The pathogenic role of anti-thyroglobulin antibody on pregnancy: evidence from an active immunization model in mice

Shelly Tartakover Matalon1,2, Miri Blank1, Yair Levy1, Howard J.A. Carp1, Ayala Arad3, Lynne Burek4, Eyal Grunebaum1, Yaniv Sherer1, Asher Ornoy2, Samuel Refetoff5,6, Roy E. Weiss5, Noel R. Rose4 and Yehuda Shoenfeld1,7

1 Center for Autoimmune Diseases, Department of Medicine ‘B’, Sheba Medical Center, Tel Hashomer & Sackler Faculty of Medicine, Tel-Aviv, 2 Department of Anatomy, Hadassah Faculty of Medicine, Hebrew University, Jerusalem, 3 Department of Pathology, Rambam Medical Center, Technion University, Haifa, Israel, 4 Departments of Pathology and of Molecular Microbiology and Immunology, Johns Hopkins Medical Institutions, Baltimore, 5 Department of Medicine, and 6 Department of Pediatrics, University of Chicago, Chicago, Illinois, USA

7 To whom correspondence should be addressed at: Department of Medicine ‘B’, Sheba Medical Center, Tel Hashomer, 52621, Israel. e-mail: shoenfel{at}post.tau.ac.il


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: The presence of antibodies to thyroglobulin (Tg) is associated with fetal loss even in the absence of thyroid dysfunction. The aim of this study was to examine whether active immunization with Tg could elicit anti-Tg autoantibodies and reproductive failure without interfering with thyroid function. METHODS: BALB/c mice that were immunized with human Tg in complete Freund’s adjuvant (CFA) or injected with only CFA were studied for the development of antibodies to Tg, T4, dsDNA, ssDNA and cardiolipin. Total T4, free T4 and thyroid-stimulating hormone (TSH) levels were also assessed before and during pregnancy. Percentages of resorbed fetuses (the equivalent to human missed abortion) were compared and autoantibody presence on the placentae and fetuses was examined. RESULTS: Following immunization, high levels of anti-Tg were observed in mice immunized with Tg, compared with mice injected with CFA [0.83 ± 0.23 versus 0.012 ± 0.016 respectively; mean ± SD optical density (OD) at 405 nm; P < 0.001]. The specificity of binding to Tg was confirmed by competition assay. Although total T4 levels were increased in comparison with control mice, this was associated with the presence of antibodies to T4. Indeed, free T4 levels and TSH were similar to control mice. Mice were killed after 14 days of pregnancy. The thyroid function and the histology of the thyroid glands were normal. Increased fetal wastage was found among the Tg-immunized mice compared with the CFA-injected mice (P = 0.04), with lower fetal and placental weights (fetal weights: 194 ± 4 mg versus 240 ± 6 mg; placental weights: 105 ± 2 mg versus 130 ± 3; P < 0.001 for both). Antibodies to Tg were demonstrated only on the placentae of Tg-immunized mice. CONCLUSION: Immunization with Tg results in the production of Tg antibodies and fetal resorption. These effects occur in the absence of thyroid dysfunction.

