Proteogenomic Data Integration Reveals CXCL10 as a Potentially Downstream Causal Mediator for IL-6 Signaling on Atherosclerosis
Abstract
BACKGROUND:
Genetic and experimental studies support a causal involvement of IL-6 (interleukin-6) signaling in atheroprogression. Although trials targeting IL-6 signaling are underway, any benefits must be balanced against an impaired host immune response. Dissecting the mechanisms that mediate the effects of IL-6 signaling on atherosclerosis could offer insights about novel drug targets with more specific effects.
METHODS:
Leveraging data from 522 681 individuals, we constructed a genetic instrument of 26 variants in the gene encoding the IL-6R (IL-6 receptor) that proxied for pharmacological IL-6R inhibition. Using Mendelian randomization, we assessed its effects on 3281 plasma proteins quantified with an aptamer-based assay in the INTERVAL cohort (n=3301). Using mediation Mendelian randomization, we explored proteomic mediators of the effects of genetically proxied IL-6 signaling on coronary artery disease, large artery atherosclerotic stroke, and peripheral artery disease. For significant mediators, we tested associations of their circulating levels with incident cardiovascular events in a population-based study (n=1704) and explored the histological, transcriptomic, and cellular phenotypes correlated with their expression levels in samples from human atherosclerotic lesions.
RESULTS:
We found significant effects of genetically proxied IL-6 signaling on 70 circulating proteins involved in cytokine production/regulation and immune cell recruitment/differentiation, which correlated with the proteomic effects of pharmacological IL-6R inhibition in a clinical trial. Among the 70 significant proteins, genetically proxied circulating levels of CXCL10 (C-X-C motif chemokine ligand 10) were associated with risk of coronary artery disease, large artery atherosclerotic stroke, and peripheral artery disease, with up to 67% of the effects of genetically downregulated IL-6 signaling on these end points mediated by decreases in CXCL10. Higher midlife circulating CXCL10 levels were associated with a larger number of cardiovascular events over 20 years, whereas higher CXCL10 expression in human atherosclerotic lesions correlated with a larger lipid core and a transcriptomic profile reflecting immune cell infiltration, adaptive immune system activation, and cytokine signaling.
CONCLUSIONS:
Integrating multiomics data, we found a proteomic signature of IL-6 signaling activation and mediators of its effects on cardiovascular disease. Our analyses suggest the interferon-γ–inducible chemokine CXCL10 to be a potentially causal mediator for atherosclerosis in 3 vascular compartments and, as such, could serve as a promising drug target for atheroprotection.
Clinical Perspective
What Is New?
•
Integrating genomic and proteomic data, we found genetic downregulation of IL-6R (interleukin-6 receptor)–mediated signaling to be associated with changes in circulating levels of 70 proteins involved in cytokine production regulation and immune cell trafficking, which were consistent with the proteomic effects of the pharmacological inhibition of IL-6R.
•
Among the 70 proteins, genetically proxied levels of the chemokine CXCL10 (C-X-C motif chemokine ligand 10) were associated with risk of coronary artery disease, large-artery atherosclerotic stroke, and peripheral artery disease, and changes in CXCL10 levels mediated a significant proportion of the effects of IL-6 (interleukin-6) signaling on these end points.
•
Higher midlife circulating CXCL10 levels were associated with higher cardiovascular risk over 20 years.
What Are the Clinical Implications?
•
Our findings provide evidence for a downstream proteomic signature of genetic IL-6 signaling downregulation that could be used as potential biomarkers of response to drugs targeting the IL-6 signaling cascade.
•
CXCL10 might be a potentially causal mediator of the effects of IL-6 signaling on atherosclerosis and could serve as a promising drug target for atheroprotection that should be further explored in clinical trials.
Anti-inflammatory treatments have emerged as a promising approach for lowering the risk of atherosclerotic cardiovascular disease.1 The IL-6 (interleukin-6) pathway has attracted major attention because of converging evidence supporting its relevance in atherosclerosis.2 First, pharmacological inhibition of IL-6 or IL-6R (IL-6 receptor) leads to reductions in atherosclerotic lesion formation in mouse models of atherosclerosis.3,4 Second, prospective cohort studies have consistently found circulating levels of IL-6 to be associated with the manifestations of atherosclerotic disease, including coronary artery disease5 and ischemic stroke.6 Third, human genetic studies have shown that polymorphisms in the gene encoding the IL-6R resulting in downregulated IL-6 signaling activity are associated with a lower lifetime risk of coronary artery disease,7 atherosclerotic ischemic stroke,8 peripheral artery disease,9 and abdominal aortic aneurysm,10 as well as a favorable cardiometabolic profile.11
In line with these results, the CANTOS trial (Canakinumab Anti-Inflammatory Thrombosis Outcomes Study), which tested a monoclonal antibody targeting IL-1β (interleukin-1β), an upstream regulator of IL-6 signaling, showed considerable reductions in major adverse cardiovascular events among patients with a recent history of myocardial infarction.12 Post hoc analyses of CANTOS showed that canakinumab was particularly protective among patients achieving substantial decreases in IL-6 levels, thus providing indirect evidence that interfering with IL-6 signaling could reduce cardiovascular risk.13 On the basis of these findings, a recent phase 2 trial (RESCUE [Trial to Evaluate Reduction in Inflammation in Patients With Advanced Chronic Renal Disease Utilizing Antibody Mediated IL-6 Inhibition]) tested ziltivekimab, a monoclonal antibody that is directed against IL-6, and found that it effectively and safely reduces biomarkers of inflammation and thrombosis among patients with chronic kidney disease and evidence of inflammation (high-sensitivity C-reactive protein [CRP] levels ≥2 mg/L).14 The cardiovascular benefit of this approach remains unknown, as an ongoing large-scale phase 3 cardiovascular outcomes trial testing ziltivekimab will not be completed before 2025.15
However, any benefits from anti–IL-6 treatments would need to be balanced against an impaired host response. Canakinumab has been associated with a higher risk of fatal infections,12 and genetic analyses support that downregulation of IL-6 signaling is associated with a higher risk of bacterial infections.11 To this end, IL-6 is known as a key component of the host innate immune defense system.16 As such, although trials directly targeting IL-6 signaling are ongoing, it would be useful to identify alternative drug targets in the same signaling pathway that might mediate the atheroprotective effects of IL-6 inhibition. Such molecules might be more specific and therefore better anti-inflammatory treatment targets for atherosclerotic cardiovascular disease.
We previously leveraged genetic data from large-scale population-based studies and detected 26 variants in the gene encoding IL-6R that mimic the effects of pharmacologically inhibiting this protein.17,18 Here, we expand these analyses to blood proteomic data derived from the INTERVAL study (n=3301) to explore: (1) the causal circulating signature of interfering with IL-6 signaling that could inform the discovery of biomarker signatures of drug response to IL-6 inhibitors, and (2) potential mediators of the favorable cardiovascular effects of IL-6 inhibition that could serve as alternative and perhaps more selective drug targets. Following up on our findings, we compared our results with those of clinical trials applying pharmacological inhibition of IL-6 signaling. Triangulating the evidence, we explore in observational studies associations between circulating levels of potential protein drug targets and incident cardiovascular disease, as well as the histological, transcriptomic, and cellular profile of atherosclerotic lesions associated with these targets.
METHODS
Data Availability and Ethical Approval
Details for access to summary statistics used for the current analyses are provided in the Extended Methods of the Supplemental Material. All included studies received ethical approval from the respective institutional review boards, and all participants provided informed consent, as detailed in the Extended Methods of the Supplemental Material.
This study adheres to the MR-STROBE (Strengthening the Reporting of Observational Studies in Epidemiology using Mendelian Randomization) guidelines.19 The checklist is provided in the Extended Methods of the Supplemental Material.
