Lack of humoral response against Helicobacter pylori peptides homologous to human ZnT 8 in Hashimoto ’ s thyroiditis patients

Introduction: The Helicobacter pylori (HP) reinfection rate seems to be higher in developing countries than in developed ones. An increased seroprevalence of HP has also been reported in patients with type 1 diabetes (T1D) and Hashimoto’s thyroiditis (HT). Mycobacterium avium subsp. paratuberculosis (MAP) has been linked to both T1D and HT. Quite a few lines of evidence indicate that autoantibodies against several epitopes belonging to human zinc transporter 8 (ZnT8) cross-recognize the homologous MAP3865c epitopes in both T1D and HT patients. HP may play a role in HT disease, most likely acting through a molecular mimicry mechanism that targets ZnT8 as reported for MAP and the two autoimmune diseases. Methodology: An enzyme-linked immunosorbent assay (ELISA) has been developed for the detection of antibodies against several epitopes deriving from HP proteins, which are highly homologous to the immunodominant ZnT8 peptides previously identified: ZnT8178–186 and ZnT8186–194. Results: None of the HP peptides tested were significantly recognized when the humoral responses of 92 HT patients and 91 healthy volunteers were analyzed. Conclusions: These findings do not support a triggering role for HP (through ZnT8 mimicking) in HT. If a molecular mimicry phenomenon is taking place, it involves a different self-antigen. Moreover, the negative outcome of the experiments performed stresses the fact that sharing stretches of sequence homology is relevant, but not enough to trigger an antibody-mediated cross-recognition.


Introduction
Autoimmune diseases, such as Hashimoto's thyroiditis (HT) and type 1 diabetes (T1D), stem from the interaction of genetic susceptibility and environmental exposures [1][2][3][4].A distinctive marker of HT is the presence of autoantibodies (aAbs) against thyroglobulin (Tg), thyroperoxidase (TPO), and thyrotropin receptor (TR) [5].Infectious agents such as bacteria and viruses may trigger autoimmunity [6,7] through a molecular mimicry mechanism, being able to mimic the antigenic profile of the epithelial cell lining the thyroid gland.Among the environmental factors activated is Helicobacter pylori (HP), a chronic pathogen living in the human stomach, which typically colonizes and infects the gastric mucosa.This Gramnegative bacterium is responsible for gastritis, which can progress to gastric/duodenal ulcers and carcinomas.
The overall prevalence of HP infection is higher in developing countries compared to developed nations [8].
Contradictory conclusions have been reported for HP infection and HT disease, both confirming and denying the existence of an association [9,10].
What is more, the existence of a remarkable correlation between Mycobacterium avium subsp.paratuberculosis (MAP) infection/seroprevalence and autoimmune diseases such as T1D and HT was recently demonstrated [1].Researchers found a very strong humoral response mounted against epitopes belonging to one MAP membrane protein (MAP3865c), and against the homologous region present in human ZnT8, which is a relevant T1D autoantigen specifically expressed in the pancreatic βcell and in the cubical epithelium that lines thyroid follicles [1][2][3].
The reported increased prevalence of anti-MAP3865c and anti-ZnT8 antibodies (Abs) in Sardinian T1D and HT subjects indicates that MAP is a possible environmental trigger of both autoimmune diseases.MAP and HP may express proteins capable of mimicking self-antigens through a common mechanism, leading to the production of aAbs, which in turn may cause self-aggression.
To determine if this was true, and to identify in human ZnT8 the most likely common target for molecular mimicry in both HT and T1D, we investigated the prevalence anti-HP Abs developed against four HP peptides highly homologous to ZnT8 immunodominant (ZnT8 178-186 and ZnT8 186-194 ) epitopes in a Sardinian population cohort of 92 HT patient and 91 healthy volunteers.

Subjects
A group of 92 HT patients (female : male ratio, 10.75 : 1; mean age, 47 ± 13 years) followed at the University Hospital of Sassari, Department of Endocrinology, and a control group consisting of 91 age and sex-matched healthy individuals were enrolled in the study.All the subjects were inhabitants of Sardinia, Italy.All healthy controls (HCs) had not undergone therapy for at least six months and had no history of autoimmune diseases or clinically inflammatory diseases.All study subjects provided signed informed consent prior to the sample and data collection, and the study protocol was approved by the University Hospital of Sassari, Italy.

