Director, Center for Biologic Nanotechnology University of Michigan Medical School Ann Arbor, Michigan 48109
Address correspondence and requests for reprints to: James R. Baker, Jr, M.D., 9220 MSRBIII; Ann Arbor, Michigan 48109-0648. E-mail: jbakerjr{at}umich.edu Internet:
![]() |
Introduction |
---|
![]() ![]() ![]() |
---|
The role of the Fas apoptosis pathway in thyroid physiology and disease has been the subject of great debate over the past few years. The Fas pathway was the first identified and most completely studied signaling pathway, and it appears to have several unique functions. It is involved in immune regulation and is also one of the pathways that activate CD8 cytotoxic T cells, used to induce apoptosis in target cells (3). FasL also has been reported to protect immune privileged sites from lymphocytes (4). For these reasons, Fas has received much attention as a potentially important pathway in autoimmune thyroid disease. Some authors have suggested that the Fas pathway plays a central role in mediating apoptotic cell death in a range of thyroid disorders, including thyroiditis (5). However, the data supporting this hypothesis has been inconsistent and is weakened by technical problems. Multiple factors that regulate Fas pathway activity in thyroid cells have been proposed, but the factors are not consistent between different studies. Further, a report, indicating that the expression of FasL on thyroid cells led to "fratricide" of thyroid cells during thyroiditis, created great interest but was criticized because it employed nodular goiter cells as controls and used antibodies that appeared to recognize a protein that was not Fas ligand (5, 6). These concerns have led to a more circumspect perspective on the function of the Fas pathway in thyroid disease. Two articles published in this issue of JCEM (see pages 0000 and 0000) clarify a number of points concerning the presence and potential function of different components of the Fas pathway in the thyroid and suggest potential roles for these molecules in thyroid disease.
The first manuscript, by Hiromatsu et al. (see page 2896), examines the expression of various components of the Fas pathway in thyrocytes with Graves disease. These investigators find that Fas is present in normal thyroid, but FasL is only expressed in thyroid tissue derived from Graves disease. Importantly, the expression of FasL on thyroid follicular cells was associated with higher levels of apoptosis in Graves disease. However, only a minority of thyrocytes in Graves tissue (521%) express FasL and could only find mRNA for ligand in two of five samples by Northern blot and five of seven samples by reverse transcriptase PCR. They were able to show that the FasL expressed on Graves thyrocytes was functional in that it induced apoptosis in target cells transfected with Fas antigen. However, thyrocyte-expressed FasL did not induce apoptosis in thyrocytes, and the inhibition of this activity by neutralizing antibody was not complete. The reason that the Fas pathway appears to be nonfunctional in thyrocytes was not investigated, but the authors suggest it might result from decreased Fas antigen expression or the presence of pathway inhibitors in thyroid cells. The authors conclude that the expression of FasL may be significant in blocking cytotoxic activity in Graves disease by destroying infiltrating, activated T cells. This would result in a predominantly TH2 response and could explain the predominance of antibody effects in this autoimmune disorder.
The other article, by Mitsiades et al. (8) (see page 2924) examines the expression of FasL in thyroid carcinomas. These investigators also show that FasL was present only in thyroid carcinomas and not in normal thyroid cells, and was expressed to varying degrees in different types of cancer. Hürthle cell cancer had the highest expression, followed by papillary, and then follicular tumors. The investigators indicate that the expression of FasL by tumors, as detected by anti-FasL or in situ hybridization for FasL, was predominantly on the periphery of thyroid tumors. This suggested the hypothesis that the thyroid cancer cells were fighting infiltrating T cells by expressing FasL at the interface between the thyroid tumor and the lymphocytes. Of interest, they confirmed an earlier report from my laboratory that the Fas pathway is usually nonfunctional in thyroid cells due to blockade by a labile protein inhibitor (9). This renders the thyrocytes and thyroid cancer cells resistant to Fas-mediated apoptosis, even from ligand expressed on thyroid cells. Finally, the investigators also showed expression of FasL in a number of tumor cell lines and documented that the FasL expressed by the tumor cell lines was functional. However, using a different technique (nuclease protection), my own investigations have been unable to confirm the expression of FasL mRNA in these lines (P. Arscott, personal communication). This suggests either that there are low levels of FasL expression in these cells or that expression is variable dependent on culture conditions.
Both reports suggest that, under certain conditions, thyroid cells can express functional FasL. The trigger for this expression is not defined by either author, although Hiromatsu suggests that DNA damage may be one trigger. Why this ligand is only seen in diseased thyroid cells is also not defined. Whether there is some physiologic role for FasL expression in these states is an open question. These two reports resolve a number of issues related to Fas expression in the thyroid. Both show definitive expression of Fas antigen by normal thyroid cells, as well as by diseased thyroid cells. This finding is in accord with the majority of most reports and consistent with what is observed in other epithelial cells (9, 10, 11). It is at odds, however, with Giordano et al. (5), who suggested that Fas antigen is only expressed in the thyroid during inflammatory conditions. This discrepancy now appears to be due, at least in part, to Giordanos use of goiterous cells as normal controls. The two current publications also confirm that FasL is not expressed on normal thyroid follicular cells. Findings of FasL protein on normal thyroid tissue appear related to technical problems, including the use of nonspecific antibodies (6). However, reports of FasL mRNA in normal thyroid cells are more difficult to dismiss, as they appear to accurately document expression (5). In this regard, it is important to remember that the finding of specific mRNA for a protein using highly sensitive techniques (such as RT-PCR) is often not associated with significant protein expression. As a result, the physiologic significance of FasL mRNA in the thyroid is unclear. Together, this indicates that thyroid cells are similar to most other types of epithelial cells in that they express Fas but not FasL (2, 12).
