sOX40 and sOX40L: Genetically motivated biomarkers with prognostic and predictive potential in SLE

Systemic lupus erythematosus (SLE) is a complex autoimmune disease that can affect many parts of the body. Many current therapies are broad and non-specific, such as steroids and general immune suppressants, and often come with significant side effects. Each person with lupus experiences a unique mix of immune system abnormalities, which makes individualized treatment essential; but further study is required to identify new targeted therapies.

Led by Dr Cunninghame Graham, the research project focuses on two immune system proteins: OX40 and OX40L. These proteins are found on the surface of certain immune cells, and their presence is associated with increased disease activity in lupus. The underlying genes for OX40 and OX40L are also associated with a higher risk of developing lupus. The project explores three key questions: whether the protein levels of OX40 and OX40L in the blood can help predict disease severity and treatment response; how these proteins may contribute to lupus at a cellular level; and whether available drugs that block the interaction between OX40 and OX40L could be effective in treating lupus. Uniquely, because these proteins are found in lupus to be freely circulating in the bloodstream, detached from their cells, they may more readily serve as biomarkers because they can be detected by readily available techniques. Ultimately, the research will determine whether serum levels of OX40 and OX40L might predict which patients are most likely to benefit from therapies that target this pathway.

What this means for people with lupus:

By identifying patients whose disease may be driven by the OX40-OX40L pathway, healthcare providers may be able to offer therapies that are more effective – potentially improving quality of life and long-term outcomes by moving toward a more precise and individualized approach.

OX40 and its exclusive receptor OX40L, play a vital role in SLE pathogenesis. Both loci are associated with lupus and risk alleles are cis-eQTLs. Preliminary protein level data showed raised serum OX40 and OX40L (sOX40/L) in SLE patients compared to healthy controls. Furthermore, higher OX40/L transcript levels are mirrored by enhanced sOX40/L protein levels in SLE patients compared with healthy controls, suggesting that serum OX40 and/or OX40L are potential prognostic biomarkers of disease activity. To accommodate disease heterogeneity, our cohort of adult, jSLE and flaring patients will have equal numbers of European and African/Asian ancestry samples. The LRA GTSA dataset (co-PI: Vyse) has a similar demographic. We have included ~150 individuals with longitudinal sera to test the prognostic capability of sOX40/L. We hypothesise that elevated OX40-OX40L co-stimulation drives increased disease activity, with the most likely mechanism that OX40-OX40L over-activity perturbs the balance between T-regulatory and T-follicular helper cells activity on B-cells during the pathological humoral immune response. We propose that sOX40/L levels are a predictive and prognostic biomarker of changes in germinal centre responses in lupus. The genetic heterogeneity of SLE arises from the dysregulation of multiple immunological processes. It is reasonable to assume that patients exhibiting dysregulated OX40-OX40L response delineate a distinct subgroup of patients from those with upregulated interferon response. These hypotheses will be tested by multi-analyte immunoassays for serum OX40/L, IFN-alpha and B-cell activity. By incorporating detailed demographic and clinical outcome data into regression/model-based approaches we will find whether patients with high sOX40/L are a distinct subgroup from those with high IFN. By integrating GWAS/WGS with global proteomic data (Olink) we will define both cis- and trans-pQTLs for sOX40/L which identify regulatory pathways for the OX40/L axis. Mendelian Randomisation will show whether any pQTLs are causal and/or improve the Genetic Risk Score for disease. We will investigate whether the strength of the OX40-OX40L co-stimulatory signal influences the levels of soluble OX40/L, using flow-cytometric analysis of co-cultured activated T and B cells, and also whether supplementation of the cultures with sOX40/L impacts the degree of co-stimulation. Finally, we will ask whether serum OX40 and/or OX40L levels predict the efficacy of blockade of the OX40-OX40L signal by anti-OX40/L monoclonal antibody. Longer term, systematic manipulation of OX40/L regulatory pathways, using CRISPR, will pinpoint key checkpoints for sOX40/L levels. We will explore whether targeted manipulation of germinal centre sOX40/L levels in mice can increase the costimulatory shut-down from the anti-OX40/L therapeutic, something that might be an important detail in strengthening the support for clinical trials of the OX40-OX40L axis in SLE.