Potential Involvement of Fas and Its Ligand in the Pathogenesis of Hashimoto's Thyroiditis

See allHide authors and affiliations

Science  14 Feb 1997:
Vol. 275, Issue 5302, pp. 960-963
DOI: 10.1126/science.275.5302.960


The mechanisms responsible for thyrocyte destruction in Hashimoto's thyroiditis (HT) are poorly understood. Thyrocytes from HT glands, but not from nonautoimmune thyroids, expressed Fas. Interleukin-1β (IL-1β), abundantly produced in HT glands, induced Fas expression in normal thyrocytes, and cross-linking of Fas resulted in massive thyrocyte apoptosis. The ligand for Fas (FasL) was shown to be constitutively expressed both in normal and HT thyrocytes and was able to kill Fas-sensitive targets. Exposure to IL-1β induced thyrocyte apoptosis, which was prevented by antibodies that block Fas, suggesting that IL-1β-induced Fas expression serves as a limiting factor for thyrocyte destruction. Thus, Fas-FasL interactions among HT thyrocytes may contribute to clinical hypothyroidism.

The interaction of Fas (CD95/APO-1) with its ligand (FasL) regulates a number of physiological and pathological processes of cell death. Triggering of Fas contributes to the regulation of the immune response and tissue homeostasis, as well as to the immunological clearance of virus or tumor cells (1).

Hashimoto's thyroiditis (HT) is an autoimmune disorder in which destructive processes overcome the potential capacity of thyroid replacement, estimated as about 5- to 10-fold in a lifetime (2). Apoptosis has been occasionally observed in histological section of normal thyroid (3). However, apoptotic cell death is abnormally accelerated during the pathologic phases leading to clinical hypothyroidism (4).

The mechanisms responsible for thyrocyte destruction remain elusive (5). Normal thyrocytes do not express Fas (6). However, Fas is inducible in some cell types upon appropriate stimulation (1). To determine the possible involvement of Fas and its ligand in autoimmune thyroid destruction, we first analyzed Fas expression in thyroid specimens from active HT and from nontoxic goiter (NTG) patients. Immunohistochemistry of frozen thyroid sections and two-color flow cytometric analysis of dispersed thyroid follicular cells, obtained by enzymatic digestion, revealed that HT thyrocytes, identified for cytokeratin (Fig. 1A) and thyroperoxidase (TPO) positivity (Fig 1B), express large amounts of Fas on their cell surface (7).

Fig. 1.

Fas expression on thyroid follicular cells from nontoxic goiter (NTG) and Hashimoto's thyroiditis (HT). (A) Immunohistochemical analysis of thyroid cryostat sections exposed to control IgG1, anti-cytokeratin, or anti-Fas and revealed by immunoenzymatic APAAP complex. (B) Flow cytometric analysis of dispersed thyroid follicular cells. Cells were labeled with control serum or control IgG1 (dotted lines), or with anti-TPO serum or anti-Fas (solid lines). Fas expression (right) was analyzed on cells gated on thyrocyte physical parameters and TPO positivity (left). (Upper panels) Nontoxic goiter. (Lower panels) Hashimoti's thyroiditis.

The expression of Fas in thyrocytes from HT patients may be a consequence of the intense inflammatory process. We therefore examined whether the exposure of normal thyrocytes to inflammatory cytokines could induce Fas expression in vitro. IL-1β was the only cytokine, among those found in HT glands (8), able to induce Fas expression in normal thyrocytes (Fig. 2A). Other cytokines, such as tumor necrosis factor-α (TNF-α), interferon-γ (IFN-γ), IL-6, and IL-12, failed to promote Fas expression (Fig. 2A). IL-1β-induced Fas expression was detectable at doses as low as 5 U/ml, reaching high levels at concentrations around 100 U/ml (Fig. 2, B and C). Both cycloheximide and actinomycin D completely prevented IL-1β-induced thyrocyte Fas expression (Fig. 2B), indicating that new RNA and protein synthesis is required in this process. As expected, IL-1β was abundantly present in HT tissue (Fig. 2, D and E).

Fig. 2.

Fas expression on normal thyrocytes induced by exposure to IL-1β. (A) Kinetics of Fas expression on NTG thyrocytes exposed to recombinant IL-1β (200 U/ml; Genzyme) (□), IL-6 (Genzyme) (▪), TNF-α (Boehringer GmbH, Mannheim, Germany) (ˆ), IL-12 (Genzyme) (▴), and IFN-γ (Sigma) (▵). Data are presented as mean fluorescence intensity (MFI) ratio (the ratio between MFI of specific and control staining). (B) Fas expression on NTG thyrocytes exposed for 72 hours to various doses of recombinant IL-1β, in the absence (□) or presence of cycloheximide (20 μg/ml; Sigma) (ˆ) or with actinomycin D (5 μg/ml; Sigma) (▴). (C) Simultaneous expression of Fas expression and TPO on NTG thyrocytes exposed for 72 hours to IL-1β (100 U/ml), as detected by two-color flow cytometric analysis. (D and E) Immunohistochemical analysis of HT thyroid cryostat sections exposed to control IgG1 (D) or IL-1β mAb (E) and revealed by immunoenzymatic APAAP complex procedure. Anti-IL-1β (IgG1) was from Genzyme, and secondary and immunoenzymatic reagents were from Dakopatts.

