Dissociation between Iodide-Induced Thyroiditis and Antibody-Mediated Hyperthyroidism in NOD.H-2h4 Mice
http://www.100md.com
内分泌学杂志 2005年第1期
Autoimmune Disease Unit (S.M.M., C.-R.C., H.A., P.N.P., B.R.), Cedars-Sinai Research Institute and University of California Los Angeles School of Medicine, Los Angeles, California 90048; and Department of Internal Medicine, Molecular Microbiology, and Immunology (H.B.-M.), University of Missouri School of Medicine and Veterans Affairs Research Service, Columbia, Missouri 65212
Address all correspondence and requests for reprints to: Sandra M. McLachlan, Cedars-Sinai Medical Center, 8700 Beverly Boulevard, Suite B-131, Los Angeles, California 90048. E-mail: mclachlans@cshs.org.
Abstract
NOD.H-2h4 mice are genetically predisposed to thyroid autoimmunity and spontaneously develop thyroglobulin autoantibodies (TgAb) and thyroiditis. Iodide administration enhances TgAb levels and the incidence and severity of thyroiditis. Using these mice, we investigated the interactions between TSH receptor (TSHR) antibodies induced by vaccination and spontaneous or iodide-enhanced thyroid autoimmunity (thyroiditis and TgAb). Mice were immunized with adenovirus expressing the TSHR A-subunit (or control adenovirus). Thyroid antibodies, histology, and serum thyroxine levels were compared in animals on a regular diet or on a high-iodide diet (0.05% NaI-supplemented water). Thyroiditis severity and TgAb levels were enhanced by iodide administration and were independent of the type of adenovirus used for immunization. In contrast, TSHR antibodies, measured by TSH-binding inhibition, thyroid-stimulating activity, and TSH-blocking activity, were induced in the majority of animals immunized with TSHR (but not control) adenovirus and were unaffected by dietary iodide. The NOD.2h4 strain of mice was less susceptible than BALB/c or BALB/k mice to TSHR adenovirus-induced hyperthyroidism. Nevertheless, hyperthyroidism developed in approximately one third of TSHR adenovirus-injected NOD.2h4 mice. This hyperthyroidism was suppressed by a high-iodide diet, probably by a nonimmune mechanism. The fact that inducing an immune response to the TSHR had no effect on thyroiditis raises the possibility that the TSHR may not be the target involved in the variable thyroiditis component in some humans with Graves’ disease.
Introduction
THYROIDITIS DEVELOPS SPONTANEOUSLY in a number of animals including Obese Strain chickens (1), BB rats (2), and diabetes-prone NOD mice (3). Mice of the NOD.H-2h4 substrain do not develop diabetes, and intriguingly, the development of thyroiditis and thyroglobulin (Tg) autoantibodies (TgAb) is markedly accelerated by increased dietary intake of iodine (4, 5, 6). In contrast to these thyroiditis models, Graves’ disease does not develop spontaneously in animals (reviewed in Ref. 7). However, Graves’-like hyperthyroidism can be experimentally induced by several novel approaches that involve in vivo expression of the TSH receptor (TSHR). For example, hyperthyroidism is induced by injecting fibroblasts, B cells, or dendritic cells expressing the TSHR (8, 9, 10). Alternatively, transient intramuscular TSHR expression by DNA vaccination induces hyperthyroidism in some outbred mice (11), and TSHR adenovirus is a particularly effective approach for generating Graves’ disease in the BALB/c strain of mice (12).
TSHR autoantibodies that stimulate the thyroid are the direct cause of Graves’ hyperthyroidism. In addition, some Graves’ patients have TgAb, and the majority of Graves’ patients has autoantibodies to thyroid peroxidase (TPO) (reviewed in Ref. 13). Thyroid microsomal (TPO) autoantibodies correlate with thyroidal lymphocytic infiltration (14, 15), and Graves’ thyroid glands typically have lymphocytic infiltration, although to a lesser degree than in Hashimoto’s thyroiditis. Therefore, TPO autoantibodies in Graves’ disease may reflect an underlying process of thyroid inflammation. However, hyperthyroidism in Graves’ disease indicates that thyroid stimulation by TSHR autoantibodies exceeds any thyroid damage associated with the autoimmune response to TPO and perhaps also to Tg. In this context, it is noteworthy that changes from hypothyroidism in Hashimoto’s thyroiditis to hyperthyroid Graves’ disease (16), or the reverse (17), have been reported.
Recently, we found that injecting adenovirus expressing the TSHR A-subunit, particularly at low doses, optimized the induction of Graves’ hyperthyroidism in BALB/c mice (18, 19). However, despite goiter and follicular hyperplasia, no thyroid lymphocytic infiltration was evident. Because NOD.H-2h4 mice are prone to develop thyroiditis, we hypothesized that an immune response to the TSHR induced by TSHR A-subunit adenovirus (TSHR-Ad) immunization in these mice would lead to thyroid lymphocytic infiltration, even without iodide administration. Also, because of the known effects of dietary iodide on this mouse strain, we compared the outcome of TSHR adenovirus immunization in animals with or without iodide supplementation in their drinking water.
Materials and Methods
Mice
NOD.H-2h4 mice were derived by crossing NOD mice with B10.A(4R) mice and backcrossing to obtain animals on the NOD background with the B10.A(4R) major histocompatibility complex (MHC) haplotype (I-Ak). The breeding stock was provided by Dr. L. Wicker (Merck Laboratories, Rahway, NJ), and the strain was subsequently bred at the University of Missouri (Columbia, MO). From 7–8 wk of age, half of the mice (all females) were maintained on 0.05% NaI in their drinking water until they were euthanized. All animal studies were approved by the local institutional animal committee and were performed in accordance with the highest standards of care in a pathogen-free facility.
TSHR adenovirus immunization
Construction and purification of adenoviruses expressing TSHR amino acid residues 1–289 (A-subunit) and ?-galactosidase (control) have been described previously (12, 18). In brief, TSHR-Ad and control adenovirus (Con-Ad) were propagated in HEK293 cells (American Type Culture Collection, Manassas, VA) and purified by CsCl density gradient centrifugation, and viral particle concentration was determined by measuring the absorbance at 260 nm (20). All viruses used in this study were from the same preparation and were stored in aliquots at –80 C.
Mice on regular drinking water or on water supplemented with 0.05% NaI (18 animals in each group) were injected im with TSHR-Ad (108 particles in 50 μl PBS) as described previously (19). Concurrently, mice on NaI-supplemented water (n = 11) or regular water (n = 10) were injected with the same dose of Con-Ad. Immunization was performed three times at 3-wk intervals. Blood was drawn 1 wk after the second injection, and mice were euthanized 4 wk after the third injection to obtain blood and thyroid glands.
Assays for TSHR antibodies
TSH-binding inhibition (TBI) was measured with a commercial kit according to the manufacturer’s protocol (Kronus, Boise, ID). Duplicate 25-μl aliquots of test mouse serum (plus 5 μl of normal human serum as carrier protein) were incubated with detergent-solubilized TSHR, [125I]TSH was added, and the TSHR antibody complexes were precipitated with polyethylene glycol. TBI values were calculated from the following formula:
Values greater than the normal range (mean + 2 SD) of TBI activity in Con-Ad-immunized mice were considered positive.
Thyroid-stimulating antibodies (TSAb) and TSH-blocking antibodies (TBAb) were measured and calculated as previously described (18). Monolayers of Chinese hamster ovary cells expressing the wild-type TSHR in 96-well plates were incubated with 3% test serum in 100 μl Hank’s buffer without NaCl and supplemented with 20 mM HEPES (pH 7.4), 1.5 mM isobutylmethylxanthine, 220 mM sucrose, 4% polyethylene glycol 4000, and 0.3% BSA (all from Sigma Aldrich, St. Louis, MO). After 3 h at 37 C, cells were frozen and thawed, and total cAMP content (medium and cells) was measured by RIA.
TSAb was expressed as a percentage of basal cAMP generated in the presence of serum from Con-Ad-immunized mice. TBAb was calculated as follows:
Values greater than the normal range (mean + 2 SD) of TSAb (or TBAb) activity in Con-Ad-immunized mice were considered positive.
Tg antibodies and IgG subclasses
Antibodies to mouse Tg were measured as previously described (5). In brief, ELISA plates were coated with mouse Tg, prepared as described previously (21), and ovalbumin as a control. Antibody binding in sera diluted 1:50 was detected using alkaline phosphatase-conjugated antibodies to total mouse IgG or mouse IgG1 and IgG2b (Southern Biotechnology Inc., Birmingham, AL). Binding data are reported as OD at 410 nm.
Serum T4 and TSH
Total T4 was measured in undiluted serum (25 μl) by RIA using a kit (Diagnostic Products Corporation, Los Angeles, CA). Serum TSH was determined by Dr. Roy Weiss (Thyroid Unit, University of Chicago, Chicago, IL) in undiluted serum (50 μl) by RIA (22).
Thyroid pathology
Histology was examined on multiple formalin-fixed sections of thyroid tissue stained with hematoxylin and eosin. The extent of thyroid lymphocytic infiltration was assessed on a score of 0 to 4, as previously reported (5). Briefly, 0 indicates a paucity of lymphocytes; 1+ thyroiditis is defined as an infiltrate of at least 125 cells in one or several foci; 2+ represents 10–20 foci of cellular infiltration, each the size of several follicles and involving up to one fourth of the gland; 3+ indicates that one fourth to half of the gland is infiltrated by lymphocytes; and 4+ indicates that greater than half of the gland is destroyed.
