Figure 10-1.1 In this remarkable x-ray crystallographic study, an HLA Class I molecule is seen from above. The two interlocking subunits form an antigen binding cleft into which the peptide epitope must fit and remain if it is to be recognized by the T cell receptor(a).    In (b), fortuitously, a peptide epitope is found occupying the cleft, fitting like the hot-dog in a bun.

CTLA4: Recently another gene was found to be related to the propensity to develop Graves' disease, and this gene also is involved in immune responses ^72,73. When an immune reaction begins, the "first signal" is through the recognition by the T cell receptor of its cognate epitope presented in an HLA molecule. However, if only the first signal is received by a T cell, the T cell tends to be turned off or anergized. In order for a progressive immune response to recur, there must be a "second signal" provided by one of several adhesion molecules which exist on the APC and T cell, and which tend to augment the affinity of the interaction ⁷⁴. Of these, one of the most important is "B7", which exists in two forms, B7.1 and B7.2, present on the surface of APCs. These molecules interact with their cognate receptors on the T cell, CD28 for B7.1 and CTLA4 for B7.2. In many situations interaction between B7.1 and CD28 give a positive stimulus to growth of the T cell, whereas interaction with CTLA4 provides a negative signal ⁷⁵. CTLA4 exists as a gene with several isoforms. These are due to AT dimer polymorphisms in the 3’-untranslated region of exon 3, which are also closely linked with a polymorphism in exon 1 ^75.1. It has been found that inheritance of the 106 base pair AT polymorphism is associated with a greater incidence of Graves' disease, especially in males ^72,73. The G (alanine ) position 49 allele was found to be linked to Graves' Disease by Heward et al (75.11), and Vaidya et al found in a linkage study that inheritance of a specific CTLA-4 allele along with MHC allele was responsible for 50% of the genetic influence in Graves' disease(75.12). CTLA gene polymorphism studies indicate that the G allele, associated with the development of Graves’ disease, also influences higher production of TPO and Tg antibodies (75.14).   
      Kouki and De Groot investigated the relation of the CTLA-4 alleles to proliferation of T cells in patients with Graves’ disease and Hashimoto’s thyroiditis.  They found that T cells from all subjects that have the G polymorphism, including normal controls, proliferate to a greater degree than do lymphocytes bearing the CTLA-4 A polymorphism.  This is presumed to give individuals carrying the G polymorphism a mild but important greater propensity for development of  a functional auto-reactive lymphocyte clone^75.13.  
    Interestingly, the HLA association suggests a relationship to disease specificity, since it has to do with the presentation of specific antigen epitopes, whereas the CTLA-4 polymorphism appears to be a general phenomenon, allowing one population group to have augmented lymphocyte proliferation, but is not specifically related to the disease.  These observations also fit with the concept that development of Graves’ disease is mediated by a set of genes rather than one specific gene (23). It is reasonable that a variation in the function of the CTLA4 gene makes it  less effective, as a suppressive signal controlling autoimmunity.
    The genetic effects of DR  genes and CTLA-4 interact. Specifically, it has been found that the positive effect of  CTLA-4 predisposition mitigates in part the negative effect of DRB1*0701, but does not interact with the positive influence of DRB1*0301(Kula D, Bednarczuk T, Jurecka-Lubieniecka B, Polanska J, Hasse-Lazar K, Jarzab M, Steinhof-Radwanska K, Hejduk B, Zebracka J, Kurylowicz A, Bar-Andziak E, Stechly T, Pawlaczek A, Gubala E, Krawczyk A, Szpak-Ulczok S, Nauman J, Jarzab B  Interaction of HLA-DRB1 alleles with CTLA-4 in the predisposition to Graves' disease: the impact of DRB1*07. Thyroid. 2006 May;16(5):447-53.).

