THERE ARE NO spontaneous animal models of Graves

Transcription

1 /01/$03.00/0 Endocrinology 142(8): Printed in U.S.A. Copyright 2001 by The Endocrine Society Naked TSH Receptor DNA Vaccination: A TH1 T Cell Response in Which Interferon- Production, Rather than Antibody, Dominates the Immune Response in Mice PAVEL PICHURIN, XIN-MIN YAN, LOREDANA FARILLA, JIN GUO, GREGORIO D. CHAZENBALK, BASIL RAPOPORT, AND SANDRA M. MCLACHLAN Autoimmune Disease Unit, Cedars-Sinai Research Institute and UCLA School of Medicine, Los Angeles, California THERE ARE NO spontaneous animal models of Graves disease. Further, conventional immunization with soluble TSH receptor (TSHR) together with adjuvant does not lead to hyperthyroidism, although thyroiditis develops in some mouse strains (reviewed in Ref. 1). Novel immunization protocols have been developed that elicit antibodies in mice resembling those in Graves patients, including the Shimojo approach of injecting live fibroblasts coexpressing the TSHR and major histocompatibility complex (MHC) antigen class II (2, 3), naked TSHR-DNA vaccination (4) and, most recently, injecting combinations of TSHR-transfected B cells, soluble TSHR and adjuvant (5). In the Shimojo approach, about 25% of AKR/N mice develop thyrotoxicosis (2, 6, 7). TSHR-DNA vaccination induced hyperthyroidism in some outbred mice (8) but not in an inbred strain (BALB/c) (4). Genetic differences among mouse strains can influence the outcome of immunization, as illustrated for DNA vaccination (above) and as previously reported for injection of TSHR, MHC class II fibroblasts (3). In addition, different immunization protocols within a particular mouse strain may affect the characteristics of the immune response. Both the Shimojo approach (2, 6, 7) and naked TSHR-DNA vaccination (4, 8) induce TSHR antibodies with TSH binding inhibition (TBI) activity. On the other hand, examination of these different studies suggests that thyroid-stimulating antibodies (TSAb) arise more frequently in the Shimojo model than after DNA vaccination. Abbreviations: CHO, Chinese hamster ovary; IFN, interferon; MHC, major histocompatibility complex antigen; TBI, TSH binding inhibition; TSAb, thyroid-stimulating antibody; TSHR, TSH receptor. Two approaches have been developed to induce TSH receptor antibodies in mice with properties resembling those in Graves disease, the Shimojo model of injecting live fibroblasts coexpressing the TSH receptor and major histocompatibility complex antigen Class II, and TSH receptor-dna vaccination. Thyroid-stimulating antibodies appear to occur less commonly after DNA vaccination, but there has been no direct comparison of these models. We performed a three-way comparison of 1) AKR/N and 2) BALB/c mice vaccinated with TSH receptor-dna and 3) AKR/N mice injected with fibroblasts expressing the TSH receptor and the major histocompatibility complex antigen class II of AKR/N mice. TSH receptor-dna vaccinated mice had low or undetectable levels of TSH receptor antibodies determined by ELISA or flow cytometry. Nonspecific binding precluded comparisons with sera from Shimojo mice by these assays. TSH binding inhibition and thyroid-stimulating antibody were undetectable in TSH receptor-dna vaccinated mice. In Shimojo mice, TSH binding inhibition was positive in approximately 60%, and thyroidstimulating antibodies were positive in hyperthyroid animals. Unlike the negative antibody data, splenocytes from TSH receptor-vaccinated (but not Shimojo) mice proliferated and produced the Th1 cytokine interferon- in response to TSH receptor antigen. In conclusion, DNA vaccination is less effective at inducing TSH receptor antibodies than the Shimojo approach, but it permits the future characterization of TSH receptor-specific T cells generated without adjuvant. (Endocrinology 142: , 2001) However, there has been no direct comparison of TSHR antibodies induced by these two protocols. Such information is important for future studies using mouse models to understand the immune response to the TSHR. In the present study, we directly compared the responses of inbred mice to vaccination with TSHR-DNA vs. injection with TSHR, MHC class II fibroblasts (Shimojo approach). Because the RT4.15HP fibroblasts (9) used in the Shimojo model express the MHC class II molecule I-A k, this model is limited to mouse strains (such as AKR/N) with the same I-A allele. On the other hand, data on TSHR-DNA vaccination of an inbred strain are only available for