Induction of Disease by a Molecularly Cloned Highly Pathogenic Simian Immunodeficiency Virus/Human Immunodeficiency Virus Chimera Is Multigenic

Abstract

One of three full-length infectious molecular clones of SHIV_DH12R, designated SHIV_DH12R-CL-7 and obtained from productively infected rhesus monkey peripheral blood mononuclear cells, directed rapid and irreversible loss of CD4⁺ T cells within 3 weeks of its inoculation into Indian rhesus monkeys. Induction of complete CD4⁺ T-cell depletion by SHIV_DH12R-CL-7 was found to be dependent on inoculum size. The acquisition of this pathogenic phenotype was accompanied by the introduction of 42 amino acid substitutions into multiple genes of parental nonpathogenic SHIV_DH12. Transfer of the entire SHIV_DH12R-CL-7 env gene into the genetic background of nonpathogenic SHIV_DH12 failed to confer the rapid CD4⁺ T-lymphocyte-depleting syndrome; similarly, the substitution of gag plus pol sequences from SIV_smE543 for analogous SIV_mac239 genes in SHIV_DH12R-CL-7 attenuated the pathogenic phenotype. Amino acid changes affecting multiple viral genes are necessary, but insufficient by themselves, to confer the prototypically rapid and irreversible CD4⁺ T-cell-depleting phenotype exhibited by molecularly cloned SHIV_DH12R-CL-7.

Simian immunodeficiency virus (SIV)/human immunodeficiency virus (HIV) chimeras (SHIVs) have proven to be useful reagents for studies of primate lentivirus pathogenesis and vaccine development (2, 3, 10, 21). Because they bear the HIV type 1 (HIV-1) envelope glycoprotein, SHIVs have been valuable for assessing vaccine-induced anti-HIV-1 neutralizing antibodies (2, 3, 5, 18, 24). Furthermore, highly pathogenic SHIVs induce an unusually rapid, complete, and irreversible depletion of CD4⁺ T lymphocytes in rhesus monkeys, thereby providing a readily demonstrable endpoint in vaccine experiments. Some highly pathogenic SHIVs were initially isolated from rhesus monkeys following multiple animal-to-animal passages of nonpathogenic viruses (14, 17). We previously reported that highly pathogenic SHIV_DH12R emerged during a single in vivo passage in a rhesus monkey treated with an anti-human CD8 monoclonal antibody at the time of its primary infection with nonpathogenic SHIV_DH12 (11). Because the intrinsic genetic heterogeneity of available uncloned highly pathogenic SHIV stocks could be contributing to the inconsistent disease phenotypes observed in some vaccine experiments, a molecularly cloned SHIV, capable of replicating to high levels in vitro and in vivo and consistently inducing rapid and complete loss of CD4⁺ T cells in inoculated Indian rhesus monkeys, was constructed.

Cloning strategy.

Lambda phage vectors have previously been used to obtain full-length molecular clones of unintegrated HIV-1 DNA from productively infected cells (1, 6, 22). The identification of a restriction enzyme that cuts the viral DNA a single time, thereby permitting the cloning of one and two long terminal repeat circular DNA molecules produced in newly infected cells is an initial step in this process. Following transfer to plasmids, cloned circularly permuted DNA molecules can be readily converted to linear, correctly oriented forms of retroviral DNA (16, 22).

