Differential Outcomes Following Optimization of Simian-Human Immunodeficiency Viruses From Clades AE, B and C.

We sought to enhance the infectivity of three SHIV stocks by optimization of a key residue in human immunodeficiency virus type 1 (HIV-1) Env (Env375). We developed the following three new simian-human immunodeficiency virus (SHIV) stocks: SHIV-SF162p3S/wild type, SHIV-AE16W, and SHIV-325cH. SHIV-SF162p3S could not be optimized, SHIV-AE16W proved comparable to the parental virus, and SHIV-325cH demonstrated markedly enhanced replicative capacity compared with the parental virus.

rhesus monkeys (7). Recently, Shaw and colleagues described a new strategy to produce SHIVs with improved binding to rhesus CD4 and increased replication in vivo (8). The phenylalanine at position 43 (F43) of CD4 engages position 375 in Env in the gp120 binding pocket (9,10). Env375 is, thus, a critical component of the binding pocket that aids in stabilization of the CD4-Env-bound conformations during viral entry. Furthermore, sequence analyses between SIV and HIV-1 Env revealed that the naturally occurring residues in SIV at Env375 are bulky and/or hydrophobic residues such as M, H, W, Y, and F, whereas HIV-1 Env375 typically has an S residue (11). Shaw and colleagues showed that mutating Env375 to the naturally occurring amino acids found in SIV Env at this position (M, H, W, Y, and F) resulted in SHIVs with a higher in vivo replicative capacity (8).
Here, we applied this optimization strategy to SHIV-SF162p3 (clade B), SHIV-AE16 (clade AE), and SHIV-325c (clade C) challenge stocks (12)(13)(14). We introduced hydrophobic and/or bulky amino acid mutations into Env375 (11), and we generated 6 variants for each SHIV. We performed a pool competition study to determine the optimal variant for each SHIV, and we observed the following three distinct outcomes with this optimization procedure: SHIV-SF162p3S could not be improved, SHIV-AE16W was comparable to the parental virus, and SHIV-325cH showed greatly enhanced replicative capacity compared with the parental virus.

RESULTS
Generation of SHIV Env375 variants. Our lab has previously generated SHIV-SF162p3 (12), SHIV-AE16 (13), and SHIV325c (14) challenge viruses. These SHIVs were constructed using the conventional KB9 SHIV design strategy (Fig. 1A). Here, we designed a new panel of SHIVs constructed with HIV-1 env sequences from SHIV-SF162p3, SHIV-AE16, and SHIV325c and cloned them into a replication-competent, pathogenic SIVmac766-based SHIV.D.191859.dCT clone (Fig. 1B) (8). This clone was previously shown to have a higher replicative capacity in vivo than KB9-derived viruses (8). The vpu and env genes were exchanged for the corresponding regions in SHIV.D.191859.dCT. Site-directed mutagenesis was used to substitute the wild-type amino acid at Env375 to mimic the larger and/or hydrophobic amino acids at Env375 in SIV. Env375 sequences were modified from S/wild type to M, H, W, Y, and F for SHIV-SF162p3 and SHIV-325c and from H/wild type to S, M, Y, W, and F for SHIV-AE16. A total of 6 variants for each of the three SHIVs were used for transfection in 293T cells to generate viruses. 293T cell cultures were then used to propagate each virus in either rhesus or human peripheral blood mononuclear cells (PBMCs) ( Table 1). These values were established for the virus stocks after 12 to 15 days in culture with PBMCs. As we have previously reported (13,15,16), SHIV-SF162p3 replicated well in both rhesus and human PBMCs, whereas SHIV-AE16 and SHIV-325c replicated more efficiently in human PBMCs (Table  1). These data are consistent with growth characteristics of the parental viruses constructed in the KB9 backbone (13)(14)(15). Interestingly, the new Env375 S/wild-type, M, and H variants for SHIV-SF162p3 showed 1-to 2-log greater replication in vitro than the parental SHIV-SF162p3 stock (Table 1). For SHIV-AE16, all Env375 variants displayed 1to 3-log greater replication than the parental stock. For SHIV-325c, all Env375 variants exhibited 2-to 3-log greater replicative capacity than the parental stock, and the S/wild-type, M, H, and W variants were the most replicative in vitro. Taken together, these findings suggest that the SHIV.D.191859.dCT backbone and the Env375 substitutions enhanced in vitro growth kinetics for SHIV-SF162p3, SHIV-AE16, and SHIV-325c compared with the parental viruses constructed in the KB9 backbone.
