INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

In both human immunodeficiency virus type 1 (HIV-1)- and simian immunodeficiency virus (SIV)-infected individuals, high viral load and generalized immune activation herald CD4⁺ T-cell decline and progression to AIDS (22, 36, 42, 49, 54). This correlation most likely reflects the fact that these lentiviruses require cellular activation signals to productively infect CD4⁺ T lymphocytes (7, 46, 57, 64). It also raises the possibility that chronic immune stimulation plays a central role in HIV and SIV pathogenesis. Support for this hypothesis comes from studies of HIV-1 that have demonstrated transient increases in plasma viremia and proviral burden in infected individuals immunized against hepatitis B or tetanus toxoid, or having an intercurrent secondary infection (11, 13, 55), and from recent studies showing that continuous rounds of virus infection and replication drive rapid CD4⁺ T-cell turnover (28, 62). On the other hand, the persistent replication of HIV and SIV that occurs throughout the course of infection could be the primary driving force underlying the maintenance of a chronic immune activation state (22). In fact, a unique strain of SIVsmm, SIVPBj14, causes massive T-cell activation both in vivo and in vitro, demonstrating that a lentivirus is capable of inducing an immune activation state (52). Together, these data not only demonstrate the importance of virus-host interactions in the regulation of HIV and SIV replication but also suggest that a determinant of virulence may be defined by a virus's ability to induce and utilize immune activation signals for efficient replication.

The requirement for cellular activation signals in the initiation and maintenance of HIV-1 and SIV replication is well established in vitro (7, 45, 46, 57, 64), but the mechanism by which the virus interacts with cellular signalling pathways to enhance viral replication remains poorly understood. In infected T cells, activation signals (e.g., via T-cell receptor/CD3 and CD28 and cytokine receptors) release the arrest in viral replication that occurs at postentry steps prior to integration of the provirus, including reverse transcription and nuclear translocation of the viral preintegration complex (7, 45, 46, 57, 64). Furthermore, in productively infected cells, cellular activation is important for upregulating viral transcription from the long terminal repeat (LTR) (2, 27, 29). Additionally, there is evidence that HIV and SIV proteins such as envelope (Env) gp120 and Nef can promote or suppress viral replication through the induction of cytokines or modulation of T-cell activation, respectively (6, 20, 26, 37). In the context of these data, virulence may potentially be influenced by genetic variation in viral determinants that affect infectivity, postentry steps in the virus life cycle, and transcription.

Genetic variation has been shown to influence the phenotype of HIV-1 (34). Indeed cytopathic variants of HIV-1 that emerge in the peripheral blood (PB) during the course of infection may play an active role in determining the rate of disease progression. These viruses are associated with an increase in viral load, accelerated CD4⁺ T-cell decline, and onset of AIDS; they are frequently able to infect T-cell lines, and they may be rapidly replicating, highly cytopathic, and syncytium inducing (T-tropic, rapid-high/SI) in tissue culture (1, 12, 14, 15, 23, 59, 60). By contrast, viruses isolated early after infection, when CD4⁺ T-cell counts are stable or slowly declining, are commonly macrophage-tropic, slowly replicating, minimally cytopathic, and non-syncytium inducing (M-tropic, slow-low/NSI) (1, 12, 14, 15, 23, 59, 60). We have observed a similar shift in the phenotype of viruses isolated from the PB of SIVMne-infected macaques at early and late stages of infection (50). Together, these data have led to the hypothesis that T-tropic, rapid-high/SI viruses are more pathogenic than M-tropic, slow-low/NSI viruses. However, immune dysfunction is evident early in infection (53), when the virus population tends to be M-tropic, slow-low/NSI, suggesting that these viruses may also have an impact on disease progression.

