Retinoic acid inducible gene I (RIG-I) is a viral RNA sensor crucial in defense against several viruses including measles, influenza A and hepatitis C. RIG-I activates type-I interferon signalling through the adaptor for mitochondrial antiviral signaling (MAVS). The E3 ubiquitin ligase, tripartite motif containing protein 25 (TRIM25), activates human RIG-I through generation of anchored K63-linked polyubiquitin chains attached to lysine 172, or alternatively, through the generation of unanchored K63-linked polyubiquitin chains that interact non-covalently with RIG-I CARD domains. Previously, we identified RIG-I of ducks, of interest because ducks are the host and natural reservoir of influenza viruses, and showed it initiates innate immune signaling leading to production of interferon-beta (IFN-β). We noted that K172 is not conserved in RIG-I of ducks and other avian species, or mouse. Because K172 is important for both mechanisms of activation of human RIG-I, we investigated whether duck RIG-I was activated by TRIM25, and if other residues were the sites for attachment of ubiquitin. Here we show duck RIG-I CARD domains are ubiquitinated for activation, and ubiquitination depends on interaction with TRIM25, as a splice variant that cannot interact with TRIM25 is not ubiquitinated, and cannot be activated. We expressed GST-fusion proteins of duck CARD domains and characterized TRIM25 modifications of CARD domains by mass spectrometry. We identified two sites that are ubiquitinated in duck CARD domains, K167 and K193, and detected K63 linked polyubiquitin chains. Site directed mutagenesis of each site alone, does not alter the ubiquitination profile of the duck CARD domains. However, mutation of both sites resulted in loss of all attached ubiquitin and polyubiquitin chains. Remarkably, the double mutant duck RIG-I CARD still interacts with TRIM25, and can still be activated. Our results demonstrate that anchored ubiquitin chains are not necessary for TRIM25 activation of duck RIG-I.

Wild waterfowl, including mallard ducks, are the natural reservoir of avian influenza A virus and they are resistant to strains that would cause fatal infection in chickens. Here we investigate potential involvement of TRIM proteins in the differential response of ducks and chickens to influenza. We examine a cluster of TRIM genes located on a single scaffold in the duck genome, which is a conserved synteny group with a TRIM cluster located in the extended MHC region in chickens and turkeys. We note a TRIM27-like gene is present in ducks, and absent in chickens and turkeys. Orthologous genes are predicted in many birds and reptiles, suggesting the gene has been lost in chickens and turkeys. Using quantitative real-time PCR (qPCR) we show that TRIM27-L, and the related TRIM27.1, are upregulated 5- and 9-fold at 1 day post-infection with highly pathogenic A/Vietnam/1203/2004. To assess whether TRIM27.1 or TRIM27-L are involved in modulation of antiviral gene expression, we overexpressed them in DF1 chicken cells, and neither show any direct effect on innate immune gene expression. However, when co-transfected with duck RIG-I-N (d2CARD) to constitutively activate the MAVS pathway, TRIM27.1 weakly decreases, while TRIM27-L strongly activates innate immune signalling leading to increased transcription of antiviral genes MX1 and IFN-β. Furthermore, when both are co-expressed, the activation of the MAVS signalling pathway by TRIM27-L over-rides the inhibition by TRIM27.1. Thus ducks have an activating TRIM27-L to augment MAVS signalling following RIG-I detection, while chickens lack both TRIM27-L and RIG-I itself.

We propose a biochemical mechanism for celiac disease and non-celiac gluten sensitivity that may rationalize many of the extradigestive disorders not explained by the current immunogenetic model. Our hypothesis is based on the homology between the 33-mer gliadin peptide and a component of the NMDA glutamate receptor ion channel - the human GRINA protein - using BLASTP software. Based on this homology the 33-mer may act as a natural antagonist interfering with the normal interactions of GRINA and its partners. The theory is supported by numerous independent data from the literature, and provides a mechanistic link with otherwise unrelated disorders, such as cleft lip and palate, thyroid dysfunction, restless legs syndrome, depression, ataxia, hearing loss, fibromyalgia, dermatitis herpetiformis, schizophrenia, toxoplasmosis, anemia, osteopenia, Fabry disease, Barret's adenocarcinoma, neuroblastoma, urinary incontinence, recurrent miscarriage, cardiac anomalies, reduced risk of breast cancer, stiff person syndrome, etc. The hypothesis also anticipates better animal models, and has the potential to open new avenues of research.

RIG-I plays an essential role in the duck innate immune response to influenza infection. RIG-I engages the critical adaptor protein mitochondrial antiviral signaling (MAVS) to activate the downstream signaling pathway. The influenza A virus non-structural protein PB1-F2 interacts with MAVS in human cells to inhibit interferon production. As duck and human MAVS share only 28% amino acid similarity, it is not known whether the influenza virus can similarly inhibit MAVS signaling in avian cells. Using confocal microscopy we show that MAVS and the constitutively active N-terminal end of duck RIG-I (2CARD) co-localize in DF-1 cells, and duck MAVS is pulled down with GST-2CARD. We establish that either GST-2CARD, or duck MAVS can initiate innate signaling in chicken cells and their co-transfection augments interferon-beta promoter activity. Demonstrating the limits of cross-species interactions, duck RIG-I 2CARD initiates MAVS signaling in chicken cells, but works poorly in human cells. The D122A mutation of human 2CARD abrogates signaling by affecting MAVS engagement, and the reciprocal A120D mutation in duck 2CARD improves signaling in human cells. We show mitochondrial localization of PB1-F2 from influenza A virus strain A/Puerto Rico/8/1934 (H1N1; PR8), and its co-localization and co-immunoprecipitation with duck MAVS. PB1-F2 inhibits interferon-beta promoter activity induced by overexpression of either duck RIG-I 2CARD, full-length duck RIG-I, or duck MAVS. Finally, we show that the effect of PB1-F2 on mitochondria abrogates TRIM25-mediated ubiquitination of RIG-I CARD in both human and avian cells, while an NS1 variant from the PR8 influenza virus strain does not.

Influenza A viruses are a major cause of human and animal disease. Periodically, avian influenza viruses from wild waterfowl, such as ducks, pass through intermediate agricultural hosts and emerge into the human population as zoonotic diseases with high mortality rates and epidemic potential.