High frequencies of resting CD4⁺ T cells containing integrated viral DNA are found in rhesus macaques during acute lentivirus infections

Abstract

We and others have reported that the vast majority of virus-producing CD4⁺ T cells during the acute infection of rhesus macaques with simian immunodeficiency virus (SIV) or CXCR4 (X4)-using simian/human immunodeficiency viruses (SHIVs) exhibited a nonactivated phenotype. These findings have been extended to show that resting CD4⁺ T lymphocytes collected from SIV- or X4-SHIV-infected animals during the first 10 days of infection continue to release virus ex vivo. Furthermore, we observed high frequencies of integrated viral DNA (up to 5.1 × 10⁴ DNA copies per 10⁵ cells) in circulating resting CD4⁺ T cells during the first 10 days of the infection. Integration of SIV DNA was detected only in memory CD4⁺ T cells and SHIVs preferentially integrated into resting naïve CD4⁺ T cells. Taken together, these results show that during the acute infection large numbers of resting CD4⁺ T cells carry integrated nonhuman primate lentiviral DNA and are the major source of progeny virions irrespective of coreceptor usage. Prompt and sustained interventions are therefore required to block the rapid systemic dissemination of virus and prevent an otherwise fatal clinical outcome.

During the first 15 years of the AIDS epidemic, the prevailing dogma was that HIV and simian immunodeficiency virus (SIV) were unable to replicate in quiescent CD4⁺ T lymphocytes (1, 2). This failure to infect nonactivated T cells in culture was attributed to multiple mechanisms including inefficient reverse transcription (RT) (1, 3–5), impaired nuclear import of reverse transcripts (6, 7), and low intracellular dNTP pools (8). It was also believed that HIV was unable to integrate into the genome of resting CD4⁺ T cells (9). Because unintegrated DNA has been reported to be inefficiently transcribed (10–12), impaired integration could also contribute to the failure of lymphocytes in the G_0/1a stage of the cell cycle to support productive infections (13).

Although these results have been confirmed in numerous laboratories, studies conducted during the past decade have shown that resting T cells, recovered from human lymphoid tissue explants maintained ex vivo, release substantial quantities of progeny virions (14, 15). Another report demonstrated that resting CD4⁺ T lymphocytes, in specimens obtained from HIV-infected individuals or acutely SIV-infected macaques, are producing viral RNA (16). We reported that 32–88% of circulating naïve CD4⁺ T lymphocytes from rhesus monkeys infected with CXCR4-tropic (X4-tropic) simian/human immunodeficiency viruses (SHIVs) were productively infected by day 10 postinfection (PI) (17). Immunophenotyping by flow cytometry and combined in situ hybridization/immunohistochemistry revealed that the vast majority of the virus-producing CD4⁺ T lymphocytes collected from these SHIV-infected animals were not expressing activation markers.

It could be argued that the prodigious levels of virus production by quiescent naïve cells measured by endpoint dilution/cocultivation reflected the activating effects induced by the monoclonal antibodies used for positive cell sorting and/or the subsequent incubation with MT4 cells (17). Accordingly, negatively-selected resting CD4⁺ T lymphocytes were purified from SIV-infected (day 7 PI) and SHIV-infected (day 10 PI) rhesus monkeys and cultivated in the presence of autologous serum and the absence of IL-2 for 4 days. Infectious virus was, in fact, detected in the supernatant medium from each ex vivo culture. In a separate series of experiments, the integration status of viral DNA in resting naïve and memory CD4⁺ T cells, recovered from SIV- and SHIV-infected macaques on days 7 and 10 PI, respectively, was determined by limiting dilution Alu-LTR PCR. High frequencies of integrated SIV DNA were detected in memory but not naïve CD4⁺ T cells, in agreement with the differential expression of the CCR5 coreceptor in these 2 lymphocyte subsets (18, 19). In animals inoculated with X4-tropic SHIV, 13–51% of the circulating naïve CD4⁺ T cells contained integrated DNA, assuming each cell carries a single integrated provirus. These results indicate that during the acute infection, integration of viral DNA occurs in a prototypical resting cell.