Other genetic associations have been reported. It has been indicated ⁷⁶, and denied ⁷⁷, that an association exists between a specific polymorphism of the TSH receptor (PRO52THR), and susceptibility to Graves' disease.   Alleles of intron 7 of the TSH-R gene were found associated with GD in Japanese patients(77a). Linkage to the TSH-R gene has recently been confirmed by Dechairo et al (77b).The TG gene has been linked to Graves and other AITD, but to date evidence for this relation is uncertain (77.01)  A Vitamin D receptor exon 2 initiation codon polymorphism has been associated with Graves’ disease in a Japanese population.  A similar association has been reported with IDDM and multiple sclerosis ^77.1. Vitamin D and its receptor are involved in control of autoimmunity, so an association is not surprising, but the mechanism remains unknown. Inheritance of specific V genes coding for immunoglobulins may carry the same kind of risk ⁷⁸. Several possible genes linked to Graves’ disease or autoimmune thyroid disease have been found by linkage studies, including one recently described at a locus on chromosome 18q21 that is also associated with IDDM. (78a).  
    Linkage studies of Graves’ family members have suggested the susceptibility locus on chromosome 20q11, which has been named GD-2 by Davies et al, is related to the gene CD-40, expressed on B cells and other immune cells.  Recently this group identified a polymorphism in the Kozak sequence of the CD40 gene at position –1 from the translational start site.  The CC genotype was associated with Graves’ disease and gave a relative risk of 1.6.  Previous studies using a mixed population of patients had not supported such a linkage, but in a Caucasian population, this association and linkage were shown(78.11).  A promoter polymorphism of the CYP27B1 gene has been associated with Graves' Disease and other autoimmune diseases (78.12). Interestingly, this gene catalyzes the conversion of 25-OH-D to 1,25-OH-D, the active metabolite of D. A possible relation of Vit D receptor to Graves' was noted above. A B cell specific gene ("ZFAT") has been linked to autoimmune thyroid disease, though not specifically to Graves (78.13).
     IL-13 gene polymorphisms were studied in Japanese GD patients (n = 310) and healthy control subjects without antithyroid autoantibodies or a family history of autoimmune disorders (n = 244). A C/T polymorphism at position -1112 of the promoter region was measured using the direct sequencing method, and an Arg(130)Gln (G2044A) polymorphism in exon 4 was examined using the PCR-restriction fragment length polymorphism method. IL-13 gene polymorphisms are associated with GD susceptibility in Japan(78.14).
    The lymphoid tyrosine phosphatase, encoded by the protein tyrosine phosphatase-22 (PTPN22) gene, is a powerful inhibitor of T cell activation. Recently, a single-nucleotide polymorphism (SNP), encoding a functional arginine to tryptophan residue change at PTPN22 codon 620 in Caucasians has been shown to be associated with GD and other autoimmune diseases.  This SNP inhibits function of the gene, which is normally to down-modulate signaling via binding to Csk and phosphorylation of Lck. Using a polymerase chain reaction (PCR)-restriction fragment (XcmI) assay to examine genotypes at the codon 620 polymorphism in 334 unrelated patients with AITD and 179 controls, none of the patients with AITD and controls had the tryptophan allele. The codon 620 polymorphism of the PTPN22 gene does not have a causal role for AITD in the Japanese. Of interest, knockout mice deficient in this gene do not develop signs of autoimmunity, suggesting it may not be important in etiology of AITD (78.15).
    The frequency of C/C genotype of CD40 was increased in GD compared to controls, but the difference was not significant (60.5% versus 55.8%, p = 0.062, odds ratio [OR] = 1.21, 95% confidence interval [CI]: 0.96-1.53). In a meta-analysis with the data from previous studies, the combined OR for the C/C genotype as a risk factor for GD was 1.22 (95% CI: 1.08-1.38, p = 0.001). There was no interaction between CD40 genotypes and other GD susceptibility alleles. No significant genotype-phenotype associations were found. The CD40 C-T polymorphism appears to have a modest effect on genetic susceptibility to sporadic GD(78.16).
    Although linkage analysis has often been considered to be superior to gene association studies for determining genetic effects in autoimmune diseases, in fact linkage analysis may be limited in defining such loci, and large-scale association studies may prove to be more useful in identifying genetic susceptibility factors for AITD. A genome-wide screen was performed on affected relative pairs with autoimmune thyroid disease. 1119 Caucasian relative pairs affected with autoimmune thyroid disease (GD or AIH) were recruited into the study. The study aimed to identify regions of genetic linkage to AITD. Three regions of suggestive linkage were obtained on chromosomes 18p11 (maximum LOD score 2.5), 2q36 (maximum LOD score 2.2) and 11p15 (maximum LOD score 2.0). No linkage to HLA was found. The absence of significant evidence of linkage at any one locus in such a large dataset argues that genetic susceptibility to AITD reflects a number of loci each with a modest effect (78.17).
      The final result of this kind of research is not clear at present. The most logical idea is that there may be several -- probably many -- genes which augment the possibility of developing immunity to the thyroid gland protein components. Thus it is suggested that the inheritance is polygenic. Rather than inheriting one gene which, in a dominant fashion, induces Graves' disease, probably individuals inherit several different genes which are related to the development of thyroid autoimmunity. If a sufficient load of the positively acting genes are inherited, they support the development of Graves' disease or other AITD, especially if other factors are present such as injury to the thyroid. A dramatic illustration of the genetic influence is provided in a recent report of "adoptive" hyperthyroidism following allogenic stem cell transplant from an HLA-identical twin with Graves" (78.1)

Gender: Perhaps the clearest association with autoimmune disease is being a member of the female sex, which carries a 10 - 20-fold risk compared to the male sex. Despite this obvious association, the mechanism has remained obscure. The association carries through not only for autoimmune thyroid disease but also for the development of multinodular goiter, and even differentiated thyroid carcinoma, but not undifferentiated thyroid carcinoma. Thus female gender may endow a generally greater reactivity of the thyroid gland, or may subject it to greater stress in some manner. It has been suggested that there may be specific receptors on the promoter for DR genes, which makes them responsive to the estrogen receptor. This sort of association would be very convincing, if proven.

