Red boxes indicate R5-tropic virus. For P1 through P3a and P3b, see the legend to Fig. 2A. (B) Table summarizing the inocula, their sources, and routes of transfer during passages 4 and 5 of SHIV-E1. (C) Virus replication kinetics in animals R798 and R974. Red down arrow with asterisk indicates inoculation of RM R798 with proviral SHIV-E1p2c183 IMC DNA i.m. at two sites on day 0. Red down arrows (without asterisks) indicate inoculation of cell-free virus on three different days. Red up arrow for passage 5 for animal R974 indicates the reisolation of SHIV-E1p5 on day 42 after blood transfer, circles with P4 and P5 indicate the passage numbers, red boxes indicate R5-tropic virus, red lines with red circles indicate viral RNA copies per milliliter on a log scale, and axes on the right indicate absolute cell numbers (103).
SHIV-E1p5, an R5, mucosally transmissible, and pathogenic strain.
Next, we sought to ensure that the final SHIV-E1p5 isolate had retained its R5 tropism. We examined the coreceptor tropism of SHIV-E1p5 in CEMx174-GFP cells; no green fluorescence was observed, consistent with the lack of X4 coreceptor use (Fig. 7A). To confirm this finding, we also performed extended coreceptor assays. SHIV-E1p5 was able to infect cells expressing only the CCR5 coreceptor, indicating exclusive R5 use; SHIV-E1p5 specifically did not replicate in GPR15/BOB- and CXCR6/BONZO/STRL33-expressing cells (Fig. 7B).
SHIV-E1p5 is exclusively R5 tropic. (A) CEMx174-GFP cell infectivity assay with SHIV-E1p5 (i and iv), the X4-tropic positive control HIV-1NL4-3 (ii and v), and R5-tropic SHIV-1157ipd3N4 (iii and vi). (Top) Images acquired in bright field; (bottom) corresponding GFP fluorescence. (B) Infectivity of SHIV-E1p5 in U87.CD4 and GHOST.CD4 cells expressing various coreceptors. HIV-196USSN20 was used as a control.
To determine whether SHIV-E1p5 was well adapted to RM hosts, we examined its replication efficiency in vitro and in vivo. PBMC of six randomly selected naive RM donors were incubated with SHIV-E1 isolated from passage 1, passage 2, and passage 5. SHIV-E1p5 was able to replicate in all six RM donor PBMC, while SHIV-E1p1d203 and SHIV-E1p2d14 replicated only in two of six and four of six animals, respectively (Fig. 8A). An additional 10 random naive donors were chosen for additional screening of SHIV-E1p5 replication in RM PBMC. Eight out of 10 RM PBMC cultures supported SHIV-E1p5 replication (Fig. 8B). These results indicate that SHIV-E1p5 is better adapted for replication in RM PBMC than virus isolated from earlier passages. Finally, to assess mucosal transmissibility, a RM was challenged intrarectally (i.r.) with a single, high dose of SHIV-E1p5 (Fig. 8C) (1.4 × 105 50% tissue culture infective doses [TCID50]). This resulted in viremia of >108 vRNA copies/ml at week 2 postexposure, followed by persistent viremia (Fig. 8C). The number of CD4+ T cells dropped to <200 cells/mm3 by day 84 postinoculation, indicating pathogenicity (Fig. 8C). Clearly, SHIV-E1p5 is a mucosally transmissible, R5 clade E SHIV that depletes CD4+ T cells to low levels, consistent with AIDS.
SHIV-E1p5 replication in vitro and in vivo. (A) Ability of PBMC from six randomly selected naive RM donors to support the replication of SHIV-E1 from passage 1 (SHIV-E1p1d203), passage 2 (SHIV-E1p2d14), and passage 5 (SHIV-E1p5). (B) SHIV-E1p5 replication in PBMC of 10 random naive RM donors. (C) SHIV-E1p5 replication kinetics in a rhesus macaque challenged intrarectally (red down arrow). Plasma vRNA copies per milliliter were assessed by using nucleic acid sequence-based amplification (NASBA) technology. Absolute numbers of blood cells were monitored by FACS analysis. At time zero, the absolute CD4+ T-cell number for the intrarectally challenged animal was 441 cells/mm3, which is within normal limits for adult Indian-origin RMs, according to Autissier et al. (69).
SHIV-E1p5 has a tier 2 neutralization phenotype.
Next, we assessed the susceptibility of SHIV-E1p5 to a panel of anti-HIV mAbs as well as polyclonal serum or plasma samples in parallel with reference viruses, including tier 1 clade C SHIV-1157ipEL-p (17), its tier 2 counterpart SHIV-1157ipd3N4 (10), and tier 1A and tier 2 HIV CRF01_AE strains TH023.6 (18) and CNE55 (19) (Table 1). The polyclonal sera were standardized pools used to determine the neutralization sensitivity of primary HIV isolates. SHIV-E1p5 was not sensitive to neutralization by various Center for HIV/AIDS Vaccine Immunology (CHAVI) serum pools isolated from South African HIV-infected individuals (clade not determined). However, five of seven clade-matched plasma samples from HIV AE-infected individuals neutralized SHIV-E1p5 (Table 1, top red box), similar to that seen with CNE55. SHIV-E1p5 showed some neutralization by polyclonal anti-HIV clade C Abs in the HIVIG-C (HIV-specific IgG from HIV clade C-infected individuals) pool and soluble CD4 (sCD4) (Table 1, middle red box). While SHIV-E1p5 was relatively difficult to neutralize with known broadly neutralizing mAbs (nmAbs), it was sensitive to nmAbs targeting the membrane-proximal external region (MPER), and mAb LN01, a CD4 binding site-specific nmAb, neutralized SHIV-E1p5 (Table 1, middle and bottom red boxes). Overall, this neutralization profile is consistent with a tier 2 neutralization phenotype.