Immune activation in HIV-1- and SIV-infected individuals is prominent in secondary lymphoid tissues such as lymph nodes (LNs), which serve as major reservoirs for the viruses (10, 21, 22, 44). In LNs, manifestations of disease are evident at early and late stages of infection (10, 22, 44), perhaps due to viral replication associated with chronic immune stimulation within this tissue. Moreover, it has been shown, at least for SIV, that the predominant variants harbored in lymphoid tissue are genetically distinguishable from those found in PB (8). While there is a high level of viral replication and evidence for disease in LNs of HIV-1- or SIV-infected individuals, primary full-length LN-derived molecular variants of HIV-1 or SIV have not been directly cloned (i.e., without prior selection in culture) and characterized. Thus, in this study we describe the direct cloning and in vitro characterization of a LN-derived molecular clone of SIVMne.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Isolation of a SIVMne molecular clone from LN tissue. DNA was extracted from mesenteric LN tissue of a pig-tailed macaque (Macaca nemestrina, animal T78027) infected with an uncloned isolate of SIVMne (5). This isolate reproducibly causes an AIDS-like syndrome in pig-tailed and rhesus (M. mulatta) macaques in 1 to 3 years (5). The mesenteric LN tissue was biopsied at 16 months postinfection when the animal had declining CD4⁺ T-cell counts and early signs of AIDS. The DNA was digested with EcoRI, layered onto a 10 to 40% continuous sucrose gradient in STE buffer (1 M NaCl, 20 mM Tris [pH 8], 5 mM EDTA [pH 8]), and fractionated by centrifugation at 26,000 rpm for 20 h at 15°C in a Beckman (Fullerton, Calif.) SW41.1 rotor. Fractions containing DNA fragments 10 to 20 kb in size were concentrated by ethanol precipitation, ligated into the EcoRI sites of the  Dash II arms, and packaged by using the Gigapack II XL system as specified by the manufacturer (Stratagene, La Jolla, Calif.). Approximately 10⁷ plaques were screened for SIV, using ³²P-labeled gag and env DNA probes derived from a pathogenic clone of SIVMne (SIVMneCl8 [4, 38, 43]). Plaques that scored positive for both probes were isolated and purified. One phage clone encoded a full-length provirus which we designated SIVMne027. The provirus was excised from the  Dash II vector with EcoRI and ligated into the EcoRI site of pUC18. The full DNA sequence of SIVMne027 was determined by manual sequencing using Sequenase version 2.0 (United States Biochemical, Cleveland, Ohio) as well as an ABI automated sequencer.

Construction of recombinant viruses. To construct a chimeric virus containing the 5' half of SIVMne027 (R-U5-gag-pol-5'vif) and 3' half of SIVMneCl8 (3'vif-vpx-vpr-rev-tat-env-nef-U3-R) or the reciprocal recombinant virus, BstBI-SalI fragments from the plasmid proviral clones of SIVMne027 and SIVMneCl8 (pMneCl8) were exchanged. BstBI cleaves both proviruses in vif (position 5343, which is 536 bp into the vif gene), and SalI cuts in the polylinker region of both proviral plasmid clones at a site downstream of the cellular sequences flanking the 3' LTR of each provirus. The chimeric virus that consisted of the 5' half of SIVMne027 and 3' half of SIVMneCl8 was designated SIVMne027/Cl8, and the reciprocal virus was designated SIVMneCl8/027.

Generation of virus stocks. Ten million CEMx174 cells were transfected with 10 µg of each plasmid proviral construct by the DEAE-dextran method and cultured for 10 days in RPMI complete medium (RPMI 1640 supplemented with 10% heat-inactivated [56°C for 30 min] fetal bovine serum, 2 mM L-glutamine, penicillin [100 U/ml], streptomycin [100 µg/ml], and amphotericin B [250 ng/ml]). Conditioned supernatants were clarified by centrifugation at 1,500 rpm in a Beckman clinical centrifuge, filtered through 0.22-µm-pore-size filters (Millipore, Bedford, Mass.), and stored in 1-ml aliquots at 70°C. The tissue culture infectious dose (TCID) of each virus supernatant stock per milliliter was determined by the sMAGI assay (9). Briefly, sMAGI indicator cells were infected with 1 to 50 µl of viral supernatant and stained for -galactosidase expression 3 days postinfection (p.i.) as previously described (9).

Infection of human CD4⁺ cell lines. The human CD4⁺ T-cell lines Jurkat, CEM, MT4, and Molt4 clone 8, and a CD4⁺ T-B-cell hybrid cell line, CEMx174, were maintained in RPMI complete medium. To examine viral tropism for these cell lines, triplicate cultures of 5 × 10⁵ cells from each line were infected with 1,000 TCID of each virus derived from the molecular clones of SIVMne and propagated in 2 ml of RPMI complete medium. Every 3 days, the total cell number in each culture was adjusted to 5 × 10⁵, and fresh medium was added. If there were cytopathic effects in a culture that reduced the total cell number below 5 × 10⁵, then no cells were removed, but the culture medium was replaced with 2 ml of fresh medium. At 6, 12, and 18 days p.i., 1 ml of supernatant from each culture was removed and stored at 70°C. Viral replication was assessed by examining the supernatants for cell-free SIV p27^gag by antigen enzyme-linked immunosorbent assay (ELISA) specific for the SIV p27^gag capsid protein (Immunotech, Westbrook, Maine). Cultures were scored positive for viral replication if the cell-free supernatants taken at 12 and 18 days p.i. were positive for SIV p27^gag (>50 pg/ml).

