The majority of autoimmune diseases including GD display strong associations with HLA variants [ 21 ]. At present it is not clear which locus is responsible for the observed association.
In parallel with susceptibility conferred by Arg it was proposed that Gln protects against GD [ 29 ]. However, the authors concluded that the observed associations to HLA class I alleles could not be attributed to LD within this haplotype [ 32 ]. Whereas Simmonds et al. Since in Caucasians the search for association between GD and DPB1 alleles has been relatively limited and inconclusive [ 33 - 35 ], it should be interesting to include this locus in adequately powered future studies.
The data from the study of Chen et al. In addition to the novel observations Chen et al. Notably, none of the studies analyzed the DPB1 locus.
Since Blacks display the largest HLA diversity, well powered studies of GD in this population may be particularly worth performing in the future. It is likely that HLA associations with autoimmune diseases including GD are caused by the physiological function of these molecules, i. The frequently postulated hypothesis posits that disease-associated variants but not other variants efficiently bind peptides derived from autoantigen s such as thyroglobulin or TSHR and thus have a permissive role in the development of immune response.
It should be emphasized that these mechanisms remain speculative. In conclusion, although the HLA region clearly contains genes predisposing to GD, the exact identity of involved variants or even loci remains unknown. It should be emphasized that HLA associated risk for GD may well reflect superimposing effects also protection of many alleles of different loci including loci encoding other molecules than HLA.
Clearly high resolution HLA studies across various ethnic groups are needed. Such studies are particularly warranted since the available evidence indicates that similarly as in other autoimmune diseases the HLA linked gene s may be the strongest of all genetic factors predisposing to GD. Lymphoid tyrosine phosphatase LYP encoded by the protein tyrosine phosphatase PTPN22 gene was originally associated with T1DM based on candidate gene approach [ 43 ] and subsequently shown to increase the risk for a number of autoimmune diseases including GD [ 44 ], rheumatoid arthritis RA , juvenile idiopathic arthritis JIA and autoimmune Addison disease AAD [ 45 ].
PTPN22 is involved in limiting the adaptive response to antigen by dephosphorylating and inactivating T cell receptor TCR associated kinases and their substrates. The W variant disrupts the interaction between PTPN22 and Csk [ 48 ] and also increases phosphatase activity, which in turn suppresses TCR signaling more efficiently than the wild-type protein [ 49 , 50 ].
In vitro experiments have shown that T-cells expressing the W allele may be hyperresponsive, and therefore carriers of this allele may be prone to autoimmunity [ 43 , 48 ]. A role of PTPN22 in T-cell regulation has been confirmed by the results of knocking out the murine homologue of PTPN22 , which lowered thresholds for T-cell-receptor signaling and inhibited production of IL-2 in these animals [ 51 ].
However, in another study, significantly higher numbers of IL-2 producing cells in W carriers after autoantigen stimulation have been found, suggesting that the W variant is rather a loss-of-function variant [ 52 ]. One explanation of these discrepant results is that the decreased levels of IL-2 associated with the W variant may be caused, rather than by an increase in activity, by an alteration in the cellular location of PTPN22 which impairs the binding of PTPN22 to the protein tyrosine kinase Csk leading to a phenotypic change [ 52 , 53 ].
Interestingly, there are reports indicating that human PTPN22 also inhibits the activity of B-cell antigen receptor [ 50 , 54 ]. Although it seems counterintuitive that the blunted cell activation should lead to increased risk of autoimmunity, there is evidence for such a mechanism also in other models [ 55 ]. While a full explanation of these findings is not yet possible, the dysfunction of regulatory cells and deregulation of lymphocyte maturation in the thymus have been invoked as mechanisms [ 56 ].
A gene dose-dependent effect of PTPN22 'T' allele on the age of onset of Graves' disease has been observed in a Polish population [ 57 ] but was not replicated in a cohort from the UK [ 60 ].
Although most of the studies focused on the CT variant, other polymorphisms of the PTPN22 gene may also be relevant. The variant RQ has been shown to reduce the risk of systemic lupus erythematosus SLE due to reduced phosphatase activity [ 63 ].
