Abstract
Objective
Hashimoto's thyroiditis (HT) is a common disease characterized by autoimmune injury of the thyroid. Its pathogenesis entails complex interactions among hereditary predisposition, immune disorders and environmental factors. In recent years, viral infection has attracted much attention as a potential environmental trigger, but the role genes associated with HT remain unclear.
Methods
In this study, COVID-19-related genes were combined with transcriptome data (GSE29315, GSE138198) of HT patients in the GEO database. Key genes were selected using machine learning (LASSO, SVM, and RF), GO/KEGG enrichment, and GSEA. Its function was confirmed by single-cell sequencing and ssGSEA immunoinfiltration analysis.
Results
A total of 16 co-expressed genes of HT and viral infection were identified. KEGG and GO enrichment results showed that these genes were significantly enriched in inflammatory signaling, viral defense and immune cell activation pathways. After screening by machine learning algorithm, four key genes (IFITM3, IFI44L, CCL3, OAS1) were finally identified as the common diagnostic markers of HT and viral infection, and the ROC curve also showed good diagnostic performance. In addition, single cell sequencing further confirmed its high expression in thyroid tissue and immune infiltrating cells.
Conclusion
A virus-triggered autoimmune cascade involving IFITM3, IFI44L, CCL3, and OAS1 may precipitate HT. These four genes constitute robust, multi-omics biomarkers for early diagnosis and targeted therapy of HT following viral infection.
Keywords: Hashimoto's thyroiditis, Viral infection, Bioinformatics, Machine learning
Highlights
-
•
Revealing a potential link between Hashimoto's thyroiditis and viral infection.
-
•
Thyroid function indicators should be monitored in patients with viral infection.
-
•
Four diagnostic biomarkers identified: IFITM3, IFI44L, CCL3, OAS1.
-
•
Combination of multiple omics.
1. Introduction
Hashimoto's thyroiditis (HT) is an autoimmune thyroid disease (AITD) marked by lymphocyte infiltration in the thyroid tissue, destruction of thyroid follicles, and increased levels of thyroid-specific autoantibodies [[1], [2], [3]]. As thyroid tissue continues to be damaged, patients with HT frequently develop either subclinical or overt hypothyroidism [4] and face a higher risk of metabolic disorders, cardiovascular issues [5], and cancers of the thyroid [6,7] and colon [8]. While research indicates that the development of HT is closely linked to genetic factors (such as polymorphisms in the HLA-DR, CTLA-4, and PTPN22 genes) [9], as well as environmental influences (like radiation and infections) [10] and epigenetic changes [11], the fundamental mechanisms of the disease are still not fully understood, particularly regarding how environmental factors disrupt immune tolerance in susceptible individuals.
Recent studies suggest that viral infection may play a “second strike” role in autoimmune thyroid disease. For example, Epstein-Barr virus (EBV) can trigger thyroid antigen cross-reaction by molecular mimicry [12,13]. Hepatitis C virus (HCV) infection can lead to the presence of thyroid autoantibodies and disrupt thyroid function [[14], [15], [16]]. The COVID-19 pandemic has further underscored the link between viral infection and thyroid dysfunction [17], with patients suffering from severe COVID-19 being 3.77 times more likely to experience thyroid dysfunction compared to those with mild cases [17,18]. SARS-CoV-2 can directly infect thyroid cells via ACE2 receptors or cause cytokine storms that indirectly harm the thyroid [19]. Additionally, human papillomavirus (HPV) and parvovirus B19 have been found in the thyroid tissue of patients with HT [20], indicating that local latent viral infections may contribute to the disease's progression.
To date, several biomarkers associated with HT have been identified, including the oxidative stress index and increased levels of inflammatory factors [[21], [22], [23], [24]]. By analyzing the GSE29315 dataset, Zheng et al. discovered ten significant genes linked to HT and suggested that pathways involving IFN-γ, IFN-α, IL6/JAK/STAT3, and inflammation may play a role in the progression of HT [25]. Further research utilizing CMap analysis indicates that SRRM1, a splicing factor responsive to viral activity, could be a potential marker for HT [10]. However, most current research is primarily focused on transcriptome-level analysis, and there has not been a comprehensive examination of the dynamic regulatory network of key virus-related genes in HT and their interactions with the immune microenvironment.
This study will integrate multiple omics data and machine learning algorithms to uncover key genes that trigger HT in viral infections. The datasets of HT patients were obtained from the Gene Expression Omnibus (GEO) [26], and differentially expressed genes (DEG) were detected using the Limma package. Co-expressed genes were identified by integrating DEG with genes related to viral infections. Lasso, SVM, and RF models were employed to further identify diagnostic markers for HT induced by viral infections. We also evaluated the levels of immune cell infiltration using ssGSEA. Lastly, single-cell analysis was performed to determine the locations of common diagnostic markers.
