INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: ST2; lung cancer; pleural effusion; helper T lymphocyte type 2

An accumulation of lymphocytes, especially helper T lymphocytes, in pleural fluid frequently occurs in neoplastic effusions secondary to direct pleural involvement and/or metastases from malignancies (1, 2). It is thought that the cytokines released from the accumulated lymphocytes may play some role in the interaction between lymphocytes and tumor cells in malignant pleural effusions (3, 4). Shimokata and colleagues demonstrated that low concentrations of interleukin (IL)-2 and interferon (IFN)- were found in malignant effusions, especially when compared with concentrations of these cytokines in tuberculous pleurisy (TB) (5). Chen and coworkers reported impairment of local cellular immunity with increased IL-10 and undetectable IL-12 in neoplastic pleural effusions (6). In addition, Nakamura and associates demonstrated that eosinophilia in the pleural fluid induced by intrapleural administration of IL-2 to patients with malignant pleurisy was due to the eosinophil colony-stimulating factor activities of various components, including IL-5, IL-3, and granulocyte/ macrophage colony-stimulating factor (7). These findings suggest that in malignant pleural effusions, local immune reactions may likely favor the T-helper type 2 (Th2) pathway over the Th1 pathway.

Three distinct types of ST2 gene products, a soluble secreted form (ST2), a transmembrane receptor form (ST2L), and a variant form (ST2V), have been cloned (8-10). The ST2 gene, also known as T1, Fit-1, and DER4, was originally identified as a gene induced by serum stimulation of fibroblasts (8, 11-13) and has recently been demonstrated to be overexpressed preferentially on Th2 effector cells but not on Th1 cells (14, 15). Furthermore, in a murine model of Th2-dependent allergic airway inflammation, administration of a neutralizing antibody against ST2 partially inhibited the development of Th2 effector functions in vivo (16, 17). We recently demonstrated that serum human ST2 concentrations in subjects with asthma were more significantly increased than those in healthy control subjects (18). Therefore, ST2 seems to play an essential role in the development of Th2 responses in pathologic conditions.

These findings suggesting that ST2 expression is critical for Th2 responses led us to evaluate the relationship between ST2 expression and Th2 dominance of effusion-associated lymphocytes in malignant effusion. We recently generated an enzyme-linked immunosorbent assay (ELISA) system to quantify the soluble form of ST2 protein (19). In the present study, using this system, we measured ST2 protein concentrations in the sera and pleural effusions of patients with carcinomatous pleurisy (CA) and compared them with those of patients with TB, which obviously favors a Th1-like immune response (20, 21). The results demonstrated that pleural ST2 concentrations in CA were significantly higher than those in TB. Th2 dominance and ST2 gene expression of CD4⁺ lymphocytes isolated from malignant pleural effusions were also analyzed in this study.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

The study protocol was approved by our institutional review board for human studies, and informed consent was obtained from all subjects. The study sample consisted of 30 patients, for whom a definitive diagnosis was ultimately reached (CA, n = 17; TB, n = 8; congestive heart failure [HF], n = 5). The diagnostic criteria used were described previously (21). The primary lesions in all cases of CA proved to be histologically confirmed lung carcinoma (adenocarcinoma, 9 cases; squamous cell carcinoma, 4; small cell carcinoma, 2; large cell carcinoma, 2). Pleural effusions were defined as exudative or transudative, using Light's criteria (22).

Preparation of Mononuclear Cells and Determination of Lymphocyte Subpopulations in Pleural Effusion

Pleural effusion specimens were collected from all subjects. Each sample was examined for the total cell number and differential cell counts. Lymphocytes were separated from pleural effusion cells by centrifugation on a discontinuous Percoll (Pharmacia, Uppsala, Sweden) density gradient as previously described (23) and were processed for flow cytometry to determine lymphocyte phenotype. Fluorescein isothiocyanate or phycoerythrin-conjugated monoclonal antibodies (anti-IgG1, -IgG2a, -CD3, -CD4, -CD8, -CD20, -CD56) were purchased from the supplier and used according to the manufacturer's protocol to detect the corresponding surface antigens (Becton Dickinson, Mountain View, CA). The antigen immunofluorescence of the cells was analyzed by a FACstar Plus Cytofluorometer (Becton Dickinson).

