Murine FLIP Transgene Expressed on Thyroid Epithelial Cells Promotes Resolution of Granulomatous Experimental Autoimmune Thyroiditis in DBA/1 Mice

Generation of cFLIP_L Transgenic Mice

The rat thyroglobulin (RTg) promoter on pTgCAT7 was provided by Dr. Motoyasu Saji (Ohio State University, Columbus, OH), and the 1.5-kb FLAG-tagged murine cFLIP_L on pBSKII (2.96 kb) was provided by Dr. Susanne Lens (Netherlands Cancer Institute, Amsterdam, The Netherlands). The 900-bp fragment of RTg promoter was obtained using HindIII and EcoRI (Invitrogen, Carlsbad, CA) and blunted. The pBSKII containing murine cFLIP_L was linearized by NotI (Invitrogen), blunted, and then ligated with the RTg promoter using T4 DNA ligase. The correct orientation of the promoter in the recombined plasmid was confirmed by sequence determination using T7 primer (DNA Core Facility, University of Missouri, Columbia, MO). The recombined construct (3.5 kb) containing 1.5-kb FLAG-tagged mouse cFLIP_L cDNA, β-globin polyA (500 bp), β-globin intron (600 bp), and RTg promoter (900 bp) was amplified in Escherichia coli and transfected into the thyroid PCCL3 cell line (provided by Dr. Daniel Altschuler, University of Pittsburgh, Pittsburgh, PA) using Fugene6 (Invitrogen), and FLIP protein expression was confirmed by immunohistochemical staining. The recombinant plasmid was digested using BssHII (Invitrogen), and the 3.5-kb fragment containing FLAG-tagged mouse cFLIP_L cDNA, β-globin polyA, β-globin intron, and RTg promoter was microinjected into fertilized oocytes from FVB female mice (Transgenic Core Facility, University of Missouri). Transgenic founders were screened by polymerase chain reaction (PCR) of tail DNA using the following primers specific to RTg promoter and cFLIP_L recombined construct: sense, 5′-CTGGTCATCATCCTGCCTTTCTC-3′; and antisense, 5′-TGCAGCCAGGTTCTCAGTCAC-3′. Dot blot of tail DNA was also used to confirm expression of the transgene in some mice. Two FVB transgenic founders (one male and one female) were obtained. Because FVB mice are less EAT susceptible than DBA/1 mice, even though both strains express the same H-2q MHC haplotype (our unpublished results), the transgenic female FVB founder was crossed with an EAT-susceptible DBA/1 male, and Tg⁺ F1 offspring were selected by PCR analysis of tail DNA. Tg⁺ F1 mice were backcrossed to DBA/1 mice. After backcrossing three times to DBA/1 mice and selecting for expression of the transgene and the DBA/1 coat color at each generation, Tg⁺ and Tg⁻ littermates used as recipients of activated DBA/1 splenocytes (see below) developed maximal G-EAT lesions 20 days after cell transfer that were indistinguishable histologically and in severity from those developing in WT DBA/1 recipients of the same cells. FLIP Tg⁺ and Tg⁻ mice from the third backcross generation also transferred severe G-EAT to DBA/1 recipients when immunized and used as donors for adoptive transfer of G-EAT (our unpublished results). Mice used for these experiments were from the third and fourth backcross generations, and identical results were obtained with both backcross groups. In all experiments, Tg⁺ and Tg⁻ littermates were used as recipients of sensitized donor cells from WT DBA/1 mice. Immunohistochemistry (IHC) using rabbit anti-FLIP polyclonal antibody (Ab) (Abcam, Cambridge, MA) and anti-FLAG polyclonal Ab (Abcam) and Western blot using anti-FLIP monoclonal antibody (mAb) (Alexis, San Diego, CA) and anti-FLAG were used to confirm expression of the transgene on TECs but not on other cell types.

