Background/Purpose: The strong clinical association between the anti-MDA5 autoantibodies and the development of severe lung disease in patients with dermatomyositis (DM) suggests an active role for the autoantibodies in the pathogenesis. The MDA5-specific autoreactive immune cells are understudied and it remains unclear if (and how) the autoantibodies contribute to the pathogenesis.

We aimed to get insight in the repertoire of the antigen-specific autoreactive B cells from patients with anti-MDA5(+) DM and confirm their antigen-specificity by producing anti-MDA5 monoclonal autoantibodies (mAb).

Methods: Circulating lymphocytes were available from MDA5(+) patients (n=3) at Karolinska Institutet (Sweden). All patients satisfy the EULAR/ACR IIM classification criteria. B cells were enriched using a negative selection kit and stained with a live/dead dye, with Streptavidin (SA)-PE-DL755 decoy probe to label non-specific B cells and with MDA5- SA-PE to label MDA5-specific B cells. Other B cell markers included CD19, IgD, CD27 and CD38. Single MDA5-SA-PE(+) B cells were index sorted using Influx (BD Bioscience) and recombinant mAbs were generated from the B cell receptor sequences as described previously and expressed as IgG1. Quality controls included SDS-PAGE, ELISA, and size-exclusion chromatography. Unspecific autoreactivity was assessed in a soluble membrane protein polyreactivity assay. MDA5-specificity of the produced mAbs was evaluated with ELISA and Western Blot. Apparent affinity was determined using a competition ELISA.

Results: Out of 234 MDA5(+)B cells sorted, 23 paired heavy and light chain sequences were selected to be expressed whereof 9 were originally IgGs, 3 IgAs and 11 IgMs. Two out of 23 mAbs were highly reactive towards the MDA5 protein in ELISA. The sequences originated from IgG-producing switched memory B cells (CD19+CD27loIgD-CD38+) and contained a low number of heavy and light variable region somatic hypermutations (SHM, 4 and 1 resp.). ELISA moreover showed that both mAbs specifically recognize the Hel2i domain of the MDA5 protein, but they differ in apparent affinity. Notably, one clone displayed sub-nanomolar apparent affinity, despite low SHM. The patient from whom the mAbs originated showed radiographic signs of ILD, had skin involvement and was lymphopenic.

Conclusion: We successfully cloned anti-MDA5 mAbs from two different B cell receptor sequences from an MDA5(+) patient with DM. Both mAbs specifically recognize the Hel2i domain of the MDA5 protein, but with different apparent affinity. These results are in line with previous findings, where we found the helicase domains to be the main immunogenic domains in 30 MDA5(+) serum samples. Therefore we hypothesize the anti-MDA5 autoantibodies potentially compete with the natural ligand of the MDA5 protein and interfere with the canonical function of MDA5 as a sensor of viral RNA. The successful production of anti-MDA5 mAbs allows us to further investigate this hypothesis, starting with setting up an in vitro competition assay to assess if the presence of anti-MDA5 mAbs can affect the downstream MDA5-signalling in lung fibroblasts.

« Back to ACR Convergence 2023

ACR Meeting Abstracts - https://acrabstracts.org/abstract/monoclonal-anti-mda5-autoantibodies-derived-from-a-patient-with-dermatomyositis-target-the-hel2i-domain-of-the-mda5-protein/