Argonaute, Vault, and Ribosomal Proteins Targeted by Autoantibodies in Systemic Lupus Erythematosus

Abstract

Objective: To expand, in an unbiased manner, our knowledge of autoantigens and autoantibodies in patients with systemic lupus erythematosus (SLE) and evaluate their associations with serological and clinical variables.

Methods: Human proteome arrays (> 21,000 proteins) were screened with serum from patients with SLE (n = 12) and healthy controls (n = 6) for IgG and IgA binding. Top hits were validated with 2 cohorts of patients with SLE (cohort 1, n = 49; cohort 2, n = 46) and other rheumatic diseases by ELISA. Clinical associations of the autoantibodies were tested.

Results: Ro60 was the top hit in the screen, and the 10 following proteins included 2 additional known SLE autoantigens plus 8 novel autoantigens involved in microRNA processing (Argonaute protein 1 [AGO1], AGO2, and AGO3), ribosomes (ribosomal protein lateral stalk subunit P2 and ovarian tumor deubiquitinase 5 [OTUD5]), RNA transport by the vault (major vault protein), and the immune proteasome (proteasome activator complex subunit 3). Patient serum contained IgG reactive with these proteins and IgA against the AGO proteins. Using the 95th percentile of healthy donor reactivity, 5-43% were positive for the novel antigens, with OTUD5 and AGO1 showing the highest percentages of positivity. Autoantibodies against AGO1 proteins were more prevalent in patients with oral ulcers in a statistically significant manner. IgG autoantibodies against AGO proteins were also seen in other rheumatic diseases.

Conclusion: We discovered new autoantigens existing in cytosolic macromolecular protein assemblies containing RNA (except the proteasome) in cells. A more comprehensive list of autoantigens will allow for a better analysis of how proteins are targeted by the autoimmune response. Future research will also reveal whether specific autoantibodies have utility in the diagnosis or management of SLE.

Keywords: autoantibodies; proteins; systemic lupus erythematosus.

Figure 1.. Selected top hits in the HuProt screen for IgG binding to 21,000 human proteins.

A., RO60, B., Smith/RNP, C., XRCC5, D., RPLP2, E., AGO1, F., AGO2, G., AGO3, H., MVP, I., PSME3, J., OTUD5, K., NCKIPSD, L., heatmap visualization of SLE patient reactivity as adjusted values (the highest value for each autoantigen set at 1.0) for the 10 proteins. * denotes p<0.05, ** p<0.01, *** p<0.005, n.s., not significant.

Figure 2.. Selected top hits in the HuProt screen for IgA binding to 21,000 human proteins.

A., HNRNPD, B., AGO1, C., AGO2, D., AGO3, E., heatmap visualization of SLE patient reactivity against the 4 proteins as adjusted values (the highest value for each autoantigen set at 1.0). * denotes p<0.05, ** p<0.01, *** p<0.005, n.s., not significant.

Figure 3.. ELISA for IgG binding to the indicated autoantigens using serum from the cohort 1 (Swedish) of SLE patients and healthy controls (HC).

A., XRCC5, B., RPLP2, C., MVP, D., AGO1, E., AGO2, F., AGO3, G., PSME3, H., NCKIPSD, I., OTUD5. Statistical significance was determined using the Mann-Whitney U test and positivity was determined using the 95^th percentile of the healthy donor data set (dotted line). ** denotes p<0.01, *** p<0.005, and **** p<0.001, n.s., not significant.

Figure 4.. ELISA for IgG binding to the indicated autoantigens using serum from the cohort 2 (UW) of SLE patients and healthy controls (HC).

A., XRCC5, B., RPLP2, C., MVP, D., AGO1, E., AGO2, F., AGO3, G., PSME3, H., NCKIPSD, I., OTUD5. Statistical significance was determined using the Mann-Whitney U test and positivity was determined using the 95^th percentile of the healthy donor data set (dotted line). ** denotes p<0.01, *** p<0.005, and **** p<0.001, n.s., not significant.

Figure 5.. Autoantibody repertoires of individual SLE patients, AGO1/AGO2 correlations, and clinical associations for AGO1.

A., hierarchal clustering and heat map of the ELISA results obtained with cohort 2 (UW) expressed as adjusted values. B., correlation between anti-AGO1 titers with anti-AGO2 titers in cohort 1. C., the same correlation in cohort 2. D., the presence or absence of oral ulcers in SLE patients negative (AGO1-; n=17) or positive (AGO1+; n=12) for IgG autoantibodies against AGO1 (as determined in Fig. 4).

Figure 6.. Presence of autoantibodies in disease controls.

ELISAs with A., AGO1, B., AGO2, C., RPLP2, D., MVP, with healthy controls (HC; n=48) and patients with rheumatoid arthritis (RA; n=20), Sjögren’s syndrome (SjS; n=12), ANCA-associated vasculitis (AAV; n=11), large-vessel vasculitis (LVV; n=9), psoriatic arthritis (PsA; n=8), and SLE cohort 2 (n=46). The percentages of positive values (above 95^th percentile of HC) are shown.