Skip to main content
Download PDF

Genetically predicted metabolite mediates the causal relationship between immune cells and autoimmune diseases

Arthritis Research & Therapy volume 26, Article number: 207 (2024) Cite this article

Abstract

Background

This study investigates the causal role of metabolites mediating immune cells in rheumatoid arthritis (RA) and ankylosing spondylitis (AS) through a Mendelian randomization (MR) study.

Methods

The two-sample and two-step MR methods were used for the current analysis: (1) causal effects of immune cells on RA and AS; (2) mediation effects of metabolites. Inverse variance weighted (IVW) is the main method to analyze causality, and MR results are verified by several sensitive analyses.

Results

This study first identified the immune cells and metabolites that are causally associated with RA and AS, respectively. Subsequent mediation analyses revealed that of the 61 metabolic factors that were causally associated with RA, 6 were identified as mediators of the relationship between immune cells and RA, including 4-cholesten-3-one levels (mediation ratio: 8.91%), N-lactoyl isoleucine levels (13%), 3- phosphoglycerate to glycerate ratio (12.9%, 2.31%, respectively), Gamma-glutamyl histidine levels (9.54%), and Citrulline to phosphate ratio (15.6%). Among the 52 metabolic factors that were causally associated with AS, 2 were identified as mediators of the relationship between immune cells and AS, including salicylate levels (10.4%) and Glucose to N-palmitoyl-sphingosine (d18:1 to 16:0) ratio (8.72%). These results performed well in sensitivity analysis.

Conclusions

Genetic predictions show causal relationships between immune cells and autoimmune diseases, and that these causal relationships can be mediated by certain metabolites as mediators.

Introduction

A heterogeneous group of diseases, autoimmune diseases (ADs), are characterized by damage and dysfunction of specific or multiple organs and tissues caused by abnormal immune responses at various functional parts of the body, including rheumatoid arthritis (RA), ankylosing spondylitis (AS), system lupus erythematosus (SLE), Graves’ disease (GD), etc [1,2,3]. The diversity of ADs influences the prevalence of the condition, which is also influenced by a number of variables including genetic, environmental, infectious, and epigenetic factors [4, 5]. According to epidemiologic research, 5–10% of individuals worldwide may be affected by ADs, and the incidence is increasing gradually [6, 7].

As common ADs, the development of RA and AS is intimately associated with tissue damage and immune system dysfunction [8]. Following an autoimmune response, antibodies combine with self-tissues to become autoantibodies and mediate the autoimmune process along with immune cells and other mediators, which is the quintessential molecular mimicry hypothesis for the initiation of autoimmune disease [9, 10]. As an important component of the immune system, immune cells produce cytokines or interact with lymphocytes under the influence of factors of pathogenicity to regulate the dynamic balance and participate in tissue repair [11]. In general, functional and phenotypic defects in certain immune cells mediate the development of ADs, as well as large accumulations of adaptive cells (e.g., T-cell subpopulations, B-cells, plasma cells, and plasmacytoid cells) and innate immune cells (e.g., monocytes, dendritic cells, mast cells, and innate lymphocytes) that contribute to the development of ADs through the activation of cytokines that create positive feedback [12, 13]. For example, a study revealed that initial CD4 T cells from RA patients accelerate senescence and reprogram the metabolic machinery [14]. Upon stimulation, these initial T cells in RA differentiate into hyperproliferative and tissue-invasive effector cells that initiate and maintain aggressive tissue inflammation in the synovium [14]. It has been shown that the percentage of CD8 + CD122 + T cells in the peripheral blood of AS is higher than that of healthy controls, and that the number of CD27 + B cells is decreased, while the number of CD86 + and CD27 + CD95 + B cells is increased [15]. In addition, immunometabolism suggests that associated core metabolic pathways including glycolysis, lipid metabolism, and mitochondrial function can also affect the development of ADs [16].

Human plasma metabolites are the end products of a group of cellular regulatory processes that can be utilized to identify pathophysiological components of complex diseases and to illuminate the complex relationships between genotype and phenotype [17]. Understanding the causal role of metabolites in disease etiology can offer actionable intervention points for treatment. Several epidemiological studies have elucidated the underlying biological mechanisms of the association between specific metabolic profiles and various diseases by investigating the identification of specific metabolic characteristics [18]. The interactions between the metabolic system and the immune system, which are regulated by genetics, nutritional status, and the gut microbiome, have recently attracted considerable attention [19]. Metabolic epigenetics suggests that metabolites can induce permanent changes in cellular organization and genetic structure by modulating the epigenome [20]. A plausible hypothesis is that an aberrant metabolic process within the body may disrupt immune tolerance, with the subsequent aberrations of metabolites in immune cells leading to autoimmune responses [21]. Metabolites are not only involved in the metabolic processes of immune cells, but also transmit information between immune cells through mechanisms such as receptors, transporter proteins, and post-translational modifications [22]. Modified immune-metabolic interactions can lead to the development of ADs [23]. Although observational studies have attempted to elucidate the causal relationship between immune cells, metabolites, and ADs, it is difficult to establish a clear causal relationship because the results may be susceptible to bias due to unanticipated confounding factors or reverse causation [24].

To address confounders and determine causal relationships in observational studies, Mendelian randomization (MR) studies were developed based on Mendelian independent classification rules [25]. In MR studies, genetic variation follows the principle of random assignment of alleles to offspring, similar to randomized controlled experiments. According to the law of segregation assortment and the law of independent assortment, the assignment of genotypes from parent to offspring is random, and therefore the association between genetic variation and outcome is unlikely to be influenced by the environment [26]. Therefore, methods that utilize genetic variation as instrumental variables (IVs) to assess potential causal associations between exposures and outcomes can be effective in addressing confounders and reverse causation in observational epidemiology [27]. The objective of this study was to use a MR study aimed at revealing the potential causal relationship between immune cells and Ads. And importantly, we assessed the mediating role of metabolites on exposure-outcome through mediation analysis. This research approach enhances the comprehension of the pathogenesis and metabolic pathways of RA and AS, while also offering dependable evidence for the development of feasible screening and prevention strategies for ADs.

