Causal effects of immune cell phenotypes on the risk of autoimmune liver diseases: a bidirectional two-sample Mendelian randomization study

Introduction

Autoimmune liver diseases (AILDs) constitute a group of chronic immune-mediated inflammatory conditions primarily targeting hepatocytes and bile duct cells. This category includes autoimmune hepatitis (AIH), primary biliary cholangitis (PBC) and primary sclerosing cholangitis (PSC). Epidemiological data indicate a steady rise in the prevalence and incidence of AIH and PSC across European regions, while PBC demonstrates a relatively higher prevalence in North America, the Asia-Pacific region, and other geographical areas (1). A substantial body of research suggests the associations between AILDs and factors such as smoking, medication usage and other substances (2). Although AILDs have been among the earliest autoimmune diseases investigated, their underlying mechanisms have only begun to be unveiled in recent years (3).

The predominant perspective asserts that immune regulatory genetic defects, coupled with specific environmental triggers, can instigate aberrant immune system activation, leading to chronic inflammation and tissue damage, thereby precipitating the onset of AILDs (4). Recent research highlights immune cell dysregulation, autoantibody mechanisms, and their links to AILDs. Autoantibodies significantly damage tissues by altering functions, triggering complement damage, forming immune complexes, and indirectly inducing cytokine release. Notably, antinuclear antibodies, antimitochondrial antibodies, and anti-neutrophil cytoplasmic antibodies are key contributors to AIH, PBC, and PSC progression, respectively. Studies highlight dysregulations in innate immune system activation and immune tolerance breakdown, resulting in abnormal activation of cells like Th17 and natural killer (NK) cells, as well as reduced regulatory T (Treg) cells. This complexity impedes full understanding of AILDs, hindering stage-specific therapeutic development. Encouragingly, research teams have conclusively shown the immunologically mediated aetiology of these three diseases using genetic investigations. This highlights the potential for further immune-related investigations into AILDs using genome-wide association study (GWAS) and related methods.

Currently, the lack of effective curative measures for AILDs leaves these conditions without definitive therapeutic solutions. Consequently, patients are frequently required to undergo lifelong administration of immunosuppressive drugs or bile acids to manage hepatic or biliary inflammation (5). Promising research, such as Treg cell therapy, offers hope by suggesting that enhancing Treg cells in the liver could mitigate disease severity (6). Therefore, further investigation is crucial to understand the intricate relationship between AILDs and immunity. We need a deeper understanding of the immune system’s role in AILDs pathogenesis.

Mendelian randomization (MR) is a statistical tool that boldly investigates whether a specific factor has a causal impact on a particular health outcome, like disease risk (7). This method harnesses genetic variants as instrumental variables (IVs), capitalizing on the fact that genetic variants are biologically linked to certain factors but remain unaffected by environmental influences. Consequently, MR has the capacity to mimic the effect of a randomized trial on disease risk without being confounded by extraneous factors (8). In recent years, GWAS advancements have provided researchers with extensive genetic variation data, enhancing the comprehensiveness and precision of MR studies. Our investigation aims to shed light on the causal relationship between immune cell characteristics and AILDs through a thorough MR analysis. We present this article in accordance with the STROBE-MR reporting checklist (available at https://tgh.amegroups.com/article/view/10.21037/tgh-25-2/rc).

Methods

Study design

This study employed a bold approach: the two-sample MR analysis, a robust causal inference method. By using genetic variants as IVs, this analysis establishes causal relationships between exposure and outcome, effectively mitigating biases arising from confounding and reverse causation issues. The selection of IVs adheres to three fundamental principles in MR analysis: (I) direct association of genetic variants with the exposure (relevance assumption); (II) independence of genetic variants from potential confounders between exposure and outcome (independence assumption); and (III) exclusion of genetic variants’ influence on the outcome through pathways other than the exposure (exclusion-restriction assumption) (8). We conducted a daring two-sample MR analysis, utilizing summary-level data from extensive GWAS, to evaluate the causal links between 731 immune cell phenotypes and AIH, PBC and PSC, respectively (Figure 1). Ethical approval was diligently obtained from the respective institutions for all GWAS studies included in this research. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Figure 1 The design of MR analysis on the relationship between immune cell phenotypes and AILDs. AIH, autoimmune hepatitis; AILD, autoimmune liver disease; GWAS, genome-wide association study; MR, Mendelian randomization; PBC, primary biliary cholangitis; PSC, primary sclerosing cholangitis; SNP, single nucleotide polymorphism.