These results were suggestive for a role of Fas in thyrocyte destruction. Because Fas expression does not always correlate with its apoptotic function (9), we examined whether Fas was able to transduce a death signal in thyrocytes. Triggering of Fas, after its in vitro induction by IL-1β on normal thyrocytes, resulted in massive apoptotic cell death, indicating that Fas is functional and induces apoptosis in thyrocytes (Fig. 3, A and C). Moreover, IL-1β-induced Fas expression resulted in appreciable thyrocyte apoptosis, in a dose-dependent fashion (Fig. 3, A to C). Importantly, immunohistochemical analysis of HT thyroid specimens revealed several apoptotic cells among the Fas-positive thyrocytes (Fig. 3D). By contrast, Fas-negative NTG thyrocytes did not show any sign of apoptosis (Fig. 3D) (10).

Fig. 3.

Thyrocyte apoptosis induced by IL-1β and Fas stimulation. (A) Percentage of hypodiploid nuclei of freshly separated NTG thyrocytes cultivated for various times with medium alone (▴) or with the addition of recombinant IL-1β (50 U/ml) at day −1, and anti-Fas (200 ng/ml) (•) or control IgM (□) at day 0. (B) Effect of 36-hour exposure of NTG thyrocytes to various doses of IL-1β. (C) Flow cytometric profiles of DNA content of nuclei from NTG thyrocytes exposed for 72 hours to IL-1β and IL-1β plus anti-Fas. CH-11 (anti-Fas IgM) was from Upstate Biotechnology (New York, New York). The percentage of hypodiploid nuclei was determined by hypotonic fluorochrome solution staining (50 μg of PI per milliliter in 0.1% sodium citrate plus 0.1% Triton X-100) (Sigma) and flow cyto metric analysis, as described (9). (D) Detection of in situ DNA breaks by terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) and BCIP (black) staining onNTG and HT section labeled with anticytokeratin or anti-Fas, with an avidin-biotin complex and 3-amino-9-ethylcarbazole as a substrate (red). Arrowheads indicate some apoptotic nuclei among CK- and Fas-expressing cells.

FasL expression by activated lymphocytes dictates the fate of activated Fas-sensitive lymphocytes (11). However, FasL expression is not confined to the immune system, as mouse eye and testis produce high levels of this cytokine (12). Moreover, FasL expression has been detected in hepatocytes from patients with alcoholic liver damage (13). We therefore investigated whether FasL was expressed in HT and nonautoimmune thyroid glands. Unexpectedly, FasL was constitutively expressed on both nonautoimmune and HT thyrocytes, as demonstrated by immunohistochemical analysis of frozen sections (Fig. 4A) and by two-color flow cytometric analysis of dispersed TPO-positive thyrocytes (Fig. 4B) (14). Accordingly, reverse transcriptase-polymerase chain reaction (RT-PCR) analysis performed on mRNA isolated from NTG or HT glands revealed that FasL mRNA was present in both autoimmune or nonautoimmune conditions (Fig. 4C) (15). FasL expression by HT thyrocytes was abundant (Fig. 4B), about four- or fivefold higher than that found on the small percentage of FasL-positive lymphocytes infiltrating HT glands (16). The presence of FasL was further investigated on purified normal thyrocytes. RT-PCR and protein immunoblot analysis confirmed the constitutive expression of FasL, both at the mRNA (Fig. 4D) and protein levels (Fig. 4E), in thyrocytes depleted of hematopoietic cells (15).

Fig. 4.

Constitutive expression of FasL by thyrocytes. (A) Immunohistochemical analysis of NTG, HT thyroid, or normal pancreatic (NP) cryostat sections exposed to control rabbit IgG or rabbit anti-FasL and visualized by immunoperoxidase. (B) Flow cytometric analysis of dispersed TPO-positive thyrocytes. Cells were labeled with anti-FasL (solid lines) or control IgG (dotted lines). (C) RT-PCR in thyroid tissues. Lanes 1 and 2, samples of NTG tissues from two different individuals; lanes 3 and 4, samples from two different HT patients. ϕ X174 DNA-Hae III digest was used as a size marker (M). (D) RT-PCR in normal human pancreatic islets (lane 1), Jurkat (human T cell leukemia) cells (lane 2), Jurkat cells activated with phorbol 12-myristate 13-acetate (PMA, 10 ng/ml) and ionomycin (400 ng/ml) for 4 hours (lane 3), normal thyroid tissue (lane 4), or hematopoietic cell-depleted thyrocytes (lane 5). (E) Immunoblot detection of FasL on cell lysates from Jurkat cells (lane 1), Jurkat cells activated with PMA (10 ng/ml) and ionomycin (400 ng/ml) for 4 hours (lane 2), normal thyroid tissue (lane 3), or hematopoietic cell-depleted thyrocytes (lane 4). FasL was visualized by staining with anti-FasL and the ECL detection system. Molecular sizes are indicated on the right (in kilodaltons).

