Increased Expression of Human T Lymphocyte Virus Type I (HTLV-I) Tax11-19 Peptide–Human Histocompatibility Leukocyte Antigen A*201 Complexes on CD4⁺ CD25 ⁺ T Cells Detected by Peptide-specific, Major Histocompatibility Complex–restricted Antibodies in Patients with HTLV-I–associated Neurologic Disease

The HTLV-I is an exogenous human retrovirus that infects ∼10–20 million people worldwide (1). The majority of infected individuals remain healthy lifelong asymptomatic carriers while ∼0.25–3% develop an inflammatory disease of the central nervous system termed HTLV-I–associated myelopathy/tropical spastic paraparesis (HAM/TSP; 2–5). HTLV-I is also the etiologic agent in adult T cell leukemia and other inflammatory diseases including uveitis, arthritis, polymyositis, Sjögren syndrome, and alveolitis (6).

In patients with HAM/TSP, increased HTLV-I provirus load and augmented immune responses to HTLV-I have been reported. One of the most striking features of the cellular immune response in HAM/TSP patients is the highly increased numbers of HTLV-I–specific CTLs, which are lower or absent in asymptomatic carriers (7–11). CTL activity is predominantly restricted to the HTLV-I Tax protein, where in HLA-A*201 patients the HTLV-I Tax11-19 peptide (LLFGYPVYV) is defined as an immunodominant epitope (12, 13). In HAM/TSP patients these CTLs have been shown to produce IFN-γ and TNF-α that are proinflammatory and neurotoxic cytokines (9, 14). Moreover, in some HLA-A*201 HAM/TSP patients, the frequency of Tax11-19–specific CTLs can be as high as 30% of all CD8⁺ T cells in peripheral blood (11) and even higher in cerebrospinal fluid (8, 15, 16). Neuropathological findings in HAM/TSP have demonstrated that CD8⁺ T cells were frequently observed in active chronic lesions (17–19). Collectively, these findings support the view that HTLV-I Tax–specific CTLs play a central role in the immunopathogenesis of HAM/TSP.

Therefore, it was of interest to address which cells are stimulating the expansion of these pathogenic HTLV-I–specific CD8⁺ CTLs in vivo in HAM/TSP patients. It has been reported that HTLV-I infects both memory (CD45RO⁺) CD4⁺ T cells (20) and effector/memory (CD27⁻ CD45RA⁻) CD4⁺ T cells in vivo (21), although recent work indicated that CD8⁺ T cells were also an in vivo cellular reservoir for HTLV-I (22, 23). Because a significant positive correlation between HTLV-I proviral load and the proportion of CD4⁺ CD25⁺ T cells has been demonstrated in PBMCs of HTLV-I–infected individuals (24), it has been hypothesized that CD4⁺ CD25⁺ T cells may also be infected with HTLV-I (25).

Virus-infected cells are known to present viral peptides on HLA class I molecules that are recognized by virus-specific CD8⁺ T cells through their antigen-specific TCR (26). Because the frequency of HTLV-I Tax–specific CD8⁺ CTLs in peripheral blood is proportional to the amount of HTLV-I proviral DNA load and the levels of HTLV-I tax mRNA expression (10, 11, 27), it has been suggested that the stimulus for these HTLV-I Tax–specific CD8⁺ CTLs are immunodominant Tax peptide–HLA complexes expressed on autologous HTLV-I–infected T cells. However, because there has never been a direct way to analyze such peptide–HLA complexes, it is unclear as to which subset of T cells dominantly express the HTLV-I Tax peptide–HLA complexes and at what frequency in HAM/TSP. Recently, using a large human Ab phage display library, novel human recombinant Fab Abs have been isolated that specifically bind to a particular peptide–HLA-A*201 complex like TCR (28, 29). The use of these Abs allowed, for the first time, the identification and/or quantification of peptide–HLA complexes on APCs. Using these newly described peptide-specific, MHC-restricted Abs, which specifically bind to Tax11-19 peptide–HLA-A*201 complexes (TaxA2-Ab; 29), we measured the level of Tax11-19 peptide–HLA-A*201 complexes in peripheral blood T cells from HAM/TSP patients. We analyzed which subset of T cells are predominantly expressing these complexes and correlated the HTLV-I proviral DNA load of different T cell subsets using the quantitative PCR technique. In addition, we tested the capability of the Tax11-19 peptide–HLA-A*201–expressing HTLV-I–infected T cells to stimulate the proliferation of HTLV-I–specific CD8⁺ T cells. The data presented here demonstrate that CD4⁺ CD25⁺ T cells are the major reservoir of HTLV-I provirus and predominantly express HTLV-I tax mRNA as well as HTLV-I Tax peptide–HLA complexes that stimulate and expand HTLV-I Tax–specific CD8⁺ T cells. To our knowledge, this is the first report that directly measured peptide–HLA complexes ex vivo using these unique peptide-specific, MHC-restricted Fab Abs.

