Share this article on:

Unintegrated Circular HIV-1 DNA in the Peripheral Mononuclear Cells of HIV-1-Infected Subjects: Association With High Levels of Plasma HIV-1 RNA, Rapid Decline in CD4 Count, and Clinical Progression to AIDS

Panther, Lori A.*; Coombs, Robert W.*†; Zeh, Judith E.‡; Collier, Ann C.*; Corey, Lawrence*†

Journal of Acquired Immune Deficiency Syndromes & Human Retrovirology: 1 April 1998 - Volume 17 - Issue 4 - pp 303-313
Clinical Science

We observed 36 HIV-infected patients to evaluate whether the presence of tandem 2-long terminal repeat circular unintegrated HIV-1 DNA (2-LTR) in peripheral blood mononuclear cells (PBMC) at baseline was associated with acceleration of HIV disease. Detection of 2-LTR at baseline correlated with high plasma HIV-1 RNA levels (p < .01), recovery of culturable HIV-1 from plasma (p = .02), and progression to AIDS during follow-up (p = .01). More patients with 2-LTR (68%) than without 2-LTR (31%) had a decline in CD4 levels of >50 cells/mm3 over the first 18 months of follow-up (p = .04), and the average annual CD4 decline was 35% in patients with 2-LTR compared with 16% in those without 2-LTR (p = 0.06). Detection of 2-LTR in PBMC at baseline was an independent predictor of high plasma HIV-1 RNA levels and subsequent CD4 cell decline in this cohort of patients with predominantly nonsyncytium-inducing (NSI) isolates at baseline. The presence of 2-LTR in PBMC appears to be reflective of ongoing HIV-1 replication, as measured by plasma HIV-1 RNA levels, and identifies persons at risk for immunologic and clinical decline.

Departments of *Medicine, †Laboratory Medicine, and ‡Statistics, University of Washington, Seattle, Washington, U.S.A.

Address correspondence and reprint requests to Lori Panther, Harvard Medical School, West Roxbury Veterans' Administration Hospital, 1400 V.F.W. Parkway, Mailcode 11A, West Roxbury, MA 02132 U.S.A.

Manuscript received April 25, 1997; accepted September 29, 1997.

Several studies have suggested that increased viral load is associated with clinical progression of HIV-1 disease (1-6). Assessment of viral load by quantitation of cell-associated, integrated HIV-1 DNA may not accurately reflect the level of viral replication (7). A recent longitudinal study has indicated that a high level (>105 copies) of plasma HIV-1 RNA was a strong predictor of rapid clinical progression and was independent of CD4 count (8). However, the number of HIV-1 RNA copies in the plasma may not directly reflect infectious particles because a substantial proportion of virions in the plasma are probably defective.

In vitro studies of retroviruses have shown that the first evidence of reverse transcription is unintegrated viral DNA appearing in the cytoplasm and transported to the nucleus within hours after infection of a cell (9-13). This unintegrated HIV-1 DNA exists in several forms, including incompletely or completely reverse-transcribed linear DNA, circular DNA containing one long terminal repeat (LTR) copy (1-LTR circular HIV-1), circular DNA containing two copies of the LTR in tandem resulting from end-to-end joining of linear viral DNA (2-LTR circular HIV-1), and monomeric or multimeric circular HIV-1 containing two or more discontinuous LTR sequences as a result of autointegration within the HIV-1 DNA template (14-15). The majority of unintegrated HIV-1 DNA forms are of the linear type, with <10% of forms existing as 1-LTR or tandem 2-long terminal repeat unintegrated HIV-1 DNA (2-LTR) circles (16). In T-cell lines, unintegrated HIV-1 DNA is felt to have a transient existence in an infected cell. Although evidence is present that unintegrated HIV-1 DNA may serve as an extrachromosomal template for transcription and translation of some viral proteins (17), it is only the completely reverse-transcribed integrated viral genome that is capable of producing infectious virions.

Multiple copies per cell of unintegrated retroviral DNA have been demonstrated in vitro with several cytopathic retroviruses (12,18-24). Unintegrated HIV-1 DNA has been demonstrated in in vitro infections of lymphocytic (15,25,26), monocytic (27-28), and chronically infected cell lines (29). The studies noted previously suggest that unintegrated cellular HIV-1 may indicate either de novo infection or reinfection of cells and may be a marker of ongoing production of infectious HIV-1.

Cross-sectional studies have indicated the presence of unintegrated HIV-1 DNA in the peripheral blood mononuclear cells (PBMC) and tissues of patients in advanced clinical stages of HIV-1 infection (30-34) and other studies have demonstrated decreases in unintegrated HIV-1 DNA in the PBMC of patients after initiation of antiretroviral therapy (35-37). Few data indicate whether unintegrated HIV-1 DNA is a good predictor of immunologic deterioration. In this study, we investigated the value of unintegrated HIV-1 DNA in PBMC in association with subsequent CD4 decline, plasma HIV-1 RNA levels, presence of culturable virus in the blood, and clinical progression in a cohort of patients.

Back to Top | Article Outline



A cohort of 36 HIV-infected patients enrolled in a prospective study of the natural history of HIV infection (38) were selected for analysis on the basis of availability of stored blood specimens for analysis. According to the study protocol, each patient was assessed at 6-month intervals with physical and neurologic examinations. At each visit, blood was taken for T-cell subsets, plasma and PBMC HIV-1 cultures, and storage of PBMC and plasma. Each patient's record was reviewed for clinical stage according to the Centers for Disease Control and Prevention (CDC) 1993 revised classification (39), presence of HIV-1 plasma viremia, CD4 count, occurrence of opportunistic infections, and antiretroviral medications.