In preliminary experiments to generate molecular clones of highly pathogenic SHIV_DH12R (11), DNA was prepared from rhesus monkey peripheral blood mononuclear cells (PBMC), infected with the SHIV_DH12R-PS1 derivative (24), and subjected to digestion by several restriction enzymes. Southern blot hybridization analyses revealed that EcoRI converted the circular forms of viral DNA into linear molecules approximately 10 kbp in size, consistent with the presence of a single EcoRI site within unintegrated SHIV_DH12R DNA (data not shown). Accordingly, Hirt-fractionated (9), SHIV_DH12R-PS1-infected monkey PBMC DNA was digested with EcoRI and ligated to similarly cleaved lambda phage EMBL-4 DNA (Stratagene/Biocrest; Cedar Creek, Tex.) as previously described (22). Positive recombinant phage plaques, identified by using the 8.1-kbp HaeII-XhoI fragment (encompassing the gag through the env sequences) from SHIV_DH12 (20) as a probe, were expanded, and the insert was transferred to pBR322. Two long terminal repeat linear forms of SHIV DNA were reconstituted from SalI-EcoRI and EcoRI-NarI subfragments of each clone as previously described (16, 22). Three of the full-length SHIV_DH12R clones (SHIV_DH12R-CL-7, SHIV_DH12R-CL-8, and SHIV_DH12R-CL-10) obtained generated progeny virions following transfection of HeLa cells, each of which exhibited robust and indistinguishable infection kinetics in rhesus monkey PBMC (data not shown). During these spreading infections, the three cloned SHIVs were highly cytopathic, inducing large syncytia similar to those observed with parental uncloned SHIV_DH12R.

Pathogenicity of cloned SHIVs in rhesus monkeys.

In the initial test of pathogenicity, 5 ml of HeLa cell supernatant from each of the three cloned SHIV transfections was subjected to ultracentrifugation (1.5 × 10⁵ × g for 30 min) and the pelleted virions were resuspended in 500 μl of Dulbecco's minimal essential medium. Equal volumes (250 μl) of the suspended virus preparations were combined, and 300 μl of this mixture was inoculated intravenously into two rhesus monkeys. As shown in Fig. 1A and B, animals RhAL30 and RhI030 rapidly lost CD4⁺ T cells, reaching levels of less than 20 cells/μl of plasma by week 3, generated high levels of peak and postpeak plasma viremia, and were euthanized at weeks 8 and 57 because of intractable diarrhea and marked weight loss, respectively. This result indicated that one or more of the three SHIV_DH12R molecular clones conferred the highly pathogenic phenotype. In a follow-up experiment, virus stocks of each clone were prepared in rhesus monkey PBMC as previously described (13) and high doses of each were inoculated individually into three different macaques. Figure 1C shows that only the recipient of SHIV_DH12R-CL-7 (inoculum size, 3 × 10⁴ 50% tissue culture infective doses [TCID₅₀]) experienced the complete depletion of CD4⁺ T lymphocytes (8 CD4⁺ T cells/μl of plasma at week 3) typically observed with parental uncloned SHIV_DH12R (7, 10). The two monkeys inoculated with SHIV_DH12R-CL-8 or SHIV_DH12R-CL-10 (receiving 1.6 × 10⁴ and 2.5 × 10⁴ TCID₅₀, respectively) had marked but transient declines (to 170 and 113 cells/μl of plasma at week 3, respectively) of their CD4⁺ T cells, which gradually returned to preinoculation levels (929 and 1,139 cells/μl of plasma at week 120, respectively). While all three virus-inoculated animals experienced high peak plasma virus loads (>10⁷ RNA copies/ml) at weeks 2 to 3 postinfection, only monkey RhWBJ, which received SHIV_DH12R-CL-7, continued to produce high and sustained postpeak levels of viral RNA (Fig. 1D). Animal RhWBJ was euthanized at week 26 with severe anorexia and marked weight loss.

FIG. 1.

One of three SHIV_DH12R molecular clones induces rapid and complete depletion of CD4⁺ T cells in rhesus monkeys. Rhesus macaques RhAL30 and RhIO30 were inoculated intravenously (300 μl) with a mixture of undiluted supernatants from transfected HeLa cells, and levels of CD4⁺ T lymphocytes (A) and plasma viral RNA (B) were determined as previously described (7, 10, 11, 19). Animals RhWBJ, RhWBV, and RhWBW were individually inoculated with SHIV_DH12R-CL-7 (3 × 10⁴ TCID₅₀), SHIV_DH12R-CL-8 (1.6 × 10⁴ TCID₅₀), and SHIV_DH12R-CL10 (2.5 × 10⁴ TCID₅₀), respectively. The resulting CD4⁺ T-cell levels (C) and plasma viremia (D) were determined.