TABLE 1
Sensitivity of SHIV-E1p5 to broadly neutralizing mAbs and polyclonal antibodiesj
aResults are from SHIV-E1p5 prepared in human PBMC, consistent with the standard protocol that uses human serum/plasma and HIV isolates as reference strains.
bValues are the dilution for serum/plasma samples at which relative luminescence units (RLU) were reduced 50% (ID50).
cValues are the concentration for nmAbs or sCD4 at which RLU were reduced 50% (ID50).
dTier 2 clade C control.
eTier 1 clade C control.
fTier 1A CRF01_AE strain.
gTier 2 CRF01_AE strain.
hSerum samples from South African HIV+ individuals during chronic infection (infecting clade not determined). “Pool” refers to combining different bleed dates from the same individual.
iGeometric mean titer used to determine the neutralization tier compared to standard tier 1A, tier 1B, and tier 2 reference viruses. Tier cutoffs for AE plasma dilutions were >1,000 for tier 1A, 250 to 1,000 for tier 1B, 51 to 250 for tier 2, and ≤50 for tier 3.
env mutations arising during serial passage of SHIV-E1.
To examine the molecular evolution of the HIV-E envelope during SHIV-E1 adaptation, RNA isolated from the SHIV-E1p1d203, SHIV-E1p2d14, and SHIV-E1p5 isolates was analyzed by deep sequencing; nucleic acid mutations were used to predict amino acid changes (Fig. 9). Predicted amino acid changes were compared to the envelope sequence of the parental clone, SHIV-E1. Of note, the entire evolution of SHIV-E1p2d14 occurred in the absence of selective pressure from host CD8+ cells and B cells due to double immunodepletion with cytotoxic mAbs. By the time when SHIV-E1p2d14 was isolated, the virus had replicated for a total of 67 days in the first two animals and represents the first intermediate isolate time-wise. Not only did the immunodepletion preclude the generation of anti-HIV Env Ab responses, but it also prevented adaptive cell-mediated immunity as well as innate immunity mediated by CD8+ NK cells. Therefore, mutations that arose in this early SHIV-E1 isolate (Fig. 9, second panel) can be ascribed to the selection of viral quasispecies with improved replicative fitness in the new host environment, rather than antiviral host responses.
Predicted Env amino acid (AA) changes during the adaptation of SHIV-E1p1d203, SHIV-E1p2d14, the “dead-end” isolate SHIV-E1p2d57, and SHIV-E1p5 by deep sequencing using Illumina technology. The four panels show predicted amino acid changes of the three R5 isolates (black boxes) and dual-tropic SHIV-E1p2d57 (navy blue box) compared to the parental infectious molecular clone, SHIV-E1. The height of each bar represents the percentage of sequence reads containing a given mutation. Sequences with a prevalence of ≥5% are shown. Salmon-colored bars indicate predicted amino acid changes found during adaptation but not associated with a coreceptor switch, and navy blue bars indicate mutations found exclusively in the dual-tropic SHIV-E1p2d57 isolate. CD4i, amino acid mutation G426R, known to be associated with a CD4i epitope and in the region of the chemokine receptor binding site (23, 24). Areas shaded in dark gray indicate Env domains that represent HIV clade E sequences (not drawn to scale), and areas shaded in light gray indicate gp41 regions derived from the SHIV backbone used for cloning, SHIV-1157ipd3N4; these sequences were derived from either HIV clade B or the SIVmac239 Env/Nef overlap (10).
A number of mutations became fixed quickly, as indicated by the 100% prevalence in all sequence reads (Fig. 9). One of the gp41 mutations, A790V, was found in all reads of the SHIV-E1p1d203 isolate. Interestingly, the prevalence of A790V in the passage 2 day 14 isolate, SHVI-E1p2d14, was only 65%. It should be kept in mind that virus isolated from passage 1 (SHIV-E1p1d203) had replicated for a total of 203 days in the first recipient animal. As mentioned above, SHIV-E1p2d14 had replicated for a total of 67 days in RMs, i.e., 53 days in the first animal until blood transfer and an additional 14 days in the second recipient. Consequently, more time had elapsed overall for SHIV-E1p1d203 (Fig. 9, top panel) than for SHIV-E1p2d14 (Fig. 9, second panel). In the final isolate, SHIV-E1p5, A790V was found at a prevalence of 100%. This mutation is located in the intracellular portion of gp41 that is not part of the HIV clade E insert. The same temporal pattern was also seen for I853T, located in the same gp41 region.
The pattern of predicted Env amino acid changes compared to the parental SHIV-E1 sequence was noticeably different in the last isolate, SHIV-E1p5. This is the only one of the four isolates shown in Fig. 9 where adaptive immune responses exerted their influence on Env evolution during serial passage. Interestingly, a number of low-level mutations appeared, especially in the ectodomain of gp41 (Fig. 9, bottom). Of note, SHIV-E1p5 is sensitive to neutralization by anti-MPER human nmAbs (Table 1, bottom red box). It is possible that the low-frequency mutations seen in gp41 represent early neutralization escape variants. Likewise, cell-mediated immunity could have participated in the selection of certain mutations. In this context, it is noteworthy that plasma of the passage 4 animal, R798, from the day of blood transfer to the last animal, R974, did not recognize HIV Env bands by Western blotting. Since SHIV-E1p5 was isolated from animal R974 on day 42, we sought to determine anti-Env Western blot reactivity on this day. Only a weak band was noted for gp160 but not for the other Env bands, indicating the beginning of host anti-Env antibody responses (data not shown).
Predicted amino acid changes in Env found exclusively in X4/dual-tropic SHIV-E1p2d57 quasispecies.