Statistical analyses
Differences between the magnitude of responses were determined by t test or Mann-Whitney rank sum test (as appropriate). Comparison of multiple groups was assessed by ANOVA. Fisher’s exact test was used to test for significant differences between the number of mice positive or negative for a particular parameter.
Results
The following four groups of NOD.H-2h4 mice were studied: mice immunized with TSHR-Ad maintained without iodide or on water containing 0.05% NaI and mice immunized with Con-Ad maintained without iodide or on iodized water. Adenovirus injections were performed on three occasions at 3-wk intervals, and blood was drawn 1 wk after two injections and at euthanasia 4 wk after the final injection. Sera were characterized for induced TSHR antibodies and for spontaneously arising TgAb. Thyroid function was assessed in terms of total T4 levels, TSH, and thyroid pathology.
Induced TSHR antibodies
A high proportion of mice had TBI antibodies after two TSHR-Ad injections, including 14 of 18 mice (78%) on regular water and 13 of 17 mice (67%) on iodized water (Fig. 1A). In contrast, zero (0%) of 11 Con-Ad-immunized mice on iodized water had TBI activity, and only one of 10 mice (10%) on regular water was borderline TBI positive. After three injections, TBI levels remained negative (with one exception) in Con-Ad-immunized mice, whereas 94% of TSHR-Ad-injected animals became TBI positive regardless of whether they were given regular or iodized water (Fig. 1B). Moreover, exposure to iodide had no effect on the levels of TBI activity in TSHR-Ad-immunized animals.
FIG. 1. TSHR antibodies measured by TBI develop in NOD.H-2h4 female mice immunized with TSHR-Ad, regardless of iodide exposure. The following four groups of mice were studied: mice immunized with TSHR-Ad that were maintained on regular drinking water (n = 18) or on water containing NaI (n = 18) and mice immunized with Con-Ad that were maintained on regular water (n = 10) or on water containing NaI (n = 11). Three injections of adenovirus were administered at 3-wk intervals. TBI values were measured in sera obtained 1 wk after two injections (A) and at euthanasia 4 wk after the third injection (B). TBI values are shown for individual mice without iodide in open circles and with iodide in black circles. The shaded area represents the mean ± 2 SD for Con-Ad immunized mice on regular water. *, Values significantly greater than for Con-Ad-immunized mice without iodide (P < 0.05, ANOVA on ranks).
After three injections, sufficient serum was available from approximately half the mice in each group for bioassay of TSAb and TBAb. None of the Con-Ad-injected mice had TSAb activity in their sera, regardless of exposure to iodide. In contrast, the majority of TSHR-Ad-injected mice was TSAb positive (eight of 10 mice on regular water and 10 of 10 mice on iodized water; Fig. 2A). In terms of TBAb activity, seven of 10 TSHR-Ad mice were positive vs. zero of five Con-Ad-immunized mice maintained on regular water (Fig. 3B). For mice on iodized water, five of 10 TSHR-Ad animals were TBAb positive vs. one of five Con-Ad mice. As for TBI, administration of iodide to the TSHR-Ad-immunized mice did not influence the magnitude of TSAb (or TBAb) activity or the proportions of animals positive for TSAb (or TBAb).
FIG. 2. TSAb and TBAb activity in NOD.H-2h4 mice immunized with TSHR-Ad. As described for Fig. 1, mice immunized with TSHR-Ad were maintained on regular water (n = 10) or NaI-supplemented water (n = 10), and mice immunized with Con-Ad were maintained on regular water (n = 5) or NaI-supplemented water (n = 5). TSAb (A) and TBAb (B) activities were measured at euthanasia 4 wk after the third injection. Data are shown for individual animals without iodide in open circles and with iodide in black circles. The dotted line represents the upper limit of TSAb or TBAb values (mean + 2 SD) in Con-Ad-immunized animals on regular water. *, Values significantly greater than for Con-Ad-immunized mice without iodide (P < 0.05, ANOVA on ranks).
FIG. 3. Autoantibodies to murine Tg arise spontaneously in NOD.H-2h4 mice, and their levels are enhanced by iodide, regardless of immunization with TSHR-Ad (or Con-Ad). Antibody levels, measured 4 wk after the third injection, are shown as ELISA OD410 for individual mice without iodide in open circles and with iodide in black circles. Mice immunized with TSHR-Ad were maintained on regular water (n = 8) or NaI-supplemented water (n = 8), and mice immunized with Con-Ad were maintained on regular water (n = 5) or NaI-supplemented water (n = 5). IgG class TgAb were assessed by measuring serum IgG binding to ELISA wells coated with murine Tg or, as a control, with ovalbumin (OVA) (A). TgAb of subclasses IgG1 and IgG2b were measured using ELISA wells coated with murine Tg and subclass-specific antisera (B). Significance of the indicated differences is as follows: A: *, P = 0.011; and **, P = 0.002 (t tests); B: #, P = 0.038; and ##, P = 0.008 (rank sum tests); and *, P = 0.023 (t test).
Spontaneous TgAb
Unlike TSHR antibodies that were generated in response to TSHR-Ad immunization, TgAb developed in NOD.H-2h4 mice regardless of the type of adenovirus injected. In particular, mice injected with either Con-Ad or TSHR-Ad had detectable IgG class antibody binding to mouse Tg-coated ELISA wells, but not to wells coated with ovalbumin (Fig. 3A). Moreover, TgAb levels were higher in mice on iodized water, irrespective of immunization with Con-Ad or TSHR-Ad. Furthermore, TgAb of subclass IgG1 and, to a lesser extent, subclass IgG2b were higher in mice exposed to iodide than in mice on regular water (Fig. 3B). Overall, TgAb levels were directly related to iodide exposure but were not influenced by the type of adenovirus used for immunization.
Thyroid function
Mice on regular water immunized with Con-Ad had normal serum T4 levels when measured after the second (Fig. 4A) and third (Fig. 4B) injections, with a single exception (one mouse after the third injection). Chronic administration of iodide to Con-Ad-immunized mice tended to reduce T4 levels, reaching significance after the third injection (Fig. 4, A and B).
FIG. 4. Serum T4 levels in NOD.H-2h4 mice immunized with TSHR-Ad or Con-Ad and maintained on regular drinking water or on iodized water. Serum T4 levels were measured 1 wk after two injections and at euthanasia 4 wk after three injections. Data shown are T4 values for individual mice immunized with TSHR-Ad and maintained on regular drinking water (n = 18; open circles) or on NaI-supplemented water (n = 18; black circles) and mice immunized with Con-Ad and maintained on regular water (n = 10; open circles) or NaI-supplemented water (n = 11; black circles). The shaded area represents the mean ± 2 SD for T4 levels in Con-Ad-immunized mice on regular water. Mice with T4 levels above these values are considered hyperthyroid. T4 levels significantly different between the indicated groups are indicated by the following: A: **, P = 0.002 (rank sum test); B: #, P = 0.022 (rank sum test); and *, P = 0.036 (t test).
The effect of iodide in reducing serum T4 levels was also observed in the TSHR-Ad-immunized mice. Thus, after the second injection, five of 18 mice on regular water but only one of 18 mice on iodized water had elevated serum T4 levels (Fig. 4A). The same pattern was seen after the third TSHR-Ad injection (five of 18 and one of 17 mice were hyperthyroid on regular or iodized water, respectively; Fig. 4B).
Although serum was available for TSH analysis on only a limited number of animals, these data provided additional insight into the thyroid status of the NOD.H-2h4 mice. TSH levels were only elevated in some animals exposed to iodide (Fig. 5). It should be appreciated that the mouse TSH assay lacks the sensitivity of present human TSH assays. Nevertheless, the only sera with TSH levels less than 10 mU/liter were from mice immunized with TSHR-Ad on regular water.
FIG. 5. TSH levels are elevated in some NOD.H-2h4 mice exposed to iodide and reduced in some animals immunized with TSHR-Ad and maintained on regular water. Values (mU/liter) are shown for individual mice immunized with TSHR-Ad and maintained on regular drinking water (n = 9; open circles) or NaI-supplemented water (n = 10; black circles) and mice immunized with Con-Ad on regular water (n = 4; open circles) or NaI-supplemented water (n = 5; black circles). The solid line indicates the upper limit (mean + 2 SD) of TSH in Con-Ad-immunized mice on regular water. The lower limit of the assay is less than 10 mU/liter TSH (dashed line). *, P < 0.001 (rank sum test) for TSH levels significantly different between the indicated groups.
Thyroid pathology
Thyroid pathology was characterized in terms of lymphocytic infiltration (thyroiditis) and thyrocyte morphology. The thyroiditis score was related to iodide exposure and was significantly higher in TSHR-Ad-immunized mice exposed to iodide than mice immunized in the same way on regular water (Fig. 6, P = 0.026). Similar changes occurred in Con-Ad-immunized animals, but the difference was not statistically significant. Two Con-Ad-immunized mice on regular water were atypical in that they had much greater lymphocytic infiltrates than previously reported for nonimmunized mice on normal iodide intake (5).
FIG. 6. Thyroid pathology in NOD.H-2h4 mice immunized with TSHR-Ad or Con-Ad and maintained on regular water or NaI-supplemented water. The mean thyroiditis score (+ SEM) is shown for mice immunized with TSHR-Ad on regular drinking water (n = 17) or NaI-supplemented water (n = 18) and for mice immunized with Con-Ad on regular water (n = 10) or NaI-supplemented water (n = 10). *, Values significantly different between the indicated groups (P < 0.026, rank sum test).