We previously reported that the rapid and irreversible loss of CD4⁺ T lymphocytes induced by uncloned SHIV_DH12R was dose dependent (7, 12). This unusual phenotype could reflect the effect of a small subpopulation of highly pathogenic virions, present in the uncloned virus stock, which was solely responsible for the rapid and complete loss of CD4⁺ T cells in animals inoculated with larger amounts of virus (e.g., 625 TCID₅₀ or greater [7]). To ascertain whether this dose-dependent phenotype was also an intrinsic property of molecularly cloned SHIV_DH12R-CL-7, four monkeys were inoculated with decreasing amounts of SHIV_DH12R-CL-7. As can be seen in Fig. 2A, the two macaques receiving 30,000 and 5,000 TCID₅₀ of SHIV_DH12R-CL-7 both sustained rapid declines of their CD4⁺ T cells and were euthanized at weeks 25 and 26 because of diarrhea and marked weight loss, a clinical outcome that was accompanied by high plasma virus loads (Fig. 2B). Three other macaques inoculated with 5,000 TCID₅₀ also experienced complete depletion of CD4⁺ T lymphocytes within weeks of inoculation and were sacrificed between weeks 13 and 25 because of their deteriorating clinical condition (data not shown). The systemic depletion of CD4⁺ T lymphocytes by SHIV_DH12R-CL-7 was similar histopathologically to that previously described for the uncloned parental SHIV_DH12R virus stock (10). CD4⁺ T-cell loss from peripheral and internal (mesenteric and colonic) lymph nodes began by day 10 postinoculation and predominantly affected lymphocytes in paracortical regions; by day 21 postinfection, an 80 to 90% reduction of this T-cell subset was observed in multiple lymph node specimens. Virus-producing cells peaked on day 10 in the T-cell-rich regions of lymph nodes and on day 14 in the thymic medulla, coincident with the observed slightly delayed loss of CD4⁺ T cells from the thymus. Postmortem examination revealed diffuse lymphoid depletion, marked thymic atrophy, and evidence of cryptosporidiosis and disseminated candidiasis in several of the SHIV_DH12R-CL-7-infected monkeys.

FIG. 2.

The rapid and complete loss of rhesus monkey CD4⁺ T cells induced by SHIV_DH12R-CL-7 is inoculum size dependent. Monkeys RhWBJ, RhR46, RhWBP, and RhBC66 were inoculated intravenously with the indicated amounts of SHIV_DH12R-CL-7, and the levels of CD4⁺ T lymphocytes (A) and plasma viral RNA (B) were determined. Monkeys RhBH30, RhWBV, and Rh429 were inoculated intravenously with the indicated amounts of SHIV_DH12R-CL-8, and the levels of CD4⁺ T lymphocytes (C) and plasma viral RNA (D) were determined.

Rapid and severe loss of CD4⁺ T cells (51 cells/μl of plasma at week 3) also occurred in the animal (RhWBP) inoculated with 500 TCID₅₀ of SHIV_DH12R-CL-7 (Fig. 2A), but in this case, the depletion did not reach baseline levels (<1% of CD4⁺ T lymphocytes, <20 cells/mm³) until week 63 postinfection. Monkey WBP had to be euthanized at week 79 because of anorexia and intractable diarrhea. Finally, the inoculation of macaque RhBC66 with 100 TCID₅₀ of SHIV_DH12R-CL-7 resulted in only a modest and transient CD4⁺ T-cell decline (557 cells/μl of plasma at week 5), with a return to preinfection levels by week 40 (1,070 cells/μl of plasma) (Fig. 2A). This moderate CD4⁺ T-lymphocyte depletion was nonetheless associated with a high peak plasma virus load (6.3 × 10⁶ viral RNA copies/ml) but low to undetectable postpeak viremia (Fig. 2B). Thus, as was the case for uncloned SHIV_DH12R, the clinical outcome of SHIV_DH12R-CL-7 infection was also dependent on inoculum size, indicating that this was an intrinsic property of both the uncloned and cloned viruses.

