Genetics-informed new causal proteins in atopic dermatitis pathogenesis and treatment

doi:10.21203/rs.3.rs-4693021/v1

Download PDF

Article

Genetics-informed new causal proteins in atopic dermatitis pathogenesis and treatment

https://doi.org/10.21203/rs.3.rs-4693021/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Atopic dermatitis (AD) is an immune-related skin disease with a genetic background. Numerous loci have been identified associated with AD to better comprehend its complicated genetic mechanisms by genome‑wide association studies (GWASs). However, current studies reveal the underlying mechanisms of these loci in the pathogenesis of AD inadequately. To explore the plasma proteins genetically correlated with AD via the GWAS data. Herein, we adopted recent AD GWAS data (N = 796,661) and the dataset of plasma protein quantitative trait loci (pQTLs), comprising 1,348 proteins from individuals of European descent. We first conducted the AD-related proteome-wide association studies (PWASs) (N = 7,213) by integrating pQTLs with the AD GWAS statistics and identified six significant plasma proteins by PWAS (P < 0.05). Then, the potential causal proteins of AD were identified via Mendelian randomization (MR), and four causal proteins of AD were discovered afterward. Following this, Bayesian colocalization analysis then explored proteins sharing the same causal variants. Three causal proteins strongly associated with the pathogenesis of AD were eventually pinpointed. Finally, we discovered drugs that could be repurposed for AD with the plasma proteins that might contribute to the pathogenesis of AD in the Drug Gene Interaction Database.

Biological sciences/Genetics/Genetic association study/Genome wide association studies

Biological sciences/Immunology/Autoimmunity

Atopic dermatitis

Genome‑wide association studies

Proteome‑wide association studies

Mendelian randomization

Bayesian colocalization

Atopic dermatitis (AD) is an inflammatory skin disease (affecting up to 24% of children^1,2) and is characterised by recurrent, pruritic, and localised eczema.³ Besides, AD involves complex interactions between genetics and the environment,^4,5 and is highly heritable⁶. Thus, effective treatments are crucial. Currently, topical immunosuppressive therapies are commonly employed for AD.⁷ However, some patients with moderate to severe AD do not respond well to topical corticosteroids.^8–10 Recently, a shift to personalised AD treatments has been requested urgently.¹¹ As a result, it is imperative to develop more revolutionary therapies, such as developing new targeted small-molecule therapies.

Nevertheless, developing new drugs is challenging because of the lack of preclinical data predicting drug efficacy and toxicity, and the eventually approved candidate new molecular drugs account for only 10% of clinical trial entities.¹² By using genome-wide association study (GWAS) data in a computational model, King et al. unexpectedly confirmed that drug targets with human genetic evidence of disease association doubled the probability of drug approval.¹³ GWAS is a powerful tool for exploring the contributions of multiple genetic factors to complex traits and has been used to successfully identify genetic loci associated with diseases. Therefore, several AD-related GWAS have identified numerous loci and Sliz et al. recently identified two novel loci (desmocollin 1 and serpin family B member 7) crucial to epidermal strength and integrity to AD in 22,474 AD cases and 774,187 controls.¹⁴ The significant variants identified in the AD GWAS, however, were mainly in the non-coding regions, and some of these variants had unknown functional mechanisms, making it challenging to interpret gene expression changes and identify real causal single nucleotide polymorphisms (SNPs) in risk loci.^14–16 Therefore, an improved approach to implement gene expression imputation, namely the transcriptome-wide association study (TWAS), has been suggested recently.¹⁵ For example, Song et Al. conducted TWASs (N cases = 9831, N controls = 279,476) on AD using skin tissue and blood, which identified four novel genes and identified their contribution to the abnormal activation of immune responses and the development of AD skin lesions.¹⁷ Despite being able to illustrate the potential mechanisms through expression quantitative trait loci (QTL) effects, the candidate loci identified by TWAS only associate genetically predicted expression and disease risk at the messenger RNA level, ignoring the protein abundance and thus being unable to demonstrate direct links with drug targets.¹⁸

It is becoming increasingly common to use the links between human genetic variation and proteins to find more targets as most small molecules and biological drugs are targeted in proteins.^19,20 Proteins, representing the major functional components of cellular and biological processes and the end products of gene expression, are the most common indicators that can be easily identified. More importantly, they serve as the perfect targets for drug conjugation. The proteome-wide association study (PWAS),^21,22 is a newly developed protein-based technique for identifying proteins associated with studied diseases or phenotypes. Zhang et al. used this method to investigate plasma proteins related to gout and discovered the interleukin (IL) 1 receptor (R) (IL-1R) antagonist protein, leading to the conclusion that anakinra could be used as a drug repurposing treatment for acute gout flares.²³ Thus, PWAS can identify risk loci that have genetic correlations with AD at the protein level and contribute to the discovery of more precise drug target proteins.