Materials and methods

Study design

Based on a two-sample MR study, we first assessed causality between exposure (731 immune cell characteristics) and outcome (RA and AS). In addition, the mediated (two-step approach) MR study was employed to delve deeper into the mediating effects between exposure and outcome mediated by metabolites as mediators (Fig. 1). Risk factors are represented by genetic variation also called IVs in MR study. Three key assumptions must be met by IVs in causal inference: (1) IVs must be strongly associated with exposure; (2) IVs are independent of possible confounders between exposure and outcome; and (3) Genetic variation does not affect outcome through pathways other than exposure (no horizontal pleiotropy). (Fig. 1)

Fig. 1
figure 1

Flowchart of the analysis performed. ADs, autoimmune diseases; RA, rheumatoid arthritis; AS, ankylosing spondylitis

Full size image

Data sources

The Genome-Wide Association Study (GWAS) catalog (GCST90199621 to GCST90201020) contains summary data for 731 immunophenotypes included in the study [32]. The data from 3757 Sardinians was used to calculate around 22 million genetic variations and validate relationships with autoimmune illnesses in the GWAS data on immunological characteristics. This data identifies molecules and mechanisms involved in the regulation of 459 cellular features by detecting 122 significant independently associated signals at 70 loci, including 118 absolute cell counts (AC), 192 relative cell counts (RC) representing the ratio of cellular levels, 389 median fluorescence intensity (MFI) representing surface antigen levels, and 32 morphological characteristics (MP). The GWAS data for plasma metabolites were retrieved from the GWAS Catalog (GCST0001391 to GCST0002121). The plasma metabolite data comprised 1,091 blood metabolites and 309 metabolite ratios [28]. Summary data can be obtained by visiting: https://ftp.ebi.ac.uk/pub/databases/gwas/summary_statistics/.

Summary data for both AS and RA were obtained from the MRC-IEU consortium [IEU OpenGWAS project (https://gwas.mrcieu.ac.uk)]. RA included 417,256 participants of European ancestry (https://gwas.mrcieu.ac.uk/datasets/ebi-a-GCST90018910/). AS included 166,144 participants of European ancestry (https://gwas.mrcieu.ac.uk/datasets/finn-b-M13_ANKYLOSPON/) [29].

Instrument variables selection

For IVs, we utilized a stringent quality control procedure to ensure data reliability and that sufficient SNPs were available for MR analysis. Therefore, based on previous studies, we set the genome-wide threshold for SNPs associated with two exposures (immune cells and metabolites) to 1 × 10− 5 [30, 31]. In addition, we excluded SNPs that are not required using strand linkage disequilibrium (LD, R2 < 0.001 and clumping distance = 10,000 kb) to ensure that IVs were independent of each other [32]. We selected SNPs with a minor allele frequency (MAF) > 0.01 and excluded palindromic SNPs [33]. Finally, we calculated and excluded F-statistics (F = R2(n-k-1)/k(1-R2); R2, exposed variance explained by the selected instrumental variable, we obtained the value of R2 in the MR Steiger directionality test; n, sample size; k, number of IVs) < 10 SNPs to avoid biasing the results by weak instrumental variables [34].

Statistical analysis

Mendelian randomization statistical analysis

A variety of methods were used to determine whether a causal relationship exists between immune cells with RA and AS, including inverse variance weighted (IVW) [35], MR-Egger [33], weighted median [36], weighted model [37], and simple model [38] methods. Of these, IVW employs meta-analysis to integrate the Wald estimates for each SNP to provide an overall estimate of the effect of exposure on outcome. When horizontal pleiotropy is absent, unbiased causal estimates can be developed by IVW linear regression. Thus, the IVW method is the most direct and important method for evaluating the presence of causality, with the remaining four modalities as auxiliary analysis methods. P < 0.05 indicated that there was a significant causal relationship between exposure and outcome. In this study, we ensured that the results of all methods were in the same direction as IVW (i.e., the beta and odds ratio (OR) values were in the same direction) [39].

Mediation statistical analysis

In the present study, positive MR analysis was conducted to identify immune cells that were identified as having a potential relationship with RA and AS, respectively (causal effect value of BetaAll). In the subsequent mediation analysis, metabolites with a potential relationship with RA and AS were first identified (causal effect value of Beta2), followed by MR analysis of the above obtained immune cells and metabolites (causal effect value of Beta1). The mediating effect was calculated as Beta12 = Beta1 × Beta2; the ratio of the mediating effect to the total effect: Beta12Ratio = Beta12/Beta All×100%; and the direct effect: BetaDirect = BetaAll - Beta12 (Fig. 1) [40]. We also calculated confidence (CI) intervals and P-values for the mediating ratio.

Sensitivity statistical analysis

By employing the MR-Egger regression [41] intercept analysis and MR-PRESSO test [33], we tested the results for horizontal pleiotropy. The intercept in MR-Egger regression can be used to assess horizontal pleiotropy, which if significant indicates the presence of horizontal pleiotropy, and vice versa, indicates the absence of horizontal pleiotropy. There are three components of MR-PRESSO, which include: horizontal pleiotropy test (MR-PRESSO global test), correction of horizontal pleiotropy by removing outliers (MR-PRESSO outlier test), and a test for significant differences in causal estimates before and after correcting for outliers (MRPRESSO distortion test).The MR-PRESSO outlier test requires that at least 50% of the genetic variance be valid IVs and relies on the instrumental strength to be independent of the direct effect (InSIDE) condition [33]. After excluding results with horizontal pleiotropy, we conducted subsequent analyses. Heterogeneity of SNP was tested using Cochran’s Q statistic, with P-values < 0.05 indicating significant heterogeneity [42]. In addition, to avoid horizontal pleiotropy caused by a single SNP, we performed “leave-one-out” sensitivity analysis to ensure the reliability of the results [43].

