Patients and Methods

Study patientsWe evaluated 35 asymptomatic HIV-1–infected patients with plasma HIV-1 RNA levels <400 copies/mL (reverse transcriptase [RT]–PCR [Roche] or branched chain DNA analysis [Chiron]) for 5–59 months. Each patient had <50 HIV-1 RNA copies/mL when evaluated in this study by clinically ultrasensitive RT-PCR (Roche) and another determination of <50 HIV-1 RNA copies/mL by a laboratory-based supersensitive RT-PCR assay [11]. The study patients were 32 men and 3 women (patients 11, 12, and 32)

The patients in this study were identified from a larger group of >500 HIV-1–infected persons who were receiving treatment in our various clinics. The patients were those referred to our laboratories with <50 copies/mL of plasma viral RNA while receiving suppressive HAART from November 1998 to March 2000. All patients were immunologically (i.e., had CD4⁺ T lymphocytes) and virologically stable with stable maintenance HAART regimens. None had active opportunistic infections or other significant medical conditions before or during the study. Some cohort members were evaluated for other viral parameters in previous studies [8, 11]. For study purposes, we obtained 80–100 mL of peripheral blood from each patient via peripheral phlebotomy

Coculture viral outgrowth assaysPBMC were separated from plasma by discontinuous ficoll centrifugation. CD8⁺ T lymphocytes were depleted from isolated PBMC by binding to magnetic beads conjugated with the anti–CD8 antibody (Biosource). This process decreases the fraction of CD8⁺ T lymphocytes in the PBMC from ∼20%–30% to 3%–5%, as determined by flow cytometry [8]. The remaining PBMC were stimulated with 5 μg/mL phytohemagglutinin (Sigma), which lacks significant cellular toxicity at this concentration [12], and 50 U/mL interleukin-2 (Gibco-BRL). We mixed the PBMC from the study patients in a 1:1 ratio (10⁷ cells each) with those of HIV-1–seronegative subjects, to allow for wild-type HIV-1 growth when infected, and then cultured the PBMC in RPMI 1640 media with 10% fetal calf serum and penicillin plus streptomycin at 37°C for 8 weeks. HIV-1 p24 antigen was measured twice a week in supernatants by ELISA (DuPont)

Quantitation of HIV-1 2-LTR circular DNAPBMC, separated from plasma by discontinuous ficoll centrifugation, without CD8⁺ T lymphocyte depletion or in vitro stimulation, were used directly for HIV-1 2-LTR circular DNA analysis. The PBMC for 2-LTR circular DNA analysis were obtained from each patient on the same date as PBMC used for in vitro viral outgrowth assays. DNA was extracted from these PBMC, and HIV-1 2-LTR circular DNA then was analyzed by a quantitative DNA PCR technique [13], with modifications in the amplification parameters to optimize sensitivity and specificity. Amplification was done with an initial denaturation step at 95°C for 10 min, followed by amplification for 35 cycles at 95°C for 1 min, 60°C for 1 min, 72°C for 1 min, and a final extension of 72°C for 5 min. The sense and antisense primers used in this DNA PCR for 2-LTR circular forms were 5′-GTAACTAGAGATACCCTCAAC-3′ and 5′-CAGATCTGGTCTAACCAGAGA-3′, respectively [13]. As negative controls, PBMC DNA from HIV-1–seronegative subjects was used. The amplified PCR products were transferred to a nylon membrane (Gene Screen Plus; DuPont) for Southern blotting that used a specific probe that overlaps the 2-LTR DNA circle junction: 5′-AGTGGCGAGCCCTCAGATGCTGC-3′ labeled with ³²P. We used a PhosphorImager (Molecular Dynamics) to visualize and quantify the specific amplicon bands. To normalize the cell number used for 2-LTR circular forms in this DNA PCR procedure, the human β-globin gene was quantitated from PBMC of each patient sample

To prepare the standard curves for this assay, we used acutely infected H9 T cells with HIV-1 strain NL4-3. DNA extracted from these cells served as a template to amplify 2-LTR circular DNA. The isolated band in the positive control, by using the above-described primers and probe for PCR, was 531 base pairs long. The copy numbers of 2-LTR amplicons were calculated on the basis of the quantity of amplified DNA, as measured spectrophotometrically. This positive control amplicon DNA was serially diluted and was analyzed via Southern blotting. As such, copy numbers of 2-LTR DNA circular forms in the study samples could be quantitated by using this standard curve. The 2-LTR circular DNA was expressed as copies per 10⁶ PBMC

