Efficient Replication of Human Immunodeficiency Virus Type 1 in Resting CD4⁺ T Lymphocytes Is Induced by Coculture with Autologous Dendritic Cells in the Absence of Foreign Antigens

Abstract

Dendritic cells (DC) are considered to be important contributors to human immunodeficiency virus type 1 (HIV-1) transmission and pathogenesis. As the first target cells in mucosal tissues, they can be become productively infected and can also capture virions and transfer them efficiently to CD4⁺ T cells located within lymphoid tissues. Resting CD4⁺ T cells appear to be another major target of HIV-1 in vivo, yet several blocks restrict replication in such cells. We report here that physical contact between virus-infected quiescent CD4⁺ T cells and uninfected autologous immature DC in the absence of any foreign antigen relieves these restrictions, allowing a highly productive HIV-1 replication.

Dendritic cells (DC) are crucial in generating a virus-specific immune response since they are recognized as the most potent antigen-presenting cells of the immune system, yet they also play a pivotal role in establishment and dissemination of human immunodeficiency virus type 1 (HIV-1) infection. Immature DC that reside in peripheral mucosal tissues are among the first targets for the virus. DC migrate to lymphoid tissues, where their contact with CD4⁺ T cells promotes a highly potent viral transfer, resulting in vigorous HIV-1 replication. In fact, CD4⁺ T-cell infection through transfer from DC leads to a more robust replication than direct infection of such cells (4, 12, 13). In addition, infection studies with sorted cells from skin explants have revealed that only DC-T-cell conjugates gave rise to strong viral production, as opposed to separated DC or CD4⁺ T cells (14).

It has been shown that HIV-1 cannot productively infect resting CD4⁺ T cells due to several blocks in the virus life cycle (3, 19, 23), and DC display a very efficient T-cell-priming ability under in vivo conditions. Therefore, we assessed the effect of autologous uninfected DC on HIV-1 replication in quiescent CD4⁺ T cells. To this end, CD4⁺ T cells were purified from peripheral blood mononuclear cells by negative selection and kept quiescent, whereas autologous immature monocyte-derived DC (iDC) were generated as described previously (8). Flow cytometry analyses revealed that the isolated CD4⁺ T-cell population is under a resting state, and a coculture step with autologous iDC leads to cellular activation (Table 1). Purified quiescent or phytohemagglutinin (PHA)-activated CD4⁺ T cells (CD4) were first inoculated with NL4-3 (X4-tropic) (5 ng of p24 per 10⁵ cells), washed three times with phosphate-buffered saline, and then either cultured alone or cocultured with autologous iDC. The presence of autologous iDC significantly enhanced viral production in cocultured cells (Fig. 1A). Replication of HIV-1 was increased in a dose-dependent manner and even by very small quantities of iDC (e.g., 1 iDC per 25 CD4 cells) (Fig. 1B). Similar observations were made when using a R5-tropic strain of HIV-1 (i.e., NL4-3BalEnv) (data not shown). Given that no virus production was seen in iDC infected alone with NL4-3 (Fig. 1A), it is very unlikely that the observed phenomenon might be due to a productive infection of the iDC subset. To validate that the increase in virus replication was really taking place in resting CD4⁺ T cells, such cells were first infected with single-cycle reporter virus pseudotyped with vesicular stomatitis virus glycoprotein (i.e., NL4-3-Luc⁺ Env⁻/VSV-G) before initiation of the coculture step with autologous iDC. A significant induction of virus gene expression was seen in the presence of iDC (Fig. 1C). Similar results were obtained when quiescent CD4⁺ T cells were infected with luciferase reporter virus pseudotyped with an R5-tropic envelope (i.e., JR-FL) (data not shown). To confirm that the observed enhancement of virus production was not caused by an antigenic response to some xenoantigen(s) present in the fetal bovine serum (FBS), cocultures were maintained either in 10% FBS or 10% heat-inactivated autologous human plasma. A comparable iDC-induced stimulation of viral replication was observed regardless of culture conditions (Fig. 1D, left panel). Moreover, in cultures maintained in autologous plasma, no productive virus infection could be detected in quiescent CD4⁺ T cells in the absence of iDC (Fig. 1D, right panel). Altogether these data indicate that a coculture with iDC allows productive infection of resting CD4⁺ T cells in an antigen-independent manner.

TABLE 1.

Flow cytometry analyses of cell surface activation markers

iDC/CD4 cell ratioa	% of cells positive for surface markerb
	Donor 1		Donor 2
	CD69	CD25	CD69	CD25	HLA-DR
0	0.20	0.69	0.73	2.5	8.75
1:10	0.89	3.78	1.79	6.25	17.6
1:4	2.88	6.64	5.16	7.6	21.2

^aResting CD4⁺ T cells were cultured either alone or with autologous iDC for 3 days. Cells were then stained with phycoerythrin-conjugated anti-CD3 and fluorescein isothiocyanate-conjugated anti-CD69, anti-CD25, or anti-HLA-DR and analyzed by flow cytometry.