Replication and cytopathicity of SIVMne variants in CEMx174 cells. To compare the replication rates and cytopathicity of the different viruses in the CEMx174 cell line, duplicate cultures of 8 × 10⁵ CEMx174 cells were infected with 200 TCID of each virus for 4 h in RPMI complete medium. Following the incubation, the cells were pelleted by centrifugation, washed twice with phosphate-buffered saline (PBS) to remove residual free virions, and resuspended in 2 ml of fresh RPMI complete medium. Every 2 to 3 days, both the viable and total cell numbers were determined by trypan blue dye exclusion, and the total syncytia per culture were counted by visual inspection. Only syncytia larger than 5 cell diameters were scored. Viral replication was monitored by assaying dilutions of the culture supernatants for cell-free SIV p27^gag by antigen ELISA. All p27^gag values were obtained in the linear range of the assay. The cell number in each culture was adjusted to 8 × 10⁵, and fresh RPMI complete medium was added to a final volume of 2 ml. If the cell number was below 8 × 10⁵, then no cells were removed, but the culture medium was replaced with 2 ml of fresh RPMI complete medium.

Isolation and infection of macaque PBMCs, monocytes/macrophages, and T cells. Macaque PB mononuclear cells (PBMCs) were isolated from whole blood of SIV- and simian type D retrovirus-negative pig-tailed macaques (M. nemestrina) by Ficoll-Hypaque centrifugation as previously described (50). To examine viral replication in nonstimulated PBMCs (i.e., in the absence of exogenous mitogens) or interleukin-2 (IL-2)-stimulated PBMCs, 3 × 10⁶ PBMCs were infected with 6,000 TCID of each virus in duplicate in 1 ml of RPMI complete medium. Following a 24-h incubation, the cells were pelleted by centrifugation, washed twice with PBS, and resuspended in RPMI complete medium. Nonstimulated PBMC cultures were grown continuously in the absence of exogenous IL-2, while IL-2-stimulated PBMC cultures were grown continuously in the presence of exogenous IL-2 (20 U/ml) for the duration of the experiment. Every 3 to 4 days, supernatants were harvested, stored at 70°C, and used to monitor viral replication by antigen ELISA for p27^gag.

Nucleotide sequence accession number. The complete nucleotide sequence of SIVMne027 was entered into GenBank under accession no. U79412.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Isolation of an infectious molecular clone of SIVMne from LN tissue. To study the properties of viruses found in LNs, we first isolated a proviral clone directly from unpassaged LN tissue of a pig-tailed macaque (animal T78027) that had been infected with an uncloned, pathogenic isolate of SIVMne (5). Macaque T78027 had been infected with SIVMne for 16 months and had declining CD4⁺ T-cell counts and early signs of AIDS at the time of the LN biopsy. Using recombinant lambda phage cloning, we obtained a single full-length proviral variant of SIVMne, designated SIVMne027, from this animal's LN DNA sample. CEMx174 cells transfected with the proviral clone of SIVMne027 expressed virus that was infectious for macaque and human cells (Table 1). The host cell range of SIVMne027 was similar to that of SIVMneCl8, a pathogenic virus that was cloned from the SIVMne isolate that was inoculated into macaque T78027 (4, 38, 43). SIVMne027 replicated in pig-tailed macaque PBMCs and to low levels in macaque monocytes/macrophages. It also replicated in the human CD4⁺ cell lines CEMx174 and MT4. However, the host range of SIVMne027 was different from those of uncloned mixtures of late variant viruses previously isolated from PBMCs of other pig-tailed macaques with AIDS. These uncloned mixtures of late variants had an expanded host range for CD4⁺ human cell lines that included Molt4 clone 8, but they replicated poorly or not at all in monocyte/macrophage cultures (Table 1 and reference 50).