To assess whether FasL expressed on thyrocytes is functional, we compared the ability of purified normal thyrocytes to kill a Fas-sensitive subline of the human T cell lymphoma HuT78 and a Fas-insensitive HuT78 variant, which expresses “death domain”-defective Fas receptors (17). Thyrocytes were able to lyse Fas-sensitive but not Fas-insensitive HuT78 cells (Fig. 5A). Moreover, cytotoxicity against Fas-sensitive HuT78 was completely prevented by the addition of monoclonal antibodies (mAbs) blocking Fas (18) (Fig. 5A), demonstrating that FasL on thyrocytes is able to induce apoptotic cell death through the engagement of functional Fas.

Fig. 5.

IL-1β-induced thyrocyte apoptosis induced by Fas-FasL interaction. (A) Cytotoxic activity of thyrocytes toward Fas-sensitive HuT78 cells (□), “death domain”-defective HuT78 cell variants (▵), or Fas-sensitive HuT78 cells in the presence of ZB4 (10 μg/ml; anti-Fas blocking mAb MBL, Japan) (18) (•). (B) Apoptotic cell death in NTG thyrocytes exposed for 36 hours to IL-6 (200 U/ml), IL-1β (200 U/ml) plus control IgG1 (10 μg/ml), or IL-1β (200 U/ml) plus ZB4 mAb. Data represent the mean ± 1 SD of five different experiments. Cell viability was determined by orange acridine-ethidium bromide staining and fluorescence microscopy analysis. (C) Flow cytometric analysis of nuclei from thyrocytes treated as in (B). Hypodiploid nuclei were evaluated by PI staining and flow cytometric analysis.

Normal thyrocytes do not undergo apoptosis, because they express negligible amounts of Fas. However, simultaneous expression of functional Fas and FasL in HT or in IL-1β-stimulated thyrocytes might result in apoptotic cell death. To address this hypothesis, we examined whether IL-1β-induced apoptosis could be blocked by interfering with Fas-FasL interactions. As shown by flow cytometry of propidium iodide (PI)-labeled thyrocytes and ethidium bromide-acridine orange staining and fluorescence microscopy analysis, the addition of mAbs that block Fas completely suppressed IL-1β-induced apoptotic cell death in vitro (Fig. 5, B and C). These findings demonstrated that IL-1β-mediated induction of Fas on thyrocytes, which constitutively express FasL, results in their apoptotic suicide or fratricide.

Our results show that both Fas and its ligand are present in HT glands and that their concomitant expression on thyrocytes is responsible for the induction of programmed cell death. Because FasL is constitutively expressed in normal thyroids, IL-1β-induced Fas expression may represent a critical limiting factor for the acceleration of thyrocyte destruction during the course of the inflammatory process. IL-1β, probably released by infiltrating monocytes or macrophages (or both) or by activated endothelial cells, can interact directly with thyrocytes, and thyrocyte destruction can proceed in a manner relatively independent of infiltrating T lymphocytes. Although autoreactive T lymphocytes may contribute to the HT thyroid-infiltrating cells (19), there is no evidence that cytotoxic T lymphocytes are directly involved in thyrocyte destruction. Attempts to localize cytotoxic T lymphocytes in situ revealed that only a few T lymphocytes, among those which infiltrate HT thyroids, contain perforin (20). Moreover, we found that the expression of FasL on infiltrating T lymphocytes is negligible compared to FasL expression on HT thyrocytes. These observations suggest a minor role for cytotoxic T lymphocytes and a prevailing involvement of FasL expressed by thyrocytes in tissue destruction. It is likely that this process is then further amplified by autoantibody-mediated cytotoxicity, as new antigenic determinants are exposed after initial tissue destruction.

In conclusion, we provide evidence suggesting that Fas-FasL interactions among thyrocytes may contribute to the pathogenesis of HT. Although infiltrating T lymphocytes may participate to this process, our results demonstrate that thyrocytes undergo Fas-mediated apoptosis after IL-1β exposure without the contribution of other cells. This mechanism might be operating in all forms of hypothyroidism that follow inflammation.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
View Abstract

Navigate This Article