Back to Top | Article Outline


HIV cultures of plasma and PBMC were performed using previously described techniques (1) and AIDS Clinical Trials Group (ACTG) procedures. In brief, 30 ml of heparin-treated whole blood was diluted 1:1 with sterile phosphate-buffered saline (PBS) and centrifuged with lymphocyte separation medium (LSM, Organon Teknika, West Chester, PA, U.S.A.) to separate plasma from PBMC. Plasma cultures were performed by inoculating 2 ml of the plasma supernatant passed through a 0.45-μm filter (Corning, Indianapolis, IN, U.S.A.) into 107 phytohemagglutinin-stimulated donor PBMC and incubated in 20 ml of medium at 5% CO2 and 37°C for 28 days, with 10 ml of culture medium changed and 3 × 106 donor PBMC added at weekly intervals. Culture supernatants were assayed on days 7, 14, 21, and 28 for p24 antigen (Abbott, North Chicago, IL, U.S.A.). All p24 antigen measurements that were >30 U/ml were considered positive.

PBMC cultures for HIV-1 were performed by washing the cells with PBS and placing 107 PBMC into flasks containing 5 × 106 phytohemagglutinin-stimulated donor PBMC in 20 ml of culture medium. Cultures were incubated and sampled as described earlier.

Back to Top | Article Outline

Phenotypic Assay

In 24 of the 34 subjects with positive PBMC cultures, aliquots of the culture supernatants were available for phenotypic assay using a method described previously (40). Briefly, the supernatant aliquots were retrieved from −70°C storage and melted in a 37°C water bath. Fifty μl of supernatant from each specimen was passed into duplicate wells containing 5 × 104 MT-2 cells and examined for syncytia every 3 days under light microscopy. The presence of syncytia at any time point over a 2-week sampling period met criteria for a syncytium-inducing (SI) viral isolate.

Back to Top | Article Outline

CD4 Cell Counts

Records were reviewed for each subject and all CD4 counts were recorded for the length of their participation in the study. CD4 cell counts were measured using standard flow cytometry techniques in an ACTG-certified laboratory at the University of Washington (41). CD4 counts were performed at each patient visit during which blood was taken and entered into a longitudinal analysis of CD4 count change over time.

Back to Top | Article Outline

Preparation of PBMC Samples for Analysis

An aliquot of primary PBMC taken at the baseline visit for each patient was placed into 10% dimethyl sulfoxide (DMSO) and stored in liquid nitrogen for subsequent polymerase chain reaction (PCR). These aliquots were thawed in a 37°C water bath and washed twice in PBS. The cells were counted with a hemocytometer and 2 × 106 cells were placed into 1.5 ml Eppendorf tubes, pelleted at 2500 rpm for 5 minutes, and the supernatant was discarded.

DNA extraction of the PBMC pellets was performed by adding equal amounts of a buffer composed of 100 mM KCl, 10 mM Tris-HCl (pH 8.3), and 2.5 mM MgCl2, and a detergent solution composed of 10 mM Tris-HCl (pH 8.3), 2.5 mM MgCl2, 1% Tween 20, 1% Nonidet P-40, and proteinase K at 0.12 mg/ml, for a final concentration of cells in suspension of 6 × 106 PBMC/ml. The samples were mixed using a vortex for 15 seconds, incubated in a 60°C water bath for 1 hour then placed in a 95°C heating block for 10 minutes (42).

Back to Top | Article Outline

Polymerase Chain Reaction

PCR was used to detect a segment of HIV-1 gag gene as a marker of total intracellular HIV-1 DNA, 2-LTR circular HIV-1 DNA as a marker of unintegrated provirus, and single-copy human leukocyte antigen deterioration quotient (HLA-DQ) locus as an internal control to measure the fidelity of the DNA extraction procedure. Because the detection of 2-LTR circular HIV-1 DNA depended on an intact tandem 2-LTR circle junction, it had the advantage of being a unique marker for unintegrated HIV-1 DNA in the presence of both integrated and unintegrated linear HIV-1 DNA.

Triplicate 50 μL aliquots of DNA sample, each containing DNA equivalent to 300,000 PBMC, were used for PCR. Fifty picomoles of each indicated oligonucleotide primer and 50 μL of master mix was added to each aliquot such that final reaction conditions were 50 mM KCl, 10 mM Tris (pH 8.4), 2.5 mM MgCl2, 0.01% gelatin, 0.2 mM each of deoxynucleotides deoxyadenosine triphosphate (dATP), deoxycytydine triphosphate (dCTP), deoxyguanosine triphosphate (dGTP), and deoxythyminidine triphosphate (dTTP), and 2.5 U Taq polymerase (U.S. Biochemical Corp., Cincinnati, OH, U.S.A.) in a total volume of 100 μl. The primer pairs used to amplify HIV-1 gag were SK38 and SK39 (43); for 2-LTR circular HIV-1 DNA primer sequences were U3 and U5 (44); and for HLA-DQ locus, primers GH26 and GH27 (45) were used. The samples were placed in a Perkin-Elmer Cetus DNA Thermal Cycler (Norwalk, CT, U.S.A.) and the PCR reaction was performed using a step-cycle file. For amplification of the gag gene and the HLA-DQ locus, the thermocycler was programmed for 30 cycles of 95°C for 30 seconds, 55°C for 30 seconds, and 72°C for 1 minute. For amplification of 2-LTR circular HIV-1 DNA, the thermocycler program was 30 cycles of 89°C for 40 seconds and 62°C for 80 seconds.