Key words: animal model/autoimmunity/fetal loss/thyroglobulin/thyroid


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Many different conditions may cause pregnancy loss, such as chromosomal abnormalities, anatomical malformations and autoimmune diseases including systemic lupus erythematosus or the antiphospholipid syndrome (Branch, 1990Go; Cowchock, 1991Go). Abnormal hormone level such as hypothyroidism or hyperthyroidism can also cause fetal demise (Gardiner-Hill, 1929Go; Hamburger and Stoffer, 1981Go; Davis et al., 1988Go), and fetal loss has been demonstrated in mice with thyroiditis (Imaizumi et al., 2001Go). Several studies have also found an association between spontaneous abortions and autoantibodies to the thyroid gland, such as thyroid peroxidase or thyroglobulin (Tg) (Davis et al., 1988Go). As thyroid dysfunction is not always detected in the presence of thyroid abnormalities (Lejeune et al., 1993Go; Pratt et al., 1993aGo;b; Geva et al., 1997Go), the higher miscarriage rate in women with autoimmune thyroiditis could be due to the autoantibodies rather than an impaired hormonal state. However, as other studies have not demonstrated an association between the presence of thyroid antibodies and pregnancy loss (Esplin et al., 1998Go; Rushworth et al., 2000Go), the direct role of thyroid autoantibodies in fetal loss is debatable. In order to study the specific role of autoantibodies to the thyroid in pregnancy loss, we utilized a murine model of active immunization with human Tg. The aim of the study was to induce the production of Tg autoantibodies without thyroid hormone abnormalities or histological evidence of thyroiditis, in order to study the effects on pregnancy outcome.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Active immunization with thyroglobulin
BALB/c female mice (8–10 weeks old) were purchased from the Sackler Faculty of Medicine, Tel-Aviv University, Israel. Human Tg (10 µg) (Biogenesis, Poole, UK) was emulsified in complete Freund’s adjuvant (CFA) and injected intradermally into the hind foot-pads of 27 mice. The control group included 29 mice that were injected with CFA alone. The mice were bled 3 weeks after immunization and prior to mating. Both mouse groups were mated overnight, and the presence of a vaginal mucus plug the following morning was taken as an evidence of mating, and designated as day 0 of pregnancy. Mice were bled on the 14th day of pregnancy and killed. All animal procedures were performed according to protocols approved by the Institutional Animal Care and Use Committees of the Sheba Medical Center and Tel-Aviv University.

Screening of mice sera for autoantibodies
Anti-Tg
ELISA plates were coated with human Tg (10 µg/ml in carbonate buffer) and blocked with 5% BSA. Following extensive washing, mouse sera [diluted 1:200 in tris-buffered saline (TBS)] were added to the wells. Bound antibodies were detected using anti-mouse IgG rabbit serum conjugated to alkaline phosphatase (Jackson Immunoresearch 115–055–146) and the intensity of the colour in the plates was read in a Titertrek ELISA reader using a 405 nm filter and a 620 nm reference filter.

Anti-dsDNA, anti-ssDNA and anti-cardiolipin
ELISA plates were coated with poly-L-lysin, calf thymus DNA (250 ng/ml in TBS) and cardiolipin (50 µg/ml in ethanol), for the detection of double stranded DNA (dsDNA), single stranded DNA (ssDNA) and cardiolipin (Cl) respectively, as previously described (Blank et al., 2002Go). Mouse sera (diluted 1:200 in TBS) were added to the wells and bound antibodies were detected as described above.

Anti-thyroxine (T4)
T4 antibodies were tested prior to pregnancy in 10 mice injected with CFA and 12 mice immunized with Tg, and during pregnancy in 18 mice injected with CFA and 19 immunized with Tg (all of which were randomly chosen). The percentage of T4 bound to IgG was measured. Serum (20 µl) was incubated with 15 000 cpm of 125I-thyroxine in 30 µl of barbital buffer for 6 h at 20°C. Bound 125I-thyroxine was precipitated with 450 µl of 20% PEG 6000 at 4°C for 30 min. The amount of labelled thyroxine in the pellet was quantified and is expressed as the percentage of the total.

Thyroid function tests
T4 levels were determined 3 weeks after injection/immunization and also immediately prior to killing. Total T4 levels were measured in each mouse from blood spotted on filter paper using the neonatal kit (Diagnostic Products Corporation, Los Angeles, USA) according to the manufacturer’s instructions. Free T4 was measured by equilibrium dialysis using a kit (Quest Diagnostics, Nichols Institute, San Juan Capistrano, USA). Pools of sera with antibodies to T4 >15% (range 15.8–55.0) and <2% (range 0.6–1.8) were prepared. Serum TSH was measured in 50 µl of serum using a sensitive, heterologous, disequilibrium, double antibody precipitation RIA as described elsewhere (Pohlenz et al., 1999Go). The sensitivity of this assay was 5–10 mIU/l. The intra- and inter-assay coefficients of variations were 16 and 27% at 20 mIU/l, 6.3 and 8.2% at 200 mIU/l, 5.4 and 9.8% at 850 mIU/l and 10 and 24% at 2000 mIU/l respectively. Thyroid-stimulating hormone (TSH) levels were tested prior to pregnancy in 12 mice injected with CFA and 14 immunized with Tg, and also in 18 randomly selected pregnant mice from both groups.