Genetic Instrument Selection
To construct an instrument for IL-6R–mediated signaling,20 we used variants within the IL6R gene or a region 300 kB upstream or downstream of it. In accordance with an approach we described previously,17,18 we selected variants associated in a genome-wide association study meta-analysis of 522 681 individuals with circulating levels of CRP, a downstream biomarker of IL-6 signaling7 that was used as a functional readout for IL-6 signaling activity. The methodology and criteria used to obtain this instrument are outlined in the Extended Methods of the Supplemental Material and the genetic variants comprised in this study are listed in Table S1. As previously described, this genetic instrument is associated with circulating IL-6, CRP, and fibrinogen levels, consistent with results of clinical trials testing tocilizumab, a monoclonal antibody against IL-6R, which have been previously described.17
Proteome-Wide Mendelian Randomization Analyses
To detect the downstream proteomic effects of genetically proxied IL-6 signaling, we applied a proteome-wide 2-sample Mendelian randomization approach using plasma data from the INTERVAL study (n=3301 with SOMAscan aptamer-based proteomics; see Extended Methods of the Supplemental Material) Using these data, after excluding exposure instruments that were also used as instruments for IL-6 signaling, we applied 2-sample Mendelian randomization (MR) analysis, applying the inverse variance weighted (IVW) method as our primary approach. As a measure of pleiotropy, heterogeneity was quantified with the I2, and significance was tested with the Cochran Q statistic (P<0.05 was considered significant).21 To correct for multiple comparisons across the proteome, we corrected the P values derived from the IVW approach using the false discovery rate method.22 P values <0.05 were considered statistically significant. To validate our results against bias arising from horizontal pleiotropy, for significant proteins, we also applied the weighted median approach23 and MR-Egger regression,24 which are more robust against the use of pleiotropic instruments. Furthermore, for significant proteins, we applied a colocalization analysis to ensure that the same genetic signal influencing IL-6 signaling also influences protein levels, and that the detected association was not the result of pleiotropic effects of variants in linkage disequilibrium with the selected instruments.25 We applied a Bayesian framework for pairwise colocalization, within 300 kB of the IL6R gene, using the coloc package in R and tested competing models (H0: the genomic region does not include any variant influencing either the exposure or the outcome; H1: the genomic region includes a variant influencing only the exposure; H2: the genomic region includes a variant influencing only the outcome; H3: the genomic region includes separate variants influencing the exposure and the outcome; and H4: the genomic region includes a variant that influences both the exposure and the outcome).26 The previous probability of any random variant being associated with both traits was set at 10–5. A posterior probability of association of ≥0.8 for the last model (Η4) was defined as providing evidence of colocalization. The effects of the selected genetic variants on proteins were extracted from the PhenoScanner database through the phenoscanner function in R,27 and MR analyses were performed in R using the TwoSampleMR package.28 Pooling the significant proteins, we explored enrichment in biological pathways using the Gene Ontology resource.29
Comparisons of the Proteomic Effects of Genetically Proxied IL-6 Signaling Downregulation With the Pharmacological Effects of IL-6R Inhibition
To test the validity of our approach for detecting the downstream proteomic signature of a signaling pathway with proteome-wide MR, we aimed to correlate the effects derived from our MR approach with the effects derived from pharmacological IL-6R inhibition. We used data from a follow-up study of the Norwegian tocilizumab NSTEMI trial (non–ST-segment–elevation myocardial infarction), which tested IL-6R blockade with tocilizumab in patients with NSTEMI. In the original randomized control trial, patients with NSTEMI received a single intravenous dose of 280 mg of tocilizumab or placebo (n=117).30 In a follow-up analysis, the proteomic profile of serum samples was quantified at baseline and 4 days after treatment among a subgroup of 48 patients from the tocilizumab arm using the SOMAscan aptamer-based proteomics assay.31 After matching the proteins available in both data sets, we quantified the Pearson r correlation between the MR effects and the pharmacological effects across the proteome.
Associations With Clinical End Points in Mediation MR Analyses
After detecting proteins significantly associated with genetically proxied IL-6 signaling, we then explored associations between genetically proxied levels of these proteins and the 3 primary manifestations of atherosclerosis to identify which of these proteins could mediate the effects of IL-6 signaling on clinical end points. Clinical outcomes included coronary artery disease (CAD), peripheral artery disease (PAD), and large artery atherosclerotic stroke (LAAS). For CAD, we used data from 63 746 CAD cases and 130 681 controls of predominantly European ancestry (≈80%) from the CARDIoGRAMplusC4D Consortium (Coronary Artery Disease Genome wide Replication and Meta-analysis plus the Coronary Artery Disease [C4D] Genetics).32 For PAD, we used summary statistics from the Million Veteran Program cohort, a longitudinal cohort study containing electronic health records and genetic data across 50 veterans facilities in the United States. This cohort includes 31 307 cases and 211 753 controls of European, African, and Hispanic ancestry.33 For LAAS, we leveraged data from the MEGASTROKE consortium of primarily individuals of European ancestry, consisting of 6688 patients with stroke and 245 201 controls from 29 studies.34
Again, we applied 2-sample IVW MR analyses using genetic variants associated with the proteins that were significantly associated with genetically proxied IL-6 signaling in the previous step as instruments. To avoid overlap in the used data sets, for this step, we used data from the deCODE study (Diabetes Epidemiology: Collaborative analysis of Diagnostic criteria in Europe) in the Icelandic population.35 This study explored the genetic architecture of 4907 circulating proteins quantified with the aptamer-based SOMAScan approach among 35 559 Icelanders.35 For the proteins significantly associated with genetically proxied IL-6 signaling, we selected genetic variants throughout the genome associated with the levels of these proteins at a P<5×10–8 and clumped at an r2<0.01. For all proteins, beyond the IVW approach, we applied the MR-Egger and the weighted median methods. Furthermore, for significant proteins, we applied cis-MR analyses, selecting genetic variants associated with the proteins within or 300 kB upstream and downstream of the genes encoding these proteins at P<10–5 after clumping at r2<0.1. Significant results were screened on the basis of a false discovery rate–corrected P<0.05.
For proteins showing significant associations with both genetically proxied IL-6 signaling and risk of cardiovascular end points, we then applied 2-step mediation MR analyses36 to explore whether any of the effects of IL-6 signaling on the outcomes could be explained by changes in circulating protein levels. We first performed multivariable MR exploring associations of genetic proxies for significant proteins on the risk of the cardiovascular end points adjusting for the effects of the respective genetic instruments on sIL-6R (soluble IL-6R) concentration. Then, by multiplying effects of the genetic proxies for IL-6 signaling on protein levels with the multivariable MR association estimates between genetically proxied protein levels and cardiovascular outcomes, we obtained indirect effects of the genetic proxies for IL-6 signaling on the outcomes mediated through the tested protein.36 We divided these estimates by the total effects of the genetic proxies of IL-6 signaling on risk of CAD, PAD, and LAAS and obtained proportions of the effects mediated through the tested proteins.
Phenome-Wide Association Study
To explore potential adverse effects associated with promising proteins, we tested associations with the full range of clinical phenotypes encoded in the UK Biobank. We used the Phecode Map 1.2 to map UK Biobank International Classification of Diseases, 10th Revision (ICD-10) codes to Phecodes37 using all ICD-10 codes (main position, secondary position, and death records) from the UK Biobank. Individuals were assigned a case status if more than one ICD-10 code was mapped to the respective phecode. To approximate effect size in a logistic regression framework, we used minor allele carrier status as an independent variable and age at baseline, sex, and 5 ancestry principal components as covariates.
Replication in an Observational Population-Based Setting
CXCL10/interferon-γ–inducing protein 10 (IP-10) showed particularly promising associations as a potential mediator of genetically proxied IL-6 signaling on all tested atherosclerotic outcomes (as shown in Results). To replicate these associations in an observational setting, we used data from the population-based MONICA (Monitoring Trends and Determinants in Cardiovascular Disease)/KORA cohort (Cooperative Health Research in the Region Augsburg),38 a prospective population-based study of inhabitants of Augsburg, Germany, who were 25 to 74 years of age at baseline (1984–1995) and followed them up until 2016. Further information on the study design, outcome assessment, characteristics of the selected study participants, and the CXCL10 measurement methodology can be found in the Extended Methods of the Supplemental Material.
To explore associations between baseline CXCL10 levels with risk of future cardiovascular events (stroke or CAD), we applied a Cox proportional hazards model adjusting for age, sex, baseline survey (recruitment period: 1984–1985, 1989–1990, and 1994–1995; model 1), and additionally for vascular risk factors (body mass index [1 kg/m2 increment], smoking [current versus noncurrent], estimated glomerular filtration rate [1 mL·min–1·1.73 m–² increment], history of CAD, diabetes, total cholesterol, high-density lipoprotein cholesterol, and hypertension defined as blood pressure >140/90 mm Hg, or use of antihypertensive medications if participants were aware of having hypertension; model 2). Based on the assumption that CXCL10 is a downstream effector of IL-6 signaling on atherosclerosis, we explored correlations between baseline IL-6 and CXCL10 levels and, in sensitivity analyses, included IL-6, CXCL10, or both in our models to see how the effects of each cytokine on cardiovascular events changed after adjusting for the other.