The gag and 2-LTR PCR products were detected by liquid hybridization to nested probes that were 5′-labeled with 32P (45). The oligonucleotide probe SK19 was used to detect SK38/SK39 gag PCR product (43); and probe P2 (5′ GGAAAATCTCTAGCAGTACTGGAAGGGCTAATTCA3′), spanning the 2-LTR circle junction, was designed to detect the 2-LTR PCR product. The radiolabeled gag and 2-LTR PCR products were detected by electrophoresis on a 10% polyacrylamide gel. Semiquantitation of the PCR products was done by comparing band intensities with a standard curve included in each PCR run; copy number was estimated as either equal to or exactly in between two dilutions of the standard curve depending on band intensity. The HLA-DQ locus PCR product was detected by electrophoresis on a 3% agarose gel by ethidium bromide staining and ultraviolet transillumination.

The standard curve for HIV gag DNA was derived from 10-fold dilutions of ACH-2 cells (AIDS Research and Reference Program, Rockville, MD, U.S.A.), an immortal cell line containing one integrated copy of the HIV-1 genome per cell (46). The sensitivity of our PCR using this method was 1 to 10 copies of HIV-1 gag. The standard curve for 2-LTR was derived from 10-fold dilutions of a pTZ19U plasmid into which had been subcloned a 590-bp DNA sequence containing the 2-LTR circle junction (provided by Jeffrey S. Smith, Robert Wood Johnson Medical School, University of Medicine and Dentistry of New Jersey, Piscataway, NJ, U.S.A.) (47). Using this standard, PCR sensitivity for 2-LTR was 1 to 10 copies. PCR on each sample was performed twice, and identical results were obtained for each sample in this study. All virologic assays were performed retrospectively and without knowledge of either clinical status or presence, absence or changes in antiretroviral therapy, or other drug therapies.

Back to Top | Article Outline

Plasma HIV-1 RNA Load Measurement

Plasma HIV-1 RNA levels were measured according to the protocol used by Chiron (48). Briefly, aliquots of whole plasma were centrifuged at 23,500 × g for 1 hour at 4°C to pellet the virus, the supernatant was discarded, and the pellet was lysed. Viral RNA was captured through oligonucleotide probes complementary to portions of the HIV-1 pol gene. Complementary bDNA molecules were hybridized to the captured viral RNA, followed by hybridization of alkaline phosphatase-labeled probes to the bound bDNA molecules, then detected by chemoluminescence. This procedure was performed in duplicate for each sample and reported as # copies/ml of plasma.

Back to Top | Article Outline

Statistical Analysis

Groups were compared using Fisher's exact test for proportions, t-tests for normally distributed data, and Mann-Whitney rank sum tests for nonparametric comparisons. Spearman rank correlation coefficients were used to assess correlations where appropriate. All p values presented are from two-tailed tests.

CD4 count changes over time and plasma RNA levels over time were analyzed using a random-effects model for longitudinal data (49). CD4 and RNA values were log-transformed, and the models were fitted by restricted maximum likelihood using the software application BMDP5V (50). These models allow for missing data and for correlations among measurements from the same person at different time points. To check that the missing data assumptions of the models were satisfied, we conducted separate analyses of groups of patients with differing amounts of follow-up and verified that similar results were obtained in each group.

Back to Top | Article Outline


The median age of the 36 patients was 35 years old (range, 23-48 years) and median estimated duration of HIV seropositivity was 28 months (range, 1-96 months). All patients were gay or bisexual men, and 5 patients were also injection drug users. The median CD4 count on enrollment into the study was 473 cells/mm3 (range, 25-1244 cells/mm3). The median duration of follow-up for the 36 patients was 36 months (range, 12-48 months). Only 3 patients were on antiretroviral therapy at baseline, but 23 patients received antiretroviral therapy at some point during follow-up.

Back to Top | Article Outline

Polymerase Chain Reaction

2-LTR DNA was detected in 19 (53%) of the 36 patients. Age, duration of HIV-1 seropositivity, and length of follow-up did not differ between those with and without 2-LTR (Table 1). HIV-1 gag DNA was present in all 36 PBMC samples. An example of the PCR results for 7 representative subjects is shown in Figure 1. No correlation was found between level of gag HIV-1 and CD4 count at baseline (Fig. 2A), although the gag copy number was significantly higher in the 19 patients positive for 2-LTR in their PBMC (median, 1000 copies/300,000 PBMC) versus those patients without 2-LTR (median, 500 copies/300,000 PBMC; p = .02) (Fig. 2B; Table 2). The median 2-LTR copy number in the 19 patients positive for 2-LTR was 50 copies/300,000 PBMC (range, 5-500 copies/300,000 PBMC) (Table 2).