SIVMne027 is rapidly replicating and highly cytopathic but non-syncytium inducing. We previously demonstrated that CEMx174 cells are highly sensitive to the cytopathic effects of SIV and can be used as an indicator for the identification of rapid-high/SI variants of SIVMne (50). To determine the biological characteristics of SIVMne027, we compared its replication kinetics, cytopathicity, and syncytium-inducing ability in CEMx174 cells with those of SIVMneCl8, a slow-low/NSI virus (50), and SIVMne170, a rapid-high/SI virus molecularly derived from an uncloned mixture of late variant viruses isolated by cocultivation from PBMCs (31). The phenotype of SIVMneCl8 is typical of viruses present early in infection, while the phenotype of SIVMne170 is representative of variants found late in infection (50). When CEMx174 cells were infected with each virus at a low MOI (0.00025), SIVMne027 replicated with delayed kinetics compared to SIVMneCl8 and SIVMne170, reaching a peak SIV p27^gag antigen later than SIVMneCl8 and SIVMne170, respectively (Fig. 1A). Furthermore, SIVMne027 was minimally cytopathic for CEMx174 cells, a phenotype similar to that of SIVMneC18 (Fig. 1B and C); specifically, it reduced the viable cell number fivefold and percentage of cells that were viable to 40%. In contrast, the rapid-high/SI virus, SIVMne170, reduced the viable cell number approximately 100-fold and the percentage of viable cells to less than 10%. Increasing the MOI to 0.025 or greater did not enhance the cytopathicity of SIVMne027 (data not shown). Finally, like SIVMneC18, SIVMne027 induced few syncytia compared to SIVMne170 (Fig. 1D). Syncytia were not observed in any other infected cell lines, including MT4 (data not shown). By these criteria, SIVMne027 is a slow-low/NSI variant of SIVMne.

View larger version (23K): [in this window] [in a new window]   FIG. 1.   Replication and cytopathicity of SIVMne027 in CEMx174 cells. Eight hundred thousand CEMx174 cells were infected with 200 TCID of virus. At 2-day intervals p.i., supernatants were harvested to monitor viral replication by antigen ELISA for SIV p27gag. The viable cell number was determined by trypan blue dye exclusion, and syncytia were scored as described in Materials and Methods. (A) SIV p27gag levels versus days p.i.; (B) extrapolated viable cell number versus days p.i.; (C) percentage of the total number of cells that are viable versus days p.i.; (D) total number of syncytia versus days p.i. The value shown for each time point is the average of duplicate cultures. Similar data were obtained in three independent experiments.

View larger version (37K): [in this window] [in a new window]   FIG. 2.   Replication and cytopathicity of SIVMne027 in PHA-stimulated macaque PBMCs. (A and C) SIV p27gag levels versus days p.i. Two million PHA-stimulated PBMCs were infected with 2,000 TCID of virus in duplicate. Virus production was monitored by antigen ELISA for SIV p27gag. The average antigen value for the duplicate cultures at each time point is shown, and the viruses used for infection are indicated. (B and D) Decrease in CD4+ T cells versus days p.i. Four million PHA-stimulated PBMCs were infected with 40,000 TCID of virus. The decrease in CD4+ T cells at the indicated time point (6 and 13 days p.i.) was monitored by FACS analysis as described in Materials and Methods. The values in panels B and D represent the decrease in the percentage of CD3+ CD4+ cells in the virus-infected cultures relative to the uninfected culture at each time point.