Analyses in Human Atherosclerotic Samples
We used data from the Athero-EXPRESS study, a biobank of carotid endarterectomy samples in which the expression levels of CXCL10 have been quantified in transcriptomic analyses to assess associations with histological features of plaque vulnerability. For 700 samples, RNA was isolated and libraries were prepared for sequencing as previously described.39 We tested associations between the normalized expression values of CXCL10 and standard histological features of plaque vulnerability (large lipid core, intraplaque hemorrhage, extensive collagen content, and plaque calcification), as determined in sections from the plaque segment with the highest atherosclerosis burden. Methods for the evaluation of the histological images have been described previously in detail.40
To explore the changes in atherosclerotic plaque associated with higher expression of CXCL10, we used data from STARNET (Stockholm-Tartu Atherosclerosis Reverse Networks Engineering Task). We used transcriptomics data from atherosclerotic aortic root samples obtained with informed consent during coronary artery bypass grafting from 514 individuals with CAD.41 We explored genes coexpressed with CXCL10 by estimating Spearman correlations with the expression of 16 214 genes in this tissue. Genes significantly coexpressed with CXCL10 were moved forward to pathway analyses in Reactome.42 To explore the cell landscape of the transcriptomic profile associated with upregulated CXCL10 expression, we then used publicly available single-nuclei transcriptomic data from aortic tissue and tested which of the identified cell groups were enriched for these genes.43
RESULTS
Effects of Genetically Proxied IL-6 Signaling on Circulating Proteins
The overall study design is presented in Figure 1. The 26 CRP-lowering genetic variants within or close to the gene encoding IL-6R that were used as proxies for IL-6 signaling downregulation are presented in Table S1. First, we explored associations between genetically proxied IL-6R–mediated signaling as captured by these variants, with proteomic changes across 3281 plasma proteins quantified with the aptamer-based SOMAScan assay among 3301 participants of the INTERVAL study, as illustrated in Figure 2A. A detailed list of the associations as derived from IVW MR analyses is provided in Table S2. After correction for multiple comparisons (false discovery rate–adjusted P value<0.05), we found that genetically proxied downregulation of the IL-6 signaling pathway is associated with lower levels of 43 and higher levels of 27 circulating proteins, explaining between 1% and 15% of their variance (Figure 2B; Table S3). Of those proteins, 54 also showed evidence of colocalization, as determined by a posterior probability of association >0.80 for at least one common genetic variant in the examined genomic region influencing both sIL-6R and the respective protein levels. It should be noted that a low probability of association for the hypothesis of no common variant (probability of association 3) for specific proteins might indicate lack of power of colocalization to detect a common signal rather than lack of a common genetic variant. Among these proteins, 45 also showed significant and directionally consistent associations in the weighted median approach and no evidence of directional pleiotropy in MR-Egger regression (Pintercept>0.05). Of those, only one showed minimal evidence of heterogeneity (MYO6, I2=27%; Table S3).
Among the 70 proteins, there were well-known components of the pathway, such as IL-6 itself or downstream mediators and effectors, such as CRP and the fibrinogen C-domain–containing protein-1 (Figure 2B; Table S3). A Gene Ontology pathway enrichment analysis with the significant proteins (Figure 2C; Table S4) showed enrichment in several cytokine production pathways, such as regulation of IL-12 and IL-1β production and regulation of the ΙFN-γ pathway, as well as in immune cell-trafficking pathways, such as myeloid cell differentiation and mononuclear cell or leukocyte proliferation. Furthermore, there was enrichment in tyrosine phosphorylation of STAT in line with the known signaling transduction pathway of IL-6R activation.44
Correlation Between MR-Derived Proteomic and Pharmacological Effects
We compared the proteomic effects derived from a genetically proxied downregulation of IL-6 signaling in MR analyses with the effects of pharmacological IL-6 signaling inhibition on the same proteins. Leveraging data from 48 patients with NSTEMI from the Norwegian tocilizumab NSTEMI study (24 who underwent treatment with tocilizumab and 24 in the control arm),31 we obtained differences between tocilizumab and placebo across 1074 proteins also assessed with a different version of the aptamer-based SOMAScan assay within 48 hours after randomization and treatment. Matching the available proteins with the proteins available in the INTERVAL study, we were able to explore correlations across 785 proteins between the MR and the trial effects (Table S5). We found a significant correlation across the proteome (Pearson r 0.16, P=1.1×10–5; Figure 2D). This correlation was even stronger when restricting the analysis to the 44 proteins reaching a P<0.05 in the trial analysis (r=0.48, P=8×10–4).
Effects of Significant Proteins on Cardiovascular End Points and Mediation Analysis
Given the well-established associations of genetically downregulated IL-6 signaling with CAD, PAD, and ischemic stroke,8,9,11 we next explored to what extent the 70 proteins associated with genetically proxied IL-6 signaling are also associated with these cardiovascular end points. We proxied the levels of 58 of these proteins in the deCODE proteogenomic database from 35 559 Icelanders, in which 4907 circulating proteins were also quantified with the aptamer-based SOMAScan assay.35 Across these 58 proteins, we found the genetically proxied circulating levels of 4 to be associated with the examined cardiovascular outcomes in IVW MR analyses after correction for multiple comparisons (false discovery rate–corrected P<0.05). Genetically proxied levels of CXCL10 were associated with all 3 atherosclerosis end points, whereas CBS (cystathionine β-synthase), CRP, and IL-23R (interleukin-23 receptor) showed significant associations only with CAD (Figure 3A; Table S6). The results for CXCL10 and IL-23R were consistent in sensitivity analyses, including the weighted median approach and MR-Egger regression.
To minimize the risk of pleiotropic effects, we next performed a sensitivity analysis based solely on cis-acting variants for these proteins. Using genetic variants in minimal linkage disequilibrium (r2<0.1) associated with the circulating levels of these proteins at P<10–5 and within 300 kB upstream or downstream of the encoding genes, we confirmed the association between genetically proxied CXCL10 levels and odds for PAD (Table S7). Figure 3B illustrates the colocalization of the genomic signals between circulating sIL-6R and CXCL10 around the same genetic variants at the IL6R locus. The posterior probability H4 ranged from 99.97% for a prior P12 of 10–3 to 19% for a prior P12 of 10–8. Given the emergence of CXCL10 as the most promising mediator in the examined associations, we performed mediation MR analyses. There was no overlap in genetic instruments for IL-6R signaling and circulating CXCL10 levels (Tables S1, S8, and S9). We found changes in circulating CXCL10 levels to mediate 39%, 67%, and 25% of the effects of genetically proxied IL-6 signaling on PAD, LAAS, and CAD, respectively (all P<0.05; Figure 3C).
There was no evidence of a reverse effect of genetically proxied CXCL10 on IL-6 levels (β: 0.10 [95% CI, –0.04 to 0.24], P=0.15). Furthermore, we performed a sensitivity analysis in a population-based study of 876 participants from the Lothian Birth Cohort 1936, where CXCL10 was quantified with the Olink technology.45 The IVW MR analyses showed an association similar in direction and magnitude to INTERVAL (βOlink: –0.05 [95%CI, –0.11 to 0.01] versus βSOMAScan: –0.06 [95% CI, –0.09 to –0.03]). An analysis of the larger UK Biobank data set, in which CXCL10 was also quantified with Olink, showed no association (β: –0.01 [95% CI, –0.02 to 0.01], P=0.55). In a PheWAS analysis, we found no evidence of significant associations between genetically proxied CXCL10 and infectious diseases (Table S10).
Circulating CXCL10 Levels and Risk of Cardiovascular Events in a Population-Based Prospective Cohort Study
Following up on the signal for CXCL10, we used data from the population-based MONICA/KORA study,38 in which circulating levels of CXCL10 have been quantified with a Luminex cytokine multiplex assay at midlife in 1704 individuals (47% women; median age, 53 years [interquartile range: 44–61]) without overt cardiovascular disease at baseline. The baseline characteristics of the study participants are presented in Table S11. There was a significant positive correlation between serum IL-6 and CXCL10 levels among study participants (r=0.19, P=1.2×10–14; Figure 4A). Next, we tested associations of circulating CXCL10 levels with the risk of a composite end point of stroke and CAD over a median follow-up of 20.9 years (interquartile range: 12.9–26.1). After adjustments for age, sex, and baseline survey (model 1) in a Cox regression model, we found circulating CXCL10 levels to be associated with a higher risk of the composite end point (hazard ratio per SD increment in log-transformed CXCL10 levels: 1.20 [95% CI, 1.05–1.36]; Figure 4B and 4C). These results remained significant in a model further adjusting for common vascular risk factors at baseline (model 2), and when including both circulating CXCL10 and IL-6 levels in the same model (model 3; Figure 4C; Table S12).