Histology of the thyroid gland
Thyroid glands of 10 Tg-immunized and 10 CFA-injected mice were excised after killing. The thyroids were fixed in 4% paraffin blocks, cut into 5 µm thick slices and stained with haematoxylin and eosin. Three slices were taken from each gland, the slices contained three or more sections separated by >50 µm apart, and these were evaluated by an observer unfamiliar with the study protocol and the different mouse groups.

Competition of binding to anti-Tg antibodies
Sera of three Tg immunized mice were pooled, and the dilution of the sera giving 50% of maximal binding to Tg-coated plates (prepared as described above) was employed for the competition assay. Different concentrations of Tg (100, 50, 25, 12, 6 and 3 µg/ml) were preincubated with the diluted sera for 1 h at room temperature, or overnight at 4°C. The mixtures of Tg and sera were further incubated for 2 h on the Tg-coated plates. The assay was continued as described for Tg ELISA.

Evaluation of pregnancy outcome
The uterine horns were examined to determine the number of live fetuses and number of resorbed pregnancies in order to calculate the percentage index of fetal resorption. Fetal resorption was identified as growth arrest and regression into an oval-shaped mass smaller than expected for the 14th day of pregnancy (with a size similar to 7–8 days pregnancy). Placental and fetal weights of randomly chosen mice (15 Tg and 15 CFA immunized mice) were determined. The person who selected the mice from both groups was blinded regarding the treatment groups.

Elution of antibodies from the placentae and the embryos
Six placentae and embryos were separated, hand-homogenized and washed five times with 50 ml phosphate buffered saline (PBS). The last wash was retained in order to confirm that no IgG was found in them. Immunoglobulins were eluted from the washed homogenates with glycine HCl buffer (0.1 mol/l pH 2.7, 15 min, room temperature) that were later neutralized with TBS. Following centrifugation, the supernatants were separated and dialyzed against maltose. The supernatants were analysed for the presence of antibodies to Tg and Cl.

Statistical analysis
Data were analysed by Student’s t-test and the {chi}2-test. P < 0.05 was considered statistically significant.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Autoantibody production in mice immunized with Tg
Mice immunized with human Tg developed high levels of anti-Tg autoantibodies (Figure 1). The levels of anti-Tg on day 14 of pregnancy were 0.83 ± 0.23 (mean ± SD OD at 405 nm) in the mice immunized with human Tg, compared with 0.012 ± 0.016 in the mice injected with CFA (P < 0.001). No mice from either group developed high levels of antibodies to dsDNA, ssDNA or cardiolipin.



View larger version (10K):
[in this window]
[in a new window]
 
Figure 1. Antibodies to thyroglobulin, ssDNA, dsDNA and cardiolipin in mice immunized with Tg. Antibody levels were determined on the 14th day of pregnancy in mice immunized with thyroglobulin (Tg) and compared to those in mice injected with complete Freund’s adjuvant (CFA). Antibody levels are expressed as optical density (OD) at 405 nm. The horizontal solid lines represent the average. T = mice immunized with Tg. C = mice immunized with CFA. anti-Tg = antibodies to thyroglobulin. ssDNA = antibodies to single strand DNA. dsDNA = antibodies to double strand DNA. Cl = antibodies to cardiolipin.

 
Inhibitory effect of Tg on binding of anti-Tg
The specificity of anti-Tg autoantibodies to Tg was demonstrated by the dose-dependent inhibition of binding of anti-Tg to Tg-coated plates by the human Tg (Figure 2; three separate experiments). Whereas 50 µg/ml of ovalbumin resulted in 20% inhibition of anti-Tg binding to Tg, the same concentration of Tg resulted in 80% inhibition of anti-Tg binding to Tg.



View larger version (16K):
[in this window]
[in a new window]
 
Figure 2. Inhibition of binding of Tg antibodies to Tg-coated plates. Sera at a concentration giving 50% of maximal binding to thyroglobulin (Tg) were incubated with increasing concentrations of human Tg or ovalbumin (Ova). Binding was evaluated as described in Figure 1. anti-Tg = antibodies to Tg.