The two other points in common between these manuscripts are significant. The fact that the Fas pathway is blocked in the basal state in thyroid cells suggests that, if Fas is involved in thyrocyte destruction, the pathway must be activated or the inhibition reversed. The presence of the inhibitor makes thyroid cancer cells resistant to the induction of apoptosis through this pathway. These studies indicate that all of the components of the Fas pathway are present on thyroid cells, but the pathway is not functionally active unless unique physiologic conditions occur that remove the labile protein inhibitor. This finding may have therapeutic significance, as blockade reversal is important to facilitate immune-mediated apoptosis in these cells. While the conditions for reversal of Fas pathway inhibition are not known, preliminary data from our laboratory has suggested that certain combinations of inflammatory cytokines may be one such facilitating condition (13). This might suggest that the Fas pathway is not normally involved in apoptosis in the thyroid but can, under certain inflammatory conditions, be activated and used by the immune system to destroy thyroid cells. The regulation of Fas signal transduction in the thyroid is therefore important to evaluate.
While both publications raise important questions concerning the role of the Fas pathway in different thyroid diseases, there are a number of caveats that must be observed for each study. In the Hiromatsu study, it is intriguing that only a minority of thyroid cells expresses FasL, and the trigger for this expression is not defined. It is disconcerting that FasL mRNA was found in only a portion of specimens, even when probed with extremely sensitive techniques such as RT-PCR. This suggests that the expression of FasL may be transient and present only in a minority of cells under unknown unique conditions in Graves disease. It also argues that the regulation of this process would be important to understand its significance. The other study, by Mitsiades et al. (8), proposes that, in thyroid cancer, poorly differentiated tumors might escape immune surveillance in part because of the expression of FasL. This hypothesis is consistent with the finding that the Fas pathway is blocked in thyroid cells, thus allowing FasL on thyroid cells to mediate apoptosis of infiltrating lymphocytes. This would be similar to what is observed in certain privileged sites in the body. However, there are a number of concerns to be resolved concerning this hypothesis. The authors suggest that there is more expression of FasL in the periphery of thyroid tumors. However, this could also be the result of shed FasL from lymphocytes or lymphoid infiltrating cells. Although the thyroid cancer cell line studies would seem to indicate that the increased ligand was definitely coming from thyroid tumors, all of these cell lines were passed for long periods of time before analysis, and the relevance of this finding to tumors in vivo is, therefore, not clear. It is also of interest that there is an intense lymphocytic response in many of the tumor samples the authors evaluated. This suggests that, at least at the time of sampling, the lymphoid cells were not markedly deleted in vivo, despite the expression of FasL.
Thus, the real issue raised by both these publications is whether or not either Graves thyrocytes or thyroid cancer cells are using expression of FasL to escape immune attack. This question is still open, as the killing of Fas transfected or Fas expressing targets in vitro is not the same as killing invading activated, cytotoxic T-cells in vivo. Given this, it is not clear that the FasL on the thyrocytes is able to induce death in lymphocytes in a manner similar to the targets that were tested. It is possible, however, that even a minority of the cells expressing functionally active FasL could delete a large number of infiltrating thyroid cells and seriously alter the immune response in either Graves disease or thyroid tumors. Given the uniqueness of the immune responses in Graves disease and to some types of thyroid tumors, a mechanism like this may be important. However, other apoptosis pathways analogous to Fas may be more important in particular diseases. An example of this is the TRAIL pathway, which our group has demonstrated to be present in the thyroid, and which is thought to be particularly important in inducing apoptosis in cancer cells (14). Therefore, the Fas pathway cannot be examined without also evaluating its relative contribution compared with these other death pathways.
In conclusion, these two studies provide important and provocative information showing that the Fas pathway, its ligand and receptor, are expressed in the thyroid in a highly regulated and variable manner. These works also suggest that a more fundamental understanding of the role of Fas in thyroid cell physiology and disease could be obtained by evaluating the regulation of this pathway and its expression in thyroid cells. Clearly, further studies are necessary to determine whether or not thyroid cells can destroy infiltrating lymphocytes in vivo in a manner similar to other Fas positive target systems in vitro. Most importantly, these studies suggest that the thyroid seems to be an excellent model for evaluating the role of Fas in different diseases, and that transgenic models where particular pathways are altered or augmented in the thyroid might provide further insights to these processes.
Received June 18, 1999.
Accepted June 21, 1999.
![]() |
References |
---|
![]() ![]() ![]() |
---|