CXCL10 Expression in Atherosclerotic Tissue Associated With Plaque Vulnerability
To explore whether CXCL10 expression in human atherosclerotic lesions is associated with plaque vulnerability, we used data from 623 individuals from the Athero-EXPRESS Biobank in Utrecht, the Netherlands, who underwent carotid endarterectomy because of symptomatic or asymptomatic stenosis and for whom both transcriptomic and histological analyses were available (Table S13; Figure 5A). After adjustments for age and sex, we found that higher normalized CXCL10 expression in carotid atherosclerotic plaques was associated with a larger lipid core (odds ratio, 1.19 [95% CI, 1.02–1.40], P=0.01; Figure 5B). No associations were detected with intraplaque hemorrhage, extensive collagen content, or plaque calcification (Figure 5B).
To test in more depth the phenotypic plaque profile associated with elevated CXCL10 expression, we used transcriptomic data from atherosclerotic aortic root tissue from 514 samples from patients with CAD in the STARNET network and explored the genes that were correlated with the expression of CXCL10 in a coexpression analysis. A total of 98 genes were positively correlated with CXCL10 expression and 6 genes were negatively correlated with CXCL10 expression (Figure 5C; Table S14). The positively correlated genes were enriched in immune system pathways, as captured by Reactome, and included CXCL9 and CXCL11, which are also ligands of the CXCR3 (C-X-C motif chemokine receptor 3), the main CXCL10 receptor.46 These pathways included cytokine signaling pathways, and, in particular, the interferon and interleukin pathways, such as interferon-γ, IL-6, and IL-1 signaling, as well as pathways related to the adaptive immune system (Figure 5D; Table S15). As a final step, we explored whether specific cell types, as captured by single-nuclei RNA sequencing from aortic tissue,43 are enriched for expressing the top genes positively coexpressed with CXCL10 (r>0.3) to infer the cellular landscape associated with higher CXCL10 expression. According to these data, endothelial cells, lymphocytes, and macrophages were the cell types primarily expressing these genes (PARP14 and STAT1 most strongly enriched in endothelial cells, NLRC5 and PARP14 most strongly enriched in lymphocytes, and PARP14 and TYMP most strongly enriched in macrophages [Figure 5E; Table S16]).
DISCUSSION
Integrating data from large-scale genomic and proteomic studies, we explored the downstream effects of genetic downregulation of the IL-6R–mediated signaling on the blood proteome and their mediating role in lowering the risk of atherosclerotic cardiovascular disease. We found changes in circulating levels of 70 proteins involved in regulation of cytokine production and immune cell trafficking, which were consistent with the proteomic effects of the pharmacological inhibition of IL-6R with tocilizumab in a clinical trial.
Across the 70 proteins affected by IL-6 signaling, genetically proxied levels of the chemokine CXCL10 were associated with risk of CAD, LAAS, and PAD, and changes in CXCL10 levels mediated a significant proportion of the effects of IL-6 signaling on these end points. In follow-up analyses, we found circulating CXCL10 levels at midlife in a population-based setting to be associated with the long-term risk of CAD or stroke. Higher expression of CXCL10 in human atherosclerotic plaques was correlated with a larger lipid core and a transcriptomic profile consistent with macrophage and lymphocyte infiltration, as well as activation of the adaptive immune system and cytokine signaling.
Our results linking CXCL10 levels with risk of atherosclerotic disease are consistent with experimental data showing a key role for this chemokine in atherogenesis and atheroprogression. The administration of a monoclonal antibody against CXCL10 produced a histologically more stable atherosclerotic plaque phenotype with increased collagen and smooth muscle cell content and a smaller necrotic lipid core in the common carotid arteries of atheroprone Apoe–/– mice fed a high-fat diet.47 This is in line with our results suggesting an association of higher intraplaque CXCL10 expression with a larger lipid core, a hallmark of unstable plaques. In another model, both sexes of Cxcl10–/– Apoe–/– mice exhibited significantly smaller atherosclerotic lesion areas across the aortic arch, thoracic and abdominal aorta, and different durations of treatment.48 The atheroprotective effect of CXCL10 blockade in translational experiments is further corroborated by evidence of decreased atherosclerotic burden in the aortic arches and carotid arteries of Ldlr–/– mice after pharmacological targeting of CXCR3, the receptor of CXCL10, with a specific small molecule antagonist,49 and a CXCR3/CCR5 dual inhibitor.50 Furthermore, experimental data also suggest that IL-6, and particularly IL-6 trans-signaling through STAT3, exerts an effect on T-cell immigration directly through regulation chemokine secretion, including CXCL10.51 CXCL10/CXCR3 interaction induces TH1 cell differentiation and migration into the atherosclerotic lesion where they release TNF-α (tumor necrosis factor-α) and interferon-γ, increasing the recruitment and activation of macrophages and contributing toward a proinflammatory lesion environment.52 Within the plague the T cells directly interact with foam cells through the binding of the CD40 to the CD40 receptor and induce foam cells to produce inflammatory molecules and uptake more lipids, exacerbating the formation of the lipid core.53 Current evidence collectively indicates that CXCL10 binding to CXCR3 drives atherosclerotic plaque progression through differentiation and homing of proinflammatory T-lymphocyte subsets; 2 isoforms (CXCR3-A and CXCR3-B) of this receptor have been identified more specifically in humans, which lead to different downstream effects. Through binding to the CXCR3-A receptor, CXCL10 has chemoattractant properties and induces the migration, proliferation, and survival of leukocytes. On the other hand, the CXCR3B isoform of the receptor has antiangiogenic properties and inhibits cell proliferation and migration while also stimulating apoptosis.46 It should be noted that CXCR3 also binds the chemokines CXCL9 and CXCL11.46
Beyond detecting mediators of the effects of IL-6 signaling on atherosclerosis, our findings also provide evidence for a downstream proteomic signature of genetic IL-6 signaling downregulation that could be correlated with trial data from pharmacological IL-6R inhibition. This finding has important implications for a more widespread use of a proteome-wide MR analysis in the context of detecting the downstream proteomic consequences of perturbing drug targets. Proteomic signatures detected with this approach could be used as potential biomarkers of response to drugs targeting the upstream targets. Although such response biomarkers are well-established for drugs targeting the IL-6 signaling cascade, our approach supports use of the method for identifying response biomarkers for drugs under earlier development. Despite the significant correlation with trial data, there were also discrepancies between genetic downregulation and pharmacological inhibition of IL-6 signaling, including a lack of effect of tocilizumab on CXCL10 levels in the NSTEMI trial. There are several reasons that might explain this discrepancy. First, genetic downregulation reflects a lifelong effect, whereas the trial captured the effects of a single dose of tocilizumab 4 days after treatment. Second, the effects of genetic downregulation were studied in a general population setting, whereas the trial tested tocilizumab in a cohort of patients with NSTEMI shortly after the event, when an inflammatory response to the infarct might influence cytokine levels. Third, tocilizumab blocks both the classical and trans-IL-6 signaling.54 Whether this is the case for most of the variants used as proxies for IL-6R signaling (selected based on their effects on CRP, primarily reflecting classical signaling) remains untested. Animal studies have shown decreases in CXCL10 levels and T-cell recruitment after targeted pharmacological inhibition of the trans-signaling cascade.51 Alternative or even competing downstream effects of the classical and trans-signaling inhibition with tocilizumab versus genetic downregulation might explain part of the discrepancies. Fourth, the uncertainty in the effects of tocilizumab because of the small trial sample size (20 treated versus 20 placebo arm) should also be considered. Future studies should systematically explore how proteomic alterations detected in clinical trials correlate with those detected in MR analyses.