All MR analyses were conducted in R (version 4.3.2; R Foundation for Statistical Computing, https://www.r-project.org/foundation/) using the “TwoSampleMR”, “VariantAnnotation”, “gwasglue” and “data.table” packages [44].

Result

Summary data on immune cells and metabolites were screened for SNPs and subsequently analyzed by MR using threshold (1 × 10− 5), LD standards and F-statistics. Detailed IVs data are shown in Supplementary Tables 1 and 2. The results of the causal relationship between immune cells with RA and AS are shown in Supplementary Tables 3 and 4, Visualization of scatterplots in Supplementary Figs. 1 and 2.

Two‑sample MR analysis

Causal effect of immunophenotypes on RA and AS risk

By MR analysis, we obtained a total of 32 immunophenotypes and 16 immunophenotypes causally associated with RA and AS, respectively (Fig. 2, Supplementary Table 5). The causal relationship between immune cells and RA showed that 19 immunophenotypes were genetically predicted to be associated with a reduced risk of RA and 13 immunophenotypes were associated with an increased risk of RA. The causal relationship between immune cells and AS showed that the genetic prediction of 4 immunophenotypes was associated with decreased risk of AS and 12 immunophenotypes were associated with increased risk of AS. Among them, CD11b on basophil was a common risk factor for RA (OR 1.050, 95% CI, 1.016–1.086; P = 0.004) and AS (OR 1.111, 95% CI, 1.010–1.222; P = 0.03).

Fig. 2
figure 2

Visualization of the causal relationship between immune cells and ADs. (A), rheumatoid arthritis; (B), ankylosing spondylitis

Full size image

Two-step MR analysis

Causal effect of metabolites on RA and AS

In the two-step MR analysis, we first performed MR analysis of metabolites with RA and AS to look for metabolites that had a causal relationship with outcome. The causal relationship between metabolites and RA showed that 18 metabolites were genetically predicted to be associated with a reduced risk of RA and 43 were associated with an increased risk of RA. These included 30 currently known metabolites, 11 metabolites not yet studied, and 20 metabolite ratios (Supplementary Table 6). The causal relationship between metabolites and AS showed that 24 metabolites were genetically predicted to be associated with a reduced risk of AS and 28 were associated with an increased risk of AS. These included 27 currently known metabolites, 8 metabolites that have not been studied and 17 metabolite ratios (Supplementary Table 7). Among them, Sulfate of piperine metabolite C18H21NO3 (1) levels was a common risk factor for RA (OR 1.105, 95% CI, 1.018-1.200; P = 0.017) and AS (OR 1.306, 95% CI, 1.015–1.682; P = 0.017). In addition, Adenosine 5’-monophosphate (AMP) to serine ratio was positively correlated with RA (OR 1.113, 95% CI, 1.016–1.219; P = 0.021) but negatively correlated with AS (OR 0.734, 95% CI, 0.582–0.926; P = 0.009). Caffeine to linoleate (18:2n6) ratio was negatively correlated with RA (OR 0.891, 95% CI, 0.796–0.999; P = 0.048) and positively correlated with AS (OR 1.246, 95% CI, 1.004–1.545; P = 0.046). It should be mentioned that we excluded metabolites that had not been studied for subsequent MR analysis.

Causal effect of immune cells on metabolites

Specific immunophenotypes (IVW, P < 0.01) were used for subsequent MR analysis with metabolites (Fig. 3). The immunophenotypes and outcome-related metabolites described above were subjected to MR analysis to determine mediating effects. In subsequent mediation analysis, we found that of the 11 immunophenotypes causally associated with RA, there were 9 immunophenotypes causally associated with a total of 15 metabolites. Three immunophenotypes causally associated with AS were causally associated with a total of 7 metabolites. (Supplementary Table 8)

Fig. 3
figure 3

Forest plot of the causal relationship between immune cells and ADs. (Inverse variance weighted, IVW, P < 0.01)

Full size image

Mediation analysis of metabolites

Finally, we performed mediation analysis and excluded metabolites with a P value of > 0.05 for mediation effects (Table 1). The results showed that CD28 + CD45RA + CD8dim AC mediated its effect on AS through Salicylate levels and Glucose to N-palmitoyl-sphingosine (d18:1 to 16:0) ratio with proportions of 10.4% (P = 0.013) and 8.72% (P = 0.034), respectively. CD3 on TD CD4 + mediated its effect on RA through 4-cholesten-3-one levels with a proportion of 8.91% (P = 0.040); PDL-1 on CD14- CD16 + monocyte mediated its effect on RA through N-lactoyl isoleucine levels with a proportion of 8.91% (P = 0.040); CD16 on CD14 + CD16 + monocyte and HLA DR on CD14- CD16- mediated their effects on RA through the common metabolite 3-phosphoglycerate to glycerate ratio at 12.9% (P = 0.021) and 2.31% (P = 0.015); CD14 on CD33dim HLA DR + CD11b + mediated its effect on RA through Gamma-glutamylhistidine levels and Citrulline to phosphate ratio, which were 9.54% (P = 0.018) and 15.6% (P = 0.008). (Fig. 4) These ratios emphasize the complex dynamic relationship between specific immune cells, metabolites with RA and AS.

Table 1 The mediation effect of immune cells on RA and AS via metabolites
Full size table
Fig. 4
figure 4

Mediating proportions of each mediator in the causal relationship between immune cells and ADs

Full size image

Sensitivity analysis

The analysis performed using the MR-Egger regression intercept method and MR-PRESSO did not reveal any outliers, indicating that there was no cross-sectional pleiotropy (P > 0.05) (Supplementary Tables 9 and 10). The possibility of horizontal pleiotropy was ruled out by MR-egger intercept (Supplementary Tables 11 and 12). In addition, Cochran’s Q statistic, “leave-one-out” sensitivity analysis and funnel plots demonstrated the stability of the results (Supplementary Table 13, Supplementary Figs. 3, 4, 5 and 6).