 
Thyroid function and histology
Tg immunization was associated with increased serum T4 levels compared with mice injected with CFA (mean ± 2 SD: 6.5 ± 3.6 µg/dl versus 5.1 ± 1.2 µg/dl respectively, P = 0.03). During pregnancy, serum T4 levels decreased in both groups (Table I), but remained higher in the Tg-immunized group than in the controls (4.7 ± 3.0 µg/dl versus 2.3 ± 0.5 µg/ml respectively, P < 0.001). Tg immunization was associated with increased binding of T4 to IgG in non-pregnant (8.5 ± 2.3%) and pregnant mice (20.3 ± 3.3%), while this was very low in CFA-injected animals, which suggests that the presence of the elevated serum T4 may be due to anti-T4. Indeed, when free T4 was assessed in pooled sera from mice with a high percentage of T4 bound to IgG (>15%, range 15.8–55.0) and in those with a low percentage of T4 bound to IgG (<2%; range 0.6–1.8), no significant difference was found (8.8 ng/ml versus 8.1 ng/ml respectively). Further evidence that the Tg-immunized mice were euthyroid was shown in the normal serum TSH levels and there was no significant difference in serum TSH levels between mice immunized with Tg or with CFA (Table I). Histological examination of the thyroids did not reveal any pathological changes in either the mice immunized with Tg or those injected with CFA (Figure 3). There were no lymphocytic infiltrates, which typically develop in experimental thyroiditis.


View this table:
[in this window]
[in a new window]
 
Table I. T4, thyroid-stimulating hormone (TSH) and antibodies to T4 in mice immunized with thyroglobulin (Tg) or complete Freund’s adjuvant (CFA) before and during pregnancy.
 


View larger version (93K):
[in this window]
[in a new window]
 
Figure 3. Thyroid sections from thyroglobulin (Tg)-immunized or complete Freund’s adjuvant (CFA)-injected mice.

 
Pregnancy outcome in Tg immunized mice
Mice were killed on day 14 of pregnancy. There were differences in fetal resorption rate between the Tg-immunized mice and the CFA-injected mice (Table II). The mice were divided into categories according to the fetal resorption rate of each mouse (<10, 10 to <20, 20 to <30, and >30%; Table III), yielding a significant difference between the two groups (P = 0.04).


View this table:
[in this window]
[in a new window]
 
Table II. Fetal resorption rate (%) of mice immunized with thyroglobulin (Tg) or injected with complete Freund’s adjuvant (CFA) alone
 

View this table:
[in this window]
[in a new window]
 
Table III. Fetal resorption rate according to categories
 
There were also lower mean fetal and placental weights in the Tg-immunized mice (Figure 4) compared with the control group (fetal weights: 194 ± 4 mg versus 240 ± 6 mg; placental weights: 105 ± 2 mg versus 130 ± 3 mg; P < 0.001 for both). All of the placentae and fetuses from the 15 mice were measured, there were 105 embryos in the Tg group and 122 embryos in the CFA group and a similar number of placentae.



View larger version (15K):
[in this window]
[in a new window]
 
Figure 4. Embryos and placentae weights of mice immunized with Tg and complete Freund’s adjuvant (CFA). Evaluation of the weights of embryos and placentae on day 14 of pregnancy in 15 mice immunized with thyroglobulin (Tg) or with CFA alone. The horizontal solid lines represent the average weight of each group.

 
Elution of antibodies from placentae and embryos
After separating the placentae from the embryos, anti-Tg autoantibodies were detected in increasing concentrations of the supernatant eluted only from the placentae of the Tg-immunized mice (Figure 5). These autoantibodies were not found on embryos of Tg-immunized mice, nor were they eluted from the placentae or embryos of CFA-injected mice. No other autoantibodies could be detected on the placentae or embryos (data not shown).



View larger version (17K):
[in this window]
[in a new window]
 
Figure 5. Antibodies to thyroglobulin (Tg) on the placentae and embryos. Evaluation for the presence of antibodies to Tg on the placentae or the embryos from mice immunized with Tg or with complete Freund’s adjuvant (CFA).