GWAS resource

We harnessed publicly available GWAS datasets as our primary data source. The exposure data, comprising immune cell phenotypes, were derived from a study conducted in 2020 involving 3,757 Sardinian individuals. This study delved into 731 immune cell phenotypes, exploring their natural genetic variations. The dataset encompassed 118 absolute counts (ACs), 389 mean fluorescence intensities (MFIs), 32 morphological parameters (MPs), and 192 relative cell counts (RCs). After analyzing over 20 million single nucleotide polymorphisms (SNPs), the research team identified 763 trait-variant associations and 53 novel loci (9).

Similarly, outcome data were sourced from publicly available GWAS analysis summaries. Data for AIH were obtained from a comprehensive analysis of 485,234 Europeans conducted in 2021, encompassing over twenty-four million SNPs (10). Data for PBC were obtained from a meta-analysis of a similar scale conducted in the same year. This meta-analysis utilized data from 24,510 European subjects and represented the most extensive GWAS analysis of PBC to date (11). Data for PSC were obtained from a comprehensive GWAS analysis conducted in 2017, which included 14,890 cases of European ethnicity. It’s important to note that all original studies obtained ethical approval and informed consent from all participants (12). Hence, all anonymized patient data utilized in this study strictly adhered to standard ethical guidelines. To address potential population stratification, we confirmed that exposure and outcome GWAS datasets were derived from independent cohorts, minimizing sample overlap.

IVs selection

In this study, the abundance of immune phenotype sample data provided ample opportunity to extract significantly independent SNPs. To ensure robustness, the significance threshold was boldly adjusted to 5e−8, setting stringent criteria for SNP selection. Additionally, within a 10,000 kb distance, linkage disequilibrium (LD) analysis was conducted with an r² threshold <0.001 in PLINK 2.0, based on the 1000 Genomes European reference panel, further fortifying the independence of each SNP and mitigating the influence of genetic pleiotropy on the results (13). Furthermore, to mitigate analytical errors stemming from weak IVs, the F-statistic for each IV was calculated. IVs with an F-statistic surpassing 10 were considered robust and thus incorporated into the study. The F-statistic formula utilized for this computation, F = R² × (N − 2)/(1 − R²), R² = 2EAF (1 − EAF) β², where R² represents the cumulative explained variance of the selected IV, EAF denotes the effect allele frequency, β signifies the SNP’s estimated effect, and N denotes the sample size of the GWAS, was employed for this purpose (14). SNPs identified with such factors were subsequently excluded from the study. Importantly, since the GWAS data of exposure and outcomes originated from different studies, concerns regarding population overlap were effectively nullified.

Reverse MR analysis

We conducted a reverse MR analysis utilizing SNPs that have been established to be associated with AILDs, with the objective of elucidating the potential causal influence of AILDs on the previously identified significant associations with immune cells. SNPs significantly associated with AILDs (P value <5e−8) were selected as IVs, following the same LD clumping (r²<0.001, 10,000 kb distance) and F-statistic criteria (F>10) as the primary MR analysis. The criteria for selecting the IVs in this reverse MR analysis adhered strictly to those applied in the primary MR analysis, ensuring consistency and rigor in our methodology.

Statistical analysis

In our MR analysis of the relationship between immune cell phenotypes and ALDs, we utilized five common analytical methods, including the inverse variance weighted (IVW) method, the MR Egger method, the weighted median method, the simple mode method, and the weighted mode method. The IVW method is simple and powerful but may be biased by genetic heterogeneity (15). To enhance the significance of our MR analysis results, we adjusted the significance threshold to 0.001, surpassing the conventional threshold of 0.05. Furthermore, we ensured consistency by selecting exposure SNPs that produced consistent odds ratios (ORs) across all five analytical methods, thereby enhancing the credibility of our analysis results. Sensitivity analyses were then conducted to scrutinize horizontal pleiotropy and heterogeneity, examining additional impacts of IVs bypassing exposure pathways or significant differences in variable effects among different subgroups (16). For a positive outcome, the P values of both sensitivity analysis and Cochran’s Q test exceeding 0.05 indicated the absence of pleiotropy or heterogeneity. Conversely, P values below 0.05 signaled the presence of pleiotropy or heterogeneity (17), mandating individual analyses to determine inclusion in the study scope. Moreover, the MR-Pleiotropy Residual Sum and Outlier (MR-PRESSO) method was also utilized to evaluate the potential pleiotropic effects by comparing the actual residual sum of squares to the anticipated residual sum of squares. Additionally, we utilized various visual analyses to reinforce the reliability of our study. These comprised leave-one-out analysis to scrutinize the influence of individual SNPs on overall results, scatter plots to confirm the absence of anomalous impacts on outcomes, and funnel plots to evaluate correlation robustness and heterogeneity presence. For this analysis, we utilized the R statistical software, specifically version 4.1.2. All analyses were performed using the TwoSampleMR package in R. Data preprocessing and visualization were conducted with R packages dplyr, ggplot2, and forestplot.