Statistical analysesWe used 2-sample, 2-tailed Student t tests to analyze the correlations between in vitro viral outgrowth and CD4⁺ T lymphocyte counts and correlations between CD4⁺ T lymphocyte counts and 2-LTR circular DNA detection. The Spearman’s&amp;rank correlation coefficient was used to compare 2-LTR circular DNA copy numbers with CD4⁺ T lymphocyte counts. The Mann-Whitney U test was used to compare 2-LTR DNA circle copy numbers with viral outgrowth in vitro. The 3 viral outgrowth patterns were compared simultaneously with 2-LTR DNA circle copy numbers by the Kruskal-Wallis test, and Fisher’s exact tests were used to determine correlations between 2-LTR circular DNA detection and patterns of viral outgrowth in vitro. All analyses were done with Systat software (SPSS)

Results

The characteristics of the 35 patients are shown in table 1. In addition to analysis of viral outgrowth from CD8⁺ T lymphocyte–depleted PBMC, 30 patients also had PBMC from the same date available for HIV-1 2-LTR circular DNA analyses; 5 patients had insufficient PBMC. Of note, all 35 patients had low viral RNA detected by a supersensitive laboratory-based assay for plasma viral RNA. The assay can quantitate ⩽5 copies/mL and detect <5 copies/mL (not illustrated), as we reported elsewhere [11]. Plasma viral RNA levels of all patients were confirmed by this laboratory-based assay to be <50 copies/mL

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 1

Quantitative DNA polymerase chain reaction (PCR) assay for the detection of human immunodeficiency virus type 1 (HIV-1) 2–long-terminal repeat (2-LTR) circular DNA in peripheral blood mononuclear cells (PBMC) of patients receiving suppressive highly active antiretroviral therapy (HAART). Equal quantities of DNA from freshly isolated PBMC of HIV-1–infected persons receiving suppressive HAART and PBMC of HIV-1–seronegative donors (negative controls) were amplified by quantitative DNA PCR to evaluate HIV-1 2-LTR circular DNA (see Patients and Methods). The human β-globin gene was used as a DNA loading standard in the assays. Positive controls with standard serial dilutions were developed from PCR-amplified fragments of 2-LTR DNA circles of acutely infected (HIV-1_NL4-3) H9 T cells. Some clinical virus isolates have variable LTR lengths, as shown by modest differences in size of 2-LTR DNA circle amplicons in some clinical PBMC samples. This PCR technique reproducibly quantitated 2-LTR DNA circles of 10 copies/10⁶ cells and detected but could not precisely quantitate below this level. These autoradiographs illustrate Southern blots that used a radiolabeled internal probe that overlapped the 2-LTR DNA circle junction

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1

Viral outgrowth in vitro and residual human immunodeficiency virus type 1 (HIV-1) replication in vivo

Viral outgrowth and CD4⁺ T lymphocytes in subgroups of HIV-1–infected persons on suppressive HAARTViral outgrowth within the cohort showed 3 patterns of in vitro growth. Seventeen (49%) CD8⁺ T lymphocyte–depleted PBMC cultures did not grow virus in vitro, whereas 18 (51%) yielded viral outgrowth, a percentage consistent with findings by others with a similar CD8⁺ T lymphocyte–depleted PBMC coculture system, rather than solely isolated and resting CD4⁺ T lymphocyte cultures [14]. The positive cultures had 2 quite different viral replication patterns, termed productive and abortive viral replication. Productive viral growth was considered to be HIV-1 p24 antigen expression in the supernatant of these cultures above 250 pg/mL, with increasing viral growth over the next several weeks in culture. In the productive viral growth, cultures continued to increase to nanogram per milliliter levels of HIV-1 p24 antigen. Abortive viral growth was considered as HIV-1 p24 antigen production >30 but <250 pg/mL in these culture supernatants. This concentration of HIV-1 p24 antigen (250 pg/mL) was used because it is the top limit of an ELISA (DuPont) used to quantitate very low levels of viral expression. Most of the p24 antigen levels in these cultures dropped to below detection or near the lowest limit of detection after several weeks in culture. Of interest, in recent studies of 3 patients with abortive replication, compared with 3 patients with productive viral replication patterns, the viral growth phenotypes were fully maintained after passage of the virus isolates onto CD8⁺ T lymphocyte–depleted PBMC from HIV-1–seronegative persons after normalization of viral input (i.e., equal input of HIV-1 p24 antigen; not illustrated). Further examination of these viral phenotypes is ongoing

