INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Human T-cell leukemia virus type 1 (HTLV-1) was the first human retrovirus to be characterized and is implicated in the pathogenesis of adult T-cell leukemia (ATL), HTLV-1-associated myelopathy or tropical spastic paraparesis (HAM/TSP), as well as other inflammatory diseases (10, 13, 32, 43; M. Osame, K. Usuku, S. Izumo, N. Ijichi, H. Amitani, A. Igata, M. Matsumoto, and M. Tara, Letter, Lancet 1:1031-1032, 1986). Despite such differential manifestation of clinical symptoms, no consistent differences have been observed among HTLV-1 strains isolated from these patients (47). Segregation in HLA haplotype between ATL and HAM/TSP patients suggests the presence of host-regulated determinants (44).

There is a clear difference in HTLV-1-specific immune responses between ATL and HAM/TSP patients. HAM/TSP patients show higher levels of immune responses to HTLV-1 than ATL patients (16, 20). This may be a consequence of the relatively active HTLV-1 expression in HAM/TSP patients (11, 46) and may contribute to the pathogenesis of the neurological disorder. On the other hand, HAM/TSP patients rarely develop ATL despite a heavy viral load. In this respect, host immune responses against HTLV-1 may prevent the development of leukemia.

In general, cytotoxic T lymphocytes (CTL) play an important role not only in viral clearance but also in tumor eradication. HTLV-1-specific, HLA class I-restricted CD8⁺ CTL can be found in HAM/TSP patients and asymptomatic HTLV-1 carriers but are scarcely detectable in ATL patients (16, 19, 20, 28, 31). These CTL often recognize HTLV-1 Tax (16, 18), a critical protein for HTLV-1 leukemogenesis (1, 37). Moreover, ATL cells are susceptible to such CTL in vitro (20). These observations support the notion that HTLV-1-specific CTL may be an important effector of host immunosurveillance against HTLV-1-transformed tumor development. However, such in vitro data do not exclude the possibility that the induction of HTLV-1-specific immunity is merely a consequence of infection and has no effect on tumor development in vivo.

HTLV-1 can immortalize cultured normal T lymphocytes from humans and other species including rats (35, 40). However, attempts to induce tumor development in experimental animals by HTLV-1 infection have been unsuccessful (39). Instead, a certain strain of rats developed HAM/TSP-like disease after a long-term HTLV-1 carrier state following inoculation of HTLV-1-producing cells (14). Inoculation of HTLV-1-immortalized cells into immunocompetent syngeneic rats generally fails to cause tumor formation, except for a few cases of fully transformed HTLV-1-infected cells with additional mutations (30, 34). This suggests that some clonal evolution of infected cells may be required for HTLV-1 leukemogenesis.

However, Tateno et al. (40) demonstrated in vivo growth of HTLV-1-immortalized cells in syngeneic newborn but not adult rats. We have also recently demonstrated tumor growth in athymic rats inoculated with HTLV-1-immortalized cell lines (29). These results suggest that HTLV-1-infected cells can cause tumor formation when the host immunity is impaired. Therefore, we hypothesized that the in vivo mechanisms for HTLV-1-transformed tumor development involve both clonal evolution of infected cells and immune impairment in the host. This may explain the situation of human HTLV-1 carriers, most of whom are asymptomatic and only a few of whom develop ATL.

In this study, we verified the contribution of host cellular immune responses to protection against development of HTLV-1 tumors. We investigated whether HTLV-1 tumors could develop in adult euthymic rats by blocking T-cell immune responses with monoclonal antibodies (MAbs) against CD80 and CD86. These molecules provide a critical costimulatory signal for T-cell immune responses via their interaction with CD28 on T cells (2, 8, 9, 12, 17, 21, 24, 26, 45). Stimulation via the T-cell receptor in the absence of costimulatory signal renders T cells anergic or in a long-lasting state of unresponsiveness (4, 25, 36). We demonstrate in this study that administration of these antibodies in adult euthymic rats results in impairment of HTLV-1-specific cellular immune responses and the development of HTLV-1 tumors. To our knowledge, this is the first animal model of inducible HTLV-1 tumor by suppression of antigen-specific cellular immunity in vivo.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Rats. Four-week-old female WKA/HKm (WKAH) rats were purchased from Shizuoka Animal Laboratory Center (Shizuoka, Japan). These rats were maintained at the experimental animal facilities of Tokyo Medical and Dental University, and the experimental protocol was approved by the animal care committee of the university.

