| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
CASE REPORT |
Research Laboratories, Center of Surgery, 1 Interdisciplinary Center for Clinical Research (IZKF), and 2 Institutes of Transfusion Medicine, 3 Pathology, 4 Legal Medicine and 5 Anatomy, University of Leipzig, Liebigstr. 20, Leipzig, 04103, Germany and 6 Leipzig Research Department, B.R.A.H.M.S AG, Berlin, Germany
(Correspondence should be addressed to G Aust; Email: ausg{at}medizin.uni-leipzig.de)
Abstract
Objective: To report on the rare simultaneous occurrence of Graves’ disease (GD) and Hashimoto’s thyroiditis (HT) in monozygotic twins.
Design: We compared the pattern of thyroid tissue-derived cDNAs to gain insight into previous and ongoing immune destruction and reconstruction processes using microarrays. The results were confirmed by immunohistology and real-time PCR.
Results: Destruction of thyroid tissue in HT reduced levels of thyrocyte-related cDNAs and cDNAs encoding extracellular matrix components, but increased levels of proteases involved in extracellular matrix degradation compared with GD. Lymphocytic infiltrates forming ectopic follicles replaced the thyroid tissue almost completely in HT. Thus, lymphocyte-related cDNA levels were higher in HT than in GD. The same was true for many chemokines and their receptors, which not only enable migration towards the thyroid but also maintain the lymphocytic infiltrate. HT also showed increased levels of cDNAs encoding molecules related to apoptosis than did GD. Surprisingly, the Th1- and Th2-specific cytokine profiles suggested for HT and GD respectively could not be confirmed. cDNAs encoding factors and receptors involved in angiogenesis were increased in GD compared with HT.
Conclusions: Comparison of gene expression reflects the cellular differences between the two types of autoimmune thyroid disease in twins with identical genetic and similar environmental background.
Introduction
Autoimmune thyroid disease (AITD) manifests itself in various clinical forms such as classical Hashimoto’s thyroiditis (HT) and Graves’ disease (GD). Histologically, HT and GD are both characterized by lymphocytic infiltration. However, this infiltration varies in quantitative as well as qualitative respects, possibly reflecting the different underlying pathogenic mechanisms of the two diseases (1). In HT, lymphocytic infiltrates replace thyroid follicles, resulting in a loss of hormone-producing cells and thus hypothyroidism. The follicles remain intact in GD. Here, stimulating autoantibodies against the thyroid-stimulating hormone receptor TSHR (thyroid stimulating antibodies, TSAbs) are the direct cause of hyperthyroidism.
Although GD and HT have very different phenotypes and the mechanisms leading to the dichotomy of GD and HT are unknown, they are generally believed to share a number of common etiological factors. GD and HT frequently occur in the same family; some individuals progress from one form to another. Whole-genome linkage studies on multiplex families with AITD have shown that multiple genes are responsible for a predisposition towards GD and HT, and that some are common to both diseases and some are unique (2). Although rare, there have been reports on monozygotic twins in which one twin had GD and the other had HT (3–6).
Studies on twins help in understanding the close relationship between various forms of AITD and reveal the genetic influence on its development (7–9). The relatively low concordance in monozygotic twins indicates that AITD requires non-heritable etiological determinants as well. Environmental and hormonal factors are shown to be associated with increased risk of developing AITD in genetically predisposed individuals (10; reviewed in Prummel et al. (11)).
In the present study, we have examined thyroid tissues excised from a pair of monozygotic twins, one of whom had GD and the other HT, and compared tissue-derived cDNA patterns by microarray to gain an insight into the past and present immune destruction and reconstruction processes in the thyroid. Since the thyroid tissues were removed on the same day, not only the genetic background but also age and processing of the thyroid were the same, which compensates for the disadvantage of analyzing only one typical case of HT compared with GD.