Among the proteins affected by genetic IL-6 signaling downregulation is the circulating soluble form of IL-6R itself (sIL-6R). The IL-6R variant with the largest effect size that largely drives the reported associations is rs2228145, which leads to the substitution of aspartic acid with alanine in the extracellular domain of IL-6R. This substitution is located within the cleavage site of the ADAM17 protease, which is responsible for ectodomain shedding of the membrane-bound IL-6R, resulting in its conversion from the membrane-bound to the soluble form of the receptor (sIL-6R).55 The result of this substitution is an increased proteolytic cleavage of the membrane-bound IL-6R and a profound increase in circulating sIL-6R and IL-6. Although the levels of the sIL-6R increase, membrane-bound IL-6R expression (especially in CD4+ T cells and monocytes) decreases, which impairs IL-6 signaling.55,56
Using MR to integrate multiomics data is a method increasingly used for bridging data from different omics layers and detecting causal relationships between them.57,58 Scaling up mediation analyses to be integrated to this approach allows explorations of associations with clinically relevant outcomes, and as such the discovery of novel therapeutic targets. Furthermore, this approach enables the dissection of pathways leading from genomic alterations to cardiovascular disease, thus providing novel insights into pathophysiology. Expanding the focus of the approach to other omics layers beyond proteomics, to more drug targets and more outcomes could provide novel insights in future explorations.
Our study has limitations. First, despite leveraging the largest available samples, such integrating approaches are characterized by decreasing power when moving downstream across omics layers.59 As such, several of our analyses, especially for the outcome of LAAS, might have been relatively underpowered, highlighting the need to repeat the analyses with increasing sample sizes. Second, although the proteomic analyses across all studies were done with the SOMAscan aptamer-based assay, there were considerable differences in proteins available in the INTERVAL study, deCODE study, and the Norwegian tocilizumab NSTEMI trial, which led to an incomplete overlap between the studies. Although the associations between genetically proxied IL-6 signaling and circulating CXCL10 were directionally consistent in 2 data sets performing protein quantifications with the proximity extension assay-based Olink assay, they did not reach statistical significance and should thus be interpreted with caution. Considerable differences have been previously reported for levels of the same proteins when quantified with Somascan or Olink.60 Third, directional pleiotropy might confound the analyses exploring associations between genetically proxied circulating proteins and cardiovascular end points, especially instruments based on variants from throughout the genome may be particularly prone to directional pleiotropy and, as such, might lead to false-positive findings. Although analyses based on cis-variants are less prone to this kind of bias, such variants are rare and not directly available for proxying all proteins. Furthermore, selecting such variants might lead to weak instrument bias because of more lenient thresholds and, as a result, estimates biased to the null that might lead to false-negative findings. It is important to note that when restricting our analysis to cis-instruments we were unable to find significant associations of genetically proxied CXCL10 with CAD or LAAS and thus those results should be interpreted with caution. Fourth, the majority of individuals explored in current analyses were of European ancestry and as such our results may not be generalizable to individuals of other ancestries. Fifth, data from deCODE were derived from an Icelandic population that has substantial differences from European populations. Although integration of data from deCODE with data from European data sets has been extensively used in the past, providing important insights into disease pathogenesis, we cannot exclude the introduction of population stratification that could bias our MR analyses in the second step (Figure 2) of our approach. Sixth, an additional caveat to our study is that we used proxies of IL-6R inhibition as genetic instruments, meaning that proteins found upstream would be upregulated as a result of positive feedback, whereas downstream molecules would be downregulated. This could create confusion as to the directionality of the levels of these proteins in real-life clinical trials. Seventh, to proxy IL-6 signaling we used CRP levels, which increase in response to the activation of the IL-6 cascade. However, IL-6 signaling is a complex cascade with a classical component, exerted through the membrane-bound IL-6R, and a trans-signaling component exerted through the soluble form of IL-6R. The 2 subpathways exert unique actions,16,61 but disentangling them with genetic instruments probably goes beyond the limits of MR. Eighth, by design, MR analyses assess the effects of lifetime downregulated IL-6 signaling, which might differ from a shorter pharmacological inhibition with IL-6R blockade. Ninth, the estimates of the direct and indirect effects of genetically proxied IL-6 signaling on atherosclerotic outcomes in the mediation analyses might be biased because of noncollapsibility of odds ratios when using a binary outcome. Tenth, colocalization analyses are based on an assumption of a single causal variant per trait in the examined regions, which is probably not realistic in most scenarios.
In conclusion, integrating genomic and proteomic data, we found a proteomic signature of IL-6 signaling activation and mediators of its effect on cardiovascular disease. Our analyses suggest that CXCL10 is a potentially causal mediator for atherosclerotic end points in 3 different vascular beds and, as such, might serve as a promising drug target for atheroprotection that should be further explored in clinical trials.
ARTICLE INFORMATION
Supplemental Material
Expanded Methods
Tables S1–S16
References 62–66
Acknowledgments
The authors thank all participants for their long-term commitment to the KORA study, the staff for data collection and research data management, and the members of the KORA Study Group (https://www.helmholtz-munich.de/en/epi/cohort/kora), who are responsible for the design and conduct of the study.
Footnote
Nonstandard Abbreviations and Acronyms
- CAD
- coronary artery disease
- CANTOS
- Canakinumab Anti-Inflammatory Thrombosis Outcomes Study
- CBS
- Cystathionine Beta-Synthase
- CRP
- C-reactive protein
- CXCL10
- C-X-C motif chemokine ligand 10
- CXCR3
- C-X-C motif chemokine receptor 3
- IL-1
- interleukin-1
- IL-6
- interleukin-6
- IL-6R
- interleukin-6 receptor alpha chain
- IP-10
- interferon-γ-induced protein 10
- IVW
- inverse variance weighted
- LAAS
- large artery atherosclerotic stroke
- MR
- Mendelian randomization
- NSTEMI
- non–ST-segment–elevation myocardial infarction
- PAD
- peripheral artery disease
- sIL-6R
- soluble IL-6R
- STARNET
- Stockholm-Tartu atherosclerosis reverse networks engineering task
- TNF-α
- tumor necrosis factor-α
Supplemental Material
REFERENCES
1.
Soehnlein O, Libby P. Targeting inflammation in atherosclerosis: from experimental insights to the clinic. Nat Rev Drug Discov. 2021;20:589–610. doi: 10.1038/s41573-021-00198-1
2.
Ridker PM, Rane M. Interleukin-6 signaling and anti-interleukin-6 therapeutics in cardiovascular disease. Circ Res. 2021;128:1728–1746. doi: 10.1161/CIRCRESAHA.121.319077
3.
Schuett H, Oestreich R, Waetzig GH, Annema W, Luchtefeld M, Hillmer A, Bavendiek U, von Felden J, Divchev D, Kempf T, et al. Transsignaling of interleukin-6 crucially contributes to atherosclerosis in mice. Arterioscler Thromb Vasc Biol. 2012;32:281–290. doi: 10.1161/ATVBAHA.111.229435
4.
Huber SA, Sakkinen P, Conze D, Hardin N, Tracy R. Interleukin-6 exacerbates early atherosclerosis in mice. Arterioscler Thromb Vasc Biol. 1999;19:2364–2367. doi: 10.1161/01.atv.19.10.2364
5.
Kaptoge S, Seshasai SR, Gao P, Freitag DF, Butterworth AS, Borglykke A, Di Angelantonio E, Gudnason V, Rumley A, Lowe GD, et al. Inflammatory cytokines and risk of coronary heart disease: new prospective study and updated meta-analysis. Eur Heart J. 2014;35:578–589. doi: 10.1093/eurheartj/eht367
6.
Papadopoulos A, Palaiopanos K, Björkbacka H, Peters A, de Lemos JA, Seshadri S, Dichgans M, Georgakis MK. Circulating interleukin-6 levels and incident ischemic stroke: a systematic review and meta-analysis of population-based cohort studies. medRxiv. 2021;2021.2003.2027.21254451. doi: 10.1101/2021.03.27.21254451
7.
Swerdlow DI, Holmes MV, Kuchenbaecker KB, Engmann JE, Shah T, Sofat R, Guo Y, Chung C, Peasey A, Pfister R, et al; Interleukin-6 Receptor Mendelian Randomisation Analysis (IL6R MR) Consortium. The interleukin-6 receptor as a target for prevention of coronary heart disease: a Mendelian randomisation analysis. Lancet. 2012;379:1214–1224. doi: 10.1016/S0140-6736(12)60110-X
8.