Cell lines. The HTLV-1-infected T-cell line TARS-1, derived from splenocytes of WKAH rats (40), was cultured in RPMI 1640 (GIBCO Laboratories, Grand Island, N.Y.) with 10% heat-inactivated fetal calf serum (FCS), penicillin (100 IU/ml), streptomycin (100 µg/ml), 2-mercaptoethanol (10⁵ M), and NaHCO₃ (2 mg/ml). The simian virus 40-transformed cell line W7KSV, originated from WKAH rat kidney cells (39), was kindly provided by Y. Tanaka (Kitasato University) and cultured in the same medium.

MAbs. Mouse MAbs to rat CD80 (3H5) and CD86 (24F) were prepared as described previously (27). Equal volumes of both MAbs rehydrated with phosphate-buffered saline (PBS) at a concentration of 2 mg/ml were mixed and then stored at 4°C.

Inoculation of TARS-1 cells and administration of anti-CD80 and anti-CD86 MAbs. TARS-1 cells were inoculated subcutaneously (s.c.) into the back of each WKAH rat at a dose of 2 × 10⁷ cells/0.5 ml in PBS. Simultaneously, half of these rats were injected intraperitoneally with 1 ml of a mixed MAb solution containing 1 mg each of anti-CD80 and anti-CD86 MAbs (collectively referred to as CD80/CD86 MAbs) per animal. Thereafter, the same amounts of MAbs were injected intraperitoneally into these rats every other day until day 14 after TARS-1 inoculation. Control PBS was injected into the other TARS-1-inoculated rats at the same time intervals. The dose and schedule of administration of anti-CD80/CD86 MAbs which caused optimal inhibition of cardiac allograft rejection were determined in preliminary studies (H. Yagita and K. Okumura, unpublished results). In some experiments, a similar protocol was used except for an initial injection of the same amounts of MAbs and 2 × 10⁷ mitomycin C (MMC)-treated TARS-1 cells 3 days before s.c. inoculation of live TARS-1 cells (Fig. 1).

Growth of subcutaneous tumor. Following s.c. inoculation of TARS-1 into rats, tumor growth was measured twice every week with a caliper. In these measurements, we determined the longest surface length (in millimeters; a) and width (in millimeters; b) and calculated the tumor volume (in cubic millimeters; V) by using the following formula: V = ab²/2.

Histological examination. Rats were anesthetized and sacrificed on day 14 or 35 after TARS-1 inoculation. Each rat was examined carefully for the presence of subcutaneous tumors at the site of injection as well as metastatic tumors. The excised tissues were stored as paraffin blocks following formalin fixation or as frozen blocks in Tissue-Tec O.C.T. compound (Sakura Finetechnical Co., Ltd., Tokyo, Japan) at 80°C. Thin-sliced specimens of paraffin blocks were stained with hematoxylin and eosin and examined under the microscope. Immunohistologic staining was performed by using thin-sliced specimens of frozen blocks and the Envision system (DAKO, Glostrup, Denmark) with anti-rat interleukin-2 (IL-2) receptor -chain MAb (Chemicon International Inc., Temecula, Calif.) as the first antibody.

T-cell proliferation assay. Splenic T cells purified through a nylon wool column (10⁵ cells/well) were cocultured with MMC (Sigma Chemical Company, St. Louis, Mo.)-treated TARS-1 or W7KSV cells (5 × 10⁴ cells/well) in 96-well U-bottom culture plates at 37°C for 72 h. Cultures were pulsed with [³H]thymidine ([³H]TdR; 37 kBq/well) for the last 18 h to assess cell proliferation. Cells were harvested with a Micro 96 Harvester (Skatron, Lier, Norway), and [³H]TdR uptake into cells (reported as mean ± standard deviation [SD]) was measured in a microplate  counter (Micro Beta Plus, Wallac, Turku, Finland).

