TRAP expression in the bone marrow was also inhibited by treatment with Ya-s11

rect contrast to the K172R mutation of human RIG-I CARD domains that abrogates ubiquitination and activation, including activation by interaction with unanchored polyubiquitin. The role of anchored ubiquitin on K172 in the activation of human RIG-I has been questioned recently and it was suggested that only unanchored K63-linked polyubiquitin chains are important for RIG-I activation. We demonstrated that the attachment of ubiquitin at duck RIG-I K167 and K193 is not necessary for activation of the CARD domains, but we suggest that duck RIG-I is unlikely to have ubiquitination sites without function in the cell. Indeed, the function of attached ubiquitin at K172, or K167 and K193 remains elusive, but perhaps this function is required for conformational changes leading to the activation of the intact RIG-I. The ability to generate anchored ubiquitin could be evidence of the ancestral activity of TRIM25, while activation by unanchored polyubiquitin chains was later derived. In addition, TRIM25 can undergo autoubiquitinylation with polyubiquitin chains. Recent enzymatic PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19650037 studies suggest that TRIM25 generates the ubiquitin chains anchored to the E2 ligase and then they are transferred to the substrate, generating a plausible scenario for the coexistence of the two types of ubiquitin chains: some of the transferred chains will be attached to RIG-I and some will remain unanchored. Another possibility is that TRIM25 produces K63linked polyubiquitin chains attached to RIG-I and an unknown deubiquitinating enzyme produces the unanchored chains. Our data, with activation of duck RIG-I CARD domains in the absence of ubiquitin attachment sites makes the latter model unlikely. In an alignment of RIG-I CARD domain amino acid sequences, it is noteworthy that the lysines 1235481-90-9 equivalent to human K169 and K193 are highly conserved across species, while the lysine equivalent to human K172 is not present in birds or rodents. RIG-I from avian species, including ducks, geese and zebrafinch, all have the equivalent lysine K169 and K193, and have the motif KSLQ, in the RIG-I regulatory site equivalent to the human KT170LK172. Turtle has the conserved K169 and K193, and has the motif KTFH. Teleost species, such as zebrafish have the sequence KVLK, and the K193 is not conserved. Mouse and rat have K169 and the motif KVLQ. Among RIG-I sequences for all species to date, K169 is conserved in all, and K193 is missing only in wild boar and rat. Serine or threonine at the site of phosphorylation of human RIG-I is well conserved among most species, but is missing in rodents and bats. Clearly, the RIG-I regulatory motif has changed over evolutionary time, possibly due to selection pressure from viruses intent on disrupting this regulation. To suggest a model for duck RIG-I activation consistent with our observations, dephosphorylation at serine residues S8 and S168 allows RIG-I to interact with MAVS, where it undergoes monoubiquitination of lysine residues K167 or K193 in the RIG-I CARD domain by TRIM25, and attachment or association with polyubiquitin chains. Alternatively, additional ubiquitin moieties are attached to the monoubiquitin bound at either site. The activation of duck RIG-I in the absence of bound ubiquitin, which is in contrast to human RIG-I lacking K172, may reflect different specificity for duck RIG-I in binding unanchored polyubiquitin chains for activation. Indeed, the activation of duck RIG-I in the TRIM25 Activation of Duck RIG-I absence of attached pol