Next, we analyzed the predicted Env amino acid changes of the dual-tropic SHIV-E1p2d57 isolate by deep sequencing (Fig. 9, third panel) and compared these changes with those of the other three SHIV-E1 progeny. In parallel, we performed Sanger DNA sequencing of the V3 loop regions of various IMCs carrying clade E env (Fig. 10) since mutations in the V3 loop are known to be associated with a coreceptor switch; tropism was predicted by the PhenoSeq platform (20,–22) and compared to the experimental data (Fig. 4, ,5B,5B, ,7,7, and and10).10). Since HIV Env coreceptor switching has been reported for Thai individuals harboring HIV CRF01_AE strains (1, 12), we included Env sequences from such cases in our alignment (reference [Ref] clones) (Fig. 10).
Predicted amino acid changes in the V3 loop of HIV clade E gp120 associated with coreceptor switching. (A) Alignment of V3 loop amino acid sequences of either infectious molecular clones (IMCs) or biological SHIV isolates (designated virus1) with the consensus CRF01_AE sequence (46) and the parental SHIV-E1 IMC. env clones had been inserted into NL-LucR vectors and were shown to be infectious; env clones had been isolated on day 14 or 57 from passage 2. The sequence of SHIV-E1p2d14c183, a clone that was used for subsequent readaptation, is also listed. Sequences starting with EU and GU were derived from data reported previously and are designated reference (Ref) clones (46, 70). A red box indicates V3 loop sequences associated with R5 tropism, and a blue box indicates sequences associated with dual tropism or X4 usage. The tropism was predicted by the PhenoSeq platform (22). (B) SHIV-E1 phylogenetic tree. Env amino acid sequences of the SHIV-E1 parental IMC and SHIV-E1 isolates were aligned and analyzed for evolutionary relatedness in Mega7. The branch lengths reflect the number of substitutions per site. Red boxes indicate R5 viruses, and blue boxes indicate R5X4/X4 viruses.
The V3 loop amino acid sequence predicted to be associated with R5 tropism, TRT at positions 304 to 306, was identical in all IMCs and viral isolates that were exclusively R5 tropic by coreceptor usage assays (Fig. 10A, red box). In contrast, all IMCs encoding env from passage 2 at day 57 had the sequence IRI at these positions, while the SHIV-E1p2d57 isolate had this same sequence at a prevalence of ∼80% (Fig. 9 and and10).10). These IMCs and this virus strain were predicted to have X4-tropic phenotypes by PhenoSeq. However, with the exception of NL-LucR.E1p2d57c32, they were all found to be dual tropic in coreceptor usage assays. Notably, a third mutation seen in V3 of SHIV-E1p2d57 at an ∼20% prevalence, I323M, was present in X4-tropic NL-LucR.E1p2d57c32. Essentially, data from the deep-sequencing analysis shown in Fig. 9 were confirmed by conventional sequencing of infectious env clones in Fig. 10A. While we were initially concerned that deep sequencing may include sequence information from nonfunctional, noninfectious env sequences, both data sets yielded similar results.
We observed additional amino acid changes in regions outside V3 in SHIV-E1p2d57 Env compared to the parental IMC and all other R5 isolates (Fig. 9, third panel). These changes include mutations in the V2 loop, the C3 and C4 regions, and gp41. Interestingly, a mutation was found at amino acid position 426, which lies within the CD4-induced (CD4i) epitope, a region known for CCR5 binding (23, 24). This G426R mutation was present at a low frequency in the earlier SHIV-E1p1d203 and SHIV-E1p2d57 isolates but was highly prevalent in the dual-tropic isolate. The final R5 SHIV-E1p5 isolate did not contain this mutation.
A striking difference in amino acid mutations at the gp41 C terminus was noted. Three mutations became fixed in SHIV-E1p1d203 and SHIV-E1p5 at amino acid positions 790, 849, and 853 but were completely absent in the dual-tropic isolate SHIV-E1p2d57, implying that the latter arose directly from an archival clone that had not yet mutated at these three gp41 positions. The remarkable absence of amino acid variability in the C terminus of gp41 in SHIV-E1p2d57 prompted us to perform a phylogenetic tree analysis (Fig. 10B). Indeed, this analysis confirmed that the dual-tropic env isolate arose directly from parental SHIV-E1 rather than from intermediary R5-tropic progeny, such as the isolate taken from the same RM at an earlier time point, SHIV-E1p2d14. These data indicate that the original E1 envelope has the ability to undergo a coreceptor switch. Interestingly, SHIV-E1p1d203 was more closely related to SHIV-E1p5 than isolates from the second passage. This could be due to the fact that SHIV-E1p1d203 replicated for a significantly longer period of time in the first animal (203 days) than did SHIV-E1p2d14 in the first and second RMs combined (67 days), as mentioned above. Another possibility is that the recovery of CD8+ T cells may have caused mutations similar to those observed in SHIV-E1p5. We had stopped depleting CD8+ and B cells in animal R547 on day 100, and by day 203, recovery of CD8+ cells could have exerted selective pressure. In contrast, neutralizing antibody responses are unlikely to arise within such a short time. To summarize, the change in tropism involved not only the V3 loop but also changes in other parts of the envelope.
DISCUSSION
Here we report the successful construction, adaptation, and characterization of SHIV-E that carries an HIV CRF01_AE envelope isolated from a volunteer in the RV144 trial who had been given placebo. Our newly isolated virus, SHIV-E1p5, recapitulates key biological aspects of HIV CRF01_AE transmission in humans. SHIV-E1p5, an exclusively R5-tropic virus with a tier 2 neutralization phenotype, was mucosally transmissible, replicated robustly in RMs, and caused a severe depletion of CD4+ T cells, leading to AIDS. During the initial adaptation phase, the newly created SHIV also underwent a coreceptor switch to dual or X4 tropism. Deep sequencing revealed that dual/X4-tropic virus branched off very early during its evolution, starting from the parental SHIV-E1 construct.