After Con-Ad immunization, examples are shown in Fig. 7 of minimal thyroid lymphocytic infiltration in mice on regular water (Fig. 7A, grade 0+) vs. iodide-supplemented water (Fig. 7B, grade 2+ to 3+). After TSHR-Ad immunization, examples of minimal thyroiditis in a mouse on regular water (grade 0+) and moderate thyroiditis (grade 2+) in a mouse on iodized water are shown in Fig. 7, C and D, respectively. At higher magnification, thyroid epithelial cells appeared to be slightly flatter in euthyroid animals (Fig. 7E) than the more cuboidal cells observed in some hyperthyroid mice (Fig. 7F).
FIG. 7. Thyroid histology in NOD.H-2h4 mice immunized with TSHR-Ad or Con-Ad. Sections were stained with hematoxylin and eosin. A and B, Con-Ad-immunized mouse on regular water (A) or iodized water (B). Magnification, x100. C and D, Hyperthyroid TSHR-Ad-immunized mouse on regular water (C) and euthyroid mouse on iodized water (D). Magnification, x100. E and F, Euthyroid mouse on iodized water (E) and hyperthyroid TSHR-Ad immunized mouse on regular water (F). Magnification, x400.
Discussion
NOD.H-2h4 mice are a particularly valuable strain for investigating autoimmune thyroid disease because they do not develop diabetes but have an increased incidence of spontaneous thyroiditis compared with diabetes-prone NOD mice (23). Moreover, it has previously been shown that iodide supplementation markedly enhances the incidence of TgAb and the incidence and severity of thyroiditis in NOD.H-2h4 mice (4, 5, 6). In these mice, TPO is not a major autoantigen. TPO-induced T-cell responses have been observed in only one study (24), and TPO autoantibodies have been sought but not found (4, 24). Tg is the major autoantigen in this thyroiditis-prone mouse strain (4, 5, 6, 24).
In the present study, we explored the net effect on thyroid status in NOD.H-2h4 mice of three different factors, namely spontaneous thyroid autoimmunity, induced TSHR antibodies, and dietary iodide. TgAb levels are increased by iodide ingestion in both TSHR-Ad- and Con-Ad-immunized animals. Similarly, low-grade thyroiditis (grades 0 to 1+) is exacerbated by iodide (typically grades 2+ to 3+). Moreover, contrary to our hypothesis, inducing an immune response to the TSHR by TSHR-Ad immunization had no effect on thyroiditis. As anticipated, immunization with TSHR-Ad, but not with Con-Ad, induced TSHR antibodies as measured by TBI, TSAb, and TBAb activities. Again, iodide exposure had no statistically significant effect on these antibody levels. Consequently, iodide and TSHR immunization had independent effects on thyroiditis/TgAb vs. TSHR antibodies, respectively. Turning to thyroid function, two major effects were observed. First, iodide reduced serum T4 levels in all groups of mice, whether immunized with Con-Ad or TSHR-Ad. Although reduced, T4 levels in most mice remained within the normal range. Second, TSHR-Ad immunization induced thyrotoxicosis in approximately 30% of mice on a normal iodide diet. Taken together, the suppressive effect of iodide abrogated the thyroid stimulation induced by TSHR-Ad immunization.
Incidentally, although thyroid histology has been extensively studied in NOD mice including the NOD.H-2h4 strain, there are only two reports describing the effect of a high-iodide diet on serum T4 levels in these mice. In NOD mice, iodide administration modestly decreased serum T4 levels (25), whereas in the NOD.H-2h4 strain, iodide had no effect (4). Our present data demonstrating that iodide consistently reduced serum T4 levels are the first to show this effect in the NOD.H-2h4 strain.
What is the mechanism by which chronic ingestion of a high-iodine diet reduces serum T4 levels? There are three possible mechanisms. First, the greater degree of thyroiditis induced by iodide could reduce the capacity of the thyroid gland to synthesize and secrete thyroid hormone. During progressive thyroid failure in Hashimoto’s thyroiditis, as serum T4 levels decline, increased TSH drive prolongs euthyroidism by increasing secretion of T3 (not measured in our study). We consider this mechanism possible but unlikely because the degree of thyroid destruction was limited. In normal rats, serum T4 levels have returned to normal by 5 wk after hemithyroidectomy (26). Increasing the severity of thyroiditis by administering iodide to BB/W rats does not reduce serum T4 levels (27) unless the animals have previously undergone hemithyroidectomy (28). However, it is possible that thyrocyte function was diminished by cytokines derived from focal nondestructive infiltrates. A second possible mechanism, failure to escape from the Wolff-Chaikoff block of iodide organification (29), is also unlikely because this process leads to a vicious cycle in which overt hypothyroidism ensues (30). The third possible mechanism, which we consider most likely, is a postulated mild suppression of thyroid hormone secretion, with thyroid hormone levels generally remaining in the normal range, which is an effect that has been demonstrated in humans (31).
Several other points of interest arise from these studies. First, TSHR-Ad immunization of NOD.H-2h4 mice clearly induced an immune response to the TSHR, yet it did not increase the severity of thyroiditis. One possible explanation is that we immunized with human, not murine, TSHR A-subunit, thereby precluding homing of TSHR-specific T lymphocytes to the thyroid. This explanation seems unlikely because conventional immunization with nonmurine thyroid antigens induces severe thyroiditis. For example, porcine Tg in complete Freund’s adjuvant induced extensive thyroiditis in NOD.H-k mice (23). Despite the well-known immunogenicity of adenovirus (32), its adjuvant properties are clearly much less effective than mycobacteria for inducing immune responses leading to tissue infiltration. Also, it is unlikely that the inability of TSHR adenovirus to induce thyroiditis is attributable to the lack of a Th1 response. TSHR adenovirus immunization does induce T helper 1 (Th1)-type responses (production of interferon- but not IL-4) (33, 34). It is possible that adenovirus immunization does not induce TGF? production, which is critical for thyroiditis development in NOD.H-2h4 mice (35).
Another point of interest is that our data differ with those from a previous investigation of the NOD strain in which TSHR immunization did not induce functional TSHR antibodies (36). In the earlier study, mice injected with TSHR bacterial fusion protein plus adjuvant developed TSHR antibodies detected by ELISA, but these antibodies lacked TBI, TSAb, or TBAb activity. This lack of functional antibody production is not unexpected considering the many failed attempts to induce Graves’ disease using different TSHR proteins and different adjuvants in different mouse strains (reviewed in Ref. 7). However, the same TSHR fusion protein-adjuvant combination induced some functional antibody activities (TBI and TBAb) in BALB/c mice (37).
These earlier data could imply that BALB/c mice are more susceptible than NOD mice to the induction of functional TSHR antibodies. However, our data indicate that TBI, TSAb, and TBAb activities are readily induced in NOD.H-2h4 mice by a different immunization protocol, namely by injecting TSHR adenovirus. Nevertheless, the NOD strain of mice is more resistant than BALB/c mice (12) to developing hyperthyroidism, particularly using low doses of adenovirus encoding the TSHR A-subunit (18, 19). That is, TSAbs appear to be relatively ineffective in the NOD.H-2h4 strain. Similar resistance to TSAb activity has been observed in other strains including C57BL/6 mice (12, 38). BALB/k mice have the same MHC class II (I-Ak) as NOD.H-2h4 mice, yet they are as susceptible to hyperthyroidism as BALB/c mice with a different MHC class II (I-Ad) (39). Consequently, the non-MHC background of the NOD strain is likely to play a role in its resistance to hyperthyroidism, despite the presence of TSAb.
Incidentally, in our previous study on NOD.H-2h4 mice involving TSHR-DNA plasmid (not adenovirus) vaccination (40), we were surprised that TgAb levels were undetectable, and minimal thyroiditis developed on a high-iodide diet. The reason for the marked difference between these two outcomes was only understood after completion of the DNA plasmid study. Taconic Farms (Germantown, NY), the source of the NOD.H-2h4 mice for the DNA plasmid study, notified investigators that they had provided a re-derived strain that, on later testing in their facility, confirmed our finding of the absence of thyroiditis on iodide treatment. Our present study used the original strain (see Materials and Methods).
In summary, we used NOD.H-2h4 mice to explore the net effects on thyroid status of spontaneous thyroid autoimmunity, induced TSHR antibodies, and iodide administration. Thyroiditis and TgAb were regulated by iodide independent of TSHR immunization. In contrast, TSHR antibodies induced by injecting TSHR-Ad were unaffected by dietary iodide. Nevertheless, the hyperthyroidism that developed in approximately one third of TSHR-Ad-injected mice was suppressed by a high-iodide diet, probably by a nonimmune mechanism (iodine autoregulation of thyroid function, as already mentioned). Finally, as reported for other mouse strains (12, 18), inducing immune responses to the TSHR using TSHR-Ad immunization is not associated with thyroid lymphocytic infiltration in thyroiditis-prone NOD.H-2h4 mice. In addition to hyperthyroidism, Graves’ disease typically has a thyroiditis component, as evident by the common presence of TPO autoantibodies. Caution should always be used in extrapolating from animal models to human disease. However, our studies in NOD.H-2h4 mice raise the possibility that, at least in some individuals, the variable thyroiditis component in Graves’ disease may be related to an antigen other than the TSHR.
