Abstract
IgG4-related disease (IgG4-RD) is an autoimmune disease of unknown cause. RUNX1 is a transcription factor involved in immune responses, and its dysfunction leads to uncontrolled immune responses. We performed, to our knowledge, the first RUNX1 methylation analysis in type 1 autoimmune pancreatitis (denoted simply as AIP), a representative IgG4-RD. We conducted methylation array analysis on AIP samples using the Illumina Infinium MethylationEPIC array. We identified a potentially important RUNX1 methylation abnormality in AIP, which was further confirmed using the quantitative SYBR green methylation-specific polymerase chain reaction (QSG-MSP) method. Using immunohistochemistry, we analyzed RUNX1 expression in lymphocytes in 11 lymph nodes with AIP (LN-AIP) and 20 lymph nodes with pancreatic ductal adenocarcinoma (PDA; LN-PDA). LN-AIP and LN-PDA refer to lymph nodes from patients with AIP and PDA, respectively. All selected PDA patients lacked lymph node metastases. By array analysis, RUNX1 was more highly methylated in LN-AIP than in LN-PDA (P = 0.0275). RUNX1 was also more highly methylated in LN-AIP than in LN-PDA by QSG-MSP validation (P = 0.0331). Immunohistochemical analysis indicated that RUNX1 expression was significantly lower in LN-AIP than in LN-PDA (P = 0.0250). Aberrant RUNX1 methylation and reduced expression may be characteristic of AIP. Further validation is warranted.
Introduction
IgG4-related disease (IgG4-RD), a disease concept that has become well known in the twenty-first century, is characterized by high serum IgG4 levels and tumor formation in various organs. 1 The histopathological features of IgG4-RD are lymphoplasmacytic infiltration and obstructive phlebitis, with many IgG4-positive plasma cells identified by immunohistochemistry.2,3 Steroids are often effective for the treatment of IgG4-RD. 4 Typically, surgical intervention is unnecessary for IgG4-RD, even when mass-like lesions are present. 5 Appropriate pathologic evaluation plays an important role in differentiating IgG4-RD from malignancies that mimic IgG4-RD, such as pancreatic ductal adenocarcinoma (PDA). Misdiagnosis can lead to unnecessary surgery, underscoring the need for accurate histopathologic analysis of specimens from patients with suspected IgG4-RD. This emphasizes the importance of clinical recognition and collaboration between clinicians and pathologists in the management of this disease.
The cause of IgG4-RD is unknown, but several autoantibodies and candidate target antigens have been reported. 6 Lactoferrin, carbonic anhydrase (CA)-II, CA-IV, and pancreatic secretory trypsin inhibitor in the pancreas are among the antigen candidates.7–12 Specific polymorphisms and haplotypes of CTLA-4, an immunoregulatory molecule primarily expressed by activated memory T cells and regulatory T cells (Tregs), have been reported as susceptibility factors for IgG4-RD.13,14 Abnormal immune responses may be involved in the development of IgG4-RD, which is considered an autoimmune disease, 15 but the pathogenic mechanism remains unclear. In recent years, links have been reported between the onset of autoimmune diseases—such as rheumatoid arthritis and systemic lupus erythematosus—and abnormal methylation. We reported the abnormal methylation of SKI in type 1 autoimmune pancreatitis (denoted simply as AIP), which is a typical IgG4-RD.16,17 However, there are still few reports on abnormal methylation and IgG4-RD.