Results

Selection of IVs

After conducting significance and LD screening of SNPs on 731 immune cell phenotypes, a total of 2,803 influential SNPs were filtered, involving 607 immune cell phenotypes. These phenotypes were categorized based on their typical characteristics and cell types, yielding proportions for each cell category (Figure 2). Notably, the MP category encompasses a relatively minor array of cell types, comprising solely conventional dendritic cells (cDCs) as well as T, B and natural killer (TBNK) cells, while the remaining feature classifications encompass seven distinct cell categories. Remarkably, all IVs boasted F-statistics surpassing 10, underscoring the robustness of their selection.

Figure 2 The summary of SNPs across different cell types after selection. AC, absolute count; cDC, conventional dendritic cell; MFI, mean fluorescence intensity; MP, morphological parameters; RC, relative cell count; SNP, single nucleotide polymorphism; TBNK, T, B and natural killer cells; Treg, regulatory T cells.

Causal effects of immunophenotypes on AILDs

In the current investigation, IVs culled from exposure data underwent MR scrutiny against outcome data about AIH, PBC and PSC. In general, the standard error of the IVW method tends to be lower than that of other methods. As a result, in cases where heterogeneity and horizontal pleiotropy are absent, priority will be given to utilizing the results obtained from the IVW analysis, which is considered the gold standard of MR analysis. Therefore, we first screened the results of IVW analysis based on a P value of 0.05, obtaining 40, 24, and 61 SNPs in AIH, PBC, and PSC. To enhance significance, we further narrowed the threshold to 0.001, reducing the significant SNPs in AIH, PBC, and PSC to 15, 16, and 9. We presented these preliminary screening results in Figure 3 in the form of three volcano plots, with results generated by different thresholds displayed in different colors, and also provided an initial description of the positive or negative orientations of each SNP with OR (Figure 3).

Figure 3 Volcano plot of MR results in immune cell phenotypes on the risk of AILDs. AIH, autoimmune hepatitis; AILD, autoimmune liver disease; cDC, conventional dendritic cell; MR, Mendelian randomization; OR, odds ratio; PBC, primary biliary cholangitis; PSC, primary sclerosing cholangitis; TBNK, T, B and natural killer cells.

After further confirmation that each SNP had five different analysis methods, all of which pointed in the same direction, a collective total of seven significant associations were sustained for AIH, three for PBC, and two for PSC (Figure 4). Among the 12 immune cell phenotypes scrutinized, there were representatives of Treg, TBNK cells, myeloid cells, monocytes, cDC, and B cells (Figure 2).

Figure 4 The MR results of five methods for positive SNPs obtained after sensitivity analysis screening. (A) Characteristic phenotypes and their analysis results in AIH. (B) Characteristic phenotypes and their analysis results in PBC. (C) Characteristic phenotypes and their analysis results in PSC. AC, absolute count; AIH, autoimmune hepatitis; CI, confidence interval; DC, dendritic cell; MR, Mendelian randomization; NK, natural killer; OR, odds ratio; PBC, primary biliary cholangitis; PSC, primary sclerosing cholangitis; SNP, single nucleotide polymorphism; unsw mem, unswitched memory.

