INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

The cutaneous T-cell lymphomas (CTCL), mycosis fungoides (MF) and the Sezary syndrome (SS), are the most frequent lymphomas involving the skin (8). Patients with CTCL may have an extremely long natural history of disease, with a median 6-year duration from the onset of skin symptoms to diagnosis of MF and finally to the leukemic phase of SS (16, 46). The development of SS as a consequence of the persistent stimulation of T cells by a variety of bacterial antigens has been hypothesized (12, 13, 17, 41, 42). Additionally, persistent colonization with toxigenic bacteria can result in the expansion of epidermotropic T cells (35). Activated T cells produce cytokines and/or chemokines that perpetuate the inflammatory reaction and recruit additional T cells to the skin (36, 37).

Leukemia T cells or Sezary cells are frequently identified in the peripheral blood of patients with either MF or SS (25). Sezary cells represent an expansion of circulating memory CD4⁺ T cells that lack CD7 antigen expression (15). The defective expression of CD7 has been described on memory T cells from healthy donors (19), from patients infected with human immunodeficiency virus (2), and from patients with SS (15). Compared with CD7⁺ T cells, CD7 T cells are less responsive to activation with anti-CD3 (2, 14) due to a lower-level expression of the T-cell receptor (TCR) and the suboptimal triggering of the TCR-CD3 complex (44). Defective triggering of the TCR-CD3 complex on memory T cells may also account for the reduction in the production of type 1 cytokines in CTCL (14, 26, 32).

Type 1 and type 2 cytokines play a vital role in immunity (30, 40). Patients with MF and SS differ in their production of (26, 32, 45) and in their response to (10, 32) cytokines. Whereas a type 1 cytokine production profile consisting of interleukin-2 (IL-2) and gamma interferon (IFN-) is common in MF, a type 2 cytokine production profile, consisting of IL-4, is more likely to occur in SS (26, 32, 33, 45). Although previous studies demonstrated the cytokine production in peripheral blood mononuclear cell cultures of patients with CTCL (26, 32, 45), none of these studies was capable of identifying the phenotype of the cells producing cytokines. In the present study, T cells were activated with phorbol 12-myristate 13-acetate (PMA), which directly stimulated calcium and phospholipid-dependent protein kinase C activity (43). Furthermore, we took advantage of a flow cytometric technique (31) that is capable of identifying the phenotype of the cytokine-producing cells at the single-cell level. We believe that such an approach provides opportunities to investigate the role of these soluble factors in exacerbating tumor pathogenesis in CTCL.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Patient population and staging. A complete medical history of each patient was taken to assess previous treatments for CTCL. Each patient was assigned to a disease stage based on a physical examination to assess performance status, concurrent nonmalignant disease, and skin involvement as previously described (7, 20). The breakdown of the 25 patients by stage of disease is shown in Table 1 and consisted of four patients with stage I, four with stage II, nine with stage III, and eight with stage IV. The laboratory examination included the following: a complete blood count and a differential count of a Wright-Giemsa-stained smear, a platelet count, and T-cell immunophenotyping. The percentage of Sezary cells in the peripheral blood was determined by the presence of malignant lymphoid cells with hyperchromatic, indented (cerebriform) nuclei identifiable under light microscopy according to previously defined criteria (38).

Phenotype of T cells in CTCL. Aliquots of whole blood (100 µl) were mixed with pretitrated amounts of monoclonal antibodies, each conjugated with one of the following fluorochromes: fluorescein isothiocyanate (FITC), phycoerythrin (PE), or peridinin chlorophyll protein (PerCP), all purchased from Becton-Dickinson Immunocytometry Systems Inc., San Jose, Calif. Each sample was incubated in the dark at ambient temperature for 15 min with a combination of monoclonal antibody reagents to detect the common leukocyte antigen (CD45⁺), monocytes (CD14⁺), total T cells (CD3⁺), and CD3⁺ T-cell subsets coexpressing CD7, CD4 (helper and inducer), or CD8 (cytotoxic and suppressor) surface antigens. At the end of this incubation period, the erythrocytes in each tube were disrupted. The samples were washed three times with phosphate-buffered saline and concentrated by centrifugation. The cell pellets were fixed with 200 µl of a 1% paraformaldehyde solution, stored at 4°C, and analyzed within 4 h with a FACSCalibur flow cytometer (Becton-Dickinson). Isotype-matched negative controls conjugated to each of the fluorochromes were used to determine nonspecific binding of human leukocytes to mouse monoclonal antibodies; lymphocyte purity was assessed by setting a gate around those cells that were positive for CD45 and negative for CD14.