In a study of patients with GD from the UK population, a haplotype reducing the risk of disease, regardless of the effect of the polymorphism at position , has been identified [ 64 ]. The conclusions from this study were generally similar to an earlier one conducted among patients with RA [ 65 ].
However, it must be stressed that the protective haplotypes identified in the aforementioned studies were different [ 64 , 65 ]. The role of other variants of PTPN22 is particularly interesting in populations of Asians and Africans, in which RW occurs very rarely or not at all [ 66 , 67 ].
However, a subsequent analysis of SNP markers within the region of the PTPN22 gene revealed the presence of a protective haplotype which is different from those identified previously [ 69 ]. Overall, available results indicate that apart from the well-known polymorphism CT, the PTPN22 locus contains other functionally important variants, particularly those conferring protection.
Currently, the identity of these variants, except for RQ, remains unknown, and their effect can be estimated only indirectly through analysis of haplotypes which differ in frequency between patients and healthy controls. CD40 has been associated with GD as a positional candidate on the basis of a genome-wide linkage study in GD which implicated 20q11 chromosomal region, designated GD-2, as harboring a susceptibility locus [ 70 , 71 ].
The association between GD and rs was not replicated in two large case-control studies in the United Kingdom [ 72 , 73 ]. CD40 is a surface molecule of the TNFR tumor necrosis factor receptor family constitutively expressed on a variety of cells in the immune system, including antigen-presenting cells APCs and B cells [ 77 ], as well as on other types of cells, for example, thyroid follicular cells [ 78 ] and orbital fibroblasts [ 79 ].
CD40 plays a fundamental role in B cell activation as its ligation provides the necessary co-stimulatory signal for B cell proliferation, immunoglobulin class switching, antibody secretion, prevention of apoptosis of germinal-center B cells, affinity maturation and generation of long-lived memory cells [ 80 , 81 ].
However, despite successful lowering of CD40 expression, no effect on the rate of disease induction was observed [ 83 ]. CTLA4 consists of four exons encoding different functional domains such as a leader sequence and extracellular, transmembrane as well as cytoplasmic domains.
The CTLA-4 protein acts as a potent negative regulator of T-cell response [ 84 , 85 ] and variants of CTLA4 have been consistently associated with numerous autoimmune disorders [ 86 ]. CTLA-4 acts by delivering an inhibitory signal through its cytoplasmatic domain which can reverse the classic TCR-induced stop signal needed for physical interaction between T cell and APC, thus reducing adhesion periods between these cells which in turn decreases cytokine production and proliferation [ 87 - 89 ].
The mechanism of the cell extrinsic action could involve stimulation of regulatory T cells but the precise mechanism is not clear since elevated levels of sCTLA-4 are found in patients with autoimmunity [ 91 , 92 ]. Recently Omaer et al.
Clearly, more work is needed to decipher all physiologically relevant effects CTLA4. It has been proposed that long AT-repeat allele decreased stability of CTLA4 mRNA thus blunting the inhibitory function of the protein and thus reducing control of T-cell proliferation [ 96 ]. The second polymorphism implicated as causative was rs, a SNP A49G in the signal peptide causing a substitution Thr to Ala [ 95 , 97 - ].
This amino acid change could influence post-translational processing leading to inefficient glycosylation of the autoimmunity predisposing variant [ ]. However studies conducted by Hu et al. However, this result was not replicated in a larger Swedish study which also did not find any correlation between concentration of serum sCTLA4 and disease status or CT60 genotype [ ].
In a comprehensive meta-analysis by Kavvoura et al. There were 28 studies that reported data for the A49G polymorphism and 7 for CT60 with a total of GD cases typed for A49G and for CT60 as well as and healthy controls, respectively. The summary OR for A49G was estimated as 1. Interestingly, the magnitude of the overall OR diminished over time, from 1. The obtained results were similar for Asians and Caucasians but the Asian studies were more divergent [ ].