BALB/c mice (4), which differ from AKR/N mice in terms of MHC class II (I-A d ) and background genes. We, therefore, performed a three-way comparison of immune responses in 1) AKR/N and 2) BALB/c mice vaccinated with TSHR-DNA and 3) AKR/N mice injected with TSHR and MHC class II expressing fibroblasts. We now demonstrate that the TSHR antibody response (measured by TBI and TSAb assays) is much lower after TSHR- DNA vaccination than after injection of TSHR, MHC class II fibroblasts. Moreover, rather than antibody production, the characteristic immune outcome of TSHR-DNA vaccination is an antigen-specific lymphocyte response characterized by production of the Th1 cytokine interferon- (IFN- ). Materials and Methods Immunization of mice by DNA vaccination or by fibroblast injection The cdna for the human TSHR (psv2-neo-ece-tshr 5,3 ) (10, 11) was transferred to the vector pcdna3 (Invitrogen, Carlsbad, CA). Fe- 3530

2 Pichurin et al. IFN- in TSHR-DNA-Vaccinated Mice Endocrinology, August 2001, 142(8): male BALB/c (The Jackson Laboratory, Bar Harbor, ME) and AKR/N mice (National Cancer Institute, Bethesda, MD) aged 6 7 wk were pretreated in the anterior tibialis muscle with cardiotoxin (Naja nigricollis, 10 m, Calbiochem, La Jolla, CA). Five to 7 d later, the mice were injected in the same muscle with 100 g TSHR-DNA or empty vector DNA. The vaccination protocol was repeated 3 and 6 wk later (total 3 vaccinations) and, in some mice, after 9 10 wk (4 vaccinations). One to 2 wk after three DNA vaccinations, or 7 wk after 4 vaccinations, mice were euthanized to obtain blood (from the vena cava), thyroids, and spleens. The thyroid glands were fixed in 4% paraformaldehyde (Sigma, St. Louis, MO; ph 7.5) and paraffin-sections stained with hematoxylin and eosin. RT4.15HP fibroblasts (9), provided by Dr. Ron Germain, NIH, were stably transfected with the human TSHR cdna in pcdna3, selected with G418 and TSHR expression was determined by 125 I-TSH binding (7, 12). Fibroblasts (TSHR-expressing and untransfected controls) were expanded in DMEM with high glucose, 10% FBS, antibiotics (all from Life Technologies, Inc., Gaithersburg, MD) and HAT medium (Sigma) to preserve MHC class II expression. After pretreatment with mitomycin C (50 g/ml; Sigma), fibroblasts were injected (10 7 cells/mouse) ip on 6 occasions at two weekly intervals into 6- to 7-wk-old female or male AKR/N mice. Mice were euthanized 10 days after the sixth fibroblast injection to obtain blood (from the vena cava), thyroids, and spleens. All animal studies were approved by the Institutional Animal Care and Use Committee and performed in accordance with the highest standards of humane care in a pathogen-free facility. Purified TSHR antigen TSHR-289 is a TSHR ectodomain variant corresponding to the A subunit (13). This protein, expressed in Chinese hamster ovary (CHO) cells, contains two conformationally different forms (active and inactive) with respect to their ability to be recognized by human TSHR autoantibodies (14). Inactive TSHR-289 was isolated from culture medium by affinity chromatography using mouse mab 3BD10. Active TSHR-289 not captured by the column was then recovered using an affinity column with a mab to histidine residues in the C-terminus (15). Purity and concentrations were determined by SDS-PAGE. Before use as immunogen (see below), in ELISA or in lymphocyte cultures, both forms of TSHR-289 were dialyzed against 10 mm Tris, ph 7.4, 50 mm NaCl. To provide positive controls for serum antibody binding in ELISA and flow cytometry, BALB/c mice (The Jackson Laboratory) were immunized with purified TSHR-289 (50 g) emulsified in Complete Freund s Adjuvant (Calbiochem). For this purpose we used inactive TSHR-289 because the active form is quite labile and will be converted into the inactive form in the immunization procedure. Mice were boosted with the same antigen (50 g) in Incomplete Freund s Adjuvant (Calbiochem) after 3 wk and again after 4 wk before tail bleeding. Serum binding to purified TSHR antigen ELISA wells coated with TSHR-289 (inactive form; 1 g/ml in 10 mm Tris, ph 7.4, 50 mm NaCl) were incubated with test sera (diluted 1:100 for DNA vaccinated mice and 1:10,000 for purified TSHR adjuvant immunized mice). Antibody binding was detected with horse radish peroxidase conjugated mouse anti-igg (Sigma), the signal developed with O-phenylenediamine and H 2 O 2 and OD read at 490 nm. Flow cytometry for serum binding to TSHR-expressing cells Mouse sera (diluted 