To be certain that the partial and transient depletion of CD4⁺ T lymphocytes induced by SHIV_DH12R-CL-8 in monkey RHWBV (Fig. 1C) was an intrinsic property of this clone and did not represent an idiosyncratic animal-specific response to virus, two additional animals were inoculated with high doses of SHIV_DH12R-CL-8. In both cases, marked but incomplete loss of CD4⁺ T cells occurred during the first weeks of infection (to 270 cells/μl [RhBH30] and 170 cells/μl [Rh429] at week 3 [Fig. 2C]); this depletion was associated with prompt control of plasma viremia (Fig. 2D). The patterns of CD4⁺ T-lymphocyte depletion and restoration in the three SHIV_DH12R-CL-8-infected monkeys were similar to one another and clearly distinguishable from that observed with the animals inoculated with SHIV_DH12R-CL-7. It is also worth noting that SHIV_DH12R-CL-7 and SHIV_DH12R-CL-8 exhibited indistinguishable infection kinetics and cytopathicity following infection of cultured PBMC (data not shown).

Determinants of SHIV pathogenicity.

Complete nucleotide sequencing of SHIV_DH12R-CL-7 revealed that 42 amino acid changes, relative to the starting sequence of nonpathogenic SHIV_DH12 (previously designated SHIV_MD14YE [20]), had accompanied the acquisition of the highly pathogenic phenotype (Fig. 3, top). Although these changes were distributed throughout the viral genome, they primarily affected the env and nef genes. On the basis of increased chemokine receptor binding affinity, membrane fusion capacity, and/or neutralization resistance, earlier analyses of the molecularly cloned SHIV_KB9 derivative of SHIV_89.6P (15) and the molecularly cloned SHIV_HXBc2P-3.2 derivative of SHIV_KU-1 (4) concluded that changes in the HIV-1 envelope glycoprotein component of highly pathogenic SHIVs were the primary determinant for inducing the rapid loss of CD4⁺ T lymphocytes in vivo. To ascertain whether the 17 amino acid substitutions in SHIV_DH12R-CL-7 Env, by themselves, could confer the rapid and irreversible CD4⁺ T-cell-depleting phenotype, the 2.7-kbp HindIII fragment (positions 6333 to 9079) encoding the entire env gene of SHIV_DH12R-CL-7 was transferred into the background of original nonpathogenic strain SHIV_DH12. A stock of the resultant chimera (SHIV_{DH12[gp160-CL-7]}) was prepared in rhesus monkey PBMC and used to inoculate macaque RhWBF with 75,000 TCID₅₀. As shown in Fig. 4A, SHIV_{DH12[gp160-CL-7]} did not induce the signature rapid loss of CD4⁺ T lymphocytes, replicating to only modest levels (peak plasma viremia of 7.5 × 10⁵ RNA copies/ml) during the initial weeks of infection (Fig. 4B). The inability of the SHIV_DH12R-CL-7 env gene to confer the unique pathogenic phenotype was also consistent with nucleotide sequence analyses of the pathogenic SHIV_DH12R-CL-7 and nonpathogenic SHIV_DH12R-CL-8 genomes, which indicated that both viruses have identical gp120s at both the nucleotide and amino acid levels (Fig. 3, bottom). More importantly, the comparison of the two viruses revealed that the distinctive CD4⁺ T-lymphocyte-depleting properties of SHIV_DH12R-CL-7 versus SHIV_DH12R-CL-8 are due to sequences encoding Nef, gp41, and Vpr; no changes affecting cis-acting elements such as the Rev responsive element, transcriptional regulatory sequences, or packaging signals were noted.

FIG. 3.

Amino acid changes associated with the acquisition of a highly pathogenic phenotype. The 42 amino acid differences between nonpathogenic parental SHIV_DH12 and highly pathogenic SHIV_DH12R-CL-7 (top) and between nonpathogenic SHIV_DH12R-CL-8 and highly pathogenic SHIV_DH12R-CL-7 (bottom) are shown. Note that some of the differences between SHIV_DH12R-CL-8 and SHIV_DH12R-CL-7 are not listed in the top panel because they represent amino acid substitutions present in SHIV_DH12R-CL-8 and not in SHIV_DH12R-CL-7.

FIG. 4.