Discussion

The complex relationship between immune cells, metabolites, and immune-mediated diseases has been a topic of interest in recent years. Immune cells and metabolites have a complex relationship that affects health and disease, and disruption of the dynamic balance can lead to the development of autoimmune diseases [10]. We have taken the summary data obtained from the GWAS database and used MR methods to delve deeper into the relationship between specific immune cells with RA and AS, and explored the mediating effects of metabolites in this context, shedding light on the complex relationship between the three in terms of gene prediction, providing compelling insights into this complex interaction, and confirming that some immunophenotypes can have an effect on RA and AS mediated by metabolites. This is the first time that we have used genetic prediction to investigate the potential causal relationships and mediating effects among the above three, and our findings can better guide our understanding of RA and AS, as well as shed light on the prevention and treatment of the diseases.

The immunopathogenesis of RA, characterized by immune cell infiltration in the joints, has spanned decades [45, 46]. According to research, several immune cell types, including T cells, B cells, monocytes, and myeloid cells, have been related to the pathogenesis of RA [14]. In this work, we identified 32 immunophenotypes that are causally linked to RA genetically. More significantly, for the first time, we screened 5 immunophenotypes (CD3 on TD CD4 + T cell, PDL-1 on CD14- CD16 + monocyte, CD16 on CD14 + CD16 + monocyte, HLA DR on CD14- CD16- monocyte, CD14 on CD33dim HLA DR + CD11b + Myeloid cell) from this group, which demonstrated the potential to influence the progression of RA by modulating related metabolites. Innate blood cells, referred to as monocytes, are responsible for preserving vascular homeostasis and acting as early responders to pathogens in acute infections. Moreover, classical monocytes contain the ability to develop into macrophages in tissues, which can result in long-term illnesses [47]. Subgroups of peripheral blood monocytes can be distinguished by their CD14/CD16 surface characteristics. Our research indicates that the expression of PDL-1, CD16, and HLA DR in different monocyte subtypes influences (CD14- CD16 + monocyte, CD14 + CD16 + monocyte, CD14- CD16- monocyte) the development of RA through specific metabolites. Previous studies have shown that chronic otitis media typically coexists with increased PDL-1, a checkpoint molecule in circulating monocytes [48]. However, in RA, more studies have focused on the role of PDL-1 in regulating the activity of peripheral T cells [49, 50]. Currently, there are just a few studies on N-lactoyl isoleucine, a critical metabolite in PDL-1 on CD14- CD16 + monocyte-mediated RA illness. Only a single population-based longitudinal investigation demonstrated that N-lactoyl isoleucine is substantially linked to the development of diabetic retinopathy in individuals with type 2 diabetes mellitus [51]. Generally speaking, most human peripheral blood monocytes strongly express surface CD14, but not CD16 (CD14+ +/CD 16-) [52]. When evaluating the frequency of CD16 + cells in all blood mononuclear cells between RA patients and healthy controls, Andrew et al. observed a statistically significant increase [52]. The involvement of HLA-DR in RA was brought to the attention of researchers long before 1989, whose study revealed that, compared to normal controls, patients with active RA presented lower densities of monocytes expressing HLA-DR [53]. This provides additional evidence that HLA-DR in peripheral monocytes can be employed as a therapeutic target to influence disease progression in RA. Surprisingly, our study also revealed that the 3-phosphoglycerate to glycerate ratio is the metabolite status that mediates the effects of CD16 on CD14 + CD16 + monocyte and HLA DR on CD14- CD16- monocyte on RA progression. Many studies have demonstrated the critical role that glucose metabolism plays in RA etiology. 3-phosphoglycerate is a glycolysis intermediate metabolite. Increased expression of phosphoglycerate kinase 1 (PGK1), a glycolytic enzyme related to 3-phosphoglycerate, in RA synovial tissue and blood may be associated with the pro-inflammatory and synovial proliferation of the disease [54]. In addition, the research by Samantha et al. suggests the possibility that glycerate dysregulation is a key mechanism leading to early RA [55]. Undoubtedly, our study provides new insights into the matter.

According to this study, CD3 on TD CD4 + T cells also mediates 4-cholesten-3-one levels affecting RA progression. Growing evidence reveals that RA is a T-cell-mediated autoimmune disease. Abnormalities in the quantity and function of CD4 + T lymphocytes in the peripheral blood are strongly associated with the development and progression of RA [56]. Several studies have also indicated that CD4 + T-cells and their cytokine products represent potential therapeutic and diagnostic targets in RA [57]. Xiong et al. examined the expression of CD3 in synovial tissues of RA and osteoarthritis (OA) patients, respectively, and revealed that the synovial tissues of the RA showed increased expression of CD3, suggesting that high levels of CD3 may promote the progression of RA [58]. Several of the most comprehensive studies concerning 4-cholesten-3-one focus on the metabolism of bile acids, which are not currently recognized to perform a role in RA [59]. Furthermore, CD14 on CD33dim HLA DR + CD11b + Myeloid cell can influence RA by modulating Gamma-glutamyl histidine levels and Citrulline to phosphate ratio. The first investigation into the function of Gamma-glutamyl histidine in RA was conducted in 1984 by Dixon et al., when researchers examined plasma from patients with RA and Behçet’s syndrome and revealed biochemical variations in A between the two conditions [60]. Similar to 3-phosphoglycerate, Citrulline to phosphate ratio typically contributes significantly to the regulation of associated disorders through energy metabolism [61, 62].