Back to Top | Article Outline

Relation Between 2-LTR and Disease Stage

The detection of 2-LTR was associated with development of clinical immunodeficiency during the follow-up period. At baseline, 13 of the 19 2-LTR-positive patients and 15 of the 17 2-LTR-negative patients were in a non-AIDS CDC class (p = .24, Fisher's exact test) (Table 1). At the end of follow-up, 11 of the 19 patients 2-LTR-positive patients had a diagnosis of AIDS, whereas only 3 of the 17 2-LTR-negative patients fell into this category (p = .02, Fisher's exact test). Seven of the 13 2-LTR-positive patients and only 1 of the 15 2-LTR-negative patients who were AIDS-free at baseline progressed to AIDS during the follow-up period (p = .01, Fisher's exact test) (Table 2).

Back to Top | Article Outline

Antiretroviral Therapy

Zidovudine therapy was given at some time during follow-up in 12 of 19 patients (63%) with 2-LTR and 9 of 17 patients (53%) without 2-LTR detected at baseline (p = .74). Length of zidovudine therapy did not differ between those with (mean, 17.5 months) and without (mean, 22.7 months) 2-LTR detected. However, patients with 2-LTR at baseline were more likely to receive didanosine (ddI) or zalcitabine therapy (p = .02) at some point during follow-up, possibly representing a change in therapy in response to decreasing CD4 counts. Overall, 7 of 19 patients (37%) with 2-LTR at baseline received combination antiretroviral therapy at some point during follow-up, compared with 1 of 17 (6%) of those without 2-LTR at baseline (p = .04) (Table 2).

Back to Top | Article Outline

2-LTR, Plasma Viremia, and Viral Phenotype

An association exists between 2-LTR DNA (plasma viremia) and isolation of cell-free HIV-1 from plasma by culture. Of the 19 (63%) patients, 12 with 2-LTR were plasma viremic compared with only 4 of the 17 (24%) 2-LTR-negative patients (p = .02) (Table 1).

Baseline CD4 counts of patients who were plasma viremic (mean, 452 cells/mm3) did not differ from patients who were not plasma viremic (mean, 563 cells/mm3; p = .19). In a model of percentage-per-year CD4 decline comparing patients with and without plasma viremia at baseline and adjusted for antiretroviral treatment, patients with plasma viremia at baseline were predicted to have CD4 declines of 37% per year, compared with 16% per year for patients without plasma viremia at baseline (p = .04).

HIV-1 was isolated from PBMC coculture in 34 of 36 patients and culture supernatants were available for phenotypic assay in 24 of 34 patients. Of the 24 isolates, 13 were from 2-LTR-positive and 11 from 2-LTR-negative subjects. The majority of patients had NSI phenotype, with SI phenotype noted in only 5 patients, 4 of whom were 2-LTR-positive. Thus, 2-LTR HIV-1 DNA was detected in patients with both NSI and SI strains of HIV (Table 1). Presence of an SI strain in the plasma at baseline was not a significant predictor of CD4 cell decline (p = .08).

Back to Top | Article Outline

Unintegrated HIV-1 DNA as a Predictor of CD4 Decline

Baseline mean CD4 counts were somewhat higher in the 2-LTR-positive patients versus the 2-LTR-negative patients, but this difference was not significant (Table 1) (p = .18). The amount of 2-LTR unintegrated HIV-1 DNA at baseline did not correlate with baseline CD4 count (Fig. 2C).

All patients were observed for at least 1 year, and all were observed for at least 18 months except for 1 patient. We first compared the 2-LTR-positive and 2-LTR-negative patients over the first 18 months of follow-up. The mean decline in CD4 during the first year of follow-up was 75 cells/mm3 in the 2-LTR-positive patients, compared with 16 cells/mm3 in the 2-LTR-negative patients. However, the difference was not statistically significant because of the high variability between patients (Table 2). Of the 19 2-LTR-positive patients, 13 (68%) experienced CD4 declines of >50 cells/mm3 over the first 18 months, compared with 5 of 16 (31%) 2-LTR-negative patients (p = .04).

When CD4 decline over a 4-year follow-up period was estimated using the random effects model, adjusting for missing data and for antiretroviral therapy, 2-LTR-positive patients were predicted to have, on average, CD4 declines of 35% per year, compared with 16% for 2-LTR-negative patients (p = .06) (Fig. 3). When antiretroviral therapy was not factored into the model, the CD4 decline in the 2-LTR-positive and 2-LTR-negative patients was 29% and 12%, respectively (p = .08). Although the effect of antiretroviral therapy on CD4 count was not statistically significant (p = .08), it was in a positive direction as expected.

Back to Top | Article Outline

Relation Between Unintegrated HIV-1 DNA, Plasma RNA Level, and CD4 Cell Count

Plasma RNA levels were not consistently available at baseline (when 2-LTR was measured) and also not consistently available for every time point that CD4 counts were performed in the follow-up period. Thus, the data in Figure 2D depict comparisons between presence of 2-LTR at baseline and first recorded HIV-1 RNA load. In the follow-up period, 29 of the 36 patients (81%) had at least two time points for which data were available for both CD4 count and plasma RNA level, but none had more than four time points. It was not possible to adjust for antiretroviral therapy in some analyses, because the data set was small and antiretroviral therapy was correlated with other predictor variables.

The presence of 2-LTR at baseline did not correlate with lower baseline CD4 counts after adjustment for plasma RNA level (p = .63), but the first recorded HIV-1 RNA levels in the follow-up period were significantly higher in the 2-LTR-positive group (Fig. 2D; Table 1). Subjects with 2-LTR at baseline had a more rapid loss of CD4 cells during the follow-up period compared with those without 2-LTR at baseline (Fig. 3; Table 2). At time points in which both CD4 count and HIV-1 RNA levels were available, high HIV-1 RNA levels predicted low CD4 counts when log(CD4) was used as the response variable (p < .01).