Materials and methods
Patients and assays
The monozygotic twin sisters were 18 years old on the day of surgery. GD and HT were diagnosed on the basis of clinical, biochemical and histological characteristics as well as by scintiscan. Thyroid dysfunction was confirmed by measurements of serum thyroid-stimulating hormone (TSH), free thyroxine (fT4) and free triiodothyronine (fT3). TSH binding inhibiting immunoglobulin (TBII) was determined using DYNOtest TRAK human, thyroglobulin (Tg) antibodies using DYNOtest anti-Tg, and thyroperoxidase (TPO) antibodies using DYNOtest anti-TPO (all from B.R.A.H.M.S AG, Hennigsdorf/Berlin, Germany) in sera from the patients (12). TSAb detection was carried out as described by Evans et al. (13). Bovine TSH (1 mU/ml, Sigma) was added for thyroid blocking antibody (TBAb) detection (14).
The thyroid tissues were removed from the twins on the same day. Twin B’s tissue sample was taken from the nodule surrounding tissue. The study was approved by the local Medical Ethics Committee.
Zygosity and HLA typing
Zygosity of the twins was evaluated by short tandem repeat (STR) typing using 16 unlinked autosomal microsatellite markers (PowerPlex16 System, Promega, Madison, WI, USA). PCR products were resolved and detected by capillary electrophoresis on an ABI Prism 3100 Avant Genetic Analyzer (Perkin Elmer, Foster City, CA, USA). Molecular typing of HLA-DRB1 and HLA-DQB1 was performed using the PCR sequence-specific primer (SSP) technique with commercial primer sets (Olerup SSP; GenoVision VertriebsgesmbH, Vienna, Austria) according to the requirements of the European Federation of Immunogenetics (EFI). HLA typing for Class I was performed by complement-dependent microlymphocytotoxicity testing.
Identification of lymphocytic infiltrates and germinal centers
Thyroid sections stained with hematoxylin and eosin (HE) were screened for focal lymphocytic infiltration (15). The ectopic follicle and germinal center areas (n = 50) were measured as described (1).
Microarray analysis
Total RNA was isolated from thyroid tissue using the RNeasy kit (Qiagen, Hilden, Germany). Ten micrograms were used to prepare ds cDNA (Superscript II, Life Technologies, Gaithersburg, MD, USA) primed with oligo-dT containing a T7 RNA polymerase promoter site (Genset SA, Paris, France). Phenol chloroform extraction was used to purify the cDNA before in vitro transcription using the ENZO BioArray RNA transcript labeling kit (Affymetrix, Santa Clara, CA, USA) to synthesize cRNA. The cRNA was fragmented and hybridized to Affymetrix GeneChip U95Av2. To detect differentially regulated genes, we selected genes that were characterized by the Affymetrix MAS5 software as ‘increased’ or ‘decreased’ and had a signal log ratio (SLR) 
1.5 or 
– 1.5 (that is, these genes are expressed at levels at least 2.828 times higher or lower in the GD than in the HT sample; arbitrarily chosen). If a certain gene is represented by more than one specific probe set on the chip, we displayed the probe set with the lowest SLR.
Real time RT-PCR
Superscript II RNaseH-reverse transcriptase (Invitrogen) was used to synthesize cDNA. Quantitative PCR was performed on a Rotorgene (Corbett Research, Mortlake, Australia) real-time machine by using SYBR Green I. The PCR primers for CXCR4, CXCL13, CD20, CD3, TPO, Tg, MMP9, and TIMP3 have been published elsewhere (15–18). The primer pair for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was sense 5'-GCT TTC AAT AGC ACC TTG CC-3', antisense 5'- CTG CTC ACG TCA TCA TCA GT -3'.
Transcripts were quantified in three tissue samples from each patient as described (15). Data are expressed as ratio-specific cDNA:GAPDH cDNA, which was calculated from the same cDNA sample. Finally, log (2) ratios of the GD:HT data were calculated.
Immunohistology
Serial sections of thyroid tissue were cut at 6 µm and fixed in ice-cold methanol for 10 min. The sections were incubated using the monoclonal antibodies Ki-67 (clone MIB1), CD45R0, CD8, CD22, laminin, collagen IV and vascular endothelial growth factor (VEGF) obtained from DakoCytomation GmbH (Hamburg, Germany), Acris Antibodies GmbH (Hiddenhausen, Germany), and Sigma-Aldrich Chemie GmbH (Taufkirchen, Germany). Following incubation with the antibody at 4 °C overnight, bound antibodies were detected using a super-sensitive detection kit (Dako Cytomation).