This study employed a four-step approach to identify potential AD protein biomarkers. First, a protein QTL (pQTL) dataset obtained from plasma and findings from AD GWAS was used to perform a PWAS analysis. Subsequently, an independent two-sample Mendelian randomization (MR) analysis based on index SNPs for proteins was conducted to identify causally significant plasma proteins associated with AD. Additionally, a Bayesian colocalization analysis was performed to determine whether the PWAS results and causal proteins identified by MR had identical causal variants. Finally, repurposed drugs and druggable protein-coding genes were discovered via the Drug Gene Interaction Database (DGIdb). This is the first study to prioritise candidate causal proteins associated with the underlying mechanisms of AD and identify potential repurposed drugs, providing a clue for developing more drugs.

AD GWAS summary statistics

The largest and most recently updated GWAS of AD from Genome Reference Consortium Human Build 38 was selected. The meta-analysis of the GWAS-summary data for AD, conducted on three biobank resources (Finn Gen research project, the Estonian Biobank, and the UK Biobank) with European-descent participants, comprised 796,661 participants, including 22,474 cases and 774,187 controls.¹⁴ Further information on the study, including sample demographics and the methods used for phenotyping, data processing, and statistical analyses, are provided in the original articles.¹⁴

Human plasma pQTL data

The comprehensive analyses of cis-genetic regulation of the plasma proteome yielded pQTL data obtained from 7,213 European-descent individuals via the large cohorts of the Atherosclerosis Risk in Communities study from four communities across the United States. A total of 1,348 plasma proteins were analysed using the slow-off rate modified aptamer, a proteomic profiling platform, with common covariates, including age and sex, being adjusted.²³ Further information regarding data processing is provided in the original article.

Statistical analysis

Quality control of AD GWAS data

The original study was subject to rigorous quality control (QC), and the data was limited to European ancestry.¹⁴ Furthermore, additional QC was undertaken on the GWAS of AD summary data by (i) aligning to hg19 human reference genomes; (ii) eliminating SNPs with a missing or duplicated reference SNP cluster identification; (iii) restricting to biallelic SNPs with a minor allele frequency of > 0.01 based on the 1000 Genomes Project;²⁴ (iv) subsequent analyses also included additional data processing procedures tailored to the specific requirements of the various methods.

PWASs

We conducted a PWAS analysis by combining AD GWAS summary statistics and cis-regulated pQTL to identify proteins with genetic associations with AD. For this purpose, we employed Functional Summary-based ImputatiON (FUSION), a pipeline based on a group of built functional/molecular genetic models,¹⁵ which is used to compute the effect of SNPs on protein abundance for proteins with significant heritability (heritability P < 0.01). We considered the potential influence of linkage disequilibrium (LD) on the estimated test statistics and thus attempted to minimise it with an LD reference panel.¹⁵ We employed the elastic net model, the most precise of all the predicted models, for our analysis. Subsequently, The FUSION methodology combined the SNP effects from the model of genetically predicted expression with the effects of the same SNPs on AD. We then performed the PWAS of AD by summing the linear sum of the z-score multiplied by the weight of the independent SNPs at the locus. The P-value used in PWAS analysis was adjusted for the False Discovery Rate (FDR).²⁵

MR analysis

MR employs genetic variants associated with exposure as instrumental variables to investigate the causal relationships between risk factors and outcomes.^26,27 Additionally, it is feasible to perform the causal protein profiling of disease by integrating pQTL data. MR is based on three hypotheses²⁶—selected genetic variants must be independent and strongly associated with the exposure of interest, the genetic variants are not associated with any confounders that might affect the exposure-outcome association, and the genetic variants can only be associated with the outcome by influencing the exposure of interest. In our study, MR was used to confirm whether the proteins identified by PWAS were causally associated with AD. This MR analysis adheres to the STROBE-MR statements,²⁸ specifically focusing on instrumental variable selection, instrumental variable assessment, MR analysis, and sensitivity analysis. Independent cis-pQTLs were selected through LD clumping with an r2 threshold of < 0.01 within the 1 Mb cis-region, referencing the European LD panel from the 1000 Genomes Project. We harmonized the effect alleles of instrumental variables (IVs) in both pQTLs and outcome GWAS data.^26,29 Additionally, it's critical to ensure that IVs are appropriate for MR analysis by applying stringent inclusion criteria and removing weak IVs with an F-statistic less than 10. Furthermore, when only a single independent SNP was available, the Wald ratio was used to estimate the causality of exposure to the outcome. ³⁰ Inverse variance weighting (IVW) was regarded as the primary method for performing MR analysis. Horizontal pleiotropy was evaluated by testing whether the y-intercept of the MR-Egger analysis was zero, and P < 0.05 implied the presence of horizontal pleiotropy. The heterogeneity among proteins with multiple independent IVs was assessed using Cochran's Q statistic. Multiple testing corrections were applied with an FDR threshold of < 0.05 (Benjamini-Hochberg method). The MR analysis in our study was carried out via the ‘TwoSampleMR’ package in R.