Ex Vivo Cytokine Production of CD4⁺ T Lymphocytes Isolated from Pleural Effusion

CD4⁺ T cells were isolated from mononuclear cells in pleural effusion using a MACS CD4⁺ cell isolation system according to the manufacturer's protocol (Miltenyi Biotech., Auburn, CA), and the cell viability was greater than 95% as determined by Trypan blue dye exclusion test. The isolated CD4⁺ T cells were cultured for six hours at 5 × 10⁶ cells/ml in RPMI 1640 medium in the absence or presence of phorbol myristate acetate (PMA, 50 ng/ml) and Ca ionophore (200 ng/ml), and the supernatants were frozen at 80° C until assay.

Measurement of Human ST2 Protein and Cytokine Concentrations

The concentration of soluble human ST2 protein in serum and in pleural effusions was measured by a sandwich ELISA we developed previously (19). The quantification of IFN-, IL-4, IL-10, and IL-12 production in sera, pleural effusions, and a conditioned medium of cultured lymphocytes was performed by a sandwich ELISA (BioSource, Inc., Camarillo, CA).

Analysis of mRNA Expression for Human ST2 by Reverse Transcription-Polymerase Chain Reaction

Total cellular RNA was extracted from isolated CD4⁺ T cells using Trizol reagent (Life Technologies, Inc., Gaithersburg, MD). Five hundred nanograms of RNA was reverse transcribed and amplified by reverse transcription-polymerase chain reaction (RT-PCR; Takara, Tsukuba, Japan) under the following conditions: 94° C, one minute; 57° C, one minute; and 72° C, one minute, for 36 cycles. The PCR products of ST2 and -actin are fragments 659 and 547 bp in length, respectively.

Statistical Analysis

Data were expressed as mean ± SEM. Multiple comparisons were performed by one-way factorial analysis of variance (ANOVA) with post hoc tests. The correlation of ST2 concentrations with several variables was conducted using Pearson's correlation test. p < 0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Characteristics and Lymphocyte Subpopulations in Pleural Effusions of Each Group

Total and differential cell counts in each group of subjects in the present study are summarized in Table 1. Subjects with TB showed a marked elevation of total cell counts, and a large proportion of these cells was lymphocytes, with some macrophages and neutrophils. Subjects with CA showed a large proportion of lymphocytes and macrophages in the pleural space. Importantly, on cytologic examination, malignant cells were found in 15 subjects. Subjects with transudative effusions secondary to HF showed a predominance of lymphocytes without sequestration of neutrophils or macrophages in the pleural space.

The dominant cell type within pleural fluids is lymphocytes. Absolute lymphocyte counts evidenced the highest values in TB, showing a significant increase in comparison with absolute lymphocyte counts in CA (p < 0.0001) and in HF (p < 0.0001) (Table 2). Evaluation of the lymphocyte subpopulation in pleural effusions of different etiologies showed that T cells, especially CD4⁺ T cells, were dominant in all situations. More specifically, a significant rise of CD4⁺ T cells and a significant decrease of CD8⁺ or CD56⁺ cells were observed in CA in comparison with TB (Table 2).