SIVMne027 replicates in nonstimulated PBMCs. A few studies have investigated the ability of HIV-1 and SIV to replicate in leukocyte cultures or PBMCs that had not been stimulated with potent mitogens such as PHA or concanavalin A (17, 20, 25, 32, 33, 47, 48, 61). To determine the efficiency of replication of SIVMne027 in PBMC cultures that had not been prestimulated with exogenous mitogens, we compared the replication kinetics of SIVMne027 and SIVMneCl8 in pig-tailed macaque PBMCs in the presence or absence of exogenous IL-2. Interestingly, SIVMne027 replicated to significant levels in the nonstimulated PBMC cultures, whereas SIVMneCl8 replicated poorly (Fig. 3A and C). The addition of IL-2 at 24 h p.i. enhanced the production of p27^gag in the SIVMne027-infected cultures 5- to 20-fold. In contrast, while the production of p27^gag in the SIVMneCl8-infected cultures was also increased by IL-2, the highest level of p27^gag was still 100- to 200-fold less than that attained by SIVMne027 (Fig. 3B and D). We observed these characteristics in seven independent experiments using PBMCs from three different pig-tailed macaques (data not shown). Furthermore, the PB-derived SIVMne variant clone, SIVMne170, behaved like SIVMneCl8 in nonstimulated PBMC cultures, demonstrating that among variants of the SIVMne strain, the ability to replicate in nonstimulated PBMCs was unique to SIVMne027 (data not shown). These data demonstrate that SIVMne027 may have an increased capacity to replicate in PBMCs, and they also show that this virus is highly responsive to signals induced by IL-2. Additionally, the ability of SIVMne027 to replicate in nonstimulated PBMCs appeared to occur without the induction of cellular proliferation because we could not detect a difference in cell number between infected and uninfected control PBMCs by the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide)] colorimetric assay. Furthermore, like infection with other variants of SIVMne, infection with SIVMne027 did not increase expression of the T-cell activation marker CD69 or CD25. These data suggest that virus production may not involve CD3⁺ lymphocytes. However, selection for CD3⁺ cells from SIVMne027-infected PBMC cultures by immunomagnetic bead separation using an anti-macaque CD3 monoclonal antibody and subsequent analysis of the CD3⁺ cells by antigen ELISA for p27^gag or PCR for proviral DNA demonstrated that the T-cell population was productively infected (data not shown).

View larger version (21K): [in this window] [in a new window]   FIG. 3.   Replication of SIVMne027 in nonstimulated and IL-2-stimulated pig-tailed macaque PBMCs. Three million nonstimulated PBMCs were infected with 6,000 TCID of virus and cultured in the presence or absence of exogenous IL-2. Virus production was monitored by antigen ELISA for SIV p27gag. Each antigen value is the average of duplicate cultures. (A and C) Virus production from nonstimulated PBMCs. (B and D) Virus production from IL-2-stimulated PBMCs. IL-2 was added 24 h p.i. and was maintained at 20 U/ml for the duration of the experiment. Panels A and B and panels C and D represent independent experiments using PBMCs prepared from different pig-tailed macaques.

Comparison of the replication kinetics of SIVMne027 and other SIV clones. To further characterize SIVMne027, we compared its ability to replicate in nonstimulated PBMCs with those of a molecular clone of the immunostimulatory viral isolate, SIVPBj14 (clone SIVPBj1.9 [16]), and a highly macrophage-tropic clone, SIVmac1A11 (3), as well as SIVMneCl8. SIVMne027, SIVPBj1.9, and SIVmac1A11 replicated to substantial levels in nonstimulated PBMCs compared to SIVMneCl8 as measured by SIV p27^gag antigen production (Fig. 4A). The peak SIV p27^gag antigen level achieved in the SIVMne027- and SIVmac1A11-infected cultures was approximately 10,000 pg/ml, while the SIVPBj14-infected cultures produced a sevenfold-higher maximal level of antigen (72,000 pg/ml). In contrast, the SIVMneCl8-infected cultures reached a maximum SIV p27^gag level at 300 pg/ml.

View larger version (17K): [in this window] [in a new window]   FIG. 4.   Comparison of the replication kinetics of SIVMne027 and of other strains of SIV in nonstimulated macaque PBMCs and macaque T-cell-enriched cultures. (A) Virus production from nonstimulated macaque PBMCs. The infection and analysis were performed as described in the legend to Fig. 3. (B) Virus production from primary macaque T-cell cultures. Cultures of T cells were infected with each virus at an MOI of 0.001. (C) Virus production from macaque monocyte-derived macrophage cultures. Monocyte-derived macrophage cultures were infected with each virus at an MOI of 0.001. For panels B and C, virus production was monitored by SIV p27gag antigen ELISA, and cultures were maintained as described in Materials and Methods.