Induction of G-EAT

WT DBA/1 mice generated in our breeding colony were used as donors for all experiments. DBA/1 donors were injected intravenously twice at 10-day intervals with 150 μg of MTg prepared as previously described2 and 15 μg of lipopolysaccharide LPS (Sigma, St. Louis, MO). Seven days later, donor spleen cells were activated in vitro with 25 μg/ml MTg and 5 ng/ml IL-12 (Peprotech, Rocky Hill, NJ) as previously described.5 Cells were harvested after 72 hours and washed twice, and 3 × 10⁷ cells were transferred intravenously to 500-Rad irradiated FLIP Tg⁺ or Tg⁻ littermate recipients. Recipient thyroids were evaluated 20 (peak of disease), 40, or 60 days after cell transfer.5,23 Both male and female mice were used for these experiments. Donor DBA/1 mice were generally 8 to 10 weeks old at the time of immunization, and recipients were 7 to 9 weeks old. All mice were bred and maintained in accordance with University of Missouri institutional guidelines for animal care.

Evaluation of G-EAT Histopathology and Fibrosis

Thyroids were removed from groups of six to eight FLIP Tg⁺ or Tg⁻ recipients 20, 40, or 60 days after cell transfer, and one lobe of each thyroid was fixed in formalin. Formalin-fixed paraffin-embedded tissues were sectioned and stained with hematoxylin and eosin. Thyroids were scored quantitatively for G-EAT severity using a scale of 1+ to 5+ according to previously established criteria.4,5 Thyroiditis of 1+ is defined as an infiltrate of at least 125 cells in one or several foci, 2+ is 10 to 20 foci of cellular infiltration involving up to 25% of the gland, 3+ indicates that 25 to 50% of the gland is infiltrated, 4+ indicates that >50% of the gland is destroyed, and 5+ indicates virtually complete destruction of the gland, with few or no remaining follicles. At day 20, thyroids with 5+ severity scores are greatly enlarged (10- to 20-fold), with granuloma formation, large numbers of histiocytes, lymphocytes, and neutrophils, microabscess formation, necrosis, and fibrosis.5 At days 40 to 60, fibrosis is extensive, and thyroids are very small and atrophic (approximately one-fifth to one-tenth the size of a normal thyroid). For evaluation of collagen deposition (fibrosis), some thyroid sections from FLIP Tg⁻ and Tg⁺ recipients were stained using Masson’s Trichrome.34 All slides were evaluated by at least two investigators, one of whom had no knowledge of the experimental treatments. Differences in interpretation were very rare.

Serum Thyroxine (T₄) Assay

Serum T₄ levels were determined using a T₄ enzyme immunoassay kit (Biotecx Labs, Houston, TX) according to the manufacturer’s instructions. Results are expressed as micrograms of T₄ per decaliter of serum. Using this assay, T₄ values for normal mouse serum ranged from 4 to 10 μg/dl; values <3 μg/dl are considered low.35

IHC

IHC staining for FLIP was described previously.23 The same protocol was used for FLAG staining. IHC staining for CD4 and CD8 was described previously.24,35 Briefly, cryostat sections were fixed in acetone for 10 minutes at 4°C. After treatment for 30 minutes with 0.1% saponin in 1% bovine serum albumin (FLIP and FLAG) or 1% bovine serum albumin alone (CD4 and CD8), slides were incubated with polyclonal anti-FLIP or anti-FLAG, rat anti-CD4 (GK1.5; American Type Culture Collection, Manassas, VA) and rat anti-CD8 mAb (53.6; American Type Culture Collection) for 60 minutes at room temperature. After incubation with a secondary biotinylated Ab (Jackson ImmunoResearch Laboratories Inc., West Grove, PA), immunoreactivity was demonstrated using the avidin-biotin complex immunoperoxidase system (Vector, Burlingame, CA) and developed using NovaRED (Vector) as the chromogen. Slides were counterstained with hematoxylin. As a negative control, primary Ab was replaced with an equal amount of normal rabbit, goat, or rat IgG; these controls were always negative. To quantify the number of CD4⁺ or CD8⁺ cells, all CD4⁺ or CD8⁺ cells in five to six randomly selected high-power fields (magnification, ×400) were manually counted, and the average number of cells per field for three individual animals per group were determined (Figure 3G).