Cytokine Concentrations in the Sera and Pleural Fluid

The concentrations of IFN-, IL-4, IL-10, and IL-12 in the sera and pleural fluid in the CA, TB, and HF groups were measured by ELISA. IFN- concentrations in pleural fluids (p-IFN-) were significantly higher in the TB group than in the CA or HF group (p < 0.0005, Figure 1A), whereas no significant differences in serum IFN- (s-IFN-) concentrations were observed among the three groups (mean s-IFN- ± SEM: CA, 3.5 ± 0.9 pg/ml; TB, 3.0 ± 1.0 pg/ml; HF, 3.5 ± 0.8 pg/ml). No significant differences in the IL-10 concentrations in sera (s-IL-10) or pleural fluid (p-IL-10) were observed among the three groups (mean s-IL-10 ± SEM: CA, 2.3 ± 1.1 pg/ml; TB, 2.0 ± 1.2 pg/ml; HF, 1.5 ± 0.9 pg/ml; p-IL-10: Figure 1B). The IL-4 concentrations in sera (s-IL-4) and pleural fluid (p-IL-4) were very low in all three groups, and no statistically significant differences were found in s-IL-4 or p-IL-4 concentrations among the groups (mean s-IL-4 ± SEM: CA, 2.2 ± 0.7 pg/ml; TB, 2.1 ± 0.2 pg/ml; HF, 2.1 ± 0.6 pg/ml; p-IL-4: Figure 1C). The IL-12 concentrations were below the minimal limits of detection in the sera and the pleural fluid in all groups (data not shown).

View larger version (12K): [in this window] [in a new window]   Figure 1.   Distribution of IFN- (A), IL-10 (B), and IL-4 (C) concentrations in the pleural effusions of subjects with CA, TB, and HF. Multiple comparisons were performed with the use of one-way factorial ANOVA followed by post hoc tests. Horizontal bars indicate mean values. *p < 0.0005.

Ex Vivo Cytokine Production by CD4⁺ T Cells Isolated from Pleural Effusions

Cytokine production in a conditioned medium of CD4⁺ T cells (5 × 10⁶ cells isolated from pleural effusions) cultured with or without PMA/Ca ionophore was compared among all groups (Figure 2). The concentration of IFN- produced by CD4⁺ T cells from the TB group was significantly higher than that from the CA or HF group (p < 0.0001) when these cells were stimulated with PMA/Ca ionophore (Figure 2A). In contrast, the level of IL-10 secretion from CD4⁺ T cells in the presence of PMA/Ca ionophore was considerably greater in the CA group than in the TB or HF group (p < 0.0001, Figure 2B). Likewise, the IL-4 level produced by these cells (stimulated with PMA/Ca ionophore) derived from the CA group was statistically higher than that from the TB (p < 0.005) or HF group (p < 0.05) (Figure 2C). IL-12 concentrations in the supernatant were undetectable in all groups (data not shown).

View larger version (17K): [in this window] [in a new window]   Figure 2.   Ex vivo cytokine production by CD4+ T lymphocytes isolated from the pleural effusions of subjects with CA, TB, and HF. CD4+ T lymphocytes (5 × 106) were isolated from pleural effusions and cultured in 1 ml of RPMI medium 1640 for six hours with or without PMA/Ca ionophore. Cytokine concentrations in the supernatant were measured by ELISA. Multiple comparisons were performed using one-way factorial ANOVA with post hoc tests. *p < 0.0001, p < 0.005, p < 0.05. Ca = Ca ionophore.

Distribution of Soluble ST2 Concentrations in the Sera and Pleural Effusions

A total of 60 samples (1 serum and 1 pleural fluid sample per subject) was taken from the 30 subjects (CA, n = 17; TB, n = 8; HF, n = 5). The distributions of soluble ST2 concentrations in the sera and pleural effusions of each group are shown in Figure 3. No significant differences in serum ST2 (s-ST2) concentrations were observed among the three groups (Figure 3A), but the ST2 concentration in the pleural effusions (p-ST2) of the CA group was significantly higher than that in those of the TB (p < 0.001) or HF group (p < 0.01) (Figure 3B).

View larger version (18K): [in this window] [in a new window]   Figure 3.   Distribution of soluble ST2 concentrations in the sera and pleural effusions of subjects with CA, TB, and HF. "Pleural ST2" means ST2 concentrations in the pleural effusion. Multiple comparisons were performed by one-way factorial ANOVA with post hoc tests. Horizontal bars indicate mean values. *p < 0.01, p < 0.001.

Correlation of p-ST2 Concentrations with Cell Populations or Cytokine Concentrations in Pleural Effusions

The correlation of p-ST2 concentrations with cell populations or cytokine concentrations in subjects with pleural effusion was analyzed using Pearson's correlation coefficient (Figure 4). The p-ST2 concentrations did not correlate with the percentage and absolute cell number of lymphocytes in pleural effusions. p-ST2 concentrations were significantly correlated with the percentage of CD4⁺ T cells, but not with the absolute cell number of CD4⁺ T cells, in pleural effusions. Further, a statistically significant inverse correlation was observed between the p-ST2 and the p-IFN- concentrations, whereas no correlation was found between the p-ST2 and p-IL-10 concentrations.