^bThe percentage of cells positive for each surface marker among the CD3-positive cell population from two healthy donors is depicted.

FIG. 1.

Productive HIV-1 infection is seen in resting CD4⁺ T cells upon a coculture with autologous iDC in the absence of antigen. (A) Resting or PHA-activated CD4⁺ T cells (10⁵), as well as iDC (2.5 × 10⁴), were inoculated with NL4-3 for 2 h, washed three times, and cultured alone for the indicated time periods. In some instances, virus-exposed resting CD4⁺ T cells were cocultured with iDC. HIV-1 replication was monitored by measuring the p24 content. (B to D) Resting CD4⁺ T cells were first inoculated either with fully infectious NL4-3 (B and D) or single-cycle pseudotyped reporter viruses (C) for 4 h, washed, and cocultured with autologous iDC at the indicated cell/cell ratios for 5 (B), 2 (C), or 8 (D) days. In some experiments, cocultures were maintained in culture medium supplemented with 10% autologous plasma or 10% FBS (D, left insert) or 10% autologous plasma only (D, right insert). Virus replication was estimated either by measuring the p24 content or monitoring luciferase activity. The results are representative of three (A, B, and D) or five (C) independent experiments with different donors.

Viral particles and proteins that are produced during the coculture step could be taken up by iDC and then presented in a major histocompatibility complex class II context. Therefore, there is a possibility that specific T cells responding to such antigens were present within purified quiescent CD4⁺ T cells used in our coculture system. To rule out this option, experiments were performed in which no viral products could be produced. In brief, freshly isolated purified CD4⁺ T cells were transfected by the Nucleofector technology with an HIV-1 long terminal repeat (LTR)-driven luciferase construct (pLTRX-LUC) (17), either used alone or in combination with a Tat expression vector (pCEP4-Tat) (5), and then cocultured with autologous iDC for 14 h. An antigen-independent increase in HIV-1 LTR-directed transcriptional activity was seen in the presence of Tat (Fig. 2A). A very faint enhancement of HIV-1 LTR-mediated reporter gene expression was detected in the absence of Tat, thus suggesting that coculture with iDC allows production of a threshold level of Tat that is sufficient to lead to a potent transcription and translation of viral proteins in quiescent CD4⁺ T cells. Nevertheless, it is very likely that coculture with iDC not only increases HIV-1 transcriptional activity but also relieve other blocks in HIV-1 life cycle, such as at the level of reverse transcription (19, 23, 24). Indeed, it has been shown that the viral genome is not completely reverse transcribed in quiescent CD4⁺ T cells (20, 23). To gain additional information on the mechanism(s) by which a coculture step with iDC can enhance HIV-1 replication in quiescent CD4⁺ T lymphocytes, resting CD4⁺ T cells were infected with HIV-1 and virus reverse transcripts were quantified over time. As illustrated in Fig. 2B, reverse transcription proceeds with an extremely slow kinetics in resting CD4⁺ T cells, and a coculture step with iDC accelerates the kinetics of the reverse transcription process. Our data thus suggest that a coculture step with autologous iDC promotes completion of reverse transcription events in resting CD4⁺ T cells.

FIG. 2.

Coculture with autologous iDC relieves blocks to HIV-1 reverse transcription in quiescent CD4⁺ T cells. (A) Resting CD4⁺ T cells were cotransfected with pLTRX-LUC and either an empty control vector or the TAT expression vector pCEP4Tat by the nucleofection method. Cells were then cocultured with autologous iDC for 14 h before assessing luciferase activity. (B) Resting CD4⁺ T cells were pulsed with NL4-3 for 2 h, washed and treated with trypsin to remove uninternalized virions, and either cultured alone or cocultured with autologous iDC at a 1:10 ratio for 72 h. Reverse transcription (RT) was assessed at various time periods by a real-time PCR assay which detects incomplete (R-U5) or complete HIV-1 reverse transcripts. Results are representative of three (A) or two (B) independent experiments with different donors.

To investigate whether a physical contact between the two cell types is required, resting CD4⁺ T cells were first exposed to HIV-1 and next cocultured with autologous iDC, either in the same well, allowing contacts between the two different cell types, or in a Transwell system in which both cell types were separated by a polycarbonate membrane (0.4 μm pore size). When the latter culture system was used, no significant viral production was detected (Fig. 3A), therefore suggesting that a contact is necessary between the two cell types. These data are in agreement with a previous study showing that HIV-1-infected DC can efficiently transfer the virus to T cells across a semipermeable membrane only if those CD4-expressing T cells were in contact with autologous DC (21).