In the AIH group, our research has identified seven immune cell types that may have causal links with AIH (Figure 4). Among these, HLA-DR on plasmacytoid DC and HLA-DR on CD33⁻ HLA-DR⁺ myeloid cell demonstrated a positive association with AIH, suggesting that these immune cell phenotypes could heighten the risk of AIH. Conversely, the remaining immune cell phenotypes, including three phenotypes of NK cells (HLA-DR⁺ NK AC, HLA-DR⁺ NK %NK, and HLA-DR⁺ NK %CD3⁻ lymphocyte) and two phenotypes of monocytes (CD14⁺ CD16⁺ monocyte AC, CD16 on CD14⁻ CD16⁺ monocyte), exhibited a negative correlation.

Within the PBC enclave, our study indicated that three phenotypes of B cells, such as CD27 on CD24⁺ CD27⁺ B cell, CD27 on IgD⁺ CD38⁻ unswitched memory (unsw mem) B cell, and CD27 on memory B cell, might have causal connections with PBC (Figure 4). All of the three B cell phenotypes considered in our investigation showed a positive correlation with PBC, suggesting that heightened levels of surface markers expressed by these cells could elevate the likelihood of developing PBC.

Similarly, amidst the PSC cohort, our study results indicated that two immune cell phenotypes, i.e., CD28 on CD45RA⁻ CD4 not Treg cell and FSC-A on CD4⁺ NK cell, might be associated with potential causality in PSC (Figure 4). Both of the two immune cell phenotypes examined in our analysis exhibited inverse correlations with PSC, suggesting that an augmentation in the abundance of surface markers expressed by these immune cells could reduce the risk of PSC development.

Reverse MR analysis

After conducting a reverse MR analysis between AILDs and the previous 12 significant immune cells, we found no evidence of reverse causal relationships among them across all applied MR methods. Notably, Cochran’s Q test and the MR-Egger intercept test indicated partial heterogeneity and horizontal pleiotropy in the reverse MR analysis, indicating variability in the effects of individual SNPs (Table 1).

Sensitivity analysis

The P values obtained from the level of pleiotropy and heterogeneity analyses in each group are illustrated in Table 1. Although the P values for pleiotropy and heterogeneity of CD28⁻ CD8br AC of Treg cell in the AIH group meet the screening criteria, the 95% CI in its MR Egger method ranges from 0.002 to 1,846, indicating an extreme outlier. To ensure data reliability, this analysis opts to exclude this SNP. Additionally, as depicted in Table 1, in the PSC group, some SNPs fail to meet the requirements of the pleiotropy analysis’s Cochran’s Q value, where P value is less than 0.05, namely: HLA-DR on CD14⁻ CD16⁻ monocyte, HLA-DR on monocyte, HLA-DR on CD33⁻ HLA-DR⁺ myeloid cell. Therefore, according to the screening rules, the results from MR Egger and IVW methods cannot be explained due to the level of pleiotropy, violating the assumptions of MR. Consequently, these three SNPs were excluded from the results. As depicted in Table 1, the MR-PRESSO global test revealed no evidence of horizontal pleiotropy effects, with all P values greater than 0.05, for all exposures except for three exposures (CD27 on CD24⁺ CD27⁺ B cell, CD28 on CD45RA⁻ CD4 not Treg cell, FSC-A on CD4⁺ TBNK cell) where testing was not feasible due to insufficient SNPs (less than 4 SNPs).

In the end, we created a heatmap to illustrate the potential associations between the MR analysis results and sensitivity analysis outcomes for each significant SNP. In Figure 5, this heatmap incorporates the results of IVW analysis, as well as pleiotropy and heterogeneity analyses, with the shade of each color block representing the significance of each SNP (Figure 5). Scatter plots and leave-one-out analyses conducted on the filtered data consistently demonstrate stability, further affirming the robustness of the results (Figures S1-S3).

Figure 5 The heatmap of heterogeneity statistics results based on the IVW method for positive SNPs (P value ≤0.05). AIH, autoimmune hepatitis; IVW, inverse variance weighted; PBC, primary biliary cholangitis; PSC, primary sclerosing cholangitis; SNP, single nucleotide polymorphism.

Discussion

To our knowledge, this study is the first to explore the causal link between immune cell phenotypes and AILDs by using the MR method. Our findings, validated through rigorous checks for horizontal pleiotropy, reveal significant and independent associations. These results provide new insights into potential immunotherapy targets for addressing AILDs.

