Detection and Enumeration of HIV-1-Producing Cells by ELISPOT (Enzyme-Linked ImmunoSpot) Assay

The enzyme-linked immunospot (ELISPOT) assay was adapted to detect and enumerate HIV-1-producing cells at the single cell level. With CEM cells or peripheral blood mononuclear cells (PBMC) infected in vitro with HIV-1, the ELISPOT assay detected cells that produced HIV-1 antigens and showed that between 5.4% and 9.5% of the p24 antigen-positive CEM cells and 11.1% to 23.6% of the p24 antigen-positive PBMC were productively infected. In HIV-1-infected patients in early stage of the disease and without antiretroviral therapy, up to 4.54 HIV-1-producing cells per 10⁶ CD4⁺ T lymphocytes were detected in peripheral blood and up to 277.75 HIV-1-producing cells per 10⁶ CD4⁺ T lymphocytes were detected in splenic lymphoid tissue. Our results indicate that the ELISPOT assay could represent a new tool to study HIV-1 replication in vivo.

HIV-1 infects CD4⁺ T lymphocytes and cells of the monocyte/macrophage lineage ^(1-3). Understanding of HIV pathogenesis can be improved by an accurate evaluation of the numbers and types of latently infected and virus-producing cells. Different molecular and virus culture techniques have been used for this purpose and results that varied according to the procedure and the clinical stage of the disease have been observed ^(4-16). For example, in situ hybridization studies have demonstrated that 1/350 to 1/10⁵ peripheral-blood mononuclear cells (PBMC) expressed HIV-1-specific RNA ^(4,5). Ho et al. ⁽⁶⁾, using end-point-dilution cultures, have also shown that infectious HIV-1 can be isolated from only 1/400 to 1/50,000 PBMC. Conversely, however, studies based on polymerase chain reaction (PCR) have detected HIV-1 DNA in 0.01% to 20% of PBMC ^(7-10) and 0.01% to 69% of CD4⁺ T lymphocytes in the peripheral blood ^(11-14). In lymphoid tissue, 5 to 10 times greater frequency of infected cells have been detected by PCR and up to 34% CD4⁺ T lymphocytes have been shown to contain HIV-1 DNA ^(15,16).

The enzyme-linked immunospot (ELISPOT) assay, which is based on solid phase immunoenzyme technology, has been initially developed for enumerating antibody-secreting cells ^(17,18). This method involves coating of a membrane with an antigen, blocking of nonspecific binding sites, incubation with cells, incubation with a secondary antibody, and addition of substrate. Reverse modifications of the original technique by coating the membrane with antibodies have allowed detection of cells producing various immunoreactive substances such as cytokines ^(19,20). In the present study, a variation of the reverse ELISPOT assay has been adapted to detect and enumerate cells that secrete HIV-1 antigens. This assay has been developed with CEM line cells and PBMC from healthy donors infected in vitro with HIV-1. Then, the numbers of HIV-producing CD4⁺ T lymphocytes have been determined in peripheral blood or splenic tissue from HIV-1-infected patients.

MATERIALS AND METHODS

Cells and Virus Infection

Human T-lymphoblastoid CEM cells were maintained at a density of 10⁶ cells/ml in RPMI-1640 (Sigma, St. Louis, MO, U.S.A.) supplemented with 10% fetal calf serum (Institut J. Boy, Rheims, France), 2 mM of L-glutamine, 100 U/ml of penicillin, and 50 μg/ml of streptomycin at 37°C in a 5% CO₂ atmosphere. PBMC from HIV-seronegative healthy donors were isolated by Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) density gradient centrifugation, stimulated with 10 μg/ml phytohemagglutinin A (Difco Laboratories, Detroit, MI, U.S.A.) for 72 hours, and maintained in RPMI-1640 culture medium with 20% fetal calf serum, 10 U/ml of recombinant human interleukin-2 (Boehringer, Mannheim, Germany), 100 U/ml of penicillin, and 100 μg/ml of streptomycin. A total of 10⁷ cells was incubated for 60 minutes at 37°C with HIV-1_LAI strain at a multiplicity of infection (MOI) of 0.1/cell. After washings with RPMI-1640, cells were resuspended in medium culture at a concentration of 10⁶/ml. The number of HIV-1-infected CEM cells or PBMC was determined by flow cytometry. Briefly, 10⁶ cells were fixed by adding 1 ml of 20 μg/ml lysolecithin in 1% paraformaldehyde (Sigma) for 2 minutes at room temperature. Thereafter, cells were treated with 2 ml of absolute methanol for 15 minutes on ice and with Nonidet P-40 (Sigma) for 5 minutes on ice. Cells were then incubated with anti-p24 mouse monoclonal antibody labelled with fluorescein isothiocyanate (KC57-RD1, Coulter, Hialeah, FL, U.S.A.). Cells treated with anti-IgG1 mouse monoclonal antibody served as control. The percentage of p24 antigen-positive cells were determined by flow cytometry (Epics XL, Coulter) ^(21-23).