In the random effects model with log(HIV-1 RNA) as the response variable, no change in RNA level over time was present, so the year of follow-up was dropped from the model. The strongest predictor of high plasma HIV-1 RNA at a given time point was low CD4 count at that time point (p = .001). Patients with 2-LTR at baseline had significantly higher plasma HIV-1 RNA loads over the follow-up period than patients without 2-LTR, even after adjustment for CD4 count (p = .003). When antiretroviral therapy was added to the model, it resulted in slightly lower plasma HIV-1 RNA levels, and the effects of CD4 count and 2-LTR at baseline remained highly significant (Table 3).

Back to Top | Article Outline


This preliminary study of a small cohort of patients suggests an in vivo association between the detection of unintegrated 2-LTR circular HIV-1 in PBMC and clinical stage, plasma viremia, and rate of CD4 count decline. Although 2-LTR in PBMC and positive plasma cultures for HIV-1 were both predictors of subsequent CD4 cell loss, the random effects model suggested that detection of 2-LTR in PBMC at baseline was a slightly better predictor of subsequent CD4 cell decline. Our observation that most of the 2-LTR-positive patients had HIV-1 isolates of NSI phenotype indicates that measurement of 2-LTR in PBMC appears to be a more universal marker of subsequent CD4 decline compared with HIV-1 phenotype. Although the age and duration of seropositivity were slightly higher in the 2-LTR-positive patients, these parameters did not differ significantly from the 2-LTR-negative patients. However, if the detection of 2-LTR as a marker of unintegrated HIV-1 DNA is a "threshold event" in the natural history of HIV-1 infection as it begins to accelerate, these slight differences may have explained the detection of 2-LTR in these subjects.

The length of follow-up and similar percentages of patients with no antiretroviral therapy in both 2-LTR-negative and 2-LTR-positive groups are aspects of the study that add strength to our observations. Our findings of higher gag copy numbers in patients with AIDS is in agreement with Bukrinsky et al. (51) and although we found a qualitative association between presence of 2-LTR in PBMC at baseline and progression to AIDS, we found no correlation between 2-LTR copy number and CDC class at baseline. This qualitative but not quantitative relationship may result from the fact that 2-LTR unintegrated DNA comprises a minority of unintegrated forms and thus the failure to detect other forms of unintegrated HIV-1 DNA may have obscured a quantitative relationship between amount of 2-LTR and stage of disease. Our findings are in agreement with those of Jurriaans et al. (52), who found a correlation between presence of 1-LTR circular unintegrated HIV-1 DNA in the PBMC and subsequent clinical progression in 21 subjects, as well as Zazzi et al. (53) who recently reported both immunologic and clinical decline in 23 patients with 2-LTR in PBMC at baseline compared with 2-LTR-negative controls. Our study strengthens these observations by correlating the presence of 2-LTR with plasma viremia (a measure of infectious virus production), high levels of plasma HIV-1 RNA (a measure of total HIV-1 particle production), and decline in CD4 counts over time.

Our study showed significantly more patients with 2-LTR in PBMC at baseline had initiation of combination antiretroviral therapy during follow-up compared with 2-LTR-negative patients, which suggests a more marked pattern of clinical and virologic progression in the 2-LTR-positive patients. Thus, the presence of unintegrated HIV-1 DNA in PBMC may be a molecular marker for continued CD4 cell infection and thus may indicate an increased risk for more rapid clinical progression. Two studies have indicated unintegrated HIV-1 DNA as a dynamic marker of HIV-1 downregulation on beginning antiretroviral therapy (36,37). Both of these studies observed the patients ≤8 weeks, within the time period of greatest virologic response to institution of therapy. Longer-term studies are warranted to assess the utility of 2-LTR to indicate failure of antiretroviral therapy as resistant HIV-1 strains emerge.

Semiquantitative measurement of HIV-1 gag DNA was not predictive of either CD4 count at baseline or stage of disease. Baseline CD4 counts were similar for both 2-LTR-positive and 2-LTR-negative groups; however, the 2-LTR-positive group had a faster decline in CD4 count, which indicates that measurement of a structural gene, such as gag, is not a physiologic indicator of the level of replicative capacity of the HIV-1 pool. Furthermore, no association was found between 2-LTR and CD4 count at baseline, but a relationship was observed between 2-LTR and subsequent CD4 decline. This suggests that unintegrated DNA in PBMC may be detectable during times of HIV-1 viremia and productive infection of susceptible immune cells, heralding a subsequent decline in CD4 count. Further investigation could focus on the ability of 2-LTR to distinguish patients with relatively pathogenic HIV-1 variants from patients with less fit variants.

Detection of 2-LTR DNA was not the sole predictor of disease progression, which illustrates the complex biology of HIV-1 infection. Inasmuch as only 5 patients had the SI phenotype, a solid conclusion regarding phenotype and CD4 cell loss cannot be made from this data set. Patients with either culturable virus in the plasma or 2-LTR in PBMC samples at baseline had accelerated CD4 decline. In addition, plasma HIV-1 RNA levels were inversely correlated with CD4 count at time points in which both variables were measured. The presence of high levels of HIV-1 viremia as measured by culture or RNA level suggests a failure of lymphoid tissues to contain HIV-1, whereas presence of 2-LTR in PBMC may reflect infection of lymphocytes prior to a decline in CD4 count. Larger studies may confirm 2-LTR as a marker of unintegrated HIV-1 DNA in PBMC as a valid predictor of immune decline in HIV-infected patients.