Results
Case report
The monozygotic twin sisters were born in 1984 after an uncomplicated pregnancy. They lived in the same house until the age of fourteen. Thus, for the last four years the twins may be discordant for a number of environmental risk factors such as iodine intake, number of negative life events, and infections. Both patients were smokers (10 cigarettes/day). There may have been a family history of AITD. In 2001, the mother was diagnosed with HT (TSH 50.0 mU/l, anti-TPO > 1000 U/ml, TBII negative). Information on other family members was not available.
After birth, goiter was diagnosed in one of the twins (twin A; GD), which was treated with iodine. At the age of 15, hyperthyroidism was diagnosed in this twin. Endocrine orbitopathy was diagnosed three years later. Twin A was treated with methimazole (20 mg/day). Patient compliance was poor, however. Since hyperthyroidism was diagnosed in twin A, attention was given to her twin sister (twin B, HT) who was found to have mild transient hypothyroidism. A cold nodule was detected by scintiscan, which was the indication for thyroid surgery. Twin A (GD) was strongly positive for autoantibodies against the TSH receptor, which had TSAb quality, whereas no TBII was detectable in the HT patient (Table 1
). Both patients’ sera showed very high levels of antibodies against TPO, and were positive for antibodies to Tg.
|
The short tandem repeat (STR) typing result confirmed monozygosity in the twin sisters examined. The 16 STR markers investigated yielded completely identical allele profiles (data not shown). The twins bore HLA-A 3,11, HLA-B 7, 35 (Bw6,-), and HLA-Cw 3,7. HLA-DRB1 *11, *15 and HLADQB1*03, *06 were determined by PCR-SSP.
Microarray analysis, immunohistology and real-time PCR
The two thyroid tissues showed differences in expression for a total of 2866 transcripts; 1088 showed higher expression in the GD patient, 1778 in the HT sample. Setting a cut-off at the signal log ratio, 522 cDNAs were increased in GD and 407 were increased in HT.
Thyrocytes
Thyrocytes remain intact in GD, while normal thyroid parenchyma disappears in HT (Fig. 1a). We found cDNAs from thyrocyte-related molecules such as the TSHR, TPO, Tg, and thyrocyte transcription factors 1 (TITF1) and 2 (FOXE1) or of molecules known to be present in thyrocytes such as type I iodothyronine 5' deiodinase (DIO1), at much lower levels - if present at all - in HT compared with GD tissue (Table 2). Real-time PCR results for Tg and TPO are shown as examples in Fig. 2.
|
|
|
The levels of cDNAs for several collagen and laminin chains, including collagen IV (COL4), which is a component of the basal membrane, were lower in HT compared with GD tissue (Table 2
). This result was confirmed by staining for collagen IV and laminin (Fig. 1g, h). On the other hand, the cDNA encoding matrix-metalloproteinase-9 (MMP9), which enables degradation of extracellular matrix especially collagen IV, was overexpressed in HT, whereas higher levels of TIMP3, a tissue inhibitor of MMPs, were expressed in GD. PCR results for MMP9 and TIMP3 are shown in Fig. 2. Thus, not only were thyrocytes destroyed in HT tissue, the basal membrane also disappeared (Fig. 1g, h).
Intrathyroidal lymphocytes
Lymphocytic infiltrates almost completely replaced thyroid tissue in HT (Table 1, Fig. 1a). Although lymphocytes formed ectopic follicles with germinal centers in both HT and GD tissue, they were higher in number and spread in HT than in GD tissue (Table 1). Thus, many transcripts specific for T-cells and B-cells or other molecules preferentially found in these cells, such as several cell surface receptors, transcription factors or cell signaling and activation molecules, were present at higher levels in HT than in GD tissue (data not shown). For example, only transcripts for the different CD3 and the T-cell receptor (TCR) chains are included in Table 2. As examples for increased lymphocytic accumulation in HT, the staining results for CD45R0, CD8 as a T-cell subpopulation, and CD22 as a B-cell marker are shown in Fig. 1b–d. PCR results for CD3 and CD20 are shown in Fig. 2.