Env375 variant pool infection study in rhesus monkeys. We performed a pooled competition experiment in vivo to define the Env375 variant from each SHIV strain with the most robust replication in rhesus monkeys. Three SHIV variant pools were constructed with Env375 variants from SHIV-SF162p3 (S/wild type, M, H, W, Y, and F), SHIV-AE16 (S, M, H/wild type, W, Y, and F), and SHIV-325c (S/wild type, M, H, W, Y, and F). Twelve adult rhesus monkeys (n ϭ 4/group) were inoculated with a single intravenous (i.v.) equimolar inoculum of 10 8 RNA copies of each SHIV pool. In this case, we used viral load rather than 50% tissue culture infective dose (TCID 50 ) measurements to create pools since the in vitro infectivity to in vivo infectivity relationship was not clear since some stocks that did not grow in rhesus PMBCs grew well in rhesus monkeys. An additional 11 adult rhesus monkeys were challenged with the parental virus stocks derived from the KB9 backbone for comparison. Animals inoculated with the parental SHIV-SF162p3 stock exhibited peak viral loads ranging from 6.9-to 7.9-log RNA copies/ml at 2 weeks postinoculation, but only 3 of 4 animals had detectable viremia by week 20 ( Fig. 2A). Animals that received the SHIV-SF162p3 Env375 variants had comparable peak viral loads to the parental stock at 2 weeks postinoculation, but all animals (parental and Env327 variants) maintained detectable viremia at week 20 ( Fig. 2A).
Monkeys that received the parental SHIV-AE16 stock displayed peak viral loads ranging from 5.1-to 7.3-log RNA copies/ml at 2 weeks postinoculation but only showed detectable viremia in 3 of 4 animals at week 20 ( Fig. 2B). Animals that received the SHIV-AE16 Env375 variants displayed peak viral loads ranging from 6.5-to 7.2-log RNA copies/ml at 2 weeks postinoculation and detectable viremia in 3 of 4 animals at week 20 ( Fig. 2B).
Animals that received the parental SHIV-325c stock displayed low peak viral loads, as we previously reported, ranging from 3.7-to 5.5-log RNA copies/ml and a late peak viremia at 4 weeks postinoculation, and 2 of 3 animals had low or undetectable viremia at week 20 (Fig. 2C). In contrast, animals that were infected with the SHIV-325c Env375 variants showed median viral loads that were 3.9-log RNA copies/ml higher at peak viremia and 24-fold higher at set point viremia than the controls. Furthermore, peak viral loads ranged from 6.8-to 7.3-log RNA copies/ml at week 2 (P Ͻ 0.001 at week 1, P Ͻ 0.05 at week 2, comparing variant pool versus parental stock). Moreover, all animals that received the SHIV-32c Env375 variants were still viremic at week 20 (Fig. 2C). Taken together, these findings suggest that the pooled Env375 variants of SHIV-325c replicated to substantially higher levels than the parental SHIV-325c challenge stock. In contrast, the pooled Env375 variants of SHIV-SF162p3 or SHIV-AE16 were only modestly different than the parental SHIV challenge stocks. Viral sequencing. To define the optimal SHIV variants in vivo, we utilized singlegenome amplification (SGA) to analyze env sequences from animals infected with the pooled Env375 variants for SHIV-SF162p3, SHIV-AE16, and SHIV-325c. We assessed plasma from week 2, week 8, and week 20 postinfection in each animal. In the animals infected with the SHIV-SF162p3 variants, parental Env375S was the predominant circulating variant comprising 99/128 (77%) of the total variants sequenced at week 2, with the remainder 22/128 (17%) for M, 4/128 (3%) for H, and 3/128 (2%) for F (Fig. 3A). By week 8 and 20, however, Env375S was the predominant or exclusive variant observed in 249/250 (Ͼ99%) of the total sequences (Fig. 3A). These findings indicate that the Env375S/wild-type variant of SHIV-SF162p3 had the highest replicative capacity in vivo. These data demonstrate that the Env375 residue in the parent SHIV-SF162p3 stock (serine) was already the optimal Env375 residue.