Genetic comparison of SIVMne027 with other SIVs. We determined the sequence of the entire SIVMne027 genome and compared the predicted amino acid sequences of the gene products with the homologous regions of SIVMneCl8, SIVPBj14 (16), SIVmac1A11 (35), and a prototype AIDS-inducing SIV clone, SIVmac239 (30, 35) (Table 2). SIVMne027 encodes a complete open reading frame for each gene. SIVMne027 had the strongest sequence similarity to SIVMneCl8; the percentage of amino acid differences ranged from a low value of 0.7% in reverse transcriptase (RT) and integrase (IN) to a high value of 4.0 to 5.3% in Vpr, Tat, Rev, and Nef. In contrast, SIVMne027 differed overall from SIVPBj14 by 12.7%. SIVMne027 was most similar to SIVPBj14 in IN (3.8% different) and most different in Env, Vif, Tat, Rev, and Nef (17.3 to 24.6%). Our sequence analysis also revealed that SIVMne027 did not have an additional Src homology 2 (SH2) binding motif (YXXL) in Nef or a duplication of the NF-B binding site in the U3 LTR region (data not shown), two of the known determinants of SIVPBj14 that contribute to its ability to replicate in nonstimulated PBMCs (16-19). Additionally, SIVMne027 differed from SIVmac239 and SIVmac1A11 by a greater percentage than SIVMneCl8 in each viral protein, except Vpx, which was identical. However, both SIVmac239 and SIVmac1A11 had fewer amino acid differences with SIVMne027 than did SIVPBj14. Finally, SIVMne027 was also found to be genetically distinguishable from other clones of SIV and HIV-1 (data not shown).

Replication of SIVMne027 in nonstimulated PBMCs requires a determinant located within the 5' half of the viral genome. To identify the determinant in SIVMne027 that confers the ability to replicate in nonstimulated PBMC cultures, we constructed reciprocal recombinant proviruses that exchange the 5' and 3' regions of SIVMne027 and the parental virus, SIVMneCl8 (Fig. 5A), and compared the abilities of viruses derived from these constructs to replicate in nonstimulated and IL-2-stimulated PBMCs. When nonstimulated PBMCs were infected with these viruses, similar patterns of replication were observed in the cultures infected with the wild-type SIVMne027 and the chimeric virus, SIVMne027/Cl8, which encodes the R-U5-gag-pol-5'vif region of SIVMne027 and 3'vif-vpx-vpr-tat-rev-env-nef-U3-R region from SIVMneCl8 (Fig. 5B). In contrast, the reciprocal clone, SIVMneCl8/027, resembled SIVMneCl8 in that it did not show an appreciable level of virus replication as measured by p27^gag expression. Furthermore, the SIVMne027/Cl8 chimeric virus was responsive to the addition of IL-2. However, its peak production of p27^gag was approximately threefold lower than that of SIVMne027 (Fig. 5C). The SIVMneCl8/027 chimera, like SIVMneCl8, produced only a low level amount of p27^gag under these conditions. These results indicate that the 3' region of SIVMne027, including nef and the U3-R region of the LTR, does not contain the primary determinants for replication in nonstimulated or IL-2-stimulated PMBCs. Instead, the primary determinant(s) that confers SIVMne027's ability to replicate in nonstimulated and IL-2-stimulated PBMCs lies in a region of the 5' half of the virus, which includes U5-gag-pol and the 5' end of vif.

View larger version (22K): [in this window] [in a new window]   FIG. 5.   Construction and analysis of SIVMne027 and SIVMneCl8 chimeric viruses. (A) Schematic diagram of the chimeric viruses. B, the BstBI restriction site used for generation of the chimeric viruses. (B) Virus production from nonstimulated PBMCs. (C) Virus production from IL-2-stimulated PBMCs. For panels B and C, infection of PBMCs and the analysis of virus production were carried out as described in the legend to Fig. 3 and Materials and Methods.

View larger version (47K): [in this window] [in a new window]   FIG. 6.   Comparison of the predicted amino acid sequences of Pol from SIVMne027 and SIVMneCl8. The predicted amino acid sequence of SIVMne027 Pol serves as the reference. The regions encoding protease (Pr), RT, and IN are shown above the sequences. Similarities and differences between SIVMne027 Pol and the other SIV Pol proteins are shown with the same notation as described in the legend to Fig. 5.

View larger version (33K): [in this window] [in a new window]   FIG. 7.   Comparison of the predicted amino acid sequences of Gag from SIVMne027 and SIVMneCl8. The regions coding for MA, CA, and NC are indicated above the sequences. Amino acid similarities and differences are noted as described in the legend to Fig. 5.