2. Methods
2.1. Data source
Two datasets [GSE29315, GSE138198] were downloaded from the GEO database [27] (https://www.ncbi.nlm.nih.gov/geo/) to obtain microarray expression data of thyroid tissues of HT patients and controls. In this study, 8 cases of thyroid hyperplasia (TPH, as controls) and 6 HT samples from GSE29315 were selected. The GSE138198 dataset consisted of 3 controls and 13 HT samples. The detailed information of these datasets was summarized in Table 1. COVID-19 genes from Disgenet database (https://disgenet.com/) and Genecard database (score >10, https://www.genecards.org/). The expression profiles of other viral infections were derived from GSE157240 and detailed information can be found in Table 2. The GSE29617 dataset presents the expression profiles of peripheral blood mononuclear cells in healthy adults before and after influenza vaccination. The detailed grouping information is summarized in Table 3.
Table 1.
Information on datasets.
| Items | GSE29315 | GSE138198 | GSE163203 |
|---|---|---|---|
| Platform | GPL8300 | GPL6244 | GPL20795 |
| Sequencing type | Expression profiling by array | Expression profiling by array | Single-cell sequencing |
| Species | Homo sapiens | Homo sapiens | Homo sapiens |
| Disease | HT | HT | Thyroid carcinoma |
| Tissue | Thyroid | Thyroid | thyroid carcinoma/normal adjacent tissue |
| Samples in the disease group | 6 | 13 | 8 |
| Samples in the control group | 8 | 3 | 2 |
| Reference | Missing | 32603365 | 34805166 |
Table 2.
Detailed information of the GSE157240 dataset.
Table 3.
Detailed information of the GSE29617 dataset.
2.2. Identification of DEG
Based on the two datasets [GSE29315, GSE138198] of control and HT thyroid tissue samples, a post-correction batch effect pooled analysis was performed using the sva R (version 3.52.0) package. Limma R package [28] (version 3.62.1) was used to identify DEG. The threshold of statistical significance was set as |logFC| >0.585 and adjusted p-value < 0.05. To visualize the representation of DEG, the volcano map is drawn using the “ggplot2” package (version 3.5.1) and the heatmap is generated by the “pheatmap” package (version 1.0.12). Finally, overlapping differentially expressed genes in COVID-19 and DEG were obtained using the ggvenn R package (version 0.1.10).
2.3. GO and KEGG pathway enrichment analyses
The GO and KEGG [29] pathways of differentially expressed genes of HT and COVID-19 were analyzed using the ClusterProfiler R package (version 4.12.1) to analyze the biological processes and key pathways involved in differentially expressed genes [30]. There are three main processes for GO analysis: biological process (BP), molecular function (MF), and cellular component (CC) [31]. The term adjusted p-value < 0.05 in both analyses was considered statistically significant.
2.4. Gene set enrichment analysis (GSEA)
GSEA is a microarray data analysis tool for analyzing biological information [32]. In this study, GSEA analysis was performed for all ranked genes using GseaVis (version 0.0.5) and fgsea package (version 1.30.0). The pathways with adjusted p-value < 0.05 are considered as significant signaling pathways.
2.5. Machine learning
In order to dig into the hub genes of HT caused by COVID-19, we used three machine learning algorithms of LASSO, SVM and RF for gene screening, using glmnet (version4.1-8), e1071 (version1.7-14) and randomForest (version 4.7-1.1) packages respectively [33]. Finally, we identified four overlapping genes as hub genes (IFITM3, IFI44L, CCL3, OAS1), which were validated by the ROC curve (pROC R package, version 1.18.5).
2.6. Immunoinfiltration and correlation analysis
ssGSEA is a method commonly used in the analysis of immune cell infiltration [34,35]. Using the ssGSEA algorithm in the GSVA package (version 1.52.3), RNA-seq data downloaded from the GEO database were analyzed to calculate immune infiltration for 28 immune cell markers. Finally, the correlation between the four key genes and these immune cells was analyzed.
2.7. Identification of drug candidates
Enrichplot package (version 1.24.2) and clusterProfiler package were used to analyze key genes. Then, we use the drug signature database (DSigDB, https://dsigdb.tanlab.org/DSigDBv1.0/) [36]screening hub genes related to drug candidate. Finally, Cytoscape software (version 3.10.3) was used to construct the gene-drug interaction network.
2.8. Single-cell RNA sequencing data processing
We downloaded a single-cell dataset from GSE163203 that included 8 papillary thyroid cancer samples and 2 adjacent tissues (five samples from PTC patients without concurrent HT (PTC without HT), three samples from PTC patients with concurrent HT (PTC with HT), and two paired samples from normal adjacent tissues (Normal). The detailed information of this dataset can be found in Table 1. The preprocessing of single-cell transcriptome data followed the method described in the previous literature [37], using the seurat package (version 4.2.1). In short, quality control is performed by filtering cells with nCount RNA>1000, nFeature_RNA >200, nFeature_RNA <10000, percent.MT < 20, and percent.RB < 20.