Chemokines and chemokine receptors
The cDNA levels of many chemokines and their corresponding receptor(s) were much higher in HT compared with those in the GD thyroid (Table 3). For example, we found higher cDNA expression for CCR7 and its ligand, CCL19. Of the unique chemokine/chemokine receptor pairs CXCL13/CXCR5 and CXCL12/CXCR4, the transcript of one binding partner showed a signal log ratio 1.5 (CXCL13, CXCR4; Table 3). Real-time PCR results for CXCL13 and CXCR4 are shown as examples in Fig. 2.
|
There were no differences in interferon (IFN)-
cDNA levels between the thyroid samples. Interestingly, interleukin (IL)-4, IL-5, IL-10, IL-12p35, and IL-13 cDNA were absent in both thyroids. We only found differences between GD and HT tissue in some cytokine receptor cDNAs (Table 3). In most cases, these transcripts were present in the HT but not in the GD sample.
Adhesion molecules
The cDNAs of adhesion molecules involved in migration, localization and retention of lymphocytes into tissues such as CD62L (SELL), CD54 (ICAM1), and integrin 
4 (ITGA4) were present at higher levels in HT than in GD (Table 3). However, the cDNAs of adhesion molecules known to be involved in angiogenesis such as integrin ß3 (ITGB3), integrin ß5 (ITGB5), and ephrin A3 (EPHA3) were higher in GD than in HT.
Apoptosis
The cDNAs of many apoptosis-related genes such as the key regulator caspase 3 (CASP3) were more heavily expressed in HT compared with GD (Table 4). In contrast, the cDNAs of repressors of cell death such as Bcl-2-associated athanogene 1 (BAG1) were found at higher levels in the GD thyroid.
|
Higher levels of VEGF-A and VEGF-C cDNAs were present in GD thyroid tissue; this was confirmed by staining for VEGF (Fig. 1f). Similarly, we found a higher cDNA level of one of the VEGF receptors, FLT-1, in the GD thyroid. Transcripts of placental growth factor, another angiogenic stimulator, were also increased (Table 4).
Discussion
AITD is a multistep process initiated by the interaction of genetic and environmental factors. Thyroid samples obtained at surgery in most cases represent the chronic phase of AITD; lymphocytes were accumulated, ectopic lymphoid follicles had formed and, in the case of HT, thyrocytes had disappeared. However, maintenance of the lymphocytic infiltrate in AITD tissues is not static. The presence of activated T-cells in the thyroid (19, 20) and increasing variability of T-cell clones and auto-antibody affinity during the course of AITD (1, 21, 22), even after treatment, reflects the longstanding active immunological processes in the thyroid. Thus, comparison between HT and GD thyroid transcript patterns within identical genetic background and management may give an insight into different immune and remodeling processes. However, comparison between HT and GD transcript expression primarily reflects the varied cellular composition of the lesioned tissue. Thyrocytes were replaced by enlarged lymphocytic infiltrates in HT. Thus, thyrocyte-related transcripts were lower and lymphocyte-related cDNAs higher in HT than in GD. Thyrocytes remain intact in GD, which unsurprisingly showed higher levels of basal membrane component cDNAs than did HT. Increased apoptosis has been proposed as the mechanism for thyroid cell destruction in HT. Besides thyrocytes, lymphocytes located in the ectopic follicles undergo apoptosis during somatic mutation in their immunoglobulin variable-region genes to provide antibodies with higher affinity. Interestingly, we found higher cDNA levels for the major components of the TRAIL/Apo2L, caspase 8 and caspase 3 apoptosis signaling pathway in HT compared with GD. Thus, an alternative pathway killing cells in HT besides that initiated by the activation of the Fas death receptor by binding Fasligand (FasL), (reviewed in Salmaso et al. (23)), might be involved. Transcripts of chemokines/chemokine receptors involved in lymphocyte migration into the thyroid and in the maintenance of infiltrates were more heavily expressed in HT than in GD thyroid. We found higher cDNA levels for CCR7 and one of its ligands, CCL19. This chemokine, together with CCL21, enables