A dose-effect association was observed since carriers of two copies of the risk haplotype had significantly higher odds for disease development when compared to single copy carriers OR 1. Kavvoura et al. The current state of knowledge does not indicate clearly the causal mechanism behind the association of CTLA4 with GD. However, CTLA4 polymorphism is among those most consistently associated with thyroid autoimmune diseases in the majority of populations.
It is primarily found on the surface of the thyroid epithelial cells [ ]. TSH is central to the regulation of the thyroid gland. TSHR is located on 14q31 [ , ] and consists of 13 exons [ ]. Despite the positive results of some studies [ - ], subsequent linkage and case-control studies have largely shown no association of GD with either of these TSHR SNPs in Caucasians [ - ]. Nevertheless, genome wide linkage analysis subsequently suggested a GD susceptibility locus in chromosomal region 14q31 [ ].
This encouraged extending the search for susceptibility loci to non-coding sequences within the TSHR gene. Among the Japanese large scale analyses of SNPs which were prompted by initial associations found by typing of microsatellites [ ] showed evidence for three haplotypes within TSHR intron 7 that were strongly associated with GD [ ]. The multiple logistic regression suggested these two SNPs in strong LD explained the association signal in the region [ ].
Functional analyses suggested that rs and rs could be associated with reduced expression of full length TSHR mRNA relative to two truncated splice variants which in turn could lead to increase in shedding of a part of the TSHR receptor called the A-subunit [ ]. The results obtained in the UK population were recently replicated in a Polish and extended British cohorts [ ]. Logistic regression indicated that association at rs may be secondary to rs or that rs is in stronger LD with the etiological variants within the region [ ].
Genotype-phenotype correlations provided no clear evidence of association between rs and any specific clinical characteristics of GD [ ].
Although the most strongly associated subregion of TSHR gene is different in Caucasians intron 1 [ , , ] and Asians intron 7 [ ] there may be some sharing of genetic susceptibility factors between these populations. Notably, rs is a perfect proxy of rs which was associated with GD in Caucasians [ , ]. On the other hand rs, a nonsynonymous SNP in the distal part of the gene exon 9 has been associated with GD in a large GWAS study of nonsynonymous variants, albeit only after the control group was expanded to include a total of subjects [ ].
As this association was not replicated, it could be a false positive, although there remains a possibility that the lack of replication was due to insufficient power [ ]. To summarise, whereas it is safe to conclude that the TSHR gene polymorphism is associated with GD, further fine mapping, perhaps using next-generation sequencing, as well as functional studies will be required to determine the exact location of the etiological DNA variants and to determine to what extent TSHR linked susceptibility factors for GD differ between Caucasian and Oriental populations.
The induction of experimental autoimmune thyroiditis in mice using Tg as an antigen provides further evidence of their major role in the pathogenesis of HT and not GD in humans. In several studies, association of TG polymorphisms with GD as well as with the relapse of Graves' hyperthyroidism after antithyroid treatment was found [ - ]. Unfortunately, these results have to be interpreted with caution because of rather small sizes of studied groups. Possibly the observed inconsistencies reflect differences in the autoimmune response to Tg in various forms of AITD.
The extent of iodination, changes within the amino acid sequence and posttranslational modifications of Tg may lead to changes in antigenicity and binding to HLA. Especially exon 33 SNP causing a change from a hydrophobic amino acid Trp to a positively charged hydrophilic amino acid Arg would be expected to change the structure of Tg. In vitro studies suggest that the disease-associated G allele confers increased promoter activity through the binding of the interferon regulatory factor-1 IRF-1 , a major interferon-induced transcription factor.
Finally, an interesting hypothesis which has not been studied posits that differences in transcription of TG in the thymus could influence the negative selection of autoreactive T-lymphocytes. FCRL3 is a receptor of unknown function with structural homology to classical receptors for immunoglobulin constant chains Fc receptors.
FCRL3 is expressed in lymphoid organs, particular strongly on the surface of the B-cells [ ]. Presence of FCRL3 was also demonstrated on the surface of a subset of T reg cells characterized by lower relative response to antigenic stimulation and reduced suppressor activity [ ]. In the original report, rs was suggested to have functional significance as the disease-associated C allele increased affinity for NFKB transcription factor and showed enhanced transcription rate in luciferase assay [ ].
