Comparison of DRB1/DQB1 alleles between patients and healthy controls
This study involved the analysis of HLA class II alleles, genotypes, and haplotypes in 100 patients with Hashimoto's Thyroiditis (HT) and 330 ethnically matched healthy individuals. Our findings are in line with similar studies conducted in Caucasian [33], Hungarian [34], and Italian [35] populations, which have highlighted the predisposing role of HLA-DRB1*03 and HLA-DRB1*04 alleles in the development of HT. Consistently, we observed that HLA-DRB1*03:01 and specific HLA-DRB1*04 subtypes, including HLA-DRB1*04:02 and HLA-DRB1*04:05, may be associated with an increased susceptibility to HT. Furthermore, our study revealed a potentially predisposing role for HLA-DQB1*02:01 and HLA-DQB1*03:02 alleles, along with a potential protective effect of HLA-DRB1*13:01 and HLA-DQB1*06:03 alleles in our HT patients. In contrast to a study by Zeitlin et al. [14], our analysis did not find a significant association for the HLA-DRB1*11 allele group. However, when examining specific DRB1*11 subtypes, we observed a susceptibility role for HLA-DRB1*11:04 alleles in our HT patients. Consistent with a study on the Greek population by Kokaraki et al. [36], our study demonstrated that HLA-DRB1*04:05, HLA-DQB1*02:01, and HLA-DQB1*03:02 alleles are associated with an increased risk of developing HT in Iranians. Additionally, our study identified more predisposing alleles, including DRB1*03:01 and DRB1*04:02, and suggested a possibly neutral effect of DRB1*07 alleles in our patients. These findings differ from the results of the study conducted in the Greek population [36].
Furthermore, our findings corroborate the results of Zeitlin et al.'s study [14] in terms of highlighting the predisposing effect of HLA-DRB1*11:04 ~ DQB1*03:01 and HLA-DRB1*03:01 ~ DQB1*02:01 haplotypes, as well as the marginally protective effect of the HLA-DRB1*13:01 ~ DQB1*06:03 haplotype on the development of HT. Among the major predisposing diplotypes in our patients, we observed DR3/DR4, DR3/DRX, and DR4/DRX diplotypes, where 'X' represents any alleles except for DRB1*03, *04, *11 and *13. Notably, within the DR3/DRX and DR4/DRX diplotypes, DR3/DR11 and DR4/DR11 were identified as two significantly predisposing diplotypes in our patients. Conversely, the potentially protective diplotype was DRX/DR13. Additionally, estimation of GRS and ROC curve analysis indicated that the risk alleles for HT in our population can be used for screening and improving the preventive actions among high risk individuals.
Insilco analysis
One of the probable hypotheses for the onset of autoimmunity is molecular mimicry [18]. Building upon this hypothesis, a potential connection has been demonstrated between coxsackie B virus infection and the initiation of type 1 diabetes [37]. Recent studies have also suggested that certain bacteria and some common human viruses might be associated with the onset of Hashimoto's thyroiditis [23, 27–30]. With this in mind, epitopes derived from four microorganisms, which have the capability to bind to the DRB1*03:01 molecule, a risk allele for HT, were analyzed for their similarity to epitopes derived from TPO and Tg, potential autoantigens in HT. Amino acid sequence alignment of the selected epitopes, based on their percentile rank from NetMHCІІpan-4.0, revealed that there is no sequence homology between epitopes derived from Tg and the microorganisms. Similarly, there was no sequence homology between epitopes derived from Helicobacter pylori and TPO. Based on three studies [27, 28, 38] that aimed to elucidate a potential link between Helicobacter pylori infection and the development of HT, it is noteworthy that despite the absence of sequence homology between the confirmed epitopes of Helicobacter pylori, TPO, and Tg, Helicobacter pylori may still contribute to HT through mechanisms other than molecular mimicry, such as bystander activation [39]. Therefore, we excluded the epitopes of Helicobacter pylori from the next step of our analysis.
The alignment of TPO-derived epitopes with epitopes from microorganisms (herpesvirus, enterovirus, and yersinia enterocolitica) revealed a common amino acid sequence at specific positions. Specifically, positions 5 and 8 were shared between TPO and Herpesvirus, residues 4 and 7 were common between TPO and Enterovirus, and residues 3 and 6 were shared between TPO and Yersinia enterocolitica. Based on these anchor residues, the epitopes were categorized into three distinct groups. Furthermore, in order to compare the presence of amino acids within these epitopes, sequence logo diagrams were created for both TPO-derived epitopes and the non-self-epitopes. These diagrams aligned with the results of the sequence alignment. Notably, aspartic acid was found to be predominant at the 6th, 7th, and 8th positions in both self and non-self-epitopes. Meanwhile, the 3rd, 4th, and 5th positions were primarily occupied by three amino acids: valine, leucine, and isoleucine.