We had to overcome severe host restriction: the initial parental SHIV-E1 construct replicated well in human cells but failed to grow in any rhesus PBMC cultures from multiple donors. Other investigators also reported that newly constructed SHIVs replicate inefficiently in rhesus PBMC (25, 26). We devised a novel strategy: in vivo transfection by i.m. inoculation of infectious proviral DNA and repeated administration of cell-free virus under ablation of both arms of adaptive immunity. This was achieved by the depletion of CD8+ T cells and NK cells by an anti-CD8 mAb and of B cells by an anti-CD20 mAb, respectively. This combined approach resulted in high levels of sustained viremia that led to severe losses of CD4+ T cells. Of note, all virus passages were designed to circumvent the pressure of neutralizing antibodies to prevent the selection of neutralization escape viral variants. An additional benefit of B-cell depletion was the prevention of RM antibody responses to the anti-CD8 mAb, which allowed us to give this mAb repeatedly, thereby extending the time window of CD8 depletion to several weeks.
Intramuscular inoculation of SIV proviral plasmid DNA has been used previously (27, 28) and resulted in sustained viremia and disease progression. Subsequently, i.m. DNA inoculation was used to vaccinate animals with nef-deleted, live attenuated SIV (29, 30). To our knowledge, the use of in vivo i.m. transfection to jump-start SHIV replication in a new host species, where none of the PBMC cultures have previously supported virus replication, has not been attempted previously. Virions budding from transfected RM muscle cells acquire host cell membranes into which HIV clade E Env spikes are embedded together with other host molecules. Possibly, such particles are more likely to enter RM target cells to increase the intensity of virus replication in subsequent cycles. Given that our goal was to expeditiously generate a RM replication-competent, pathogenic SHIV-E strain for use in preclinical vaccine trials, the individual contributions of i.m. inoculation with infectious proviral DNA and repeated i.v. inoculations of cell-free virus to the overall success in achieving high replication levels in the first animal are unknown because they were carried out simultaneously. We can say with certainty, however, that removing host defenses allowed unbridled viral replication. Overall, our combined strategy of proviral DNA inoculation, repeated inoculations with virus produced in transfected 293T cells, and sustained ablation of host adaptive immune responses led to the rapid adaptation of the new virus.
In retrospect, the second recipient, animal R551, initially had clinically silent T. cruzi infection. Due to the double immunodepletion with anti-CD8 and anti-CD20 mAbs and the resulting high levels of SHIV viremia that depleted CD4+ T cells, RM R551 had severe immunodeficiency, which led to parasitemia detectable by microscopy. This came as a surprise; after all, the primate study was conducted in Maryland. It turned out that this animal came from a breeding facility in southern Texas, where the arthropod vectors for T. cruzi, Triatoma species or “kissing bugs,” are widespread. A recent survey across Texas revealed that ∼50% of these vectors, including those found near human dwellings, harbor T. cruzi (31). When a baboon colony housed outdoors at the Southwest National Primate Research Center was examined for T. cruzi, 57% of the animals tested were seropositive, and among these, 24% had echocardiographic evidence of Chagas' heart disease (32, 33). In the context of HIV infection, T. cruzi is considered to be an opportunistic infection, and there is evidence that the two pathogens mutually upregulate each other (13, 34,–36). A recent epidemiological survey in Brazil found a 5% coinfection prevalence in HIV-positive individuals, 3.8 times higher than previously estimated (37).
We previously showed that preexisting parasite infection in RMs increased the risk of R5 SHIV acquisition as well as peak viremia (38). Similarly, parasite inoculation of RMs with chronic SHIV infection exacerbated viremia (39). However, to our knowledge, an association of parasite coinfection and HIV Env coreceptor switching has not been reported. To date, the exact mechanisms driving HIV Env coreceptor switching remain to be determined.
Cheng-Mayer et al. previously reported HIV Env coreceptor switching in the context of SHIV infection in RMs (40,–43). Their work involved SHIVSF162P3N, a late-stage, R5 clade B SHIV. An X4 strain, SHIVBR24N, was recovered at necropsy, at 28 weeks postinoculation, from the SHIVSF162P3N-infected RM (41). When comparing the evolutionary changes during coreceptor switching in the clade B SHIV to our case, common mutations were noted, as follows: T to I in V3 (T304I in SHIV-E1p2d57) and two mutations in C4 (next to G426R in SHIV-E1p2d57). The C4 changes are within a conserved region previously shown to be involved in binding to HIV coreceptors (44, 45). The shared T-to-I change in V3 is predicted to cause the loss of a potential N-linked glycosylation site, which has been linked to X4 usage in the case of HIV CRF01_AE Envs (46). Other amino acid changes in SHIVBR24N Env were not shared in our case. Phylogenetic tree analysis showed that X4 SHIVBR24N variants branched off late in the disease course from one of two R5 SHIVSF162P3N clusters. In contrast, the coreceptor switch that we observed arose independently from the parental IMC construct as a separate branch.
Env sequences and the evolution of viral variants from an earlier time point, week 20 after inoculation of the SHIVSF162P3N-infected RM, were also determined; such viruses were R5X4 intermediates between the initial R5 SHIV and the final X4 variant (43). In our case, we observed the emergence of a minor fraction of X4-only virus at necropsy; most SHIV-E1p2d57 variants were dual tropic.
The Cheng-Mayer group reported a second case of coreceptor switching in a SHIVSF162P3N-infected, CD8-depleted RM (42). Replacing only the V3 loop in the R5 parental strain with that of an X4-only isolate revealed that V3 mutations were necessary for the coreceptor switch, but additional changes present in V1/V2, C4, and gp41 in the X4 isolate yielded optimal virus entry into indicator cell lines. V3 loop mutations in this second case of SHIVSF162P3N coreceptor switching differed from those in the first case, indicating convergent evolution. The Cheng-Mayer group then documented molecular evolution resulting in the coreceptor switch of a clade B SHIV; our data expand these findings to a clade E SHIV. Clearly, HIV envelopes of different clades have the capacity to alter coreceptor use in the context of SHIV infection in RMs, akin to different HIV clades in humans.
