Background: TREX1 is a negative regulator of the type I interferon response and TREX1 variants are widely considered to confer strong risk for non-monogenic lupus. The lack of validated instruments to capture peripheral type I interferon activity in large biobanks, however, limits the replication of these genetic associations. As such, there is a pressing need for a peripheral type I interferon signature measurement which can be used in large biobanks to adequately characterise immune endotypes and re-evaluate what is considered to be a canonical genetic association.

Objectives: We aimed to i) build a proteomic type I interferon signature, ii) define type I interferon-related autoimmunity, and iii) re-examine the association of rare TREX1 variants with non-monogenic lupus and type I interferon-related autoimmunity.

Methods: We used data from UK Biobank (UKB), a population-based study of 502k adult volunteers recruited in 2006-2010. We first built and validated the Markers of Interferon-α Response in Olink (MIRO) score defined from 12 interferon-stimulated gene products captured in UKB. Second, we assessed the association of MIRO with 20 autoimmune conditions to define type I interferon-related autoimmunity. Third, we leveraged this proteomic instrument to re-examine TREX1 lupus risk variants. Whole-exome sequencing was used to identify rare TREX1 variants reported as non-monogenic lupus risk factors or causing TREX1 -related monogenic disease in ClinVar, LOVD and prior publications. To further characterise 10 reported lupus risk variants, we assessed their functional impact using recombinant protein studies in Trex1 ^-/- mouse embryonic fibroblasts. We tested associations with Firth penalized logistic and linear regressions and replicated results in European ancestry participants using SKAT-O and deleteriousness predictions (LOFTEE). We finally performed a systematic review and meta-analysis of TREX1 -lupus genetic association studies using concepts “genetic variants”, “TREX1” and “lupus”.

Results: A peripheral type I interferon signature (proxied by MIRO; Figure 1A) was observed in UKB participants on interferon therapy and with chronic viral hepatitis (positive controls) but not in those with haemorrhoids (negative control; Figure 1B). MIRO scores were markedly correlated with known interferon-stimulated genes (strongest correlation: galectin 9, r=0.692, p<0.0001). A substantial type I interferon signature was observed for nine conditions including lupus and Sjogren disease, which we defined as type I interferon-related autoimmunity (Figure 1C). These results were then used to comprehensively re-examine TREX1 lupus risk variants. Among 469,229 UKB participants, we identified 1,216 carriers (no homozygote) of any of 35 TREX1 variants (Figure 2A) and 1,306 lupus cases (69.5% from hospital admissions). Participants carrying any of the 10 previously reported TREX1 lupus risk variants did not have an increased risk of lupus (OR=0.178; 95% CI: 0.001, 1.207; p=0.090). In functional assays, only two of these 10 variants (R114H, P212Hfs) had any effect on TREX1 nuclease activity. Carriers of rare TREX1 variants did not have an increased risk of type I interferon-related autoimmunity, other autoimmune conditions or neurological comorbidities of lupus (Figure 2B), and did not have higher MIRO scores (β=0.01; 95% CI: -0.08, 0.10; p=0.829). These neutral results were reproduced with SKAT-O using very rare and high-confidence loss-of-function variants. In a meta-analysis of three published TREX1 -lupus studies and UKB data, we found no significant association of rare TREX1 variants with lupus (OR=2.16; 95% CI: 0.63, 7.34; p=0.219), with substantial heterogeneity (I ² =76.6%) driven by two studies (including the first-in-topic candidate gene study) that did not account for population stratification (R ² =76.8%; Figure 2C).

Conclusion: We built a proteomic type I interferon signature (the MIRO score) to identify type I interferon-related autoimmune diseases in UKB. Leveraging this proteomic signature, we found no significant association between TREX1 variants, non-monogenic lupus or type I interferon-related autoimmunity in UKB. Taken together, these results do not support the reputed risk of lupus conferred by rare TREX1 variants which has become embedded in narrative reviews. At a practical level, we suggest there is no merit in sequencing TREX1 in patients with lupus, where monogenic disease is not suspected.

Figure 1.

Derivation and validation of a proteomic interferon score and identification of interferonopathic autoimmune diseases in UK Biobank. Panel A : Flowchart of Olink Explore 3072 targets in MIRO score. Panel B : Peripheral type I interferon activation using MIRO in linear regressions adjusted for age and sex. P-values with bold characters indicate statistical significance (p<0.0021 or 0.05/24 tests). Scores available for 44,843 UKB participants. Panel C : Type I interferon-related autoimmune diseases (significant increase with β≥0.2; colour: β, diameter: sample size). Abbreviations : Ad, Addison’s disease; AS, ankylosing spondylitis; DM, dermatomyositis; IBD, inflammatory bowel disease; MG, myasthenia gravis; MS, multiple sclerosis; NPX, normalized protein expression; PBC, primary biliary cirrhosis; PMR, polymyalgia rheumatica; RA, rheumatoid arthritis; SjD, Sjogren disease; SLE, systemic lupus erythematosus; SSc, systemic sclerosis.

Figure 2.

Association of rare TREX1 variants with autoimmunity in UK Biobank and systematic review. Panel A : Lollipop plot of carrier counts in UK Biobank for rare TREX1 variants. Panel B : Clinical associations of rare TREX1 variants in UK Biobank. Panel C : Forest plot of previously reported TREX1 -lupus genetic association studies meta-analysed with UK Biobank results (random effects).

REFERENCES: NIL.

Acknowledgements: NIL.

Disclosure of Interests: None declared.

© The Authors 2025. This abstract is an open access article published in Annals of Rheumatic Diseases under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ ). Neither EULAR nor the publisher make any representation as to the accuracy of the content. The authors are solely responsible for the content in their abstract including accuracy of the facts, statements, results, conclusion, citing resources etc.