Results

In earlier experiments evaluating virus production during the acute infection of 3 rhesus monkeys with the highly-pathogenic X4-tropic SHIV_DH12R, we reported that resting naïve CD4⁺ T cells were massively infected and, as shown in Fig. 1, were the major source of infectious virus circulating in the blood on day 10 PI (17). In that study, activated CD8⁺ T lymphocytes were observed in the blood (CD69⁺, HLA-DR⁺, and Ki-67⁺) and lymph nodes (Ki-67⁺) by week 2 PI, but the activation status of CD4⁺ T cells recovered from these sites remained unchanged from their preinoculation resting state. Confocal microscopy of day 10 lymph node specimens, using SIV riboprobes and the anti-Ki-67 mAb, revealed that >96% of SHIV-producing cells did not express Ki-67. Taken together, these results indicated that the vast majority of virus-producing cells during the acute in vivo infection exhibited a nonactivated phenotype.

Fig. 1.

Naive CD4⁺ T cell are the major source of infectious virus circulating in the blood during the acute infection of rhesus monkeys with SHIV_DH12R. The frequencies of productively-infected naïve or memory cells collected on day 10 PI from 3 SHIV_DH12R-infected animals were determined by cocultivation with MT4 cells for 14 days as described (17).

As a follow-up to this initial study, we wanted to answer 2 questions: (i) would virus production by resting CD4⁺ T lymphocytes, after their isolation from acutely infected animals, continue ex vivo and (ii) what was the integration status of viral DNA in resting virus-producing CD4⁺ T cells in vivo? Accordingly, individual rhesus monkeys were inoculated intravenously with 1 × 10⁴ TCID₅₀ SIVmac239 (macaque CN30) or 1 × 10⁵ TCID₅₀ SHIV_DH12R (macaque WGG) and the levels of plasma viremia and circulating CD4⁺ T cells were determined during the first 10 days of infection (Fig. 2A). The SIV-infected animal experienced a more rapid increase of its plasma virus load compared with the SHIV-infected macaque and both monkeys sustained partial depletions of their peripheral blood CD4⁺ T cells. Peripheral blood mononuclear cells (PBMCs) were prepared from both monkeys on days 7 and 10 PI and a portion of each PBMC sample was sorted by FACS for quantitative cell-associated viral DNA content in naive (CD95^lowCD28^high) or memory (CD95^highCD28^{low and high}) CD4⁺ T lymphocytes. As shown in Fig. 2B, high levels (3.1 or 3.8 × 10⁴DNA copies per 10⁵ cells) of memory CD4⁺ T cells from the SIV-infected macaque contained viral DNA on days 7 and 10 PI, whereas the frequency of infected naïve cells was significantly lower. This difference is consistent with the expression levels of the CCR5 coreceptor used by SIV in these 2 CD4⁺ T lymphocyte subsets. In contrast, SHIV_DH12R DNA was detected in both naïve and memory CD4⁺ T cells, reaching levels of 1.6 × 10⁴ and 1.7 × 10⁴ DNA copies per 10⁵ cells, respectively, on day 10 PI, consistent with the surface expression of CXCR4 on both CD4⁺ T cell subsets (20).

Fig. 2.

Peripheral blood CD4⁺ T cell profiles and viral nucleic acid levels during the acute SIV and SHIV infections of macaques. (A) Circulating CD4⁺ T cell numbers and plasma viral RNA loads were measured in a SIV (CN30)-infected and a SHIV_DH12R (WGG)-infected rhesus monkey at the indicated times. (B) PBMC samples, collected on days 7 and 10 PI from macaques CN30 and WGG, were stained with anti-CD3, CD4, CD28, and CD95 mAbs to distinguish naïve from memory CD4⁺ T cells and sorted by FACS for determinations of the cell-associated viral DNA copies per 10⁵ cells by DNA PCR.

Spontaneous Virus Production by Resting CD4⁺ T Cells ex Vivo.