Other Suggested Causative Factors
A variety of other ideas have been presented as the "cause of Graves' disease", but remain unlikely or unproven. (Table 10-3)  Mutation of T or B cells to produce a specific reactive clone has been suggested ⁸⁰. Of course somatic mutation is a known part of the development of B cell clones, whereas specific mutations of B cells or T cells can produce tumors rather than a disease producing autoimmunity.

Treg abnormalities-The idea that there could be a deficiency of suppressor cells has been much reported, and there is considerable evidence that something like this is true in autoimmune thyroid disease ^82-85. Exactly what the deficiency is remains uncertain. It has been shown that there are fewer CD8+ cells which are specific for thyroid antigens ⁸⁶. If suppressor cells play a role in controlling autoimmunity, presumably it is by the elaboration of cytokines which affect other autoreactive cells. In a certain sense, since the function of the immune system is always delicately balanced between factors which stimulate a reaction and factors which tend to suppress it, a deficiency of a suppressor function must, in a certain way, be related to the development of clinical autoimmunity. Defining the defect as a specific suppressor cell deficiency has, however, been difficult.
    Current research centers on the role of  T cells which express CD4, CD25 and Foxp3 proteins and CD127, and appear to be important in suppressing autoimmune reactions(78.2). A deficiency of such cells, or in their function, has been related to coeliac disease, rheumatoid arthritis and autoimmune diabetes, and is probably related to GD although no current evidence proves this relation. It is possible that a relative deficiency of such cells explains the appearance of GD during immune system recovery following medical treatment of patients with advance HIV disease (78.3) and after therapy with the lymphocyte depleting antibody CAMPATH. Regulatory T cells expressing CD25, GITR and Fox p3 are increased in blood of patients with AITD, compared to controls, and further increased in the thyroid.  However tests for suppressive action of these cells suggests they are deficient in this function(78.2). Depletion of CD4+ CD25+ T cells rendered  C57BL/6 mice more susceptible to induction of a model of Graves disease by immunization with an adenovirus expressing TSHR (78.21).

Fetal cell microchimerism-Intrathyroidal fetal cell microchimerism has been suggested as a possible etiologic agent in autoimmunity.  During pregnancy, fetal and maternal cells are transferred between mother and fetus.  It has been shown that fetal cells from male infants can persist in the maternal circulation for up to 20 years.  Male fetal origin cells were studied in human thyroids by identifying the male specific region of the SRY region of the Y chromosome, and were detected in 6 of 7 frozen thyroid tissue specimens from patients with Graves’ disease, and one of four with thyroid nodules.   Fetal male cells are possible candidates for modulating autoimmune thyroid disease, since they might either induce an immune response, or  develop a sort of graft-versus-host immune response to the mother (86.1). 

Development of expression of Class I or Class II MHC molecules on the thyroid epithelial cell was suggested as a factor in the causation of Graves' disease by Bottazzo et al ⁸⁷. It is now apparent that exposure of thyroid epithelial cells to Interferon, presumably elaborated by infiltrating lymphocytes or other immune cells, can lead to the expression of Class II molecules on the thyroid cell surface ⁸⁸. Expression of these molecules does allow the thyroid epithelial cell to function as a weak antigen presenting cell ⁸⁹. Class II expression is secondary to the effect of an autoimmune lymphocyte attack and is induced by Interferon ⁹⁰. Culture of human thyroid cells from patients with Graves' disease in vitro shows that Class II expression disappears ⁹¹, as it does when the cells are transplanted into nude mice ⁹². It is also possible that the Class II expression is a defensive response ⁹³ Antigen presentation to a T cell, in the absence of a second signal, could lead to anergy of the attacking T cell. If the original hypothesis is proven to be correct, it may be that the Class II expression is secondary but may play a role in continuing or strengthening the autoimmune reactivity to thyroid antigens. Antithyroid drugs may have an immunosuppressive effect on autoimmune thyroid disease through inhibition of HLA-DR expression on thyrocytes.  Follicular cells of patients with overt thyrotoxicosis express HLA-DR, while those in remission, or those under medication with antithyroid drugs, did not (92a). Recently it was reported that transgenic mice expressing MHC Class II on their thyroid cells do not develop autoimmunity (93.1). This is a strong argument indicating that Class II MHC expression on Graves thyroid cells is secondary and not a causative event. 

