Multiple Effector Functions Mediated by Human Immunodeficiency Virus-Specific CD4⁺ T-Cell Clones

Study subjects.

Four persons with vigorous p24-specific T helper cell proliferative responses were selected for study. Subject CTS-01 is a 50-year-old African-American male infected with HIV-1 for at least 20 years. Without antiretroviral therapy his viral load has been always less than 1,000 RNA copies/ml and his CD4⁺ T-cell count above 500 cells/ml. Subjects AC-01, AC-25, and AC-36 were treated with antiretroviral therapy during acute HIV-1 infection (43), and clones were isolated 11 to 18 months after initiation of therapy. Clones from AC-01 and AC-36 were isolated before a supervised therapy interruption and from AC-25 after treatment interruption and reinstitution. Thirty-five HIV-1-seronegative individuals' peripheral blood mononuclear cells (PBMC) were used as controls for proliferative responses to p24 protein (43).

Peptides and antibodies.

Recombinant p24 protein (amino acids 133 to 373) derived from the NY-5 strain of HIV-1 was produced in a baculovirus expression system with 90 to 95% purity (Protein Science, Meriden, onn.T). Shorter p24 peptides were generated as free acids with an Advanced ChemTech 396Ω peptide synthesizer (44). Flow cytometry antibodies were obtained from Becton Dickinson (San Jose, Calif.).

T-cell clones.

Culture media (R+) consisted of RPMI 1640 (Sigma, St. Louis, Mo.) with penicillin-streptomycin (Mediatech, Herndon, Va.), HEPES (Mediatech), and l-glutamine (Mediatech). T-cell clones were maintained in R+ and 10% heat-inactivated human AB serum (R10H). Clones were generated by limiting dilution. Freshly isolated PBMC (10⁷) were suspended in 10 ml of R10H in a T25 flask and stimulated with p24 (1 μg/ml) and IL-2 (100 U/ml; Hoffmann-La Roche). Indinavir (Merck, 0.4 μM), zidovudine (AZT; Glaxo Wellcome; 0.5 μM), and lamivudine (Glaxo Wellcome; 3 μM) were added to the media for the first 4 weeks of culture. After 2 weeks the PBMC were restimulated with p24 protein (1 μg/ml), IL-2 (100 U/ml), and 10⁷ irradiated (30 Gy), autologous PBMC. Three days after restimulation the PBMC were plated at limiting dilution with 10, 3, or 1 cell per well. Clones were screened in a 2-day proliferation assay using autologous B-lymphoblastoid cell lines (B-LCL) as antigen-presenting cells and p24 protein as the stimulus. Clones that proliferated in response to p24 protein were restimulated every 2 to 3 weeks with an anti-CD3-specific antibody 12F6 (obtained from Johnson Wong, Massachusetts General Hospital), IL-2 (100 U/ml), and 10⁷ irradiated allogeneic PBMC.

TCR Vβ chain analysis.

In order to determine if the T-cell lines were monoclonal, T-cell receptor (TCR) Vβ expression was determined by an established method (12). Twenty-four sets of primers specific for TCR Vβ RNA were used. PBMC (10⁵) were used as positive controls, and distilled water was used for negative controls.

Proliferation assays.

To assay T helper cell responses in fresh PBMC, 10⁵ cells were incubated with p24 protein at 1 μg/ml for 6 days and then pulsed with [³H]thymidine for 6 h before harvesting as previously described (44). To test T-cell clones, antigen was presented by autologous or partially HLA matched B-LCL. The B-LCL were irradiated (120 Gy) and incubated overnight in R10H with the appropriate antigen at 1 μg/ml. The following day B-LCL and T-cell clones were plated in triplicate wells at 50,000 cells/well each in 96-well plates in R10H. After 48 h, 1 μCi of [³H]thymidine (Dupont NEN, Boston, Mass.) in 50 μl of R10H was added per well. Plates were harvested onto glass fiber filters after 18 h. Results were expressed as stimulation index, the ratio of counts from wells with antigen divided by the counts obtained from wells without antigen. Alternately, results were shown as net counts per minute, the difference between the counts. Based on previous studies of HIV-1-specific proliferative responses, stimulation indexes greater than 5 were considered significant (44).

IFN-γ Elispot assays.