To preserve the nonactivated status of CD4⁺ T cells during purification, PBMC samples, collected on day 7 from SIV-inoculated macaque CN30 and on day 10 from SHIV-inoculated macaque WGG, were initially stained with non-CD4⁺ T cell markers (CD8, CD20, CD14, and CD16) and markers of activation (CD25, CD69, and HLA-DR). These collection times were selected because: (i) circulating memory CD4⁺ T lymphocytes had declined to low levels in the SIV-infected monkey (CN30) by day 10 PI; and (ii) larger numbers of SHIV_DH12R-infected cells were present in animal WGG on day 10 compared with day 7 PI. Lymphocytes, negative for non-CD4 and activation markers, were then sorted by FACS. The purity of the resulting CD4⁺ T cell populations from both animals was >96%; these cells expressed minimal levels of CD25, CD69, or HLA-DR (Fig. 3A).

Fig. 3.

Resting CD4⁺ T cells from acutely SIV and SHIV-infected macaques continue to release progeny virions ex vivo. (A) Nonactivated CD4⁺ T cells were purified by negative FACS sorting from PBMCs collected on day 7 from macaque CN30 (SIV_mac239) or day 10 from macaque WGG (SHIV_DH12R) as described in Methods. The FACS profiles of CD4⁺ HLA-DR⁺ cells from each monkey are shown. (B) Purified resting CD4⁺ T cells were treated with Pronase to eliminate any surface-bound virions and cultured for 4 days in the absence of any stimulation in RPMI medium 1640 supplemented with 15% autologous serum. Culture supernatants were collected daily, filtered, and spinoculated on MT-4 T cells, in duplicate, to amplify low levels of released progeny virions. Virus production was assessed by measuring the 32P_RT activity released into the medium of duplicate MT-4 cultures on day 14 PI by autoradiography.

The negatively-sorted resting CD4⁺ T cell preparation from each animal was cultured in duplicate for 4 days in the absence of any stimulation, including exogenous IL-2, as described in Methods. The production of progeny virus was monitored daily by measuring RT activity released into the medium. No RT activity was measured in these culture supernatants. The detection limit of the ³²P-RT-assay used (21) ranged from 2.5 × 10³ to 8 × 10³ TCID₅₀/mL, based on RT measurements of SIV and SHIV stocks of known infectious titer. To amplify any infectious virus released into the medium, filtered supernatants, collected daily (days 1–4) from the RT-negative duplicate cultures of resting SIV- or SHIV-infected CD4⁺ T cells, were spinoculated onto MT-4 T cells, and the duplicate cultures of MT-4 T cells were maintained for an additional 14 days. As shown in Fig. 3B, RT activity was detected as early as day 1 (1 of 2 wells) in the supernatants from each of the original nonactivated CD4⁺ T cell cultures. Both duplicate wells were positive for SIV and SHIV production in supernatants collected on days 3–4. These results indicate that resting CD4⁺ T cells, collected from macaques during acute SIV and SHIV infections, continued to produce and release progeny virions during 4 days of ex vivo cultivation.

Integration Status of Viral DNA in CD4⁺ T Lymphocytes During Acute SIV and SHIV Infections.

Because of the long-standing belief that integration of reverse transcripts did not occur in quiescent cells, we next investigated whether integrated DNA could be detected during the primary infection. A modified Alu-LTR PCR approach, using pairs of outward-facing PCR primers binding to conserved regions of human and rhesus macaque AluDNA sequences, in conjunction with a cloned cell line containing a single copy of integrated SIV DNA, were used to analyze genomic DNA prepared from sorted and endpoint-diluted naïve and memory CD4⁺ T cells from acutely SIV- and SHIV-infected animals.

The 3D8 cell line, carrying 1 copy of integrated SIV DNA, was derived from chronically-infected CEM×174 cells, 8 weeks after infection with SIVmac316, as described in SI Text. Cells carrying viral DNA were initially identified by DNA PCR and then examined by Southern blot analysis to ascertain SIV proviral DNA copy number, using EcoRI, BamHI, and SacI, which cleave SIVmac316 DNA 0, 1, and 2 times, respectively. An analysis of 3D8 cellular DNA, carried out in triplicate and shown in Fig. S1, is consistent with the presence of a single integrated copy of SIV DNA in this cloned line.