Acknowledgments
We thank Dr. L. Wicker (Merck Laboratories, Rahway, NJ) for providing the breeding stock of NOD.H-2h4 mice. In addition, we are grateful for contributions by Dr. Boris Catz (Los Angeles, CA).
References
Wick G, Most J, Schauenstein K, Kromer G, Dietrich H, Ziemicki A, Fassler R, Schwarz S, Neu N, Hala K 1985 Spontaneous autoimmune thyroiditis: a bird’s eye view. Immunol Today 6:359–365
Colle E, Guttmann RD, Seemayer TA 1985 Association of spontaneous thyroiditis with the major histocompatibility complex of the rat. Endocrinology 116:1243–1247
Bernard NF, Ertug F, Margolese H 1992 High incidence of thyroiditis and anti-thyroid autoantibodies in NOD mice. Diabetes 41:40–46
Rasooly L, Burek CL, Rose NR 1996 Iodine-induced autoimmune thyroiditis in NOD-H2h4 mice. Clin Immunol Immunopathol 81:287–292
Braley-Mullen H, Sharp GC, Medling B, Tang H 1999 Spontaneous autoimmune thyroiditis in NOD.H-2h4 mice. J Autoimmun 12:157–165
Hutchings PR, Verma S, Phillips JM, Harach SZ, Howlett S, Cooke A 1999 Both CD4(+) T cells and CD8(+) T cells are required for iodine accelerated thyroiditis in NOD mice. Cell Immunol 192:113–121
Rapoport B, Chazenbalk GD, Jaume JC, McLachlan SM 1998 The thyrotropin receptor: interaction with thyrotropin and autoantibodies. Endocr Rev 19:673–716
Shimojo N, Kohno Y, Yamaguchi K-I, Kikuoka S-I, Hoshioka A, Niimi H, Hirai A, Tamura Y, Saito Y, Kohn LD, Tahara K 1996 Induction of Graves-like disease in mice by immunization with fibroblasts transfected with the thyrotropin receptor and a class II molecule. Proc Natl Acad Sci USA 93:11074–11079
Kaithamana S, Fan J, Osuga Y, Liang SG, Prabhakar BS 1999 Induction of experimental autoimmune Graves’ disease in BALB/c mice. J Immunol 163:5157–5164
Kita-Furuyama M, Nagayama Y, Pichurin P, McLachlan SM, Rapoport B, Eguchi K 2003 Dendritic cells infected with adenovirus expressing the thyrotropin receptor induce Graves’ hyperthyroidism in BALB/c mice. Clin Exp Immunol 131:234–240
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 105:803–811
Nagayama Y, Kita-Furuyama M, Ando T, Nakao K, Mizuguchi H, Hayakawa T, Eguchi K, Niwa M 2002 A novel murine model of Graves’ hyperthyroidism with intramuscular injection of adenovirus expressing the thyrotropin receptor. J Immunol 168:2789–2794
McLachlan SM, Rapoport B 2004 Why measure thyroglobulin autoantibodies rather than thyroid peroxidase autoantibodies? Thyroid 14:510–520
Yoshida H, Amino N, Yagawa K, Uemura K, Satoh M, Miyai K, Kumahara Y 1978 Association of serum antithyroid antibodies with lymphocytic infiltration of the thyroid gland: studies of seventy autopsied cases. J Clin Endocrinol Metab 46:859–862
Paschke R, Vogg M, Swillens S, Usadel KH 1993 Correlation of microsomal antibodies with the intensity of the intrathyroidal autoimmune process in Graves’ disease. J Clin Endocrinol Metab 77:939–943
Takasu N, Yamada T, Sato A, Nakagawa M, Komiya I, Nagasawa Y, Asawa T 1990 Graves’ disease following hypothyroidism due to Hashimoto’s disease: studies of eight cases. Clin Endocrinol (Oxf) 33:687–698
Tamai H, Kasagi K, Takaichi Y, Takamatsu J, Komaki G, Matsubayashi S, Konishi J, Kuma K, Kumagai LF, Nagataki S 1989 Development of spontaneous hypothyroidism in patients with Graves’ disease treated with antithyroidal drugs: clinical, immunological, and histological findings in 26 patients. J Clin Endocrinol Metab 69:49–53
Chen C-R, Pichurin P, Nagayama Y, Latrofa F, Rapoport B, McLachlan SM 2003 The thyrotropin receptor autoantigen in Graves’ disease is the culprit as well as the victim. J Clin Invest 111:1897–1904
Chen C-R, Pichurin P, Chazenbalk GD, Aliesky H, Nagayama Y, McLachlan SM, Rapoport B 2004 Low-dose immunization with adenovirus expressing the thyroid-stimulating hormone receptor A-subunit deviates the antibody response toward that of autoantibodies in human Graves’ disease. Endocrinology 145:228–233
Mittereder N, March KL, Trapnell BC 1996 Evaluation of the concentration and bioactivity of adenovirus vectors for gene therapy. J Virol 70:7498–7509
Braley-Mullen H, Johnson M, Sharp GC, Kyriakos M 1985 Induction of experimental autoimmune thyroiditis in mice with in vitro activated splenic T cells. Cell Immunol 93:132–143
Pohlenz J, Maqueem A, Cua K, Weiss RE, Van Sande J, Refetoff S 1999 Improved radioimmunoassay for measurement of mouse thyrotropin in serum: strain differences in thyrotropin concentration and thyrotroph sensitivity to thyroid hormone. Thyroid 9:1265–1271
Damotte D, Colomb E, Cailleau C, Brousse N, Charreire J, Carnaud C 1997 Analysis of susceptibility of NOD mice to spontaneous and experimentally induced thyroiditis. Eur J Immunol 27:2854–2862
Verma S, Hutchings P, Guo J, McLachlan S, Rapoport B, Cooke A 2000 Role of MHC class I expression and CD8(+) T cells in the evolution of iodine-induced thyroiditis in NOD-H2(h4) and NOD mice. Eur J Immunol 30:1191–1202
Tani J, Mori K, Hoshikawa S, Nakazawa T, Satoh J, Nakagawa Y, Ito S, Yoshida K 2002 Prevention of lymphocytic thyroiditis in iodide-treated non-obese diabetic mice lacking interferon regulatory factor-1. Eur J Endocrinol 147:809–814
Clark OH, Lambert WR, Cavalieri RR, Rapoport B, Hammond ME, Ingbar SH 1976 Compensatory thyroid hypertrophy after hemithyroidectomy in rats. Endocrinology 99:988–995
Allen EM, Appel MC, Braverman LE 1986 The effect of iodide ingestion on the development of spontaneous lymphocytic thyroiditis in the diabetes-prone BB/W rat. Endocrinology 118:1977–1981
Allen EM, Appel MC, Braverman LE 1987 Iodine-induced thyroiditis and hypothyroidism in the hemithyroidectomized BB/W rat. Endocrinology 121:481–485
Ingbar SH 1972 Autoregulation of the thyroid. Response to iodide excess and depletion. Mayo Clin Proc 47:814–823
Braverman LE, Ingbar SH, Vagenakis AG, Adams L, Maloof F 1971 Enhanced susceptibility to iodide myxedema in patients with Hashimoto’s thyroiditis. J Clin Endocrinol Metab 32:515–521
Vagenakis AG, Downs P, Braverman LE, Burger A, Ingbar SH 1973 Control of thyroid hormone secretion in normal subjects receiving iodides. J Clin Invest 52:528–532
Geutskens SB, van der Eb MM, Plomp AC, Jonges LE, Cramer SJ, Ensink NG, Kuppen PJ, Hoeben RC 2000 Recombinant adenoviral vectors have adjuvant activity and stimulate T cell responses against tumor cells. Gene Ther 7:1410–1416
Pichurin PN, Chen C-R, Nagayama Y, Pichurina O, Rapoport B, McLachlan SM 2004 Evidence that factors other than particular thyrotropin receptor T cell epitopes contribute to the development of hyperthyroidism in murine Graves’ disease. Clin Exp Immunol 135:391–397
Nagayama Y, Mizuguchi H, Hayakawa T, Niwa M, McLachlan SM, Rapoport B 2003 Prevention of autoantibody-mediated Graves’-like hyperthyroidism in mice with IL-4, a Th2 cytokine. J Immunol 170:3522–3527
Braley-Mullen H, Chen K, Wei Y, Yu S 2001 Role of TGF? in development of spontaneous autoimmune thyroiditis in NOD.H-2h4 mice. J Immunol 167:7111–7118
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 115:199–206
Costagliola S, Many M-C, Stalmans-Falys M, Tonacchera M, Vassart G, Ludgate M 1994 Recombinant thyrotropin receptor and the induction of autoimmune thyroid disease in BALB/c mice: a new animal model. Endocrinology 135:2150–2159
Chen CR, Aliesky H, Pichurin PN, Nagayama Y, McLachlan SM, Rapoport B 2004 Susceptibility rather than resistance to hyperthyroidism is dominant in a thyrotropin receptor adenovirus-induced animal model of Graves’ disease as revealed by BALB/c-C57BL/6 hybrid mice. Endocrinology 145:4927–4933
Nagayama Y, McLachlan SM, Rapoport B, Niwa M 2003 A major role for non-MHC genes, but not for micro-organisms, in a novel model of Graves’ hyperthyroidism. Thyroid 13:233–238
Pichurin P, Schwarz-Lauer L, Braley-Mullen H, Paras C, Pichurina O, Morris JC, Rapoport B, McLachlan SM 2002 Peptide scanning for thyrotropin receptor T-cell epitopes in mice vaccinated with naked DNA. Thyroid 12:755–764(Sandra M. McLachlan, Hele)
Address all correspondence and requests for reprints to: Sandra M. McLachlan, Cedars-Sinai Medical Center, 8700 Beverly Boulevard, Suite B-131, Los Angeles, California 90048. E-mail: mclachlans@cshs.org.