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This study shows that active immunization of mice with human Tg results in the production of anti-Tg autoantibodies and pregnancy failure manifested by an increased fetal resorption rate (equivalent to human missed abortions). In order to examine the isolated effect of anti-Tg on pregnancy outcome, we utilized BALB/c female mice which have been previously shown to develop high titres of anti-Tg following immunization with Tg, without accompanying thyroiditis or thyroid hormone abnormalities (Vladestine and Rose, 1971Go; Tomer et al., 1996Go). The specificity of anti-Tg was confirmed by competition assays, in which the antibody binding to Tg-coated plates was inhibited by soluble Tg, but not by ovalbumin. Although the mice immunized with Tg developed higher T4 levels than the mice injected with CFA, the increased T4 levels were due to an increased amount of T4-bound to IgG. Indeed, free T4 levels did not differ between mice with variable total T4 levels, and there was no significant difference in serum TSH levels between mice immunized with Tg or with CFA. Furthermore, thyroid histology was completely normal in mice immunized with Tg, indicating that this murine model is suitable to assess the isolated effect of anti-Tg.

Some reports support an association between Tg antibodies and fetal loss (Stagnaro-Green et al., 1990Go; Glinoer et al., 1991Go; Pratt et al., 1993bGo; Takashi et al., 1997Go; Kutteh et al., 1999aGo; Abramson and Stagnaro-Green, 2001Go; Matalon-Tartakover et al., 2001Go); however, a causative role has not been established. Fetal loss among patients with thyroid antibodies could be induced by several putative mechanisms. The most obvious one is thyroid dysfunction, as is commonly seen in Hashimoto’s thyroiditis. However, the increase in miscarriages cannot always be explained by thyroid dysfunction alone (Dendrinos et al., 2000Go). This suggests that the higher rate of miscarriages observed in women with autoimmune thyroid disturbances reflect primarily an autoimmune phenomenon, rather than, or in addition to, a consequence of overt thyroid hormone abnormalities. Increased pregnancy loss is found among patients suffering from different autoimmune diseases, and organ specific antibodies to the thyroid have been found in parallel to non-organ specific autoantibodies (Magaro et al., 1992Go). Therefore, the presence of antibodies to the thyroid could represent a secondary marker of a predisposition for an autoimmune disease rather than the actual cause of fetal loss (Coulman et al., 1999Go; Sherer and Shoenfeld, 1999Go). However, we failed to identify other non-organ specific autoantibodies among the mice that developed anti-Tg and increased fetal resorption rates. Other studies have also reported discordance between the presence of thyroid autoantibodies and non-organ specific autoantibodies (Pratt et al., 1993Go; Bussen and Steck, 1997Go; Mecacci et al., 2000Go), thus excluding polyclonal activation as the cause for thyroid autoantibody production.

Immunization with Tg was associated with lower placental and embryonic weights. The finding of anti-Tg binding to the placenta and not to the embryo-body might suggest that these autoantibodies have a direct effect on the placenta. It is well known that autoantibodies can pass into the amniotic fluids (Cohen et al., 2000Go) and interact with the syncytiotrophoblast and cytotrophoblast (Kutteh et al., 1999bGo). However, it is possible that anti-Tg bound passively to the placenta but exhibited no harmful effect, as many other antibodies and immune complexes do. Moreover, a direct pathogenic effect of anti-Tg antibodies is only a possible explanation for the reproductive failure, as other possibilities which interfere with reproduction can occur following immunization, such as a switch to Th1. Further studies will clarify the precise effect that anti-Tg may exert on placental function, and whether these antibodies have a pathogenic role on pregnancy. These should include, for example, passive transfer of anti-Tg antibodies, and transfer of sera from mice immunized with Tg into naive mice.

In conclusion, mice immunized with Tg developed high titres of Tg antibodies without thyroid dysfunction, had a higher incidence of fetal resorptions and reduced placental and embryo weights. Hence, anti-Tg might exert a direct pathogenic effect on pregnancy outcome in the absence of thyroid dysfunction, but more studies are required in order to confirm this assumption.


    Acknowledgements
 
We are indebted to Neal Scherberg, Xiao Hui Liao and Joseph Sack for their help. This work was supported by the W.Schrieber Foundation, The Sackler Faculty of Medicine, Tel-Aviv University, by grants from the USA Public Health Service Grants DK 58281 (R.E.W.) and DK 15070 (S.R.) and the Seymour J.Abrams Thyroid Research Center. Y.S. is an incumbent of the Laura Schwartz Kipp Chair for Research into autoimmune diseases.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Abramson, J. and Stagnaro-Green, A. (2001) Thyroid antibodies and fetal loss: an evolving story. Thyroid, 11, 57–63.[CrossRef][ISI][Medline]