Confocal Laser Scanning Double-Immunofluorescence Microscopy

To detect differential expression of FLIP, FasL, active caspase-8, and active caspase-3 by CD4⁺ and CD8⁺ T cells or TECs, dual-color immunofluorescence and confocal laser scanning microscopy was performed using the same protocol as previously described23,24,34,35,36 using rat anti-CD4 and -CD8 mAb and polyclonal anti-FLIP (described above), anti-FasL mAb (Santa Cruz Biotechnology, Santa Cruz, CA), anti-active caspase-8 (IMGENEX, San Diego, CA), anti-active caspase-3 (BD Pharmingen, San Diego, CA), and anti-CD3 (DakoCytomation, Carpinteria, CA) for T cells or pan-cytokeratin-26 (PCK; Sigma) for TECs. Frozen sections of thyroids were used except for dual staining of FasL and CD3 or PCK (formalin-fixed paraffin sections were used). FLIP, FasL, active caspase-8, and active caspase-3 were visualized with Alexa 568 (red; Molecular Probes, Eugene, OR). Cytokeratin (TECs), CD3, CD4, and CD8 molecules were visualized by Alexa 488 (green; Molecular Probes). Apoptosis was detected using an in situ cell death kit (red; Roche, Nutley, NJ) with formalin-fixed paraffin sections of thyroids as previously described.36 Slides were observed with a Bio-Rad Radiance 2000 confocal system (Bio-Rad, Hercules, CA) coupled to an Olympus IX70 inverted microscope (Olympus, Tokyo, Japan). To quantify the number of double positive cells, all double positive cells in four to five randomly selected high-power fields (magnification, ×600) were manually counted, and the average number of cells per field for three individual animals per group was determined (Figure 4F and Figure 5, K and L).

Reverse Transcriptase (RT)-PCR

Thyroid lobes were removed from individual mice at different times after adoptive transfer, and one lobe was stored at −80°C before processing. Frozen thyroids were homogenized in TRIzol (Invitrogen), and RNA was extracted and reverse transcribed as previously described in detail.37,38 Diluted cDNA was amplified using the following conditions: 94°C for 30 seconds for denaturing, 60°C for 30 seconds for annealing, and 72°C for 1 minute for extension. To determine the relative initial amounts of target cDNA, each cDNA sample was serially diluted one-fifth and one-twenty-fifth and amplified with specific primers.37,38 Hypoxanthine phosphoribosyltransferase was used as a housekeeping gene to verify that the same amount of RNA was amplified. PCR products were electrophoresed in 2% agarose gel, visualized by UV light after staining with ethidium bromide, and normalized between samples relative to levels of hypoxanthine phosphoribosyltransferase using an IS-1000 Digital Imaging System (Life Sciences, St. Louis, MO). Most gene primers used in this study have been described previously.34,37,38 IL-17 primers were: sense, 5′-GGTCAACCTCAAAGTCTTTAACTC-3′, and antisense, 5′-TTAAAAATGCAAGTAAGTTTGCTG-3′. Foxp3 primers were: sense, 5′-GGCCCTTCTCCAGGACAGA-3′, and antisense, 5′-GCTGATCATGGCTGGGTTGT-3′.

Western Blot

Thyroids were homogenized in lysis buffer (50 mmol/L HEPES, 250 mmol/L NaCl, 5 mmol/L ethylenediamine tetraacetic acid, and 0.1% Triton). After centrifugation at 15,000 rpm for 15 minutes at 4°C, supernatants were collected and electrophoresed by adding 30 μg of protein to a 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis gel. After transfer onto nitrocellulose membranes, they were incubated overnight at 4°C in Tris-buffered saline containing 5% nonfat dry milk and 0.05% Tween 20 and probed 2 hours at room temperature with rat anti-FLIP mAb or rabbit anti-FLAG polyclonal Ab (Sigma). After washing, membranes were incubated with a peroxidase-conjugated goat anti-rat or goat anti-rabbit IgG (1/5000) for 1 hour at room temperature, and the reaction was detected with a chemiluminescence detection kit (Pierce, Rockford, IL).24 For normalization of signals, the membranes were stripped by incubating at 60°C for 30 minutes in stripping buffer (62.5 mmol/L Tris-HCl, 100 mmol/L 2-ME, and 2% sodium dodecyl sulfate), washed in Tris-buffered saline/Tween 20, and reprobed with 1/2000 rabbit anti-actin primary Ab (Santa Cruz Biotechnology) and 1/2000 horseradish peroxidase-conjugated anti-rabbit IgG as secondary Ab. Bands were scanned and quantitated using Quality One-4.41 software. Each lane in the figures represents protein extracted from individual thyroids of normal FLIP transgenic mice or Tg⁻ littermate controls (Figure 1D) or G-EAT Tg⁻ and G-EAT Tg⁺ recipients (Figure 6B).