Georgakis MK, Malik R, Gill D, Franceschini N, Sudlow CLM, Dichgans M; INVENT Consortium, CHARGE Inflammation Working Group. Interleukin-6 signaling effects on ischemic stroke and other cardiovascular outcomes: a Mendelian randomization study. Circ Genom Precis Med. 2020;13:e002872. doi: 10.1161/CIRCGEN.119.002872
9.
Levin MG, Klarin D, Georgakis MK, Lynch J, Liao KP, Voight BF, O’Donnell CJ, Chang KM, Assimes TL, Tsao PS, et al. A missense variant in the IL-6 receptor and protection from peripheral artery disease. Circ Res. 2021;129:968–970. doi: 10.1161/CIRCRESAHA.121.319589
10.
Cai T, Zhang Y, Ho YL, Link N, Sun J, Huang J, Cai TA, Damrauer S, Ahuja Y, Honerlaw J, et al; VA Million Veteran Program. Association of interleukin 6 receptor variant with cardiovascular disease effects of interleukin 6 receptor blocking therapy: a phenome-wide association study. JAMA Cardiol. 2018;3:849–857. doi: 10.1001/jamacardio.2018.2287
11.
Georgakis MK, Malik R, Li X, Gill D, Levin MG, Vy HMT, Judy R, Ritchie M, Verma SS, Nadkarni GN, et al; Regeneron Genetics Center. Genetically downregulated interleukin-6 signaling is associated with a favorable cardiometabolic profile: a phenome-wide association study. Circulation. 2021;143:1177–1180. doi: 10.1161/CIRCULATIONAHA.120.052604
12.
Ridker PM, Everett BM, Thuren T, MacFadyen JG, Chang WH, Ballantyne C, Fonseca F, Nicolau J, Koenig W, Anker SD, et al; CANTOS Trial Group. Antiinflammatory therapy with canakinumab for atherosclerotic disease. N Engl J Med. 2017;377:1119–1131. doi: 10.1056/NEJMoa1707914
13.
Ridker PM, Libby P, MacFadyen JG, Thuren T, Ballantyne C, Fonseca F, Koenig W, Shimokawa H, Everett BM, Glynn RJ. Modulation of the interleukin-6 signalling pathway and incidence rates of atherosclerotic events and all-cause mortality: analyses from the Canakinumab Anti-Inflammatory Thrombosis Outcomes Study (CANTOS). Eur Heart J. 2018;39:3499–3507. doi: 10.1093/eurheartj/ehy310
14.
Ridker PM, Devalaraja M, Baeres FMM, Engelmann MDM, Hovingh GK, Ivkovic M, Lo L, Kling D, Pergola P, Raj D, et al; RESCUE Investigators. IL-6 inhibition with ziltivekimab in patients at high atherosclerotic risk (RESCUE): a double-blind, randomised, placebo-controlled, phase 2 trial. Lancet. 2021;397:2060–2069. doi: 10.1016/S0140-6736(21)00520-1
15.
Novo Nordisk A/S. ZEUS: A research study to look at how ziltivekimab works compared to placebo in people with cardiovascular disease, chronic kidney disease and inflammation (ZEUS). 2021. NCT05021835
16.
Rose-John S, Winthrop K, Calabrese L. The role of IL-6 in host defence against infections: immunobiology and clinical implications. Nat Rev Rheumatol. 2017;13:399–409. doi: 10.1038/nrrheum.2017.83
17.
Georgakis MK, Malik R, Burgess S, Dichgans M. Additive effects of genetic interleukin-6 signaling downregulation and low-density lipoprotein cholesterol lowering on cardiovascular disease: a 2×2 factorial Mendelian randomization analysis. J Am Heart Assoc. 2021;11:e023277. doi: 10.1161/JAHA.121.023277
18.
Georgakis MK, Malik R, Richardson TG, Howson JMM, Anderson CD, Burgess S, Hovingh GK, Dichgans M, Gill D. Associations of genetically predicted IL-6 signaling with cardiovascular disease risk across population subgroups. BMC Med. 2022;20:245. doi: 10.1186/s12916-022-02446-6
19.
Skrivankova VW, Richmond RC, Woolf BAR, Davies NM, Swanson SA, VanderWeele TJ, Timpson NJ, Higgins JPT, Dimou N, Langenberg C, et al. Strengthening the reporting of observational studies in epidemiology using Mendelian randomisation (STROBE-MR): explanation and elaboration. BMJ. 2021;375:n2233. doi: 10.1136/bmj.n2233
20.
Georgakis MK, Malik R, Gill D, Franceschini N, Sudlow CLM, Dichgans M. Interleukin-6 signaling effects on ischemic stroke and other cardiovascular outcomes: a Mendelian randomization study. Circ Genom Precis Med. 2020;13:e002872. doi: 10.1161/CIRCGEN.119.002872
21.
Bowden J, Hemani G, Davey Smith G. Invited commentary: detecting individual and global horizontal pleiotropy in Mendelian randomization: a job for the humble heterogeneity statistic? Am J Epidemiol. 2018;187:2681–2685. doi: 10.1093/aje/kwy185
22.
Aggarwal S, Yadav AK. False discovery rate estimation in proteomics. Methods Mol Biol. 2016;1362:119–128. doi: 10.1007/978-1-4939-3106-4_7
23.
Hartwig FP, Davey Smith G, Bowden J. Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption. Int J Epidemiol. 2017;46:1985–1998. doi: 10.1093/ije/dyx102
24.
Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44:512–525. doi: 10.1093/ije/dyv080
25.
Zuber V, Grinberg NF, Gill D, Manipur I, Slob EAW, Patel A, Wallace C, Burgess S. Combining evidence from Mendelian randomization and colocalization: review and comparison of approaches. Am J Hum Genet. 2022;109:767–782. doi: 10.1016/j.ajhg.2022.04.001
26.
Giambartolomei C, Vukcevic D, Schadt EE, Franke L, Hingorani AD, Wallace C, Plagnol V. Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. PLoS Genet. 2014;10:e1004383. doi: 10.1371/journal.pgen.1004383
27.
Kamat MA, Blackshaw JA, Young R, Surendran P, Burgess S, Danesh J, Butterworth AS, Staley JR. PhenoScanner V2: an expanded tool for searching human genotype-phenotype associations. Bioinformatics. 2019;35:4851–4853. doi: 10.1093/bioinformatics/btz469
28.
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:e34408. doi: 10.7554/eLife.34408
29.
Gene Ontology Consortium. The Gene Ontology resource: enriching a GOld mine. Nucleic Acids Res. 2021;49:D325–D334. doi: 10.1093/nar/gkaa1113
30.
Kleveland O, Kunszt G, Bratlie M, Ueland T, Broch K, Holte E, Michelsen AE, Bendz B, Amundsen BH, Espevik T, et al. Effect of a single dose of the interleukin-6 receptor antagonist tocilizumab on inflammation and troponin T release in patients with non-ST-elevation myocardial infarction: a double-blind, randomized, placebo-controlled phase 2 trial. Eur Heart J. 2016;37:2406–2413. doi: 10.1093/eurheartj/ehw171
31.
George MJ, Kleveland O, Garcia-Hernandez J, Palmen J, Lovering R, Wiseth R, Aukrust P, Engmann J, Damas JK, Hingorani AD, et al. Novel insights into the effects of interleukin 6 antagonism in non-ST-segment-elevation myocardial infarction employing the SOMAscan proteomics platform. J Am Heart Assoc. 2020;9:e015628. doi: 10.1161/JAHA.119.015628
32.
Nikpay M, Goel A, Won HH, Hall LM, Willenborg C, Kanoni S, Saleheen D, Kyriakou T, Nelson CP, Hopewell JC, et al. A comprehensive 1,000 Genomes-based genome-wide association meta-analysis of coronary artery disease. Nat Genet. 2015;47:1121–1130. doi: 10.1038/ng.3396
33.
Klarin D, Lynch J, Aragam K, Chaffin M, Assimes TL, Huang J, Lee KM, Shao Q, Huffman JE, Natarajan P, et al; VA Million Veteran Program. Genome-wide association study of peripheral artery disease in the Million Veteran Program. Nat Med. 2019;25:1274–1279. doi: 10.1038/s41591-019-0492-5
34.
Malik R, Chauhan G, Traylor M, Sargurupremraj M, Okada Y, Mishra A, Rutten-Jacobs L, Giese AK, van der Laan SW, Gretarsdottir S, et al. Multiancestry genome-wide association study of 520,000 subjects identifies 32 loci associated with stroke and stroke subtypes. Nat Genet. 2018;50:524–537. doi: 10.1038/s41588-018-0058-3
35.