Altered gag, pol, and env sequences in SHIV_DH12R-CL-7 are necessary but not sufficient to confer the highly pathogenic phenotype. Macaque RhWBF was inoculated intravenously with SHIV_{DH12[gp160-CL-7]} (75,000 TCID₅₀), in which only the env gene from SHIV_DH12R-CL-7 was inserted into nonpathogenic SHIV_DH12 (A and B). Animal Rh8350 was inoculated with SHIV_{DH12R-E543(GPV)} (16,000 TCID₅₀), in which the 5-kbp NarI-BstBI fragment bearing gag, pol, and vif sequences from SIV_smE543 was introduced into SHIV_DH12R-CL-7 (C and D). Plasma viremia and CD4⁺ T-cell levels were determined as indicated.

The amino acid comparison of SHIV_DH12R-CL-7 with its nonpathogenic parent, SHIV_DH12, shown in Fig. 3 (top) indicates that the acquisition of a more aggressive pathogenic phenotype was accompanied by 10 amino acid substitutions mapping to the SIV_mac239 gag and pol genes present in SHIV_DH12R-CL-7. The contribution of these SIV changes was evaluated by substituting the 5,041-bp NarI-BstBI fragment, carrying gag, pol, and vif sequences, from pathogenic SIV_smE543 (8) for analogous SIV_mac239 sequences in SHIV_DH12R-CL-7. The molecular clone, designated SHIV_{DH12R-E543(GPV)}, was transfected into HeLa cells, and the culture supernatant was used to prepare a virus stock in rhesus monkey PBMC. Despite receiving 16,000 TCID₅₀ of SHIV_{DH12R-E543(GPV)} intravenously, monkey Rh8350 experienced only a modest loss of CD4⁺ T cells (248 cells/μl of plasma at week 2) and a relatively low peak plasma viremia (5 × 10⁵ viral RNA copies/ml) (Fig. 4C and D). This attenuated in vivo response very likely reflects differences between the gag and pol genes of SIV_smE543 and SIV_mac239, as well as differences between SIV_smE543 and the 10 amino acid changes introduced into the gag and pol sequences attending the evolution of SHIV_DH12R-CL-7 from its nonpathogenic SHIV_DH12 parent. Taken together, these results indicate that changes affecting gag-pol and env are necessary, but insufficient by themselves, to confer the prototypically rapid and irreversible disease phenotype exhibited by SHIV_DH12R-CL-7.

SHIV_DH12R-CL-7 uses CXCR4 for entry into rhesus monkey PBMC.

The env gene of nonpathogenic SHIV_DH12 was derived from HIV-1_DH12 (19), previously reported to use CXCR4 and CCR5 for entry into the human glioblastoma cell line U87MG expressing CD4 (27). To assess the chemokine receptor used by SHIV_DH12R-CL-7 for replication in rhesus monkey PBMC, cells were infected in the presence of small-molecule coreceptor-targeted inhibitors specific for CCR5 or CXCR4. The production of progeny virus was measured as reverse transcriptase activity released into the medium on day 5 postinfection. As shown in Fig. 5A, infection of monkey PBMC by SHIV_DH12R-CL-7 was blocked by the CXCR4 inhibitor AMD3100 and not by two CCR5 inhibitors (TAK-779 and AD-101). The opposite result was obtained with SIV_mac239, which was blocked by the two CCR5 inhibitors and not by AMD3100 (Fig. 5B). The latter result is in agreement with a previous report showing that CCR5 is the coreceptor used by SIV_mac239 for infection of macaque PBMC (26). SHIV_DH12R-CL-7 also appears to be exclusively T cell tropic since it was unable to establish spreading infections in alveolar macrophage (13).

FIG. 5.

Coreceptor usage of SHIV_DH12R-CL-7 and SIV_mac239 for entry into macaque PBMC. SHIV_DH12R-CL-7 and SIV_mac239 were spinoculated onto rhesus PBMC in the presence of the indicated small-molecule coreceptor inhibitors. Rhesus monkey PBMC (5 × 10⁴ cells) were dispensed into 96-well round-bottom plates. The inhibitor concentrations used were 0.05, 0.1, 0.5, 1.0, 5.0, and 10 μM. Reverse transcriptase activity released into the medium on day 5 postinfection was determined in the absence (dashed line) or presence of inhibitor.