Dot Blot Analysis

Dot blot analysis was used to assess the level of FLIP transgene in FLIP Tg⁺ mice. After denaturation, 20 μg of tail DNA was immobilized onto nitrocellulose membranes. FLAG-tagged murine FLIP Klenow (1.5 kb) was used as the template, [α-³²P]dCTP-labeled DNA probes were generated using strip-EZ probe synthesis and removal kits (Ambion, Austin, TX). Prehybridization was performed for 2 to 4 hours, hybridization was performed overnight at 42°C, and washes were done according to instructions. Blots were exposed either 8 hours or 2 days at −80°C to blue X-ray films. For normalization of signals, the membranes were stripped by incubating at 90°C for 30 minutes in stripping buffer, washed, and reprobed with a β-actin probe.

Statistical Analysis

All experiments were repeated at least three times. Statistical analysis of data was performed using an unpaired two-tailed Student’s t-test or/and the Mann-Whitney rank sum test. A value of P < 0.05 was considered significant.

Generation of Transgenic Mice Expressing Murine cFLIP_L on TECs

To test the hypothesis that overexpression of FLIP by TECs would promote earlier resolution of G-EAT, we generated transgenic mice expressing mouse cFLIP_L under control of the RTg promoter (Figure 1A). The FVB transgenic founder female was crossed with an MHC-compatible, EAT-susceptible DBA/1 male, and the Tg⁺ offspring were backcrossed to DBA/1 as described in Materials and Methods. Both IHC (Figure 1, B and C) and Western blot (Figure 1D) showed that neither FLIP nor FLAG-tagged protein was detected in thyroids of naïve Tg⁻ mice. However, FLIP and FLAG were both highly expressed by TECs in the founder and all Tg⁺ mice at each backcross generation (Figure 1, B–D; data not shown) but not in other tissues such as spleen and cervical lymph nodes (data not shown). Some mice were typed for FLIP transgene expression by both PCR analysis of tail DNA and dot blot analysis, and results using both assays were consistent. Dot blot indicated that the FLIP transgene level in normal FLIP Tg⁺ mice was nearly six times higher than in Tg⁻ littermates (Figure 1E).

Distinct Outcomes of G-EAT in FLIP Tg⁺ and FLIP Tg⁻ DBA/1 Recipients

G-EAT was induced by adoptive transfer of spleen cells from MTg-sensitized DBA/1 donors activated with MTg and IL-12 in vitro. G-EAT lesions in both Tg⁺ and Tg⁻ recipients of DBA/1 spleen cells reached maximal severity 20 days after cell transfer, and disease severity scores were similar (4 to 5+) in both FLIP Tg⁺ and Tg⁻ recipients (Figure 2). At day 40, severity scores of 4 to 5+ were maintained in both groups (data not shown), although thyroids of most Tg⁺ recipients were larger (less atrophic) and had less fibrosis (Table 1; data not shown). By day 60, thyroid lesions in all Tg⁻ recipients were fibrotic, with an average severity score of 4.5, whereas thyroid lesions in all but four of the Tg⁺ recipients had less fibrosis and had begun to resolve with an average severity score of 3.3 (Figure 2). The difference between the two groups at day 60 was highly significant (P < 0.001, using Student’s t-test and the Mann-Whitney rank sum test). Collagen deposition in the thyroid, an indicator of fibrosis, was assessed by Masson’s Trichrome staining. As shown in Figure 3, A and B, there was extensive deposition of collagen (blue) at day 60 in thyroids of FLIP Tg⁻ recipients (left panels), whereas collagen deposition was considerably reduced in thyroids of FLIP Tg⁺ recipients (right panels) at day 60. The decreased fibrosis in thyroids of Tg⁺ versus Tg⁻ recipients was also evident at days 20 and 40, as summarized in Table 1. Despite almost complete destruction of thyroids at day 20, most Tg⁺ and Tg⁻ recipients had normal serum T₄ at day 20 (Table 1). Serum T₄ was low in almost all mice with 4 to 5+ severity scores at days 40 and 60 but returned to normal in Tg⁺ recipients with 1 to 3+ severity scores at day 60 (Table 1; data not shown). Because high TSH levels could promote thyroid regeneration in mice with low serum T₄, it is important to emphasize that both Tg⁺ and Tg⁻ mice developed severe G-EAT at day 20 and had low serum T₄ at day 40. T₄ levels returned to normal only in mice with substantial resolution of their thyroid lesions (Table 1; data not shown). Therefore, thyroid regeneration in this model actually occurs more quickly when T₄ levels have returned to normal (Table 1) or when mice develop less severe G-EAT and maintain normal serum T₄ throughout the course of disease.6,7,39 These results indicate that the outcome of G-EAT differed in FLIP Tg⁻ and Tg⁺ littermate recipients given the same pool of WT donor cells; ie, thyroid lesions resolved earlier, and there was less fibrosis in FLIP Tg⁺ recipients, whereas thyroids of FLIP Tg⁻ recipients had ongoing inflammation and extensive fibrosis.