Blank, M., Krause, I., Fridkin, M., Keller, N., Kopolovic, J., Goldberg, I., Tobar, A. and Shoenfeld, Y. (2002) The infectious origin of the antiphospholipid syndrome: induction by passive transfer of anti-{beta}2GPI antibodies induced by common bacteria. J. Clin. Invest., 109, 797–804.[Abstract/Free Full Text]

Branch, D.W. (1990) Autoimmunity and pregnancy loss. JAMA, 264, 1453–1454.[CrossRef][ISI][Medline]

Bussen, S.S., and Steck, T. (1997) Thyroid antibodies and their relation to antithrombin antibodies, anticardiolipin antibodies and lupus anticoagulant in women with recurrent spontaneous abortions (antithyroid, anticardiolipin and antithrombin autoantibodies and lupus anticoagulant in habitual aborters). Eur. J. Obstet. Gynecol. Reprod. Biol., 74, 139–143.[CrossRef][ISI][Medline]

Cohen, S.B., Goldenberg, M., Rabinovici, J., Lidor, A.L., Dulitzky, M., Gilburd, B., Shoenfeld, Y., and Sciff, E. (2000) Anti-cardiolipin antibodies in fetal blood and amniotic fluid derived from patients with anti-phospholipid syndrome. Hum. Reprod., 15, 1170–1172.[Abstract/Free Full Text]

Coulman, C.B., Branch, D.W., Clark, D.A., Gleicher, N., Kutteh, W., Lockshin, M.D., and Rote, N.S. (1999) American Society for Reproductive Immunology Report of the Committee for establishing criteria for diagnosis of Reproductive Autoimmune Syndrome. Am. J. Reprod. Immunol., 41, 121–132.[ISI][Medline]

Cowchock, S. (1991) Autoantibodies and pregnancy wastage. Am. J. Reprod. Immunol., 26, 38–41.[ISI][Medline]

Davis, L.E., Leveno, K.J., and Cunningham, F.G. (1988) Hypothyroidism complicating pregnancy. Obstet. Gynecol., 72, 108–112.[Abstract]

Dendrinos, S., Papasteriades, C., Tarassi, K., Christodoulakos, G., Prasinos, G., and Creatsas, G. (2000) Thyroid autoimmunity in patients with recurrent spontaneous miscarriages. Gynecol. Endocrinol., 14, 270–274.[ISI][Medline]

Esplin, M.S., Branch, D.W., Silver, R., and Stagnaro-Green, A. (1998) Thyroid autoantibodies are not associated with recurrent pregnancy loss. Am. J. Obstet. Gynecol., 179, 1583–1586.[ISI][Medline]

Gardiner-Hill, H. (1929) Pregnancy complicating simple goiter and Graves’ disease. Lancet, i, 120–124.[CrossRef]

Geva, E., Lessing, J.B., Lerner-Geva, L., Azem, F., Yovel, L., and Amit, A. (1997) The presence of antithyroid antibodies in euthyroid patients with unexplained infertility and tubal obstruction. Am. J. Reprod. Immunol., 3, 184–186.

Glinoer, D., Soto, M.F., Bourdoux, P., Lejeune, B., Delange, F., Lemone, M., Kinthaert, J., Robijn, C., Grun, J.P., and de Nayer, P. (1991) Pregnancy in patients with mild thyroid abnormalities: maternal and neonatal repercussions. J. Clin. Endocrinol. Metab., 73, 421–427.

Hamburger, J.I., and Stoffer, S.S. (1981) Is prevention of hyperthyroidism complicating pregnancy justification for routine ablative therapy for hyperthyroidism in women in the child bearing age. In J.I.Hamberger, and J.M.Miller (Eds) Controversies in Clinical Thyroidology. Springer–Verlag, New York, USA. pp 107–108.