Comparison of DNA methylation array results from AIP and obstructive pancreatitis (OP) of pancreatic ductal adenocarcinoma (PDA), and comparison of DNA methylation array results from lymph nodes near AIP and lymph nodes near OP of PDA, showed that the most methylated transcription factor in the promoter region in both AIP and the lymph nodes near AIP was RUNX1. However, they were in different regions of the RUNX1 promoter, so the region that was more highly methylated in both AIP and lymph nodes near AIP was selected as the candidate gene. Several reports have implicated decreased RUNX1 expression in immune dysregulation.18–20 The IgG4-RD microenvironment has been analyzed, including the lymph nodes, and similar changes have been observed in the lymph nodes, pancreas, and lungs. 21 We analyzed lymph nodes because our focus was on the aberrant methylation of RUNX1 expressed in lymphocytes. Thereafter, we verified RUNX1 methylation levels using quantitative SYBR green methylation-specific polymerase chain reaction (PCR; QSG-MSP) and protein expression via immunohistochemistry. We further analyzed the relationship between RUNX1 methylation levels and the clinicopathological data of the AIP patients in this study. To determine the role of RUNX1 expression in immune regulation during AIP, we compared RUNX1 expression in lymphocytes from the lymph nodes of patients with AIP or PDA.
Patients and Methods
Patients and Tissue Samples
We identified 11 patients that met the diagnostic criteria for AIP and 20 PDA patients without a history of AIP at Shinshu University Hospital or one of its affiliated hospitals from 2011 to 2021. We retrieved specimens from the archived tissue blocks of these patients: 11 lymph nodes with AIP (LN-AIP) and 20 lymph nodes with PDA (LN-PDA). This study was approved by the Ethics Committee of Shinshu University, Japan. The clinicopathological data from the patients are provided in Table 1. All lesions were reviewed by two pathologists (TU and MI) for pathologic diagnosis using hematoxylin and eosin (HE)-stained tissue sections.
Clinical features of patients with autoimmune pancreatitis (AIP) and pancreatic ductal adenocarcinoma (PDA).
*Normal value for IgG4: < 70 mg/dL (cut-off: 135 mg/dL).
†n = 7.
‡n = 11.
AIP, autoimmune pancreatitis; PDA, pancreatic ductal adenocarcinoma.
Methylation Array
An Illumina Infinium MethylationEPIC array (Illumina, San Diego, CA) was performed on 4 LN-AIP and 5 LN-PDA. Images were obtained using GenomeStudio software (Illumina). The methylation score for each CpG group was represented as a beta value according to the fluorescence intensity ratio. Beta values range from 0 (non-methylated) to 1 (completely methylated). Using the EPIC system, a high beta value is indicative of hypermethylation. Preparation of DNA for the Illumina Infinium Methylation EPIC array was performed according to the Illumina Infinium HD Methylation Assay Protocol Guide. For bisulfite treatment, genomic DNA was extracted from formalin-fixed and paraffin-embedded unstained tissue sections (4 µm) of LN-AIP and LN-PDA samples, using the QIAamp DNA Mini Kit (Qiagen, Valencia, CA). DNA was modified with sodium bisulfite using the EZ DNA Methylation Kit (Zymo Research, Orange, CA), according to the manufacturer's instructions.
QSG-MSP
QSG-MSP was performed to quantify the level of RUNX1 CpG DNA methylation in 5 LN-AIP and 13 LN-PDA using real-time PCR as previously reported. 22 The primers for QSG-MSP were designed using Methyl Primer Express Software v1.0 (Applied Biosystems). The presence of CpG islands was also determined using Methyl Primer Express v1.0 software (Applied Biosystems). The sequences of the methylated primers were 5′- GGTTTTAGGTAAGCGGTTC-3′ (forward) and 5′-ACGAAAATATAACGACTCCAA-3′ (reverse). The sequences of the primers for the control ACTB (beta-actin) gene were 5′-TGGTGATGGAGGAGGTTTAGTAAGT-3′ (forward) and 5′-AACCAATAAAACCTACTCCTCCCTTAA-3′ (reverse), as previously described. 22 Quantitative PCR was performed in a 25-µL reaction volume with 12.5 µL of 2× SYBR Green PCR Master Mix (Applied Biosystems), 2.5 pmol of each primer, and 25 ng of bisulfite-treated DNA sample. Thermal cycling was as follows: 95′°C for 10 min, 40 cycles of 95′°C for 15 s, and 60′°C for 1 min. Only samples with amplification at the correct melting temperature were used for further methylation analyses. The ΔΔCT method was used to analyze the relative expression level of RUNX1.