1:50) were examined for antibody binding to CHO cells expressing high numbers of TSH receptors ( TSHR/ cell) (16) and detection with fluorescein isothiocyanate (FITC)-conjugated, affinity-purified goat antimouse IgG (Caltag Laboratories, Inc., Burlingame, CA) as previously described (17). All assays included cells incubated with second antibody alone and with normal mouse serum. Flow cytometry was performed (10,000 events) using a FACScan with Cellquest Software (Becton Dickinson and Co., San Jose, CA). Inhibition of 125 I-TSH binding to its receptor (TBI) TBI (18) was measured using a kit (Kronus, Boise, ID). Duplicate serum aliquots (50 l; undiluted) were incubated with detergent solubilized porcine TSHR; 125 I-TSH (bovine) was added and the TSHRantibody complexes were precipitated with polyethylene glycol. The TBI values are calculated from the formula: TSH binding in the presence of test serum nonspecific binding TSH binding in the presence of normal serum nonspecific binding TSAb assay TSAb activity was assayed following the approach of Costagliola et al. (8). CHO cells expressing approximately 150,000 TSHR/cell (11) were grown to confluence in 96-well plates and incubated (4 h at 37 C) with test sera diluted 1/30 in hypotonic buffer (19) containing 10 mm HEPES, ph 7.4, 1 mm isobutylmethylxanthine and 0.3% BSA. Supernatants (500 l diluted 1:400) were acetylated (20 l triethylamine and 10 l acetic anhydride) and camp levels were measured by RIA using 2 -O-succinyl-[ 125 I] iodotyrosine methyl ester (NEN Life Science Products, Boston, MA) and a rabbit anti-camp antibody (Fitzgerald, Concord, MA). Results are expressed as % basal camp released in the presence of normal mouse serum. Proliferation in response to purified TSHR antigen Splenocytes (quadruplicate aliquots; cells; 200 l) were incubated in 96-well round bottomed plates in the presence or absence of TSHR 289 (see above, 4 or 20 g/ml). Culture medium was RPMI 1640, 10% heat inactivated FBS, 2 mm glutamine, 1 mm sodium pyruvate, 50 g/ml gentamycin, 50 m -mercaptoethanol, and 100 U/ml penicillin. After 6 days (37 C, 5% CO 2 ), 150 l medium was removed from each well to determine cytokine production (see below). Each well was then supplemented with the same volume of fresh medium with or without murine IL-2 (20 ng/ml, R&D,Minneapolis, MN). After approximately 8 h incubation, 1 Ci (3) H-thymidine ( 3 H-TdR; NEN Life Science Products) was added to each well and cultures were harvested approximately 18 h later for scintillation counting using a Tomtec Harvester 96 (Orange, CO). This approach for using the same set of cultures to determine both proliferation and cytokine production is based on the protocol of Hafler and colleagues (20). The data for lymphocyte proliferation are reported as 3 H-TdR incorporated (mean cpm sem for quadruplicate wells). Cytokine production in response to TSHR antigen Supernatants from quadruplicate aliquots were pooled, centrifuged to remove cell debris and stored ( 80 C). Duplicate aliquots (100 l) were assayed for IFN- and IL-4 by ELISA using capture and biotinylated detection-antibodies from BD PharMingen (San Diego, CA) and following the manufacturer s protocol. Cytokine production was expressed as pg/ml using standard curves of recombinant murine INF- and IL-4 (PharMingen). Results TSHR antibodies in DNA vaccinated mice TSHR antibodies generated in DNA vaccinated BALB/c and AKR/N mice were characterized in four different ways, namely binding to TSHR-coated ELISA plates, flow cytometry with mammalian cells expressing TSHRs, as well as two functional assays, inhibition of radiolabeled TSH binding to its receptor (TBI) and the ability to stimulate camp production (TSAb). The high, nonspecific serum binding by sera from mice injected with RT4.15HP fibroblasts (17) precludes the use of ELISA or flow cytometry to compare TSHR antibodies in fibroblast-injected AKR/N mice and in TSHR- DNA vaccinated mice. Consequently, as positive controls for these two direct binding assays (ELISA and flow cytometry), we used sera from BALB/c mice immunized with purified TSHR antigen and adjuvant. It should be emphasized that conventional immunization generates high antibody titers