Ferkingstad E, Sulem P, Atlason BA, Sveinbjornsson G, Magnusson MI, Styrmisdottir EL, Gunnarsdottir K, Helgason A, Oddsson A, Halldorsson BV, et al. Large-scale integration of the plasma proteome with genetics and disease. Nat Genet. 2021;53:1712–1721. doi: 10.1038/s41588-021-00978-w
36.
Carter AR, Sanderson E, Hammerton G, Richmond RC, Davey Smith G, Heron J, Taylor AE, Davies NM, Howe LD. Mendelian randomisation for mediation analysis: current methods and challenges for implementation. Eur J Epidemiol. 2021;36:465–478. doi: 10.1007/s10654-021-00757-1
37.
Wu P, Gifford A, Meng X, Li X, Campbell H, Varley T, Zhao J, Carroll R, Bastarache L, Denny JC, et al. Mapping ICD-10 and ICD-10-CM codes to phecodes: workflow development and initial evaluation. JMIR Med Inform. 2019;7:e14325. doi: 10.2196/14325
38.
Herder C, Baumert J, Thorand B, Martin S, Lowel H, Kolb H, Koenig W. Chemokines and incident coronary heart disease: results from the MONICA/KORA Augsburg case-cohort study, 1984–2002. Arterioscler Thromb Vasc Biol. 2006;26:2147–2152. doi: 10.1161/01.ATV.0000235691.84430.86
39.
Mokry M, Boltjes A, Cui K, Slenders L, Mekke JM, Depuydt MAC, Timmerman N, Waissi F, Verwer MC, Turner AW, et al. Transcriptomic-based clustering of advanced atherosclerotic plaques identifies subgroups of plaques with differential underlying biology that associate with clinical presentation. medRxiv. 2021;2021.2011.2025.21266855. doi: 10.1101/2021.11.25.21266855
40.
Georgakis MK, van der Laan SW, Asare Y, Mekke JM, Haitjema S, Schoneveld AH, de Jager SCA, Nurmohamed NS, Kroon J, Stroes ESG, et al. Monocyte-chemoattractant protein-1 levels in human atherosclerotic lesions associate with plaque vulnerability. Arterioscler Thromb Vasc Biol. 2021;41:2038–2048. doi: 10.1161/ATVBAHA.121.316091
41.
Franzen O, Ermel R, Cohain A, Akers NK, Di Narzo A, Talukdar HA, Foroughi-Asl H, Giambartolomei C, Fullard JF, Sukhavasi K, et al. Cardiometabolic risk loci share downstream cis- and trans-gene regulation across tissues and diseases. Science. 2016;353:827–830. doi: 10.1126/science.aad6970
42.
Fabregat A, Sidiropoulos K, Viteri G, Forner O, Marin-Garcia P, Arnau V, D’Eustachio P, Stein L, Hermjakob H. Reactome pathway analysis: a high-performance in-memory approach. BMC Bioinformation. 2017;18:142. doi: 10.1186/s12859-017-1559-2
43.
Pirruccello JP, Chaffin MD, Chou EL, Fleming SJ, Lin H, Nekoui M, Khurshid S, Friedman SF, Bick AG, Arduini A, et al. Deep learning enables genetic analysis of the human thoracic aorta. Nat Genet. 2022;54:40–51. doi: 10.1038/s41588-021-00962-4
44.
Heinrich PC, Behrmann I, Haan S, Hermanns HM, Muller-Newen G, Schaper F. Principles of interleukin (IL)-6-type cytokine signalling and its regulation. Biochem J. 2003;374:1–20. doi: 10.1042/BJ20030407
45.
Hillary RF, Trejo-Banos D, Kousathanas A, McCartney DL, Harris SE, Stevenson AJ, Patxot M, Ojavee SE, Zhang Q, Liewald DC, et al. Multi-method genome- and epigenome-wide studies of inflammatory protein levels in healthy older adults. Genome Med. 2020;12:60. doi: 10.1186/s13073-020-00754-1
46.
van den Borne P, Quax PH, Hoefer IE, Pasterkamp G. The multifaceted functions of CXCL10 in cardiovascular disease. Biomed Res Int. 2014;2014:893106. doi: 10.1155/2014/893106
47.
Segers D, Lipton JA, Leenen PJM, Cheng C, Tempel D, Pasterkamp G, Moll FL, de Crom R, Krams R. Atherosclerotic plaque stability is affected by the chemokine CXCL10 in both mice and humans. Int J Inflam. 2011;2011:936109. doi: 10.4061/2011/936109
48.
Heller EA, Liu E, Tager AM, Yuan Q, Lin AY, Ahluwalia N, Jones K, Koehn SL, Lok VM, Aikawa E, et al. Chemokine CXCL10 promotes atherogenesis by modulating the local balance of effector and regulatory T cells. Circulation. 2006;113:2301–2312. doi: 10.1161/CIRCULATIONAHA.105.605121
49.
van Wanrooij EJA, de Jager SCA, van Es T, de Vos P, Birch HL, Owen DA, Watson RJ, Biessen EAL, Chapman GA, van Berkel TJC, et al. CXCR3 antagonist NBI-74330 attenuates atherosclerotic plaque formation in LDL receptor–deficient mice. Arterioscler Thromb Vasc Biol. 2008;28:251–257. doi: 10.1161/ATVBAHA.107.147827
50.
van Wanrooij EJA, Happé H, Hauer AD, de Vos P, Imanishi T, Fujiwara H, van Berkel TJC, Kuiper J. HIV entry inhibitor TAK-779 attenuates atherogenesis in low-density lipoprotein receptor–deficient mice. Arterioscler Thromb Vasc Biol. 2005;25:2642–2647. doi: 10.1161/01.ATV.0000192018.90021.c0
51.
McLoughlin RM, Jenkins BJ, Grail D, Williams AS, Fielding CA, Parker CR, Ernst M, Topley N, Jones SA. IL-6 trans-signaling via STAT3 directs T cell infiltration in acute inflammation. Proc Natl Acad Sci U S A. 2005;102:9589–9594. doi: 10.1073/pnas.0501794102
52.
Szentes V, Gazdag M, Szokodi I, Dézsi CA. The role of CXCR3 and associated chemokines in the development of atherosclerosis and during myocardial infarction. Front Immunol. 2018;9:1932. doi: 10.3389/fimmu.2018.01932
53.
Wolf D, Ley K. Immunity and inflammation in atherosclerosis. Circ Res. 2019;124:315–327. doi: 10.1161/CIRCRESAHA.118.313591
54.
Rose-John S, Jenkins BJ, Garbers C, Moll JM, Scheller J. Targeting IL-6 trans-signalling: past, present and future prospects. Nat Rev Immunol. 2023;23:1–16. doi: 10.1038/s41577-023-00856-y
55.
Garbers C, Monhasery N, Aparicio-Siegmund S, Lokau J, Baran P, Nowell MA, Jones SA, Rose-John S, Scheller J. The interleukin-6 receptor Asp358Ala single nucleotide polymorphism rs2228145 confers increased proteolytic conversion rates by ADAM proteases. Biochim Biophys Acta. 2014;1842:1485–1494. doi: 10.1016/j.bbadis.2014.05.018
56.
Ferreira RC, Freitag DF, Cutler AJ, Howson JM, Rainbow DB, Smyth DJ, Kaptoge S, Clarke P, Boreham C, Coulson RM, et al. Functional IL6R 358Ala allele impairs classical IL-6 receptor signaling and influences risk of diverse inflammatory diseases. PLoS Genet. 2013;9:e1003444. doi: 10.1371/journal.pgen.1003444
57.
Haglund A, Zuber V, Yang Y, Abouzeid M, Feleke R, Ko JH, Nott A, Babtie AC, Mills JD, Muhammed L, et al. Single-cell Mendelian randomisation identifies cell-type specific genetic effects on human brain disease and behaviour. bioRxiv. 2011;2022:2028–517913. doi: 10.1101/2022.11.28.517913
58.
Zheng J, Haberland V, Baird D, Walker V, Haycock PC, Hurle MR, Gutteridge A, Erola P, Liu Y, Luo S, et al. Phenome-wide Mendelian randomization mapping the influence of the plasma proteome on complex diseases. Nat Genet. 2020;52:1122–1131. doi: 10.1038/s41588-020-0682-6
59.