We used Ficoll-Hypaque (BioWhittaker) centrifugation to separate PBMCs from six HLA-A*201⁺ HAM/TSP patients (HAM 1-6) and six HLA-A*201⁺ HTLV-I–seronegative healthy donors (A2HD). HAM/TSP was diagnosed according to the World Health Organization's guidelines. Blood samples were obtained after informed consent as part of a clinical protocol reviewed and approved by the National Institutes of Health institutional review panel. HTLV-I infection was confirmed by ELISA (Abbot Laboratories) and Western blot analysis (Genelabs). CD4⁺ T cells or CD8⁺ T cells were negatively selected from PBMCs with magnetic beads (Miltenyi Biotec), and CD25⁺ and CD25⁻ T cells or CD45RO⁺ and CD45RO⁻ T cells were separated from the selected CD4⁺ or CD8⁺ T cells using magnetic beads (Miltenyi Biotec).

The immortalized B cell line transfected with a full-length HLA-A*201 cDNA (HmyA2.1; reference 30) was used to determine the reactivity of TaxA2-Ab and a peptide-specific, MHC-restricted Ab specific for the melanoma gp100 G9-154 peptide–HLA-A*201 complex (MelanomaA2-Ab) with cell surface–expressed peptide–HLA-A*201 complexes. HmyA2.1 cells were pulsed with Tax11-19 peptide (LLFGYPVYV) or melanoma gp100 G9-154 peptide (KTWGQYWQV) at the indicated concentrations for 1 h at 37°C in DMEM medium (GIBCO BRL) supplemented with 10% fetal bovine serum (Atlanta Biological), 2 mM l-glutamine, 40 U/ml penicillin, 40 μg/ml streptomycin (all from BioWhittaker). The cells were then washed once to remove any free peptide and stained with 1 μg PE-labeled TaxA2-Ab or MelanomaA2-Ab for 1 h at 4°C, and then washed once before analysis. PBMCs from A2HD and HAM/TSP patients were also analyzed using the same procedure. For peptide pulse analysis of peripheral blood T cells, purified CD4⁺ or CD8⁺ T cells were incubated with peptides at the indicated concentrations for 1 h at 37°C in RPMI medium (GIBCO BRL) supplemented with 10% fetal bovine serum, 2 mM l-glutamine, 40 U/ml penicillin, 40 μg/ml streptomycin. The cells were stained with 1 μg PE-labeled TaxA2-Ab or MelanomaA2-Ab for 1 h, and then stained with Tricolor-labeled monoclonal Ab against CD4 and allophycocyanin-labeled monoclonal Ab to CD8 (all from Caltag) for the last 30 min at 4°C. For analyzing the level of HLA-A*201 expression, CD4⁺ T cells from A2HD and HAM/TSP patients were stained with 1 μg anti–HLA-A*201 Ab (BB7.2; reference 31) for 30 min at 4°C. After washing, the cells were stained with PE-labeled anti–mouse IgG1Ab (Southern Biotechnology Associates, Inc.) for 30 min at 4°C. For phenotypic analysis, purified CD4⁺ T cells by magnetic beads from HAM/TSP patients were cultured for 24 h at 37°C in RPMI medium supplemented with 10% fetal bovine serum, 2 mM l-glutamine, 40 U/ml penicillin, 40 μg/ml streptomycin. The cells were then stained with 1 μg PE-labeled TaxA2-Ab for 1 h, and stained with FITC-labeled monoclonal Ab against CD25, CD45RO, CD45RA (all from Caltag), or CD27 (BD Biosciences) for the last 30 min at 4°C. All cells were washed twice before analysis on a FACSCalibur™ (Becton Dickinson). Data were analyzed with CELLQuest™ software (Becton Dickinson).