Determination of cytokine synthesis by T cells. Preparation for the synthesis of cytokines and for the detection of cytokines in the cytoplasm of T cells was done as previously described (28, 31). Briefly, the cells were prepared in four sequential steps: (i) the cells were activated by PMA (25 ng/ml) for 4 h at 37°C in the presence of 1 µg of ionomycin/ml and 10 µg of brefeldin A (BFA; a nontoxic but potent inhibitor of intracellular transport)/ml; (ii) the activated cells were divided between two sets of tubes and reacted with either anti-CD3-PerCP and anti-CD7-FITC or with anti-CD3-PerCP and anti-CD8-FITC to determine the surface immunophenotype of the T cells; (iii) the cells were treated with a permeabilization solution (Becton-Dickinson) for 1 h at 37°C to allow entry of the monoclonal antibody; and (iv) the cells were stained with a cytokine-specific monoclonal antibody conjugated with PE to detect IL-2, IL-4, or IFN-. All cell preparations were fixed in a solution of 1% paraformaldehyde and stored at 4°C until they were analyzed with the FACSCalibur flow cytometer. In parallel experiments, the blood from each patient and control subject was incubated with BFA alone and served as a resting or unstimulated preparation.

Data analysis. The percentages of cytokine-producing T cells were calculated for total T cells and subsets of T cells for each patient and control subject. The data are presented as both the mean number of cytokine-producing T cells per microliter of peripheral blood and the mean percentage of cytokine-producing cells per specific T-cell subset. Statistical differences in the mean percentages and numbers of cytokine-producing T cells between study groups were determined by the unpaired t test.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

Detection of Sezary cells in peripheral blood smears of patients with CTCL. Table 1 lists the clinical characteristics and hematology parameters of patients with CTCL. Sezary cells were detected in the peripheral blood of patients with CTCL at every stage of the disease, with an increasing percentage of Sezary cells associated with disease progression. We were able to identify Sezary cells in 18 of the 25 patients with CTCL, 14 of whom had >10% Sezary cells in their peripheral blood (Table 1). An increase in the percentage of Sezary cells in patients with more severe CTCL was also supported by a recent study with PCR-based analysis (24).

Immunophenotype of peripheral blood lymphocytes. Three-color flow-cytometric analysis was performed on all blood samples to determine the percentages of total T cells (CD3⁺) and of the T-cell subsets in circulation (Table 2). The immunophenotypes of total T cells and subsets of T cells were similar for patients with MF and controls. The immunophenotype of patients with SS was significantly different from those of patients with MF and control subjects. In particular, patients with SS had significantly higher percentages of CD3⁺, CD7, CD4⁺, and CD4⁺ CD45RO⁺ T cells and significantly lower percentages of CD7⁺ and CD8⁺ T cells than both patients with MF and control subjects (Table 2).

In vitro activation of peripheral blood T lymphocytes of patients with CTCL. When the peripheral blood lymphocytes were cultured in vitro in the presence of BFA and without PMA, only negligible numbers of T cells from all groups synthesized cytokine (data not shown). The numbers of CD69⁺ CD3⁺ T-cell subsets per microliter in patients with MF (811 ± 125) was lower than those in patients with SS (1,472 ± 228) and controls (1,394 ± 155) (Table 3). In addition, in the patients with MF and the control donors, CD69 expression was significantly higher in the CD7⁺ T cells than in their autologous CD7 T cells. Compared to control donors, significantly lower percentages of CD4⁺ and CD8⁺ T cells of patients with MF and SS expressed CD69 (Table 3).