3 3532 Endocrinology, August 2001, 142(8): Pichurin et al. IFN- in TSHR-DNA-Vaccinated Mice but the induced antibodies do not resemble those arising spontaneously in human thyroid autoimmunity disease (reviewed in Refs. 1, 12) Serum IgG antibody binding was studied using ELISA wells coated with a truncated TSHR ectodomain variant, TSHR-289, secreted by mammalian cells (13). We used the inactive form of TSHR-289 (15) because coating on ELISA wells converts the active to the inactive form (data not shown). Sera from only a few BALB/c and AKR/N mice vaccinated three or four times with TSHR-DNA gave ELISA OD values 2 sd greater than the mean for control DNA vaccinated mice (Fig. 1). As expected, the positive signals obtained for these sera (1:100 dilution) were very low compared with the high values obtained for sera from BALB/c mice conventionally immunized with TSHR-289 protein and adjuvant (1:10,000 dilution) (Fig. 1). TSHR-289 is a secreted TSHR A subunit and lacks amino acids present in the B subunit region of the ectodomain to which antibodies could arise following TSHR-DNA vaccination. Therefore, as a second approach, we used flow cytometry to examine antibody binding to CHO cells expressing high numbers of full length TSHR ( / cell) (16). However, only a small proportion of TSHRvaccinated BALB/c and AKR/N mice had serum IgG antibody that exceeded the median fluorescence values for sera from vector-treated mice (Fig. 2). There was no correlation between serum antibody positivity by flow cytometry and ELISA (not shown). As with TSHR-289 coated ELISA wells, conventional immunization with FIG. 2. DNA vaccination generates low levels of TSHR antibodies detected by flow cytometry. BALB/c and AKR/N mice were vaccinated three or four times with TSHR DNA. TSHR antibodies were detected by flow cytometry using CHO cells stably expressing the TSHR. Binding is reported as median immunofluorescence units for sera (1:50 dilution) from individual mice. As mentioned above for ELISA, sera from AKR/N mice injected with TSHR, MHC class II fibroblasts cannot be studied by flow cytometry because of the high, nonspecific binding induced by injection of fibroblasts TSHR (17). Included as positive controls are the values for BALB/c mice conventionally immunized with TSHR-289 and adjuvant. The numbers of mice in each group are in parentheses. TSHR protein and adjuvant was more effective than DNA vaccination in generating TSHR antibodies that recognize the full-length TSHR (Fig. 2). Although the levels of TSHR antibodies induced by the Shimojo approach could not be studied by ELISA or flow cytometry (see above), they could be compared with those arising after DNA vaccination using two functional (and more clinically relevant) assays, TBI and TSAb. None of the sera from TSHR-vaccinated mice were TBI positive (Fig. 3). In contrast, as previously reported, sera from approximately 60% AKR/N mice injected with TSHR-transfected fibroblasts were TBI positive compared with untransfected fibroblastinjected controls. Finally, TSAb was undetectable in sera of BALB/c or AKR/N mice vaccinated with TSHR-DNA (Fig. 4). In contrast, sera from all three AKR/N recipients of TSHR-fibroblasts had TSAb activity, consistent with their elevated T4 levels (of 8.2, 9.6, and 12.6 g/dl), as previously reported (17). FIG. 1. DNA vaccination generates low levels of TSHR antibodies detected by ELISA. BALB/c and AKR/N mice were vaccinated three or four times with TSHR-DNA, and sera were tested by ELISA using, as antigen, TSHR-289 (an ectodomain variant equivalent to the TSHR A subunit). Data are shown as OD values for sera (1:100 dilution) from individual mice after three or four vaccinations. Sera from AKR/N mice injected with TSHR, MHC class II fibroblasts cannot be studied in this assay because of the high, nonspecific binding induced by injection of fibroblasts, regardless of expression of the TSHR (17). Alternative positive controls are sera from BALB/c mice conventionally immunized with TSHR-289 and adjuvant (1:10,000 dilution). The vertical broken line indicates the mean 2 SD for mice vaccinated with the empty vector (Vector). The numbers of mice in each group are in parentheses. Lymphocytic infiltration of the thyroid gland Thyroid histology was examined in BALB/c and in AKR/N mice after three or four vaccinations with TSHR- or control-dna. No thyroid lymphocytic infiltration was observed regardless of whether the mice did or did not have TSHR antibodies as detected by ELISA (TSHR-289) or by flow cytometry (TSHR-CHO cells). Similarly, and consistent with other data (2, 6), lymphocytes were not detected in thyroid glands from TSHR-fibroblast injected mice, whether or not the mice were euthyroid or had elevated thyroxine levels (17).