CD4⁺ CD25⁺ and CD4⁺ CD25⁻ cells were separated from HAM/TSP patients and cultured for 24 h, and the cells were irradiated (3,000 rad). 10⁵ autologous CD8⁺ T cells were then mixed with 10⁵ of the irradiated CD4⁺ CD25⁺ or CD4⁺ CD25⁻ cells and cultured for 4 d in round-bottom, 96-well culture plates. Incorporation of [³H]thymidine (1 μCi/well) by proliferating cells was measured during the last 16 h of culture. In addition, proliferation of HTLV-I Tax–specific T cells and CMV pp65–specific T cells was analyzed using flow cytometry. The cultured cells were stained with PE-labeled Tax11-19 peptide–loaded HLA-A*201 tetramer (provided by National Institutes of Health AIDS Research and Reference Reagent Program) or CMV pp65 peptide–loaded HLA-A*201 tetramer (Beckman Coulter) for 30 min at 4°C. The cells were stained with Tricolor-labeled anti-CD4 Ab and APC-labeled anti-CD8 Ab (Caltag) for the last 15 min at 4°C. The cells were then fixed and permeabilized with Cytofix/CytoPerm kit (BD PharMingen), and stained with FITC-conjugated anti-Ki67 Ab (BD Biosciences). All cells were washed twice before analysis on a FACSCalibur™ (Becton Dickinson).

HTLV-I proviral DNA load and HTLV-I tax mRNA load were measured using an ABI PRISM 7700 Sequence Detector (Applied Biosystems) as previously described (10, 11). DNA was extracted using Puregene DNA Isolation Kit (Gentra) and 100 ng sample DNA solution per well was analyzed. The HTLV-I proviral DNA load was calculated by the following formula: copy number of HTLV-I (pX) per 100 cells = (copy number of pX) / (copy number of β actin/2) × 100. RNA was extracted using RNeasy Mini Kit (QIAGEN) according to the manufacturer's instructions. cDNA was synthesized using TaqMan Gold RT-PCR Kit (Applied Biosystems). Sample cDNA from 300 ng RNA per well was applied and analyzed. We used the human housekeeping gene hypoxanthine ribosyl transferase (HPRT) primers and probe set (Applied Biosystems) for internal calibration. The relative HTLV-I tax mRNA load was calculated by the following formula: HTLV-I tax mRNA load = (value of tax) / (value of HPRT) × 10,000.

To demonstrate the specificity of peptide-specific, MHC-restricted Fab Abs that bind to Tax11-19 peptide–HLA-A*201 complexes (TaxA2-Ab), a human immortalized B cell line expressing HLA-A*201 (HmyA2.1) was pulsed with 10 μM HTLV-I Tax11-19 peptide or control melanoma gp100–derived G9-154 peptides, and incubated with PE-labeled TaxA2-Ab. The ability of TaxA2-Ab to bind to Tax11-19 peptide–HLA-A*201 molecule was then monitored by flow cytometry. As shown in Fig. 1 A, TaxA2-Ab reacted only with Tax11-19 peptide–loaded HmyA2.1 cells, but not with cells loaded with the control peptide. Conversely, a peptide-specific, MHC-restricted Fab Ab specific for the melanoma gp100 G9-154 peptide–HLA-A*201 complex (MelanomaA2-Ab) specifically bound HmyA2.1 cells pulsed with the melanoma gp100 peptide, but not Tax11-19 (Fig. 1 B). Peptide titration studies demonstrated that the level of TaxA2-Ab staining on HmyA2.1 cells correlated with the concentration of loaded Tax11-19 peptide (Fig. 1 C).