In the AIH cohort, HLA-DR on plasmacytoid DC and HLA-DR on CD33⁻ HLA-DR⁺ myeloid cell has robustly demonstrated a positive correlation with AIH incidence, emphasizing the pivotal role of HLA-DR in immune cell function. In Europeans and North American Caucasians, AIH shows a correlation with various HLA-DR alleles, while in Japan and Argentina, the presence of different HLA-DR alleles highlights the diversity in AIH susceptibility among populations (18). The exact pathogenic mechanisms of HLA-DR in AIH remain elusive. Current research indicates that HLA-DRB1 might facilitate antigen recognition by CD4⁺ cells and the formation of effector cells, thus contributing to immune response processes within liver tissue. These processes may entail the outbreak of interferon-gamma (IFN-γ) by effector Th1 cells, ultimately resulting in liver tissue damage (19). Also, other research reveals that interleukin (IL)-17 from HLA-DR-activated Th17 cells worsens AIH by inducing hepatocyte pyroptosis via the STAT3-IFI16 pathway, boosting IL-1β production (20). Blocking IL-17 in animal models reduces liver damage, suggesting anti-IL-17 therapies may benefit AIH patients upon clinical validation (21).

The CD33⁺ HLA-DR⁻ phenotype represents a highly distinctive characteristic according to current classifications of myeloid-derived suppressor cells (MDSCs). It is trustworthy that MDSCs can be significant in immunosuppression, immune evasion, and tumor progression through both immunomodulatory and non-immunomodulatory mechanisms. In this context, we propose the intriguing notion that cells exhibiting surface antigen expression contrary to that of MDSCs may potentially lead to immune hyperactivity or even autoimmunity (22). Salivary proteomics revealed elevated levels of cystatin A, statherin, and histatins in AIH, suggesting a heightened antimicrobial defence mechanism that may contribute to this immune hyperactivity (23). We hope this concept will inspire further exploration among researchers in the field.

However, the analysis results for three other immune cell phenotypes seem to “challenge” our understanding of HLA-DR. With HLA-DR, three distinct phenotypes of NK cells all correlate negatively with AIH, implying a specific protective association with the disease. Current research shows that HLA-DR⁺ NK cells possess the distinctive features of both NK cells and dendritic cells (DCs), fulfilling a pivotal function within the innate immune system. Liver-resident NK cells, a unique subset, demonstrate enhanced secretion of cytokines and chemokines, potentially fostering liver immune tolerance (24). High tetraantennary sialylation in plasma N-glycans emerged as a unique marker for AIH, potentially aiding non-invasive diagnosis and reflecting altered immune regulation by liver-resident NK cells (25). Although direct associations with AIH have not been reported, liver-resident NK cells have been linked to other autoimmune diseases, suggesting a potentially similar relationship with AIH.

In human monocyte classification, CD14⁺ CD16⁺ and CD14⁻ CD16⁺ belong to intermediate and non-classical monocyte subsets, respectively. Non-classical monocytes and CD16⁺ monocytes are recognized for their anti-inflammatory effects (26). They contribute to wound healing and enhance neutrophil adhesion by secreting tumor necrosis factor-alpha (TNF-α). These monocytes migrate and participate in immunosurveillance, exhibiting dual anti-inflammatory and pro-inflammatory functions (27). Studies on similar autoimmune diseases suggest their role in attenuating inflammation progression through pathways involving Treg recruitment and immune suppression via CXCL12 and transforming growth factor beta (TGF-β). Moreover, CD16⁺ monocytes express higher levels of CX3CR1, indicating the induction of more fractalkine (CX3CL1) into endothelial cells. CX3CL1 facilitates leukocyte migration and adhesion, potentially preventing liver fibrosis in animal models. Therefore, understanding the impact of the axis on liver function, especially in AIH patients, may lead to new therapeutic strategies.

The analysis reveals that CD27 expression on B cells indicates a substantial connection with the progression of PBC. CD27⁺ B cells are generally defined as memory B cells (28). Autoreactive B cells, among them, can generate autoantibodies upon interaction with self-antigens. In humans, CD27⁺ B cells have been identified as marginal zone (MZ) B cells with autoreactivity, which are prevalent in patients with autoimmune diseases and mouse models. These cells may exert harmful effects due to their rapid response to antigens (29). Additionally, CD24hiCD27⁺ B cells, known for their robust TNF-α synthesis, influence the progression of PBC and the transition from Th1 to Th17 cells, which may be crucial for inflammation in PBC (30,31).