Furthermore, this study discovered that AS is related to CD28 + CD45RA + CD8dim AC Treg cell by modulating Salicylate levels and Glucose to N-palmitoyl-sphingosine (d18:1 to 16:0) ratio. AS is an autoimmune disease characterized by clonal proliferation of T-cells. Previous research has demonstrated that patients with AS exhibit different CD8 + Treg cell differentiation patterns from normal participants, which may contribute to the pathogenesis of AS [63]. Compared to male controls, male patients with AS exhibited a 1.5- to 2-fold increase in Th17 (P = 0.013) among CD45 + lymphocytes [64]. A copper-salicylate preparation, permalon, earlier employed as an effective drug in the treatment of rheumatoid and other degenerative diseases [65]. It has been demonstrated that the metabolism of salicylic acid regulates immunity to promote both growth and defense in plants, which implies that it could have an impact on AS progression via the same pathways [66]. A sphingolipid involved in cell signaling, ceramide, frequently accumulates in tissues and plasma during metabolic dysfunction, dyslipidemia and inflammation. N-palmitoyl-sphingosine (d18:1 to 16:0) is a special type of ceramide. Increased plasma N-palmitoyl-sphingosine (d18:1 to 16:0) may be substantially correlated with inflammatory bowel disease, according to a study of mucosal and plasma metabolomes in new-onset pediatric inflammatory bowel disease [67]. However, the above metabolite profiles have not been investigated in AS patients. Undoubtedly, our findings will enrich the theory of metabolite pathogenesis in AS patients.

Overall, this study has several strengths. Firstly, a total of 32 immunophenotypes and 16 immunophenotypes may be causally associated with RA and AS employing MR analysis. Secondly, our study also identified 61 and 52 metabolites at the genetic level that may contribute to the progression of RA and AS, respectively. Importantly, we also found that specific immunophenotypes may influence disease progression by regulating specific metabolites, which were not mentioned in previous studies. Without a doubt, this research offers fresh perspectives on the etiology of particular autoimmune diseases, such as RA and AS. Our study is not without limits, though. First, given that the majority of those participating in the data aggregated in this study were of European descent, the ability to extrapolate findings to other ethnic groups may be limited. Second, the current research samples were not evaluated by additional subgroup stratification analysis since individual patient information is not currently included in the GWAS data. Third, while immunophenotypes and metabolites that might be associated with the development of RA and AS were identified by MR analysis, biological mechanisms should also be taken into account when interpreting the data. Biological mechanisms are not disclosed by the statistical effect values alone. This means, of course, that stronger biological evidence is needed to support this study.

Conclusions

In conclusion, we employed MR analysis to further clarify the patterns of RA and AS interactions with the immune system as well as provide insight into the associations between immunological phenotypes, metabolites, and diseases. In addition to offering believable genetic evidence, this finding also provides valuable guidelines for future mechanistic and clinical research on ADs.

Data availability

No datasets were generated or analysed during the current study.

References

  1. Schett G, McInnes IB, Neurath MF. Reframing Immune-mediated inflammatory diseases through signature cytokine hubs. N Engl J Med. 2021;385(7):628–39.

    Article  CAS  PubMed  Google Scholar 

  2. Schett G, Mackensen A, Mougiakakos D. CAR T-cell therapy in autoimmune diseases. Lancet. 2023;402(10416):2034–44.

    Article  CAS  PubMed  Google Scholar 

  3. Wang L, Wang FS, Gershwin ME. Human autoimmune diseases: a comprehensive update. J Intern Med. 2015;278(4):369–95.

    Article  CAS  PubMed  Google Scholar 

  4. Wang Q, Feng D, Jia S, Lu Q, Zhao M. B-Cell receptor repertoire: recent advances in Autoimmune diseases. Clin Rev Allergy Immunol. 2024;66(1):76–98.

    Article  PubMed  Google Scholar 

  5. Xiao ZX, Miller JS, Zheng SG. An updated advance of autoantibodies in autoimmune diseases. Autoimmun Rev. 2021;20(2):102743.

    Article  CAS  PubMed  Google Scholar 

  6. Tsoukalas D, Fragoulakis V, Papakonstantinou E, Antonaki M, Vozikis A, Tsatsakis A, Buga AM, Mitroi M, Calina D. Prediction of Autoimmune Diseases by Targeted Metabolomic Assay of Urinary Organic Acids. Metabolites 2020, 10(12).

  7. Miller FW. The increasing prevalence of autoimmunity and autoimmune diseases: an urgent call to action for improved understanding, diagnosis, treatment, and prevention. Curr Opin Immunol. 2023;80:102266.

    Article  CAS  PubMed  Google Scholar 

  8. Harroud A, Hafler DA. Common genetic factors among autoimmune diseases. Science. 2023;380(6644):485–90.

    Article  CAS  PubMed  Google Scholar 

  9. Wucherpfennig KW, Strominger JL. Molecular mimicry in T cell-mediated autoimmunity: viral peptides activate human T cell clones specific for myelin basic protein. Cell. 1995;80(5):695–705.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Blanco LP, Kaplan MJ. Metabolic alterations of the immune system in the pathogenesis of autoimmune diseases. PLoS Biol. 2023;21(4):e3002084.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Fiorillo MT, Haroon N, Ciccia F, Breban M. Editorial: Ankylosing Spondylitis and Related Immune-mediated disorders. Front Immunol. 2019;10:1232.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Liu J, Ting JP, Al-Azzam S, Ding Y, Afshar S. Therapeutic advances in diabetes, autoimmune, and neurological diseases. Int J Mol Sci 2021, 22(6).

  13. Ohno H. [Gut microbiota, immunity, and autoimmune diseases]. Rinsho Ketsueki. 2021;62(8):900–8.

    PubMed  Google Scholar 

  14. Weyand CM, Goronzy JJ. Immunometabolism in the development of rheumatoid arthritis. Immunol Rev. 2020;294(1):177–87.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Fei Y, Yu H, Wu Y, Gong S. The causal relationship between immune cells and ankylosing spondylitis: a bidirectional mendelian randomization study. Arthritis Res Ther. 2024;26(1):24.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Zhao L, Kostic AD, Zhang X. Editorial: gut microbiome, metabolites, and immune response in lupus, rheumatoid arthritis, and Sjogren syndrome. Front Immunol. 2022;13:1126800.