As these peptide-specific, MHC-restricted Fab Abs have never been used for the staining of ex vivo human T lymphocyte samples, we validated the ability of these Abs to detect peptide–HLA complexes using ex vivo T cells from HLA-A*201 HTLV-I–seronegative healthy donors (A2HD) pulsed with Tax11-19 peptide. CD4⁺ and CD8⁺ T cells were purified using magnetic beads from A2HD PBMCs, pulsed with or without Tax11-19 peptide, and stained with TaxA2-Ab followed by flow cytometry analysis. As shown in Fig. 1, D and E, TaxA2-Ab could detect the Tax11-19 peptide–HLA-A*201 complexes on A2HD CD4⁺ T cells (Fig. 1 D) as well as on CD8⁺ T cells (Fig. 1 E) when pulsed with 100 μM Tax11-19 peptide, but not with control G9-154 peptide. The binding of TaxA2-Ab to Tax11-19 peptide pulsed A2HD T cells were also peptide dose dependent (Fig. 1 F).

Using this TaxA2-Ab, the expression of Tax11-19 peptide–HLA-A*201 complexes was analyzed on T cells from HAM/TSP patients. It has been reported that HTLV-I infects both CD4⁺ and CD8⁺ T cells (20, 22). Although HTLV-I antigen expression ex vivo is negligible, infected cells can start to express HTLV-I antigen after short-term culture (32). Therefore, CD4⁺ and CD8⁺ T cells were separated using MACS beads and the expression of Tax11-19 peptide–HLA-A*201 complexes on each T cell subset before and after culture was investigated using the TaxA2-Ab. Before culture, the Tax11-19 peptide–HLA-A*201 complexes could not be detected both on ex vivo CD4⁺ and CD8⁺ T cells (Fig. 2 A). However, after 24 h of culture, the expression of Tax11-19 peptide–HLA-A*201 complexes was demonstrated on CD4⁺ T cells, but not on CD8⁺ T cells (Fig. 2 A). These results indicate that the peptide-specific, MHC-restricted Fab Abs are capable of detecting the specific peptide–HLA complexes after natural endogenous intracellular antigen processing as previously reported (29), and in HAM/TSP patients CD4⁺ T cells present more Tax11-19 peptide–HLA-A*201 complexes than CD8⁺ T cells.

As there are no direct methods for the enumeration and phenotyping of individual cells bearing physiological levels of peptide–HLA complexes in mixed cell populations, the peptide-specific, MHC-restricted Fab Abs are ideally suited to conduct such an analysis. Therefore, to investigate which subset of CD4⁺ T cells preferentially express endogenous Tax11-19 peptide–HLA-A*201 complexes, the TaxA2-Ab staining on 24-h cultured CD4⁺ T cells was analyzed in combination with flow cytometric staining of cell surface antigens (CD25, CD27, CD45RO, and CD45RA). As shown in Fig. 2 B, Tax11-19 peptide–HLA-A*201 complexes were preferentially expressed on CD25⁺, CD27⁻, CD45RO⁺, CD45RA⁻ cells (i.e., phenotypically defining an activated, effector/memory population). Among these cell surface markers, CD25 was shown to be the best marker to discriminate Tax11-19 peptide–HLA-A*201 complex–positive cells from –negative cells. Identical results were obtained from all six HAM/TSP patients tested (not depicted).

Increased detection of Tax11-19 peptide–HLA-A*201 complexes on CD4⁺ CD25⁺ T cells from HAM/TSP patients suggested two possibilities: HTLV-I preferentially infects CD4⁺ CD25⁺ T cells, or HTLV-I–infected T cells expressed CD25 during short-term culture as previously reported (33). Although these two possibilities are not mutually exclusive, to address whether CD4⁺ CD25⁺ T cells are preferentially infected with HTLV-I, we separated peripheral blood T cells from HAM/TSP patients into CD4⁺ CD25⁺, CD4⁺ CD25⁻, CD4⁺ CD45RO⁺, CD4⁺ CD45RO⁻, CD8⁺ CD25⁺, CD8⁺ CD25⁻ T cells using MACS beads, and measured HTLV-I proviral DNA load within each population using a real-time quantitative PCR method (TaqMan). As shown in Fig. 3 A (means of six HAM/TSP patients), HTLV-I proviral DNA was highest in CD4⁺ CD25⁺ T cells, although HTLV-I also showed a tropism for memory CD4⁺ T cells (CD4⁺ CD45RO⁺) and is consistent with a previous report (20). This preferential infection for CD25⁺ population was also observed in CD8⁺ T cells (Fig. 3 A).