Bayesian colocalisation analysis

We performed colocalisation analysis using the COLOC R package to explore the probability of the causally identical SNP which drives pQTL and AD.³¹ We used the default COLOC priors of \(\:{p}_{1}\) =\(\:{\:10}^{-4}\), \(\:{p}_{2}\) = \(\:{10}^{-4}\), and \(\:{p}_{12}\) = \(\:{10}^{-5}\), where \(\:{p}_{1}\) is the probability that a given variant was associated with AD, \(\:{p}_{2}\:\)was the probability that a given variant was a significant pQTL, and \(\:{p}_{12}\) was the probability that a given variant is an AD GWAS result and a pQTL. COLOC uses computed approximation Bayes factors and summary association data; the five mutually exclusive hypotheses were tested, which were as follows: (1) no causal SNP was found for either AD GWAS or pQTL (H0); (2) only AD GWAS had a causal SNP (H1); (3) only pQTL had a causal SNP (H2); (4) AD GWAS and pQTL had a causal SNP; however, the two causal SNPs differed (H3); (5) AD GWAS and pQTL had a causal SNP and shared the same SNP (H4). We primarily focused on the last hypothesis (H4), and posterior probability (PP) was used to quantify the support for H4 (denoted as PPH4). We defined strong evidence of colocalisation at PPH4 ≥ 0.75.

Drug-gene interactions

The DGIdb functions as a central repository for data on drug-gene interaction and druggability from various new and updated sources, which cover 41,102 genes and 14,449 drugs.³² Proteins identified by Bayesian colocalisation analysis were used, referencing the open-access database DGIdb, in search of more future therapies to investigate the potential drugs that could treat patients with AD.^32–35

PWAS verified six proteins associated with AD

Using the FUSION pipeline to integrate human plasma proteomes and GWAS results to identify proteins whose abundance is correlated with genetic risk for AD, We identified six proteome-wide significant proteins (IL-18R1, IL-6R, IL-1R like 1 [IL-1RL1], S100 calcium-binding protein A7 [S100A7], matrix metallopeptidase 12 [MMP12], and argininosuccinate lyase [ASL]) associated with AD (FDR of P < 0.05) (IL-18R1 P-value = 7.14×10^− 17, IL-6R P-value = 1.12×10^− 8, IL-1RL1 P-value = 2.46×10^− 6, S100A7 P-value = 1.90×10^− 5, MMP12 P-value = 2.40×10^− 5, and ASL P-value = 1.96×10^− 4) (Fig. 1, Table 1).

MR identified four proteins associated with AD using plasma pQTL

MR analysis was used for validation to identify whether the six cis-regulated blood proteins were causal risk factors. MR estimates of most of the analysed proteins could be based on the IVW. We further confirmed four protein biomarkers that revealed significant evidence of an association in the AD GWAS. The IVW verified that AD risk was associated with IL-18R1 (odds ratio [OR]: 1.071, P = 2.58×10 − 2), IL-6R (OR: 1.054, P = 1.49×10 − 5), S100A7 (OR: 1.057, P = 1.65×10 − 3), and MMP12 (OR: 1.085, P = 2.05×10 − 5). (Fig. 2, Table 2).

Colocalisation between AD risks and pQTLs in the plasma

Statistical Bayesian colocalisation analysis examined the PP for a shared causal variant between a pQTL and AD for the proteins, referred to as H4 ≥ 0.75 to research whether a common causal SNP existed between the AD risk gene and pQTL and identified four proteins which proved colocalisation. It finally suggests that three proteins (IL-18R1, IL-6R, and MMP12) played a crucial role in the pathophysiology of AD (Fig. 3, Fig. 4, Table 2).

Drug-gene interactions of plasma protein-coding genes

Three plasma proteins identified after colocalisation were included to further explore the potential drugs available for AD. By using the plasma protein-coding genes, we found that eight drugs could be repurposed for treating AD based on the IL-6R target, and four drugs could be repurposed for the MMP12 target. All the interacted drugs discovered in the DGIdb were antagonists which inhibited IL-6R and MMP12 targets, bringing positive effects because the two targets were associated with an increased risk of AD (Table 3). Additionally, IL-18R1 was identified as a druggable gene, which suggests it is a good target for developing newer drugs.

