The epithelium represents a unique lymphoid compartment containing a distinct population of mucosal lymphocytes, the intraepithelial lymphocytes (1, 2). These T cells are nearly exclusively found in epithelial tissues, where they reside in intimate association with epithelial cells (3, 4). Most CD8+ T cells in or adjacent to the intestinal epithelium and 40–50% of the CD4+ T cells in the intestinal lamina propria express the integrin CD103 that mediates binding to E-cadherin at the basolateral side of the epithelium (5). CD103 is also expressed by lymphocytes within the bronchial epithelium, by some alveolar wall lymphocytes, and by bronchoalveolar fluid T cells (6, 7). However, in bronchoalveolar lavage fluid, the relative amount of CD103-expressing T cells is very different between CD4+ and CD8+ T cells. Most of the CD8+ T cells express this integrin independently of the disease of the lung (8). In contrast, the proportion of CD4+ T cells expressing CD103 is significantly higher in diseases associated with pulmonary fibrosis than in non-fibrotic diseases or healthy controls (8, 9). However, very little is known about the phenotype of CD103+ T cells and the conditions that lead to the striking accumulation of these cells in pulmonary fibrosis.
Activation of T cells is followed by the sequential expression of mostly well-defined activation markers. For instance, CD69 and CD25 are expressed within 24 h after activation, whereas the expression of human leukocyte antigen (HLA)–DR requires approximately 4 days. In contrast, VLA-1 and VLA-2 can be detected only several weeks after activation (10). The integrin CD103 also has been reported to be an activation marker on T cells and is expressed on the cell surface starting 2 days after in vitro activation (8). It is known that the costimulatory action of transforming growth factor (TGF) β1 and T-cell receptor–mediated signals are required for the expression of CD103. Functionally, this integrin plays an essential role as a homing receptor for epithelial tissue because CD103 knockout mice show a significant reduction in the number of intraepithelial lymphocytes (11). In addition, CD103 is critically important for the migration of CD8+ effector T cells into epithelial compartments (12).
Because the CD103+/CD4+ T lymphocyte population of the lung preferentially expands in fibrotic lung diseases, we hypothesized that this T-cell population might play a unique role in these disorders. Therefore, it was of interest to analyze the activity and surface receptor expression of these cells. We found that the CD103+ and CD103− T-cell subpopulations express a marker for memory T cells, CD45RO. To further characterize these cells, we also assessed their activity by analyzing activation markers by flow cytometry and cytokine production pattern with respect to the differentiation into T-helper 1 (Th1) or Th2 cells.
- Top of page
- MATERIALS AND METHODS
- LITERATURE CITED
The integrin CD103 functions to promote T-cell migration into the epithelium and is the requisite homing molecule of T lymphocytes to the epithelium (11, 12). Although it is expressed on only a small proportion of blood T cells, it is induced upon activation of the T cells in the presence of transforming growth factor (TGF) β1 (8, 18). In previous work we reported that the regulation of CD103 expression differs between CD4+ and CD8+ T cells. Moreover, we and others reported that the proportion of CD103+ T cells in the BAL within the CD4+ T-cell population increases significantly in interstitial lung diseases (8, 9).
The data presented show that the CD103+ and CD103− CD4+ BAL T-lymphocyte subpopulations are highly activated CD69+ and CD45RO+ T lymphocytes. There were, however, some remarkable differences. CD25, CD27, and CD62L were almost only expressed on CD103− in comparison with CD103+ CD4+ T cells. CD28 expression was high on CD103− but low on CD103+ T cells. In contrast, CD101, VLA-1, and VLA-2 were expressed on CD103+ and significantly less on CD103− T cells. Despite these differences, the CD103+ and CD103− CD4+ BAL T-cell populations showed characteristics of Th1 cells, with IFN-γ and no IL-4 mRNA production. In addition to the production of IFN-γ mRNA, we detected IL-10 mRNA. These data propose that the CD103+ CD4+ T lymphocytes are a phenotypic distinct memory or effector T-cell population with respect to the expression of this integrin and with respect to the expression of receptors characteristic for naive or activated CD4+ T lymphocytes.
During the response to antigen, the numbers of activated cells increases rapidly. This expansion is driven by engagement of T-cell receptors and costimulatory proteins such as CD28, CD27, CD134, 4-1BB, LFA-1, and CD2 on the surface of antigen-engaged T cells (19). Activated antigen-specific T cells are short-lived, and death of these massively expanded activated cells probably serves to control inflammation. However, a small proportion of the activated T cells differentiated into memory T cells characterized by the expression of CD45RO. In the present study the CD103− and the CD103+ were CD45RO+, which classifies them as memory cells. It has been described that memory cells can be distinguished by their expression of CD62L and CCR7. CD62LhiCCR7+ are central memory cells and CD62LloCCR7− are effector memory cells (20, 21). The expression of CD62L was low on the CD103− CD4+ T-cell population and nearly absent on the CD103+ T cells. This finding and the one that both cell populations produce mRNA for IFN-γ and no IL-2 suggest that both cell populations are effector memory cells. However, we did not analyze the expression of CCR7.
