Quantitation of Human Immunodeficiency Virus Type 2 DNA in Peripheral Blood Mononuclear Cells by Using a Quantitative-Competitive PCR Assay

Abstract

A new quantitative-competitive PCR-based human immunodeficiency virus type 2 (HIV-2) proviral DNA assay (QC-PCR) was developed and used to determine the proviral load in HIV-2-infected individuals. Proviral load varied considerably, with means of 1,831 copies per 10⁶ peripheral blood mononuclear cells for asymptomatic subjects (n = 19) and 2,587 for AIDS patients (n = 2). HIV-2 viral and proviral loads also varied significantly over time in asymptomatic patients. These data suggest that a high level of virus replication occurs throughout the asymptomatic phase of HIV-2 infection.

AIDS is caused by human immunodeficiency virus type 1 (HIV-1) or HIV-2. HIV-1 is etiologically associated with the worldwide pandemic of AIDS, whereas HIV-2 is endemic to restricted geographic areas such as West Africa, Portugal, France, and India (10, 20). Individuals with HIV-2 infection show a lower rate of progression to AIDS than HIV-1-infected individuals and may also have better chances of survival (13). The rates of vertical and horizontal transmission of HIV-2 are significantly lower than those of HIV-1 (4, 11). The number of reported cases of AIDS in areas where HIV-2 is endemic is much lower than that in areas where HIV-1 is endemic (14, 20). Both viral and host-related factors may contribute to the differences between the natural histories of HIV-2 and HIV-1 infection (3, 9, 12, 15, 21).

Quantitative methods have been developed to allow the absolute quantitative evaluation of HIV-1 RNA molecules and proviral DNA sequences in peripheral blood mononuclear cells (PBMCs). A significant correlation between viral replication, as measured by the number of plasma RNA copies, and disease progression has been demonstrated in HIV-1-infected patients (16, 18, 22). Quantification of HIV-1 plasma RNA and proviral DNA in PBMCs and/or lymphoid tissues is also essential to ascertain the efficacy of antiretroviral therapy (5, 8, 23, 24).

Much less is known about the natural history of HIV-2 infection. Simon et al. (21) have shown that the cellular and plasma viral loads of asymptomatic HIV-2-infected individuals with >200 CD4⁺ cells/mm³ were lower than those found in a similar group of HIV-1-infected individuals. Cross-sectional studies involving patients from Guinea-Bissau and The Gambia have indicated that proviral DNA loads in PBMCs of HIV-2-infected subjects are similar to those found in HIV-1-infected subjects and are inversely correlated with CD4⁺- lymphocyte counts and clinical status (1, 2, 17). However, prospective studies of viral load in HIV-2 infection are lacking, as are cross-sectional studies of proviral DNA load in HIV-2-infected individuals from other areas where HIV-2 infection is endemic besides Africa.

In Portugal 5.2% of the AIDS cases recorded since 1983 were caused by HIV-2 infection (20). To characterize HIV-2 infection in Portugal, we have designed, developed, and optimized a quantitative-competitive PCR (QC-PCR) assay and have used it to quantify the proviral loads in PBMCs of HIV-2-infected individuals residing in Portugal and attending a Lisbon hospital.

Thirty-four blood samples were obtained from 21 HIV-2-infected individuals. The subjects’ age range was 15 to 64 years (mean, 40.8 years). HIV-2 seropositivity was determined by an enzyme-linked immunosorbent assay (ELISA) (ELAVIA-II; Pasteur) and confirmed by Western blot assay. Except for patients H8 and H9, who were diagnosed with AIDS, all subjects were asymptomatic at the time of blood collection. Five patients were receiving antiretroviral therapy with the following drugs: zidovudine (AZT) (patients H1 and H3), AZT plus zalcitabine (patient H7), or AZT plus didanosine (patients H8 and H9). PBMCs were prepared by Ficoll-Hypaque gradient centrifugation. CD4⁺ cell counts were determined by FACScan analysis. High-molecular-weight chromosomal DNA was extracted from 2 × 10⁶ cells as described previously (19).