4 Pichurin et al. IFN- in TSHR-DNA-Vaccinated Mice Endocrinology, August 2001, 142(8): FIG. 3. TSHR antibodies measured by TBI. BALB/c and AKR/N mice were vaccinated three or four times with TSHR DNA. In addition, AKR/N mice were injected with fibroblasts coexpressing the TSHR and MHC class II (Shimojo model). Negative controls were mice injected with the empty vector (Vector), or with fibroblasts expressing MHC class II but not the TSHR. Data are presented as the % inhibition for individual mice and the number of mice in each group is included in parentheses. TBI values for the Shimojo model include unpublished observations as well as data from previous studies (7, 17). Initially, we used the active form of TSHR-289. Cell proliferation was measured by 3 H-thymidine incorporation after 6 7 d incubation. One day before the end of these incubations medium was taken for cytokine analysis (see below) and was replaced with medium containing IL-2. We focused on BALB/c mice because previous investigations of TSHR- DNA vaccination were performed with this inbred strain (4). Variable splenocyte thymidine incorporation from individual mice was observed in the absence of TSHR-289 (Fig. 5, clear bars in left and right panels). TSHR-289 (4 and/or 20 g/ml) induced a significant proliferative response in splenocytes from all 5 TSHR-vaccinated BALB/c mice (Fig. 5, left panel; P 0.05). Splenocyte proliferation was unrelated to serum TSHR antibody levels. In contrast, splenocytes from only one of 4 vector-vaccinated mice showed a significant response to TSHR-289 (Fig. 5, right panel). Inactive TSHR-289 was as effective as the active form at stimulating splenocyte proliferation, regardless of whether IL-2 was, or was not, added to the culture (Fig. 6A). Finally, we studied the proliferative response to TSHR-289 of splenocytes from AKR/N mice previously immunized with TSHR-expressing fibroblasts (the Shimojo model). Basal proliferation in the absence of TSHR-289 was extremely high, particularly when the cultures were supplemented with IL-2 one day before harvesting (Fig. 6B; note the ordinate values compared with Fig. 6A). In the absence of IL-2, there was no response to either active or inactive TSHR-289 (Fig. 6B). However, proliferation was decreased in IL-2 supplemented cultures containing TSHR-289 (active or inactive). The reason for this apparent suppression is unclear. It was unrelated to cytokines (see below) and may simply reflect erratically high proliferation levels seen for splenocytes from Shimojo mice. FIG. 4. TSAb activity in mouse sera. BALB/c and AKR/N mice were vaccinated three or four times with TSHR DNA. In addition, AKR/N mice were injected with fibroblasts coexpressing the TSHR and MHC class II (Shimojo model). Negative controls were mice injected with the empty vector (Vector), or with fibroblasts expressing MHC class II but not the TSHR. TSAb values are reported as % basal camp relative to normal mouse serum. The three sera from AKR/N recipients of TSHR-expressing fibroblasts had elevated serum thyroxine levels (8.2, 9.6, and 12.6 g/dl) (17). Splenocyte proliferation in response to TSHR antigen Several studies indicate that DNA vaccination is more effective at immune priming than inducing antibody production (for example, see Ref. 21). Therefore, it was possible that, despite low or absent TSHR antibody levels, lymphocytes from TSHR-vaccinated mice would proliferate when exposed in vitro to TSHR antigen. To investigate this possibility, we incubated spleen cells with purified TSHR-289. FIG. 5. Splenocytes from BALB/c mice vaccinated with TSHR DNA proliferate in response to TSHR-289. Control animals were vaccinated with the empty vector. 3 H-thymidine ( 3 H-TdR) uptake is presented as the mean SEM (quadruplicate cultures) for splenocytes incubated with active TSHR-289 (0, 4, or 20 g/ml) for 6 d. About 24 h before harvesting, cultures were supplemented with IL-2. Splenocytes were cultured from TSHR DNA vaccinated mice (nos. 1 through 5) and vector DNA vaccinated mice (a d). TSHR antibody levels detected by flow cytometry (FACS) or ELISA are indicated as 2 (clearly positive), (positive), or (negative). *, Values significantly higher in the presence than in the absence of TSHR-289; P 0.05 (t test).