Abstract
NOD.H-2h4 mice are genetically predisposed to thyroid autoimmunity and spontaneously develop thyroglobulin autoantibodies (TgAb) and thyroiditis. Iodide administration enhances TgAb levels and the incidence and severity of thyroiditis. Using these mice, we investigated the interactions between TSH receptor (TSHR) antibodies induced by vaccination and spontaneous or iodide-enhanced thyroid autoimmunity (thyroiditis and TgAb). Mice were immunized with adenovirus expressing the TSHR A-subunit (or control adenovirus). Thyroid antibodies, histology, and serum thyroxine levels were compared in animals on a regular diet or on a high-iodide diet (0.05% NaI-supplemented water). Thyroiditis severity and TgAb levels were enhanced by iodide administration and were independent of the type of adenovirus used for immunization. In contrast, TSHR antibodies, measured by TSH-binding inhibition, thyroid-stimulating activity, and TSH-blocking activity, were induced in the majority of animals immunized with TSHR (but not control) adenovirus and were unaffected by dietary iodide. The NOD.2h4 strain of mice was less susceptible than BALB/c or BALB/k mice to TSHR adenovirus-induced hyperthyroidism. Nevertheless, hyperthyroidism developed in approximately one third of TSHR adenovirus-injected NOD.2h4 mice. This hyperthyroidism was suppressed by a high-iodide diet, probably by a nonimmune mechanism. The fact that inducing an immune response to the TSHR had no effect on thyroiditis raises the possibility that the TSHR may not be the target involved in the variable thyroiditis component in some humans with Graves’ disease.
Introduction
THYROIDITIS DEVELOPS SPONTANEOUSLY in a number of animals including Obese Strain chickens (1), BB rats (2), and diabetes-prone NOD mice (3). Mice of the NOD.H-2h4 substrain do not develop diabetes, and intriguingly, the development of thyroiditis and thyroglobulin (Tg) autoantibodies (TgAb) is markedly accelerated by increased dietary intake of iodine (4, 5, 6). In contrast to these thyroiditis models, Graves’ disease does not develop spontaneously in animals (reviewed in Ref. 7). However, Graves’-like hyperthyroidism can be experimentally induced by several novel approaches that involve in vivo expression of the TSH receptor (TSHR). For example, hyperthyroidism is induced by injecting fibroblasts, B cells, or dendritic cells expressing the TSHR (8, 9, 10). Alternatively, transient intramuscular TSHR expression by DNA vaccination induces hyperthyroidism in some outbred mice (11), and TSHR adenovirus is a particularly effective approach for generating Graves’ disease in the BALB/c strain of mice (12).
TSHR autoantibodies that stimulate the thyroid are the direct cause of Graves’ hyperthyroidism. In addition, some Graves’ patients have TgAb, and the majority of Graves’ patients has autoantibodies to thyroid peroxidase (TPO) (reviewed in Ref. 13). Thyroid microsomal (TPO) autoantibodies correlate with thyroidal lymphocytic infiltration (14, 15), and Graves’ thyroid glands typically have lymphocytic infiltration, although to a lesser degree than in Hashimoto’s thyroiditis. Therefore, TPO autoantibodies in Graves’ disease may reflect an underlying process of thyroid inflammation. However, hyperthyroidism in Graves’ disease indicates that thyroid stimulation by TSHR autoantibodies exceeds any thyroid damage associated with the autoimmune response to TPO and perhaps also to Tg. In this context, it is noteworthy that changes from hypothyroidism in Hashimoto’s thyroiditis to hyperthyroid Graves’ disease (16), or the reverse (17), have been reported.
Recently, we found that injecting adenovirus expressing the TSHR A-subunit, particularly at low doses, optimized the induction of Graves’ hyperthyroidism in BALB/c mice (18, 19). However, despite goiter and follicular hyperplasia, no thyroid lymphocytic infiltration was evident. Because NOD.H-2h4 mice are prone to develop thyroiditis, we hypothesized that an immune response to the TSHR induced by TSHR A-subunit adenovirus (TSHR-Ad) immunization in these mice would lead to thyroid lymphocytic infiltration, even without iodide administration. Also, because of the known effects of dietary iodide on this mouse strain, we compared the outcome of TSHR adenovirus immunization in animals with or without iodide supplementation in their drinking water.
Materials and Methods
Mice
NOD.H-2h4 mice were derived by crossing NOD mice with B10.A(4R) mice and backcrossing to obtain animals on the NOD background with the B10.A(4R) major histocompatibility complex (MHC) haplotype (I-Ak). The breeding stock was provided by Dr. L. Wicker (Merck Laboratories, Rahway, NJ), and the strain was subsequently bred at the University of Missouri (Columbia, MO). From 7–8 wk of age, half of the mice (all females) were maintained on 0.05% NaI in their drinking water until they were euthanized. All animal studies were approved by the local institutional animal committee and were performed in accordance with the highest standards of care in a pathogen-free facility.
TSHR adenovirus immunization
Construction and purification of adenoviruses expressing TSHR amino acid residues 1–289 (A-subunit) and ?-galactosidase (control) have been described previously (12, 18). In brief, TSHR-Ad and control adenovirus (Con-Ad) were propagated in HEK293 cells (American Type Culture Collection, Manassas, VA) and purified by CsCl density gradient centrifugation, and viral particle concentration was determined by measuring the absorbance at 260 nm (20). All viruses used in this study were from the same preparation and were stored in aliquots at –80 C.
Mice on regular drinking water or on water supplemented with 0.05% NaI (18 animals in each group) were injected im with TSHR-Ad (108 particles in 50 μl PBS) as described previously (19). Concurrently, mice on NaI-supplemented water (n = 11) or regular water (n = 10) were injected with the same dose of Con-Ad. Immunization was performed three times at 3-wk intervals. Blood was drawn 1 wk after the second injection, and mice were euthanized 4 wk after the third injection to obtain blood and thyroid glands.
Assays for TSHR antibodies
TSH-binding inhibition (TBI) was measured with a commercial kit according to the manufacturer’s protocol (Kronus, Boise, ID). Duplicate 25-μl aliquots of test mouse serum (plus 5 μl of normal human serum as carrier protein) were incubated with detergent-solubilized TSHR, [125I]TSH was added, and the TSHR antibody complexes were precipitated with polyethylene glycol. TBI values were calculated from the following formula:
Values greater than the normal range (mean + 2 SD) of TBI activity in Con-Ad-immunized mice were considered positive.
Thyroid-stimulating antibodies (TSAb) and TSH-blocking antibodies (TBAb) were measured and calculated as previously described (18). Monolayers of Chinese hamster ovary cells expressing the wild-type TSHR in 96-well plates were incubated with 3% test serum in 100 μl Hank’s buffer without NaCl and supplemented with 20 mM HEPES (pH 7.4), 1.5 mM isobutylmethylxanthine, 220 mM sucrose, 4% polyethylene glycol 4000, and 0.3% BSA (all from Sigma Aldrich, St. Louis, MO). After 3 h at 37 C, cells were frozen and thawed, and total cAMP content (medium and cells) was measured by RIA.
TSAb was expressed as a percentage of basal cAMP generated in the presence of serum from Con-Ad-immunized mice. TBAb was calculated as follows:
Values greater than the normal range (mean + 2 SD) of TSAb (or TBAb) activity in Con-Ad-immunized mice were considered positive.
Tg antibodies and IgG subclasses
Antibodies to mouse Tg were measured as previously described (5). In brief, ELISA plates were coated with mouse Tg, prepared as described previously (21), and ovalbumin as a control. Antibody binding in sera diluted 1:50 was detected using alkaline phosphatase-conjugated antibodies to total mouse IgG or mouse IgG1 and IgG2b (Southern Biotechnology Inc., Birmingham, AL). Binding data are reported as OD at 410 nm.
Serum T4 and TSH
Total T4 was measured in undiluted serum (25 μl) by RIA using a kit (Diagnostic Products Corporation, Los Angeles, CA). Serum TSH was determined by Dr. Roy Weiss (Thyroid Unit, University of Chicago, Chicago, IL) in undiluted serum (50 μl) by RIA (22).
Thyroid pathology
Histology was examined on multiple formalin-fixed sections of thyroid tissue stained with hematoxylin and eosin. The extent of thyroid lymphocytic infiltration was assessed on a score of 0 to 4, as previously reported (5). Briefly, 0 indicates a paucity of lymphocytes; 1+ thyroiditis is defined as an infiltrate of at least 125 cells in one or several foci; 2+ represents 10–20 foci of cellular infiltration, each the size of several follicles and involving up to one fourth of the gland; 3+ indicates that one fourth to half of the gland is infiltrated by lymphocytes; and 4+ indicates that greater than half of the gland is destroyed.
Statistical analyses
Differences between the magnitude of responses were determined by t test or Mann-Whitney rank sum test (as appropriate). Comparison of multiple groups was assessed by ANOVA. Fisher’s exact test was used to test for significant differences between the number of mice positive or negative for a particular parameter.