Ratio of CD4⁺ versus CD8⁺ T Cells Differs in G-EAT Thyroids That Will Resolve or Progress to Fibrosis

Our previous studies showed that thyroids of mice with 3 to 4+ G-EAT severity scores that would resolve had more CD8⁺ T cells than CD4⁺ T cells at day 20, whereas CD4⁺ T cells outnumbered CD8⁺ T cells at day 20 in thyroids with 5+ severity scores that would progress to fibrosis.39 IHC was used to determine whether the ratio of CD4⁺ versus CD8⁺ T cells differed in thyroids of FLIP Tg⁺ and Tg⁻ recipients. Consistent with our previous study,39 CD4⁺ T cells outnumbered CD8⁺ T cells at day 20 in FLIP Tg⁻ recipients, which had ongoing inflammation and fibrosis at day 60, whereas CD8⁺ T cells outnumbered CD4⁺ T cells at day 20 in FLIP Tg⁺ recipients whose lesions would resolve by day 60 (data not shown). These differences were even more striking at day 60 (Figure 3). CD4⁺ T cells (Figure 3, C and D) outnumbered CD8⁺ T cells (Figure 3, E and F) in FLIP Tg⁻ recipients (Figure 3, C–F, left panels) with ongoing inflammation and fibrosis at day 60 (Figure 3G, left panel), whereas there were fewer T cells in thyroids of FLIP Tg⁺ recipients, and most were CD8⁺ (Figure 3, C–F, right panels) at day 60 (Figure 3G, right panel). Numbers of CD4⁺ and CD8⁺ T cells were assessed semiquantitatively by counting five to six random high-power fields on slides of three mice in each group with comparable 4 to 5+ G-EAT severity scores (Figure 3G). The more completely resolved thyroids (1 to 3+) had many fewer inflammatory cells at day 60, but most were CD8⁺ T cells (data not shown). These results are consistent with our previous studies indicating that the ratio of CD4⁺ versus CD8⁺ T cells in G-EAT thyroids correlates with and can be a good predictor of the outcome of the autoimmune inflammatory response.39

Outcome of G-EAT in FLIP Tg⁺ and Tg⁻ Recipients Correlates with the Site of Expression of FLIP and FasL