In this study, we used pQTL and GWAS data to ascertain the proteins associated with AD that might become potential targets for further investigation and therapeutic research. Following independent PWAS and MR validation analyses, four potential risk proteins (IL-18R1, IL-6R, S100A7, and MMP12) of AD were identified. Further colocalisation analysis confirmed three potential risk proteins, namely IL-18R1, IL-6R, and MMP12. The DGIdb analysis revealed that IL-6R and MMP12 are available for pharmaceutical targeting, with eight (monoclonal antibody [mAb] tocilizumab, sirukumab, sarilumab, vobarilizumab, satralizumab, bazedoxifene, raloxifene, and fluorouracil) and four (zileuton, acetazolamide, captopril, and deferoxamine) drug types, respectively.

The process of discovering drug targets in research is of great significance for clarifying and understanding the mechanism of a disease. Despite increased research efforts in recent years, most omics-level data is centred around genomics and transcriptomics in AD.³⁶ PWAS, a novel functional protein-centric approach to association studies, is based on the concept that causal variants in coding regions affect phenotypes by altering the biochemical functions of the proteins produced by the genes. PWAS not only lowers the need for multiple testing but also provides concrete functional interpretations for the protein-coding genes it discovers.²¹ Zhang et al. proposed the first PWAS framework, demonstrating the potential target curative proteins for disease.²³ Therefore, we employed PWAS to combine the signals of all variants that jointly affect a protein-coding gene and determine their collective impact on the protein's function using machine learning and probabilistic models.²¹ Our analysis implicated proteins that had previously been studied in relation to AD, such as IL-18R1 and IL-6R, as well as new candidates, such as MMP12. These proteins play a crucial role in AD progression by the way of cis-regulation of plasma protein concentration, thereby influencing the deficiency of cell-mediated immune response and the process of inflammatory reaction.

Cytokine imbalance observed in AD can modify cell-mediated immune responses.⁵ The protein encoded by IL-18R1 binds IL-18 specifically, which is critical to signalling transduction mediated by IL-18 and is overexpressed on the skin of patients with AD. It has been suggested that IL-18 is produced by keratinocytes and Langerhans cells,^5,37 and can foster Th1 or Th2 lineage maturation, as well as acting as an important regulator of innate and acquired immunity.³⁷ Research conducted by Shubhada et al. found a significantly higher IL-18R1 expression compared with normal donors. This supports our findings that an increase in IL-18R1 expression will increase the risk of AD. Furthermore, a GWAS previously reported that a functional IL-6R variant is a risk factor for persistent AD.³⁸ AD is a multifactorial disorder associated with CD4 T cells, which can result in IL-6 overexpression. IL-6, bound to IL-6R regulates the immune response, inflammation, and pathogen responses. Our study findings also demonstrate the significance of IL-6R on AD progression. The use of mAb tocilizumab, which has been identified as a promising repurposed drug for AD, has demonstrated the potential to interrupt IL-6R signalling.³⁹ Additionally, sarilumab,⁴⁰ vobarilizumab,⁴¹ and satralizumab⁴² are human mAbs to IL-6R, which is used for treating rheumatoid arthritis or other conditions. A human mAb, sirukumab, selectively binds to the IL-6 cytokine, thereby blocking IL-6 signalling pathways by binding to membrane-bound and soluble IL-6 receptors.⁴² Bazedoxifene,⁴³ raloxifene,⁴⁴ and fluorouracil⁴⁵ produce effects via IL-6/IL-6R/signal transducer and activator of transcription 3 signalling inhibition. The above-targeted drugs are likely to serve as new drugs for AD, improving the quality of life of patients with AD. However, further experimental validation is warranted.

Recent studies have suggested that T helper 17 (Th17) cells play a significant role in AD pathogenesis.^46,47 Consistent with previous studies, this study determined that MMP12, which is associated with Th17 cells, was highly expressed in patients with AD.⁴⁸ MMP12 is a proteinase secreted by macrophages into the wound extracellular space.^49,50 Macrophage-specific MMP-12 can lower the influx of polymorphonuclear leukocytes that is induced by lipopolysaccharides. In addition, MMP12 can cause Th17 cells to aggregate in inflammatory tissue, thereby creating a positive feedback cycle that maintains an inflammatory state in skin lesions.⁵¹ Furthermore, it can proteolyse and cleave chemokines, altering their bioactive properties, and thereby regulating their activity in signalling pathways.⁵² Altogether, the underlying mechanism of MMP12 might be elucidated by the inhibition of leukocytes, the recruitment of more Th17 cells, and the modifications of bioactive chemokines to foster AD progression. The effect of metallopeptidase inhibitors on MMP12 protein might prevent AD progression. Yanes et al. identified that zileuton, a 5-lipoxygenase Inhibitor, significantly improved AD.⁵³ However, no studies have reported the therapeutic effects of acetazolamide (a carbonic anhydrase inhibitor),⁵⁴ captopril (angiotensin-converting enzyme inhibitor),⁵⁵ and deferoxamine (versatile metajl chelator)⁵⁶ on AD. Based on the mechanism of action of MMP12 protein on AD, we believe that the therapeutic effects of these drugs on AD can be further studied.