The mean CD4⁺ T lymphocyte count in the patients yielding productive viral growth in vitro was 455 cells/mm³, with 613 and 768 cells/mm³ in abortive growth and in cultures negative for viral growth, respectively. These data yielded a statistically significant difference for productive versus no growth as related to CD4⁺ T lymphocyte counts (P<.02). Although it did not reach statistical significance, there was also a clear trend for increasing CD4⁺ T lymphocyte counts from productive to abortive and abortive to no viral growth. Also, comparison of CD4⁺ T lymphocyte counts in no viral growth versus all positive cultures (i.e., productive plus abortive) yielded statistically significant differences in these patients, 768 versus 543 cells/mm³, respectively (P⩽.02; (table 1)

HIV-1 2-LTR circular DNA and cryptic viral replication in vivoThirty patients in this cohort were evaluated for 2-LTR circular DNA in isolated PBMC with the same samples used for viral outgrowth cultures (table 1; figure 1). Twenty of these 30 patients (67%) were positive for 2-LTR DNA circles by a quantitative PCR technique. By group, there were dramatic differences in 2-LTR circular DNA detection in productive 7 of 7, abortive 9 of 10, and no viral outgrowth 4 (31%) of 13 persons. In addition, if productive and abortive infection were combined (e.g., positive growth) 16 (94%) of 17 were positive for 2-LTR circular DNA in isolated PBMC. When we compared abortive to negative, productive to negative, and positive (abortive plus productive) to negative, the differences were statistically significant for 2-LTR circular DNA detection at P<.01, P<.01, and P<.001, respectively. There was no statistically significant difference in 2-LTR circular DNA detection in abortive versus productive infections

DNA copy numbers in each viral outgrowth group by the 2-LTR method were as follows: productive growth, <10–303 copies/10⁶ cells with only 1 sample with <10 copies/10⁶ cells; abortive growth, negative to 334 copies/10⁶ cells with 6 samples being <10 copies/10⁶ cells. Samples without in vitro viral growth showed a 2-LTR DNA circle range from negative to <10 copies/10⁶ cells. All 4 positive 2-LTR DNA samples in the subgroup with no viral outgrowth had <10 copies/10⁶ cells (i.e., below the limit of quantitation but detectable; figure 1). The 2-LTR DNA circle copy numbers were correlated inversely with CD4⁺ T lymphocytes at a highly significant level (r=.505; P⩽.01), although detection solely of 2-LTR DNA circles did not correlate significantly with CD4⁺ T lymphocyte counts. Of note, there appeared to be a threshold level of ∼500/mm³ CD4⁺ T lymphocytes at which the copy numbers of 2-LTR DNA circles increased as CD4⁺ T lymphocyte counts decreased (not illustrated). The 2-LTR DNA circle copy numbers were remarkably different between no viral growth to abortive growth, no viral growth to productive growth, and positive (i.e., abortive plus productive) and no growth groups; P⩽.003, P⩽.001 and P⩽.001, respectively. A 3-way simultaneous comparison of negative, abortive, and productive viral growth to 2-LTR DNA circle copy numbers was also highly significant (P⩽.001)

Discussion

These studies reveal that patients with divergent clinical and virologic parameters can be clearly delineated from HIV-1–infected persons with viral RNA in peripheral blood plasma level <50 copies/mL while receiving suppressive HAART. Viral outgrowth assays from CD8⁺ T lymphocyte–depleted PBMC cultures correlated in a statistically significant manner with levels of CD4⁺ T lymphocytes in the peripheral blood of these patients. Lower levels of CD4⁺ T lymphocytes were demonstrated in patients with viral outgrowth. In addition, a trend for progressively lower CD4⁺ T lymphocyte counts was observed in patients for whom negative, abortive, and finally productive viral outgrowth cultures were compared