View larger version (30K): [in this window] [in a new window]   Figure 4.   Correlation of p-ST2 concentrations with the cell populations and cytokine concentrations in pleural effusions. Pearson's correlation coefficient (r) was calculated for the relation of ST2 concentrations in pleural effusions to several variables. p < 0.05 was considered statistically significant.

mRNA Expression of ST2 in CD4⁺ T Cells Isolated from Pleural Effusions

We analyzed the expression of human ST2 mRNA in CD4⁺ T cells isolated from malignant and tuberculous pleural effusions (Figure 5A). The expression of ST2 mRNA in the CA group was significantly upregulated in comparison with that in the TB group (p < 0.001, Figure 5B).

View larger version (33K): [in this window] [in a new window]   Figure 5.   (A) Detection of ST2 mRNA expression of CD4+ T lymphocytes isolated from pleural effusions of subjects with CA and TB by RT-PCR. The cDNA samples obtained by RT-PCR for ST2 and -actin were loaded on a 2.5% agarose gel. (B) Densitometric analysis of ST2 mRNA expression. The mean DNA intensity levels ± SEM in each group. The intensity of DNA bands was quantified by densitometry (National Institute of Health IMAGE1.16). The relative level designates the ratio obtained from the density of the band of ST2 mRNA expression standardized with that of the -actin mRNA expression level. *p < 0.001.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study, we investigated the dominant subset of helper T cells in malignant pleural effusions and determined the soluble ST2 concentrations in the sera and pleural fluid of subjects with CA, TB, and HF, respectively. The results demonstrated that effusion-associated CD4⁺ lymphocytes in CA, which produced IL-4 and IL-10 ex vivo but did not produce IFN- or IL-12, thus polarized preferentially to Th2 cells and that p-ST2 concentrations were significantly higher in the CA group than in the TB or HF group.

Human immunity has been found to have two major components, cellular immunity and humoral immunity. The Th1 pathway favors cellular immunity, whereas the Th2 pathway favors humoral immunity (24). Th1 cells produce IL-2, IFN-, and tumor necrosis factor-, whereas Th2 cells produce IL-4, IL-5, and IL-10 (24). Although lack of cellular immunity has been reported in malignant pleural effusions (6), thus far there have been very few reports analyzing lymphocyte subtypes in these effusions from the viewpoint of Th1/Th2 balance. Some reports have demonstrated that the IL-10 concentrations in malignant pleural effusions increased significantly as compared with those in sera, whereas the IL-12 concentrations were below minimal detectable concentrations in both the sera and effusions, and suggested that helper T cells in malignant pleural effusions may be Th2-dominant (7, 23). However, comparative studies have found no significant difference between subjects with tuberculosis and cancer regarding p-IL-10 concentrations, though p-IFN- concentrations were found to be higher in TB than in malignant pleurisy (25, 26), as was also shown in our results (Figure 1B). Furthermore, because cancer cells and macrophages as well as helper T cells may produce IL-10 (27, 28), increased concentrations of IL-10 in an effusion do not directly indicate Th2 dominance in lymphocytes. To address these issues, we analyzed ex vivo cytokine production by CD4⁺ T cells isolated from effusions caused by CA, TB, and HF and compared the concentrations of these cytokines among these three groups. Our results clearly demonstrated that CD4⁺ T cells isolated from CA effusions, stimulated with PMA/Ca ionophore, produced higher amounts of IL-4 and IL-10 than did the effusions from the TB or HF group, whereas IFN- was more significantly produced by CD4⁺ T cells from effusions of the TB group than from those of the CA or HF group. These results strongly suggest that CD4⁺ T cells present in malignant pleural effusions are Th2-dominant (but may not be activated enough to produce Th2 cytokines spontaneously). In contrast, CD4⁺ T cells in tuberculous pleural effusions are Th1-dominant and may be activated enough to produce Th1 cytokines, because p-IFN- concentrations were higher in the TB group, and ex vivo production of IFN- by CD4⁺ T cells in the TB group was more significant even without stimulation, compared with those in the CA or HF group.