Elispot plates were coated with anti-IFN-γ antibody (Endogen, Woburn, Mass.) and incubated overnight at 4°C. The following day plates were washed six times with phosphate-buffered saline (PBS; Mediatech). B-LCL (5 × 10⁴) and antigen (1 μg/ml) were added in 100 μl of R10H. T-cell clones were added at 100 cells/well in 100 μl of R10H. After overnight incubation the cells were discarded and plates were washed six times with PBS, and biotinylated anti- IFN-γ (Endogen) was added for 1.5 h at 25°C. The plates were then washed six times with PBS. Streptavidin (Bio-Rad, Hercules, Calif.) (100 μl/well) was added, and plates were incubated 45 min at 25°C. After six more PBS washes 100 μl/well of coloring reagent (nitro blue tetrazolium–5-bromo-4-chloro-3-indolylphosphate [NBT/BCIP]; Bio-Rad) was added. Once dots appeared, the reaction was stopped with three water washes. Background responses to wells with no antigen or irrelevant antigen ranged from 0 to 1.5 spots per well, or 0 to 15,000 spot-forming cells (SFC) per 10⁶ cells. Responses greater than 50,000 SFC per 10⁶ cells and 5 times maximum background were considered significant.

Cytotoxicity assays.

Target cells (B-LCL) were incubated overnight with whole p24 or the cognate 22-amino-acid peptide at 1 μg/ml and Na₂⁵¹CrO₄ (60 μCi/ml; Dupont NEN). The following day the targets were washed twice with cold R10H before effectors were added. Effectors and targets were incubated together in triplicate wells for 4 h at 37°C, then the plates were centrifuged at 1,000 rpm for 5 min at 4°C. A 25-μl aliquot of reaction supernatant was assayed for ⁵¹Cr release. Maximal release was obtained by mixing the targets with 100 μl of 1% Triton X-100. The percent specific ⁵¹Cr release was calculated as 100 × [(cpm experimental release − cpm spontaneous release)/(cpm maximal release − cpm spontaneous release)]. Some lysis inhibition experiments were performed in the presence of 2 mM EGTA and MgCl₂. These reagents chelate extracellular calcium and were added at the time clones and targets were combined. We also performed cytolytic assays in the presence of inhibitors with concentrations varying from 0 to 10 μg/ml of concanamycin A (Sigma), brefeldin A (Sigma), or anti-Fas antibody (Coulter Immunotech, Miami, Fla.). The effector cells were incubated with inhibitors for 2 h prior to plating with B-LCL, and the inhibitors remained present throughout the assay. Spontaneous lysis was less than 30% unless otherwise noted. For the purpose of interpretation, specific lysis of greater than 10% was considered significant.

Lytic granule release assay.

To quantify lytic granule release, autologous B-LCL were pulsed with various concentrations of cognate peptide for 1 h at 37^oC and combined with CD4⁺ T cells at an effector-to-target cell (E:T) ratio of 1:1 in round-bottomed 96-well plates. The density of the CTL in the mixture was 7 × 10⁵ per ml (1.5 × 10⁵ per well). After 4 h at 37^oC the culture supernatant was collected, and activity of serine esterase, an essential component of the secreted granules, was measured using the BLT (N-α-benzyloxycarboxyl-l-lysinethiobensyl ester) substrate in the presence of DTNB [5,5′-dithio-bis(2-nitrobenzoic acid)] by reading the optical density (OD) at 405 nm (method communicated by Pierre Henkart). In our experiments DTNB was added prior to the addition of BLT to measure the background OD at 405 nm. The background was then subtracted from the final reading for each sample.

Intracellular staining for IFN-γ production.

Quantitation for intracellular production of IFN-γ upon antigen stimulation was performed as described elsewhere (51). Briefly, 5 × 10⁵ clone cells were incubated with B cells pulsed with peptide at 5 μg/ml and anti-CD28 and anti-CD49 antibodies at 1 μg/ml each for 2 h at 37C at a 5° slant in 1 ml of R10H in a fluorescence-activated cell sorting (FACS) tube. At 2 h brefeldin A was added at 10 μg/ml, and incubation was continued as above for 4 more h. The samples were then refrigerated overnight. The next day the samples were washed in PBS plus 1% fetal calf serum, then stained with surface antibodies for 20 min. After washing in PBS the cells were fixed and permeabilized with two 15-min incubations with Caltag reagents according to the manufacturer's instructions. The cells were then stained with intracellular IFN-γ–fluorescein isothiocyanate (FITC), washed twice in PBS, and analyzed on the Becton Dickinson FACS machine. For the concanamycin A inhibition experiments, the cloned cells were first incubated with concanamycin A for 1 h prior to the addition of B-LCL and antigen. At least 50,000 total events were counted for each condition.

Proliferative responses of PBMC to p24 Gag protein.