There were significant direct correlations between the ability to detect HIV-1 2-LTR circular DNA and viral outgrowth in either productive, abortive, or combined productive and abortive growth, compared with patients for whom no viral outgrowth in vitro was demonstrated. Quantities of 2-LTR DNA circle copy numbers correlated strongly with productive, abortive, or positive viral outgrowth, compared with cultures lacking viral outgrowth. This suggests that viral outgrowth in vitro correlates with ongoing low-level HIV-1 replication in PBMC in vivo. Finally, lower CD4⁺ T lymphocyte counts were correlated with higher copy numbers of 2-LTR DNA circles

HIV-1 2-LTR circular DNA is an end-stage form of retroviral DNA in which cellular nuclear ligases catalyze the joining of the 2-LTRs located at both the 5′ and 3′ ends of the virus from linear viral DNA into a circular episome after progression of the viral preintegration complex into the nucleus from the cytoplasm. The 2-LTR DNA moieties are a direct correlate of new viral infections of cells [13]. Low levels of covert lentiviral replication may be detected by analyzing 2-LTR DNA circles. As such, 2-LTR circular DNA represents the molecular marker of covert viral replication. Although the half-life of 2-LTR DNA circles is controversial, studies indicate these moieties may have a short half-life of hours or days [13]. This suggests that their detection would correlate directly with new cellular infection events in the recent past. If 2-LTR circular DNA is shown to have a longer half-life, this would still be useful for evaluating viral replication over a significant but finite time period, as these DNA structures might represent cumulative temporal events indicating ongoing viral replication. If this were the case, measurement of 2-LTR circular DNA might be akin to the measurement of hemoglobin A_1C in diabetes for evaluation of glucose control over a more prolonged period

One study showed persistence of episomal HIV-1 moieties in patients receiving suppressive HAART [13]. In this analysis, most patients had 2-LTR DNA circles in PBMC but there were no clear correlations evaluated with critical clinical parameters (e.g., CD4⁺ T lymphocyte counts). In a small group (n=13), Sharkey et al. [13] demonstrated an association between the frequency of numbers of positive virus cultures in vitro with the presence of 2-LTR DNA circles

The present analyses also suggest the clear need for longitudinal studies in cohorts such as this one, to determine whether covert HIV-1 replication below the detection levels of most clinical assays leads over relatively long periods to eventual CD4⁺ T lymphocyte depletion. There are now cases in the literature in which this has been well described [15]. The relatively low level of CD4⁺ T lymphocytes in patients with ongoing in vivo viral replication (i.e., detectable 2-LTR circular DNA) below the detection limit of most clinical assays and in those with viral outgrowth by the in vitro culture system might be based on continuing depletion of these immune cells in vivo. It is also possible that persons with prior presence of relatively few CD4⁺ T lymphocytes (i.e., before HAART and/or after HAART-induced new CD4⁺ T lymphocyte levels) have the propensity for ongoing viral replication because of poor immune function in containing viral replication in the CD4⁺ T lymphocyte pool

In summary, subgroups of HIV-1–patients can now be defined and stratified within cohorts of patients receiving suppressive HAART by virologic and clinical parameters. This suggests the need for further studies to evaluate the significant diversity of the increasingly large numbers of patients with low levels of plasma viral RNA on HAART (i.e., <50 copies/mL), who are not virologically homogenous. Longitudinal studies should allow for the determination of whether “intensification” therapy might be added to the suppressive HAART regimens of certain patients with ongoing but covert viral replication and with depleted CD4⁺ T lymphocytes to intervene in halting potentially ongoing immune cell depletion [15]. This is important as treatment of HIV-1–infected patients with virally suppressive HAART is maintained over increasingly long time periods. Finally, clinical protocols underway and being designed to attempt long-term “remission” of HIV-1 infection may need to subdivide patients into differing groups to tailor clinical trials toward treating some persons with more aggressive regimens, in an attempt to halt ongoing covert viral replication in CD4⁺ T lymphocytes before attacking the latent proviral reservoir