Our previous studies suggested that up-regulation of FLIP by inflammatory cells might contribute to chronic inflammation by blocking apoptosis of FasL-positive inflammatory cells,24 whereas expression of FLIP by FasL-positive TECs might protect TECs from apoptosis, contributing to resolution.24 Confocal microscopy was used to determine whether the site of expression of FasL (paraffin sections) and FLIP (frozen sections) by TECs versus inflammatory cells differed in thyroids of FLIP Tg⁺ and Tg⁻ recipients. Our unpublished data showed that FasL staining did not work well on frozen sections, and FLIP staining worked well both on frozen and paraffin sections. Because CD4 and CD8 cannot be stained on paraffin sections, CD4⁺ and CD8⁺ T cells, the predominant inflammatory cells in G-EAT thyroids, were identified by cell surface expression of CD3 on paraffin sections (Figure 4A, green), and TECs were identified by PCK staining (Figure 4, B and C, green). Although TECs and inflammatory cells both expressed FasL (Figure 4, A and B, red) in FLIP Tg⁻ (left panels) and Tg⁺ recipients (right panels), FasL was mainly expressed by CD3⁺ inflammatory cells in FLIP Tg-recipients (Figure 4A, left panel, yellow, overlay) and by TECs in FLIP Tg⁺ recipients (Figure 4B, right panel, yellow, overlay). Although TECs and inflammatory cells both expressed FLIP (Figure 4C, red) in FLIP Tg⁻ (left panel) and Tg⁺ recipients (right panel), FLIP was mainly expressed by PCK + TECs in FLIP Tg⁺ recipients (Figure 4C, right panel, yellow, overlay). Confocal staining of CD4 (Figure 4D, green) or CD8 (Figure 4E, green) on frozen sections with FLIP (Figure 4, D and E, red) showed that, consistent with IHC (Figure 3), CD4⁺ T cells were predominant in thyroids of FLIP Tg⁻ recipients (Figure 4, D and E, left panels), whereas CD8⁺ T cells were predominant in thyroids of FLIP Tg⁺ recipients (Figure 4, D and E, right panels) at day 60. FLIP was mainly expressed by CD4⁺ T cells in FLIP Tg⁻ recipients (Figure 4D, left panel, yellow, overlay), whereas very few CD4⁺ T cells in Tg⁺ recipients expressed FLIP (Figure 4D, right panel, yellow, overlay). Double positive cells (yellow, overlay) in four to five randomly selected high-power fields of three individual mice per group (magnification, ×600) were manually counted, and the results are summarized in Figure 4F (bars a–e correspond to Figure 4, A–E). Results are representative areas on slides from at least three individual mice examined per group.

Evaluation of Active Caspase-8, Active Capase-3, and Apoptosis in Thyroids of FLIP Tg⁺ and Tg⁻ Recipients

Fas/FasL-mediated apoptosis involves activation of caspase-8 and its downstream effector, caspase-3, and FLIP inhibits Fas/FasL-mediated apoptosis through its ability to block activation of caspase-8.12,15,16,22,25 Confocal microscopy was used to determine whether the site of expression of active caspase-8 and active caspase-3 by TECs versus inflammatory cells differed in thyroids of FLIP Tg⁺ and Tg⁻ recipients. CD4⁺ and CD8⁺ T cells were identified by cell surface expression of CD4 and CD8 (Figure 5, B, C, E and F, green), and TECs were identified by PCK (Figure 5, A and D, green). Although at day 20, TECs and inflammatory cells both expressed active caspase-8 (Figure 5, A–C, red) in FLIP Tg⁻ (left panels) and Tg⁺ recipients (right panels), active caspase-8 was mainly expressed by TECs in FLIP Tg⁻ recipients (Figure 5A, left panel, yellow, overlay) and by CD4⁺ T cells in FLIP Tg⁺ recipients (Figure 5B, right panel, yellow, overlay). Active caspase-3 was also expressed primarily by TECs in FLIP Tg⁻ recipients (Figure 5D, left panel, yellow, overlay) and by CD4⁺ T cells in FLIP Tg⁺ recipients at day 20 (Figure 5E, right panel, yellow, overlay). Very few active caspase-8 or caspase-3⁺ cells were detected at day 60 (data not shown).

We previously showed that resolution of G-EAT was at least in part due to increased apoptosis of CD4⁺ effector T cells by a FasL-dependent mechanism.23,24,39 A fluorescence apoptosis kit was used to determine whether there was increased apoptosis of T cells in thyroids of FLIP Tg⁺ versus Tg⁻ recipients and to determine the distribution of apoptotic cells in these thyroids (Figure 5, G–J). Because detection of apoptotic cells was reduced on frozen sections (our unpublished results), it was necessary to use paraffin sections for these experiments. TECs were identified by PCK (green), T cells were identified by CD3 (green), and apoptosis was detected as red nuclear staining. Many apoptotic cells were detected in thyroids of both FLIP Tg⁺ and Tg⁻ recipients with G-EAT at day 20 (Figure 5, G and H), whereas only a few apoptotic cells were detected at day 60 (Figure 5, I and J). Apoptosis (red) was mainly in inflammatory cells in FLIP Tg⁺ recipients both at days 20 (Figure 5H, right panel) and 60 (Figure 5J, right panel) and mainly in TECs in FLIP Tg⁻ recipients at days 20 (Figure 5G, left panel) and 60 (Figure 5I, left panel). Double positive cells (yellow, overlay in Figure 5, A–F; or red circled by green in Figure 5, G–J) in four to five randomly selected high-power fields of three individual mice per group (magnification, ×800) were manually counted and summarized in Figure 5, K and L (bars a–j correspond to Figure 5, A–J). Shown are representative areas on slides of thyroids of at least three individual mice per group. These results are consistent with cell distribution patterns of active caspase-8 and active caspase-3 and support the idea that transgenic FLIP on TECs may function, at least in part, to promote earlier resolution of G-EAT by protecting TECs from apoptosis and inducing increased apoptosis of effector CD4⁺ T cells through the caspase cascade pathway.