These results directly demonstrated that CD4⁺ CD25⁺ T cells are an important virus reservoir. To investigate how much of the virus load of PBMCs was accounted for by CD4⁺ CD25⁺ T cells, HTLV-I proviral DNA load in PBMCs and the percentage of CD4⁺ CD25⁺ T cells in PBMCs in each HAM/TSP patient were analyzed (Fig. 3 B). Surprisingly, virus load and percentage of CD4⁺ CD25⁺ T cells were similar in HAM/TSP patients with medium to low virus load (HAM 3–6). Furthermore, in the patients with high virus load (HAM 1 and 2), the number of virus-infected cells (virus load) was demonstrated to be higher than the percentage of CD4⁺ CD25⁺ T cells, suggesting that other T cell populations were also infected with HTLV-I in the patients with high virus load. As previously reported (24), a positive correlation between HTLV-I proviral load and the proportion of CD4⁺ CD25⁺ T cells has also been observed.

Although both CD4⁺ CD25⁺ and CD4⁺ CD25⁻ T cells were infected with HTLV-I (Fig. 3 A), it was shown that CD4⁺ CD25⁺ T cells preferentially express Tax11-19 peptide–HLA-A*201 complexes (Fig. 2 B), suggesting that HTLV-I–infected CD4⁺ CD25⁺ T cells inherently express virus relative to HTLV-I–infected CD4⁺ CD25⁻ cells. To test this possibility, CD4⁺ CD25⁺ and CD4⁺ CD25⁻ T cells were isolated from peripheral blood of five HAM/TSP patients, cultured for 24 h, and the level of HTLV-I tax mRNA expression in each population was compared using real-time quantitative RT-PCR (Fig. 3 C). Interestingly, CD4⁺ CD25⁺ T cells expressed extremely high HTLV-I tax mRNA levels (approximately eight times) compared with CD4⁺ CD25⁻ T cells. These results are consistent with the observations that in HAM/TSP, CD4⁺ CD25⁺ T cells are the main reservoir for HTLV-I and express endogenous virus mRNA and Tax11-19 peptide–HLA-A*201 complexes.

Having observed that CD4⁺ CD25⁺ T cells are preferentially infected with HTLV-I and express Tax11-19 peptide–HLA-A*201 complexes, we tested whether these cells can stimulate the proliferation of autologous CD8⁺ T cells (and more specifically, HTLV-I–specific CD8⁺ T cells) in HAM/TSP patients. Peripheral blood T cells from three HAM/TSP patients were separated into CD4⁺ CD25⁺ and CD4⁺ CD25⁻ cells and cultured for 24 h. After culture, both populations were irradiated. Autologous CD8⁺ T cells were then added at a 1:1 ratio where proliferation of CD8⁺ T cells was analyzed after 4 d of culture. As shown in Fig. 4 A, CD4⁺ CD25⁺ T cells from HAM/TSP patients preferentially stimulated the proliferation of autologous CD8⁺ cells.