The HIV-2 ROD env gene was cloned into pSK₂₉ (Stratagene); this generated the recombinant plasmid pFB₅₀. The internal standard used in this work, plasmid pPG, was constructed as follows. A 64-bp DNA fragment was excised from pFB₅₀ by digestion with the Tth111I restriction enzyme (which cuts at positions 7752 and 7816 in the HIV-2 ROD env gene). The deleted plasmid was gel purified, treated with Klenow polymerase to generate blunt ends, and ligated to a 72-bp gel purified DNA fragment from phage φX174, obtained by digestion of its replicative DNA form with HaeIII restriction enzyme (which cuts at positions 4878 and 4949 in the phage genome). Sequence analysis confirmed the correct substitution of the DNA fragment.

To study the competition kinetics between pPG and wild-type HIV-2 DNA and determine the sensitivity of the QC-PCR assay, reconstruction experiments were performed with the pROD1.12 plasmid as input DNA. Amplifications were carried out in 100-μl reaction mixtures containing the following: 50 μM concentrations of dATP, dGTP, dCTP, and dTTP; primers TM1 (5′-biotin-ACACCAATTGGCTTCGCACCTAC-3′ [positions 7614 to 7636 in HIV-2 ROD]) and TM2 (5′-CCCATGGTACAGTAGTGTGGCAGAC-3′ [positions 7974 to 7950]) at 0.2 μM each; 1.5 mM MgCl₂; 10 mM Tris-HCl (pH 8.3); 90 mM KCl; and 2.5 U of Taq DNA polymerase. Forty cycles of PCR amplification were carried out with the following set of temperatures: 95°C for 30 s, 62°C for 60 s, and 72°C for 90 s. Amplified products were detected either by ethidium bromide staining after agarose gel electrophoresis (Fig. 1A), in which case DNA bands were quantified by scanning densitometry, or by an ELISA-based assay. For the ELISA detection, 5 μl of each amplified product was added to two parallel microplate wells coated with 1 mg of streptavidin per ml and incubated for 2 h at 37°C. DNA was then denatured by adding 0.25 M NaOH and incubating for 10 min at room temperature. For hybridization, the pHIV probe (5′-Dig-CTGTTGCTGTTGCTGCACTATCC-3′ [positions 7789 to 7805 in HIV-2 ROD]) was added to one of the plate wells and the internal standard probe, pPhi (5′-Dig-TATGGTTACAGTATGCCCATCGCAG-3′ [positions 4920 to 4944 in φX174 phage replicative DNA form]), was added to the other well. Incubation with the probes continued for 3 h at 42°C. Plates were then incubated with alkaline phosphatase-conjugated anti-digoxigenin Fab (Boehringer Mannheim) for 1 h at 37°C. A chromogenic substrate, p-nitrophenyl phosphate, was added, and the plates were allowed to incubate for 1 h at 37°C in the dark. The optical densities (OD), read at 405 nm, ranged from 0.09 to 3.0, with a linear dynamic scale ranging from 0.1 to 2.5 after subtraction of the background. The OD cutoff of 0.1 was calculated by adding 2 standard deviations to the mean OD obtained from 10 DNA samples from HIV-seronegative individuals. The ratio between the OD of pPG plasmid and that of the HIV-2 unknown (y axis) was plotted against predetermined DNA copy numbers of pPG (x axis). The copy number of the unknown, expressed as the number of copies per 10⁶ PBMCs, was calculated from the linear regression analysis (y = 1) (Fig. 1B).

FIG. 1.

Reconstruction experiment of the competitive amplification of 1,000 copies of pROD1.12 plasmid. (A) Gel electrophoresis analyses of the amplified products. A total of 1,000 copies of pROD1.12 were coamplified with 10 (lane 1), 100 (lane 2), 1,000 (lane 3), or 10,000 (lane 4) copies of the internal standard, pPG plasmid. (B) Graphic representation of results obtained with the ELISA detection method.