This study was conducted using the largest and most comprehensive human proteome and summary statistics from the most recent AD GWAS. Although this offers high credibility, some limitations must be acknowledged. Firstly, the scale of pQTL data is restricted to Europeans, not incorporating other ethnic groups. Secondly, the data of our study, limited to plasma proteins, might yield discrepant results for other relevant tissue systems. Furthermore, our research lacks a comprehensive understanding of the molecular mechanism of AD development through more experimental research, which could provide accurate treatment for patients with AD in the future. Future research should benefit from resources in cell types and cell state gene expression, which would augment the accuracy of PWAS, potentially uncovering distinct insights into genetic and functional processes and identifying potential druggable targets for novel AD treatments.

In conclusion, our study provided strong evidence of a causal relationship between three plasma proteins (IL-18R1, IL-6R, MMP12) and AD. These findings help elucidate the genetic mechanisms underlying AD and provide potential AD therapeutic targets for its treatment.

Competing interests:

The authors disclose no competing interests.

Funding:

This work was not sponsored by any funding.

Author Contribution

Study design and major manuscript writing: CL., YJZ., and QFF. CL. contributed to data extraction, quality assessment, analysis and interpretation of data. KXY., LLZ., YY., and WXZ. were involved in drafting the article and revising it critically for important intellectual content. All authors approved the final version to be published. Dr. Li, Dr. Lv, and Dr. Cai had full access to all of the data in the study and took responsibility for the integrity of the data and the accuracy of the data analysis. All authors reviewed the article.

Acknowledgement

We would like to thank the corresponding authors Feng Li, YongQiang Lv, and Yue Cai for providing writing and editorial support.

Data Availability

All analyses conducted in the article used publicly available data. The GWAS summary data for atopic dermatitis is available in the GWAS Catalog, at https://www.ebi.ac.uk/gwas/studies/GCST90027161. The data of plasma protein quantitative trait loci is available at http://nilanjanchatterjeelab.org/pwas.