Limiting dilution Alu-LTR PCR, performed on specimens recovered from acutely-infected macaques, was used to quantitate the frequency of integrated viral DNA. The values obtained from these samples were then compared with those derived from 3D8 cells, which were contemporaneously examined. A representative analysis of 3D8 cellular DNA performed in triplicate is shown in Fig. 4A. The mean number of cells required to detect the single copy of SIV DNA in this cell line was 3.3, based on several independent experiments. This level of sensitivity is similar to that reported for the ACH-2 cell line, which also contains a single copy of integrated HIV-1 DNA (22).

Fig. 4.

Quantitation of integrated viral DNA in naïve or memory CD4⁺ T cells collected from acutely SIV or SHIV-infected macaques. (A) Genomic DNA from 3D8 cells was serially diluted 2-fold, in triplicate, starting with 500 cell DNA equivalents and analyzed by Alu-LTR PCR. (B) Genomic DNAs, purified from sorted naïve or memory CD4⁺ T cells from SHIV_DH12R-infected macaque 95P005 on day 10 PI, were serially diluted, and integrated proviral DNA was measured by Alu-LTR PCR. (C) The frequencies of naïve or memory CD4⁺ T cells carrying integrated SIV or SHIV DNA were determined by Alu-LTR PCR and normalized to the values obtained with the 3D8 cell line.

Genomic DNAs, prepared from positively-sorted naïve or memory CD4⁺ T cells on day 7 PI from the SIV (CN30), day 10 PI from the SHIV (WGG)-infected macaque, or day 10 PI from the 3 previously-studied SHIV-infected animals (94E057, 95P005, and H482) shown in Fig. 1, were endpoint-diluted and evaluated by Alu-LTR PCR. A representative analysis of the integrated SHIV DNA present in CD4⁺ T lymphocytes from monkey 95P005 is shown in Fig. 4B. Because a mean of 3.3 3D8 cells was required to detect the single copy of SIV DNA present per cell, the values obtained from the naïve or memory cells, recovered from SHIV/SIV-infected animals, were normalized accordingly. Based on this determination, the number of naïve CD4⁺ T cells needed to amplify a single copy of integrated DNA from macaque 95P005 was 2.0 cells; the number of memory CD4⁺ T cells required was higher (7.9 cells). As shown in Fig. 4C and Table S1, 2.0–7.9 naïve CD4⁺ T lymphocytes collected from the 4 SHIV-infected monkeys on day 10 PI were required to detect integrated proviral DNA. Thus, the frequency of integrated SHIV DNA in circulating naive CD4 T cells was surprisingly high, ranging from 1.27 to 5.08 × 10⁴ DNA copies per 10⁵ cells, with a mean of 2.43 × 10⁴ DNA copies per 10⁵ cells. The numbers of memory CD4⁺ T cells from these 4 animals needed to amplify a single copy of integrated viral DNA were higher (7.9–15.8 cells). The frequency of integrated SHIV DNA in memory CD4⁺ T cells ranged from 6.34 × 10³ to 1.27 × 10⁴ DNA copies per 10⁵ cells with a mean 1.01 × 10⁴ DNA copies per 10⁵ cells.

In the single SIV-infected animal (CN30) evaluated, the number of memory CD4⁺ T cell needed to amplify a single copy of integrated DNA and the frequency of integrated viral DNA were 15.8 cells and 6.34 × 10³ DNA copies per 10⁵ cells, respectively, for samples collected on day 7 PI (Fig. 4C and Table S1). As expected from its coreceptor usage, integrated SIV DNA was undetectable in naïve CD4⁺ T cells (even when >500 cells were analyzed per reaction).

The total cell-associated viral DNA in naïve or memory CD4⁺ T cells recovered from each animal was measured by quantitative DNA PCR using SIV gag primers (Table 1). Although the different PCR assay systems used to quantitate integrated and total SIV and SHIV DNAs in the CD4⁺ T cells subsets make direct comparisons of viral DNA frequencies problematic, higher levels of total viral DNA were generally measured. Nonetheless, the frequencies of viral DNA copies per 10⁵ cells measured were very high in both assays.

Table 1.