We could reliably detect from 10 to 100,000 copies of HIV-2 DNA (Table 1 and data not shown). A comparison of the numbers of copies of pROD1.12 obtained by the two detection methods, ELISA and ethidium bromide staining, showed a high level of correlation (Spearman rank correlation, r = 0.92, P < 0.001). Five QC-PCR experiments were performed in which the same amount (∼200 ng) of cellular DNA from HIV-2 ALI-infected PBMCs was used as an unknown. Amplicons were detected with the ELISA assay. The results of the amplification were similar in all experiments, with a coefficient of variation (CV) of 13.5%. Taken together, these experiments demonstrated that the QC-PCR assay is highly reproducible and that it can accurately quantitate a wide range of copies of HIV-2 DNA.

TABLE 1.

Coamplification of various levels of pROD1.12^a

Input no. of copies	No. of copies calculated by:
	ELISA		Gel electrophoresis
	Meanb	CV (%)	Meanb	CV (%)
10	19	33.5	11	34.5
100	187	12.8	89	13.0
1,000	1,427	4.3	1,755	5.6
10,000	10,667	18.0	8,454	19.4

^aA total of 10, 100, 1,000, or 10,000 copies of pROD1.12 were each coamplified with 10, 100, 1,000, or 10,000 copies of pPG. Following analysis, the levels of pROD1.12 were treated as unknowns, and the copy number was determined as described in the text. 

^bThe results shown are means of six independent experiments performed in duplicate. 

The proviral DNA load in PBMCs was determined for 21 HIV-2-infected subjects. Each sample (1 μg of DNA, equivalent to 1.5 × 10⁵ cells) was coamplified in four parallel reactions, each containing 10, 100, 1,000, or 10,000 copies of pPG plasmid. Amplicons were detected and quantified with the ELISA assay, as described above. HIV-2 DNA sequences were detected in all HIV-2-infected subjects and not in 10 DNA samples from HIV-seronegative individuals or in 10 DNA samples from HIV-1-infected patients. These results indicated that the QC-PCR assay was both sensitive and specific in its ability to detect HIV-2 DNA in clinical samples. The quantitative results are summarized in Tables 2 and 3. For asymptomatic subjects the mean proviral copy number per 10⁶ PBMCs was 1,831 (∼1 HIV-2 DNA copy per 547 PBMCs). This is similar to the average proviral load found in HIV-1-infected asymptomatic individuals (1 HIV-1 DNA copy per 1,000 PBMCs) (6) and is also similar to the data previously reported for HIV-2-infected asymptomatic individuals of African origin (1, 2, 17). For AIDS patients the mean proviral copy number per 10⁶ PBMCs was 2,587. For subjects with >500 cells/mm³, the mean proviral copy number per 10⁶ PBMCs was 2,330 (range, 270 to 5,655); for subjects with 200 to 500 CD4⁺ cells/mm³, the mean proviral copy number per 10⁶ PBMCs was 1,957 (range, 261 to 5,455). In contrast to other studies, no significant correlation was found between proviral load and CD4⁺ cell count (Spearman rank correlation, r = −0.004, P = 0.983) (1, 2, 17).

TABLE 2.

Cross-sectional analysis of proviral DNA load, infectious viral load, and CD4⁺ cell counts in HIV-2-infected individuals^a

Patient	CD4+ cells per mm3	Proviral copy no. per 106 PBMCs	Viral load (TCID/106 PBMCs)b	Virus phenotypec
M19	NDd	57	0
M15	210	261	ND
H0	713	270	0.5	NSI
H13	1,416	459	0
M18	ND	460	0
H8	67	1,345	0
H15	ND	1,643	0.5	NSI
H12	640	1,656	0
H6	570	1,693	0
H11	630	1,768	0
H14	735	2,770	0
H1	1,143	3,578	0
H9	86	3,828	ND
H4	659	4,547	5	NSI
H10	954	5,419	0

^aExcept for patients H8 and H9, diagnosed with AIDS, all subjects were asymptomatic at the time of blood collection. 