Lastly, in the animals infected with the SHIV-325c variants, there was a diversity of Env375 variants observed at week 2, including 23  predominant Env375 variant was H at 158/226 (70%) sequences, while the remaining sequences were 33/226 (15%) for M, 24/226 (11%) for Y, and 11/226 (5%) for F (Fig. 3C). These data show that the Env375H variant of SHIV-325c had enhanced replicative capacity in vivo compared with the parental and other variants of SHIV-325c.
Generation of large-scale SHIV challenge stocks. We selected the three most prevalent Env375 variants (SHIV-SF162p3S/wild type, SHIV-AE16W, and SHIV-325cH) in the SHIV.D.191859.dCT backbone identified by SGA (Fig. 3) for the generation of large-scale challenge stocks. We inoculated 293T transfection-derived supernatant into rhesus PBMCs for SHIV-SF162p3S/wild type or into human PBMCs for SHIV-AE16W and SHIV-325cH. High titer stocks could not be generated for SHIV-AE16W or SHIV-325cH in rhesus PBMCs, despite improved replicative capacity in vivo. The large-scale stocks all had comparable viral load titers in the 10 9 RNA copies/ml range, high SIV p27 levels ranging from 72.3 to 110 ng/ml, and TCID 50 /ml infectivity titers of 10 6 for SHIV-SF162p3S/wild type and SHIV-AE16W and 10 8 for SHIV-325cH (Table 2).
We next evaluated the infectivity of the SHIV-SF162p3S/wild-type, SHIV-AE16W, and SHIV325cH challenge stocks following repetitive intrarectal (i.r.) inoculations in rhesus monkeys. Twelve adult rhesus monkeys (n ϭ 4/group) received six repetitive i.r. challenges with a 1:100 dilution of each SHIV challenge stock. For SHIV-SF162p3S/wild type, all four animals became productively infected by the second challenge, with peak viral loads ranging from 4.9-to 6.7-log RNA copies/ml and detectable viremia in 3 of 4 animals at week 26 (Fig. 4A). For SHIV-AE16W, 3 of 4 animals became infected after the first, second, and fifth challenge, respectively, with peak viral loads ranging from 5.6-to 7.3-log RNA copies/ml and with detectable viremia throughout the course of the study (Fig. 4B). For SHIV-325cH, all animals became infected after two to six challenges, with peak viral loads from 5.1-to 7.1-log RNA copies/ml, and all animals showed detectable viremia at the end of the study (Fig. 4C). These data show that the 1:100 dilutions of  SHIV-SF1623S/wild-type, SHIV-AE16W, and SHIV325cH stocks infected most animals by the i.r. route and that a 1:100 dilution may be an appropriate dose for repetitive i.r. challenge regimens.
We compared peak and week 20 setpoint plasma viral loads for the SHIV-325c parental, pool, and 325cH variant challenge stocks. At peak viremia, animals that received the pool and 325cH variant displayed a 2.9-log (P Ͻ 0.05) and 2.6-log increase, respectively, in viral loads compared with the parental stock (Fig. 5A). At week 20, animals that received the pool and 325cH variant displayed a 1.7-log and 1.6-log (P Ͻ 0.05) increase, respectively, compared with the parental stock (Fig. 5B). These data demonstrate that the optimized SHIV-325cH challenge stock showed markedly enhanced in vivo replication compared with the parental challenge stock.