Peripheral “memory” CD45RO+ human CD4+ T cells express CD28 (22). In this study the BAL “memory” CD45RO+ CD4+ T cells varied with respect to CD28 expression and allowed separation into two groups of cells: one that expressed CD28 but not CD103 and another that expressed low CD28 and high CD103. Low expression of CD28 has been described on CD4+ T cells (23). The significance of this finding is not known. However, T cells from CD28-deficient mice showed a reduced proliferative response to antigen in vitro (24). Therefore, CD103+ CD28low T cells may represent a differentiated population of low proliferation rate. This possibility would be in agreement with the notion that memory T lymphocytes are slowly but continuously dividing cells (19). However, we did not teste the proliferative capacity of the BAL CD4+ T cells, and the relative expansion in fibrotic lung diseases remains to be analyzed. CD101 and CD103 have been described as costimulatory molecules for the activation of T cells. High expression of these molecules therefore may substitute for the lack of CD27 and CD28. It has been shown that the antibodies 2E7 and M290 against CD103 act as costimulatory signals with anti–T-cell receptor antibodies to increase the lytic machinery of intraepithelial CD8+ cells (25). This finding was confirmed by the experiments of Sarnacki et al. who described a strong synergistic effect of CD3-induced activation and CD103 crosslinking on the proliferative response of intraepithelial lymphocytes (26, 27). In addition, the molecule CD101 seems to be a strong costimulatory molecule for T cells (28). A gene that predicts a seven-immunoglobulin domain chain-like structure encodes CD101. It has restricted expression predominantly on mucosal T lymphocytes and appears to have a costimulatory function of special relevance for CD28− T cells and for mucosal lymphocytes. The expressions of CD101 and CD103 may explain the strong activation status of the CD103+ T cells in the absence of CD25, CD27, and CD28. More experiments are required to identify the difference in activity between the CD103+ and CD103− T lymphocytes.
Another finding of the study was that the CD103+/CD4+ T cells express VLA-1 and VLA-2, activation markers expressed several weeks after activation (10). These molecules are expressed after prolonged activation of T cells and indicate the presence of “older” or differentiated cells. This is supported by the observation that repeated activation induces the loss of CD28 (29). Therefore, CD103+/CD4+ T cells appear to participate in fibrogenic inflammation and may have developed as consequence of recurring contact with activation signals such as cytokines or other triggers. However, it is not clear whether these cells represent pro-inflammatory or protective cells. Data from work with experimental colitis indicated that the pro-inflammatory component might be dominating (30). In contrast to this finding, CD103-deficient mice showed increased skin inflammation, suggesting a protective role of these cells (31). Aside from the phenotypic differences between CD103+ and CD103− CD4+ T lymphocytes, we did not identify factors that distinguish between these two populations, even when cells obtained from non-fibrotic conditions were analyzed. Both subpopulations showed expression of mRNA for inflammatory cytokines and receptors. There were, however, differences between some diseases with regard to the CD103+ and CD103− CD4+ T lymphocytes. We found IL-2 mRNA only in patients with sarcoidosis, and TNF-α mRNA was absent in IPF patients. But the small number of patients does not allow generalizing this observation.
In addition to the loss of CD25 and CD28 expression, the CD103+ CD4+ T cells in the BAL fluid did not express CD27. It has been reported that loss of CD27 expression is irreversible and represents terminal effector T-cell differentiation (32, 33). This observation may suggest that the CD103+ CD4+ T lymphocytes are terminally differentiated effector T cells. Therefore, the increase in CD103+/CD4+ T cells may be a consequence of expansion and accumulation of effector CD4+ T cells that might be involved in the process of lung fibrosis. However, the CD103− CD4+ T-cell population also showed signs of high activation and expresses CD27. Consistent with the work from Ludvikson et al. (30), this finding suggests that CD103+ T lymphocytes are effector cells and CD103− comprise the memory cell population. More data are required to confirm this hypothesis because we could not identify differences between the CD103+ and CD103− populations with respect to pro-inflammatory parameters.
In summary, the results described in this paper demonstrate that the CD103− and the CD103+ BAL CD4+ T-cell populations are highly active cells. The CD103+ population displays a distinct phenotype of a continuously activated long-living cell that may exert effector function and may be distinct from the CD103− subpopulation. Newly discovered markers may provide an opportunity to distinguish memory from effector T cells (34, 35). Although detection of these cells may provide a diagnostic marker when interstitial lung diseases are expected, their functional involvement in this process warrants further investigation.