3. Results
3.1. Identification of DEG and co-expression network construction
This study integrated two datasets, GSE29315 and GSE138198, comprising a total of 11 control samples (Control) and 19 HT samples (Treat). Considering differences between the two datasets, we performed batch effects correction (Fig. 1A). Based on the corrected data, a total of 754 DEG were identified using the limma R package. Out of these DEG, 473 genes were found to be upregulated while 281 genes were found to be downregulated in HT. A volcano plot was employed to visualize the distribution of these DEG (Fig. 1B), and the expression patterns of the top 50 up-regulated and down-regulated genes were depicted using heatmaps generated by the pheatmap R package (Fig. 1C). Subsequently, COVID-19-related genes were retrieved from two authoritative databases, DisGeNET and GeneCards. Intersection analysis of the two databases identified 65 consensus genes associated with COVID-19（Fig. 1D）. Comparative analysis between the DEG cluster and the COVID-19-related gene revealed 16 overlapping genes (Fig. 1E).
Fig. 1.
Identification of DEG and co-expression network construction. (A)Principal Component Analysis (PCA) of GSE138198 and GSE293135 datasets before and after batch correction. (B) Volcano map of DEG. (C) Heatmap of DEG. (D) Venn diagram of COVID-19 related genes from two different websites. (E) Venn diagram between DEG and COVID-19 genes.
3.2. Functional enrichment study revealed an association between HT and COVID-19
In our quest to unravel the underlying molecular mechanisms of HT, we conducted an in-depth analysis of overlapping genes of HT and COVID-19, including GO and KEGG analyses. GO is divided into the following three parts: BP, MF and CC (Fig. 2A). In BP and MF, these differential genes are mainly concentrated in leukocyte activation, signal transduction of cell surface receptor connections, regulation of lymphocyte chemotaxis, and production of tumor necrosis factor. These genes are also enriched in the outer and extracellular regions of the plasma membrane. We then turned our attention to KEGG pathway analysis, which mainly includes viral protein interaction with cytokine and cytokine receptor, cytokine−cytokine receptor interaction, influenza A, Epstein-Barr virus infection, rheumatoid arthritis, IL-17 and TNF signaling pathways, measles, herpes simplex virus 1 infection, and yersinia infection (Fig. 2B). Overall, the function of these differential genes was significantly associated with immune response-related pathways and viral infection.
Fig. 2.
Functional and pathway enrichment analysis. (A) Bubble plot of GO analysis of intersection genes between HT and COVID-19. (B) Bar plot of KEGG pathway of HT and COVID-19 intersection genes. (C) GSEA of enriched pathways in HT.
3.3. GSEA enrichment analysis
We performed GSEA enrichment analysis to further explore the key pathways that influence HT. Through the GseaVis and fgsea packages, based on normalized enrichment scores (NES), we selected the top five most significantly enriched signaling pathways, as shown in Fig. 2C. GSEA analysis confirmed that viral protein interaction with cytokine and cytokine receptor, hematopoietic cell lineage, staphylococcus aureus infection, natural killer cell mediated cytotoxicity, and cytokine - cytokine receptor interaction pathways play an important role in the pathogenesis of HT.
3.4. Construction of the lasso model, SVM model, and RF model
We constructed three algorithms to choose potential genes affecting HT in COVID-19 infection from a total of 16 differential genes. The results of the lasso model showed that HT induced by COVID-19 infection was linked to the expression of six genes, including IFITM3, IFI44L, CCL3, OAS1, IL1B, and BCL11A (Fig. 3A). In the course of our research, we also used a powerful machine learning technique called Support Vector Machine (SVM) to generate feature vectors from the collected data. Through this complex set of data filtering and analysis, we revealed seven of the most important genes strongly associated with HT. These genes identified by SVM, as shown in Fig. 3B, have great potential for further study and characterization. In the random forest algorithm, 9 marker genes with good prediction effect were screened, which were CD84, BCL11A, CCL2, CCL3, OAS1, IFITM3, IFI44L, IFI44 and BTK (Fig. 3C).
Fig. 3.
Construction of the lasso model, SVM model, and RF model. (A) Coefficient path diagram of Lasso regression model. (B) A SVM approach for the selection of feature genes. (C) The correlation between total number of trees and error rate in the random forest. Sequence of genes based on their relative importance.
