Fig. 4: A phospho-dependent foldback loop in Axin LIR.
a Top, cartoon of Axin1 (as in Fig. 1a), with sequence of LIR underneath whose evolutionary conservation is indicated by JackHMMER plot; below, predicted GSK3 and CK1 phospho-acceptor sites mutated in S4 > A (green); S325, priming residue for CK1; green, orange, mutated S1 and S2 triplet residues; bold, other mutated residues (including S3 and S4 doublets). b, c CoIP assays as in Fig. 1, following co-expression of wt or mutant FLAG-Axin1, LRP6-GFP and GSK3β-HA, as specified in panels; colors as in (a); numbers underneath blots indicate mean values relative to wt (= 1.00) as obtained by densitometry of IP blots normalized to corresponding input blots, each showing similar results across eight independent experiments; left, positions of molecular weight markers (in kDa). An enhanced coIP signal for the S1 > A, S2 > A and S1 > D mutants was consistently observed compared to their wt Axin control. d, e Structural predictions of Axin LIR by AF2 or AF3 as indicated, with conserved residues colored as in (a) in stick representations (above), revealing striking clustering of negative (red) and positive (blue) charges in surface representation (below); note also kink induced by tandem prolines (P344, P345). f ITC profiles of 50 μM unphosphorylated (43mer) or phosphorylated (43PPP) Axin LIRup peptides titrated with 500 μM wt or RKR mutant Axin Lip-LIRα, revealing phospho-dependent interaction between phosphorylated LIRup and positively-charged LIRα; in each panel, the KD value was calculated from three independent experiments, and its standard deviation is given. Note that the LIRup peptides include 7 residues of LIRα (to render them soluble in aqueous buffer) which may allow residual intramolecular folding back of the carboxy terminus in the case of 43PPP, in which case the calculated KD value for 43PPP may represent an underestimate. g Cartoon depicting intramolecular interaction within Axin LIR (‘foldback loop’) which depends on phosphorylation of Axin LIRup.