    Article  CAS  PubMed  Google Scholar 

  17. Shi H, Kichaev G, Pasaniuc B. Contrasting the Genetic Architecture of 30 complex traits from Summary Association Data. Am J Hum Genet. 2016;99(1):139–53.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Zhang J, Fang XY, Leng R, Chen HF, Qian TT, Cai YY, Zhang XH, Wang YY, Mu M, Tao XR, et al. Metabolic signature of healthy lifestyle and risk of rheumatoid arthritis: observational and mendelian randomization study. Am J Clin Nutr. 2023;118(1):183–93.

    Article  PubMed  Google Scholar 

  19. Medina G, Vera-Lastra O, Peralta-Amaro AL, Jimenez-Arellano MP, Saavedra MA, Cruz-Dominguez MP, Jara LJ. Metabolic syndrome, autoimmunity and rheumatic diseases. Pharmacol Res. 2018;133:277–88.

    Article  CAS  PubMed  Google Scholar 

  20. Dai Z, Ramesh V, Locasale JW. The evolving metabolic landscape of chromatin biology and epigenetics. Nat Rev Genet. 2020;21(12):737–53.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Yu XH, Cao RR, Yang YQ, Lei SF. Identification of causal metabolites related to multiple autoimmune diseases. Hum Mol Genet. 2022;31(4):604–13.

    Article  CAS  PubMed  Google Scholar 

  22. Zhang B, Vogelzang A, Fagarasan S. Secreted immune metabolites that mediate immune cell communication and function. Trends Immunol. 2022;43(12):990–1005.

    Article  CAS  PubMed  Google Scholar 

  23. Zmora N, Bashiardes S, Levy M, Elinav E. The role of the Immune System in Metabolic Health and Disease. Cell Metab. 2017;25(3):506–21.

    Article  CAS  PubMed  Google Scholar 

  24. Liao HT, Tsai CY. Cytokines and regulatory T cells in ankylosing spondylitis. Bone Joint Res. 2023;12(2):133–7.

    Article  PubMed  PubMed Central  Google Scholar 

  25. Bowden J, Holmes MV. Meta-analysis and mendelian randomization: a review. Res Synth Methods. 2019;10(4):486–96.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Davey Smith G, Hemani G. Mendelian randomization: genetic anchors for causal inference in epidemiological studies. Hum Mol Genet. 2014;23(R1):R89–98.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Sanderson E, Glymour MM, Holmes MV, Kang H, Morrison J, Munafo MR, Palmer T, Schooling CM, Wallace C, Zhao Q et al. Mendelian randomization. Nat Rev Methods Primers 2022, 2.

  28. Chen Y, Lu T, Pettersson-Kymmer U, Stewart ID, Butler-Laporte G, Nakanishi T, Cerani A, Liang KYH, Yoshiji S, Willett JDS, et al. Genomic atlas of the plasma metabolome prioritizes metabolites implicated in human diseases. Nat Genet. 2023;55(1):44–53.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Sakaue S, Kanai M, Tanigawa Y, Karjalainen J, Kurki M, Koshiba S, Narita A, Konuma T, Yamamoto K, Akiyama M, et al. A cross-population atlas of genetic associations for 220 human phenotypes. Nat Genet. 2021;53(10):1415–24.

    Article  CAS  PubMed  Google Scholar 

  30. Liu X, Tong X, Zou Y, Lin X, Zhao H, Tian L, Jie Z, Wang Q, Zhang Z, Lu H, et al. Mendelian randomization analyses support causal relationships between blood metabolites and the gut microbiome. Nat Genet. 2022;54(1):52–61.

    Article  CAS  PubMed  Google Scholar 

  31. Sanna S, van Zuydam NR, Mahajan A, Kurilshikov A, Vich Vila A, Vosa U, Mujagic Z, Masclee AAM, Jonkers D, Oosting M, et al. Causal relationships among the gut microbiome, short-chain fatty acids and metabolic diseases. Nat Genet. 2019;51(4):600–5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Liu X, Qi X, Han R, Mao T, Tian Z. Gut microbiota causally affects cholelithiasis: a two-sample mendelian randomization study. Front Cell Infect Microbiol. 2023;13:1253447.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from mendelian randomization between complex traits and diseases. Nat Genet. 2018;50(5):693–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Burgess S, Thompson SG, Collaboration CCG. Avoiding bias from weak instruments in mendelian randomization studies. Int J Epidemiol. 2011;40(3):755–64.

    Article  PubMed  Google Scholar 

  35. Palmer TM, Lawlor DA, Harbord RM, Sheehan NA, Tobias JH, Timpson NJ, Davey Smith G, Sterne JA. Using multiple genetic variants as instrumental variables for modifiable risk factors. Stat Methods Med Res. 2012;21(3):223–42.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estimation in mendelian randomization with some Invalid instruments using a weighted median estimator. Genet Epidemiol. 2016;40(4):304–14.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Li C, Niu M, Guo Z, Liu P, Zheng Y, Liu D, Yang S, Wang W, Li Y, Hou H. A mild causal relationship between tea consumption and obesity in General Population: a two-sample mendelian randomization study. Front Genet. 2022;13:795049.

    Article  PubMed  PubMed Central  Google Scholar 

  38. Hartwig FP, Davey Smith G, Bowden J. Robust inference in summary data mendelian randomization via the zero modal pleiotropy assumption. Int J Epidemiol. 2017;46(6):1985–98.