To determine whether proliferation of HTLV-I–specific CD8⁺ T cells were MHC restricted and HTLV-I specific, intracellular Ki-67 staining combined with HTLV-I Tax11-19–HLA-A*201 tetramer or CMV pp65–HLA-A*201 tetramer staining was performed in a HAM/TSP patient who harbored detectable levels of both HTLV-I and CMV tetramer⁺ CD8⁺ T cells. As shown in Fig. 4 C, when HAM/TSP patient CD8⁺ T cells were cultured with autologous CD4⁺ CD25⁺, the majority of HTLV-I Tax11-19–specific CD8⁺ T cells (90.1%) was shown to be proliferating (Ki67⁺). By comparison, stimulation with autologous CD4⁺ CD25⁻ cells resulted in the proliferation of only 37.4% of the HTLV-I Tax11-19–specific CD8⁺ T cells. Surprisingly, CMV pp65–specific CD8⁺ T cells showed a similar degree of proliferation when stimulated with autologous CD4⁺ CD25⁺ or CD4⁺ CD25⁻ T cells. These results directly demonstrate that HTLV-I–infected CD4⁺ CD25⁺ T cells can preferentially stimulate the expansion of HTLV-I–specific CD8⁺ T cells through peptide–HLA complexes in HAM/TSP patients.

The inability to detect Tax11-19 peptide–HLA-A*201 complexes directly from ex vivo T cells of HAM/TSP patients (Fig. 2 A) could be reflective of a low concentration of Tax11-19 peptide bound to endogenous HLA-A*201 molecules. As we had shown that the staining intensity of the TaxA2-Ab was dependent on the concentration of the peptide used for pulsing T cells in A2HD (Fig. 1 F), it was of interest to determine whether this TaxA2-Ab staining of peptide-pulsed ex vivo T cells was different between A2HD and HLA-A*201 HAM/TSP patients. CD4⁺ T cells were separated from ex vivo peripheral blood T cells of six A2HD and six HLA-A*201 HAM/TSP patients. These ex vivo T cells were pulsed at a concentration of 10 μM Tax11-19 peptide, previously determined to be in a range where the TaxA2-Ab was unable to detect Tax11-19 peptide–HLA-A*201 complexes (Fig. 1 F). In contrast to HTLV-I Tax11-19 peptide–pulsed A2HD T cells (Figs. 5 A and 1 F), staining intensity of the TaxA2-Ab on comparably peptide-pulsed HAM/TSP T cells was significantly higher (Fig. 5, A and B). To determine if this difference between A2HD and HAM/TSP patients was specific for the amount of Tax11-19 peptide–HLA-A*201 complexes as defined by the TaxA2-Ab, CD4⁺ T cells from the same individuals were pulsed with the same concentration of control melanoma gp100 G9-154 peptide, and the level of G9-154 peptide–HLA-A*201 complexes was analyzed using the MelanomaA2-Ab. As shown in Fig. 5 A, staining intensity with the MelanomaA2-Ab in HAM/TSP patients was similar to that in A2HD. These observations were confirmed using a wider range of peptide concentrations (Fig. 6, A and B). As the staining sensitivity of the TaxA2-Ab might be affected by the amount of expressed HLA-A*201 complexes, we compared the amount of HLA-A*201 expression on CD4⁺ T cells between A2HD and HAM/TSP using anti–HLA-A*201–specific Ab (BB7.2). There was no statistically significant difference in the amount of HLA-A*201 expression between A2HD and HAM/TSP (Fig. 5 C), suggesting that differential levels of HLA-A*201 expression do not contribute to the TaxA2-Ab staining intensity between A2HD and HAM/TSP patients. In addition, we were able to identify two HLA-A*0201 asymptomatic carriers from which sufficient cells were available to analyze with the TaxA2-Ab. TaxA2-Ab staining intensities on CD4⁺ T cells of these two HLA-A*0201 asymptomatic carriers were similar with those on A2HD (unpublished data).

Because the level of staining intensity for Tax11-19 peptide–HLA-A*201 complexes was different between HAM/TSP patients, controls, and two HLA-A*0201 asymptomatic carriers, it was likely that increased detection of these complexes on ex vivo CD4⁺ T cells from HAM/TSP patients might reflect the amount of endogenously processed Tax11-19 peptide. In support of this hypothesis, we determined if there was a correlation between the level of Tax11-19 peptide–HLA-A*201 complexes and HTLV-I proviral DNA load, HTLV-I tax mRNA load, and HTLV-I Tax11-19 peptide–specific CTL frequencies. As shown in Fig. 5 D, all three parameters significantly correlated with the level of expression of Tax11-19 peptide–HLA-A*201 complexes in HAM/TSP patients. However, the level of TaxA2-Ab staining did not correlate with the level of HLA-A*201 expression (unpublished data), again suggesting that the differential levels of Tax11-19 peptide–HLA-A*201 complex expression is not reflective of the amount of HLA expression. Collectively, these results suggest the existence of endogenously processed Tax11-19 peptide–HLA-A*201 complexes on ex vivo CD4⁺ T cells from HAM/TSP patients, which are stimulating HTLV-I Tax11-19–specific CD8⁺ T cells.