Previous studies have shown dominant HIV-1-specific T helper cell responses directed against the p24 Gag protein and that these responses are inversely correlated to HIV-1 viral load (22, 44). To determine the breadth of this response, overlapping peptides were used to map discrete targeted regions. An example is shown for subject CTS-01, an individual with long-term nonprogressive HIV-1 infection and a robust p24-specific T helper cell response. The dominant epitopes targeted were defined using synthetic 22-amino-acid peptides overlapping by 12 amino acids and spanning the p24 protein. Significant peptide-specific proliferative responses were detected to 3 of the 23 peptides tested (Fig. 1A), indicating a polyclonal response. The dominant response was directed against peptide 9, sequence DRVHPVHAGPIAPGQMREPRGS (44). Furthermore, responses of 10 HIV-1-seronegative controls to the panel of proteins were negligible (Fig. 1B).

Generation of CD4⁺ T-cell clones and mapping.

In order to define functional characteristics of T helper cell responses and fine map the epitopes targeted, clones were established. Limiting-dilution cloning of PBMC from subject CTS-01 was performed to derive 10 T-cell lines that proliferated to p24 protein. The fine specificity of the proliferative response was mapped using a set of overlapping p24 peptides. All T-cell lines but one were found to proliferate exclusively in response to peptide 9, DRVHPVHAGPIAPGQMREPRGS, with stimulation indices ranging from 25 to 230, net cpm 1,675 to 9,530 (data not shown). Clonality was determined by RT-PCR for Vβ region T-cell receptor usage. Clone 16 was selected for in-depth analysis because it expressed T-cell receptor Vβ4 exclusively and was presumed to be clonal (Fig. 2A). The remaining T-cell lines all expressed Vβ4 but also expressed one to three additional Vβ regions and are likely not clonal. T-cell lines were all functionally clonal and are referred to as clones in the remainder of the article. Each of the T-cell clones was >99.2% CD4⁺, and the majority were less than 0.5% CD8⁺ by FACS analysis (Fig. 2B).

p24-specific cytotoxic responses mediated by CD4⁺ CD8⁻ T-cell clones.

CD4⁺ T cells are classically thought to assist in antiviral control by providing help to both CD8⁺ CTL and B cells. CD4⁺ T cells have also been shown to secrete a number of antiviral chemokines, such as IFN-γ, RANTES, MIP-1α, and MIP-1β. We queried whether or not these CD4⁺ T-cell clones would also possess direct cytolytic function. The 10 CD4⁺ T cell clones derived from subject CTS-01 were assessed for their ability to lyse B-LCL incubated with peptide 9. All 10 clones assayed lysed autologous B-LCL incubated with peptide 9 at an E:T ratio of 100:1 (17 to 68% lysis), but not B-LCL incubated without peptide or with an irrelevant control peptide (Fig. 3). Seven of 10 clones also exhibited specific lysis in response to whole p24 protein at an E:T ratio of 100:1 (13 to 61% lysis; Fig. 3).

Fine epitope mapping of proliferative, cytotoxic, and IFN-γ-secreting responses.

T helper cell responses are typically targeted at exogenously processed proteins 10 to 17 amino acids in length. Truncations of peptide 9 were synthesized to define the minimal epitope recognized by the clones from subject CTS-01. Three clones were examined, and data for one representative clone are shown. All three clones recognized the same minimal 9-amino-acid peptide, VHAGPIAPG (Fig. 4A). Deletion of either the N-terminal valine or C-terminal glycine resulted in complete abrogation of the proliferative response. However, longer peptides containing these residues were all recognized, and some were associated with stronger proliferative responses.

We next compared proliferation to cytolytic activity and IFN-γ secretion using these truncated peptides. Fine mapping of the cytolytic response revealed the same minimal epitope (Fig. 4B). The clone was found to secrete IFN-γ in response to stimulation by B-LCL pulsed with cognate peptide (2.8 × 10⁵ SFC/10⁶; Fig. 4C), but not by B-LCL incubated with an irrelevant peptide (0 to 1.5 × 10⁴ SFC/10⁶; data not shown). The minimal epitope for stimulation of IFN-γ secretion was identical to the minimal epitope required to induce proliferation and cytolysis, VHAGPIAPG.

Clonal responses were DQ7 restricted.

The HLA restriction of clones targeting the minimal peptide VHAGPIAPG was determined using partially matched B-LCL as antigen-presenting cells in proliferation, cytolytic, and IFN-γ Elispot assays. B-LCL that could present antigen effectively in the three assays all had allele DQ7, while those lacking DQ7 could not present antigen effectively (Fig. 5, left panel). Multiple other B-LCL were assayed, and the results consistently identified DQ7 as the only allele associated with effective antigen presentation (Fig. 5 and data not shown). These data thus confirm that the effector functions mediated by these HIV-1-specific T helper cell clones are HLA class II restricted.