Laughter, M. R. et al. The global burden of atopic dermatitis: lessons from the Global Burden of Disease Study 1990-2017. Br J Dermatol 184, 304-309; 10.1111/bjd.19580 (2021).
Tian, J. et al. Global epidemiology of atopic dermatitis: a comprehensive systematic analysis and modelling study. Br J Dermatol 190, 55-61; 10.1093/bjd/ljad339 (2023).
Stander, S. Atopic Dermatitis. N Engl J Med 384, 1136-1143; 10.1056/NEJMra2023911 (2021).
Sroka-Tomaszewska, J. & Trzeciak, M. Molecular Mechanisms of Atopic Dermatitis Pathogenesis. Int J Mol Sci 22; 10.3390/ijms22084130 (2021).
David Boothe, W., Tarbox, J. A. & Tarbox, M. B. Atopic Dermatitis: Pathophysiology. Adv Exp Med Biol 1027, 21-37; 10.1007/978-3-319-64804-0_3 (2017).
Bataille, V., Lens, M. & Spector, T. D. The use of the twin model to investigate the genetics and epigenetics of skin diseases with genomic, transcriptomic and methylation data. J Eur Acad Dermatol Venereol 26, 1067-1073; 10.1111/j.1468-3083.2011.04444.x (2012).
Frazier, W. & Bhardwaj, N. Atopic Dermatitis: Diagnosis and Treatment. Am Fam Physician 101, 590-598; (2020).
Wollenberg, A. et al. Consensus-based European guidelines for treatment of atopic eczema (atopic dermatitis) in adults and children: part I. J Eur Acad Dermatol Venereol 32, 657-682; 10.1111/jdv.14891 (2018).
Wollenberg, A. et al. Consensus-based European guidelines for treatment of atopic eczema (atopic dermatitis) in adults and children: part II. J Eur Acad Dermatol Venereol 32, 850-878; 10.1111/jdv.14888 (2018).
Weidinger, S., Beck, L. A., Bieber, T., Kabashima, K. & Irvine, A. D. Atopic dermatitis. Nat Rev Dis Primers 4, 1; 10.1038/s41572-018-0001-z (2018).
Facheris, P., Jeffery, J., Del Duca, E. & Guttman-Yassky, E. The translational revolution in atopic dermatitis: the paradigm shift from pathogenesis to treatment. Cell Mol Immunol 20, 448-474; 10.1038/s41423-023-00992-4 (2023).
Schuhmacher, A., Gassmann, O. & Hinder, M. Changing R&D models in research-based pharmaceutical companies. J Transl Med 14, 105; 10.1186/s12967-016-0838-4 (2016).
King, E. A., Davis, J. W. & Degner, J. F. Are drug targets with genetic support twice as likely to be approved? Revised estimates of the impact of genetic support for drug mechanisms on the probability of drug approval. PLoS Genet 15, e1008489; 10.1371/journal.pgen.1008489 (2019).
Sliz, E. et al. Uniting biobank resources reveals novel genetic pathways modulating susceptibility for atopic dermatitis. J Allergy Clin Immunol 149, 1105-1112 e1109; 10.1016/j.jaci.2021.07.043 (2022).
Gusev, A. et al. Integrative approaches for large-scale transcriptome-wide association studies. Nat Genet 48, 245-252; 10.1038/ng.3506 (2016).
Brown, S. J. What Have We Learned from GWAS for Atopic Dermatitis? J Invest Dermatol 141, 19-22; 10.1016/j.jid.2020.05.100 (2021).
Song, J. et al. Integrative transcriptome-wide analysis of atopic dermatitis for drug repositioning. Commun Biol 5, 615; 10.1038/s42003-022-03564-w (2022).
Wainberg, M. et al. Opportunities and challenges for transcriptome-wide association studies. Nat Genet 51, 592-599; 10.1038/s41588-019-0385-z (2019).
Santos, R. et al. A comprehensive map of molecular drug targets. Nat Rev Drug Discov 16, 19-34; 10.1038/nrd.2016.230 (2017).
Daghlas, I. & Gill, D. Mendelian randomization as a tool to inform drug development using human genetics. Cambridge Prisms: Precision Medicine 1; 10.1017/pcm.2023.5 (2023).
Brandes, N., Linial, N. & Linial, M. PWAS: proteome-wide association study-linking genes and phenotypes by functional variation in proteins. Genome Biol 21, 173; 10.1186/s13059-020-02089-x (2020).
Wingo, A. P. et al. Integrating human brain proteomes with genome-wide association data implicates new proteins in Alzheimer's disease pathogenesis. Nat Genet 53, 143-146; 10.1038/s41588-020-00773-z (2021).
Zhang, J. et al. Plasma proteome analyses in individuals of European and African ancestry identify cis-pQTLs and models for proteome-wide association studies. Nat Genet 54, 593-602; 10.1038/s41588-022-01051-w (2022).
Genomes Project, C. et al. An integrated map of genetic variation from 1,092 human genomes. Nature 491, 56-65; 10.1038/nature11632 (2012).
Wingo, A. P. et al. Large-scale proteomic analysis of human brain identifies proteins associated with cognitive trajectory in advanced age. Nat Commun 10, 1619; 10.1038/s41467-019-09613-z (2019).
Davey Smith, G. & Hemani, G. Mendelian randomization: genetic anchors for causal inference in epidemiological studies. Hum Mol Genet 23, R89-98; 10.1093/hmg/ddu328 (2014).
Burgess, S. et al. Guidelines for performing Mendelian randomization investigations. Wellcome Open Res 4, 186; 10.12688/wellcomeopenres.15555.2 (2019).
Skrivankova, V. W. et al. Strengthening the Reporting of Observational Studies in Epidemiology Using Mendelian Randomization: The STROBE-MR Statement. JAMA 326, 1614-1621; 10.1001/jama.2021.18236 (2021).