3.5. Identification of characteristic genes
We analyzed the genetic data using three machine learning models and finally obtained four key genes through intersection screening: IFITM3, IFI44L, CCL3, and OAS1 (Fig. 4A). Further analysis revealed that the expression levels of these genes were significantly increased in the thyroid tissues of HT patients compared to the control group (Fig. 4B), suggesting that they may play an important role in the development of HT. To gain a deeper understanding of the relationships among these key genes, we constructed a correlation heatmap (Fig. 4C). The results indicated a strong positive correlation between IFI44L and OAS1, suggesting that they may act synergistically in the pathological process of HT. Additionally, to further clarify the chromosomal distribution of the aforementioned genes, we visualized the co-expressed genes (Fig. 4D).In terms of validating the diagnostic efficacy of the key genes, we constructed a prediction model based on these four genes and used the area under the ROC curve (AUC) as the evaluation metric (Fig. 4E). Notably, the CCL3 gene exhibited an AUC value as high as 0.971, indicating its excellent reliability in differentiating HT patients from healthy controls. Considering the AUC values of all genes, our prediction model demonstrated high diagnostic performance (Fig. 4F), providing potential biomarkers for the early diagnosis of HT. Moreover, calibration curves and decision curve analysis (DCA) further confirmed the clinical diagnostic value of these genes in predicting HT related to COVID-19 infection (Fig. 4G and H). We comprehensively incorporated all the risk factors of the key genes to construct the prediction model (Fig. 4I), ensuring the completeness and accuracy of the model.
Fig. 4.
Identification of characteristic genes. (A) Venn diagram showing characteristic genes shared by LASSO, SVM, and RF. (B) Expression of key genes in HT (Treat) and control samples. (C) Correlation heatmap of key genes. (D) Circos plot of co-expressed genes. (E) ROC curves show the effectiveness of four key genes as diagnostic tools. (F) Evaluation prediction model of the ROC. (G) Evaluate the calibration curve of the predictive model. (H) The decision curve analysis of the prediction model. (I) Nomogram of gene expression levels predicting HT risk.
3.6. Immune cell infiltration and drug candidate prediction
In the above functional cluster analysis, it was shown that the activation of immune cells was closely related to the progression of HT. ssGSEA algorithm was used to infer immune cell characteristics and explore the correlation between four key genes in immune cell infiltration. We observed that compared with the control group, the proportion of 21 kinds of immune cells in HT group was significantly increased (Fig. 5A), and the association analysis with 28 immune cells showed that the co-expressed gene OAS1 was positively correlated with most immune cells (Fig. 5B). In addition, we also searched for drugs that may act on these four key genes, among which acetohexamide, mephentermine and other drugs are more sensitive (Fig. 6A and B).
Fig. 5.
Immune cell infiltration. (A) ssGSEA algorithm analyzed 28 kinds of immune cells. (B) Correlation matrix between key genes and multiple immune cell types.
Fig. 6.
Drug candidate prediction. (A) Bar chart of drug candidates. (B) Map of the network of associations between different drugs and genes.
3.7. Single-cell sequencing reveals the location of co-expressed genes
The GSE163203 single-cell RNA sequencing dataset was obtained from the GEO database, comprising five samples from PTC without HT, three samples from PTC with HT, and two normal samples. Following stringent quality control procedures (Fig. S1), cellular clustering was performed based on clustertree analysis, and the integrated single-cell transcriptomic profiles were visualized using UMAP dimensionality reduction (Fig. 7A and B). Cell type annotation using established markers broadly identified six major populations: epithelial cells, B cells, dendritic cells (DC), macrophages, monocytes, and endothelial cells (Fig. 7C), with their relative proportions across samples quantified (Fig. 7D). Notably, key gene expression analysis showed relatively high expression levels of IFITM3 and CCL3 in six major cell clusters (Fig. 7E and F). To further characterize the spatial expression patterns of key genes, we performed UMAP visualization of 1536 cells, highlighting the expression profiles of IFITM3, IFI44L, CCL3, and OAS1 (Fig. 8A and B). Given that the GSE163203 dataset contains three groups of normal tissues, PTC with HT, and PTC without HT, we systematically analyzed the expression characteristics of key genes in different cell clusters (Fig. 8C) and different groups (Fig. 8D). The results showed that IFITM3 and CCL3 were not only dominant in cell clusters, but also significantly up-regulated in thyroid tissue of HT samples. In summary, single-cell transcriptome analysis revealed the spatial distribution characteristics and expression dynamics of these four co-expressed genes in thyroid tissue from the cell resolution level, offering valuable perspectives for in-depth understanding of their biological functions in HT.
Fig. 7.
The definition of cell clusters. (A) Resolution ranged from 0.01 to 3 for Seurat clustering results. (B) UMAP of different cell subpopulations. (C) UMAP dimensionality reduction cluster diagram. (D) The proportions of the six major cell clusters in different samples. (E) The dotplot of key genes expression in different cell subsets. (F) The featureplot of key genes expressed in different cell subsets.