Results
The following four groups of NOD.H-2h4 mice were studied: mice immunized with TSHR-Ad maintained without iodide or on water containing 0.05% NaI and mice immunized with Con-Ad maintained without iodide or on iodized water. Adenovirus injections were performed on three occasions at 3-wk intervals, and blood was drawn 1 wk after two injections and at euthanasia 4 wk after the final injection. Sera were characterized for induced TSHR antibodies and for spontaneously arising TgAb. Thyroid function was assessed in terms of total T4 levels, TSH, and thyroid pathology.
Induced TSHR antibodies
A high proportion of mice had TBI antibodies after two TSHR-Ad injections, including 14 of 18 mice (78%) on regular water and 13 of 17 mice (67%) on iodized water (Fig. 1A). In contrast, zero (0%) of 11 Con-Ad-immunized mice on iodized water had TBI activity, and only one of 10 mice (10%) on regular water was borderline TBI positive. After three injections, TBI levels remained negative (with one exception) in Con-Ad-immunized mice, whereas 94% of TSHR-Ad-injected animals became TBI positive regardless of whether they were given regular or iodized water (Fig. 1B). Moreover, exposure to iodide had no effect on the levels of TBI activity in TSHR-Ad-immunized animals.
FIG. 1. TSHR antibodies measured by TBI develop in NOD.H-2h4 female mice immunized with TSHR-Ad, regardless of iodide exposure. The following four groups of mice were studied: mice immunized with TSHR-Ad that were maintained on regular drinking water (n = 18) or on water containing NaI (n = 18) and mice immunized with Con-Ad that were maintained on regular water (n = 10) or on water containing NaI (n = 11). Three injections of adenovirus were administered at 3-wk intervals. TBI values were measured in sera obtained 1 wk after two injections (A) and at euthanasia 4 wk after the third injection (B). TBI values are shown for individual mice without iodide in open circles and with iodide in black circles. The shaded area represents the mean ± 2 SD for Con-Ad immunized mice on regular water. *, Values significantly greater than for Con-Ad-immunized mice without iodide (P < 0.05, ANOVA on ranks).
After three injections, sufficient serum was available from approximately half the mice in each group for bioassay of TSAb and TBAb. None of the Con-Ad-injected mice had TSAb activity in their sera, regardless of exposure to iodide. In contrast, the majority of TSHR-Ad-injected mice was TSAb positive (eight of 10 mice on regular water and 10 of 10 mice on iodized water; Fig. 2A). In terms of TBAb activity, seven of 10 TSHR-Ad mice were positive vs. zero of five Con-Ad-immunized mice maintained on regular water (Fig. 3B). For mice on iodized water, five of 10 TSHR-Ad animals were TBAb positive vs. one of five Con-Ad mice. As for TBI, administration of iodide to the TSHR-Ad-immunized mice did not influence the magnitude of TSAb (or TBAb) activity or the proportions of animals positive for TSAb (or TBAb).
FIG. 2. TSAb and TBAb activity in NOD.H-2h4 mice immunized with TSHR-Ad. As described for Fig. 1, mice immunized with TSHR-Ad were maintained on regular water (n = 10) or NaI-supplemented water (n = 10), and mice immunized with Con-Ad were maintained on regular water (n = 5) or NaI-supplemented water (n = 5). TSAb (A) and TBAb (B) activities were measured at euthanasia 4 wk after the third injection. Data are shown for individual animals without iodide in open circles and with iodide in black circles. The dotted line represents the upper limit of TSAb or TBAb values (mean + 2 SD) in Con-Ad-immunized animals on regular water. *, Values significantly greater than for Con-Ad-immunized mice without iodide (P < 0.05, ANOVA on ranks).
FIG. 3. Autoantibodies to murine Tg arise spontaneously in NOD.H-2h4 mice, and their levels are enhanced by iodide, regardless of immunization with TSHR-Ad (or Con-Ad). Antibody levels, measured 4 wk after the third injection, are shown as ELISA OD410 for individual mice without iodide in open circles and with iodide in black circles. Mice immunized with TSHR-Ad were maintained on regular water (n = 8) or NaI-supplemented water (n = 8), and mice immunized with Con-Ad were maintained on regular water (n = 5) or NaI-supplemented water (n = 5). IgG class TgAb were assessed by measuring serum IgG binding to ELISA wells coated with murine Tg or, as a control, with ovalbumin (OVA) (A). TgAb of subclasses IgG1 and IgG2b were measured using ELISA wells coated with murine Tg and subclass-specific antisera (B). Significance of the indicated differences is as follows: A: *, P = 0.011; and **, P = 0.002 (t tests); B: #, P = 0.038; and ##, P = 0.008 (rank sum tests); and *, P = 0.023 (t test).
Spontaneous TgAb
Unlike TSHR antibodies that were generated in response to TSHR-Ad immunization, TgAb developed in NOD.H-2h4 mice regardless of the type of adenovirus injected. In particular, mice injected with either Con-Ad or TSHR-Ad had detectable IgG class antibody binding to mouse Tg-coated ELISA wells, but not to wells coated with ovalbumin (Fig. 3A). Moreover, TgAb levels were higher in mice on iodized water, irrespective of immunization with Con-Ad or TSHR-Ad. Furthermore, TgAb of subclass IgG1 and, to a lesser extent, subclass IgG2b were higher in mice exposed to iodide than in mice on regular water (Fig. 3B). Overall, TgAb levels were directly related to iodide exposure but were not influenced by the type of adenovirus used for immunization.
Thyroid function
Mice on regular water immunized with Con-Ad had normal serum T4 levels when measured after the second (Fig. 4A) and third (Fig. 4B) injections, with a single exception (one mouse after the third injection). Chronic administration of iodide to Con-Ad-immunized mice tended to reduce T4 levels, reaching significance after the third injection (Fig. 4, A and B).
FIG. 4. Serum T4 levels in NOD.H-2h4 mice immunized with TSHR-Ad or Con-Ad and maintained on regular drinking water or on iodized water. Serum T4 levels were measured 1 wk after two injections and at euthanasia 4 wk after three injections. Data shown are T4 values for individual mice immunized with TSHR-Ad and maintained on regular drinking water (n = 18; open circles) or on NaI-supplemented water (n = 18; black circles) and mice immunized with Con-Ad and maintained on regular water (n = 10; open circles) or NaI-supplemented water (n = 11; black circles). The shaded area represents the mean ± 2 SD for T4 levels in Con-Ad-immunized mice on regular water. Mice with T4 levels above these values are considered hyperthyroid. T4 levels significantly different between the indicated groups are indicated by the following: A: **, P = 0.002 (rank sum test); B: #, P = 0.022 (rank sum test); and *, P = 0.036 (t test).
The effect of iodide in reducing serum T4 levels was also observed in the TSHR-Ad-immunized mice. Thus, after the second injection, five of 18 mice on regular water but only one of 18 mice on iodized water had elevated serum T4 levels (Fig. 4A). The same pattern was seen after the third TSHR-Ad injection (five of 18 and one of 17 mice were hyperthyroid on regular or iodized water, respectively; Fig. 4B).
Although serum was available for TSH analysis on only a limited number of animals, these data provided additional insight into the thyroid status of the NOD.H-2h4 mice. TSH levels were only elevated in some animals exposed to iodide (Fig. 5). It should be appreciated that the mouse TSH assay lacks the sensitivity of present human TSH assays. Nevertheless, the only sera with TSH levels less than 10 mU/liter were from mice immunized with TSHR-Ad on regular water.
FIG. 5. TSH levels are elevated in some NOD.H-2h4 mice exposed to iodide and reduced in some animals immunized with TSHR-Ad and maintained on regular water. Values (mU/liter) are shown for individual mice immunized with TSHR-Ad and maintained on regular drinking water (n = 9; open circles) or NaI-supplemented water (n = 10; black circles) and mice immunized with Con-Ad on regular water (n = 4; open circles) or NaI-supplemented water (n = 5; black circles). The solid line indicates the upper limit (mean + 2 SD) of TSH in Con-Ad-immunized mice on regular water. The lower limit of the assay is less than 10 mU/liter TSH (dashed line). *, P < 0.001 (rank sum test) for TSH levels significantly different between the indicated groups.
Thyroid pathology
Thyroid pathology was characterized in terms of lymphocytic infiltration (thyroiditis) and thyrocyte morphology. The thyroiditis score was related to iodide exposure and was significantly higher in TSHR-Ad-immunized mice exposed to iodide than mice immunized in the same way on regular water (Fig. 6, P = 0.026). Similar changes occurred in Con-Ad-immunized animals, but the difference was not statistically significant. Two Con-Ad-immunized mice on regular water were atypical in that they had much greater lymphocytic infiltrates than previously reported for nonimmunized mice on normal iodide intake (5).
FIG. 6. Thyroid pathology in NOD.H-2h4 mice immunized with TSHR-Ad or Con-Ad and maintained on regular water or NaI-supplemented water. The mean thyroiditis score (+ SEM) is shown for mice immunized with TSHR-Ad on regular drinking water (n = 17) or NaI-supplemented water (n = 18) and for mice immunized with Con-Ad on regular water (n = 10) or NaI-supplemented water (n = 10). *, Values significantly different between the indicated groups (P < 0.026, rank sum test).