Virus-specific T cells recognize antigens by engaging the antigen-specific TCR with peptide–HLA complexes displayed on the surface of APCs. A major advance in immunology has been the description and use of polyvalent, soluble peptide–HLA complexes that specifically bind the TCR (tetramers; 34). Quantitative detection of antigen-specific T cell populations by such peptide–HLA tetramers has been useful for monitoring virus-specific T cell immunity in laboratory and clinical settings (35). However, there has been a tremendous shortage of reagents to study and visualize the ligand for the TCR, the peptide–HLA complex. Recently, by using large human Ab phage libraries, unique Abs with peptide-specific, HLA-restricted recognition pattern, termed peptide-specific, MHC-restricted Abs, have been isolated (28, 29). In this study we used these Abs, which specifically recognize HTLV-I Tax11-19 peptide–HLA-A*201 complexes (TaxA2-Ab), in analyzing ex vivo T cells from HLA-A*201 healthy individuals and HTLV-I–infected HAM/TSP patients. We have demonstrated specific binding of this Ab to exogenous peptide-pulsed as well as naturally processed endogenous HTLV-I Tax11-19 peptide–HLA-A*201 complexes on peripheral T cells from HAM/TSP patients, but not from HLA-A*201 healthy donors or two HLA-A*201 asymptomatic carriers. We feel this peptide-specific, MHC-restricted Ab will be useful in the analysis of other infectious agents, tumors, and autoantigens, and will further enhance our understanding of antigen–host cell immunological interactions.

This study also has defined the in vivo cellular tropism of HTLV-I using a reliable and accurate real-time quantitative PCR technique (TaqMan). HTLV-I was originally isolated from cultured CD4⁺ T cells of a patient with cutaneous T cell lymphoma (36). Soon after, HTLV-III/lymphadenopathy–associated virus was identified (37) and subsequently renamed HIV-I. These retroviruses predominantly infect CD4⁺ cells (38, 39), a seminal observation that directly led to defining CD4 as a receptor component for HIV-I (40, 41). Although the receptor for HTLV-I has not yet been identified, it must clearly be ubiquitous because HTLV-I infects a wide range of cells in vitro (42). Since 1990, HTLV-I has been thought to preferentially infect memory CD4⁺ (CD4⁺ CD45RO⁺) T cells in vivo (20). A recently established real-time quantitative PCR technique (TaqMan) has demonstrated that CD8⁺ T cells were also infected with HTLV-I (22). Because HTLV-I Tax has been known to transactivate the expression of IL-2Rα chain (CD25; 33), and a significant positive correlation between HTLV-I provirus load and the proportion of CD4⁺ CD25⁺ T cells has been demonstrated in PBMCs of HTLV-I–infected individuals (24), it has been hypothesized that CD4⁺ CD25⁺ T cells might be a reservoir for HTLV-I (25). This study, again using real-time quantitative PCR, supported this hypothesis but suggested two possibilities: HTLV-I preferentially infects CD4⁺ CD25⁺ T cells, or HTLV-I–infected T cells could induce CD25 in vivo. Although these possibilities are not mutually exclusive, the report that the HTLV-I receptor is an early T cell activation marker (43) may support the former possibility because CD25 is also a T cell activation marker, whereas a study demonstrating that HTLV-I Tax protein transactivates the promoters of IL-2 and CD25 (33) may support the latter.