Fig. 8.
The location of the co-expressed gene. (A) UMAP showing the total number of cells and the number of cells in each subpopulation. (B) UMAP of cell density. (C) Expression profiles of key genes in different cells. (D) Expression profiles of key genes in different groups.
3.8. Validation of candidate genes
Based on the above results, we confirmed four candidate genes related to HT caused by COVID-19 infection. To further verify the prevalence of these candidate genes in patients with different viral infections, we analyzed the GSE157240 dataset. This dataset contains peripheral blood expression profiles of patients infected with dengue virus, DNA viruses (adenovirus, cytomegalovirus, Epstein-Barr virus and herpes simplex virus), enterovirus/rhinovirus, influenza virus, Metapneumovirus, and other respiratory RNA viruses (parainfluenza virus and respiratory syncytial virus), as well as healthy adults. We compared the transcriptome data of different viral pathogens with those of healthy adults respectively. The results showed that the expressions of IFITM3, IFI44L and OAS1 were significantly upregulated in all viral infection groups compared with healthy adults (Fig. 9A–F), indicating that they have broad reactivity to viral infections. In addition, we also analyzed the possible impact of vaccination on the expression of four candidate genes. In the GSE29617 dataset, there are sequencing data of peripheral blood mononuclear cells from 28 healthy adults on the 0th day (before vaccination), the 3rd day and the 7th day after vaccination against influenza. Fig. 9G revealed the expression profiles on day 0 and day 3 of vaccination, and no differences were found. However, on the 7th day after vaccination, it can be observed that the expression of CCL3 has significantly decreased (Fig. 9H), indicating that the vaccination has a certain protective effect on the virus-infected body. In conclusion, we believe that these four candidate genes can be used to identify HT patients infected with the virus and serve as targets for future treatment.
Fig. 9.
Validation of key genes. (A–F) Expression analysis of key genes in blood samples of dengue virus infection(A), DNA virus infection(B), Entero/Rhinovirus infection(C), Influenza virus infection(D), Metapneumovirus infection(E), Other respiratory RNA virus infection(F) and healthy patients. (G) Expression analysis of candidate genes on day 0 (pre-vaccination) and day 3 after influenza vaccination. (H) Expression analysis of candidate genes on day 0 and day 7 after influenza vaccination.
4. Discussion
In this study, by integrating multiple omics data with machine learning algorithms, we systematically revealed the potential mechanism of viral infection-related genes in HT. We identified four key genes (IFITM3, IFI44L, CCL3, OAS1) and found that their expression patterns were closely related to immune microenvironment disorders and viral infection pathways in HT patients. These findings provided new insights into the molecular mechanism of HT and laid a solid foundation for the development of targeted intervention drugs.
A total of 754 HT differential genes overlapped with 65 COVID-19 related genes, and 16 differential genes were obtained. KEGG signaling pathway indicated that these differential genes were mainly enriched in viral protein-cytokine interaction, cytokine-cytokine interaction, IL17 and TNF signaling pathways. Studies have shown that the presence of some viral antibodies in HT patients may mimic the occurrence of an abnormal immune response [38]. IL17 and TNF, as pro-inflammatory cytokines, are also believed to be involved in the progression of AITD in the current study. The explosion of inflammatory factors can not only up-regulate the transcriptional turnover of nuclear factor-κB (NF-κB), Janus kinase/signal transduction and transcriptional activator (JAK/STAT) and mitogen-activated protein kinase (MAPK), but also stimulate the production of autoantibodies against thyroid antigen, resulting in increased thyroid tissue damage [22,39]. However, comprehensive studies are still needed to elucidate the key genes of viral infection and immunoinflammation in the development of HT.
The hub genes screened in this study were all highly associated with antiviral immunity and inflammatory responses. Interferon Induced Transmembrane Protein 3 (IFITM3) is a key effector molecule downstream of interferon signal, which can inhibit the fusion of virus envelope and host cell membrane, and plays an important role in EBV and SARS-CoV-2 infection [40]. Its significant upregulation suggests that sustained interferon response in HT may promote autoantigen exposure by enhancing the immunogenicity of thyroid cells. Interferon (IFN)-induced protein 44-like (IFI44L) gene has also been associated with respiratory viral infection and is involved in the innate immune response induced after viral infection [41,42]. Similarly, OAS1 (2′-5' -oligoadenylate synthetase 1) is a central gene in RNA virus defense, and its activation can lead to apoptosis and the release of inflammatory factors [43], which may exacerbate thyroid follicular destruction. In addition, chemokine CCL3 has been shown to play a key role in the pathogenesis of HT in organoid models [44]. Notably, the expression patterns of these genes were further validated in single-cell sequencing data, especially in thyroid epithelial cells and infiltrating immune cells, suggesting that the virus may be involved in HT process by directly infecting the thyroid gland or indirectly activating the systemic immune response.