Immunohistochemistry
The level of RUNX1 protein in the lymph nodes was verified using the Human Protein Atlas (https://www.proteinatlas.org), which suggested that RUNX1 expression was strong in the paracortical region of the lymph node. Immunohistochemical and HE staining were performed on formalin-fixed, paraffin-embedded tissue sections (4 µm) from 11 LN-AIP and 20 LN-PDA. Immunohistochemical staining for CD20 and CD3 was conducted using an automated slide preparation system (BenchMark ULTRA automated slide stainer, Ventana Medical Systems, Tucson, AZ). The following primary antibodies were used in accordance with the manufacturer's instructions: CD20 (clone: L26, Leica, Wetzlar, Germany) and CD3 (clone: LN10, Leica). Slides were manually immunostained with anti-RUNX1 antibody (1:100; Cell Signaling Technology, Danvers, MA) using the following method. For antigen retrieval, sections were microwaved in 0.45% Tris/5 mM EDTA for 30 min. Detection of the primary antibodies was performed using Novolink Polymer Detection Systems (Leica Biosystems, Nussloch, Germany). Immunostaining for FOXP3 was also performed for Treg identification. The samples were autostained for FOXP3 using the Bond-III system (Leica, Wetzlar, Germany), and BOND Epitope Retrieval Solution 2 was utilized for antigen retrieval. The primary antibody used was anti-FOXP3 (dilution 1:100, Clone 236A/E7; Abcam).
In paracortical areas with CD3 and FOXP3 positivity, the Allred scoring method was used for RUNX1 staining analysis. 23 The intensity score ranges from 0 to 3, with 0 for no staining, 1 for weak staining, 2 for intermediate staining, and 3 for intense staining. The proportion score refers to the percentage of immuno-positive nuclei after categorization into six subgroups (0 for no staining, 1 for <1%, 2 for 1%–10%, 3 for 11%–33%, 4 for 34%–66%, and 5 for 67%–100%). The semi-quantitative Allred score (total score) was calculated by adding the proportion score to the intensity score.
Statistical Analysis
We performed the Chi-squared test or Wilcoxon rank-sum test with JMP software version 14 (SAS Institute, Inc, Cary, North Carolina, United States) to analyze the clinical data. P values of <0.05 were considered significant.
Results
Clinicopathological Features
All AIP patients had lymphoplasmacytic inflammation, storiform fibrosis, and abundant IgG4-positive cells (≥10 cells/high-power field) in the pancreas. No PDA patients exhibited the features of IgG4-RD. According to the classification defined by Sato et al, 24 LN-AIP was type I in 0 patients, type II in 5 patients, type III in 6 patients, type IV in 0 patients, and type V in 0 patients.
The serum concentrations of IgG4 were measured for 7 AIP patients (mean: 255 mg/dL, range: 200-403 mg/dL) and 11 PDA patients (mean: 27 mg/dL, range: 17-66 mg/dL) (Table 1).
Methylation Array Beta Values and the QSG-MSP Methylation Rate
We conducted a methylation array to identify hypermethylated genes in LN-AIP and LN-PDA. We examined the function of each gene identified in the methylation array using existing databases and publications, and classified genes with functions specifically related to the immune system as genes involved in immune regulation. We then sequentially examined the genes with the highest methylation levels in LN-AIP compared with LN-PDA and selected the immune-related genes with the highest comparative methylation levels. Among these genes, RUNX1 had the highest beta value in LN-AIP compared with LN-PDA. In our selected promoter regions, the RUNX1 beta value in LN-AIP (mean: 0.684, range: 0.619-0.728) was significantly higher than that in LN-PDA (mean: 0.537, range: 0.502-0.600) (Figure 1A; P = 0.0275). QSG-MSP analysis confirmed that the RUNX1 methylation ratio was significantly higher in LN-AIP (mean: 0.198, range: 0.043-0.3775) than in LN-PDA (mean: 0.016, range: 0-0.099) (Figure 1B; P = 0.0331).