View larger version (32K): [in this window] [in a new window]   FIG. 8.   Cytopathicity of the chimeric viruses for PHA-stimulated PBMCs. Four million PBMCs prestimulated with PHA for 3 days were infected with 40,000 TCID and cultured in RPMI complete medium and IL-2 (20 U/ml). Cytopathicity of the viruses for CD4+ T cells was monitored by FACS as described in Materials and Methods.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The predominant SIV variants in lymphoid tissue have been shown to be genetically different from those found in the peripheral blood of infected macaques (8). To begin to evaluate the contribution of LN-derived lentivirus variants to viral pathogenesis, we molecularly cloned a variant of SIV directly from DNA prepared from unpassaged LN tissue of a pig-tailed macaque inoculated with SIVMne. Importantly, this is the first infectious HIV-1 or SIV molecular clone isolated directly from unpassaged LN tissue. The virus, SIVMne027, displayed a distinct phenotype in culture compared to other SIVMne clones. SIVMne027 was macrophage-tropic, replicated poorly in T-cell lines, and was non-syncytium inducing. However, it replicated efficiently in PHA-stimulated PBMCs and was highly cytopathic for the CD4⁺ T-cell population. Interestingly, SIVMne027 also replicated efficiently in nonstimulated PBMCs in the absence of potent mitogens (e.g., PHA or concanavalin A), and it was responsive to IL-2 under these culture conditions. These data demonstrate that in SIV- and, by implication, HIV-1-infected individuals, LNs may harbor cytopathic variant viruses with an increased ability to utilize, and perhaps deregulate, cellular activation signals that are typically required for viral replication.

We previously demonstrated that SIVMne, like HIV-1, evolves from a slow-low/NSI virus to a rapid-high/SI virus population during the course of an infection (50). Interestingly, the primary determinant that conferred the rapid replication kinetics and increased cytopathic effects of a rapid-high/SI SIVMne variant molecular clone mapped to gag, but not the syncytium induction determinant, which mapped to the envelope surface protein coding region (31, 51), suggesting that rapidly replicating, highly cytopathic SIVMne variants may evolve independent of the syncytium-inducing phenotype. The data presented here demonstrate that a SIVMne variant, SIVMne027, can be rapidly replicating and highly cytopathic in the absence of an overt syncytium-inducing phenotype. The experimental inoculation of macaques with SIVMne027 will allow us to directly examine whether the increased in vitro virulence of SIVMne027 is predictive of increased in vivo pathogenicity. Furthermore, comparative studies of SIVMneCl8 and SIVMne027 may allow us to further define the viral genetic determinants and in vitro biological characteristics of virulence. In regard to the latter, it is noteworthy that the chimeric viruses, SIVMne027/Cl8 and SIVMneCl8/027, were less cytopathic for CD4⁺ T cells in PHA-stimulated PBMCs and replicated less efficiently than the parent virus SIVMne027 in IL-2-stimulated PBMCs, demonstrating that determinants in both halves of the virus may contribute to its overall virulence.

Unlike other SIVMne clones, SIVMne027 replicated efficiently in nonstimulated and IL-2-stimulated PBMCs. SIVMne027's ability to replicate in nonstimulated PBMCs initially appeared to be similar to that of the immunostimulatory virus, SIVPBj14. However, SIVMne027 was distinguishable from SIVPBj14 by several criteria. First, in contrast to the molecular clone of SIVPBj14, SIVPBj1.9, SIVMne027 did not replicate well in either macaque resting T-cell or monocyte-derived macrophage-enriched cultures; instead, it required the presence of both cell populations for efficient replication. Second, SIVMne027 did not contain either a duplication of the NF-B site in the U3 region of the LTR or an additional YXXL motif in the amino-terminal region of Nef, two mutations that have been shown to be critical for the phenotype of SIVPBj14 (16-20). We further verified that mutations in these regions of SIVMne027 were not the primary determinants conferring the ability to replicate in nonstimulated PBMCs with recombinant viruses. Third, SIVMne027 was slowly replicating and minimally cytopathic in CEMx174 cells, whereas SIVPBj14 has been shown to be highly cytopathic for this cell line (16). Fourth, while SIVPBj14 is able to induce T-cell proliferation (25), we could not detect an increase in lymphocyte proliferation in SIVMne027-infected PBMCs by the MTT assay in preliminary experiments. We also could not detect an up regulation of the T-cell activation markers CD69 and CD25 on CD4⁺ T cells in SIVMne027-infected cultures. Nevertheless, we did find infected CD3⁺ lymphocytes, demonstrating the importance of T cells in SIVMne027 replication in nonstimulated PBMC cultures. These data suggest that SIVMne027 may either induce immune activation signals that are sufficient for viral replication but not cellular proliferation or bypass the requirement for T-cell activation. However, IL-2 enhanced SIVMne027 production from infected PBMCs. This observation indirectly suggests that the SIVMne027-infected cells may be activated, because resting T cells are refractory to the effects of IL-2 (56). Our failure to detect T-cell activation in SIVMne027 infected-cultures by other methods, such as FACS, may be explained by a high rate of turnover of productively infected CD3⁺ CD4⁺ T cells. This interpretation is consistent with our data demonstrating that SIVMne027 is highly cytopathic for the CD3⁺ CD4⁺ T cells. Together, these data demonstrate that SIVMne027 is genetically and phenotypically distinct from SIVPBj14.