Because ST2 has been reported to be preferentially expressed on Th2 in murine cell lines (14, 15), Th2 dominance in lymphocytes of malignant pleural effusions led us to analyze the concentrations of the soluble form of ST2 protein in the effusions. As we expected, our results demonstrated that p-ST2 concentrations in the CA group were significantly higher than those in the TB or HF group. Further, the ST2 mRNA expression of CD4⁺ T cells isolated from CA effusions was upregulated compared with that from the TB group. These results suggest that increased production of ST2 may be derived from Th2 cells, which are dominant among the CD4⁺ T cells in malignant pleural effusions.

Upon encountering an antigen, the naive CD4⁺ T helper precursor (Thp) cells enact a specific genetic process that results in differentiation toward the Th1 or Th2 lineage. This differentiation process can be influenced by antigen concentration, through the ligation of costimulatory molecules, and by cytokines (29). It is thought that pleural effusion is a good model by which to understand the direct interaction between immune cells and antigen (pathogen or tumor antigen), and the role of cytokines in it for Th differentiation, on the basis of the fact that a pleural effusion resulting from cancer presents an environment in which both effector lymphocytes and target tumor cells coexist in a defined space. Our results demonstrating that CD4⁺ lymphocytes in a malignant pleural effusion produced IL-4 and IL-10 ex vivo, but not IFN- or IL-12, suggest that impairment of the Th1-mediated immune response, which is essential for antitumor immunity, results in a progression of cancer to the pleural cavity and a subsequent accumulation of effusion. IL-10 has various immunosuppressive effects on T cells (30), whereas IL-12 has an obligatory role for the induction of Th1 cells and for the differentiation and activation of natural killer cells and cytotoxic T cells (31, 32). Lymphocytes infiltrating the pleural cavity may be susceptible to local stimulation by these cytokines. Increased IL-10 and undetectable IL-12 in the malignant pleural effusions observed in our study are consistent with the findings of suppressed antitumor immunity in effusions (3, 4, 6, 23). Furthermore, an increase of p-ST2 suggests depression of the Th1-mediated antitumor immune response in malignant pleural effusions. This is supported by our findings here that ST2 (released dominantly from Th2 cells) concentrations inversely correlated with IFN- (for Th1-mediated immune response) concentrations in pleural effusions.

In addition to its relationship to antitumor immunity in malignant pleurisy, the Th1/Th2 cytokine balance in the pleural space plays another important role in the pathophysiologic process of pleural disease. The development of pleural effusions is associated with the presence of an increased number of inflammatory and immune cells in the pleural space (1, 2). Different disease entities are typically associated with the predominance of a certain cell type in this lesion (1, 2, 22). Antony and colleagues demonstrated that Th1 and Th2 cytokines influenced C-C chemokine (macrophage inflammatory protein 1 and monocyte chemoattractant protein 1) expression in pleural mesothelial cells and thus regulated the mononuclear cell migration into pleural space in TB (33). Various cells present in the pleural space may produce various cytokines or chemokines that contribute to the progress of pleural effusion (7, 26-28, 33, 34). It is important to further analyze the effect of the Th1/Th2 cytokine balance on the regulation of cytokine and chemokine production from the pleural effusion cells and the subsequent recruitment of inflammatory and immune cells to the pleural space.

In conclusion, our results showed that the p-ST2 concentrations in the CA group were significantly elevated and that CD4⁺ helper T cells present in malignant pleural effusions were relatively shifted to the Th2 lineage and significantly expressed ST2 mRNA. These findings suggest that elevation of ST2 may reflect Th2 dominance, which may subsequently suppress antitumor immunity in malignant pleural effusions. Further elucidation of the biologic and pathologic function of ST2 in pleural effusions will provide insight into the immunopathologic characteristics of malignant pleural effusions.