LIG_IRFs_LxIS_1

Sensing of pathogenic microbes and tissue damage by the innate immune system triggers immune cells to secrete cytokines that promote host defence. Type I Interferons (IFN), including IFN-α and IFN-β, are major cytokines secreted when immune cells sense pathogen-associated molecular patterns (PAMPs) through pattern-recognition receptors (PRRs) (Stetson,2006). Cytosolic PRRs include RIG-I-like receptors, which detect viral dsRNA and are then recruited to the mitochondrial surface where they associate with the mitochondrial antiviral-signalling protein (MAVS). Another cytosolic PRR is the cyclic-GMP-AMP synthase (cGAS) which, in response to the presence of viral DNA, catalyses the production of cGAMP. This second messenger leads to the activation of the endoplasmic reticulum membrane protein stimulator of IFN genes (STING) (Sun,2013). Toll-like receptors (TLRs) are membrane-bound PRRs that recognize different types of bacterial, viral or fungal PAMPs. For example, TLR3 responds to viral dsRNA, while TLR4 recognizes lipopolysaccharides from gram-negative cell bacteria (Akira,2006). TLRs recruit adaptor proteins such as the TIR-domain-containing adaptor protein (TRIF, also called TICAM1) to further induce the interferon response pathway. The phosphorylation of the adaptor proteins MAVS, cGAS and STING is essential for downstream signalling. The three proteins share a conserved linear motif referred to as pLxIS that is phosphorylated by TANK-binding kinase 1 (TBK1) and/or IκB kinase-ε (IKKε) (Liu,2015). Once phosphorylated and helped by other electrostatic interactions and a semi-conserved hydrophobic residue 1 to 3 positions upstream, the motif binds to positively charged surfaces of IRF-3, IRF-5, IRF-7 (Lazear,2013) and possibly some other Interferon Regulatory Factor paralogues (IRFs).

The IRF family consists of nine transcription factors (IRF1-9) with different defence roles (Savitsky,2010). Among them, IRF-3 and IRF-7 are engaged in innate immunity against bacteria and viruses as the key IRFs for IFN production (Wu,2014). While they are similar in sequence and structure and tend to form heterodimers (Schmid,2014), IRF-3 is expressed ubiquitously while IRF-7 is generally present at low concentrations (Au,1998) but responds to a feedforward loop induced by IFNs (Marie,1998). IRF-5 and IRF-6 are more structurally related, with the former linked to pro-inflammatory response and the regulation of apoptosis, and the latter involved in the regulation of cell proliferation and the differentiation of keratinocytes (Taniguchi,2001).

Regardless of their heterogeneity, the signal response (or serine-rich) domain (SRD) adjacent to the C-terminal IRF-Associated Domain (IAD) of IRF-3, IRF-5, IRF-6 and IRF-7 (PF10401) carry variants of the same pLxIS motif that can also be phosphorylated by TBK1/IKKε. Phosphorylation of these IRFs promote their dissociation from the adaptor proteins and subsequent homo- or hetero-dimerization through the pLxIS motifs, which bind the same phosphopeptide-binding domain that is then blocked for further interaction with adaptor proteins and enabling transition into the nucleus (Zhao,2016). For example, the C-terminal region of IRF-3 undergoes opening following phosphorylation, with the subsequent exposure of a binding interface for CREB binding protein (CBP) or p300 (7JFL; Qin,2005) allowing activation of innate immunity gene expression. Likewise, IRF-5 is activated by phosphorylation by IKKβ at conserved residues of the C-terminal domain. This also releases autoinhibition by a conformational rearrangement, followed by translocation of IRF-5 to the nucleus and interaction with CBP (Ren,2014). The crystallization of IRF-5 with phosphomimetic mutations reveals an elongated structure in the C-terminal region (3DSH). This conformation promotes dimerization and positions the binding interfaces to interact with the CBP/p300 dimers. However, the affinity is 1-2 orders of magnitude lower compared to IRF-3, which is expected given the relatively short length of the peptide assayed, but could nevertheless indicate a finely tuned switch between autoinhibition and dimerization in IRF-5. This difference in affinity may highlight distinct functional requirements: IRF-3 is expressed constitutively to offer a highly sensitive response to viral infection, while IRF5 functions as a more stable dimer with greater basal activity (Chen,2008). Structurally similar to IRF-5 and with a very similar pLxIS motif variant, IRF-6 has a different functional role which is not conserved in vertebrates. In mammals, IRF-6 has a developmental role not related with IFN induction (Kawasaki,2016). But it has been suggested that IRF-6 is a positive modulator of IFN expression in zebrafish, including the activation of the transcription of MAVS, RIG-I and other host genes (15939375). The pLxIS motif is needed for IRF-6 dimerization, a key step for its activation and nuclear translocation, as demonstrated by coimmunoprecipitation of phosphomimetic double serine mutants (Kwa,2014). The motif in IRF-7 is a variant, pLxLS, with Leu for the usual Ile (Lin,2000).

It is not documented that other IRF family members interact with the pLxIS motif. The IAD domain that is conserved in IRF-1 and IRF-2 differs from that of other IRFs (Yanai,2012). IRF-4 and IRF-8 have a central role in controlling the development of dendritic cells and thus are prone to interact with other partners (PMD:30984161). IRF-9 does not show autoinhibition, and the structural evidence of IRF-9 in complex with STAT2 (5OEN) shows a specific IAD interaction interface that does not involve a linear motif (Rengachari,2018).

Interestingly, the rotavirus non-structural protein 1 (NSP1) also contains the pLxIS motif which binds to the same phosphopeptide-binding region of IRFs. The interaction prevents the activation of IRF-3, IRF-5 or IRF-7 by phosphorylation, promoting their degradation instead. Phosphorylation of the Ser is not necessary for a low affinity binding (e.g. ~200 μM with IRF-3), but favours a more competitive affinity with respect to the host's other adaptor proteins (16 μM for IRF-3, against 43 to 104 μM measured for cellular adaptor proteins) (Zhao,2016). Upon binding, NSP1 can behave as an E3 ubiquitin ligase, directing IRFs with IAD regions into the ubiquitin proteasome system (Arnold,2013). In this way, rotavirus is able to escape innate immune recognition by interfering with the IRF-dependent pathways (Barro,2005). This raises the possibility that other pathogens may mimic the pLxIS motif to disrupt innate immunity.