    Article  PubMed  PubMed Central  Google Scholar 

  39. Sun D, Zhou T, Heianza Y, Li X, Fan M, Fonseca VA, Qi L. Type 2 diabetes and hypertension. Circ Res. 2019;124(6):930–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Chen L, Peters JE, Prins B, Persyn E, Traylor M, Surendran P, Karthikeyan S, Yonova-Doing E, Di Angelantonio E, Roberts DJ, et al. Systematic mendelian randomization using the human plasma proteome to discover potential therapeutic targets for stroke. Nat Commun. 2022;13(1):6143.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44(2):512–25.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Bowden J, Del Greco MF, Minelli C, Zhao Q, Lawlor DA, Sheehan NA, Thompson J, Davey Smith G. Improving the accuracy of two-sample summary-data mendelian randomization: moving beyond the NOME assumption. Int J Epidemiol. 2019;48(3):728–42.

    Article  PubMed  Google Scholar 

  43. Hemani G, Tilling K, Davey Smith G. Orienting the causal relationship between imprecisely measured traits using GWAS summary data. PLoS Genet. 2017;13(11):e1007081.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, Laurin C, Burgess S, Bowden J, Langdon R et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife 2018, 7.

  45. Weyand CM, Goronzy JJ. The immunology of rheumatoid arthritis. Nat Immunol. 2021;22(1):10–8.

    Article  CAS  PubMed  Google Scholar 

  46. van Delft MAM, Huizinga TWJ. An overview of autoantibodies in rheumatoid arthritis. J Autoimmun. 2020;110:102392.

    Article  PubMed  Google Scholar 

  47. Narasimhan PB, Marcovecchio P, Hamers AAJ, Hedrick CC. Nonclassical monocytes in Health and Disease. Annu Rev Immunol. 2019;37:439–56.

    Article  CAS  PubMed  Google Scholar 

  48. Leichtle A, Jeschke S, Plötze-Martin K, Idel C, Bruchhage KL, Pries R. 1,8-Cineol attenuates checkpoint molecule PDL-1 and adhesion molecule CX3CR1 in circulating monocytes in Otitis Media patients. J Pers Med 2024, 14(3).

  49. Samarpita S, Rasool M. Majoon Chobchini reinstates PDL-1 expression and blocks dendritic cell -T helper 17 pathogenic axis in rheumatoid arthritis animal model. Cytokine. 2023;163:156136.

    Article  CAS  PubMed  Google Scholar 

  50. Raptopoulou AP, Bertsias G, Makrygiannakis D, Verginis P, Kritikos I, Tzardi M, Klareskog L, Catrina AI, Sidiropoulos P, Boumpas DT. The programmed death 1/programmed death ligand 1 inhibitory pathway is up-regulated in rheumatoid synovium and regulates peripheral T cell responses in human and murine arthritis. Arthritis Rheum. 2010;62(7):1870–80.

    Article  CAS  PubMed  Google Scholar 

  51. Fernandes Silva L, Hokkanen J, Vangipurapu J, Oravilahti A, Laakso M. Metabolites as risk factors for Diabetic Retinopathy in patients with type 2 diabetes: a 12-Year follow-up study. J Clin Endocrinol Metab. 2023;109(1):100–6.

    Article  PubMed  PubMed Central  Google Scholar 

  52. Cairns AP, Crockard AD, Bell AL. The CD14 + CD16 + monocyte subset in rheumatoid arthritis and systemic lupus erythematosus. Rheumatol Int. 2002;21(5):189–92.

    Article  PubMed  Google Scholar 

  53. Hassan J, Feighery C, Bresnihan B, Whelan A. Expression and regulation of the HLA-DR antigen on circulating monocytes isolated from patients with rheumatoid arthritis. Arthritis Rheum. 1989;32(6):682–90.

    Article  CAS  PubMed  Google Scholar 

  54. Zhao Y, Yan X, Li X, Zheng Y, Li S, Chang X. PGK1, a glucose metabolism enzyme, may play an important role in rheumatoid arthritis. Inflamm Res. 2016;65(10):815–25.

    Article  CAS  PubMed  Google Scholar 

  55. Rodríguez-Muguruza S, Altuna-Coy A, Arreaza-Gil V, Mendieta-Homs M, Castro-Oreiro S, Poveda-Elices MJ. Del Castillo-Piñol N, Fontova-Garrofé R, Chacón MR: a serum metabolic biomarker panel for early rheumatoid arthritis. Front Immunol. 2023;14:1253913.

    Article  PubMed  PubMed Central  Google Scholar 

  56. Wang L, Wang L, Hao P, Cao Q, Zhang Z. Anti-CCL22 increases regulatory T cells in CD4(+) T cells of rheumatoid arthritis patients via STAT5 pathway. Exp Ther Med. 2020;19(3):2127–32.

    CAS  PubMed  Google Scholar 

  57. Kinne RW, Emmrich F, Freesmeyer M. Clinical impact of radiolabeled anti-CD4 antibodies in the diagnosis of rheumatoid arthritis. Q J Nucl Med Mol Imaging. 2010;54(6):629–38.

    CAS  PubMed  Google Scholar 

  58. Xiong G, Lei T, Dong S, Xu L, Li M, Wang R. Roles of CD3, CD4 and CD8 in synovial lymphocytes of rheumatoid arthritis. Pol J Pathol. 2022;73(1):21–6.

    Article  PubMed  Google Scholar 

  59. Zhao X, Zhang W, Vig P, Kostrub C, Setchell KDR. Serum bile acid profiling and Mixed Model Analysis Reveal Biomarkers Associated with Pruritus reduction in Maralixibat-treated patients with BSEP Deficiency. Metabolites 2022, 12(10).

  60. Dixon JS, Yurdakul S, Surrall KE, Yazici H, Chamberlain MA. A study of serum biochemistry in Behçet’s syndrome. Br J Rheumatol. 1984;23(4):283–7.

    Article  CAS  PubMed  Google Scholar 

  61. Saheki T, Moriyama M, Kuroda E, Funahashi A, Yasuda I, Setogawa Y, Gao Q, Ushikai M, Furuie S, Yamamura KI, et al. Pivotal role of inter-organ aspartate metabolism for treatment of mitochondrial aspartate-glutamate carrier 2 (citrin) deficiency, based on the mouse model. Sci Rep. 2019;9(1):4179.

