In recent years, some studies have proved that human plasma proteins play essential roles in the pathogenesis of SpA. However, the genetic mechanism underlying the correlation remained unknown. In this study, we conducted the LD score regression analysis to explore the genetic correlation of human plasma proteins and SpA. Several proteins were found to be associated with SpA. As far as we have known, this is the first study systematically exploring the impact of human plasma proteins on SpA.
IL-17D showed the suggestive genetic correlation evidence with SpA through using the LD score regression analysis and GEO2R application. IL-17 is a novel family of inflammatory cytokines. Excessive activation of IL-17 signaling can contribute to autoimmunity or chronic inflammatory disease. Some studies have shown that the IL-17 family plays an essential role in the development of SpA[20, 21]. For example, Taams LS et al. indicated that the IL-17A inhibitors show efficacy in treating multiple facets of SpA, including psoriasis, enthesitis, synovitis, bone erosion, and new bone formation, which illustrates the importance of IL-17A in pathophysiology of SpA. IL-17A is the most studied among the IL-17 family. IL-17A can degrade the extracellular matrix within the joint by inducing the production of matrix metalloproteinases (MMPs), and IL-17A can lead to active osteoclasts and destruct the bone by up-regulating receptor activator of nuclear factor-κB (NF-κB) expression. What’s more, IL-17A promotes angiogenesis, increasing blood flow and facilitating the influx of inflammatory cells into the joint, leading to joint inflammation. Although few studies on the pathologic mechanism effect of IL-17D on SpA, both IL-17A and IL-17D belong to the interleukin-17 family. They have homology with each other, including similar amino acid sequences and structure-function. So, IL-17D may have potential mechanisms in the pathogenesis of SpA.
Trypsin-2 showed the genetic correlation evidence with SpA. Trypsin-2 can degrade the major structural collagen of articular cartilage, type II collagen, and activates several collagenases . These mechanisms are capable of contributing to tissue-degenerative diseases such as rheumatoid arthritis . Moilanen M et al. found trypsinogen-2 can degrade type I collagen, involving the initial action of collagenolytic matrix metalloproteinases (MMP-1, -8, and − 13) activated by MMP-3. It has been reported that MMPs, and MMP-3 in particular, are produced in response to cytokines in the joints, being more highly expressed in the synovial tissues of SpA patients than in peripheral blood mononuclear cells . In addition, S Sun et al. detected a significantly increased expression of alpha 1-anti-trypsin (ATA1) in synovial membranes of patients with ankylosing spondylitis . Trypsin-2 may become a suggestive human plasma protein for the pathogenesis of SpA, though the precise genetic relationship requires further research.
Serpin A9 also showed the genetic correlation with SpA. Serpin A9 is a protease inhibitor that inhibits trypsin and trypsin-like serine proteases. The genetic correlations between the trypsin and SpA have already been discussed in our study above. STRING (http://string-db.org) is a database that aims to provide a critical assessment and integration of protein-protein interactions, including direct (physical) as well as indirect (functional) associations . By using the STRING database, we found there were functional partnerships and interactions among Serpin A9 and germinal center-associated signaling and motility protein(GCSAM), interferon regulatory factor-4 (IRF-4). Serpin A9 can affect the GCSAM, which mediates the migration-inhibitory effects of IL-6. Da-He Li et al. proved that IL-6 can up-regulate Annexin A2, which may promote ligament ossification, and down-regulating Annexin A2 can ameliorate ossification of fibroblasts from patients with ankylosing spondylitis. We found the Serpin A9 could regulate the expression of interferon regulatory factor-4 (IRF-4) by using STRING database. Chen Q et al. found IRF-4-binding protein can inhibit IL-17 through controlling the activity of the IRF-4 transcription factor. IL-17, known as a family of inflammatory cytokines, has already been discussed in our study above too. Although no researchers have studied whether Serpin A9 has a direct effect on SpA, our results provide a potential insight to study the genetic mechanism of SpA.
Neurensin-1 showed the genetic correlation evidence with SpA. Researchers found Neurensin-1 has an impact on the expression of growth arrest and DNA damage-inducible protein beta (GADD45B). GADD45B can encode a ubiquitously expressed protein that is often associated with growth arrest and apoptosis through demethylating CpG islands of representative gene targets. Some studies have demonstrated that DNA demethylation and hypermethylation play significant roles in arthritis. For example, Dániel M Tóth et l. found the DNA demethylation induced by 5′-azaC halted arthritis progression in mice, which can attribute to the inhibited production of IgG1 antibodies. This study proved that DNA hypermethylation plays a leading role in the pathogenesis of autoimmune arthritis and DNA demethylation has a therapeutic potential in arthritis management. What’s more, Keremu et al. found Neurensin-2 could promote osteosarcoma cell proliferation and growth by dysregulating Wnt/β-catenin signaling. It was reported that the Wnt signaling is likely to play essential roles in the process of ankylosis in SpA, and the activation of β-catenin signaling in cartilage tissue may be the key event leading to spine and joint destruction in patients with SpA. To sum up, the roles of Neurensin-1 in SpA are potential, and the precise relationship between Neurensin-1 and SpA requires further research.
The strength of our study is that we used the latest large-scale GWAS summary dataset of SpA to observe multiple genetic correlations between SpA and human plasma proteins through LD score regression analysis[5, 15]. The large sample size of GWAS summary data can ensure the accuracy of our research results. We verified whether human plasma proteins genetically related to SpA were differentially expressed at the gene expression level. These results were better provided the latest clues for future research on the genetic mechanism of SpA.
This study also has some limitations. Firstly, the GWAS summary data came from the UK Biobank cohort and based on European ancestry, so our results may not apply to other ancestry studies. Further LD score regression of different populations is needed to prove our results. Secondly, to validate the TWAS results, we compared the significant genes identified by TWAS of SpA with the differentially expressed genes detected by the mRNA expression profiles of psoriatic arthritis. However psoriatic arthritis is just one disease within SpA. So our results should be interpreted with caution. Further biological studies should be conducted to confirm our findings.