After Con-Ad immunization, examples are shown in Fig. 7 of minimal thyroid lymphocytic infiltration in mice on regular water (Fig. 7A, grade 0+) vs. iodide-supplemented water (Fig. 7B, grade 2+ to 3+). After TSHR-Ad immunization, examples of minimal thyroiditis in a mouse on regular water (grade 0+) and moderate thyroiditis (grade 2+) in a mouse on iodized water are shown in Fig. 7, C and D, respectively. At higher magnification, thyroid epithelial cells appeared to be slightly flatter in euthyroid animals (Fig. 7E) than the more cuboidal cells observed in some hyperthyroid mice (Fig. 7F).
FIG. 7. Thyroid histology in NOD.H-2h4 mice immunized with TSHR-Ad or Con-Ad. Sections were stained with hematoxylin and eosin. A and B, Con-Ad-immunized mouse on regular water (A) or iodized water (B). Magnification, x100. C and D, Hyperthyroid TSHR-Ad-immunized mouse on regular water (C) and euthyroid mouse on iodized water (D). Magnification, x100. E and F, Euthyroid mouse on iodized water (E) and hyperthyroid TSHR-Ad immunized mouse on regular water (F). Magnification, x400.
Discussion
NOD.H-2h4 mice are a particularly valuable strain for investigating autoimmune thyroid disease because they do not develop diabetes but have an increased incidence of spontaneous thyroiditis compared with diabetes-prone NOD mice (23). Moreover, it has previously been shown that iodide supplementation markedly enhances the incidence of TgAb and the incidence and severity of thyroiditis in NOD.H-2h4 mice (4, 5, 6). In these mice, TPO is not a major autoantigen. TPO-induced T-cell responses have been observed in only one study (24), and TPO autoantibodies have been sought but not found (4, 24). Tg is the major autoantigen in this thyroiditis-prone mouse strain (4, 5, 6, 24).
In the present study, we explored the net effect on thyroid status in NOD.H-2h4 mice of three different factors, namely spontaneous thyroid autoimmunity, induced TSHR antibodies, and dietary iodide. TgAb levels are increased by iodide ingestion in both TSHR-Ad- and Con-Ad-immunized animals. Similarly, low-grade thyroiditis (grades 0 to 1+) is exacerbated by iodide (typically grades 2+ to 3+). Moreover, contrary to our hypothesis, inducing an immune response to the TSHR by TSHR-Ad immunization had no effect on thyroiditis. As anticipated, immunization with TSHR-Ad, but not with Con-Ad, induced TSHR antibodies as measured by TBI, TSAb, and TBAb activities. Again, iodide exposure had no statistically significant effect on these antibody levels. Consequently, iodide and TSHR immunization had independent effects on thyroiditis/TgAb vs. TSHR antibodies, respectively. Turning to thyroid function, two major effects were observed. First, iodide reduced serum T4 levels in all groups of mice, whether immunized with Con-Ad or TSHR-Ad. Although reduced, T4 levels in most mice remained within the normal range. Second, TSHR-Ad immunization induced thyrotoxicosis in approximately 30% of mice on a normal iodide diet. Taken together, the suppressive effect of iodide abrogated the thyroid stimulation induced by TSHR-Ad immunization.
Incidentally, although thyroid histology has been extensively studied in NOD mice including the NOD.H-2h4 strain, there are only two reports describing the effect of a high-iodide diet on serum T4 levels in these mice. In NOD mice, iodide administration modestly decreased serum T4 levels (25), whereas in the NOD.H-2h4 strain, iodide had no effect (4). Our present data demonstrating that iodide consistently reduced serum T4 levels are the first to show this effect in the NOD.H-2h4 strain.
What is the mechanism by which chronic ingestion of a high-iodine diet reduces serum T4 levels? There are three possible mechanisms. First, the greater degree of thyroiditis induced by iodide could reduce the capacity of the thyroid gland to synthesize and secrete thyroid hormone. During progressive thyroid failure in Hashimoto’s thyroiditis, as serum T4 levels decline, increased TSH drive prolongs euthyroidism by increasing secretion of T3 (not measured in our study). We consider this mechanism possible but unlikely because the degree of thyroid destruction was limited. In normal rats, serum T4 levels have returned to normal by 5 wk after hemithyroidectomy (26). Increasing the severity of thyroiditis by administering iodide to BB/W rats does not reduce serum T4 levels (27) unless the animals have previously undergone hemithyroidectomy (28). However, it is possible that thyrocyte function was diminished by cytokines derived from focal nondestructive infiltrates. A second possible mechanism, failure to escape from the Wolff-Chaikoff block of iodide organification (29), is also unlikely because this process leads to a vicious cycle in which overt hypothyroidism ensues (30). The third possible mechanism, which we consider most likely, is a postulated mild suppression of thyroid hormone secretion, with thyroid hormone levels generally remaining in the normal range, which is an effect that has been demonstrated in humans (31).
Several other points of interest arise from these studies. First, TSHR-Ad immunization of NOD.H-2h4 mice clearly induced an immune response to the TSHR, yet it did not increase the severity of thyroiditis. One possible explanation is that we immunized with human, not murine, TSHR A-subunit, thereby precluding homing of TSHR-specific T lymphocytes to the thyroid. This explanation seems unlikely because conventional immunization with nonmurine thyroid antigens induces severe thyroiditis. For example, porcine Tg in complete Freund’s adjuvant induced extensive thyroiditis in NOD.H-k mice (23). Despite the well-known immunogenicity of adenovirus (32), its adjuvant properties are clearly much less effective than mycobacteria for inducing immune responses leading to tissue infiltration. Also, it is unlikely that the inability of TSHR adenovirus to induce thyroiditis is attributable to the lack of a Th1 response. TSHR adenovirus immunization does induce T helper 1 (Th1)-type responses (production of interferon- but not IL-4) (33, 34). It is possible that adenovirus immunization does not induce TGF? production, which is critical for thyroiditis development in NOD.H-2h4 mice (35).
Another point of interest is that our data differ with those from a previous investigation of the NOD strain in which TSHR immunization did not induce functional TSHR antibodies (36). In the earlier study, mice injected with TSHR bacterial fusion protein plus adjuvant developed TSHR antibodies detected by ELISA, but these antibodies lacked TBI, TSAb, or TBAb activity. This lack of functional antibody production is not unexpected considering the many failed attempts to induce Graves’ disease using different TSHR proteins and different adjuvants in different mouse strains (reviewed in Ref. 7). However, the same TSHR fusion protein-adjuvant combination induced some functional antibody activities (TBI and TBAb) in BALB/c mice (37).
These earlier data could imply that BALB/c mice are more susceptible than NOD mice to the induction of functional TSHR antibodies. However, our data indicate that TBI, TSAb, and TBAb activities are readily induced in NOD.H-2h4 mice by a different immunization protocol, namely by injecting TSHR adenovirus. Nevertheless, the NOD strain of mice is more resistant than BALB/c mice (12) to developing hyperthyroidism, particularly using low doses of adenovirus encoding the TSHR A-subunit (18, 19). That is, TSAbs appear to be relatively ineffective in the NOD.H-2h4 strain. Similar resistance to TSAb activity has been observed in other strains including C57BL/6 mice (12, 38). BALB/k mice have the same MHC class II (I-Ak) as NOD.H-2h4 mice, yet they are as susceptible to hyperthyroidism as BALB/c mice with a different MHC class II (I-Ad) (39). Consequently, the non-MHC background of the NOD strain is likely to play a role in its resistance to hyperthyroidism, despite the presence of TSAb.
Incidentally, in our previous study on NOD.H-2h4 mice involving TSHR-DNA plasmid (not adenovirus) vaccination (40), we were surprised that TgAb levels were undetectable, and minimal thyroiditis developed on a high-iodide diet. The reason for the marked difference between these two outcomes was only understood after completion of the DNA plasmid study. Taconic Farms (Germantown, NY), the source of the NOD.H-2h4 mice for the DNA plasmid study, notified investigators that they had provided a re-derived strain that, on later testing in their facility, confirmed our finding of the absence of thyroiditis on iodide treatment. Our present study used the original strain (see Materials and Methods).
In summary, we used NOD.H-2h4 mice to explore the net effects on thyroid status of spontaneous thyroid autoimmunity, induced TSHR antibodies, and iodide administration. Thyroiditis and TgAb were regulated by iodide independent of TSHR immunization. In contrast, TSHR antibodies induced by injecting TSHR-Ad were unaffected by dietary iodide. Nevertheless, the hyperthyroidism that developed in approximately one third of TSHR-Ad-injected mice was suppressed by a high-iodide diet, probably by a nonimmune mechanism (iodine autoregulation of thyroid function, as already mentioned). Finally, as reported for other mouse strains (12, 18), inducing immune responses to the TSHR using TSHR-Ad immunization is not associated with thyroid lymphocytic infiltration in thyroiditis-prone NOD.H-2h4 mice. In addition to hyperthyroidism, Graves’ disease typically has a thyroiditis component, as evident by the common presence of TPO autoantibodies. Caution should always be used in extrapolating from animal models to human disease. However, our studies in NOD.H-2h4 mice raise the possibility that, at least in some individuals, the variable thyroiditis component in Graves’ disease may be related to an antigen other than the TSHR.
Acknowledgments
We thank Dr. L. Wicker (Merck Laboratories, Rahway, NJ) for providing the breeding stock of NOD.H-2h4 mice. In addition, we are grateful for contributions by Dr. Boris Catz (Los Angeles, CA).