Mass spectrometry (MS) studies have significantly deepened our understanding of PBC, revealing multiple biomarkers that reflect immunological and metabolic disturbances. These studies have found that certain metabolic molecules, such as acylcarnitines and free fatty acids, are associated with the severity of fibrosis and inflammation, indicating the role of lipid metabolism disorders in disease progression (32). Additionally, MS analysis has detected new autoantigens in patients who are negative for traditional anti-mitochondrial antibodies (AMA), suggesting broader autoimmune targets (33). These metabolic and immune biomarkers, in combination with immune features such as CD27⁺ B cells, highlight the complex interplay between immune activation and metabolic dysfunction in PBC, providing new directions for non-invasive disease monitoring and potential therapeutic targets.

Within the PSC cohort, two immune phenotypes, CD28 on CD45RA⁻ CD4 not Treg cell and FSC-A on CD4⁺ NK cell, seem to correlate with a diminished risk of PSC onset. The CD45RA⁻ CD4 not Treg subset, characterized by the expression of CD4, CD25, and CD127, falls under the category of CD4⁺ memory T cells. Recent research suggests that CD25⁺CD127hi cells may play a role in regulating inflammatory responses by favoring a Th2-type bias and dampening pro-inflammatory reactions. CD28, a crucial co-stimulatory molecule, is found to be depleted in CD28-deficient T cells observed in various autoimmune disorders, possibly due to prolonged antigen stimulation and exposure to TNF-α (34). A liquid chromatography-tandem MS assays confirm CD28 dysregulation in autoimmunity, with elevated glutamine and tryptophan derivatives like kynurenine linked to altered T-cell metabolism in PSC (35). This depletion may weaken the regulatory effects on CD4⁺CD25⁺ Treg cells, leading to chronic activation of CD28⁻ T cells. These activated T cells can migrate to liver tissues, accumulate around bile ducts, and release high levels of TNF-α and IFN-γ, promoting inflammatory progression in PSC by inducing the expression of inflammatory molecules in nearby bile epithelial cells Also, recent bile proteome studies in PSC confirm heightened inflammation through elevated IL-8 and ANXA1 (36), which correlate with T cell and NK cell activities, thus reinforcing the significance of CD28 and FSC-A phenotypes in PSC risk (23,37).

Our study possesses several strengths. Firstly, we used multiple MR methods to confirm our findings’ accuracy. Moreover, the potential bias stemming from weak IVs was mitigated, as all our F values exceeded 10. Reverse MR analysis demonstrated that genetically predicted AILDs had no causal effect on significant immune cells. Furthermore, by exclusively focusing on European populations, we minimized any potential bias arising from ethnic disparities. Therefore, these measures were taken to reduce the risk of plagiarism.

However, our study has certain limitations. Despite considering numerous immune cell phenotypes, some remained unexamined due to data limitations. Furthermore, the limited statistical data on such diseases in major global disease databases might compromise the universality of the analytical results. For robustness, future validation in larger, diverse patient cohorts is essential. This study initiates exploration into the connection between immune cell phenotypes and AILDs. Subsequent research can delve deeper into this association and explore potential immunotherapeutic targets, thus paving the way for innovative approaches to treating AILDs.

Conclusions

This study highlights the potentially causal correlation between immune cell phenotypes and AILDs, providing us with further clues to explore the pathogenesis of AILDs. Through an in-depth investigation of immune cell phenotypes, we may better understand the pathophysiological processes underlying AILDs, offering valuable insights for future functional studies to validate these associations and advance understanding of AILDs mechanisms.

Acknowledgments

We acknowledge the funding and the investigators who contributed to the research in this study.

Footnote

Reporting Checklist: The authors have completed the STROBE-MR reporting checklist. Available at https://tgh.amegroups.com/article/view/10.21037/tgh-25-2/rc

Peer Review File: Available at https://tgh.amegroups.com/article/view/10.21037/tgh-25-2/prf