Tin, A. & Kottgen, A. Mendelian Randomization Analysis as a Tool to Gain Insights into Causes of Diseases: A Primer. J Am Soc Nephrol 32, 2400-2407; 10.1681/ASN.2020121760 (2021).
Teumer, A. Common Methods for Performing Mendelian Randomization. Front Cardiovasc Med 5, 51; 10.3389/fcvm.2018.00051 (2018).
Giambartolomei, C. et al. Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. PLoS Genet 10, e1004383; 10.1371/journal.pgen.1004383 (2014).
Freshour, S. L. et al. Integration of the Drug-Gene Interaction Database (DGIdb 4.0) with open crowdsource efforts. Nucleic Acids Res 49, D1144-D1151; 10.1093/nar/gkaa1084 (2021).
Cotto, K. C. et al. DGIdb 3.0: a redesign and expansion of the drug-gene interaction database. Nucleic Acids Res 46, D1068-D1073; 10.1093/nar/gkx1143 (2018).
Griffith, M. et al. DGIdb: mining the druggable genome. Nat Methods 10, 1209-1210; 10.1038/nmeth.2689 (2013).
Wagner, A. H. et al. DGIdb 2.0: mining clinically relevant drug-gene interactions. Nucleic Acids Res 44, D1036-1044; 10.1093/nar/gkv1165 (2016).
Ghosh, D., Bernstein, J. A., Khurana Hershey, G. K., Rothenberg, M. E. & Mersha, T. B. Leveraging Multilayered "Omics" Data for Atopic Dermatitis: A Road Map to Precision Medicine. Front Immunol 9, 2727; 10.3389/fimmu.2018.02727 (2018).
Naik, S. M. et al. Human keratinocytes constitutively express interleukin-18 and secrete biologically active interleukin-18 after treatment with pro-inflammatory mediators and dinitrochlorobenzene. J Invest Dermatol 113, 766-772; 10.1046/j.1523-1747.1999.00750.x (1999).
Esparza-Gordillo, J. et al. A functional IL-6 receptor (IL6R) variant is a risk factor for persistent atopic dermatitis. J Allergy Clin Immunol 132, 371-377; 10.1016/j.jaci.2013.01.057 (2013).
Navarini, A. A., French, L. E. & Hofbauer, G. F. Interrupting IL-6-receptor signaling improves atopic dermatitis but associates with bacterial superinfection. J Allergy Clin Immunol 128, 1128-1130; 10.1016/j.jaci.2011.09.009 (2011).
in LiverTox: Clinical and Research Information on Drug-Induced Liver Injury (2012).
Kerschbaumer, A. et al. Efficacy of pharmacological treatment in rheumatoid arthritis: a systematic literature research informing the 2019 update of the EULAR recommendations for management of rheumatoid arthritis. Ann Rheum Dis 79, 744-759; 10.1136/annrheumdis-2019-216656 (2020).
Heo, Y. A. Satralizumab: First Approval. Drugs 80, 1477-1482; 10.1007/s40265-020-01380-2 (2020).
Luo, P. et al. Bazedoxifene exhibits anti-inflammation and anti-atherosclerotic effects via inhibition of IL-6/IL-6R/STAT3 signaling. Eur J Pharmacol 893, 173822; 10.1016/j.ejphar.2020.173822 (2021).
Pozios, I. et al. Raloxifene inhibits pancreatic adenocarcinoma growth by interfering with ERbeta and IL-6/gp130/STAT3 signaling. Cell Oncol (Dordr) 44, 167-177; 10.1007/s13402-020-00559-9 (2021).
Zhong, L. et al. Feng-Liao-Chang-Wei-Kang Combined with 5-Fluorouracil Synergistically Suppresses Colitis-Associated Colorectal Cancer via the IL-6/STAT3 Signalling Pathway. Evid Based Complement Alternat Med 2020, 1395954; 10.1155/2020/1395954 (2020).
Esaki, H. et al. Early-onset pediatric atopic dermatitis is T(H)2 but also T(H)17 polarized in skin. J Allergy Clin Immunol 138, 1639-1651; 10.1016/j.jaci.2016.07.013 (2016).
Zheng, Y. et al. Interleukin-22, a T(H)17 cytokine, mediates IL-23-induced dermal inflammation and acanthosis. Nature 445, 648-651; 10.1038/nature05505 (2007).
Guan, J., Li, Y., Lu, F. & Feng, J. Adipose-derived stem cells ameliorate atopic dermatitis by suppressing the IL-17 expression of Th17 cells in an ovalbumin-induced mouse model. Stem Cell Res Ther 13, 98; 10.1186/s13287-022-02774-7 (2022).
Werb, Z. & Gordon, S. Elastase secretion by stimulated macrophages. Characterization and regulation. J Exp Med 142, 361-377; 10.1084/jem.142.2.361 (1975).
Chou, J., Chan, M. F. & Werb, Z. Metalloproteinases: a Functional Pathway for Myeloid Cells. Microbiol Spectr 4; 10.1128/microbiolspec.MCHD-0002-2015 (2016).
Furue, K., Ito, T., Tsuji, G., Nakahara, T. & Furue, M. The CCL20 and CCR6 axis in psoriasis. Scand J Immunol 91, e12846; 10.1111/sji.12846 (2020).
Rodriguez, D., Morrison, C. J. & Overall, C. M. Matrix metalloproteinases: what do they not do? New substrates and biological roles identified by murine models and proteomics. Biochim Biophys Acta 1803, 39-54; 10.1016/j.bbamcr.2009.09.015 (2010).
Yanes, D. A. & Mosser-Goldfarb, J. L. Emerging therapies for atopic dermatitis: The prostaglandin/leukotriene pathway. J Am Acad Dermatol 78, S71-S75; 10.1016/j.jaad.2017.12.021 (2018).
Mullens, W. et al. Acetazolamide in Acute Decompensated Heart Failure with Volume Overload. N Engl J Med 387, 1185-1195; 10.1056/NEJMoa2203094 (2022).
Zhang, Y. et al. Captopril attenuates TAC-induced heart failure via inhibiting Wnt3a/beta-catenin and Jak2/Stat3 pathways. Biomed Pharmacother 113, 108780; 10.1016/j.biopha.2019.108780 (2019).
Bellotti, D. & Remelli, M. Deferoxamine B: A Natural, Excellent and Versatile Metal Chelator. Molecules 26; 10.3390/molecules26113255 (2021).