Induction of CTL and cytotoxic assay. Splenic T cells (5 × 10⁶ cells/well) were cocultured with MMC-treated TARS-1 cells (2.5 × 10⁶ cells/well) in 2 ml of 10% FCS-RPMI 1640 per well of a 24-well plate. Six days later, cytotoxic activities against TARS-1 and W7KSV cells were measured by 6-h ⁵¹Cr release assay at various effector-to-target (E/T) ratios as described previously (5). Specific cytotoxicity was calculated as ([experimental ⁵¹Cr release  spontaneous ⁵¹Cr release]/[maximum ⁵¹Cr release  spontaneous ⁵¹Cr release]) × 100. In some experiments, [³H]TdR release assay was also used to measure cytotoxicity. In this method, target cells were incubated with 3.7 MBq of [³H]TdR per 10⁶ cells for 12 h at 37°C, followed by triplicate washing; labeled target cells (5 × 10³ cells/well) and effector cells were plated in 96-well U-bottom plates at an E/T ratio of 30. After 6 h of incubation at 37°C, cells were harvested with a Micro 96 Harvester (Skatron), and [³H]TdR remaining in target cells was measured in a microplate  counter (Micro Beta Plus). Percent specific cytotoxicity was calculated as ([cpm without effector  cpm with effector]/[cpm without effector]) × 100.

Generation of the CTL line. For induction of HTLV-1-specific CTL in long-term cultivation, splenic T cells (2.5 × 10⁶ cells/well) were cocultured with the same number of MMC-treated TARS-1 cells in 10% FCS-RPMI 1640 in the presence of 20 U of recombinant human IL-2 (rhIL-2; Shionogi Pharmaceutical Co., Osaka, Japan) per ml, with periodical stimulation using MMC-treated TARS-1 cells every 2 weeks.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Growth of HTLV-1 tumors in anti-CD80/CD86 MAb-treated rats. In preliminary experiments, we found that several HTLV-1-immortalized rat T-cell lines including TARS-1 were tumorigenic in athymic rats but not in syngeneic immunocompetent rats. We then examined whether suppression of HTLV-1-specific immune responses in vivo results in the development of HTLV-1 tumors. To achieve antigen-specific immunosuppression, we used a combination of anti-CD80 and anti-CD86 MAbs, which block a costimulatory signals for T-cell activation. HTLV-1-immortalized TARS-1 cells were subcutaneously inoculated into syngeneic WKAH rats with or without intraperitoneal administration of anti-CD80 and anti-CD86 MAbs. The animals were then periodically administered the MAbs for 14 days. As shown in Table 1 and Fig. 1, tumor development was consistently observed in all anti-CD80/CD86 MAb-treated rats. Two of the three MAb-treated rats sacrificed on day 14 after inoculation had visible lung metastasis. Histological examination of the subcutaneous mass revealed a lymphoma-like appearance with medium-sized tumor cells expressing IL-2 receptor and infiltration of tumor cells in the lungs (Fig. 2). To confirm the blockade of costimulation at antigen presentation, we also used a slightly different protocol, involving pretreatment of rats with MMC-treated TARS-1 cells and anti-CD80/CD86 MAbs 3 days before s.c. inoculation of live TARS-1 cells. Rats treated with this protocol showed tumor induction similar to that observed with the original protocol (Table 1). In contrast, control animals without MAb treatment developed no or little swelling, which reached a peak size at 5 to 6 days at the TARS-1-inoculated site but later regressed spontaneously. Among control rats, those pretreated with MMC-treated TARS-1 cells showed more efficient tumor regression than rats without pretreatment (Fig. 1).

View larger version (23K): [in this window] [in a new window]   FIG. 1.   Growth of subcutaneous TARS-1 tumors in rats injected intraperitoneally (ip) with a mixture of anti-CD80 MAb (3H5) and anti-CD86 MAb (24F) (1 mg/rat) (, ) or control PBS (, ). All tested rats were inoculated s.c. with TARS-1 cells (2 × 107/rat) on day 0. One group of rats was treated with either MAbs () or PBS () simultaneously with live TARS-1 inoculation, while the other group was pretreated (pre) with MAbs and MMC-treated TARS-1 () or MMC-treated TARS-1 alone () 3 days before inoculation of live TARS-1. MAbs were administered every other day for the next 2 weeks and then discontinued. Inoculation protocols are shown at the top. The volume of each subcutaneous tumor was calculated as described in Materials and Methods. Each symbol represents an individual rat.

View larger version (104K): [in this window] [in a new window]   FIG. 2.   Macroscopic and microscopic appearance of a TARS-1 tumor. (a) Subcutaneous tumor at the injected site on day 14 following TARS-1 inoculation in an anti-CD80/CD86 MAb-treated rat (right) but not an untreated rat (left). (b) Regression of the tumor on day 35 in a rat treated with MAbs only during the initial 14 days (right). The TARS-1-inoculated MAb-untreated rat never developed a tumor (left). (c) Frozen section of subcutaneous tumor of a MAb-treated TARS-1-inoculated rat sacrificed on day 14, stained with MAb to IL-2 receptor (brown) and hematoxylin (blue). (d) Hematoxylin-and-eosin-stained frozen section of the lung of the rat shown in panel c.