Simmonds et al. In the case of rs the conclusion was particularly strong since an opposite effect protection was reported. Although the study confirmed association of rs tagging rs with GD, this association turned out to be secondary to the effect of two other SNPs within FCRL3 : the previously implicated rs [ ] and rs [ ].
Further analysis suggested that rs may not be the etiological variant, as the effect of risk G allele of this SNP was haplotype dependent being apparent only when present together with the rs C allele. Overall the available data suggest that genetic polymorphism s modifying susceptibility for GD do exist in the FCRL3 region but the primarily associated variant s remain s to be found [ ].
The strongest association reported by Song et al. This was confirmed and extended by Song et al. Further support for functional significance of the rs A variant comes from its association with reduced concentration of UGRP1 in serum of healthy subjects as well as of patients with GD or asthma [ , ].
UGRP1 is a ligand for macrophage scavenger receptor with collagenous structure MARCO [ ] which has an important function in the innate immune system of the lung where it binds inhaled particles including microbial pathogens and facilitates their clearance by the macrophage system [ - ].
Accordingly, UGRP1 is predominantly expressed in the lung although a low level expression was also found in human thyroid and kidney [ , ]. Whereas expression of this gene in the thyroid does not preclude a local effect the well documented function of MARCO and presumably UGRP1 in lung physiology suggest that the association may be caused by systemic effects originating from the respiratory system.
Interestingly, rs A has been associated with asthma [ ]. Although this was not confirmed, an intriguing link between this variant and total IgE, although limited to healthy controls, was recently suggested [ ]. Out of 14 SNPs analyzed, seven showed association, although evidence was modest 0. The strongest of associations found was to rs although the direction of effect protection was opposite to that found in T1DM [ ].
Given the central role of IL2 in immune response it is perhaps not surprising that this region was also associated with other autoimmune diseases such as celiac disease [ ], ulcerative colitis [ ], RA [ ], JIA [ ], psoriasis and psoriatic arthritis [ ] and others [ ] although in the majority of studies another SNP rs , only weakly linked with rs, was implicated with the minor allele generally conferring protection [ ].
It is indeed possible that the 4q27 contains more than one susceptibility locus since rs but not rs is associated with RA [ ]. The second strongest association reported by Todd et al. This finding was not secondary to the effect of an established risk marker of multiple sclerosis MS located in the same region rs in IL7R [ - ] as this SNP was not associated with GD [ ].
A subsequent study in a German population patients, controls did not report statistically significant association between rs and GD although the expected trend was noted [ ]. No association between rs and GD was found among the Chinese patients, controls [ ] or the Japanese patients and controls [ ].
The association between autoimmunity and rs was first found in a GWA of nonsynonymous SNPs in T1DM with the minor allele threonine conferring protection [ ]. IFIH1 is one of a family of intracellular proteins involved in innate immunity through recognition of viral RNA [ ]. Interestingly, available data indicate that a normal i.
Whereas the rare variants identified by Nejentsev et al. It has been shown that Ala is associated with reduced IFIH1 transcription [ , ] but there are also reports conflicting with this conclusion [ , ]. Association to rs on 2q11 [ ] may indicate the role of locus AFF3 which encodes nuclear protein LAF-4 with homology to a protein involved in leukemia development and a possible function in lymphoid ontogenesis [ ].
Another novel association in GD suggested by Todd et al. The Ser variant alters splicing of the CD transcript and this has been suggested to explain the association [ ].
Another not exclusive possibility is that the amino acid change increases receptor signaling by strengthening the phosphorylation of neighboring sites [ , ]. Conversely, recent in vitro analysis indicated a direct functional effect of another SNP closely linked with rs rs , with the disease associated variant causing a decreased expression [ , ].
Consistent with this was the in vivo observation that disease-associated rsrs haplotype correlated with lower surface expression of CD in T cells and NKT Natural Killer T cells [ ].