    Article  PubMed  PubMed Central  Google Scholar 

  62. Ajaz S, McPhail MJ, Gnudi L, Trovato FM, Mujib S, Napoli S, Carey I, Agarwal K. Mitochondrial dysfunction as a mechanistic biomarker in patients with non-alcoholic fatty liver disease (NAFLD). Mitochondrion. 2021;57:119–30.

    Article  CAS  PubMed  Google Scholar 

  63. Zhang L, Jarvis LB, Baek HJ, Gaston JS. Regulatory IL4 + CD8 + T cells in patients with ankylosing spondylitis and healthy controls. Ann Rheum Dis. 2009;68(8):1345–51.

    Article  CAS  PubMed  Google Scholar 

  64. Lejon K, Hellman U, Do L, Kumar A, Forsblad-d’Elia H. Increased proportions of inflammatory T cells and their correlations with cytokines and clinical parameters in patients with ankylosing spondylitis from northern Sweden. Scand J Immunol. 2022;96(3):e13190.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  65. Sorenson JR, Hangarter W. Treatment of rheumatoid and degenerative diseases with copper complexes: a review with emphasis on copper-salicylate. Inflammation. 1977;2(3):217–38.

    Article  CAS  PubMed  Google Scholar 

  66. Liu N, Xu Y, Li Q, Cao Y, Yang D, Liu S, Wang X, Mi Y, Liu Y, Ding C, et al. A lncRNA fine-tunes salicylic acid biosynthesis to balance plant immunity and growth. Cell Host Microbe. 2022;30(8):1124–e11381128.

    Article  CAS  PubMed  Google Scholar 

  67. Nyström N, Prast-Nielsen S, Correia M, Globisch D, Engstrand L, Schuppe-Koistinen I, Halfvarson J. Mucosal and plasma metabolomes in new-onset paediatric inflammatory bowel disease: correlations with Disease characteristics and plasma inflammation protein markers. J Crohns Colitis. 2023;17(3):418–32.

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

We would like to thank the UK Biobank consortia, GWAS Catalog, and IEU OpenGWAS project. We would like to thank the anonymous reviewers for their constructive comments.

Funding

None.

Author information

Author notes

  1. Jinpeng Wei and Jian Li contributed equally to this work and share first authorship.

Authors and Affiliations

  1. Department of Orthopedics, Third Hospital of Shanxi Medical University, Shanxi Bethune Hospital, Shanxi Academy of Medical Sciences, Tongji Shanxi Hospital, Taiyuan, 030032, China

    Jinpeng Wei, Jian Li, Tianyang Li, Shuhan Yang, Hua Wu & Haihu Hao

  2. Department of Orthopedics, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, 430030, China

    Tao Xu & Yingchi Zhang

Authors
  1. Jinpeng Wei
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Jian Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Tianyang Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Tao Xu
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Yingchi Zhang
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Shuhan Yang
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Hua Wu
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Haihu Hao
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

WJP and LJ designed the study and performed the statistical analysis; WJP, LJ, LTY, and XT participated in data interpretation; WJP, LTY, and ZYC wrote the first draft, and LJ, YSH, WH and HHH revised the article. WH and HHH had direct access and responsibility for verifying all data reported in the manuscript. All authors read and approved the submitted version of the manuscript.

Corresponding authors

Correspondence to Hua Wu or Haihu Hao.

Ethics declarations

Ethics approval and consent to participate

All summarized statistics utilized in the MR analyses were generated by previous studies, for which ethical approval and individual consent were obtained for all original studies.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Data sharing statement

Summary statistics for GWAS are publicly available for download. All scripts for the analysis are available from the authors upon request.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

13075_2024_3445_MOESM1_ESM.xlsx

Supplementary Material 1: Table 1 IVs of immune cells after removal of linkage disequilibrium and weak instrumental variables; Table 2 IVs of Metabolites after removal of linkage disequilibrium and weak instrumental variables; Table 3 Causal relationship between immune cells and RA; Supplementary Table 4 Causal relationship between immune cells and AS; Supplementary Table 5 Immune cells with IVW less than 0.05 in the causal outcome of immune cells and RA and AS; Table 6 Metabolite with IVW less than 0.05 in the causal outcome of immune cells and RA; Table 7 Metabolite with IVW less than 0.05 in the causal outcome of immune cells and AS; Table 8 Causal effect of immunophenotypes on metabolites; Table 9 MR-PRESSO test of immune cells and rheumatoid arthritis; Table 10 MR-PRESSO test of immune cells and ankylosing spondylitis; Table 11 MR analytical horizontal pleiotropy test of immune cells and ADs; Table 12 MR analytical horizontal pleiotropy test of metabolite and ADs; Table 13 MR analytical Cochran’s Q statistic of immune cells and ADs.

13075_2024_3445_MOESM2_ESM.pdf

Supplementary Material 2: Fig. 1 Scatter plots for the causal association between immune cells and AS (Supplementary figures were numbered using the last four digits of the GWAS ID of the immune cells, and this applies to the rest of the supplementary file);

Supplementary Material 3: Fig. 2 Scatter plots for the causal association between immune cells and RA;

Supplementary Material 4: Fig. 3 Funnel plots for the causal association between immune cells and AS;

Supplementary Material 5: Fig. 4 Funnel plots for the causal association between immune cells and RA;

Supplementary Material 6: Fig. 5 Leave-one-out plots for the causal association between immune cells and AS;

Supplementary Material 7: Fig. 6 Leave-one-out plots for the causal association between immune cells and RA.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Wei, J., Li, J., Li, T. et al. Genetically predicted metabolite mediates the causal relationship between immune cells and autoimmune diseases. Arthritis Res Ther 26, 207 (2024). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13075-024-03445-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13075-024-03445-z

Keywords

Arthritis Research & Therapy

ISSN: 1478-6362

Contact us