5 3534 Endocrinology, August 2001, 142(8): Pichurin et al. IFN- in TSHR-DNA-Vaccinated Mice FIG. 7. IFN- production in response to TSHR-289 by splenocytes from TSHR DNA-vaccinated BALB/c mice. In contrast, AKR/N mice injected with fibroblasts expressing TSHR and MHC class II (Shimojo model), produce IFN- spontaneously and there is no response to TSHR-289. A, TSHR-vaccinated BALB/c mice (n 5) vs. BALB/c mice vaccinated with the empty vector (n 4). IFN- in culture supernatants was determined by ELISA. Data shown are the mean SEM. *, Values significantly higher in the presence than in the absence of TSHR-289, P 0.01 (Mann Whitney Rank sum test). B, AKR/N mice were injected with TSHR-expressing vs. nonexpressing fibroblasts. Data represent the mean SD for duplicate cultures. NS, Not significantly different. The data shown are representative of six separate experiments in which high levels of IFN- were produced by splenocytes from fibroblast injected mice (17). FIG. 6. Effect of active and inactive TSHR-289 and IL-2 on splenocyte proliferation. Where indicated, IL-2 was added approximately 25 h before harvesting. Data are shown as the mean SEM uptake of 3 H-TdR for quadruplicate wells. A, Splenocytes from BALB/c mice vaccinated with TSHR-DNA or the empty vector (Vector). *, Responses significantly higher than in the absence of TSHR-289, P for TSHR-vaccinated mice and P 0.01 for vector-dna vaccinated mice. B, Splenocytes from AKR/N mice injected with TSHRexpressing or nonexpressing fibroblasts. Differences between groups were not statistically significant. Cytokine production by splenocytes cultured with TSHR-289 We determined the effect of TSHR-289 on the production by spleen cells of the major Th1 and Th2 cytokines, IFN- and IL-4. As described above, supernatants from splenocyte cultures were removed after 6 d and assayed by ELISA. Addition of TSHR-289 to splenocytes from TSHR DNAvaccinated BALB/c mice greatly increased IFN- production from undetectable basal levels (Fig. 7A). Similar responses were induced with active and inactive TSHR-289 (data not shown). Control mice vaccinated with the empty vector showed no increase in IFN- production when challenged with TSHR-289. Overall, IFN- production in BALB/c mice reflected the proliferative response to TSHR-289 (Fig. 5), but the cytokine response provided a more specific indicator of prior TSHR sensitization by DNA vaccination (Fig. 7). Unlike splenocytes from the DNA-vaccinated mice, spleen cells from AKR/N mice injected with fibroblasts (Shimojo model) spontaneously generated very high levels of IFN- in the absence of TSHR-289 (compare open bars in Fig. 7, A and B). Moreover, TSHR-289 did not significantly increase IFN- production by splenocytes from mice injected with TSHRexpressing or nonexpressing fibroblasts (Fig. 7B). The Th2 cytokine IL-4 was undetectable in splenocytes cultures from DNA vaccinated mice, as well as from Shimojo model mice, regardless of the presence or absence of TSHR-289. Discussion Our studies demonstrate that the TSHR antibody response after injecting AKR/N mice with TSHR, MHC class II fibroblasts is greater than that induced by TSHR-DNA vaccination of either AKR/N or BALB/c mice. In particular, the Shimojo approach induces TSHR antibodies detectable using a TBI kit in at least 60% of mice, and TSAb activity is present in the sera of hyperthyroid animals. In contrast, TSHR antibodies were undetectable in all DNA vaccinated BALB/c or AKR/N mice by the same assays, and were only detectable at low levels by flow cytometry or ELISA in some animals. It should be emphasized that neither flow cytometry nor ELISA can be used to measure TSHR antibody levels in mice injected with TSHR-expressing fibroblasts (the Shimojo approach) because of the induction of high levels of nonspecific antibodies. Constitutive expression of the costimulatory molecule B7 1 on the RT4.15HP fibroblasts used in the Shimojo molecule is probably responsible for the generalized immune activation (17) and may well amplify antibody responses to the TSHR. Low or undetectable levels of TSHR antibody in DNA vaccinated mice did not preclude the possibility of lymphocyte sensitization to the TSHR. Indeed, splenocytes from TSHR-DNA vaccinated mice proliferated and produced large amounts of IFN- in response to challenge with puri-