Since no such genotype-dependent differences in expression were observed in NK or B cells it was suggested that CD function in T cells or NKT cells was important in disease pathogenesis [ ].
These results were later extended to ulcerative colitis [ ], JIA [ ] and possibly celiac disease [ ]. Subsequently it was suggested that there was even stronger association with T1DM at rs so that this marker should replace rs [ ] but so far there have been no data on association with GD of rs or SNPs in the PTPN2 region other than rs or rs PTPN2 encodes a classical, non-receptor protein tyrosine phosphatase, related to but distinct from PTPN22 which is a very well established risk factor for autoimmunity see above.
PTPN2 is highly expressed in hematopoietic cells and the phenotype of the knockout mouse model indicates its important role in immune system development, function and predisposition to autoimmunity [ ]. Subsequently, rs was associated with SLE [ ] and differences of ESR2 mRNA concentration suggesting that increased expression of this gene may predispose to autoimmunity [ ].
Although the association with susceptibility to GD could not be replicated in a Japanese cohort, subgroup analysis in Japanese GD patients revealed a correlation between the NFKB1 genotype and the development of GO as well as the age of disease onset [ ]. NFKB1 del ATTG has been associated with a decreased transcription [ ] and its prevalence was assessed in cohorts with various autoimmune diseases although with variable results [ ].
In female cells, one of the two X chromosomes is inactivated in early embryonic life and this epigenetic system assures that men and women have equal expression of the genes from the X chromosome, despite the difference in X chromosome copy number.
However, in accordance with normal distribution, this random process sometimes generates an unbalanced inactivation pattern with one copy of the X chromosome preferentially expressed. Skewed XCI could also result from a bias in the initial choice of the X chromosome which is inactivated due to germline XIST X-inactive-specific transcript mutations [ 84 ] or deleterious X-linked mutations, X chromosome rearrangements, ageing, twinning, or monoclonal expansion of cells [ ].
Importantly, the degree of XCI skewing may also be determined by polymorphic variants on the X chromosome, in particular those located close to the X inactivation centre or chromosomal bands Xqq26 [ - ]. These results were confirmed by three studies in other populations [ - ]. Ozcelik et al. Skewed XCI was also demonstrated in thyroid specimens [ ].
Yin et al. Similar conclusions were reached by Chabchoub et al. The explanation of the mechanism by which skewed XCI can lead to autoimmunity has been based on the loss of mosaicism hypothesis which posits that lower expression of self-antigens from one X chromosome may interfere with induction of tolerance in the thymus [ - ].
The inefficient negative selection in the thymus could result from the fact that the cells mediating it i. Owing to an overlap between HLA class I and II antigen presentation pathways these untolerized T cells could in turn provide help to a population of B cells expressing these antigens, thus causing their polyclonal activation.
Whereas originally proposed to explain pathogenesis of SLE [ ] this mechanism could conceivably decrease the threshold also for other forms of autoimmunity, including GD.
Another mechanism of autoimmunity development could be related to differences in the degree of XCI skewage found among different tissues [ ]. In particular, it could happen that antigens with low expression in the thymus are relatively highly expressed in peripheral tissues such as thyroid and induce autoimmunity [ - ].
However, this explanation is difficult to reconcile with the lack of skewed XCI observed in other autoimmune disorders with female preponderance such as MS, SLE or primary biliary cirrhosis [ - ] as well as data from animal experiments showing that female predisposition to autoimmunity is not influenced by homozygosity for X chromosome which should eliminate defective tolerance induction due to skewed XCI [ , ].
Although the association between skewed XCI and AITD has been found repeatedly [ - ] its mechanism should be regarded as speculative. Recently the same group who described the original association between AITD and skewed XCI provided preliminary evidence of its noncausality [ ]. The argument was based on lack of correlation of within-pair differences in values of XCI with differences in concentration of autoantibodies against thyroid peroxidase observed in MZ but not DZ twin pairs [ ].
Whereas the study had limited power mainly because of high within-pair similarities in the concentration of autoantibodies among MZ, its results were interpreted to suggest that skewed XCI and AITD both share genetic determinants but are not causally related [ ].
Since the degree of XCI skewing may be linked with markers on the X chromosome [ ] such an interpretation would suggest that there are as yet unidentified AITD susceptibility loci located on the X chromosome.
The existence of GD susceptibility loci on the X chromosome was suggested by early studies based on linkage analysis, but without agreement regarding the exact location [ , ]. Fetal microchimerism refers to presence within the maternal organism of a small population of cells originating from the fetus.
A convenient method to study fMC consists of looking among females for male cells or their DNA [ ]. A first link between fMC and thyroid pathology was provided by demonstration of male cells in thyroidectomy specimens from female patients but not in control autopsy samples [ ].
Subsequently, the same group extended these findings using a different methodology [ ]. Studying fresh-frozen specimens Ando et al. Some indirect support for a causative role of microchimerism in AITD has been provided by an observation that twins from opposite sex pairs have a higher frequency of thyroid autoantibodies then monozygotic twins [ ].
However, in this setting, the relevant microchimersim should result from in utero cell trafficking between twins rather than from migration of fetal cells into the mother organism and thus could be qualitatively different [ ]. In a murine experimental autoimmune thyroiditis model it was shown that inflamed but not healthy thyroid accumulated fetal cells, including T cells and dendritic cells [ ]. Although these cells could theoretically contribute to postpartum exacerbation of thyroid inflammation [ , , ] these data are compatible with high prevalence of fMC in human AITD being a consequence rather than the cause of disease.
Indeed, the hypothesis that presence of fetal cells was associated with the maternal response to injury inflammatory or other as opposed to causing disease was proposed relatively early [ ] and is consistent with observations in animal models [ - ] and humans [ ].
Whereas such an association was indeed found in one cohort [ ] it was not observed in four other studies [ - ]. Whereas the role of fMC in postpartum disease aggravation cannot be excluded, available data suggest that this and other sex related features of GD such as female preponderance will perhaps be more readily explained by hormonal differences between females and males. Interestingly, in pathogenesis of AITD a role for hyperestrogenic state due to a longer reproductive span has been proposed [ ].
The role of hormonal factors in thyroid autoimmunity is also consistent with the association discussed above between GD and a ESR2 variant [ ].
By the time Graves attended Charterhouse school he was a devoted prude, chastely and fearfully homosexual, and forthright to the point of aggression in matters sporting and academic. It was at Charterhouse that Graves was first published as a poet.
Fighting in the Battle of Loos could be endured only by consuming a bottle of Scotch a day. Later in the war he was so severely wounded that he was assumed dead. An errant telegram informed his parents of the news; a misplaced notice in the Times informed the nation. Near death, she admits her crimes to Claudius, stating that all were done in fulfillment of prophecies and for the greater glory of Rome. Claudius devotes most of his life to the art of history, and Graves himself took pride in his scholarship.
As for that vigorous fancy, Moorcroft Wilson quotes a letter from Graves in which he explains his tendency to reuse personal material from his life and the lives of those he knows without consideration of the harm this might do. When talking of himself in his poems or his memoir, he is direct and engaging.
When writing fiction, he almost invariably constructs so sturdy a framework of fact, theory, conjecture, and supposition that the conventional pleasures of fiction are elided. It recurs here, in all its ugliness, and the heroine narrator bears some responsibility for it. But Nausicaa, remember, is the author of the Odyssey and includes the killings in the poem she writes.
Other novels are less successful. The novel appeared a century after Palmer hanged, but eight years before the United Kingdom executed its last felon. And, whether or not he murdered his dissolute brother Walter, no one can deny he deliberately helped him graveward. The doctor took out a life insurance policy on the booze-soaked Walter, already a delirium tremens sufferer, and encouraged him to drink more, even going so far as to cover any pub bills Walter might incur as he staggered to his deathbed.
But how many people deliberately poison themselves as slowly as Graves alleges Annie Palmer did? Early in I, Claudius , the emperor reflects on the posthumous life of his memoirs.
0コメント