Cell-cell interactions play an important role in regulating lentivirus replication. Several recent studies have demonstrated that efficient replication of both SIV and HIV-1 requires contact between mononuclear phagocytes or antigen-presenting cells, such as macrophages or dendritic cells, and T cells (32, 33, 47, 48, 61). Furthermore, introduction of an additional SH2 binding domain into the amino-terminal region of Nef of SIVmac239 results in a virus (SIVmac239YE) that is capable of replicating in nonstimulated PBMCs in a macrophage-dependent manner (20). We demonstrate here that SIVMne027, like the SIVmac239YE mutant (20) but unlike SIVPBj1.9, depends on monocytes/macrophages for efficient replication in nonstimulated PBMCs. SIVMne027 also appeared to resemble the macrophage-tropic virus, SIVmac1A11, in its ability to replicate in nonstimulated PBMCs because SIVmac1A11 required the monocyte/macrophage population. However, SIVMne027 differed from SIVmac1A11 because it replicated poorly in cultures enriched for monocyte-derived macrophages, whereas SIVmac1A11 replicated to a high level. Indeed, replication of SIVmac1A11 in nonstimulated PBMCs could be explained entirely by macrophage infection. By contrast, the level of SIVMne027 replication in monocyte-derived macaque macrophages is too low to account for the high-level replication in nonstimulated macaque PBMCs. Therefore, it seems unlikely that replication of SIVMne027 in the monocyte/macrophage population alone can fully explain its ability to replicate to high levels in nonstimulated PBMCs. In support of this interpretation, we demonstrated here and have previously shown that both SIVMneCl8 and SIVMne170 replicated with kinetics similar to those of SIVMne027 in monocyte-derived macrophage cultures (31, 50, 51), but neither virus replicated to appreciable levels in either nonstimulated or IL-2-stimulated PBMCs, further suggesting that macrophage infection is insufficient for viral replication in nonstimulated or IL-2-stimulated PBMCs. A model consistent with our data is that SIVMne027 has a greater capacity to utilize activation signals resulting from mononuclear phagocyte-T-cell interactions for replication than the other molecular variants of SIVMne. Although the significance of the SIVMne027 phenotype is unclear, the selection for viruses during infection that are highly responsive to cellular activation signals seems plausible, especially because persistent SIV and HIV-1 replication correlates with immune activation and disease progression (22, 36, 42, 49, 54), productive infection of CD4⁺ T cells requires activation signals (7, 46, 57, 64), and vaccinations or secondary infections elevate viral loads in HIV-1-infected individuals (11, 13, 55).

The primary determinant(s) conferring the ability of SIVMne027 to replicate in nonstimulated and IL-2-stimulated PBMCs lies within gag-pol-5'vif. While we have not identified the specific mutation(s) that determine the ability of SIVMne027 to replicate in nonstimulated and IL-2-stimulated PBMCs in a functional assay, a U5, Pol, or Vif determinant seems unlikely because these regions are highly conserved between SIVMne027 and SIVMneCl8. On the other hand, the Gag polyprotein of SIVMne027 encodes nine unique mutations that distinguish it from SIVMneCl8 and other SIVs, making it a more likely candidate for the determinant that confers SIVMne027's ability to replicate in nonstimulated and IL-2-stimulated PBMCs. Interestingly, Novembre et al. showed, using recombinant viruses between SIVPBj clones and the closely related clones SIVsmmH4 and SIVsmmH9, that mutations in gag and the central region of the proviral genome which encodes for regulatory genes affect the phenotype of SIVPBj14 (40, 41). Our data further underscore the influence that mutations selected in genes located in the 5' half of the viral genome exert on viral replication. Identification of the specific mutations and the mechanism of action will be important for understanding how variation in gag, pol, or vif affects viral replication in vitro and the impact that this may have on virus replication and pathology in vivo.