Frequencies of integrated and total viral DNA in naïve or memory CD4⁺ T cells in SHIV-infected (day 10) or SIV-infected (day 7) animals

Animal ID	CD4 subset	Integrated copies/105 cells	Total DNA copies/105 cells	Integrated/total DNA ratio
94E057 (SHIV)	Naïve	1.27 × 104	7.25 × 104	0.17
	Memory	6.34 × 103	2.35 × 104	0.27
95P005 (SHIV)	Naïve	5.08 × 104	2.84 × 105	0.18
	Memory	1.27 × 104	1.07 × 105	0.12
H482 (SHIV)	Naïve	1.27 × 104	6.44 × 104	0.20
	Memory	1.06 × 104	5.67 × 104	0.19
WGG (SHIV)	Naïve	2.11 × 104	1.57 × 104	1.35
	Memory	1.06 × 104	1.71 × 104	0.62
CN30 (SIV)	Naïve	*	1.51 × 103	*
	Memory	6.34 × 103	3.13 × 104	0.20

*Below detection limit.

Discussion

Activated PBMC cultures or T cell leukemia cell lines have been used for >20 years to study HIV replication and spreading infections in vitro. It has been assumed by many that the replicative properties of diverse HIV-1 strains, measured in these in vitro systems, are reflective of virus infectivity in vivo. However, most CD4⁺ T lymphocytes in vivo are not activated and those that are may not reach the levels of activation achievable in vitro. As noted earlier, resting PBMC cultures or purified quiescent CD4⁺ T lymphocyte preparations are resistant to HIV-1 infection in vitro. This resistance has been ascribed to a generalized impairment/inefficiency of several critical steps in the virus life cycle including entry, RT, and nuclear import. Although it was once thought that the integration reaction is blocked in nonactivated CD4⁺ T cells, at least 3 recent studies (23–25) report that HIV-1 DNA integrates into the chromosomal DNA of resting cells, even when the size of the virus inoculum is low. Nonetheless, spontaneous release of progeny virions in these systems has never been observed, but can be induced after treatment with anti-CD3 plus anti-CD28 mAbs (24).

We and others have reported that during the acute infection of rhesus macaques with either SIV or X4-SHIVs, most virus-infected cells were immunophenotypically resting memory and naïve CD4⁺ T lymphocytes, respectively (16, 17, 26). We have extended this observation in the current study and found that nonactivated virus-infected cells, recovered from acutely-infected animals, continue to release low levels of infectious progeny virions ex vivo without additional stimulation. In this regard, intriguing quantitative differences have been reported when HIV-1 RT and integration reactions are evaluated during ex vivo infections of CD4⁺ T cells exhibiting a continuum of activation states. For example, when viral DNA generated by RT is measured in ex vivo-activated versus freshly-purified resting T cells, 30- to 100-fold more DNA copies per cell are detected in the former (24, 25). In contrast, the HIV-1 DNA content in “endogenously” activated CD4⁺ T cells (negatively-sorted CD4⁺ T lymphocytes expressing HLA-DR and CD69) was only 2-fold higher than that present in resting T cells. A similar hierarchy of relative integration frequencies (1:5:50) was observed for resting, endogenously-activated, and ex vivo-activated CD4⁺ T cells, respectively. These results suggest that virus infectivity in vivo, both in resting and activated CD4⁺ T cells, may be far less robust than that measured in vitro.

HIV-infected resting CD4⁺ T cells have been identified during both the acute and chronic phases of infection in vivo. Some reports (27, 28) have described the presence of integrated HIV-1 DNA and the recovery of virus from CD45RA⁺ naïve CD4⁺ T lymphocytes from chronically-infected individuals, particularly after coreceptor switching. In addition, latently HIV-1-infected, resting memory CD4⁺ T cells have been detected in individuals receiving highly active antiretroviral therapy with extremely low levels of viremia (29–31). When resting memory CD4⁺ T lymphocytes from viremic and nonviremic, latently-infected individuals were purified and cultured ex vivo in the absence of activating stimuli, only cells from viremic persons spontaneously released progeny virions (32).