Studying molecular envelope evolution by deep sequencing revealed some striking patterns. Compared to the parental construct SHIV-E1, all three R5-only isolates contained 12 amino acid changes in Env that had become fixed with a prevalence of 90 to 100%, indicating a selective advantage of R5 progeny viruses (Fig. 9). Since they were present in the early isolate, SHIV-E1p2d14, host immune responses played no part in the selection of these mutations. Of note, only 3 of these 12 amino acid changes were found at such a high prevalence in the dual-tropic isolate, SHIV-E1p2d57, thus confirming early divergence during evolution between the R5- and dual-tropic progeny. Overall, the predicted amino acid sequence conservation in the R5-only strains ranges between 98 and 99% compared to the parental envelope.
The last viral isolate, SHIV-E1p5, is the only one that had been exposed to the selective pressures of innate and adaptive host immunity, predominantly cell-mediated immune responses. The virus replicated in animal R798 (passage 4) for 28 days until blood transfer to the last recipient, R974, the source of SHIV-E1p5 isolated on day 42. Neither of these last two recipients were immunodepleted. Thus, within 28 or 42 days, cell-mediated immune responses may have selected cytotoxic-T-lymphocyte (CTL) escape mutants in SHIV-E1p5 (47). In contrast, neutralizing antibodies typically take months to develop, especially against viruses with a tier 2 neutralization phenotype. Of note, anti-V3 loop neutralizing antibodies develop early. The lack of new mutations in the SHIV-E1p5 V3 loop is an indirect sign that neutralizing antibodies had not yet developed within the short replication times allowed in the fourth and fifth recipients.
Recently, the generation of SHIV challenge stocks carrying HIV CRF01_AE env (26) was described. Our stock, SHIV-E1p5, differs in important ways. (i) env carried by SHIV-E1p5 was derived from the RV144 placebo group. As such, our SHIV provides a link to this phase III trial with a contemporaneous env. The env genes used by Tartaglia et al. were from RV217 and R254, both focused on early HIV transmission. (ii) Tartaglia et al. monitored SHIV-infected RMs maximally for only 10 to 20 weeks, during which time no significant pathogenicity was observed. (iii) Two of the three challenge stocks described by Tartaglia et al. were prepared in human PBMC. This raises important issues for preclinical vaccine efficacy trials. Repeated low-dose mucosal challenges of RMs with human PBMC-grown viruses can generate xenoresponses, which differ from alloresponses during human-to-human transmission of HIV. If candidate vaccines are also produced in human cells, human cell-grown challenge SHIV or SIV stocks are contraindicated for efficacy studies in macaques, based upon data reported previously (48,–50). (iv) The LTR in our SHIV-E1p5 is more akin to HIV LTRs that have extra NF-κB sites. In contrast, the SIVmac239 LTR has only one such site. Higher numbers of NF-κB sites are known to be more responsive to cytokine stimulation for activation signals using the NF-κB signal transduction pathway (10, 51). This may be important for studies aimed at studying virus reservoirs and eradication.
The final virus stock, SHIV-E1p5, is a biological isolate that can be used as a tool to assess prevention strategies targeted against CRF01_AE in NHPs. Preclinical vaccine efficacy studies frequently use repeated low-dose mucosal challenges. While we have been able to give a proof of concept that SHIV-E1p5 is mucosally transmissible, the limitations of our studies allowed us to enroll only one animal to test the i.r. route using a high-dose inoculum. Further studies will be required to assess repeated low-dose challenges, including other mucosal routes. The use of biological SHIV isolates in place of IMCs more accurately reflects HIV transmission among humans, where multiple quasispecies are present at each exposure. However, regarding the SHIV/RM model, one of the limitations of using a biological SHIV isolate for challenges is that virus features, such as neutralization phenotype and pathogenicity, may not be properties of all viral variants in the inoculum. As a result, the clinical outcomes in individual RM hosts with breakthrough infection may vary in situations involving low-dose challenges with the transmission of only one or a few variants. However, challenges with a biological isolate also provide the experimental advantage of indicating how many quasispecies are transmitted when testing vaccine efficacy.
In summary, SHIV-E1p5, a RM PBMC-grown strain, reflects key biological aspects of HIV transmission and disease progression in humans. As such, this virus represents a tool to evaluate anti-HIV CRF01_AE prevention strategies by passive and/or active immunization of rhesus macaques.
MATERIALS AND METHODS
Construction of infectious proviruses and generation of virus stocks.
Infectious HIV CRF01_AE env clones derived from placebo group participants in the RV144 phase III vaccine trial in Thailand (1) were used to generate several proviral SHIV-E clones. One of the env clones, 620345.2, was engineered to incorporate KpnI and BamHI restriction sites into the env sequence by PCR using primers Fwd-620345-Kpn (ST-A101; AGTTTATTATGGGGTACCTGTGTGGAGAGATGCAGA) and Rev-620345-Bam (ST-A102; CAGGCAAGTGCTAAGGATCCGCTCACTAATCG). The addition of these sites enabled the cloning of the env sequence into the SHIV-1157ipd3N4 backbone (GenBank accession no. DQ779174) (10). Clones with the correct env insert were identified by restriction enzyme digestion and confirmed by sequencing (Genewiz Inc.). The final construct, termed SHIV-620345.2, or SHIV-E1 for short, was transfected into 293T cells by using FuGENE6 transfection reagent (Promega), and supernatants were filtered to produce the SHIV-E1 stock.