Imaizumi, M., Pritsker, A., Kita, M., Ahmad, L., Unger, P., and Davies, T. (2001) Pregnancy and murine thyroiditis: thyroglobulin immunization leads to fetal loss in specific allogeneic pregnancies. Endocrinology, 142, 823–829.[Abstract/Free Full Text]

Kutteh, W.H., Rote, N.S., and Silver, R. (1999a) Antiphospholipid antibodies and reproduction: the antiphospholipid antibody syndrome. Am. J. Reprod. Immunol., 41, 133–152.[ISI][Medline]

Kutteh, W.H., Yetman, D.L., Carr, A.C., Beck, L.A., and Scott, R.T. Jr. (1999b) Increased prevalence of antithyroid antibodies identified in women with recurrent pregnancy loss but not in women undergoing assisted reproduction. Fertil. Steril., 71, 843–848.[CrossRef][ISI][Medline]

Lejeune, B., Grun, J.P., de Nayer, P., Servais, G., and Glinoer, D. (1993) Antithyroid antibodies underlying thyroid abnormalities and miscarriage or pregnancy induced hypertension. Br. J. Obstet. Gynaecol., 100, 669–672.[ISI][Medline]

Magaro, M., Zoli, A., Atomonte, L., Mirone, L., La Sala, L., Barini, A., and Scuder, F. (1992) The association of silent thyroiditis and active systemic lupus erythematosus. Clin. Exp. Rheumatol., 10, 67–70.

Matalon-Tartakover, S., Blank, M., Ornoy, A., and Shoenfeld, Y. (2001) The association between anti-thyroid antibodies and pregnancy loss. Am. J. Reprod. Immunol., 45, 72–77.[CrossRef][ISI][Medline]

Mecacci, F., Parretti, E., Coini, R., Lucchetti, R., Magrini, A., La Torre, P., Mignosa, M., Acanfora, L., and Mello, G. (2000) Thyroid autoimmunity and its association with non-organ-specific antibodies and subclinical alterations of thyroid function in women with a history of pregnancy loss or preeclampsia. J. Reprod. Immunol., 46, 39–50.[CrossRef][ISI][Medline]

Pohlenz, J., Maqueem, A., Cua, K., Weiss, R.E., Van Sande, J., and Refetoff, S. (1999) Improved radioimmunoassay for measurement of mouse TSH in serum: Strain differences in TSH concentration and thyrotroph sensitivity to thyroid hormone. Thyroid, 9, 1265–1271.[ISI][Medline]

Pratt, D., Novotny, M., Kaberlein, G., Dudkiewicz, A., and Gleicher, N. (1993a) Antithyroid antibodies and the association with non-organ-specific antibodies in recurrent pregnancy loss. Am. J. Obstet. Gynecol., 168, 837–841.[ISI][Medline]

Pratt, D.E., Kaberlein, G., Dudkiewicz, A., Karande, V., and Gleicher, N. (1993b) The association of antithyroid antibodies in euthyroid nonpregnant women with recurrent first trimester abortions in the next pregnancy. Fertil. Steril., 60, 1001–1005.[ISI][Medline]

Rushworth, F.H., Backos, M., Rai, R., Chilcott, I.T., Baxter, N., and Regan, L. (2000) Prospective pregnancy outcome in untreated recurrent miscarriages with thyroid autoantibodies. Hum. Reprod., 15, 1637–1639.[Abstract/Free Full Text]

Sherer, Y., and Shoenfeld, Y. (1999) The antiphospholipid (Hughes) syndrome: An entity to be expanded. Am. J. Reprod. Immunol., 41: 113–116.

Stagnaro-Green, A., Roman, S.H., Cobin, R.H., el-Harazy, E., Alvarez-Marfany, M., and Davies T.F. (1990) Detection of at risk pregnancy by means of highly sensitive assays for thyroid autoantibodies. JAMA, 264, 1422–1425.[Abstract]

Takashi, I., Hisato, T., Yoh, H., Nobukai, M., Yuji, M., and Nobuyuki, A. (1997) Effect of autoantibodies on course of pregnancy and fetal loss. Obstet. Gynecol., 90, 365–369.

Tomer, Y., Gilburd, B., Sack, J., Davies, T., Meshorer, A., Burek, C.L., Rose, N.R., and Shoenfeld, Y. (1996) Induction of thyroid autoantibodies in naive mice by idiotypic manipulation. Clin. Immunol. Immunopathol., 78, 180–187.[CrossRef][ISI][Medline]

Vladestine, A.O., and Rose, N.R. (1971) Autoimmune murine thyroiditis: Relation to histocompatibility (H-2) type. Science, 174, 1137–1139.[ISI][Medline]

Submitted on July 26, 2002; resubmitted on November 28, 2002; accepted on January 20, 2003.