In contrast to the chronic phase of the HIV-1 infection in which the frequency of latently-infected resting memory CD4⁺ T cells with replication competent-integrated provirus is exceedingly low (33), high levels of virus-producing resting cells have been reported during the acute infection. One study (16), using in situ hybridization in combination with staining by antibodies to activation and cell cycle markers, reported that 50% of the virus-infected cells in lymph node specimens from recently-infected individuals were HLA-DR⁻ and KI-67⁻, the resting memory subset. In the more tractable SIV/macaque system, a majority of virus-positive cells in lymph node biopsies were shown to be HLA-DR⁻ and KI-67⁻ on day 12 after intravaginal inoculation. Not unexpectedly, the intracellular levels of viral RNA measured in these experiments were 4- to 6-fold higher in activated cells present in the same microscopic fields. In a follow-up study using the same approach, >90% of SIV RNA-positive memory CD4⁺ T cells in the colon were found to be negative for CD69, CD25, and Ki-67 on day 10 PI (26).

In this study we have shown that both SIV and SHIVs are able to infect resting CD4⁺ T lymphocytes during the acute infection of rhesus monkeys and integrate DNA copies of their genomes into host cell chromosomal DNA. The observed high frequencies of resting CD4⁺ T cells carrying integrated viral DNA on day 7 PI (SIV) and day 10 PI (SHIV) is consistent with the reported high frequencies of virus-infected cells using other approaches (17, 34). In the case of acutely SHIV-infected macaques, 13–51% of the circulating naïve CD4⁺ T cells contained integrated viral DNA, which is in good agreement with the reported 30–90% frequency of naïve CD4⁺ T cells releasing progeny virions (17). The results of the integration assays also reflected the differential levels of coreceptor expression in the CD4⁺ T lymphocyte subsets targeted by SIV and X4-SHIVs (20). Integration of SIV DNA was observed only in memory CD4⁺ T cells and the X4 SHIVs preferentially integrated into naïve versus memory CD4⁺ T cells. Thus, during the acute infection of macaques, resting CD4⁺ T cells are the major targets of nonhuman primate lentiviruses, irrespective of their coreceptor usage.

Based on available evidence, the very low levels of virus production, observed in ex vivo cultures of resting CD4⁺ T cells from acutely SIV- and SHIV-infected monkeys (Fig. 3B), are difficult to reconcile with the high levels of plasma viremia (≈10⁷ RNA copies per mL) contemporaneously measured in these animals. A plausible explanation of this apparent paradox is that: (i) the high frequency of circulating infected, nonactivated CD4⁺ T lymphocytes is reflective of the extraordinarily large number of similar productively infected cells in lymphoid tissues and/or effector sites; (ii) the latter population produces small amounts of virus per cell; and (iii) the net effect of the slow release of small quantities of progeny virions by an enormous fraction of the total CD4⁺ T cell population would result in high SIV or SHIV virus loads, as suggested (26). In this regard, it has recently been reported that SIV-infected resting cells may generate 70–93% of peak virus production, based on Bayesian statistical analyses (35).

A related and currently-unresolved issue concerns the mechanism responsible for the prodigiously rapid rate of virus spread throughout a sizable fraction of total body CD4⁺ T cells during the initial weeks of the infection. It would seem unlikely that cell-free virus transmission alone could account for the rapid increase of virus-infected cells in vivo. The high concentrations of tightly-packed juxtaposed infected and uninfected resting CD4⁺ T cells in vivo suggest alternative mechanisms for virus dissemination such as cell-to-cell spread, dendritic cell-mediated transfer, and cytokine-augmented dispersal of progeny virions all may contribute to this process. The extremely-rapid systemic spread exhibited by SIV and SHIVs in macaques and presumably, by HIV-1 in humans, during the acute infection clearly requires prompt, potent, and sustained interventions to prevent the establishment of a chronic infection and an otherwise fatal clinical outcome. The rapid pace of virus dissemination presents a formidable challenge for the development of an effective prophylactic HIV vaccine.

Methods

Virus and Animals.

The origin and preparation of the tissue culture-derived SHIV_DH12R and molecular clone-derived SIV_mac239 stocks have been described (36). Rhesus macaques (Macaca mulatta) were maintained in accordance with the guidelines of the Committee on Care and Use of Laboratory Animals and were housed in a biosafety level 2 facility; biosafety level 3 practices were followed (36).