FIG. 3.

iDC-mediated increase in virus production in resting CD4⁺ T cells requires physical contact, live iDC, and apoptosis. (A) Resting CD4⁺ T cells were infected with NL4-3 for 3 h, washed, and next either cocultured with autologous iDC or cultured separated from iDC in a Transwell device. Viral production was monitored by assessing the p24 content after 7 days. (B) Resting CD4⁺ T cells were inoculated with NL4-3 for 4 h, washed, and cocultured with iDC that had been previously either left untreated, subjected to a heat treatment (i.e., 60°C for 10 min), or fixed in 2% paraformaldehyde (PFA) for 30 min. Viral production was monitored by assessing the p24 content after 7 days. (C) Resting CD4⁺ T cells were infected either with single-cycle pseudotyped reporter viruses (left insert) or fully competent NL4-3 (right insert) for 2 h, washed, and next cocultured with autologous iDC at the indicated cell/cell ratio in the presence of ZVAD-fmk (25 μM) or a similar concentration of the diluent dimethyl sulfoxide (DMSO). HIV-1 expression was monitored at 72 h (luciferase-encoding virus) or 6 days (NL4-3) following initiation of the coculture. Results are representative of three (A and B) or five (C) independent experiments with different donors. Statistical significance was assessed using a Mann-Whitney test.

It is now well accepted that antigenic activation can be achieved through engagement of the T-cell receptor complex and costimulatory molecules such as CD28 and CD40L by their cognate ligands expressed on antigen-presenting cells. This can be mimicked using specific antibodies or paraformaldehyde-fixed cells expressing the necessary ligands such as HLA-DR, CD80, or CD86 (7, 16). To define whether live iDC are needed for the increase in HIV-1 replication, resting CD4⁺ T cells were exposed to HIV-1 and then cocultured either with live, heat-treated, or paraformaldehyde-fixed autologous iDC. Virus production was augmented with live iDC only (Fig. 3B), therefore indicating that contact with live iDC activates CD4⁺ T cells by a mechanism more complex than a simple receptor engagement since enhancement of virus replication is no longer seen with paraformaldehyde-fixed iDC. This hypothesis in reinforced by the observation that addition of antibodies directed against various cell surface adhesion and costimulatory molecules such as ICAM-1, ICAM-2, ICAM-3, LFA-1, CD28, CD80, CD86, CD40, and CD40L had no modulatory effect on the iDC-induced induction of HIV-1 expression in resting CD4⁺ T cells (data not shown).

The proliferative response of T cells upon coculture with autologous DC, macrophages, or B cells in the absence of any foreign antigen has been described as the autologous mixed-lymphocyte reaction (AMLR) (9, 18, 22). Studies on AMLR have established that it is an interleukin-2 (IL-2)-independent process that relies on HLA-DR, CD4, and some costimulatory molecules (10). Coculture with autologous iDC was shown to induce increased motility, small calcium mobilization, and upregulation of CD69 in CD4⁺ T cells (15). These responses indicate that cells leave the G₀ phase and enter the cell cycle, a process that is sufficient to relieve the block to HIV-1 reverse transcription and integration (11). More recent studies have indicated that cells undergoing apoptosis could be the source of autoantigens in AMLR (2, 6). During apoptosis, self-antigens are modified by the action of caspases and can be presented by DC to autoreactive T cells (1). To assess if apoptosis could be involved in the iDC-mediated increase in virus replication in resting CD4⁺ T cells, cocultures were initiated in the presence of the caspase inhibitor Z-Val-Ala-Asp-CH₂F (ZVAD-fmk). The iDC-dependent induction of HIV-1 LTR-driven gene expression in resting CD4⁺ T cells was significantly but not completely inhibited in the presence of ZVAD-fmk (Fig. 3C, left insert). Similar results were obtained when using replication-competent NL4-3 (Fig. 3C, right insert), and no cell toxicity was observed at the concentration used (i.e., 25 μM) (data not shown). It can be proposed that presentation of self-antigens derived from apoptotic cells is partially responsible for the observed phenomenon.

In conclusion, we have established that an intimate contact between uninfected autologous iDC and virus-infected quiescent CD4⁺ T cells is sufficient to trigger an antigen-independent increase in virus gene expression partly through presentation of apoptosis-induced self-antigens. In vivo, this phenomenon could be involved in the infection of quiescent CD4⁺ T cells, especially in the gut during acute infection, as well as in the formation and maintenance of the reservoir composed of latently infected resting memory CD4⁺ T cells. The significance of our observations is provided by the fact that such long-lived cellular reservoirs can persist for years despite effective drug regimens and appear to be the major barrier to eradication of HIV-1.