It remains unclear, however, whether unintegrated HIV-1 DNA has a direct effect on cell death or is merely an inert indicator of cell infection. Bukrinsky et al. (54) showed that 2-LTR circles were not associated with active integrase and may not have a capacity for integration into the host cell genome and production of infectious virus. Laurent-Crawford and Hovanessian (55) demonstrated that HIV-infected CEM cells treated with 5 μM zidovudine had decreased amounts of unintegrated HIV-1 DNA but continued to undergo apoptosis, though the high concentration of zidovudine may have independently contributed to the apoptosis observed in their model. Other mechanisms by which unintegrated HIV-1 DNA may contribute to cell death include serving as an extrachromosomal template for viral RNA and protein synthesis (20,56,57), perturbation of cellular membrane permeability (58), or a direct toxic effect on cellular respiration from accumulation of "free ends" of unintegrated HIV-1 DNA in the nucleus (59). Additionally, because apoptosis is hastened through interaction of HIV-1 envelope glycoprotein with CD4 receptors on the cell surface membrane (60-62), reinfection of a cell by HIV-1 may further accelerate this process.

The small number of patients and the limitations of the semiquantitative PCR method may have obscured the possibility of finding a quantitative association between amount of 2-LTR in PBMC and CD4 count at baseline. Still, the qualitative presence of 2-LTR as a marker for infectious HIV-1 production and subsequent immune decline was evident. In addition, only one baseline measurement of 2-LTR in each subject was made, and it is believed that unintegrated DNA is a relatively labile phenomenon in T-cell lines. A longitudinal analysis of a larger population of patients may reveal that appearance of unintegrated DNA may mark the emergence of pathogenic viral strains, or variants resistant to the current antiretroviral therapy. The exact role of unintegrated HIV-1 DNA in the cell was not elucidated by our study. However, the association between unintegrated HIV-1 DNA and subsequent virologic and immunologic progression in this cohort of patients indicates that it may identify patients at risk for disease progression so that therapeutic intervention may be instituted before clinical decline takes place.

Our data suggest that prospective, controlled studies are warranted to assess whether quantitative measurement of unintegrated HIV-1 in PBMC would accurately predict an increased risk for disease progression. In addition, PCR analysis of cell-associated unintegrated HIV-1 DNA is a marker that may be used to monitor virologic response to antiretroviral agents in clinical trials, and to assess for escape from immunologic containment in HIV-infected patients so that changes in therapy may be instituted before clinical progression occurs.

Acknowledgments: We are indebted to Joel Gibson, James Stamateou, and Sandra Aung for technical assistance, to Susan Ross for clinical assistance, and to Barb Williams and Paul Boutin for statistical help. This work was supported by grants NS-25183 and Al-27764 from the National Institutes of Health.