Regression of tumor following discontinuation of MAb treatment. After cessation of MAb treatment on day 14, a proportion of rats were further maintained. The kinetics of tumor growth in these rats is shown in Fig. 1. In most of these rats, subcutaneous tumors regressed promptly after withdrawal of MAbs and were hardly palpable at the time of autopsy between 25 and 35 days after TARS-1 inoculation. At autopsy, no viable tumor cells were observed at the site of injection or in the lungs of these rats (data not shown). Thus, animals treated with anti-CD80/CD86 MAbs exhibited HTLV-1 tumor growth during MAb treatment but rejected the tumor following withdrawal of MAbs (Fig. 1).

Impairment of HTLV-1-specific T-cell proliferation in anti-CD80/CD86 MAb-treated rats. CD80/CD86 blockade often induces T cells unresponsive to concurrently inoculated antigens (23, 42). In the next series of experiments, we tested HTLV-1-specific proliferative T-cell responses in anti-CD80/CD86 MAb-treated and untreated rats. Splenic T cells from untreated control rats at day 14 showed high proliferation in response to MMC-treated TARS-1 cells. This proliferative response was specific to TARS-1 cells, as these splenic T cells did not respond to simian virus 40-transformed W7KSV cells. In contrast, splenic T cells from MAb-treated rats exhibited a markedly reduced level of proliferative response against TARS-1 cells, which was slightly higher than that against W7KSV cells (Fig. 3a). These results indicated that the HTLV-1-specific T-cell proliferative response was greatly impaired during anti-CD80/CD86 MAb treatment. Proliferative responses of splenic T cells were also examined on day 35 when tumor regression had occurred following cessation of MAb treatment on day 14. As shown in Fig. 3b, these T cells exhibited a level of TARS-1-specific proliferative response comparable to that exhibited by splenic T cells from untreated control rats. These results indicated that HTLV-1-specific T-cell hyporesponsiveness induced by anti-CD80/CD86 MAb treatment was reversible.

View larger version (13K): [in this window] [in a new window]   FIG. 3.   Splenic T-cell proliferative response against TARS-1. Splenic T cells isolated on day 14 (a) or day 35 (b) from TARS-1-inoculated rats with (MAb-treated) or without (control) anti-CD80/CD86 MAb treatment were cocultured without () or with MMC-treated W7KSV () or TARS-1 () cells for 3 days. [3H]TdR incorporation was measured during the last 18 h. Data represent the mean ± SD of triplicate wells. Similar results were obtained in three independent experiments.

Impairment of HTLV-1-specific CTL response in anti-CD80/CD86 MAb-treated rats. Since CTL is considered the primary effector for elimination of virus-infected cells, we examined the CTL response against TARS-1 in TARS-1-inoculated rats with or without anti-CD80/CD86 MAb treatment. Splenic T cells from each group of rats at day 14 or 35 were stimulated with MMC-treated TARS-1 cells for 5 days, and then cytotoxicity against TARS-1 was examined. Rats inoculated with TARS-1 without MAbs showed a high level of cytotoxicity against TARS-1 but not against W7KSV (Fig. 4a). In contrast, almost no significant CTL activity was induced in splenic T cells from anti-CD80/CD86 MAb-treated rats (Fig. 4b). At day 35, however, almost comparable levels of TARS-1-specific CTL activities were induced in splenic T cells regardless of anti-CD80/CD86 MAb treatment (Fig. 4c and d). The absence of tumor-specific CTL response during anti-CD80/CD86 MAb treatment and its recovery after cessation of treatment was associated with tumor development and regression in these animals, respectively. Taken together with the proliferative responses, these results strongly suggested that HTLV-1-specific T-cell responses are closely associated with elimination of HTLV-1-infected tumor cells in vivo.

View larger version (16K): [in this window] [in a new window]   FIG. 4.   CTL induction against TARS-1. Splenic T cells were isolated on day 14 (a and b) or day 35 (c and d) from TARS-1-inoculated rats treated without (a and c) or with (b and d) anti-CD80/CD86 MAbs and cocultured with MMC-treated TARS-1 for 6 days. Cytotoxic activity against W7KSV () or TARS-1 () target cells was tested by a standard 51Cr release assay at the indicated E/T ratios. Data represent the mean ± SD of triplicate wells. Similar results were obtained in three independent experiments.