Antigenic properties of SHIV challenge stocks. Finally, we evaluated the neutralization sensitivity of each SHIV stock in TZM-bl cells with a panel of broadly reactive neutralizing monoclonal antibodies (MAbs) as well as a panel of 5 chronic HIVϩ serum samples. The SHIV-SF162p3S/wild-type, SHIV-AE16W, and SHIV-325cH challenge stocks retained neutralization profiles comparable to the parental viruses (Table 3). However,   (Table 3). Overall, these data indicate that the pSHIV.D.191859.dCT backbone as well as the Env375 amino acid substitutions did not substantially impact neutralization profiles, except for higher sensitivity to CD4 inhibition.

DISCUSSION
Recent reports have demonstrated that Env375 optimization can improve SHIV pathogenicity in vivo (8,17). In this study, we confirm and extend these findings by describing the generation of three new SHIV challenge stocks based on SHIV-SF162p3, SHIV-AE16, and SHIV-325c. We demonstrate three distinct outcomes following Env375 optimization: (i) no improvement (SHIV-SF162p3), (ii) generation of a variant with comparable replicative capacity to the parental virus (SHIV-AE16W), and (iii) generation of a variant with improved replicative capacity compared with the parental virus (SHIV-325cH).
HIV-1 Env binding to CD4 is critical for viral entry and in vivo replication (8,11,18,19). Li et al. showed three SHIVs (CH505, CH848, and BG505) had low baseline CD4 binding and required Env375 substitutions to enhance CD4 binding and in vivo infectivity (8). Our previously constructed parental SHIV-325c challenge stock (14) had a relatively low replicative capacity compared with other SHIVs. For the new variant SHIV-325cH, the Env375H substitution led to markedly enhanced replication in vivo. These data are consistent with previous findings that the Env375H substitution may have a selective advantage over the naturally occurring Env375S for HIV-1 clade C Envs (17).
One important aspect of our study that warrants further evaluation is to determine whether SHIV-325cH can lead to clinical AIDS progression in rhesus monkeys. Despite significant differences in viral loads measurements compared with control animals, CD4 ϩ T cell counts were inconsequential. Li et al. showed in some cases that rhesus monkeys did exhibit AIDS-defining conditions after challenge (8). However, some of these animals received CD8 ϩ T cell-depleting drugs that may have contributed to the high-level viremia observed. Nonetheless, these strategies suggest that further research with SHIV-325cH is warranted since clade C HIV-1 causes the majority of new HIV-1 transmissions around the world (20).
In summary, we report the generation of new SHIV-SF162p3S/wild-type, SHIV-AE16W, and SHIV-325cH challenge stocks. SHIV-325cH represents a new clade C SHIV with a high replicative capacity in vivo as a result of the Env375H mutation and may be useful for preclinical studies of preventative and therapeutic interventions for clade C HIV-1. This virus did not require in vivo passaging and was infectious in a low dose, repetitive challenge protocol in rhesus monkeys. Our data demonstrate that Env375 mutations can lead to SHIVs with reduced, comparable, or enhanced replication capacity in rhesus monkeys.

MATERIALS AND METHODS
Animals. Indian-origin rhesus monkeys (Macaca mulatta) were used in the current study. Thirty-five Mamu-A*01-negative adult male and female animals were housed at Bioqual (Rockville, MD), and 12 Mamu-A*01-negative adult male and female animals were housed at Alpha Genesis (Yemassee, SC).
Ethics. All animals were maintained in accordance with the Association for Assessment and Accreditation of Laboratory Animals with approvals from the relevant Institutional Animal Care and Use Committees (IACUCs), including the Bioqual IACUC and the Alpha Genesis IACUC. The animal protocols in this study adhered to NIH standards set forth in the "Policy on Humane Care of Vertebrate Animals Used in Testing, Research, and Training" and the "Guidelines for the Care and Use of Laboratory Animals" (DHHS publication number NIH 85-23). Animal welfare was maintained by the Bioqual and Alpha Genesis animal management programs, which are accredited by the American Association for the Accreditation of Laboratory Animal Care (AAALAC) and meet all applicable federal and institutional standards for standard housing, standard monkey diet, water ad libitum, social enrichments, and steps intended to minimize suffering, such as the use of anesthesia for all procedures and analgesia for invasive procedures.