To generate the NL-Luc.R version of passage 2 virus, termed SHIV-E1p2, the envelope region of the latter was cloned into the proviral DNA backbone of pNL-LucR.T2A (16). The backbone region of pBR322 (New England BioLabs) and the env region of pNL-LucR.T2A were excised by using the EcoRI/BamHI restriction sites and ligated to produce the shuttle vector pBR322(NL4-3). The env region of SHIV-E1p2 was excised by using the KpnI/BamHI restriction sites and cloned into the backbone of the shuttle vector pBR322(NL4-3). Finally, using the EcoRI/BamHI restriction sites, the env region of pBR322-NL-E1 was swapped into the backbone of pNL-LucR.T2A to obtain NL-LucR.E1. The final construct was confirmed by PCR and DNA digestion.
To generate an infectious molecular clone of SHIV-E1p2, the env regions of NL-LucR.E1 and the 3′ half of the infectious provirus clone SHIV-1157ipd3N4, termed 3′ SHIV-1157ipd3N4, were swapped by digestion with KpnI and BamHI followed by ligation to produce 3′ SHIV-E1; the resulting plasmid was confirmed by PCR. Finally, env of 3′ SHIV-E1 and the backbone of 5′ SHIV-vpu+ were digested with SphI and NotI and ligated to produce the final SHIV-E1p2 IMC, termed SHIV-E1p2c183. The final construct was confirmed by PCR and restriction enzyme digestion, and the env sequence was confirmed by DNA sequencing. SHIV-E1p2c183 was transfected into 293T cells by using X-tremeGENE 9 DNA transfection reagent (Sigma) to produce the SHIV-E1p2c183 stock.
Virus controls.
Several virus controls were used for various in vitro assays. SHIV-1157ipd3N4 (clade C) (10) and SHIV-KNH1144p4 (clade A) (our unpublished data) were developed in our laboratory. HIV-1NL4-3 and HIV-196USSN20 (52) were obtained from the NIH AIDS Reagent Program. HIV1084i is an IMC of a recently transmitted clade C virus isolated from a Zambian infant (53).
Cells and cell culture.
CEMx174-GFP cells (gift of Barbara Felber, National Cancer Institute, Frederick, MD) contain the green fluorescent protein (GFP) gene under the regulation of the HIV LTR and express CD4 and CXCR4. CEMx174-GFP cells were grown in RPMI 1640 supplemented with 25 mM HEPES, 2 mM l-glutamine, 100 U/ml penicillin-streptomycin (Gibco), and 10% fetal bovine serum (FBS) (SAFC Bioscience). U87 and GHOST(3) cells express either only human CD4 or CD4 along with various HIV/SIV human chemokine coreceptors. U87 cells were grown in Dulbecco's modified Eagle's medium (DMEM) (Gibco) supplemented with 15% FBS, 2 mM l-glutamine, 1 μg/ml puromycin (Sigma-Aldrich), 300 μg/ml Geneticin (G418 sulfate; Gibco), and 100 U/ml penicillin-streptomycin. GHOST(3) cells were grown in DMEM supplemented with 10% FBS, 500 μg/ml Geneticin (G418 sulfate), 100 μg/ml hygromycin (Gibco), 100 U/ml penicillin-streptomycin, and 1 μg/ml puromycin. TZM-bl cells (also designated JC53-bl [clone 13] cells) are derived from a HeLa cell line (JC.53) that stably expresses CD4 and CCR5; these cells also contain integrated reporter genes that express firefly luciferase or β-galactosidase under the control of the HIV LTR (54,–58). TZM-bl cells were grown in DMEM supplemented with 10% FBS, 25 mM HEPES, and 50 μg/ml gentamicin (Gibco). 293T cells were grown in DMEM supplemented with 10% FBS, 2 mM l-glutamine, and 100 U/ml penicillin-streptomycin. All Gibco cell culture reagents were purchased from Thermo Fisher Scientific. U87, GHOST, TZM-bl, and 293T cells were obtained through the NIH AIDS Reagent Program, Division of AIDS, NIAID, NIH. RM blood for PBMC isolation and virus stock generation was obtained from the Yerkes National Primate Research Center Comparative AIDS Core (Atlanta, GA) and the Southwest National Primate Research Center (San Antonio, TX). Human PBMC were isolated from heparinized blood of three anonymous healthy donors.
Antigen capture assays for p24 and p27.
Levels of p27 (SHIV, which carries SIV gag) and p24 (HIV) were determined by using the respective antigen capture enzyme-linked immunosorbent assay (ELISA) kits (ABL Inc.). Briefly, virus samples were mixed with disruption buffer to inactivate the virus and release p27/p24. Standards and inactivated virus samples were added to 96-well microELISA plates coated with murine mAbs against p27/p24; the assays were performed according to the manufacturer's instructions.
TCID50 determination.
Virus samples were added to TZM-bl cells at serial 5-fold dilutions in medium prepared with DEAE-dextran (MP Biomedicals) at a final concentration of 15 μg/ml. The level of luciferase activity expressed in infected cells was measured at 48 h postinfection after adding the Bright-Glo luciferase substrate (Promega); plates were read in a Mithras LB 940 microplate reader (Berthold Technologies). The TCID50 was calculated as the dilution point at which 50% of the cell cultures were infected, using a Microsoft Excel macro.
Neutralization sensitivity and tier phenotyping.