HIV-1-Infected Patients and CD4⁺ T Lymphocyte Isolation

Study was carried out on cells obtained from 7 HIV-1-infected adults. Their clinical (A, B, and C) and biologic (1, 2, and 3) categories were defined according to the 1993 revised classification system for HIV infection ⁽²⁴⁾. Lymphocytes counts were done by standard automated techniques (Coulter). The HIV-1 RNA plasma level was determined with the Amplicor HIV-1 Monitor assay (Roche Diagnostics, Neuilly, France). Heparinized peripheral blood samples were collected from 5 HIV-1-infected patients (patients #1-#5). PBMC were isolated by Ficoll-Hypaque density gradient centrifugation. The CD4⁺ lymphocytes were separated from PBMC by an immunomagnetic bead separation technique (Dynabeads M-450, Dynac, Oslo, Norway). PBMC were incubated with beads attached to anti-CD19, anti-CD8, and anti-CD14 monoclonal antibodies (4:1 bead-to-cell ratio) for 30 minutes at 2° to 4°C. The unbound cell fraction was removed and washed twice in phosphate buffered saline (PBS). With this method, flow cytometry by standardized procedures showed that more than 90% of the remaining cells were CD4⁺ T. Moreover, splenic cells were obtained from 2 HIV-1-infected patients (patients #6 and #7) who had undergone splenectomy for HIV-1-associated thrombocytopenia. Splenic mononuclear cells were isolated by Ficoll-Hypaque density gradient density and suspended in culture medium. By flow cytometry, 6% (patient #6) and 23% (patient #7) of the splenic mononuclear cells were CD4⁺ T lymphocytes.

HIV-1 ELISPOT Assay

Ninety-six-well microtiter plates (Nunc, Roskilde, Denmark) with Immobilon-P membrane as solid phase (Millipore Corporation, Bedford, MA, U.S.A.) were used. Coating was done overnight at 4°C with 50 ng/well of anti-gp120 monoclonal antibody (Sanofi Diagnostics Pasteur, Marnes-1a-Coquette, France) or anti-HIV-1 polyclonal antibodies. Anti-HIV polyclonal antibodies were constituted by a pool of sera from HIV-1-infected patients with full serologic pattern in Western blot. This pool was adsorbed on CEM cells at a concentration of 5.10⁶ cells/ml for 60 minutes at 37°C under agitation and then used at a 1:250 dilution. After coating, unbound antibodies were removed by 3 washing with PBS and the plates were blocked with PBS-5% bovine serum albumin (Sigma) for 30 minutes at room temperature. Viable CEM cells or PBMC were counted in a hemocytometer after trypan blue dye exclusion staining and serially diluted in growth medium at concentrations of 10⁴, 5.10³, 10³, and 10² cells/ml. Uninfected cells and HIV-1-infected cells (10⁴/ml) incubated with 50 μg/ml cycloheximide (Sigma) for 30 minutes at 37°C in 5% CO₂ atmosphere served as controls. Cells obtained from HIV-1-infected patients were diluted in medium culture at a concentration of 10⁶ cells/ml. One hundred μl of cell suspensions was added to duplicate wells. Plates were then incubated for 24 hours at 37°C in 5% CO₂ humidified atmosphere. The plates were washed 3 times with PBS-0.05% Tween 20, 3 times with PBS, and 3 times again with PBS-0.05% Tween 20. Biotinylated anti-HIV polyclonal antibodies (Coulter) or biotinylated anti-p24 monoclonal antibody (Du Pont de Nemours, Bad Nauheim, Germany) were added and incubated for 6 hours at room temperature. After washing with PBS, a solution of phosphatase alcaline-conjugated streptavidine, diluted 1:500 in PBS, was added and the plates were incubated for 45 minutes at room temperature. The plates were washed 3 times in PBS and developed with the chromogen substrates (5-bromo-4-chloro-3-indolylphosphate toluidine salt and p-nitroblue tetrazolium chloride; Sigma). Spots appeared as purple precipitates after 10 minutes and the reaction was stopped by washing the plates with water. Spots were identified as described by Czerkinsky et al. ⁽¹⁷⁾, namely 0.05 to 0.2 mm circular foci, densely granulated with decrease of density radiating from the center, and were counted under low magnification (×16 and ×40) with a dissecting microscope.

RESULTS

Detection of HIV-1-Producing Cells

HIV-1-producing cells could be detected by ELISPOT in wells coated with HIV-1-infected CEM cells and PBMC, whereas no spot was detected in wells coated with uninfected cells (Fig. 1). As shown in Figure 2, HIV-1-producing CEM cells could be detected by ELISPOT at 4 to 7 days after infection and the higher number of spots were detected at day 6 postinfection. The higher number of spots was obtained by using anti-HIV-1 polyclonal antibodies for coating and biotinylated anti-p24 monoclonal antibody for detection (Figs. 2 and 3). Treatment with cycloheximide reduced the number of spots by 60% to 80% (Fig. 4). To determine the numbers of HIV-1-infected CEM cells and PBMC, a flow cytometry test, which detects intracellular HIV p24 antigen, was used. The ELISPOT assay, using anti-HIV polyclonal antibodies on solid phase and biotinylated anti-p24 monoclonal antibody, detected between 5.4% and 9.5% p24 antigen-positive CEM cells and 11.1% to 23.6% of p24 antigen-positive PBMC (Table 1).