Plasma Viral RNA Quantitation.

Viral RNA levels in plasma were determined by real-time RT-PCR (ABI Prism 7700 Sequence Detection System; Applied Biosystems) as described (20).

Lymphocyte Immunophenotyping and Live Cell Sorting.

EDTA-treated blood samples were stained for flow cytometric analysis as described (20). The preparation of positively-sorted naïve and memory CD4⁺ T cells negatively sorted resting and CD4⁺ T cells are described in SI Text. The purity of the resulting CD4⁺ T cell populations, negative for non-CD4 and activation markers (CD25, CD69, and HLA-DR), from both animals was >96%.

Calibration of Cell Numbers for Quantitative DNA and Alu-LTR PCR.

Cell numbers in each genomic DNA sample were determined by quantitative PCR of albumin genes, by using an ABI Prism 7700 Sequence Detection System. The rhesus macaque albumin primer and probe sequences were AlbF, TGCATGAGAAAACGCCAGTAA; AlbR, ATGGTCGCCTGTTCACCAA; and AlbP, AGAAAGTCACCAAATGCTGCACGGAATC, as described (20). The standard used was rhesus macaque genomic DNA present in known numbers of PBMCs, as determined by serial limiting dilution.

Cell-Associated DNA Levels in Naïve or Memory CD4⁺ T Lymphocytes.

Cell-associated viral DNA was measured by a quantitative PCR assay for SIV gag by using a ABI Prism 7700 Detection System as described (20). Cell numbers analyzed in each reaction had been determined by quantitative PCR of albumin DNA in each sample.

Resting CD4⁺ T Lymphocytes Culture and Amplification of Infectious Virus in Cultured T Cell Lines.

Negatively-sorted resting CD4⁺ T cells (2 × 10⁶ cells), were treated with Pronase (Roche; 20 mg/mL), and cultured for 4 days in the absence of any stimulation in RPMI medium 1640 supplemented with 15% autologous serum, collected from each animal before infection. Culture supernatants were assayed for RT activity to measure progeny virus production (21). To amplify progeny virions in supernatants of resting CD4⁺ T cell cultures, filtered supernatants, collected daily (days 1–4) were spinoculated (37) onto MT-4 T cells in duplicate. Virus production was assessed by measuring the RT activity released into the culture medium on day 14.

Alu-LTR PCR to Detect Integrated Viral DNA.

The isolation of a cloned cell line (3D8) containing a single integrated copy of SIV DNA is described in SI Text. Genomic DNA was prepared from FACS-sorted naïve and memory CD4⁺ T cells as described (17). To detect only integrated lentiviral genomes, the described Alu-LTR PCR technique (29) was modified for use with rhesus macaque Alu sequences. To improve the sensitivity of detecting integrated SIV DNA situated near genomic Alu sequences in either orientation, pairs of outward-facing PCR primers, binding to conserved regions of human and rhesus macaque AluDNA sequences were used in combination with an SIV LTR reverse primer, as described (38, 39). For the first-round PCR, genomic DNA was serially diluted in 2-fold dilutions (500 to 2 cell equivalents), and each dilution was divided into triplicate pools. The first-round rhesus Alu -SIV LTR PCR primers were: RS-Alu-1 (+), 5′-TACTCAGGAGGCTGAGGCAGGAG-3′; RS-Alu-2 (−), 5′-GCCTCCCAAAGTACTAGGATTACAG-3′, and SIV2ST3 (−, 5′-CAAACTTCCATGCTAGAACCTCTCCC-3′. The primers used for the nested PCR were SIV.LTR-315–338 (+), 5′-GGAAGAAGGCATCATACCAGATTG-3′ and SIV.LTR.427–448 (−), 5-AATGCTCCTCATCCTCCTGTGC-3′. PCR conditions for both the first-round and nested PCRs have been described (29). SIV LTR nested PCR products were visualized on 2% agarose TAE gels (SeaKem ME Agarose; Cambrex Biosciences) with ethidium bromide (0.5 μg/mL).