FLIP Protein and mRNA Expression in Thyroids of FLIP Tg⁺ and Tg⁻ Mice at Day 20

To compare the levels of FLIP in thyroids of FLIP Tg⁺ and Tg⁻ recipients with G-EAT at day 20, FLIP mRNA and protein expression levels were determined by semiquantitative RT-PCR and Western blot (Figure 6). FLIP mRNA was undetectable in normal Tg⁻ thyroids but was significantly up-regulated in thyroids of FLIP Tg⁻ recipients with severe G-EAT. FLIP mRNA expression in thyroids of normal FLIP Tg⁺ mice was similar to that in thyroids of FLIP Tg⁻ recipients with severe G-EAT and was further up-regulated in thyroids of FLIP Tg⁺ recipients with G-EAT at day 20 (Figure 6A). Consistent with the FLIP mRNA expression level, FLIP protein was undetectable in normal Tg⁻ thyroids (Figure 1D) but was significantly up-regulated in thyroids of FLIP Tg⁻ recipients with severe G-EAT and was further increased in thyroids of FLIP Tg⁺ recipients with severe G-EAT (Figure 6B).

Cytokine and Foxp3 mRNA Expression in Thyroids of FLIP Tg⁺ and Tg⁻ Mice

The balance between proinflammatory cytokines such as interferon-γ (IFN-γ), tumor necrosis factor-α (TNF-α), and IL-17 and anti-inflammatory cytokines such as IL-10, IL-13, and transforming growth factor-β (TGF-β) is important for development and progression of inflammation in autoimmune diseases including G-EAT.25,34,39,40,41,42,43,44,45,46,47,48 To determine whether proinflammatory and anti-inflammatory cytokines were differentially expressed in thyroids of FLIP Tg⁺ and Tg⁻ recipients, mRNA expression of particular cytokines in individual thyroids was determined by RT-PCR 20 (data not shown) and 40 (Figure 7) days after cell transfer. Because regulatory T cells can reduce inflammation in many autoimmune diseases including EAT,49,50,51,52 expression of Foxp3 as a marker of regulatory T cells49 was also investigated. Expression of cytokine and Foxp3 mRNA was undetectable in thyroids of normal FLIP Tg⁺ or Tg⁻ mice (Figure 7, A–G). At day 20, IL-17 mRNA was lower in thyroids of FLIP Tg⁺ recipients compared with Tg⁻ recipients, whereas expression of other cytokines and Foxp3 mRNA was similar in thyroids of both FLIP Tg⁺ and Tg⁻ recipients with G-EAT (data not shown). By day 40, expression of proinflammatory cytokines such as IFN-γ, TNF-α, and IL-17 was lower (Figure 7, A–C), whereas expression of anti-inflammatory cytokines such as IL-10 and IL-13 was higher (Figure 7, D and E) in thyroids of FLIP Tg⁺ compared with Tg⁻ recipients, even though G-EAT severity scores were similar (4 to 5+) for both groups. Expression of TGF-β and Foxp3 was similar in thyroids of FLIP Tg⁺ and Tg⁻ recipients with G-EAT (Figure 7, F and G). IL-4 expression could not be assessed because of very low expression levels at days 20 and 40. These results suggested that expression of proinflammatory and anti-inflammatory cytokines in recipient thyroids correlated with differences in the outcome of G-EAT seen at day 60 (Figure 2) in FLIP Tg⁺ versus Tg⁻ recipients, consistent with the increased apoptosis of inflammatory cells in thyroids of FLIP Tg⁺ recipients compared with Tg⁻ recipients (Figure 5).