Synthesis of IL-2 by PMA-activated T-cell subsets. The synthesis of IL-2 by T-cell subsets following stimulation with PMA is shown in Table 4. The mean percentages and counts per microliter of CD3⁺ T cells synthesizing IL-2 were similar for all groups. Within the CD3⁺ T cells of patients with MF and controls, the number of IL-2⁺ CD7⁺ T cells was significantly higher than the autologous IL-2⁺ CD7 T-cell subset (P = 0.01). On the other hand, patients with SS had significantly fewer IL-2⁺ CD7⁺ and significantly more IL-2⁺ CD7 T cells per microliter than both patients with MF and control subjects. Patients with SS also had significantly higher numbers of IL-2⁺ CD4⁺ T cells than patients with MF and significantly lower numbers of IL-2⁺ CD8⁺ T cells than both patients with MF and controls.

Synthesis of IFN- by PMA-activated T-cell subsets. The highest percentages and numbers microliter of IFN--producing T cells were observed in controls, followed by those of patients with MF and finally by those of patients with SS (Table 5). There were significant decreases in the percentages and counts of IFN-⁺ CD3⁺ T cells associated with disease progression in CTCL. The majority of IFN-⁺ CD3⁺ T cells of patients with MF belonged to the CD7⁺ T-cell subset. In contrast, less than 10% of CD7 T cells of patients with SS produced IFN-, indicating that the majority of CD7 T cells of these patients were not capable of producing IFN-. In addition, patients with SS had significantly lower percentages of IFN-⁺ CD4⁺ T cells than patients with MF and controls. Whereas patients with MF had more IFN-⁺ CD8⁺ (243 ± 33 per µl) than IFN-⁺ CD4⁺ (110 ± 22 per µl) T cells, patients with SS had more IFN-⁺ CD4⁺ (100 ± 45 per µl) than IFN-⁺ CD8⁺ (46 ± 11 per µl) T cells.

View this table: [in this window] [in a new window]   TABLE 5.   Synthesis of IFN- by T-cell subsets

Synthesis of IL-4 by PMA-activated T-cell subsets. The synthesis of IL-4 by T-cell subsets following stimulation with PMA is shown in Table 6. In patients with SS, a significantly higher percentage of CD7⁺ T cells than CD7 T cells synthesized IL-4. Conversely, fewer CD7⁺ T cells from control donors synthesized IL-4 than their autologous CD7 T cells. Equivalent percentages of CD7⁺ and CD7 T cells of patients with MF synthesized IL-4. More importantly, the total numbers of CD3⁺ T cells synthesizing IL-4 were similar for all three groups. Whereas the percentage of IL-4⁺ CD4⁺ T cells in patients with SS was significantly lower than that in patients with MF (P = 0.01) and controls (P = 0.01), the percentage of IL-4⁺ CD8⁺ T cells in control donors was significantly lower than that in patients with MF (P = 0.01) and SS (P = 0.01). In addition, patients with MF had a significantly higher number of IL-4⁺ CD8⁺ T cells per microliter than patients with SS and controls.

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

Although CD7 is expressed on major subsets of peripheral blood T cells (15), its function is unknown and no ligand has been identified. The loss of CD7 antigen expression is found on memory T cells in healthy individuals (19), in patients infected with human immunodeficiency virus (2, 22), and in the synovial fluids of patients with rheumatoid arthritis (21). In patients with SS, the lack of CD7 antigen expression on CD4⁺ T cells is a classical feature of malignant Sezary cells (19). Consistent with previous studies (3, 18), we found a significant increase in the percentage and absolute number of memory-helper T cells that lacked CD7 antigen expression in patients with SS (Table 2).

It is commonly recognized that CTCL patients are defective in triggering the TCR-CD3 complex (14, 26, 32). In the present study, we stimulated T cells with PMA via a pathway that is independent of that triggering the CD3-TCR complex (43). Compared with CD7⁺ T cells, a lower percentage of autologous CD7 T cells of patients with CTCL coexpressed the activation marker CD69 (Table 2), thereby indicating that the defect is not restricted to activation of TCR (9) but extends to nonspecific activation mechanisms. Since the malignant cells in patients with SS are predominantly CD7 T cells, their reduced responsiveness to activation (Table 3) accompanied by their reduced capacity to proliferate (14) may be a consequence of their abnormal phenotype. Because T lymphocytes in patients with SS are likely to be clonal (24) and to express activation antigens (1), further stimulation of these T cells may lead to suboptimal activation to avoid programmed cell death (9).