In summary, this study for the first time explored and identified the core genes of HT and viral infection, and analyzed the possible pathogenesis. Four key genes, IFITM3, IFI44L, CCL3, OAS1, may serve as potential biomarkers. However, there were some limitations to this study. First, the sample size was small, and future studies need a larger dataset to verify the existing results. Secondly, thyroid cancer is also associated with HT patients, which brings difficulties to the verification of core genes. In the future, we need to collect thyroid tissue samples from patients with initial HT, and increase in vivo and in vitro experiments to further verify the functional mechanism of these key genes in HT.
5. Conclusion
In conclusion, the molecular cross network between HT and viral infection was revealed through multi-dimensional analysis, and the hypothesis of “virus-triggered autoimmune cascade” was proposed. In addition, we established four pivotal genes as characteristic diagnostic markers, providing potential targets for early diagnosis and immunotherapy of HT. These findings also suggest that viral infection is the key factor triggering HT, and thyroid function indicators should be monitored in patients with viral infection.
CRediT authorship contribution statement
Yuhan Zhang: Writing – original draft, Visualization, Software, Investigation, Formal analysis. Hanyu Wang: Methodology. Mengfei Fu: Methodology, Investigation. Liu Yang: Investigation. Lu Yu: Investigation. Xiao Chen: Investigation. Siqi Wang: Investigation. Yu Wang: Investigation. Zixuan Wang: Investigation. Jiaqi Liu: Investigation. Hui Sun: Writing – review & editing, Supervision, Resources, Funding acquisition, Conceptualization.
Availability of data and materials
The datasets generated during and/or analyzed during the current study are available in the GEO database (https://www.ncbi.nlm.nih.gov/geo/).
Ethics approval and consent to participate
Not applicable.
Consent to publication
All the authors consent to publish this manuscript.
Funding
This work was supported by the National Natural Science Foundation of China (No. 82270832 to Hui Sun), Bethune Charitable Foundation (No. Z04JKM2022E036 to HanyuWang), the Wuhan Knowledge Innovation Project, Grant (No. 2023020201010161), the Principal Investigator is Hui Sun, and the Technology Innovation Project of Hubei Province, Grant (No. 2023BCB131), the Principal Investigator is Hui Sun.
Declaration of competing interest
All authors confirm that there are no conflicts of interest, including but not limited to financial, commercial, or professional affiliations, related to the research design, data interpretation, or conclusions presented in this study.
Acknowledgements
Not applicable.
Handling Editor: Dr Y Renaudineau
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.jtauto.2025.100302.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
Data availability
Data will be made available on request.
References
- 1.Ralli M., Angeletti D., Fiore M., et al. Hashimoto's thyroiditis: an update on pathogenic mechanisms, diagnostic protocols, therapeutic strategies, and potential malignant transformation. Autoimmun. Rev. 2020;19(10) doi: 10.1016/j.autrev.2020.102649. [DOI] [PubMed] [Google Scholar]
- 2.Zhang Q.Y., Ye X.P., Zhou Z., et al. Lymphocyte infiltration and thyrocyte destruction are driven by stromal and immune cell components in Hashimoto's thyroiditis. Nat. Commun. 2022;13(1):775. doi: 10.1038/s41467-022-28120-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ott J., Meusel M., Schultheis A., et al. The incidence of lymphocytic thyroid infiltration and Hashimoto's thyroiditis increased in patients operated for benign goiter over a 31-year period. Virchows Arch. 2011;459(3):277–281. doi: 10.1007/s00428-011-1130-x. [DOI] [PubMed] [Google Scholar]
- 4.Petranović Ovčariček P., GöRGES R., Giovanella L. Autoimmune thyroid diseases. Semin. Nucl. Med. 2024;54(2):219–236. doi: 10.1053/j.semnuclmed.2023.11.002. [DOI] [PubMed] [Google Scholar]
- 5.Chen W.H., Chen Y.K., Lin C.L., et al. Hashimoto's thyroiditis, risk of coronary heart disease, and L-thyroxine treatment: a nationwide cohort study. J. Clin. Endocrinol. Metab. 2015;100(1):109–114. doi: 10.1210/jc.2014-2990. [DOI] [PubMed] [Google Scholar]