Here we demonstrate that a high HTLV-I proviral load in CD4⁺ CD25⁺ T cells from HAM/TSP patients was associated with an increased ability to express HTLV-I mRNA and process HTLV-I Tax11-19–HLA-A*201 complexes on their cell surface. Of importance is the observation that HTLV-I–infected CD4⁺ CD25⁺ T cells were directly shown to stimulate the proliferation of HTLV-I–specific CD8⁺ T cells where the level of Tax11-19–HLA-A*201 expression correlated with HTLV-I Tax–specific CD8⁺ CTL frequency ex vivo. These increased antigen-presenting abilities of HTLV-I–infected CD4⁺ CD25⁺ T cells may also be associated with HTLV-I–induced overexpression of adhesion molecules and down-regulation of immunosuppressive cytokine as previously reported (44–46). These results support previous reports demonstrating that the frequency of HTLV-I Tax–specific CD8⁺ CTLs in the peripheral blood is proportional to the amount of HTLV-I proviral DNA load and the levels of HTLV-I tax mRNA expression (10, 11, 27). These results suggest the existence of continuous in vivo HTLV-I antigen presentation, which may serve to drive the increased numbers of HTLV-I–specific CD8⁺ CTLs in HAM/TSP patients. Moreover, the existence of such continuous HTLV-I antigen presentation in the presence of very high frequencies of HTLV-I–specific CD8⁺ CTLs (as high as 30% of total CD8⁺ cells in some patients), also supports the hypothesis that HTLV-I–specific CTLs might be insufficient to control persistent HTLV-I infection (47, 48).

Recently, CD4⁺ CD25⁺ T cells have been reported to include regulatory T cell populations that are engaged in the maintenance of immunologic self-tolerance by actively suppressing the activation and expansion of self-reactive T cells that may cause autoimmune disease (49, 50). This normal CD4⁺ CD25⁺ regulatory T cell population constitutes 5–10% of peripheral CD4⁺ T cells in mice and humans (49, 50). Removal of this population from normal rodents leads to the spontaneous development of various autoimmune diseases (49, 50). Importantly, CD4⁺ CD25⁺ T cells have been reported to suppress the proliferation of CD4⁺ CD25⁻ T cells as well as CD8⁺ T cells induced either by polyclonal or antigen-specific stimuli (49, 50). It has not escaped our attention that a proportion of HTLV-I–infected CD4⁺ CD25⁺ T cells in HAM/TSP may also be regulatory and infection with HTLV-I may alter function of these cell populations. It is clear from the results in this study that HTLV-I–infected CD4⁺ CD25⁺ cells are not suppressive but are stimulatory for HTLV-I Tax–specific proliferation of CD8⁺ T cells. These results suggest that CD4⁺ CD25⁺ cells infected with HTLV-I may lack regulatory T cell function in HAM/TSP patients. This hypothesis is currently being examined.

In conclusion, using a recently developed peptide-specific, MHC-restricted Ab that specifically recognizes Tax11-19 peptide–HLA-A*201 complexes coupled with a real-time quantitative PCR methodology, CD4⁺ CD25⁺ T cells from HAM/TSP patients were demonstrated to be a major reservoir of HTLV-I proviral DNA. These cells preferentially express HTLV-I mRNA and can present Tax11-19 peptide–HLA-A*201 complexes that can stimulate the proliferation of HTLV-I–specific CD8⁺ T cells. Moreover, higher sensitivity to HTLV-I Tax11-19 peptide–pulsed CD4⁺ T cells in HAM/TSP patients suggest the existence of endogenously processed Tax11-19 peptide–HLA-A*201 complexes on ex vivo CD4⁺ T cells. This novel Ab demonstrated a significant correlation between HTLV-I DNA, RNA, peptide–HLA complexes, and HTLV-I–specific T cell immune responses. These results indicate that the increased detection of HTLV-I proviral DNA and Tax11-19 peptide–HLA-A*201 complexes in HAM/TSP CD4⁺ CD25⁺ T cell subsets may serve to stimulate and expand HTLV-I Tax–specific CD8⁺ T cells that are important in the pathogenesis of HTLV-I–associated neurological disease.

This work was supported by a research grant from the Israel Science Foundation, administered by the Israel Academy for Sciences. U. Tomaru was supported by the Japanese Society for the Promotion of Science and a United States National Institutes of Health Research Fellowship.