Assessment by Sequence Conservation and Computational Analyses of SIKE's Phosphorylation Sites to Quaternary Structure Stability

Abstract

Suppressor of IKKepsilon (SIKE), a protein first identified in the antiviral innate immune response associated with TANK‐binding kinase 1 (TBK1), is associated with multiple, distinct protein complexes including TBK1, STRIPAK (striatin interacting phosphatase and kinase) and cytoskeletal proteins, tubulin and actinin. Although SIKE's function is not fully defined in these complexes, SIKE does function as a high‐affinity substrate of TBK1 with phosphorylation occurring at up to six SIKE serine residues, S133, S185, S187, S188, S190, and S198. The accompanying figure shows a cartoon diagram of SIKE dimer model with the phosphorylation sites shown in spheres (red 187, 190; orange 133; 187 green; 188 and 190 cyan). Using a phosphomimetic mutant (S133/185/187/188/190/198E) of SIKE, size exclusion chromatography and chemical crosslinking studies showed primarily a monomeric species, whereas unmodified SIKE separated as a dimeric species. These observations suggested that SIKE could undergo a phosphorylation‐induced change in quaternary structure, but how many of the six phosphorylation sites were necessary? To explore this idea, conservation of the phosphorylation sites was explored. 133 SIKE sequences (Animalia, Fungi) from OrthoDB and Protein Blast were compiled and aligned using Clustal Omega. Positions 187 and 190 were conserved as serine in >90% of sequences whereas position 133 was ~60% serine and 25% aspartate or glutamate with a 9:10 bias towards aspartate. The remaining positions had 75% or less retaining serine with position 188 registering ~32% lysine. Using a model of the SIKE dimer where the phosphorylation sites line part of the dimer interface, phosphorylated S187/190 in each subunit would be positioned to repulse one another. To gain further insight into the role of individual phosphorylation sites on the theoretical dimer interface, phosphomimetic mutants at each phosphorylation site were computationally created (PyMOL), energy minimized (GalaxyRefineComplex) to yield 10 models, and the interface stability of each model for each mutant assessed on a per residue basis (HawkDock MM/GBSA). These mutant data were then compared to a similar analysis of the wild type SIKE model to identify residues with significant changes in total free energy of binding (dimer interface). S133E showed no significant changes from WT whereas S185/187/188/190/198E individually revealed a similar pattern of residues with altered total free energy of binding adjacent to the phosphorylation sites that differed in the number (19‐30) of residues affected in the region. The sequence conservation suggested that positions 187 and 190 could serve as key phosphorylation sites to trigger a dimer to monomer transition. Although the computational analyses has not resolved the effect of multiple phosphorylation events, the primary alteration to dimer stability is localized to the regions adjacent to the phosphorylation sites.