- 6.Jiang H., He Y., Lan X., et al. Identification and validation of potential common biomarkers for papillary thyroid carcinoma and Hashimoto's thyroiditis through bioinformatics analysis and machine learning. Sci. Rep. 2024;14(1) doi: 10.1038/s41598-024-66162-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Li B., Cai Z., Zhang Y., et al. Biomarkers associated with papillary thyroid carcinoma and Hashimoto's thyroiditis: bioinformatic analysis and experimental validation. Int. Immunopharmacol. 2024;143(Pt 3) doi: 10.1016/j.intimp.2024.113532. [DOI] [PubMed] [Google Scholar]
- 8.Chen Y.K., Lin C.L., Cheng F.T., et al. Cancer risk in patients with Hashimoto's thyroiditis: a nationwide cohort study. Br. J. Cancer. 2013;109(9):2496–2501. doi: 10.1038/bjc.2013.597. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Zhao Z., Gao Y., Pei X., et al. Causal role of immune cells in Hashimoto's thyroiditis: Mendelian randomization study. Front Endocrinol (Lausanne) 2024;15 doi: 10.3389/fendo.2024.1352616. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lim D.W., Choi M.S., Kim S.M. Bioinformatics and connectivity map analysis suggest viral infection as a critical causative factor of Hashimoto's thyroiditis. Int. J. Mol. Sci. 2023;24(2) doi: 10.3390/ijms24021157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Davies T.F. Infection and autoimmune thyroid disease. J. Clin. Endocrinol. Metab. 2008;93(3):674–676. doi: 10.1210/jc.2008-0095. [DOI] [PubMed] [Google Scholar]
- 12.Robinson W.H., Younis S., Love Z.Z., et al. Epstein-Barr virus as a potentiator of autoimmune diseases. Nat. Rev. Rheumatol. 2024;20(11):729–740. doi: 10.1038/s41584-024-01167-9. [DOI] [PubMed] [Google Scholar]
- 13.Janegova A., Janega P., Rychly B., et al. The role of Epstein-Barr virus infection in the development of autoimmune thyroid diseases. Endokrynol. Pol. 2015;66(2):132–136. doi: 10.5603/EP.2015.0020. [DOI] [PubMed] [Google Scholar]
- 14.Fallahi P., Ferrari S.M., Vita R., et al. The role of human parvovirus B19 and hepatitis C virus in the development of thyroid disorders. Rev. Endocr. Metab. Disord. 2016;17(4):529–535. doi: 10.1007/s11154-016-9361-4. [DOI] [PubMed] [Google Scholar]
- 15.Colaci M., Malatino L., Antonelli A., et al. Endocrine disorders associated with hepatitis C virus chronic infection. Rev. Endocr. Metab. Disord. 2018;19(4):397–403. doi: 10.1007/s11154-018-9475-y. [DOI] [PubMed] [Google Scholar]
- 16.Shen Y., Wang X.L., Xie J.P., et al. Thyroid disturbance in patients with chronic hepatitis C infection: a systematic review and meta-analysis. J Gastrointestin Liver Dis. 2016;25(2):227–234. doi: 10.15403/jgld.2014.1121.252.chc. [DOI] [PubMed] [Google Scholar]
- 17.Lui D.T.W., Lee C.H., Woo Y.C., et al. Thyroid dysfunction in COVID-19. Nat. Rev. Endocrinol. 2024;20(6):336–348. doi: 10.1038/s41574-023-00946-w. [DOI] [PubMed] [Google Scholar]
- 18.Darvishi M., Nazer M.R., Shahali H., et al. Association of thyroid dysfunction and COVID-19: a systematic review and meta-analysis. Front Endocrinol (Lausanne) 2022;13 doi: 10.3389/fendo.2022.947594. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zhang Z., Fang T., Lv Y. Causal associations between thyroid dysfunction and COVID-19 susceptibility and severity: a bidirectional Mendelian randomization study. Front Endocrinol (Lausanne) 2022;13 doi: 10.3389/fendo.2022.961717. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Cyna W., Wojciechowska A., Szybiak-Skora W., et al. The impact of environmental factors on the development of autoimmune thyroiditis-review. Biomedicines. 2024;12(8) doi: 10.3390/biomedicines12081788. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ates I., Arikan M.F., Altay M., et al. The effect of oxidative stress on the progression of Hashimoto's thyroiditis. Arch. Physiol. Biochem. 2018;124(4):351–356. doi: 10.1080/13813455.2017.1408660. [DOI] [PubMed] [Google Scholar]
- 22.Attiq A., Afzal S., Wahab H.A., et al. Cytokine storm-induced thyroid dysfunction in COVID-19: insights into pathogenesis and therapeutic approaches. Drug Des. Dev. Ther. 2024;18:4215–4240. doi: 10.2147/DDDT.S475005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Tywanek E., Michalak A., Świrska J., et al. Autoimmunity, new potential biomarkers and the thyroid gland-the perspective of Hashimoto's thyroiditis and its treatment. Int. J. Mol. Sci. 2024;25(9) doi: 10.3390/ijms25094703. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Xu B., Huang S., Peng W., et al. Comprehensive analysis of hub biomarkers associated with immune and oxidative stress in Hashimoto's thyroiditis. Arch. Biochem. Biophys. 2023;745 doi: 10.1016/j.abb.2023.109713. [DOI] [PubMed] [Google Scholar]