Sadler MC, Auwerx C, Lepik K, Porcu E, Kutalik Z. Quantifying the role of transcript levels in mediating DNA methylation effects on complex traits and diseases. Nat Commun. 2022;13:7559. doi: 10.1038/s41467-022-35196-3
60.
Haslam DE, Li J, Dillon ST, Gu X, Cao Y, Zeleznik OA, Sasamoto N, Zhang X, Eliassen AH, Liang L, et al. Stability and reproducibility of proteomic profiles in epidemiological studies: comparing the Olink and SOMAscan platforms. Proteomics. 2022;22:e2100170. doi: 10.1002/pmic.202100170
61.
Lissilaa R, Buatois V, Magistrelli G, Williams AS, Jones GW, Herren S, Shang L, Malinge P, Guilhot F, Chatel L, et al. Although IL-6 trans-signaling is sufficient to drive local immune responses, classical IL-6 signaling is obligate for the induction of T cell-mediated autoimmunity. J Immunol. 2010;185:5512–5521. doi: 10.4049/jimmunol.1002015
62.
Ligthart S, Vaez A, Vosa U, Stathopoulou MG, de Vries PS, Prins BP, Van der Most PJ, Tanaka T, Naderi E, Rose LM, et al; LifeLines Cohort Study. Genome analyses of >200,000 individuals identify 58 loci for chronic inflammation and highlight pathways that link inflammation and complex disorders. Am J Hum Genet. 2018;103:691–706. doi: 10.1016/j.ajhg.2018.09.009
63.
Sinnott-Armstrong N, Tanigawa Y, Amar D, Mars NJ, Aguirre M, Venkataraman GR, Wainberg M, Ollila HM, Pirruccello JP, Qian J, et al. Genetics of 38 blood and urine biomarkers in the UK Biobank. bioRxiv. 2019;660506. doi: 10.1101/660506
64.
Sun BB, Maranville JC, Peters JE, Stacey D, Staley JR, Blackshaw J, Burgess S, Jiang T, Paige E, Surendran P, et al. Genomic atlas of the human plasma proteome. Nature. 2018;558:73–79. doi: 10.1038/s41586-018-0175-2
65.
Thorand B, Kolb H, Baumert J, Koenig W, Chambless L, Meisinger C, Illig T, Martin S, Herder C. Elevated levels of interleukin-18 predict the development of type 2 diabetes: results from the MONICA/KORA Augsburg Study, 1984–2002. Diabetes. 2005;54:2932–2938. doi: 10.2337/diabetes.54.10.2932
66.
Then C, Then HL, Lechner A, Thorand B, Meisinger C, Heier M, Peters A, Koenig W, Rathmann W, Scherberich J, et al. Serum uromodulin and risk for cardiovascular morbidity and mortality in the community-based KORA F4 study. Atherosclerosis. 2020;297:1–7. doi: 10.1016/j.atherosclerosis.2020.01.030
Information & Authors
Information
Published In
Copyright
© 2023 American Heart Association, Inc.
Versions
You are viewing the most recent version of this article.
History
Received: 2 April 2023
Accepted: 17 November 2023
Published online: 28 December 2023
Published in print: 27 February 2024
Keywords
Subjects
Authors
Disclosures
Disclosures Dr Anderson receives sponsored research support from the American Heart Association (AHA) (18SFRN3425000) and Bayer AG, and has consulted for ApoPharma, Inc. Dr Rosand receives sponsored research support from the National Institutes of Health, AHA and OneMind, and consults for Takeda Pharmaceuticals. Dr Koenig reports advisory board fees from AstraZeneca, Novartis, Amgen, Pfizer, The Medicines Company, DalCor, Kowa, Corvidia, OMEICOS, Daiichi-Sankyo, Novo Nordisk, New Amsterdam Pharma, TenSixteen Bio, Esperion, Genentech; lecture fees from Bristol-Myers Squibb, Novartis, Amgen, Berlin-Chemie, Sanofi and AstraZeneca; grants and nonfinancial support from Abbott, Roche Diagnostics, Beckmann, and Singulex, outside the submitted work. Dr Bernhagen is inventor on patent applications related to anti-MIF/chemokine strategies in inflammatory and cardiovascular diseases.
Sources of Funding
This work was supported by grants to Dr Georgakis from the German Research Foundation (Deutsche Forschungsgemeinschaft [DFG]) in the form of a Walter-Benjamin Fellowship (GZ: GE 3461/1-1, ID: 466957018), within the framework of the Munich Cluster for Systems Neurology (EXC 2145 SyNergy, ID 390857198), and in the form of an Emmy Noether grant (GZ: GE 3461/2-1, ID 512461526), as well as grants from the Fritz-Thyssen Foundation (Ref. 10.22.2.024MN) and the Hertie Foundation (Hertie Network of Excellence in Clinical Neuroscience, ID P1230035). The project has received funding from the German Research Foundation (DFG) as part of the Munich Cluster for Systems Neurology (EXC 2145 SyNergy – ID 390857198, to Dr Dichgans), the CRC 1123 (B3; to Dr Dichgans), DI 722/16-1 (ID: 428668490/405358801, to Dr Dichgans), DI 722/13-1 and DI 722/21-1 (to Dr Dichgans); a grant from the Leducq Foundation (to Dr Dichgans); the European Union’s Horizon Europe (European Innovation Council) programme under grant agreement No 101115381 (to Dr Dichgans); and ERA-NET Neuron (MatriSVDs, to Dr Dichgans). Dr Anderson is supported by National Institutes of Health R01NS103924, U01NS069673, American Heart Association (AHA) 18SFRN34250007, AHA-Bugher 21SFRN812095, and the MGH McCance Center for Brain Health for this work. The German Diabetes Center is funded by the German Federal Ministry of Health (Berlin, Germany) and the Ministry of Culture and Science of the state North Rhine-Westphalia (Düsseldorf, Germany) and receives additional funding from the German Federal Ministry of Education and Research (BMBF) through the German Center for Diabetes Research (DZD e.V.). Dr Bernhagen is supported by DFG project SFB1123/A3 and EXC 2145 SyNergy – ID 390857198. The KORA study was initiated and financed by the Helmholtz Zentrum München – German Research Center for Environmental Health, which is funded by the German Federal Ministry of Education and Research (BMBF) and by the State of Bavaria. Data collection in the KORA study is done in cooperation with the University Hospital of Augsburg.
Metrics & Citations
Metrics
Citations
Download Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Select your manager software from the list below and click Download.
- Translating Anti-Inflammatory Strategies for Atherosclerosis: Deep Phenotyping, Next-Generation Drug Targets, and Precision Medicine, Cells, 13, 15, (1306), (2024).https://doi.org/10.3390/cells13151306
- Rethinking of phosphodiesterase 5 inhibition: the old, the new and the perspective in human health, Frontiers in Endocrinology, 15, (2024).https://doi.org/10.3389/fendo.2024.1461642
- Circulating CXCL9, monocyte percentage, albumin, and C-reactive protein as a potential, non-invasive, molecular signature of carotid artery disease in 65+ patients with multimorbidity: a pilot study in Age.It, Frontiers in Endocrinology, 15, (2024).https://doi.org/10.3389/fendo.2024.1407396
- Mendelian Randomization Applied to Neurology, Neurology, 102, 4, (2024).https://doi.org/10.1212/WNL.0000000000209128
- Systematic Druggable Genome‐Wide Mendelian Randomization Identifies Therapeutic Targets for Functional Outcome After Ischemic Stroke, Journal of the American Heart Association, 13, 16, (2024)./doi/10.1161/JAHA.124.034749
- Association between aortic calcification and cytokine levels in patients with peripheral artery disease, Clinical and Translational Science, 17, 10, (2024).https://doi.org/10.1111/cts.70036
- Targeting immune cell recruitment in atherosclerosis, Nature Reviews Cardiology, (2024).https://doi.org/10.1038/s41569-024-01023-z
- Genetically proxied IL‐6 signaling and risk of Alzheimer's disease and lobar intracerebral hemorrhage: A drug target Mendelian randomization study, Alzheimer's & Dementia: Translational Research & Clinical Interventions, 10, 3, (2024).https://doi.org/10.1002/trc2.70000
Loading...
View Options
Get Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginPurchase Options
Purchase this article to access the full text.
eLetters(0)
eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.
Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.