Box plot of the RUNX1 beta values from the methylation array, RUNX1 methylation rates by QSG-MSP, and allred scores of the lymph nodes from patients with autoimmune pancreatitis (LN-AIP) or lymph nodes from patients with pancreatic ductal adenocarcinoma (LN-PDA). (*P = 0.0275, **P = 0.0331, and ***P = 0.0250).
Immunohistochemistry for RUNX1
RUNX1 was expressed in the nucleus of many lymphocytes in the paracortical region of LN-AIP and LN-PDA, where CD3-positive and FOXP3-positive Tregs were also often identified (Figure 2). LN-AIP had a significantly lower Allred score (mean: 6, range: 6-6) than LN-PDA (mean: 6.5, range: 6-7) (Figure 1C; P = 0.0250).
Immunohistochemical detection of RUNX1 expression in samples from lymph nodes from patients with autoimmune pancreatitis (LN-AIP) or lymph nodes from patients with pancreatic ductal adenocarcinoma (LN-PDA). Hematoxylin and eosin staining in the cortical region of LN-AIP (A) and LN-PDA (E). CD3 immunostaining in the cortical region of LN-AIP (B) and LN-PDA (F). Treg immunostaining in the cortical region of LN-AIP (C) and LN-PDA (G). RUNX1 staining in the cortical region of LN-AIP (D) and LN-PDA (H).
Discussion
As the deletion of RUNX1 in lymphocytes produces various immune abnormalities, the finding of abnormal RUNX1 methylation in LN-AIP in our study may account for the diverse causes of AIP. RUNX1 deficiency in CD4-positive T lymphocytes results in spontaneous cellular hyperactivation that causes fatal autoimmune inflammatory lung disease. 18 Deletion of RUNX1 or core-binding factor β—a cofactor of the Runx family in Tregs—causes colitis and pneumonia.25,26 In addition, Wong et al 18 found that RUNX1 deficiency in CD4-positive naive T lymphocytes results in the production of the inflammatory cytokines IL-17 and IL-21.
IL-21 has been reported to induce a humoral immune response by plasma cells.27,28 In IgG4-RD, follicular helper T (Tfh) cells mainly produce IL-21 and promote the differentiation of plasmablasts, which are involved in IgG4 production. 29 Since RUNX1 methylation abnormalities in lymph nodes are stronger in LN-AIP than in LN-PDA, the RUNX1 abnormality may be related to abnormality of the lymphocytes themselves.
The transcription factor Run RUNX1 inhibits the differentiation of naive CD4-positive T lymphocytes into the Th2 lineage by suppressing Gata3 expression, 30 so suppression of RUNX1 could theoretically promote differentiation into the Th2 lineage. It has been reported that deletion of RUNX1 results in Gata3 expression in murine T cell acute lymphoblastic leukemia, 31 which further supports the idea that GATA3 expression promotes Th2 differentiation. 32 However, recent studies suggest that Tfh cells, rather than Th2 cells, are the main source of IL-4 production in IgG4-RD. 33 Therefore, while RUNX1 suppression might still contribute to the differentiation of Th2 cells via GATA3 expression, the role of Tfh cells in IL-4 production should not be overlooked. These findings indicate that abnormalities in RUNX1 could have significant implications for the development of IgG4-RD, though the exact pathways remain to be fully elucidated.
Although IL-21 levels are elevated in IgG4-RD, 34 IL-17 levels are not significantly higher than in controls. 35 Th17-related molecules, including IL-17, were rarely expressed in patients with IgG4-retaled dacryoadenitis and sialadenitis. 35 IL-21 is produced by Th2, Th17, and Tfh cells. 35 Thus, it is possible that Th2, Tfh, or both Th2 and Tfh cells produce IL-21 in IgG4-RD.
The expression of RUNX1 in lymphocytes suggests its involvement in the regulation of immune responses, particularly in the lymphocyte microenvironment. Although we studied its expression in lymph nodes, differential RUNX1 expression in lymphocytes in individuals with PDA and AIP may reflect different immune regulatory mechanisms in these diseases. Furthermore, the role of RUNX1 in regulating lymphocyte function may influence the progression of PDA and its interaction with the immune system. These findings emphasize the importance of considering lymphocyte-specific RUNX1 expression when examining the role of RUNX1 in both immune regulation and tumor biology in PDA.