Table.1 The AD PWAS identified 6 significant proteins

Protein	CHR	P0	PWAS.Z	PWAS.P	PWAS.FDR
IL18R1	2	102927989	8.345	7.14E-17**	9.50E-14**
IL6R	1	154377669	5.711	1.12E-08**	7.45E-06**
IL1RL1	2	102927962	4.711	2.46E-06**	8.19E-04**
S100A7	1	153430220	4.276	1.90E-05**	5.06E-03*
MMP12	11	102733467	4.224	2.40E-05**	5.32E-03*
ASL	7	65540787	-3.723	1.96E-04**	3.73E-02*

* significant at p<0.05; ** significant at p<0.005

This table presents the z-scores for the PWAS association in AD with their corresponding p-values and the FDR-adjusted p-values for all important genes. These proteins contribute to AD pathogenesis via their cis-regulated plasma protein abundance. AD, atopic dermatitis; ASL, argininosuccinate lyase; CHR, chromosome; FDR, false discovery rate; IL-18R1, interleukin-18 receptor 1; IL-6R, interleukin-6 receptor; IL-1RL1, interleukin-1 receptor like 1; MMP12, matrix metalloproteinase 12; P0, chromosome position; PWAS, proteome-wide association study; PWAS.Z, PWAS z-scores; PWAS.P, PWAS p values; S100A7, S100 calcium-binding protein A7.

Table.2 Risk proteins verified by MR and colocalisation analysis.

Protein	OR(95％CI)	SE	Method	Num_SNP	Pval	FDR	PP.H4.abf	SNP
IL6R	1.054(1.029-1.079)	0.012	IVW	16	1.49E-05**	5.13E-05**	0.992	rs4129267
MMP12	1.085(1.045-1.127)	0.019	IVW	7	2.05E-05**	5.13E-05**	0.935	rs2276109
IL18R1	1.071(1.008-1.138)	0.031	IVW	8	2.58E-02*	3.23E-02*	0.918	rs11465583
S100A7	1.057(1.021-1.094)	0.018	IVW	6	1.65E-03**	2.76E-03**	-	-

* significant at p<0.05; ** significant at p<0.005

CI, confidence interval;FDR, false discovery rate; IL-18R1, interleukin-18 receptor 1; IL-6R, interleukin-6 receptor; MMP12, matrix metalloproteinase 12; MR, Mendelian randomization; Num_SNP, number of single nucleotide polymorphisms; OR, odds ratio; PP.H4, posterior probability for H4; Pval, p values; SE, standard error; SNP, single nucleotide polymorphisms; S100A7, S100 calcium-binding protein A7.

Table.3 Drugs discovered by plasma protein-coding genes and the drug-gene interaction type

Gene	Drug	Interaction Type	Query Score	Interaction Score
IL-6R	SARILUMAB	antagonist (inhibitory)	30.13	48.09
	TOCILIZUMAB	antagonist (inhibitory)	10.76	17.17
	SATRALIZUMAB	antagonist (inhibitory)	6.46	10.30
	SIRUKUMAB	antagonist (inhibitory)	4.30	6.87
	VOBARILIZUMAB	antibody (inhibitory)	4.30	6.87
	BAZEDOXIFENE	antagonist (inhibitory)	2.15	3.43
	RALOXIFENE	antagonist (inhibitory)	0.30	0.47
	FLUOROURACIL	antagonist (inhibitory)	0.08	0.13
MMP12	ZILEUTON	antagonist (inhibitory)	1.08	3.86
	ACETAZOLAMIDE	antagonist (inhibitory)	0.78	2.81
	CAPTOPRIL	antagonist (inhibitory)	0.66	2.38
	DEFEROXAMINE	antagonist (inhibitory)	0.61	2.21

IL-6R, interleukin-6 receptor; MMP12, matrix metalloproteinase 12.

No competing interests reported.

Download PDF

Reviews received at journal
12 Sep, 2024
Reviews received at journal
02 Sep, 2024
Reviewers agreed at journal
29 Aug, 2024
Reviewers agreed at journal
02 Aug, 2024
Reviewers invited by journal
29 Jul, 2024
Editor assigned by journal
29 Jul, 2024
Editor invited by journal
15 Jul, 2024
Submission checks completed at journal
12 Jul, 2024
First submitted to journal
05 Jul, 2024

You are reading this latest preprint version

Genetics-informed new causal proteins in atopic dermatitis pathogenesis and treatment

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

Statistical analysis

Results

Discussion

Conclusion

Declarations

Competing interests:

Funding:

Author Contribution

Acknowledgement

Data Availability

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1