6 Pichurin et al. IFN- in TSHR-DNA-Vaccinated Mice Endocrinology, August 2001, 142(8): fied TSHR-289, an ectodomain variant corresponding to the TSHR A subunit (14, 15). Splenocyte proliferation and INF- production were unrelated to TSHR antibodies in the donor mouse s serum. In these studies we used BALB/c mice because they had been previously used in DNA vaccination (4), as well as in numerous TSHR immunization studies (5, 22, 23). Our finding of TSHR sensitization following DNA vaccination is consistent with the studies of Many et al. (24) in which splenocytes from TSHR DNA-vaccinated mice could be reactivated in vitro with TSHR antigen before adoptive transfer to naive mice. In the Shimojo model, the generalized immune activation, particularly INF- production, precluded our ability to detect in vitro responses to TSHR antigen. Our finding of IFN-, but not IL-4, production in TSHRvaccinated mice is perhaps surprising for two reasons: 1) BALB/c mice generally display a bias toward Th2 responses, at least with respect to Leishmania infection (reviewed in Ref. 25) and 2) TSHR-primed T cells generated by DNA vaccination and transferred to naive BALB/c recipients induced thyroiditis characterized by a Th2 type response, with cells containing IL-4 and IL-10 (24). On the other hand, our data are consistent with numerous other observations that im DNA vaccination against diverse antigens induces Th1-type responses as reflected in IFN- production and antibodies of subclass IgG2a (21, 26, 27). Moreover, as discussed by Costagliola et al. (4), their isolation of three IgG2a mab following TSHR-DNA vaccination is also in agreement with the Th1 bias of im DNA vaccination. The proliferative and cytokine responses to the two alternately folded forms of native TSHR-289 antigen provide additional insight into the outcome of TSHR-DNA vaccination. Active TSHR-289 is recognized by patients autoantibodies and not by a mouse mab (3BD10), whereas the inactive form is recognized by the mab but not by patients autoantibodies (14, 15). Unlike recognition by human autoantibodies, both active and inactive TSHR-289 induced proliferation and IFN- cytokine production by splenocytes from TSHR-DNA vaccinated mice. This lack of discrimination between active and inactive TSHR-289, at least at the T cell level, is likely to be in accordance with the low or absent levels of antibody responses in the TSHR-DNA vaccinated mice. Important observations in our study were the very poor TSHR antibody response and absence of thyroiditis following TSHR DNA-vaccination. These data are in striking contrast to those of Costagliola et al. (4). We can only speculate about the basis for these differences. We used the same mouse strain and the same immunization protocol. We pretreated our mice with cardiotoxin, the agent found by Costagliola et al. (4) to provide the strongest enhancement of genetic immunization. In addition, both studies were performed using BALB/c mice. However, the animals were derived from different breeding colonies, Charles River in the study by Costagliola et al. (4) and The Jackson Laboratory in our case. Another clue to the difference between our findings may be the marked variance in diabetes in NOD mouse colonies around the world, which arises despite similar breeding protocols and may reflect environmental factors including housing conditions (28). Our animal facility is pathogen-free. From discussions at a workshop at the International Thyroid Meeting, Kyoto, Japan (October 2000), it was clear that facilities used by a number of researchers (apparently including that used by Costagliola et al.) are not pathogen-free. Obviously, humans do not live in barrier conditions, and for that unforeseen (and unchangeable) reason our facility may be less well suited than others for the development of an animal model resembling Graves disease in humans. Nevertheless, some environments are less hygienic than others. For example, infectious mononucleosis occurs in older teenagers in Europe but not in undeveloped rural Africa owing to continual exposure from infancy to the causative agent (Epstein-Barr virus) in the latter. Moreover, there is evidence that infectious organisms may play a role in the manifestation of human thyroid autoimmunity (reviewed in Ref. 29). It is possible, therefore, that infections could contribute to the expression of experimentally induced Graves disease. In conclusion, DNA vaccination is much less effective at inducing TSHR antibodies than the Shimojo approach (injecting fibroblasts coexpressing the TSHR and MHC class II), at least under pathogen-free conditions and with this particular inbred colony. The Shimojo approach has the disadvantage of generalized nonspecific immune activation. In contrast, DNA vaccination offers the opportunity for future characterization of TSHR-specific T cells generated without adjuvant. Acknowledgments We thank Dr. Ronald N. Germain, Laboratory for Immunology, Lymphocyte Biology Section, NIAID, NIH, for generously providing us with the RT4.15HP fibroblast line. Received January 31, Accepted April 3, Address all correspondence and requests for reprints to: Sandra M. McLachlan, Cedars-Sinai Medical Center, 8700 Beverly Boulevard, Suite B-131, Los Angeles, California mclachlans@cshs.org. This work was supported by NIH Grants DK and DK References 1. Rapoport B, Chazenbalk GD, Jaume JC, McLachlan SM 1998 The thyrotropin receptor: interaction with thyrotropin and autoantibodies. Endocr Rev 19: Shimojo N, Kohno Y, Yamaguchi K-I, et al Induction of Graves-like disease in mice by immunization with fibroblasts transfected with the thyrotropin repector and a class II molecule. Proc Natl Acad Sci USA 93: Yamaguchi K-I, Shimojo N, Kikuoka S, et al Genetic control of antithyrotropin receptor antibody