The neutralization sensitivity of SHIV-E1p5 was tested against a panel of nmAbs and neutralizing serum or plasma pools on TZM-bl cells, as described previously (59). Briefly, neutralizing antibody activity was measured in 96-well culture plates by using Tat-regulated luciferase (Luc) reporter gene expression to quantify reductions in virus infection in TZM-bl cells. Assays were performed with replication-competent SHIV stocks grown in human or RM PBMC. Serum/plasma samples for tier phenotyping were diluted over a range of 1:20 to 1:43,740 in cell culture medium, and purified polyclonal/monoclonal antibodies and sCD4 were assayed in concentrations ranging from 25 to 0.01 μg/ml, also diluted in cell culture medium. Test reagents were preincubated with virus (∼150,000 relative light unit equivalents) for 1 h at 37°C before the addition of cells. Following 48 h of incubation, cells were lysed, and Luc activity was determined by using a microtiter plate luminometer and BriteLite Plus reagent (PerkinElmer). Neutralization titers are the sample dilutions (for serum/plasma) or antibody concentrations (for sCD4, purified IgG preparations, and mAbs) at which relative luminescence units (RLU) were reduced by 50% compared to RLU in virus control wells after subtraction of background RLU in cell control wells. Serum/plasma samples were heat inactivated at 56°C for 1 h prior to assays. Neutralization tier phenotyping was conducted by measuring neutralization titers (50% infective dose [ID50]values) as described above, using five to seven serum/plasma samples from HIV-positive (HIV+) individuals during chronic infection. The geometric mean titer (GMT) was calculated in Microsoft Excel, and the tier phenotype was determined by comparing these values to the GMTs of standard panels of viruses representing tier 1A, tier 1B, and tier 2 viruses (60, 61), using the same five or seven HIV+ serum/plasma samples.
Coreceptor usage assays.
CEMx174-GFP cells were exposed to various passages of SHIV-E1 for 48 to 72 h, and GFP expression was determined by using confocal microscopy (Nikon). HIV-1NL4-3 and SHIV-1157ipd3N4 were used as X4- and R5-tropic controls, respectively. The U87 and GHOST(3) cell lines expressing human CD4 and HIV/SIV human coreceptors were used to study virus tropism. U87.CD4.CCR1, U87.CD4.CCR2, U87.CD4.CCR3, U87.CD4.CXCR4, U87.CD4.CCR5, GHOST(3) BOB/GPR15, and GHOST(3) Bonzo/STRL33 cells were plated at 1 × 106 total cells per well, followed by the addition of predetermined amounts of virus p24/p27 in 200 μl to each well. Cells were washed in 1 ml of fresh medium the next day. Supernatants were collected starting on day 3 and then every other day for 2 weeks and analyzed by a p24/p27 ELISA. HIV-1NL4-3, HIV1084i, and HIV-196USSN20 were used as X4, R5, and extended coreceptor usage controls, respectively.
Animals and animal care.
Seven Indian-origin RMs (Macaca mulatta) were used in this study; they were kept according to NIH guidelines on the care and use of laboratory animals at the NIH Simian Vaccine Evaluation Unit (SVEU) of Bioqual Inc., a USDA-registered facility accredited by the American Association for the Accreditation of Laboratory Animal Care International (AAALAC). Bioqual Inc. complies with all policies of the Guide for the Care and Use of Laboratory Animals (62); DHHS (NIH 85-23); Animal Welfare (DHHS-TN 73-2); NIH Manual Issuances 4206 and 6000-3-4-58; “Responsibility for Care and Use of Animals,” CDC/NIH, 4th ed.; “Biosafety in Microbiological and Biomedical Laboratories”; and Public Health Service Policy on Humane Care and Use of Laboratory Animals (63). The Animal Care and Use Committee at Bioqual Inc. approved all experiments and met all applicable federal and institutional standards. Animals were euthanized first by anesthetizing them with 0.6 ml ketamine and then by i.v. barbiturate overdose of 2.0 pentobarbital, at 1 ml/10 lb, which is consistent with the recommendations of the Panel on Euthanasia of the American Veterinary Medical Association.
Immunodepletion of B cells and CD8+ cells.
RMs R547 and R551 were immunodepleted of B cells and CD8+ cells by infusion of mAbs against CD20 and CD8, respectively. Anti-CD8 mAb (MT807R1; cM-T807 mAb with rhesus IgG1 constant and variable framework regions), obtained from the NIH Non-Human Primate Reagent Resource, was administered (50 mg/kg of body weight) at various intervals. B cells were depleted by using the anti-CD20 humanized mAb rituximab (Genentech Inc.), which was obtained through the Dana-Farber Cancer Institute Pharmacy and administered at a dose of 20 mg/kg of body weight. Animals receiving these mAbs were pretreated with diphenhydramine (1 mg/kg i.v.) to prevent possible reactions to the mAbs.
Plasma viral RNA measurement and cell counts.
The plasma viral RNA load was monitored weekly after each virus exposure. Plasma collected from virus-exposed RMs was clarified by centrifugation at 2,300 × g for 3 min and either lysed directly (0.1 ml) in lysis buffer (bioMérieux) or further centrifuged to pellet the virus from a higher volume (0.5 ml to 1 ml) by ultracentrifugation at 49,100 × g for 60 min. The virus pellet was then lysed in 1 ml lysis buffer. A fixed amount of Q calibrator RNA (105 copies or 104 copies) was added to the lysed sample, and the nucleic acid was extracted by using acidified silica, as described previously (64). Analysis of viral loads was performed by using two RNA assays with different primer-probe sets and different levels of sensitivity. The viral RNA load was quantitated by using a real-time nucleic acid sequence-based amplification assay (NASBA) for SIV RNA, as described previously (65). Samples showing undetectable RNA loads were further analyzed by using a second assay, which detects the presence of viral RNA only qualitatively, with a limit of detection of ∼25 copies of RNA. The amplification conditions for the qualitative assay used were similar to those used for the quantitative assay except that the amplification oligonucleotides used for wild-type RNA were AATTCTAATACGACTCACTATAGGGAAAGGTTATACATTCTGACACATT (P1) and AGGATCAGATATTGCAGGAA (P2). The SIV wild-type beacon (5′-FAM [6-carboxyfluorescein]-CGCGATAACCCCATACCAGTAGGCAACAATCGCG-IABkFQ-3) had a fluorophore (FAM) linked to the 5′ end and a quencher (Iowa black) at the 3′ end. Animals with undetectable viral RNA as determined by both quantitative and qualitative assays were considered uninfected.