Back to Top | Article Outline


1. Coombs RW, Collier AC, Allain JP, et al. Plasma viremia in human immunodeficiency virus infection. N Engl J Med 1989;321:1626-31.
2. Schnittman SM, Greenhouse JJ, Psallidopoulos MC, et al. Increasing viral burden in CD4+ T cells from patients with human immunodeficiency virus (HIV) infection reflects rapidly progressive immunosuppression and clinical disease. Ann Intern Med 1990;113:438-43.
3. Schecter M, Neumann PW, Weaver MS, et al. Low HIV-1 proviral burden detected by negative polymerase chain reaction in seropositive individuals correlates with slower disease progression. AIDS 1991;5:573-9.
4. Escaich S, Ritter J, Rouguer P, et al. Relevance of quantitative detection of HIV proviral sequences in PBMC of infected individuals. AIDS Res Hum Retrovir 1992;8:1833-7.
5. Gupta P, Kingsley L, Armstrong J, Ding M, Cotrill M, Rinaldo C. Enhanced expression of human immunodeficiency virus type I correlates with development of AIDS. Virology 1993;196:586-95.
6. Piatak M Jr, Saag MS, Yang LC, et al. High levels of HIV-1 in plasma during all stages of infection determined by competitive PCR. Science 1993;359:1749-54.
7. Pantaleo G, Graziosi C, Demarest JF, et al. HIV infection is active and progressive in lymphoid tissue during the clinically latent stage of disease. Nature 1993;362:355-8.
8. Mellors JW, Kingsley LA, Rinaldo CR Jr, et al. Quantitiation of HIV-1 RNA in plasma predicts outcome after seroconversion. Ann Intern Med 1995;122:573-9.
9. Fritsch E, Temin H. Formation and structure of infectious DNA of spleen necrosis virus. J Virol 1977;21:119-30.
10. Shank PR, Varmus HE. Virus-specific DNA on the cytoplasm of avian sarcoma virus-infected cells is a precursor to covalently closed circular viral DNA in the nucleus. J Virol 1978;25:104-14.
11. Clayman CH, Mosharrafa ET, Anderson DL, Faras AJ. Circular forms of DNA synthesized by Rous sarcoma virus in vitro. Science 1979;206:582-4.
12. Weller SK, Temin HM. Cell killing by avian leukosis viruses. J Virol 1981;39:713-21.
13. Stevenson M, Stanwick TL, Dempsey MP, Lamonica CA. HIV-1 replication is controlled at the level of T cell activation and proviral integration. EMBO J 1990;9:1551-60.
14. Pauza CD, Galindo J. Persistent human immunodeficiency virus type 1 infection of monoblastoid cells leads to accumulation of self-integrated viral DNA and to production of defective virions. J Virol 1989;63:3700-7.
15. Farnet CM, Haseltine WA. Circularization of human immunodeficiency virus type 1 DNA in vitro. J Virol 1991;65:6942-52.
16. Chun T-W, Carruth L, Finzi D, et al. Quantification of latent tissue reservoirs and total body viral load in HIV-1 infection. Nature 1997;387:183-8.
17. Cara A, Cereseto A, Lori F, Retiz MS. HIV-1 protein expression from synthetic circles of DNA mimicking the extrachromosomal forms of viral DNA. J Biol Chem 1996;271:5393-7.
18. Keshet E, Temin HM. Cell killing by spleen necrosis virus is correlated with a transient accumulation of spleen necrosis virus DNA. J Virol 1979;31:376-88.
19. Temin HM, Keshet E, Weller SK. Correlation of transient accumulation of linear unintegrated viral DNA and transient cell killing by avian leukosis and reticuloendotheliosis virus. Cold Spring Harb Symp Quant Biol 1979;44:773-8.
20. Weller SK, Joy AE, Temin H. Correlation between cell killing and massive second-round superinfection by members of some subgroups of avian leukosis virus. J Virol 1980;33:494-506.
21. Brahic M, Stowring L, Ventura P, Haase AT. Gene expression in visna virus infection in sheep. Nature 1981;292:240-2.
22. Chen ISY, Temin HM. Establishment of infection by spleen necrosis virus: inhibition in stationary cells and the role of secondary infection. J Virol 1982;41:183-91.
23. Mullins JI, Dhen CS, Hoover EA. Disease-specific and tissue-specific production of unintegrated feline leukaemia virus variant DNA in feline AIDS. Nature 1986;319:333-6.
24. Rasty S, Dhruva BR, Schiltz RL, Shih DS, Issel CJ, Montelaro RC. Proviral DNA integration and transcriptional patterns of equine infectious anemia virus during persistent and cytopathic infections. J Virol 1990;64:86-95.
25. Pauza CD, Galindo JE, Richman DD. Reinfection results in accumulation of unintegrated viral DNA in cytopathic and persistent human immunodeficiency virus type 1 infection of CEM cells. J Exp Med 1990;172:1035-42.
26. Robinson HL, Zinkus DM. Accumulation of human immunodeficiency virus type 1 DNA in T cells: result of multiple infection events. J Virol 1990;64:4386-841.
27. Besansky NJ, Butera ST, Sinha S, Folks TM. Unintegrated human immunodeficiency virus type 1 DNA in chronically infected cell lines is not correlated with surface CD4 expression. J Virol 1991;65:2695-8.
28. Geleziunas R, Arts EJ, Boulerice F, Goldman H, Wainberg MA. Effect of 3′-azido-3′-deoxythymidine on human immunodeficiency virus type 1 replication in human fetal brain macrophages. Antimicrob Agents Chemother 1993;37:1305-12.
29. Chowdhury MIH, Koyanagi Y, Suzuki M, Kobayashi S, Yamaguchi K, Yamamoto N. Increased production of human immunodeficiency virus (HIV) in HIV-induced syncytia formation: an efficient infection process. Virus Genes 1992;6:63-78.
30. Teo I, Veryard C, Barnes H, et al. Circular forms of unintegrated human immunodeficiency virus type I DNA and high levels of viral protein expression: association with dementia and multinucleated giant cells in the brains of patients with AIDS. J Virol 1997;71:2928-33.
31. Levy JA, Kaminsky LS, Morrow WJW, et al. Infection by the retrovirus associated with the acquired immunodeficiency syndrome. Ann Intern Med 1985;103:694-9.