- 25.Zheng L., Dou X., Song H., et al. Bioinformatics analysis of key genes and pathways in Hashimoto thyroiditis tissues. Biosci. Rep. 2020;40(7) doi: 10.1042/BSR20200759. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Barrett T., Wilhite S.E., Ledoux P., et al. NCBI GEO: archive for functional genomics data sets--update. Nucleic Acids Res. 2013;41(Database issue):D991–D995. doi: 10.1093/nar/gks1193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Edgar R., Domrachev M., Lash A.E. Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. 2002;30(1):207–210. doi: 10.1093/nar/30.1.207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ritchie M.E., Phipson B., Wu D., et al. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7):e47. doi: 10.1093/nar/gkv007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Wixon J., Kell D. The Kyoto encyclopedia of genes and genomes--KEGG. Yeast. 2000;17(1):48–55. doi: 10.1002/(SICI)1097-0061(200004)17:1<48::AID-YEA2>3.0.CO;2-H. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Yu G., Wang L.G., Han Y., et al. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284–287. doi: 10.1089/omi.2011.0118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Harris M.A., Clark J., Ireland A., et al. The Gene Ontology (GO) database and informatics resource. Nucleic Acids Res. 2004;32(Database issue):D258–D261. doi: 10.1093/nar/gkh036. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Reimand J., Isserlin R., Voisin V., et al. Pathway enrichment analysis and visualization of omics data using g:Profiler, GSEA, Cytoscape and EnrichmentMap. Nat. Protoc. 2019;14(2):482–517. doi: 10.1038/s41596-018-0103-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Guan S., Xu Z., Yang T., et al. Identifying potential targets for preventing cancer progression through the PLA2G1B recombinant protein using bioinformatics and machine learning methods. Int. J. Biol. Macromol. 2024;276(Pt 1) doi: 10.1016/j.ijbiomac.2024.133918. [DOI] [PubMed] [Google Scholar]
- 34.HäNZELMANN S., Castelo R., Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinf. 2013;14:7. doi: 10.1186/1471-2105-14-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Barbie D.A., Tamayo P., Boehm J.S., et al. Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1. Nature. 2009;462(7269):108–112. doi: 10.1038/nature08460. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Yoo M., Shin J., Kim J., et al. DSigDB: drug signatures database for gene set analysis. Bioinformatics. 2015;31(18):3069–3071. doi: 10.1093/bioinformatics/btv313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Stuart T., Butler A., Hoffman P., et al. Comprehensive integration of single-cell data. Cell. 2019;177(7) doi: 10.1016/j.cell.2019.05.031. 1888-902.e21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Di Crescenzo V., D'Antonio A., Tonacchera M., et al. Human herpes virus associated with Hashimoto's thyroiditis. Inf. Med. 2013;21(3):224–228. [PubMed] [Google Scholar]
- 39.Horie I., Abiru N., Nagayama Y., et al. T helper type 17 immune response plays an indispensable role for development of iodine-induced autoimmune thyroiditis in nonobese diabetic-H2h4 mice. Endocrinology. 2009;150(11):5135–5142. doi: 10.1210/en.2009-0434. [DOI] [PubMed] [Google Scholar]
- 40.Diamond M.S., Farzan M. The broad-spectrum antiviral functions of IFIT and IFITM proteins. Nat. Rev. Immunol. 2013;13(1):46–57. doi: 10.1038/nri3344. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Dediego M.L., Martinez-Sobrido L., Topham D.J. Novel functions of IFI44L as a feedback regulator of host antiviral responses. J. Virol. 2019;93(21) doi: 10.1128/JVI.01159-19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Lempainen J., Korhonen L.S., KantojäRVI K., et al. Associations between IFI44L gene variants and rates of respiratory tract infections during early childhood. J. Infect. Dis. 2021;223(1):157–165. doi: 10.1093/infdis/jiaa341. [DOI] [PubMed] [Google Scholar]
- 43.Boehmer D.F.R., Formisano S., De Oliveira Mann C.C., et al. OAS1/RNase L executes RIG-I ligand-dependent tumor cell apoptosis. Sci Immunol. 2021;6(61) doi: 10.1126/sciimmunol.abe2550. [DOI] [PubMed] [Google Scholar]
- 44.Xiao H., Liang J., Liu S., et al. Proteomics and organoid culture reveal the underlying pathogenesis of Hashimoto's thyroiditis. Front. Immunol. 2021;12 doi: 10.3389/fimmu.2021.784975. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated during and/or analyzed during the current study are available in the GEO database (https://www.ncbi.nlm.nih.gov/geo/).
Data will be made available on request.