Cells. Human and rhesus monkey peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Hypaque gradient purification, followed by stimulation with concanavalin (ConA; 6.25 g/ml) and human interleukin-2 (IL-2; 20 units/ml; AIDS Reagent Program) overnight. Human whole blood was purchased commercially (Research Blood Components), and rhesus whole blood was purchased from Bioqual for PBMC isolation. Cells were cultured in RPMI 1640 medium (Gibco) supplemented with 20% fetal bovine serum, 2 mM glutamine, 100 units/ml penicillin, 100 g/ml streptomycin, and 20 units/ml of IL-2. TZM-bl cells (JC53-bl clone 13 cells) are derived from a HeLa cell line (JC.53) that stably expresses CD4 and HIV coreceptors, as well as luciferase and ␤-galactosidase, under the control of the HIV-1 long terminal repeat.
Design and construction of SHIV molecular clones. HIV-1 envelope sequences from SHIV-SF162p3, SHIV-AE16, and SHIV-325c, previously constructed in the KB9 backbone (8,(13)(14)(15), were used to design a new panel of SHIVs in an optimized SHIV.D.191859.dCT backbone (8,21,22). The SHIV.D.191859.dCT backbone containing a MfeI restriction site upstream of vpu and an AvrII restriction site at the C-terminal end of env was provided by George Shaw, University of Pennsylvania, Philadelphia, PA. In order to facilitate the cloning of SHIV-SF162p3, SHIV-AE16, and SHIV325c, site-directed mutagenesis was performed with the Q5 site-directed mutagenesis kit (New England BioLabs) to introduce silent mutations into vpu and env sequences synthesized by GeneArt (GeneArt, Invitrogen, Darmstadt, Germany) to remove internal MfeI and AvrII restriction sites, leaving only MfeI and AvrII sites flanking the 5= and 3= ends. Bulky and/or hydrophobic amino acids were substituted for the wild-type amino acid at position 375 in the HIV env sequence, using primers shown in Table 4 Generation of SHIV challenge stocks. Six infectious molecular clones of SHIV-SF162p3S/wild type, M, H, Y, W, and F replicated to high levels in rhesus PBMCs, whereas six infectious molecular clones of SHIV-AE16S, M, H/wild type, W, Y, and F and six infectious molecular clones of SHIV-325cS/wild type, M, H, W, Y, and F replicated to high levels in human PBMCs. SHIV-SF162p3S/wild type, SHIV-AE16W, and SHIV-325cH were selected to generate large-scale challenge stocks. Each virus stock was produced from 120 ml of rhesus monkey or human whole blood. Cell culture supernatants harvested from transiently transfected 293T cell were used to inoculate ConA-stimulated PBMCs in the presence of 20 U/ml human IL-2 (AIDS Research and Reference Reagent Program). The medium was replaced and collected in 3-day increments over the course of 12 to 15 days. Virus was quantified by SIV p27 enzyme-linked immunosorbent assay (ELISA; Zeptometrix), 50% tissue culture infectious dose (TCID 50 ) in TZM-bl cells, and reverse transcription-quantitative PCR (RT-qPCR).
TCID 50 titer in TZM-bl cells. Virus stocks were assessed for infectivity by inoculation of TZM-bl cells using 5-fold serial dilutions in the presence of 11 g/ml of diethylaminoethyl (DEAE)-dextran hydrochloride (Sigma, St. Louis, MO) in quadruplicate wells. Virus infectivity was determined 48 h postinfection by measuring the levels of luciferase activity expressed in infected cells. The TCID 50 was calculated as the dilution point at which 50% of the cultures were infected.