By measuring cytokine proteins in individual cells rather than in culture supernatants as others have done (27, 45), we found that similar percentages and counts of total T cells from all study groups synthesized IL-2 (Table 4). Since patients with MF and healthy donors had significantly more CD7⁺ than CD7 T cells in circulation, it was expected that their IL-2⁺ CD7⁺ T cells would outnumber their IL-2⁺ CD7 T cells (Table 4). In contrast, patients with SS had a higher number of IL-2⁺ CD7 T cells per microliter and the bulk of the IL-2 was more likely to be made by CD7 than by CD7⁺ T cells, as others have suggested (45).

By blocking cytokine transport with BFA, we found no difference in the percentages of IL-2⁺ CD7⁺ (32.0% ± 4.3%) and IL-2⁺ CD7 (42.2% ± 7.2%) T-cell subsets in patients with SS (Table 4). Since there were similar numbers of IL-2⁺ CD3⁺ T cells in patients with SS and control donors (Table 4), we cannot confirm a deficiency in IL-2 production in patients with SS as previously described (27, 32, 45). Furthermore, the difference in results may be explained by a defect in the posttranslational processes or by an enhancement in the degradation of intracellular protein in malignant cells (11). Alternatively, it can be explained by the reabsorption of IL-2, acting as an autocrine growth factor for CD7 T cells (10), through membrane-bound IL-2 receptors (6, 23) and/or by the formation of complexes with soluble IL-2 receptors (5, 6).

As expected, there was a significant decline in the number of IFN-⁺ CD3⁺ T cells in patients with CTCL compared to that in healthy donors (Table 5). Between MF and SS, there was a further significant decrease in the percentage and number of IFN-⁺ CD3⁺ T cells, indicating an inverse relationship between the number of IFN-⁺ CD3⁺ T cells and disease stage. Our data support earlier observations that T cells in SS produced less IFN- following stimulation with mitogen (32, 45). Furthermore, the deficiency in IFN- synthesis in MF is more attributable to the CD4⁺ T cells than to the CD8⁺ T cells (Table 5). There were fewer IFN--producing CD8⁺ T cells in patients with SS than in other groups, and this may account for the suppressed cytotoxic-T-lymphocyte activity in these patients (39).

In healthy individuals, whereas stimulation with antigen resulted in a higher production of IFN- by CD7 T cells (34), IL-4 was preferentially produced by CD7⁺ T cells following stimulation with lectin (4). Previous reports based on a variety of methodologies have generally shown patients with SS to have a type 2 cytokine response profile (45). Accordingly, we expected to find the majority of IL-4 producers to be CD7 T cells in patients with SS. Instead, we found the total number of IL-4⁺ T cells to be similar for all three study groups (Table 6). There were few IL-4⁺ CD7 T cells in patients with SS even through the majority of T cells of patients with SS were of CD7 phenotype (Table 2). The synthesis of IL-4 by T-cell lines is known to be transient, and not all Th2 cells are primed to produce this cytokine (29). Hence, our observation of a few IL-4⁺ T cells in all study groups may be due to the low frequency of IL-4-synthesizing Th2 cells in peripheral circulation coupled with the transient synthesis of this cytokine upon PMA stimulation. Furthermore, our method of evaluating IL-4 synthesis is a snapshot of cytokine production in cells and is unlikely to reflect the amount of IL-4 accumulated over an extended period in cultures.

Our observations did not favor the previously reported Th0-Th2 profile of CD7 cells in patients with SS and raised the possibility that CD7 T cells, presumably leukemic in patients with SS (15), could be distinct from Th0-Th2-like T cells of healthy donors. Moreover, there were more IL-4⁺ CD7⁺ than IL-4⁺ CD7 T cells in controls, indicating that CD7⁺ T cells could have a greater contribution to the overall production of IL-4. In conclusion, we found no defect in the ability of CD7 T cells in patients with SS to synthesize IL-2, even though a defect in the production of IFN- in these patients was found in this and other studies. Our data expand on the observations of others by showing that both CD7⁺ and CD7 T-cell subsets are less capable of synthesizing IFN- than those of patients with MF and control subjects. These data support the hypothesis that disease progression from MF to SS in patients with CTCL may be associated with the defective IFN- production accompanied by a decrease in cytotoxic T lymphocyte activity (39). Finally, our experimental approach can be used to address the association of cytokine induction by bacterial superantigens and pathogenesis in CTCL.