Funding: This study was supported by the National Key Research and Development Program of China (No. 2022YFC2303600 to D.K.J.), the General Programs from the National Natural Science Foundation of China (No. 82272765 to D.K.J.), the Guangdong Basic and Applied Basic Research Foundation (No. 2022A1515220090 to D.K.J.), the Local Innovative and Research Teams Project of Guangdong Pearl River Talents Program (No. 2017BT01S131 to D.K.J.), the Innovative Research Team Project of Guangxi Province (No. 2017GXNSFGA198002 to D.K.J.), the Grant for Recruited Talents to Start Scientific Research from Nanfang Hospital, and the Outstanding Youth Development Scheme of Nanfang Hospital, Southern Medical University (No. 2017J001 to D.K.J.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tgh.amegroups.com/article/view/10.21037/tgh-25-2/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Heo NY, Kim H. Epidemiology and updated management for autoimmune liver disease. Clin Mol Hepatol 2023;29:194-6. [Crossref] [PubMed]
2. Lee BT, Tana MM, Kahn JA, et al. We Are Not Immune: Racial and Ethnic Disparities in Autoimmune Liver Diseases. Hepatology 2021;74:2876-87. [Crossref] [PubMed]
3. Trivedi PJ, Hirschfield GM, Adams DH, et al. Immunopathogenesis of Primary Biliary Cholangitis, Primary Sclerosing Cholangitis and Autoimmune Hepatitis: Themes and Concepts. Gastroenterology 2024;166:995-1019. [Crossref] [PubMed]
4. Neshat SY, Quiroz VM, Wang Y, et al. Liver Disease: Induction, Progression, Immunological Mechanisms, and Therapeutic Interventions. Int J Mol Sci 2021;22:6777. [Crossref] [PubMed]
5. Li L, Kang Y. The Gut Microbiome and Autoimmune Hepatitis: Implications for Early Diagnostic Biomarkers and Novel Therapies. Mol Nutr Food Res 2023;67:e2300043. [Crossref] [PubMed]
6. Richardson N, Wootton GE, Bozward AG, et al. Challenges and opportunities in achieving effective regulatory T cell therapy in autoimmune liver disease. Semin Immunopathol 2022;44:461-74. [Crossref] [PubMed]
7. Bowden J, Holmes MV. Meta-analysis and Mendelian randomization: A review. Res Synth Methods 2019;10:486-96. [Crossref] [PubMed]
8. Birney E. Mendelian Randomization. Cold Spring Harb Perspect Med 2022;12:a041302. [Crossref] [PubMed]
9. Orrù V, Steri M, Sidore C, et al. Complex genetic signatures in immune cells underlie autoimmunity and inform therapy. Nat Genet 2020;52:1036-45. [Crossref] [PubMed]
10. Sakaue S, Kanai M, Tanigawa Y, et al. A cross-population atlas of genetic associations for 220 human phenotypes. Nat Genet 2021;53:1415-24. [Crossref] [PubMed]
11. Cordell HJ, Fryett JJ, Ueno K, et al. An international genome-wide meta-analysis of primary biliary cholangitis: Novel risk loci and candidate drugs. J Hepatol 2021;75:572-81. [Crossref] [PubMed]
12. Ji SG, Juran BD, Mucha S, et al. Genome-wide association study of primary sclerosing cholangitis identifies new risk loci and quantifies the genetic relationship with inflammatory bowel disease. Nat Genet 2017;49:269-73. [Crossref] [PubMed]
13. Swerdlow DI, Kuchenbaecker KB, Shah S, et al. Selecting instruments for Mendelian randomization in the wake of genome-wide association studies. Int J Epidemiol 2016;45:1600-16. [Crossref] [PubMed]
14. Pierce BL, Ahsan H, Vanderweele TJ. Power and instrument strength requirements for Mendelian randomization studies using multiple genetic variants. Int J Epidemiol 2011;40:740-52. [Crossref] [PubMed]
15. Bowden J, Del Greco M F, Minelli C, et al. Improving the accuracy of two-sample summary-data Mendelian randomization: moving beyond the NOME assumption. Int J Epidemiol 2019;48:728-42. [Crossref] [PubMed]
16. Verbanck M, Chen CY, Neale B, et al. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet 2018;50:693-8. [Crossref] [PubMed]