Stress: Psychic trauma or psychic stress has long been considered to be a possible etiology of Graves' disease. In Perry's original report, a crippled woman, who was injured when her nanny allowed her wheelchair to fly down a flight of stairs, had rapid onset of thyrotoxicosis. This report, in 1820, has been followed by many more recent studies, some of which support the idea. The incidence of Graves' disease increased in Denmark during World War II ⁹⁴, but did not in Ireland during the sustained civil war in that country during the period of 1980 through 1990. A recent study in Yugoslavia indicated that patients with Graves' disease had suffered on average more stressful episodes than control subjects, but previous similar studies have failed to show this relationship ^95-97. .  A recent article found increased numbers of stressful life events in patients with Graves’ disease prior to onset of the disease, compared to patients in a toxic nodular goiter group who had a similar number as control patients. (97a). Stress induces a variety of physiologic responses including anxiety, tachycardia, restlessness, etc., which are not unlike symptoms of Graves' disease. Its role remains enigmatic in causation of Graves' disease to this date. A mechanistic route from stress to the development of Graves' disease is not obvious. Theoretically, stress might cause activation of the adrenal cortex or the sympathetic nervous system. Hypercortisolism would tend to suppress autoimmunity. Heightened sympathetic nervous system activity might theoretically cause stimulation of thyroid secretion, as has been shown in experimental animals ⁹⁸. Other specific stressors have been reported.
    Aggressive weight loss programs have been reported to induce Graves' disease. Administration of thyroid hormone, sometimes given for induction of weight loss, also has been followed by mini-epidemics of Graves' disease ¹⁰⁰.

Excess Iodide: Iodide itself has been thought to induce Graves' disease, thus leading to the term "Jod Basedow". This syndrome refers to the occurrence of thyrotoxicosis following supplementation of iodide in medicinals or by salt iodinization. Excess iodide clearly does induce hyperthyroidism in patients with multinodular goiter ¹⁰¹. Presumably autonomous nodules in the goiter are unable to produce an excess of hormone when their synthesis is limited by iodide , but when this iodide supply is augmented many-fold, the nodules can process it to produce an excess of hormone. The best defined epidemic of iodide-induced thyrotoxicosis occurred in Tasmania after the government introduced iodization of salt, and was clearly associated with multinodular goiters rather than typical Graves' disease ^102,103.

Possibly increased iodide intake can actually augment thyroid autoimmunity through other mechanisms. For example, increased iodide intake has been correlated with an increase in incidence of AITD ^103.1 This could in theory work by augmenting iodination of TG, and heavily iodinated TG is more immunogenic in animals than is poorly iodinated TG. Also, under special circumstances excess iodide can induce thyroid cell necrosis, and this might liberate antigens. Adding KI to the diet of the thyroiditis-prone BB strain rat and to NOD mice increases the severity of thyroiditis ^103.2-103.3.

Whether an excess of iodide can induce true Graves' disease and autoimmunity remains unknown. In fact the addition of 2 - 6 mg per day of iodide to the intake of most patients with Graves' disease, raising plasma iodide levels above 5ug/dl, causes a dramatic reduction in hormone release by the "Wolff-Chaikoff" effect, which is an inhibition of hormone synthesis and of hormone release ^104-107. Iodide is one of the most rapid acting agents in suppressing thyrotoxicosis. While it has this effect in most individuals with Graves' disease, its action tends to be partial or transient, and thus is not relied upon as an effective antithyroid agent. A recent study suggests that the ability of iodide to suppress Graves' disease may be because iodide down-regulates MHC Class I and II expression on thyroid cells ^107.1.

Smoking has been related to Graves' Disease, and more specifically to a greater propensity to develop ophthalmopathy, or to have worsening of the condition.    Salvi et al also studied cytokines in patients with Graves’ disease and found serum IL-6 concentrations higher.  Interestingly, smoking, which is associated with an adverse effect on Graves’ ophthalmopathy, appeared to have no interaction with the serum lymphokines ^107.2.

Other older ideas on the cause of Graves' disease included a role for TSH, or some fragment of TSH ¹⁰⁸, or the HCG molecule. None of these is thought to be involved in thyroid autoimmunity according to current formulations, although excess TSH production by a pituitary adenoma is an established cause of thyrotoxicosis ^109,110, and TSH produces the thyrotoxicosis seen in "Pituitary Resistance to Thyroid Hormone". Vassart and co-workers ¹¹¹ have recently recognized the cause of non-autoimmune Hereditary Familial Hyperthyroidism inherited in an autosomal dominant manner. In two sibships, activating germline mutations were found in the transmembrane segments of the TSH-R. The mutations cause persistent basal hyperfunction of the receptor and early onset of thyrotoxicosis. Other causes of thyrotoxicosis (but not Graves' Disease) are described briefly in Chapter 11 and more fully in Chapter 13. Numerous other theories regarding the cause of Graves' disease have been proposed in the past. For a critical review of earlier speculations and a superb bibliography, the reader is referred to the monograph by Iversen ¹¹².