Minimal recovery of T-cell unresponsiveness by exogenous IL-2. We next analyzed the state of HTLV-1-specific T cells in anti-CD80/CD86 MAb-treated rats. It has been proposed that stimulation via the T-cell receptor under CD80/CD86 blockade renders such cells anergic and that this anergy can be reversed by exogenous IL-2 (33). We therefore tested whether the impaired proliferative response of splenic T cells from MAb-treated rats could be restored upon TARS-1 stimulation in the presence of exogenous IL-2. As shown in Fig. 5, IL-2 (50 and 100 U/ml) only slightly enhanced the proliferative response of splenic T cells of anti-CD80/CD86 MAb-treated rats up to levels almost comparable to those of naive splenic T cells. In contrast, splenic T cells of TARS-1-inoculated rats without MAb treatment showed a high level of TARS-1-specific proliferative response, which was further enhanced by exogenous IL-2. The poor recovery by exogenous IL-2 of the T-cell response of MAb-treated rats implied that the number of anergic cells, if any, might be very low.

View larger version (12K): [in this window] [in a new window]   FIG. 5.   Minimal recovery of T-cell response to TARS-1 by exogenously added IL-2. Splenic T cells were isolated from age-matched naive rats (a) or TARS-1-inoculated rats treated without (b) or with (c) anti-CD80/CD86 MAbs on day 14 and were cocultured without () or with MMC-treated W7KSV () or TARS-1 () cells in the absence or presence of 50 or 100 U of rhIL-2 per ml for 4 days. Gray bars represent positive control wells containing concanavalin A (Con A; 10 µg/ml). [3H]TdR incorporation was measured during the last 18 h. Data represent the mean ± SD of triplicate wells. Similar results were obtained in two independent experiments.

Effect of anti-CD80/CD86 MAbs on the in vitro T-cell response. Unlike previous reports demonstrating induction of persistent tolerance against allografts by CD80/CD86 blockade (23, 42), our results showed recovery of the T-cell response against HTLV-1-infected tumor cells following termination of anti-CD80/CD86 treatment. We then tested the in vitro effects of anti-CD80/CD86 MAbs on recovered T-cell responses. On day 35, splenic T cells from MAb-treated rats, in which the tumor had already regressed following withdrawal of MAb treatment, showed a significant level of TARS-1-specific proliferation, and such proliferation was moderately inhibited by exogenously added anti-CD80/CD86 MAbs in vitro (Fig. 6c). Splenic T cells from TARS-1-inoculated rats exhibited significant levels of TARS-1-specific but also spontaneous proliferation without any stimulation. These proliferative responses were mostly inhibited by anti-CD80/CD86 MAbs (Fig. 6b). On the other hand, HTLV-1-specific CTL function was not affected by these MAbs in vitro. Figure 6d shows the results of CTL assays with effector T cells derived from a TARS-1-inoculated rat and expanded in an in vitro culture with TARS-1 stimulation for 5 weeks. These cells were able to kill target cells that expressed HTLV-1 antigens in the presence of anti-CD80/CD86 MAbs in vitro. Thus, the proliferative response but not cytotoxic function of TARS-1-specific T cells was significantly inhibited by anti-CD80/CD86 MAbs in vitro. These results suggest that treatment of rats with MAb allowed tumor growth by suppressing the expansion of TARS-1-specific T-cell clones without rendering these cells permanently anergic.