References
Wick G, Most J, Schauenstein K, Kromer G, Dietrich H, Ziemicki A, Fassler R, Schwarz S, Neu N, Hala K 1985 Spontaneous autoimmune thyroiditis: a bird’s eye view. Immunol Today 6:359–365
Colle E, Guttmann RD, Seemayer TA 1985 Association of spontaneous thyroiditis with the major histocompatibility complex of the rat. Endocrinology 116:1243–1247
Bernard NF, Ertug F, Margolese H 1992 High incidence of thyroiditis and anti-thyroid autoantibodies in NOD mice. Diabetes 41:40–46
Rasooly L, Burek CL, Rose NR 1996 Iodine-induced autoimmune thyroiditis in NOD-H2h4 mice. Clin Immunol Immunopathol 81:287–292
Braley-Mullen H, Sharp GC, Medling B, Tang H 1999 Spontaneous autoimmune thyroiditis in NOD.H-2h4 mice. J Autoimmun 12:157–165
Hutchings PR, Verma S, Phillips JM, Harach SZ, Howlett S, Cooke A 1999 Both CD4(+) T cells and CD8(+) T cells are required for iodine accelerated thyroiditis in NOD mice. Cell Immunol 192:113–121
Rapoport B, Chazenbalk GD, Jaume JC, McLachlan SM 1998 The thyrotropin receptor: interaction with thyrotropin and autoantibodies. Endocr Rev 19:673–716
Shimojo N, Kohno Y, Yamaguchi K-I, Kikuoka S-I, Hoshioka A, Niimi H, Hirai A, Tamura Y, Saito Y, Kohn LD, Tahara K 1996 Induction of Graves-like disease in mice by immunization with fibroblasts transfected with the thyrotropin receptor and a class II molecule. Proc Natl Acad Sci USA 93:11074–11079
Kaithamana S, Fan J, Osuga Y, Liang SG, Prabhakar BS 1999 Induction of experimental autoimmune Graves’ disease in BALB/c mice. J Immunol 163:5157–5164
Kita-Furuyama M, Nagayama Y, Pichurin P, McLachlan SM, Rapoport B, Eguchi K 2003 Dendritic cells infected with adenovirus expressing the thyrotropin receptor induce Graves’ hyperthyroidism in BALB/c mice. Clin Exp Immunol 131:234–240
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 105:803–811
Nagayama Y, Kita-Furuyama M, Ando T, Nakao K, Mizuguchi H, Hayakawa T, Eguchi K, Niwa M 2002 A novel murine model of Graves’ hyperthyroidism with intramuscular injection of adenovirus expressing the thyrotropin receptor. J Immunol 168:2789–2794
McLachlan SM, Rapoport B 2004 Why measure thyroglobulin autoantibodies rather than thyroid peroxidase autoantibodies? Thyroid 14:510–520
Yoshida H, Amino N, Yagawa K, Uemura K, Satoh M, Miyai K, Kumahara Y 1978 Association of serum antithyroid antibodies with lymphocytic infiltration of the thyroid gland: studies of seventy autopsied cases. J Clin Endocrinol Metab 46:859–862
Paschke R, Vogg M, Swillens S, Usadel KH 1993 Correlation of microsomal antibodies with the intensity of the intrathyroidal autoimmune process in Graves’ disease. J Clin Endocrinol Metab 77:939–943
Takasu N, Yamada T, Sato A, Nakagawa M, Komiya I, Nagasawa Y, Asawa T 1990 Graves’ disease following hypothyroidism due to Hashimoto’s disease: studies of eight cases. Clin Endocrinol (Oxf) 33:687–698
Tamai H, Kasagi K, Takaichi Y, Takamatsu J, Komaki G, Matsubayashi S, Konishi J, Kuma K, Kumagai LF, Nagataki S 1989 Development of spontaneous hypothyroidism in patients with Graves’ disease treated with antithyroidal drugs: clinical, immunological, and histological findings in 26 patients. J Clin Endocrinol Metab 69:49–53
Chen C-R, Pichurin P, Nagayama Y, Latrofa F, Rapoport B, McLachlan SM 2003 The thyrotropin receptor autoantigen in Graves’ disease is the culprit as well as the victim. J Clin Invest 111:1897–1904
Chen C-R, Pichurin P, Chazenbalk GD, Aliesky H, Nagayama Y, McLachlan SM, Rapoport B 2004 Low-dose immunization with adenovirus expressing the thyroid-stimulating hormone receptor A-subunit deviates the antibody response toward that of autoantibodies in human Graves’ disease. Endocrinology 145:228–233
Mittereder N, March KL, Trapnell BC 1996 Evaluation of the concentration and bioactivity of adenovirus vectors for gene therapy. J Virol 70:7498–7509
Braley-Mullen H, Johnson M, Sharp GC, Kyriakos M 1985 Induction of experimental autoimmune thyroiditis in mice with in vitro activated splenic T cells. Cell Immunol 93:132–143
Pohlenz J, Maqueem A, Cua K, Weiss RE, Van Sande J, Refetoff S 1999 Improved radioimmunoassay for measurement of mouse thyrotropin in serum: strain differences in thyrotropin concentration and thyrotroph sensitivity to thyroid hormone. Thyroid 9:1265–1271
Damotte D, Colomb E, Cailleau C, Brousse N, Charreire J, Carnaud C 1997 Analysis of susceptibility of NOD mice to spontaneous and experimentally induced thyroiditis. Eur J Immunol 27:2854–2862
Verma S, Hutchings P, Guo J, McLachlan S, Rapoport B, Cooke A 2000 Role of MHC class I expression and CD8(+) T cells in the evolution of iodine-induced thyroiditis in NOD-H2(h4) and NOD mice. Eur J Immunol 30:1191–1202
Tani J, Mori K, Hoshikawa S, Nakazawa T, Satoh J, Nakagawa Y, Ito S, Yoshida K 2002 Prevention of lymphocytic thyroiditis in iodide-treated non-obese diabetic mice lacking interferon regulatory factor-1. Eur J Endocrinol 147:809–814
Clark OH, Lambert WR, Cavalieri RR, Rapoport B, Hammond ME, Ingbar SH 1976 Compensatory thyroid hypertrophy after hemithyroidectomy in rats. Endocrinology 99:988–995
Allen EM, Appel MC, Braverman LE 1986 The effect of iodide ingestion on the development of spontaneous lymphocytic thyroiditis in the diabetes-prone BB/W rat. Endocrinology 118:1977–1981
Allen EM, Appel MC, Braverman LE 1987 Iodine-induced thyroiditis and hypothyroidism in the hemithyroidectomized BB/W rat. Endocrinology 121:481–485
Ingbar SH 1972 Autoregulation of the thyroid. Response to iodide excess and depletion. Mayo Clin Proc 47:814–823
Braverman LE, Ingbar SH, Vagenakis AG, Adams L, Maloof F 1971 Enhanced susceptibility to iodide myxedema in patients with Hashimoto’s thyroiditis. J Clin Endocrinol Metab 32:515–521
Vagenakis AG, Downs P, Braverman LE, Burger A, Ingbar SH 1973 Control of thyroid hormone secretion in normal subjects receiving iodides. J Clin Invest 52:528–532
Geutskens SB, van der Eb MM, Plomp AC, Jonges LE, Cramer SJ, Ensink NG, Kuppen PJ, Hoeben RC 2000 Recombinant adenoviral vectors have adjuvant activity and stimulate T cell responses against tumor cells. Gene Ther 7:1410–1416
Pichurin PN, Chen C-R, Nagayama Y, Pichurina O, Rapoport B, McLachlan SM 2004 Evidence that factors other than particular thyrotropin receptor T cell epitopes contribute to the development of hyperthyroidism in murine Graves’ disease. Clin Exp Immunol 135:391–397
Nagayama Y, Mizuguchi H, Hayakawa T, Niwa M, McLachlan SM, Rapoport B 2003 Prevention of autoantibody-mediated Graves’-like hyperthyroidism in mice with IL-4, a Th2 cytokine. J Immunol 170:3522–3527
Braley-Mullen H, Chen K, Wei Y, Yu S 2001 Role of TGF? in development of spontaneous autoimmune thyroiditis in NOD.H-2h4 mice. J Immunol 167:7111–7118
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 115:199–206
Costagliola S, Many M-C, Stalmans-Falys M, Tonacchera M, Vassart G, Ludgate M 1994 Recombinant thyrotropin receptor and the induction of autoimmune thyroid disease in BALB/c mice: a new animal model. Endocrinology 135:2150–2159
Chen CR, Aliesky H, Pichurin PN, Nagayama Y, McLachlan SM, Rapoport B 2004 Susceptibility rather than resistance to hyperthyroidism is dominant in a thyrotropin receptor adenovirus-induced animal model of Graves’ disease as revealed by BALB/c-C57BL/6 hybrid mice. Endocrinology 145:4927–4933
Nagayama Y, McLachlan SM, Rapoport B, Niwa M 2003 A major role for non-MHC genes, but not for micro-organisms, in a novel model of Graves’ hyperthyroidism. Thyroid 13:233–238
Pichurin P, Schwarz-Lauer L, Braley-Mullen H, Paras C, Pichurina O, Morris JC, Rapoport B, McLachlan SM 2002 Peptide scanning for thyrotropin receptor T-cell epitopes in mice vaccinated with naked DNA. Thyroid 12:755–764(Sandra M. McLachlan, Hele)