migration of T-cells into secondary lymphoid tissues or, as in the case of thyroid tissue, into ectopic follicles. Moreover, we have confirmed recent data demonstrating higher expression of CXCL12 or its receptor CXCR4 as well as CXCL13 or its receptor CXCR5 in HT compared with GD thyroids (15, 18, 24) by microarray and real-time PCR. In addition to chemokines, classical cytokines may reflect ongoing immune processes in AITD. IL-2, INF- and tumor necrosis factor (TNF)--secreting Th1 cells stimulated by IL-12 were proposed to be involved in cell-mediated inflammatory reactions as seen in HT. A Th2-type cytokine secretion pattern characterized by IL-4, IL-5, IL-10 and IL-13 should prevail in GD. Here, the absence of many cytokine cDNAs in one or even both thyroid samples was surprising at first sight. However, the role of the Th1/Th2 balance in AITD is still controversial, especially after the publication of data from clinical studies and newly established mouse models (reviewed in Rotondi et al. (25), Dogan et al. (26), and Nagayama et al. (27)). Moreover, older studies have investigated cytokine cDNAs in thyroids using RT-PCR, which enables detection of traces of cytokine mRNAs (28–31). If low cytokine mRNA levels reflect protein expression of cytokines, their role during pathogenesis of AIDT remains unclear.
Hypervascularity and increased blood flow in the thyroid is a diagnostic feature in GD (32, 33). Micro-vascular density per follicle is significantly higher in GD tissues than in normal thyroid (34). The mechanism of blood vessel formation – angiogenesis – is modulated by both promoters and inhibitors. In our study, upregulation of the principal promoter of endothelial cell growth and migration, VEGF, and one of its receptors, Flt-1, was seen in the GD tissue. Our data agree with those of Nagura et al. (34).
Until now, there have only been a few studies on monozygotic twins with AITD; one of these twins had either hyperthyroidism or hypothyroidism (3–6, 35, 36). All twins were female. The diagnosis of AITD in older studies was based on measurements of thyroid hormones, and, in some cases, on fine-needle aspirates. The diagnosis of GD in two of the studies was confirmed by the presence of LATS (long-acting thyroid stimulator). In two pairs of twins, the GD patient was TSAb-positive, while TBAbs were detected in the patient with hypothyroidism (5, 6). In the case presented here, TBII was only detected in the GD patient, who showed TSAb but not TBAb activity. The HLA phenotype in our twins is not associated with any increased relative risk for Caucasians to develop AITD, although the diagnosis of HT in the mother suggests a family history of AITD. Increased frequency of HLA-DR3 (DRB1*03) and HLA-DQA1*0501 haplotypes has been demonstrated in Caucasian GD patients. The relative risk for people with DR3 is approximately 2–4 (37).
Altogether, comparison of equally managed thyroid tissue from monozygotic twins reflects the ongoing destruction and reconstruction processes within the thyroid at the cellular level. These processes are clearly different between the two clinical forms of AITD, GD and HT.
Acknowledgements
We would like to thank M Steinert and P Lamesch (formerly Center of Surgery, University of Leipzig; now at the Municipal Hospital Martha-Maria Halle-Dölau and St Georg Municipal Hospital Leipzig respectively) for providing thyroid tissues. The microarray analysis was conducted at the IZKF Leipzig core facility. We are grateful to P Süptitz, D Sittig and E Schmücking for their technical assistance. This work was supported by a grant from Deutsche Forschungsgemeinschaft (DFG/AU 132-2). K K is supported by IZKF Leipzig (project Z03).
References
This article has been cited by other articles:
|
|
 |
|
  N. Manji, J. D. Carr-Smith, K. Boelaert, A. Allahabadia, M. Armitage, V. K. Chatterjee, J. H. Lazarus, S. H. S. Pearce, B. Vaidya, S. C. Gough, et al. Influences of Age, Gender, Smoking, and Family History on Autoimmune Thyroid Disease Phenotype J. Clin. Endocrinol. Metab., December 1, 2006; 91(12): 4873 - 4880. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