generation in H-2k mice immunized with thyrotropin receptor-transfected fibroblasts. J Clin Endocrinol Metab 82: Costagliola S, Rodien P, Many M-C, Ludgate M, Vassart G 1998 Genetic immunization against the human thyrotropin receptor causes thyroiditis and allows production of monoclonal antibodies recognizing the native receptor. J Immunol 160: Kaithamana S, Fan J, Osuga Y, Liang SG, Prabhakar BS 1999 Induction of experimental autoimmune Graves disease in BALB/c mice. J Immunol 163: Kita M, Ahmad L, Marians RC, et al Regulation and transfer of a murine model of thyrotropin receptor antibody mediated Graves disease. Endocrinology 140: Jaume JC, Rapoport B, McLachlan SM 1999 Lack of female bias in a mouse model of autoimmune hyperthyroidism (Graves disease). Autoimmunity 29: Costagliola S, Many MC, Denef JF, Pohlenz J, Refetoff S, Vassart G 2000 Genetic immunization of outbred mice with thyrotropin receptor cdna provides a model of Graves disease. J Clin Invest 105: Germain RN, Ashwell JD, Lechler RI, et al Exon-shuffling maps control of antibody- and T-cell-recognition sites to the NH 2 -terminal domain

7 3536 Endocrinology, August 2001, 142(8): Pichurin et al. IFN- in TSHR-DNA-Vaccinated Mice of the class II major histocompatibility polypeptide A. Proc Natl Acad Sci USA 82: Nagayama Y, Kaufman KD, Seto P, Rapoport B 1989 Molecular cloning, sequence and functional expression of the cdna for the human thyrotropin receptor. Biochem Biophys Res Commun 165: Kakinuma A, Chazenbalk G, Filetti S, McLachlan SM, Rapoport B 1996 Both the 5 and 3 non-coding regions of the thyrotropin receptor messenger RNA influence the level of receptor protein expression in transfected mammalian cells. Endocrinology 137: Jaume JC, Guo J, Wang Y, Rapoport B, McLachlan SM 1999 Cellular thyroid peroxidase (TPO), unlike purified TPO and adjuvant, induces antibodies in mice that resemble autoantibodies in human autoimmune thyroid disease. J Clin Endocrinol Metab 84: Chazenbalk GD, Jaume JC, McLachlan SM, Rapoport B 1997 Engineering the human thyrotropin receptor ectodomain from a non-secreted form to a secreted, highly immunoreactive glycoprotein that neutralizes autoantibodies in Graves patients sera. J Biol Chem 272: Chazenbalk GD, Wang Y, Guo J 1999 A mouse monoclonal antibody to a thyrotropin receptor ectodomain variant provides insight into the exquisite antigenic conformational requirement, epitopes and in vivo concentration of human autoantibodies. J Clin Endocrinol Metab 84: Chazenbalk G, McLachlan S, Pichurin P, Rapoport B 2001 A prion-like shift between two conformational forms of a recombinant thyrotropin receptor A subunit module: purification and stabilization using chemical chaperones of the form reactive with Graves autoantibodies. J Clin Endocrinol Metab 86: Chazenbalk GD, Kakinuma A, Jaume JC, McLachlan SM, Rapoport B 1996 Evidence for negative cooperativity among human thyrotropin receptors overexpressed in mammalian cells. Endocrinology 137: Yan X-M, Guo J, Pichurin P, et al Cytokines, IgG subclasses and costimulation in a mouse model of thyroid autoimmunity induced by injection of fibroblasts co-expressing MHC class II and thyroid autoantigens. Clin Exp Immunol 122: Shewring GA, Rees Smith B 1982 An improved radioreceptor assay for TSH receptor antibodies. Clin Endocrinol 17: Kasagi K, Konishi J, Iida Y, et al A new in vitro assay for human thyroid stimulator using cultured thyroid cells: effect of sodium chloride on adenosine 3,5 - monophosphate increase. J Clin Endocrinol Metab 54: Windhagen A, Anderson DE, Carrizosa A, Williams RE, Hafler DA 1996 IL-12 induces human T cells secreting IL-10 with IFN-. J Immunol 157: Laylor R, Porakishvili N, De Souza JB, Playfair JH, Delves PJ, Lund T 1999 DNA vaccination favours memory rather than effector B cell responses. Clin Exp Immunol 117: Wagle NM, Dallas JS, Seetharamaiah GS, et al Induction of hyperthyroxinemia in BALB/C but not in several other strains of mice. Autoimmunity 18: Costagliola S, Many M-C, Stalmans-Falys M, Vassart G, Ludgate M 1995 The autoimmune response induced by immunising female mice with recombinant human thyrotropin receptor varies with the genetic background. Mol Cell Endocrinol 115: Many MC, Costagliola S, Detrait M, et al Development of an animal model of autoimmune thyroid eye disease. J Immunol 162: Fowell DJ, Bix M, Shinkai K, Lacy D, Locksley RM 1998 Disease susceptibility and development of the cytokine repertoire in the murine Leishmania major model. Eur Cytokine Netw 9: Boyle JS, Silva A, Brady JL, Lew AM 1997 DNA immunization: induction of higher avidity antibody and effect of route on T cell cytotoxicity. Proc Natl Acad Sci USA 94: Feltquate DM, Heaney S, Webster RG, Robinson HL 1997 Different T helper cell types and antibody isotypes generated by saline and gene gun DNA immunization. J Immunol 158: Pozzilli P, Signore A, Williams AJ, Beales PE 1993 NOD mouse colonies around the world recent facts and figures. Immunol Today 14: Tomer Y, Davies TF 1993 Infection, thyroid disease, and autoimmunity. Endocr Rev 14: