Abstract
Graves' disease (GD) is a frequent organ-specific autoimmune disease of the thyroid gland, with environmental and genetic factors implicated in the pathogenesis. One of the histopathological hallmarks of this disease is lymphocytic infiltration, predominantly composed of CD4+ and CD8+ T-lymphocytes (1). The pattern of CD4+ T-lymphocytes involved in GD changes throughout the course of the disease. In patients with a recent onset, a predominance of T helper 1 (Th1) lymphocytes, as a sign of cell-mediated immune reaction, can be observed, whereas humoral immunity, in which Th2 lymphocytes predominate, plays a greater role in later stages of the disease (1,2). Therefore, the active phase and exacerbation of GD is characterized by the presence of proinflammatory and Th1-associated cytokines in the thyroid gland (3,4).
Chemokine CXC motif ligand 9 (CXCL9, also called MIG [monokine induced by IFN-γ]) and CXCL10 (IP-10 [IFN-γ–inducible protein of 10 kDa) are IFN-γ–inducible chemokines that preferentially attract Th1 lymphocytes through the CXC chemokine receptor 3, expressed at high levels on Th1 lymphocytes (5,6). Recently, several studies have demonstrated that CXC chemokines, particularly CXCL10, is increased in diverse autoimmune diseases including autoimmune thyroid disorders (7 –11). Thus, in thyroid glands from GD patients the expression of CXCL9 and CXCL10 mRNAs was significantly higher compared with normal thyroids. Additionally, an inverse correlation between serum concentration of CXCL10 and the duration of GD has been described, where the highest levels were found in patients with recent onset (12 –14).
So far, genetic variants in CXCL10 gene have been analyzed in different immune-mediated diseases such as seasonal allergic rhinitis, multiple sclerosis (MS), and Alzheimer's disease, but not yet in GD (15 –17). As it is conceivable that chemokines could play an important role in GD explaining at least in part why lymphocytes accumulate in the thyroid, we studied two single-nucleotide polymorphisms within the CXCL10 gene (rs8878 and rs35795399) in GD patients from Germany, Poland, and Serbia as well as in their respective healthy controls.
Altogether 1192 individuals were involved in this study: 581 patients with GD (228 Germans, 181 Poles, and 172 Serbs) and the others are healthy controls (n = 611; 316 Germans, 152 Poles, and 143 Serbs). The GD patients enrolled in this study from Germany, Poland, and Serbia were 22–89, 10–82, and 22–83 years old with a male:female ratio of 1:3.9, 1:3.8, and 1:5.7, respectively. GD was defined by hyperthyroidism, serum antithyrotropin receptor antibodies and/or diffuse goiter, or presence of ophthalmopathy, which was classified according to the NOSPECS classification. The cutoff for a severe course of GD was defined by the presence of Graves' ophthalmopathy, recurrent hyperthyroidism, thyreotoxic crisis, huge goiter (Stadium III), and/or other autoimmune diseases such as Addison's disease, Hashimoto's thyroiditis, MS, type 1 diabetes, or rheumatoid arthritis.
The control group comprised unrelated age- and gender-matched staff personal or medical students from the local universities (Germany/Frankfurt am Main, Poland/Warsaw, Serbia/Zajecar) or volunteer blood donors from the Red Cross Transfusion Centre in Frankfurt am Main/Germany without personal or family history of autoimmune diseases. All individuals were of Caucasian origin with ancestors of the same population (Germany, Poland, and Serbia) at least for two prior generations and were inhabitants from the surrounding area of Frankfurt am Main, Warsaw, and Zajecar. All subjects gave their consent to the study, which was approved by the local ethics committee.
So far, no data correlating CXCL10 polymorphisms and GD have been reported. Therefore, we focused our analysis on two polymorphisms that were previously analyzed in patients with Alzheimer's disease (rs8878) (16) and type 1 diabetes (rs35795399) (18), respectively. The single-nucleotide polymorphisms were positioned according to the NCBI (
Our results showed that all samples were in Hardy–Weinberg equilibrium. Rs8878 and rs35795399 polymorphisms were not in linkage disequilibrium with each other. Distribution of the rs8878 and rs35795399 polymorphisms (data not shown) did not show significant differences between patients with GD and healthy controls in any population and they did not differ between patients of the three populations. Interestingly, in Germans, a significant association between the rs8878 polymorphism and the severity of GD was observed. Thus the genotype “AA” was less frequent in patients with a severe course of GD compared with those with a mild GD (8.5% vs. 31.8%, p = 1 × 10−4; Table 1). In the Polish population, no differences were found comparing patients with severe course of GD to patients with mild course of GD (24.2% vs. 18.3%, p = 0.5131; Table 1). Further, no differences were observed between patients with a mild course of GD between Germans and Poles. No clinical parameters from Serbian patients were available. Therefore, the correlation between the severity of GD and genetic data was limited to the German and Polish population.
All p-values are given uncorrected. A corrected p (pc) of <0.05 was considered significant.
Significant after correction.
The distribution of the rs35795399 polymorphism did not differ between Germans and Poles and was not associated to the severity of GD in any population. No correlations were found between thyroid antibodies or gender and the CXCL10 polymorphisms in any populations (data not shown).
Our findings, similar to other studies that investigated polymorphisms within the CXCL10 gene in diverse autoimmune-mediated diseases (16,17,19), did not find a direct association between polymorphisms within the CXCL10 gene and GDs in Germans, Poles, or Serbians. Nevertheless, in line with a recent study in MS, which demonstrated that although CXCL10 polymorphisms did not predispose or increase the risk to develop MS they modulated the progression of the disease (15,20), we found a correlation between the severity of the clinical course of GD and the rs8878 polymorphism in the German population. Thus, German patients with a severe course of the disease showed a lower frequency of the genotype “AA” compared with patients with a mild course. Additionally, the distribution of the rs8878 polymorphism differed between Germans and Poles with a severe course of GD. As we definitely exclude differences in the diagnosis of GD and the estimation of severity of the disease, the inversed ratio of the rs8878 genotype frequencies in Polish and German patients with a severe course led to the indisputable influence of not yet identified cofactors that modulate the effects of polymorphism.
As it is still unclear why some patients have severe or intractable forms of the disease while others do not (21,22), it is plausible that polymorphisms within the CXCL10 gene could modulate the attraction of activated T-lymphocytes to the thyroid gland and therefore have a direct influence on the severity of the course of GD. Given that this study is the first report investigating the role of the rs8878 polymorphism within the CXCL10 as possible marker to predict severity of GD, more investigations are required to elucidate the potential role of CXCL10 polymorphisms to predict the severity of GD.
Footnotes
Acknowledgment
E.R.-L. was financed by Else Kröner-Fresenius-Stiftung.