Enzyme-linked immunosorbent assay (ELISA)
Indirect ELISAs were performed in order to verify the humoral response mounted against the selected peptides as previously described [3].Briefly, ELISA 96-well Nunc immunoplates were coated with 10 μg/mL of peptides diluted in 0.05 Μ carbonatebicarbonate buffer, pH 9.5 (Sigma, St. Louis, USA) overnight at 4°C.The remaining protein-free sites in the plate were blocked by 200 μL/well of 5% phosphate-buffered saline (PBS)-non-fat dried milk (Sigma, St. Louis, USA) solution upon incubation for two hours at room temperature.After being washed twice with PBS containing 0.05% Tween-20 (PBS-T), serum samples at 1:100 dilution in 0.05% PBS-T were added and incubated for two hours at room temperature.Extensive washing of the plate was followed by addition of 100 μL of goat anti-human immunoglobulin G polyclonal Ab (1:1000; Sigma, St. Louis, USA) and incubation for one hour at room temperature.Finally, after five washes in PBS-T, the reaction was developed by adding 100 μL/well of para nitrophenylphosphate (Sigma, St. Louis, USA) as substrate for alkaline phosphatase.Plates were incubated in the dark for about 10 minutes, and optical density (OD) was measured at 405 nm using aVERSATunable Max microplate reader (Molecular Devices, Orleans Drive Sunnyvale, USA).Negative and positive control wells were included in each plate tested, and results are expressed as means of three different experiments performed.

Statistical analysis
Statistical analyses were performed using GraphPad 6 Software.Data was presented as mean ± standard error or median (25th and 75th percentiles) as clinical and laboratory data for each parameter analyzed.The results were analyzed by Fisher's exact test, and p ≤ 0.05 was considered to be significant.The cut-off values were determined by receiver operating characteristic (ROC) curve analyses using values normalized to a strongly positive control included in each essay performed.The specificity was set at > 93.4% (i.e., Ab+ HCs < 7%), and the sensitivity was calculated accordingly.
As shown in Figure 1, all selected peptides displayed the same Ab reactivity in the HCs group and very similar percent fraction of Ab-positivity in the HT one; in any case, there was no statistically significant difference between patients and controls.

Discussion
We investigated whether HP might play a role in HT disease, probably acting through a molecular mimicry mechanism which targets human ZnT8 protein as it was reported for MAP [1][2][3].We aimed to find an explanation accounting for the interconnection found between HP infection and two autoimmune diseases, HT and T1D [11], which have been proven to be associated with MAP infection [1][2][3].
To achieve this goal, we searched for Abs against four HP peptides (Table 1), which are highly homologous to two human ZnT8 immunodominant epitopes, ZnT8 178-186 and ZnT8 186-194 , the latter being highly recognized by the sera of HT and T1D subjects [1][2][3].
Therefore, we postulated that MAP and HP may express proteins capable of mimicking the same selfantigens ZnT8.Nevertheless, this study failed to find a direct association between HP seropositivity and HT.Indeed, there was no statistically significant difference in the seroprevalence of HP infection in Sardinian HT patients compared to control participants, which were matched for age, sex, and socioeconomic status.
Accordingly, we concluded that these HP epitopes are not correlated with HT.If a molecular mimicry phenomenon is taking place, it certainly targets Table 1.Alignment of the amino acid sequences of the identified peptides by ClustalW2.

Peptide name
Sequence alignment Protein different self-antigens.What is more, the negative outcome of the experiments performed stresses the fact that sharing stretches of sequence homology is relevant, but not enough to trigger an antibodymediated cross-recognition.

Conclusions
These findings do not support a triggering role for HP (through ZnT8 mimicking) in HT disease.Indeed, we cannot discount a direct involvement of HP in HT, which might act as an environmental trigger, targeting different self-antigens via molecular mimicry.To verify this hypothesis, further studies should seek and identify new possible sources of cross-reactive Abs; moreover, the genetic polymorphism of the host needs to be investigated.

Figure 1 .
Figure 1.Dot plot showing the distribution of IgG for the four Helicobacter pylori (HP) peptides measured by indirect ELISA