View larger version (12K): [in this window] [in a new window]   FIG. 6.   (a to c) Inhibitory effect of anti-CD80/CD86 MAbs on T-cell responses against TARS-1 in vitro. Splenic T cells from age-matched naive rats (a) or TARS-1-inoculated rats without (b) or with (c) anti-CD80/CD86 treatment were isolated on day 35 and cocultured without () or with MMC-treated W7KSV () or TARS-1 () cells in the presence (+) or absence () of 20 µg each of anti-CD80 and anti-CD86 MAbs per ml for 4 days. [3H]TdR incorporation was measured during the last 18 h. Data are the mean ± SD of triplicate wells. (d) CTL activity against TARS-1 was not inhibited by anti-CD80/CD86 MAbs. Cytotoxic activity of CTL derived from a TARS-1-inoculated rat was tested against [3H]TdR-labeled W7KSV () or TARS-1 () target cells at an E/T ratio of 30 in the presence (+) or absence () of 20 µg each of anti-CD80 and anti-CD86 MAbs per ml. The effector CTL used were obtained from a culture with periodical TARS-1 stimulation as described in Materials and Methods. Data are the mean ± SD of triplicate wells.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In this study, we demonstrated that immunocompetent rats inoculated with HTLV-1-transformed cells could develop tumors after blocking CD80/CD86-CD28 interaction with anti-CD80/CD86 MAbs. During MAb treatment, the proliferative and CTL responses of splenic T cells to HTLV-1-infected cells were impaired. In contrast, following discontinuation of MAb treatment, tumor regression occurred in these rats, and this was associated with recovery of HTLV-1-specific T-cell responses. The inverse correlation between tumor growth and cellular immunity in these rats indicated that inhibition of HTLV-1-specific T-cell responses by anti-CD80/CD86 MAbs was the principal mechanism for tumor development in MAb-treated rats. These results suggested that (i) HTLV-1-specific cell-mediated immunity plays a pivotal role in immunosurveillance against the development of HTLV-1-transformed tumors and (ii) CD28-mediated costimulatory pathway plays a critical role in the elicitation of immune responses against HTLV-1.

Previous studies have proposed that a costimulatory signal provided by CD80 and CD86 is required for the full activation of T cells, and the absence of this costimulation induces T-cell anergy (4, 25, 36). However, in the present study, HTLV-1-specific T-cell responses were suppressed during anti-CD80/CD86 MAb treatment but reversed upon withdrawal of MAbs. This contrasts with the previous observations that transient blockade of CD80 and CD86 with a single injection of CTLA-4-Ig successfully induced persistent acceptance of allografts (23, 42). One possible reason for the failure to induce a state of tolerance in the present study could be constitutive expression of multiple costimulatory molecules on HTLV-1-infected cells. In this regard, HTLV-1-infected cells are known to express various costimulatory molecules in addition to CD80 and CD86, such as CD40 and OX40L, which have been implicated in CD28-independent pathways of T-cell costimulation (3, 7, 38). Therefore, even in the presence of anti-CD80/CD86 MAbs, induction of T-cell anergy might be prevented by these costimulatory molecules expressed on HTLV-1-infected cells. Alternatively, various cytokines produced by HTLV-1-infected T cells (6, 41) may also contribute to the prevention of T-cell anergy.

Although the induction of anergy by in vivo administration of anti-CD80/CD86 MAbs appeared to be incomplete in the present system, TARS-1-specific proliferative and CTL responses were abolished and tumors grew in MAb-treated rats during treatment (Fig. 3 and 4). In vitro analysis revealed that exogenously added anti-CD80/CD86 MAbs inhibited proliferation but not cytotoxicity of TARS-1-specific T cells (Fig. 6). This indicated that the in vivo blockade of CD80 and CD86 might also inhibit clonal expansion of HTLV-1-specific T cells, which was a prerequisite for tumor rejection. The poor recovery of T-cell response of MAb-treated rats by exogenous rhIL-2 (Fig. 5) also supports the presence of fewer HTLV-1-reactive T cells in these rats than in untreated rats.

It is of note that a significant level of in vitro proliferation was observed without any stimulation of T cells in rats inoculated with TARS-1 alone, which was blocked by anti-CD80/CD86 MAbs in vitro (Fig. 6b). Similar spontaneous proliferation of peripheral blood mononuclear cells was described for patients with HAM/TSP (15), and potential involvement of CD80 and CD86 in this phenomenon has also been suggested (22). Interestingly, splenic T cells obtained on day 35 from TARS-1-inoculated rats which were initially treated with MAb showed a lower level of spontaneous proliferation than those from untreated rats (Fig. 3b and 6c). This finding suggests that the effector cells inducing spontaneous proliferation might be raised in vivo through interaction with CD80 and CD86 molecules expressed on HTLV-1-infected cells.

In conclusion, we demonstrated in this study that suppression of HTLV-1-specific cellular immune responses led to the development of HTLV-1-transformed tumors in vivo. Our study emphasizes the importance of host cellular immune responses against HTLV-1-induced tumor development, indicating that immunosuppressive therapy against HTLV-1-associated inflammatory diseases might potentially favor the development of ATL.