Fig. 1	Fig. 2	Fig. 3
Fig. 4	Table 1

Detection of CD4⁺ T Lymphocytes That Produce HIV-1 in Infected Patients

Peripheral blood samples from 5 HIV-1-infected patients (patients #1 to #5) were analyzed. All patients were asymptomatic (class A1 or A2) and did not receive antiretroviral therapy. The CD4⁺ counts varied from 342 to 657 cells/mm³. HIV-1 RNA loads were determined for 4 patients and varied from 21.613 to 160.894 copies/ml (Table 2). The number of HIV-1 producing cells varied from 0 to 4.54 per 10⁶ peripheral blood CD4⁺ T lymphocytes. For each patient, results varied according to the antibodies used for coating and detection. The number of patients included in this study was too low to analyze the correlation between the number of spots and viral load or CD4⁺ cell counts.

Table 2

Splenic lymphoid tissues were obtained from 2 HIV-1-infected patients (patients #6 and #7). These patients received zidovudine therapy before and after splenectomy. These patients were classified in stages B3 and B1, respectively, of the U.S. Centers for Disease Control and Prevention (Table 2). The ELISPOT assay detected up to 16.67 HIV-1-producing cells per 10⁶ cells corresponding to 277.75 HIV-1-producing cells per 10⁶ splenic CD4⁺ T lymphocytes in patient #6 and up to 11.25 HIV-1 producing cells per 10⁶ cells corresponding to 48.92 HIV-1 producing per 10⁶ cells splenic CD4⁺ T lymphocytes in patient #7 (Table 2).

DISCUSSION

The ELISPOT assay described in this report allowed detection of cells producing HIV-1 antigens at the single-cell level. Inhibition of spot appearance by cycloheximide indicates that this assay detected cells that produced HIV-1 antigens that were synthesized de novo. Therefore, spots corresponded productively to HIV-1-infected cells that secreted HIV-1 antigens in vitro. Other techniques, such as indirect immunofluorescence assay, were reported to detect HIV producing cells ⁽²⁵⁾. The ELISPOT assay had the advantage of not requiring a fluorescence microscope.

With the ELISPOT assay, only 5.4% to 9.5% of the p24 antigen-positive CEM cells and 11.1% to 23.6% of the p24 antigen-positive PBMC produced spots, which indicates that in cell cultures, most HIV-1-infected cells did not excrete viral antigens or virions. The sensitivity of the assay depended on the antibody captured on the solid phase and on the biotinylated antibody used for the detection of the antibody-viral antigen complex when the tests were performed on CEM cells or PBMC that had been infected in vitro. The better sensitivity was obtained with anti-HIV polyclonal antibodies on solid phase and biotinylated anti-p24 monoclonal antibody for detection. This observation indicates that in vitro HIV-1-infected cells produced an excess of p24 protein as compared with HIV-1 envelope proteins. In contrast, in samples from HIV-1-infected patients, a higher number of spot-forming cells was detected with anti-gp 120 antibody for coating and biotinylated anti-HIV polyclonal antibodies for detection, which thereby indicates that HIV-1 producing cells in blood or lymphoid organs from these patients do not produce excess p24 protein.

In the peripheral blood of seropositive patients, up to 4.54 of 10⁶ CD4⁺ T lymphocytes were detected by the ELISPOT assay. All patients were in CDC classes A1 or A2 and had previously received no antiretroviral therapy. These results are in agreement with HIV RNA in situ hybridization studies ⁽⁴⁾. Most HIV-1 DNA-positive cells detected by PCR are latently infected ⁽²⁶⁾, and only a small fraction is competent for HIV-1 replication ⁽²⁷⁾. Therefore, the ELISPOT assay, in agreement with the findings of previous studies, demonstrates that a low proportion of HIV-infected cells effectively replicates the virus in the peripheral blood of asymptomatic HIV-1-infected patients.

In splenic lymphoid tissue, we detected a 10 to 60 times greater frequency of HIV-1-producing CD4⁺ T lymphocytes in comparison with peripheral blood CD4⁺ T lymphocytes; this observation confirms the preferential localization of HIV-1 expression in lymphoid tissue, as previously reported ⁽¹⁵⁾.

In summary, the ELISPOT assay described in this study allowed detection and quantitation of productive HIV-1-infected cells at the single cell level. This assay represents an powerful tool to enumerate HIV-producing cells in cellular compartments, such as the mucosa and lymphoid tissue. This could be of special interest in clinical study and for further evaluation of the antiretroviral activity of anti-HIV-1 drug therapies.

Acknowledgment: This work was supported by the Centre National de la Recherche Scientifique (CNRS), the Institut National de la Santé et de la Recherche Médicale (INSERM), the Agence Nationale de Recherche sur le SIDA (ANRS), and the Fondation pour la Recherche Médicale-SIDACTION. We thank Patrick Atger and Patrick Schuman from the Service Image Communication of the INSERM for technical help.