Additionally, based on data extracted from UniProt, it was observed that TPO-derived epitopes, containing anchor residues L-D and I-D, may be involved in binding functions, such as binding to Ca, Heme b, and proton, within this protein. In contrast, epitopes consisting of V-D are derived from the structural portion of TPO. The presence of three I-D epitopes at critical positions of the TPO protein (associated with binding sites) suggests that these regions may be more susceptible to molecular mimicry compared to other epitopes in this protein. Furthermore, it's worth noting that Enteroviruses and Yersinia enterocolitica have epitopes derived from the genomic polyprotein and chaperonin GroEL, respectively. Similar to both Herpesviruses and Yersinia enterocolitica, Enteroviruses exhibit a higher number of I-D epitopes. Additionally, two types of glycoprotein epitopes, V-D and I-D, which are involved in mediating herpesvirus entry into host cells and play an immunogenic role, appear to be of significant importance.
When comparing the sequence homology between the TPO protein and the proteins from the four microorganisms, it was found that only the sequence 145–180 of the envelope glycoprotein D in Herpesviruses exhibited homology with the TPO sequence 151–199. Additionally, data extracted from UniProt revealed that one of the essential receptors for the entry of Herpesviruses into human cells is the CD160 (HVEM) molecule, which binds to the envelope glycoprotein D of Herpesviruses. A study conducted by Weiwei et al. [40] has demonstrated that polymorphisms in the CD160 receptor can influence the entry of Herpesviruses into thyroid follicles. Furthermore, an increased load of Herpesviruses within thyroid cells can enhance the presentation of epitopes from the envelope glycoprotein D via the predisposing HLA-DRB1*03:01 molecule, leading to the induction of immune responses. Consequently, due to the presence of homology between glycoprotein D and TPO, T cells activated by the envelop glycoprotein D may also recognize the TPO autoantigen as a foreign antigen, thus leading to the development of anti-TPO antibodies through a process known as molecular mimicry.
According to a study by Baker et al. [41], specific residues within the TPO sequence 151–199, including Leu 177, Gly 194, Leu 196, Asn 198, and Gly 199, bind to the Fab region of the anti-TPO antibody's L, H, H, L, and H chains, respectively. Based on these findings, it can be postulated that the homologous sequence 145–180 of the envelope glycoprotein D in Herpesviruses serves as a critical molecular mimic. This suggests that linear epitopes, such as IREDDQPSS and VTVDSIGML, from this protein may be presented to autoreactive T cells, leading to their activation. It's important to note that antibodies typically recognize conformational epitopes, whereas the studied epitopes in this research are linear. This brings to mind the concept of linked recognition, wherein B-lymphocytes recognize epitopes in their folded, conformational forms, and then present linear epitopes derived from the same antigen to induce T cell responses [42]. In this context, the other linear epitopes studied may also be presented to autoreactive T cells, potentially triggering autoimmune cellular responses.
Based on molecular docking analysis, we found that Asnβ82, Arg β74, Glnβ70, and Serα53 and Asnα62 amino acid residues of HLA-DR3 molecule are critical for binding to the anchor residues of V-D in the epitopes. Additionally, anchor residues of L-D in the epitopes bind to Lysβ71 and Argβ74, along with Asnα62 and Asnα69 of HLA-DR3 molecule. Moreover, Lysβ71, Argβ74, and Trpβ61, as well as Asnα62 and Asnα69, are important residues for binding to I-D residues in the epitopes. HLA-ІІ molecules possess nine binding pockets inside the binding cleft termed pockets P1–P9 [43]. P1, P4, and P9 binding pockets have key residues for binding to anchor residues of the antigenic peptides. Key residues of HLA-DR3 molecule are Asnβ82 and Serα53 in P1, Lysβ71, Arg β74, and Asnα62 in P4 and Trpβ61 and Asnα69 in P9. Peptides bind to HLA molecules via establishing non-covalent bonds such as hydrogen and electrostatic bonds. In the HLA-DR3 molecule the key residues in the binding pockets are polar and positively charged amino acids, while the peptide anchor residues are polar and negatively charged that leads to a strong non-covalent binding between peptide residues and HLA pockets to increase the binding affinity of HLA to the antigenic epitope. The P4 binding pocket was found to be more important than the other pockets, as the key amino acids (Lysβ71, Arg β74, and Asnα62) located in this pocket and primarily interact with aspartic acid which commonly found in the various studied peptides. Conversely, the key amino acids in the P1 and P9 pockets bind to different anchor residues of peptides. Any substitution with aspartic acid at this position affects the binding affinity of HLA molecule to the antigenic epitope. The Lysβ71 and Argβ74 primarily form hydrogen and electrostatic bonds with the aspartate at positions 6, 7 and 8 of the peptides which indicated the critical role of aspartate at those positions.
Based on the multiple epitope alignment, it was determined that the three microorganisms-derived epitopes (herpesvirus, enterovirus, and Yersinia enterocolitica), showed homology with TPO-derived epitopes. Also, based on docking, alignment epitopes indicated the capacity for binding to the predisposing HLA-DRB1*03 allele. Therefore, the observed similarities between proteins from those pathogens especially envelop glycoprotein D from Herpesviruses and TPO autoantigen might somewhat confirm the molecular mimicry hypotheses on development of HT.