^bInfectious viral load was determined by an end point dilution culture assay in which 10-fold dilutions of patient PBMCs were cocultured with uninfected PBMCs. TCID, tissue culture infective dose. 

^cThe syncytium-inducing (SI) or non-syncytium-inducing (NSI) capacity of the viruses was determined in PBMCs and MT-2 cells. 

^dND, not done. 

TABLE 3.

Longitudinal analysis of proviral DNA load, infectious viral load, and CD4⁺ cell counts in asymptomatic HIV-2-infected individuals^a

Patient (sample)	Day of blood collectionb	CD4+ cells per mm3	Proviral copy no. per 106 PBMCs	Viral load (TCID/106 PBMCs)	Virus phenotype
H2 (1)	16.03.93	1,120	5,655	0.5	NSI
H2 (2)	23.11.93	1,339	1,418	ND
H2 (3)	21.06.94	1,190	555	0
H3 (1)	16.03.94	255	279	50	NSI
H3 (2)	13.02.95	240	5,455	ND
H5 (1)	21.03.94	1,139	2,186	5	NSI
H5 (2)	20.06.94	1,229	3,331	0.5	NSI
H5 (3)	09.01.95	1,036	459	0
H5 (4)	04.05.95	1,547	5,521	0
H5 (5)	04.10.95	1,567	271	ND
H5 (6)	22.01.96	1,327	392	ND
H7 (1)	08.06.94	300	1,687	50	SI
H7 (2)	15.09.94	390	2,848	50	SI
H7 (3)	14.02.95	370	1,212	50	SI
M17 (1)	06.01.94	ND	1,027	0
M17 (2)	12.04.94	ND	859	0
M17 (3)	18.05.94	ND	268	0
M20 (1)	26.08.93	ND	567	0
M20 (2)	10.09.93	ND	11	0

^aInfectious viral load and virus phenotype were determined as described in Table 2. TCID, tissue culture infective dose. ND, not done. SI, synctium inducing. NSI, non-synctium inducing. 

^bDates are presented as day.month.year. 

CD4⁺ cell counts and proviral load in PBMCs were determined over time for six asymptomatic patients monitored for a mean period of 10.1 months (range, 0.5 to 22 months) (Table 3). The number of CD4⁺ cells was remarkably stable over time for all patients. In contrast, proviral load varied considerably for most subjects (mean variation, 18-fold; range, 2.4 to 51.6-fold). These results contrast with those for asymptomatic HIV-1-infected patients, for whom DNA copy numbers in PBMCs are usually low and stable over time (7, 22). Except for patient H3, who received AZT therapy and for whom an increase in proviral load was observed, proviral load for all other patients either decreased or fluctuated over time. The latter observation was particularly evident for the long-term nonprogressor patient H5, who was monitored for the longest period of time, 22 months. In addition, virus was isolated from the PBMCs of four of the subjects with virus titers either decreasing or remaining stable over time (Table 3). Together, these results indicate that viral multiplication may be occurring during the asymptomatic phase of HIV-2 infection and support previous suggestions, based on cross-sectional studies, that a low viral load is not the only factor accounting for the slow disease progression of HIV-2-infected patients (1, 17). Instead, the stringent control of viral multiplication over time that we have observed in most asymptomatic patients seems to favor the hypothesis that the immune system plays the most important role in the slow disease progression observed in HIV-2-infected patients (3).

Recent studies have shown that viral DNA remains detectable in lymph nodes and PBMCs of HIV-1-infected individuals whose plasma RNA levels have been reduced by combination drug therapy to undetectable levels (5, 8, 23, 24). Plasma RNA levels in HIV-2 infection are usually too low to be detected by the currently available methods (12). Thus, the QC-PCR method described here may be useful for the continued therapeutic monitoring of HIV-2-infected patients. Finally, the QC-PCR assay should also be useful in studying the viral burden in lymphoid tissue, which is crucial in understanding the dynamics of HIV-2 infection.