17. Bowden J, Del Greco M F, Minelli C, et al. Assessing the suitability of summary data for two-sample Mendelian randomization analyses using MR-Egger regression: the role of the I2 statistic. Int J Epidemiol 2016;45:1961-74. [Crossref] [PubMed]
18. McCabe M, Rush N, Lammert C, et al. HLA-DR Mismatch and Black Race Are Associated With Recurrent Autoimmune Hepatitis After Liver Transplantation. Transplant Direct 2021;7:e714. [Crossref] [PubMed]
19. Yuming Z, Ruqi T, Gershwin ME, et al. Autoimmune Hepatitis: Pathophysiology. Clin Liver Dis 2024;28:15-35. [Crossref] [PubMed]
20. Xu W, Wang Y, Jin C, et al. IL-17 Imbalance Promotes the Pyroptosis in Immune-Mediated Liver Injury Through STAT3-IFI16 Axis. Immune Netw 2023;23:e46. [Crossref] [PubMed]
21. Berry SPD, Dossou C, Kashif A, et al. The role of IL-17 and anti-IL-17 agents in the immunopathogenesis and management of autoimmune and inflammatory diseases. Int Immunopharmacol 2022;102:108402. [Crossref] [PubMed]
22. Hegde S, Leader AM, Merad M. MDSC: Markers, development, states, and unaddressed complexity. Immunity 2021;54:875-84. [Crossref] [PubMed]
23. Olianas A, Guadalupi G, Cabras T, et al. Top-Down Proteomics Detection of Potential Salivary Biomarkers for Autoimmune Liver Diseases Classification. Int J Mol Sci 2023;24:959. [Crossref] [PubMed]
24. Zhao ZB, Lu FT, Ma HD, et al. Liver-resident NK cells suppress autoimmune cholangitis and limit the proliferation of CD4(+) T cells. Cell Mol Immunol 2020;17:178-89. [Crossref] [PubMed]
25. Pongracz T, Biewenga M, Stoelinga AEC, et al. Autoimmune hepatitis displays distinctively high multi-antennary sialylation on plasma N-glycans compared to other liver diseases. J Transl Med 2024;22:456. [Crossref] [PubMed]
26. Williams H, Mack C, Baraz R, et al. Monocyte Differentiation and Heterogeneity: Inter-Subset and Interindividual Differences. Int J Mol Sci 2023;24:8757. [Crossref] [PubMed]
27. Zhang W, Shao T, Leung PSC, et al. Dual B-cell targeting therapy ameliorates autoimmune cholangitis. J Autoimmun 2022;132:102897. [Crossref] [PubMed]
28. Grimsholm O. CD27 on human memory B cells-more than just a surface marker. Clin Exp Immunol 2023;213:164-72. [Crossref] [PubMed]
29. Lee S, Ko Y, Kim TJ. Homeostasis and regulation of autoreactive B cells. Cell Mol Immunol 2020;17:561-9. [Crossref] [PubMed]
30. Shi H, Wang Q, Liu H, et al. Gastroesophageal varices in primary biliary cholangitis with anti-centromere antibody positivity: Early onset? Open Life Sci 2024;19:20220979. [Crossref] [PubMed]
31. Bakhtiar S, Kaffenberger C, Salzmann-Manrique E, et al. Regulatory B cells in patients suffering from inborn errors of immunity with severe immune dysregulation. J Autoimmun 2022;132:102891. [Crossref] [PubMed]
32. Rogalska M, Błachnio-Zabielska A, Zabielski P, et al. Acylcarnitine and Free Fatty Acid Profiles in Primary Biliary Cholangitis: Associations with Fibrosis and Inflammation. Nutrients 2025;17:1097. [Crossref] [PubMed]
33. Deng CW, Wang L, Fei YY, et al. Exploring pathogenesis of primary biliary cholangitis by proteomics: A pilot study. World J Gastroenterol 2017;23:8489-99. [Crossref] [PubMed]
34. Tiegs G, Horst AK. TNF in the liver: targeting a central player in inflammation. Semin Immunopathol 2022;44:445-59. [Crossref] [PubMed]
35. Malkawi AK, Ohlund L, Rahman AMA, et al. Co-stimulatory pathway competitive assay development using Liquid chromatography-tandem mass spectrometry (LC-MS/MS). J Pharm Biomed Anal 2024;242:116034. [Crossref] [PubMed]
36. Kan M, Chiba T, Konno R, et al. Bile proteome analysis by high-precision mass spectrometry to examine novel biomarkers of primary sclerosing cholangitis. J Hepatobiliary Pancreat Sci 2023;30:914-23. [Crossref] [PubMed]
37. Kattelus R, Starskaia I, Lindén M, et al. Phenotypic profiling of human induced regulatory T cells at early differentiation: insights into distinct immunosuppressive potential. Cell Mol Life Sci 2024;81:399. [Crossref] [PubMed]