Following challenge, RMs were bled periodically, and absolute CD4+ T-cell measurements were performed by using the BD Biosciences TruCount platform. This method was used to quantify absolute counts and percentages of subpopulations of CD3+ CD4+ T cells, CD3+ CD8+ T cells, and total CD45+ leukocytes. For fluorescence-activated cell sorter (FACS) analysis, a minimum of 5,000 CD45+ leukocytes were acquired. Multicheck low-positive-control (stabilized human blood) samples yielded results consistent with expected values.
Virus reisolation and expansion.
The different passages of SHIV-E1 were isolated from infected animals by PBMC cocultures, as described previously (10). Briefly, RM PBMC, including cells from prescreened naive RM donors, were isolated by using Ficoll-Paque Plus (GE Healthcare Life Sciences); PBMC from naive donors were stimulated separately with 7 μg/ml concanavalin A (ConA) (Sigma-Aldrich) for 3 days and 40 U/ml of human interleukin-2 (IL-2) (Sigma-Aldrich) for 2 days prior to mixing them at a ratio of 1:1 with the PBMC from infected animals. Cocultures were maintained in medium with 10 ng/ml human TNF-α (10) and 40 U/ml of human IL-2. Supernatants were collected, every 48 to 72 h, with medium replacements at every collection, for >3 weeks, and filtered by using 0.45-μm syringe filters (EMD Millipore) to obtain cell-free virus stocks. SHIV-E1p1, SHIV-E1p2, and SHIV-E1p5 are biological isolates from animals R547 (day 203), R551 (day 14), and R974 (day 42), respectively. Large stocks of SHIV-E1p5 were developed by coculturing SHIV-E1p5-infected PBMC from animal R974 and PBMC from prescreened naive RMs, as described above. A separate stock of SHIV-E1p5 was grown in human PBMC isolated from naive healthy donors for in vitro studies, including determining the sensitivity of SHIV-E1p5 to a panel of broadly neutralizing Abs (bnAbs) and broadly neutralizing human sera.
RNA isolation and deep sequencing.
Virus preparations isolated by PBMC coculture from passage 1, passage 2 (day 14), and passage 5 as well as citrate plasma samples from passage 2 (day 57) were selected for deep sequencing. Virus stocks or plasma was thawed, transferred to ultracentrifuge tubes on top of a 2-ml 20% sucrose cushion, and centrifuged (SW 28 rotor; Beckman Coulter) (4 h at 24,000 rpm at 4°C). The supernatant was removed, each sample was resuspended in 1 ml phosphate-buffered saline (PBS), and vRNA was purified from the samples by using the QIAamp viral RNA minikit (Qiagen). Contaminating DNA was digested with the RNase-free DNase set (Qiagen). RNA samples were prepared for sequencing as previously described (66). Briefly, DNA, rRNA, and mRNA were removed from harvested nucleic acids by using a Turbo DNA-free kit (Life Technologies), a Ribo-Zero magnetic gold kit (Epicenter Biotechnologies), and RNA purification beads (Illumina), respectively. The remaining material was cleaned and concentrated by using an RNA Clean and Concentrator-5 kit (Zymo). Sequencing libraries were generated by using the Illumina TruSeq total RNA sample preparation kit, according to the manufacturer's instructions. The library was then sequenced by applying sequencing by synthesis (Illumina) using the 150-bp paired-end format on an Illumina MiSeq system. Initial data analysis occurred by using the Illumina pipeline to generate a FASTQ file containing all the reads. This was then mapped to the parental virus sequence by using Lasergene Seq-Man NGen. Quality trimming to mers was performed on reads with a minimum similarity of 93%. Single nucleotide polymorphisms (SNPs) were quantified by determining the number of reads at each position and comparing the variability to the reference sequence. This permits the expression of SNP abundance as a percentage of total reads.
SHIV-E1 molecular phylogenetic analysis.
Env amino acid sequences of the SHIV-E1 parental IMC, SHIV-E1p1d203, SHIV-E1p2d14, SHIV-E1p2d57, and SHIV-E1p5 were aligned in Mega7. The evolutionary history was inferred by using the maximum likelihood method based on the JTT matrix-based model (67). The tree with the highest log likelihood (−2,904.91) is shown. The initial tree(s) for the heuristic search was obtained automatically by applying neighbor-joining and BioNJ algorithms to a matrix of pairwise distances estimated by using a JTT model and then selecting the topology with the superior log likelihood value. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site (next to the branches). The analysis involved 5 amino acid sequences. All positions containing gaps and missing data were eliminated. There were a total of 874 positions in the final data set. Evolutionary analyses were conducted in Mega7 (68).
ACKNOWLEDGMENTS
We thank Charla Andrews for support and discussions and Nancy Miller for support through a contract of the SVEU (HHSN272201100023C). We thank Barbara Felber for the gift of CEMx174-GFP cells, Joey Wang and Tho Hua for technical support, and Asha Nabbale and Juan Esquivel for their expert assistance in the preparation of the manuscript. TZM-bl cells were obtained through the NIH AIDS Reagent Program, Division of AIDS, NIAID, NIH, from John C. Kappes, Xiaoyun Wu, and Tranzyme Inc.
This work was supported by the Henry M. Jackson Foundation for the Advancement of Military Medicine Inc. (prime award no. W81XWH-11-2-0174, subaward no. 798196, to R.M.R.) and by R01 AI100703 to R.M.R. Neutralization tier phenotyping was supported by NIH/NIAID contract HHSN27201100016C (to D.C.M.). We thank the Yerkes National Primate Research Center Comparative AIDS Core (funded in part by ORIP/OD P51OD011132) for providing naive RM blood.
REFERENCES
Articles from Journal of Virology are provided here courtesy of American Society for Microbiology (ASM)