Conclusions.

A molecular clone, designated SHIV_DH12R-CL-7, has been obtained that consistently causes rapid, complete, and irreversible loss of CD4⁺ T lymphocytes. This disease phenotype has been previously reported for three independently derived uncloned SHIVs (11, 14, 17) and is applicable to only one of the three full-length infectious SHIV_DH12R clones obtained in this study. The acquisition of these unusual pathogenic characteristics appears to be multigenic: 42 amino acid substitutions, relative to the starting nonpathogenic SHIV_DH12 clone (20), distributed among several viral genes, were present in SHIV_DH12R-CL-7. In contrast to studies of the envelope glycoproteins associated with molecularly cloned SHIV_KB9 and SHIV_HXBc2P-3.2, which were reported to be the principal determinants inducing CD4⁺ T-lymphocyte depletion (4, 15), the entire env gene of SHIV_DH12R-CL-7, by itself, failed to confer the rapid and irreversible CD4⁺ T-cell-depleting properties following its insertion into the genome of nonpathogenic parental strain SHIV_DH12. Similarly, substitution of gag-pol sequences from pathogenic SIV_E543 could not replace analogous SIV_mac239 genes in SHIV_DH12R-CL-7. This result implies that the 10 amino acid substitutions introduced into the SIV_mac239 gag and pol genes (Fig. 3) also contributed to the rapid CD4⁺ T-lymphocyte-depleting phenotype. In both cases, peak levels of plasma viremia were reduced compared to those caused by SHIV_DH12R-CL-7 and the capacity of each virus to induce CD4⁺ T-cell depletion was minimal (Fig. 4). In the studies described here, our working definition of SHIV pathogenicity was rapid, irreversible, and complete elimination of CD4⁺ T lymphocytes coupled with high and sustained levels of viral RNA in plasma. SHIV_DH12R-CL-7 was the only cloned SHIV that fulfilled these criteria.

The mechanism(s) underlying the rapid elimination of CD4⁺ T lymphocytes and induction of immunodeficiency by pathogenic SHIVs remains unknown, although the increased fusogenicity of the envelope glycoprotein and the infection of a substantial fraction of CD4⁺ T cells in lymphoid tissue during the first weeks postinoculation have been proposed to explain the unusual disease phenotype (4, 10, 15). Because the fraction of CD4⁺ T cells in blood and lymphoid tissues expressing CXCR4 is very large (>80%) (23, 25) (Y. Nishimura and M. Martin, unpublished data), complete and systemic elimination of this T-lymphocyte subset by X4-tropic SHIV_DH12R-CL-7 could simply reflect the targeting and unrelenting depletion of these cells. It is also possible that a threshold level of systemic virus production must be reached during the first week of infection to cause the overwhelming and systemic killing of virtually all CD4⁺ T-lymphocyte subsets in lymphoid tissues. A delay of only a few days in exceeding such a threshold, such as that due to infections initiated with low-dose virus inocula, may permit the development of immune responses capable of controlling virus replication. However, monkeys inoculated with large amounts of SHIV_DH12R-CL-8 generated peak levels of plasma viremia that were similar to those seen in animals infected with SHIV_DH12R-CL-7 (Fig. 2B and D), yet the CD4⁺ T-cell depletion in SHIV_DH12R-CL-8-inoculated macaques was incomplete and transient (compare Fig. 2A and C). Assuming that the concentration of virus particles circulating in the blood is a reflection of SHIV replication systemically, the level of progeny virion production, per se, during the acute infection cannot be the principal determinant causing the unrelenting and complete elimination of CD4⁺ T lymphocytes. Since SHIV_DH12R-CL-7 and SHIV_DH12R-CL-8 differ by only nine amino acids, it is likely that one or more of these residues contribute to the signature SHIV disease. Chimeric SHIVs containing individual amino acid substitutions or combinations of these amino acid substitutions are currently being constructed for inoculation into rhesus monkeys to ascertain which viral gene(s) confers the rapid CD4⁺ T-cell-depleting phenotype.