Fig. 2

Proposed model of the interaction between DKC1125 and KSRP. a DKC1125 interacts with KSRP via the third and fourth KH-domain. Schematic representation of KSRP-His deletion mutants are shown. The four gray boxes indicate the KH-domain of KSRP. Bacterially expressed proteins from His-tagged KSRP deletion constructs were purified with Ni-NTA resin and pulled down with a DKC1125 chemical probe. The interaction was examined by probing the blots with anti-His antibody. Positive bands were detected in KSRP full-length and the KH3 (residues 322–386) and KH4 (residues 420–491) domains, but not in the KH12 domain (residues 100–300) or the C-terminus of KSRP (residues 501–711). Arrows point to the right size of each protein. b Effects of wild-type KSRP and KSRP deletion mutants on cell motility and influence of treatment with DKC1125. Cell invasion was compared between Caco2 cells expressing wild-type (WT) KSRP-myc and those expressing KSRP deletion mutant lacking the KH34-domain (ΔKH34-KSRP-myc) after treatment with DKC1125 (0.5 μM). NS: not significant, *p < 0.05. c Predicted structure of KH34-domain and a possible binding pose of DKC1125. The sky-blue and pink regions indicate the KH3-domain (residue 324–394) and KH4-domain (residue 425–495), respectively. The random loop between the two domains is colored gray. Two predicted binding sites of the KH34-domain are represented in dark gray (site1; residue 395–418) and magenta (site2; residue 464–472). The spheres represent single point mutations of R411A, R415A, Q417A, and N467A. According to the computational simulation, DKC1125 (green) was put on the KH34-domain and totally covered by site2 of KH4-domain, especially hydrophobic P466 and F472. The main binding force was induced by strong hydrogen bonds (dashed line) between the ligand and protein. DKC1125 formed hydrogen bonds with R415 of the loop and S430 of the KH4-domain. The side chain aggregation of E343, K347, R411, and P407 could stabilize the ligand pocket of the KH4-domain. d Effects of KSRP deletion mutant on cell motility, and the influence of treatment of DKC1125. Cell invasion was compared between Caco2 cells expressing wild-type KSRP-myc and the KSRP deletion mutant within the KH34-domain after treatment with DKC1125 (0.5 μM). The amino acids within the putative binding pocket regions predicted to be essential to the binding of DKC1125 by the computer docking model were replaced with alanine or deleted, and the effects of overexpression of the mutants on cell invasion, with or without DKC1125 (0.5 μM), were tested and compared with those of wild-type KSRP. Data are expressed as in Fig. 1a. The increase in cell invasion by wild-type KSRP was restored by DKC1125, and the same results were observed in the Q417A, Δ395–418 (site1), and N467A mutant, but not in the R415A, Δ464–472 (site2), or R411A mutant