32. Shaw GM, Harper ME, Hahn BH, et al. HTLV-III infection in brains of children and adults with AIDS encephalopathy. Science 1985;227:177-81.
33. Pang S, Kaoyanagi Y, Miles S, Wiley C, Vinters HV, Chen ISY. High levels of unintegrated HIV-1 DNA in brain tissue of AIDS dementia patients. Nature 1990;343:85-9.
34. Li Y, Kappes JC, Conway JA, Price RW, Shaw GM, Hahn BH. Molecular characterization of human immunodeficiency virus type 1 cloned directly from uncultured human brain tissue: identification of replication-competent and -defective genomes. J Virol 1991;65:3973-85.
35. Nicholson WJ, Shepherd AJ, Aw DW-J. Detection of unintegrated HIV type 1 DNA in cell culture and clinical peripheral blood mononuclear cell samples: correlation to disease stage. AIDS Res Hum Retrovir 1996;12:315-23.
36. Bush CE, Donovan RM, Smereck SM, Strang D, Markowitz N, Saravolitz LD. Quantitation of unintegrated HIV-1 DNA in asymptomatic patients in the presence or absence of antiretroviral therapy. AIDS Res Hum Retrovir 1993;9:183-6.
37. Donovan RM, Bush CE, Smereck SM, Baxa DM, Markowitz NP, Saravolatz LD. Rapid decrease in unintegrated human immunodeficiency virus DNA after the initiation of nucleoside therapy. J Infect Dis 1994;170:202-5.
38. Collier AC, Marra C, Coombs RW, et al. Central nervous system manifestations in human immunodeficiency virus infection without AIDS. J Acquir Immune Defic Syndr Hum Retrovirol 1992;5:229-41.
39. Centers for Disease Control. 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Morbid Mortal Wkly Rep 1992;41:1-19.
40. Koot M, Vos AHV, Keet RPM, et al. HIV-1 biological phenotype in long-term infected individuals evaluated with an MT-2 cocultivation assay. AIDS 1992;6:49-54.
41. Kidd PG, Austin GD, Collier AC, Arditti DE, Coombs RW, Corey L. Interference of recombinant soluble CD4 immunoglobulin G in flow cytometric measurement of CD4+ lymphocytes. Am J Clin Pathol 1992;97:166-9.
42. Kawasaki ES. Sample preparation from blood, cells, and other fluids: a guide to methods and applications. In: Innis MA, Gelfland DH, Sninsky JJ, White TJ, eds. PCR Protocols. San Diego: Academic Press, 1990:146-52.
43. Kwok S, Erlich G, Poiesz B, Kalish R, Sninsky JJ. Enzymatic amplification of HTLV-I viral sequences from peripheral blood mononuclear cells and infected tissues. Blood 1988;72:1117-23.
44. Whitcomb JM, Kumar R, Hughes SH. Sequence of the circle junction of human immunodeficiency virus type 1: implications for reverse transcription and integration. J Virol 1990;64:4903-6.
45. Saiki RK, Scharf S, Faloona F, et al. Enzymatic amplification of β-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 1985;230:1350-4.
46. Clouse KA, Powell D, Washington I, et al. Monokine regulation of human immunodeficiency virus-1 expression in a chronically infected human T cell clone. J Immunol 1989;2:431-8.
47. Smith JS, Kim S, Roth MJ. Analysis of long terminal repeat circle junctions of human immunodeficiency virus type 1. J Virol 1990;64:6286-90.
48. Cao Y, Ho DD, Todd J, et al. Clinical evaluation of branched DNA signal amplification for quantifying HIV type 1 in human plasma. AIDS Res Hum Retrovir 1995;11:353-61.
49. Laird NM, Ware JH. Random-effects models for longitudinal data. Biometrics 1982;38:963-74.
50. Schlucter MD. Unbalanced repeated measures models with structured covariance matrices. In: Dixon WJ, Brown MB, Engleman L, Jennrich RI, eds. BMDP statistical software manual. Berkeley: University of California Press, 1990:1207-44.
51. Bukrinsky MI, Stanwick TL, Dempsey MP, Stevenson M. Quiescent T lymphocytes as an inducible virus reservoir in HIV-1 infection. Science 1991;254:243-7.
52. Jurriaans S, deRonde A, Dekker J, Cornelissen M, Goudsmit J. Increased number of single-LTR HIV-1 DNA junctions correlates with HIV-1 antigen expression and CD4+ cell decline in vivo. J Med Virol 1995;45:91-8.
53. Zazzi M, Romano L, Catucci M, et al. Evaluation of the presence of 2-LTR HIV-1 unintegrated DNA as a simple molecular predictor of disease progression. J Med Virol 1997;52:20-5.
54. Bukrinsky M, Sharova N, Stevenson M. Human immunodeficiency virus type 1 2-LTR circles reside in a nucleoprotein complex which is different from the preintegration complex. J Virol 1993;67:6863-5.
55. Laurent-Crawford AG, Hovanessian AG. The cytopathic effect of human immunodeficiency virus is independent of high levels of unintegrated viral DNA accumulated in response to superinfection of cells. J Gen Virol 1993;74:2619-28.
56. Somasundaran M, Robinson HL. Unexpectedly high levels of HIV-1 RNA and protein synthesis in a cytocidal infection. Science 1988;242:1554-7.
57. Stevenson M, Haggerty S, Lamonica CA, Meier CM, Welch S-K, Wasiak AJ. Integration is not necessary for expression of human immunodeficiency virus type 1 protein products. J Virol 1990;64:2421-5.
58. Cloyd MW, Lynn WS. Perturbation of host-cell membrane is a primary mechanism of HIV cytopathology. Virology 1991;181:500-11.
59. Temin HM. Mechanisms of cell killing/cytopathic effects by nonhuman retroviruses. Rev Infect Dis 1988;10:399-405.
60. Groux H, Torpier G, Monté D, Mouton Y, Capron A, Amiesen JC. Activation-induced death by apoptosis in CD4+ T cells from human immunodeficiency virus-infected asymptomatic individuals. J Exp Med 1992;175:331-40.
61. Meyaard L, Otto SA, Jonker RR, Mijnster MJ, Keet RP, Meidma F. Programmed death of T cells in HIV-1 infection. Science 1992;257:217-9.
62. Gougeon M-L, Garcia S, Heeney J, et al. Programmed cell death in AIDS-related HIV and SIV infections. AIDS Res Hum Retrovir 1993;9:553-63.

Key Words: HIV-1; Biologic markers; Viral DNA; T lymphocytes; Polymerase chain reaction

© Lippincott-Raven Publishers.