From a diagnostic perspective, the differential expression of RUNX1 in the lymphocytes of LN-AIP and LN-PDA suggests that RUNX1 immunostaining may be a useful tool in surgical pathology to differentiate between these two diseases. As shown in our study, lymphocytes in PDA have higher RUNX1 expression than those in AIP, so RUNX1 may serve as an immunohistochemical marker to improve diagnostic accuracy. Incorporating RUNX1 immunostaining, particularly in lymphocytes, into the diagnostic process may improve the accuracy of differentiating AIP from PDA. Further studies evaluating the sensitivity and specificity of RUNX1 as a diagnostic marker in lymphocytes will further strengthen the role of RUNX1 in routine surgical pathology.
Autoantibodies against laminin, 36 CA-II, 10 and lactoferrin 8 have been identified in IgG4-RD. It has also been reported that a single nucleotide polymorphism in RUNX1 may be linked to the etiology of the autoimmune disease rheumatoid arthritis, 37 and a deficiency in RUNX1 may facilitate the appearance of autoantibodies in IgG4-RD.
There are fewer naive Tregs and more memory T cells in patients with IgG4-RD than in healthy individuals. 38 This result is consistent with the findings of Wong et al 18 in RUNX1-deficient mice. Methylation of RUNX1 in AIP may lead to fewer naive Tregs and more memory T cells, hallmarks of IgG4-RD. RUNX1 methylation may thereby disrupt immune tolerance, leading to specific clinicopathological findings such as the appearance of multiple autoantibodies and IgG4 production in IgG4-RD. This hypothesis should be investigated using RUNX1-deficient cells.
A limitation of this study is that it only looked at a limited number of methylation abnormalities and related protein expression. Expression analysis using cultured cells is needed to determine whether RUNX1 suppression causes cytokine abnormalities associated with IgG4-RD development.
Conclusion
Aberrant methylation of RUNX1 may be a characteristic abnormality of IgG4-RD. Further analysis is warranted to elucidate the relationship between IgG4-RD and RUNX1 methylation abnormalities.
Footnotes
List of Abbreviations
Acknowledgments
We are grateful to Masanobu Momose, Yasuyo Shimojo, Chitose Arai, Marina Nuno, Kanade Wakabayashi, Naoko Yamaoka, Shotaro Komamura, and Daiki Ogura of Shinshu University Hospital for their excellent technical assistance; Shinya Hokibara (Suwa Red Cross Hospital), Yukiko Kusama (Nagano Municipal Hospital), Masato Nakaguro (Nagoya University Hospital), Koh Nakazawa (National Hospital Organization Matsumoto Medical Center), Hisashi Shimojo (Aizawa Hospital), Kenji Sano (Iida Municipal Hospital), Satoshi Shiozawa (Saku Center Hospital), and Ichiro Ito (Nagano Red Cross Hospital) for their collaboration. We thank Amanda Holland, PhD, from Edanz (
) for editing a draft of this manuscript.
Availability of Data and Materials
All data generated and analyzed during the current study are available from the corresponding author upon reasonable request.
Ethics Approval and Consent to Participate
The Ethics Committee of Shinshu University School of Medicine approved this study (Approval Code: 720). The requirement for informed consent was waived by the Ethics Committee of Shinshu University School of Medicine, and an opt-out method was used because of the retrospective design of the study. The investigation was conducted in compliance with the provisions of the Helsinki Declaration.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
This study was supported by the Hokuto Foundation for Bioscience (grant award to TU). This work was supported by JSPS KAKENHI Grant Number JP20K07405.
CRediT Authorship Contribution Statement
Trial Registration
Not applicable, because this article does not contain any clinical trials.
Informed Consent
The requirement for informed consent was waived by the Ethics Committee of Shinshu University School of Medicine, and an opt-out method was used because of the retrospective design of the study.