Back

Annexin A2 mediates RIG-I-like receptor responses to viral infection

Wray-McCann, G.; Michalik, D.; McAllan, A. L.; Gearing, L. J.; Liou, R. H.-C.; Campbell, N. K.; Straub, S.; Hosseni-Far, H.; Matthews, A.; Tate, M. D.; Davidson, S.; Fisher, R.; Bernecky, C.; Rehwinkel, J.; Hertzog, P. J.; Sampaio, N. G.

2025-10-29 immunology
10.1101/2025.10.28.684968 bioRxiv
Show abstract

RIG-I and MDA5 are critical sensors of virus infection and detect many important human pathogens including SARS-CoV-2, influenza A virus (IAV), Zika virus and West Nile virus (WNV). MDA5 is activated by binding to double-stranded RNA generated during infection and induces strong pro-inflammatory and antiviral responses mediated largely via type I interferons. Activation of MDA5 is also implicated in monogenic (e.g. Aicardi-Goutieres syndrome) and complex polygenic (e.g. systemic lupus erythematosus) autoinflammatory diseases, demonstrating the importance of appropriate regulation of this pathway. Yet, how MDA5 is regulated is poorly defined. We employed SILAC-mass spectrometry to discover MDA5 binding partners in virus-infected cells. Our screen revealed Annexin A2 (ANXA2) as an interactor of MDA5 specifically during infection. Direct binding of ANXA2 and MDA5 was recapitulated in vitro and required Ca2+, and ANXA2 promoted MDA5 oligomerisation in cells. Loss of ANXA2 in human cells significantly blunted interferon responses to EMCV, WNV and IAV, demonstrating a role for ANXA2 in activation of both MDA5 and RIG-I. Using human induced pluripotent stem cells, we found that ANXA2 was required for potent neuronal cell responses to WNV infection and macrophage responses to multiple viral pathogens. Thus, we find that ANXA2 is a critical mediator of RIG-I-like receptor pathways.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.