CD4⁺ T-cell mediated cytolysis was seen in multiple individuals.

We explored whether CD4⁺ T-cell clones from other HIV-1-infected individuals would also exhibit antigen-specific cytolytic activity. We derived clones from three individuals who were diagnosed and treated with highly active antiretroviral therapy during acute HIV-1 infection who generated vigorous T helper cell responses to p24 antigen (43, 44). Three clones were derived from one individual, two from a second, and one from a third. Each clone recognized a different epitope within p24 as mapped by overlapping 22-amino-acid peptides (Fig. 6A). Clonality was determined by RT-PCR for Vβ T-cell receptor usage as described previously. Clones 1 to 3 and 6 each expressed a unique T-cell receptor Vβ region, clone 4 expressed two Vβ T-cell receptor products, and none of the tested Vβ T-cell receptors was identified for clone 5. The six clones were tested for p24-specific cytolytic activity at an E:T ratio of 100:1 (Fig. 6B). All six clones exhibited cytolytic activity, confirming that the observation of cytolytic CD4⁺ T-cell clones was not limited to one individual. The six clones were also tested for cytolytic activity directed against the 22-amino-acid peptide each was found to recognize in a proliferation assay. Four of the six clones maintained levels of specific lysis greater than 25% at an E:T ratio of 10:1 (Fig. 6C). These results provide evidence at a clonal level that HIV-1-specific T helper cells possess multiple effector functions, including proliferation, IFN-γ secretion, and cytolysis.

Cytolytic activity occurred at doses of peptide similar to those required to induce proliferation and IFN-γ production.

The in vivo significance of these cytolytic T helper cell clones is unknown. In order to determine if the cytolytic function of these clones was artificially induced by concentrations of antigen much higher than those required for proliferation and IFN-γ production, we performed a peptide titration. Figure 7A demonstrates that a peptide concentration of 1 μg/ml was able to elicit proliferation, IFN-γ production, and cytolysis in the clone from subject CTS-01, implying that the cytolysis observed was not merely an artifact due to unusually high antigen concentrations. Identical results were obtained with a clone from subject AC-25 (data not shown).

Lysis was mediated by the perforin pathway.

The finding that these Gag-specific T helper cell clones mediated cytolytic activity prompted us to examine the mechanism of cell lysis. We focused our attention on clone 16 from subject CTS-01 and clone 4 from subject AC-25. The two main mechanisms of CD4⁺ T-cell cytotoxicity that have been reported include perforin-mediated and Fas-mediated lysis (21, 32). Additionally, CD4⁺ CTL have been shown to secrete IFN-γ, granzyme A, and tumor necrosis factor (TNF)-α and -β (11, 29, 47, 54). Serine esterase granule secretion was measured after specific stimulation in the clone from subject AC-25 (Fig 7B). Peptide 4 from p24 was able to elicit serine esterase release as measured by conversion of the substrate BLT, implying that the clones contained and could release lytic granules. Furthermore, chelation of extracellular calcium with EGTA has been used as a method of selectively blocking perforin-mediated lysis. We examined lysis mediated by the AC-25 clone in the presence of EGTA and found it to be abrogated, implying that the clone utilized the perforin pathway of lysis (Fig. 7C). Perforin staining of the clone from AC-25 was performed and was 6 to 8 times the background staining of B-LCL or staining with control immunoglobulin (data not shown).

Concanamycin A and brefeldin A have been used as inhibitors of perforin-mediated and Fas-mediated cytotoxicity, respectively, to differentiate between these two mechanisms of lysis (24). Concanamycin A, an inhibitor of vacuolar-type H⁺-ATPase, accelerates degradation of perforin by increasing the pH of lytic granules. Brefeldin A inhibits intracellular glycoprotein transport and was shown to selectively block Fas-mediated killing. Anti-Fas antibody can also block Fas-mediated killing (55). We found that concanamycin A could abrogate lysis in a dose-dependent fashion in the clone from subject CTS-01 (Fig. 8A). Brefeldin A and anti-Fas antibody over the same concentration range showed no effect on cytolytic activity. To demonstrate that the clones retained functions other than perforin secretion in the presence of concanamycin A, we quantitated IFN-γ production in the presence of concanamycin A in the clone from subject AC-25 (Fig. 8B). At the highest concentration of concanamycin A, IFN-γ production was abrogated. However, IFN-γ production was unaffected over a range of concentrations of concanamycin A that resulted in complete abrogation of lysis (Fig. 8C), supporting the notion that perforin secretion was selectively inhibited by concanamycin A. These results imply that the Gag-specific T helper cell responses utilized a perforin-dependent pathway in killing targets.