Structure of the S100A6 complex with a fragment from the C-terminal domain of Siah-1 interacting protein: a novel mode for S100 protein target recognition

Abstract

S100A6 is a member of the S100 subfamily of EF-hand Ca (2+) binding proteins that has been shown to interact with calcyclin binding protein/Siah-1 interacting protein (CacyBP/SIP or SIP), a subunit of an SCF-like E3 ubiquitin ligase complex (SCF-TBL1) formed under genotoxic stress. SIP serves as a scaffold in this complex, linking the E2-recruiting module Siah-1 to the substrate-recruiting module Skp1-TBL1. A cell-based functional assay suggests that S100A6 modulates the activity of SCF-TBL1. The results from the cell-based experiments could be enhanced if it were possible to selectively inhibit S100A6-SIP interactions without perturbing any other functions of the two proteins. To this end, the structure of the S100A6-SIP complex was determined in solution by NMR and the strength of the interaction was characterized by isothermal titration calorimetry. In an initial step, the minimal S100A6 binding region in SIP was mapped to a 31-residue fragment (Ser189-Arg219) in the C-terminal domain. The structure of the S100A6-SIP(189-219) complex revealed that SIP(189-219) forms two helices, the first of which (Met193-Tyr200) interacts with S100A6 in a canonical binding mode. The second helix (Met207-Val216) lies over the S100A6 dimer interface, a mode of binding to S100A6 that has not previously been observed for any target bound to an S100 protein. A series of structure-based SIP mutations showed reduced S100A6 binding affinity, setting the stage for direct functional analysis of S100A6-SIP interactions.

Figure 1

S100A6 inhibits down-regulation of β-catenin by SCF-TBL1. HEp-2 cells with siRNA of S100A6 (S100A6−) and of non-specific siRNA (S100A6+) were transfected with empty vector or pSiah-1-FLAG. Since Siah-1 is highly auto-ubiquitinated, a proteasome inhibitor, MG-132, was added for detection of overexpressed Siah-1.

Figure 2

Mapping of S100A6 binding region in SIP. a) Binding affinity was measured using isothermal calorimetry (ITC) for full length SIP, SIP-SGS, H-A and H-B. b) ITC of SIP(189–219) with S100A6. 400 µM SIP( 189–219) was titrated into 40 µM S100A6.

Figure 3

NMR chemical shift perturbation assay of the effect of SIP(189–219) binding to S100A6. a) ¹⁵N-¹H HSQC spectra of S100A6 acquired in absence (black) and presence (red) of SIP(189–219). The data were collected at 30 °C on a sample containing 0.1 mM S100A6 and 0.5 mM SIP(189–219) in 50 mM Tris, pH 6.5, 10 mM Ca²⁺ in 95% H₂O/5% D₂O. b) Changes in chemical shifts in a) are plotted against the protein sequence. Blue lines are drawn for the four helices of S100A6. Red squares indicate S100A6 peaks that disappear upon the addition of SIP(189–219).

Figure 4

Stoichiometry of S100A6-SIP(189–219) complex. a) Chemical cross-linking detected by SDS-PAGE. 0.1 mM S100A6 and SIP(189–219) were chemically cross-linked using BS³ and aliquots were taken over a period of 1-30 min. Lanes at right show the result of cross-linking isolated S100A6 and SIP(189–219) for 30 min. b) Electrospray mass spectroscopy. 0.1 mM S100A6 and 0.15 mM SIP(189–219) were mixed in 25 mM ammonium acetate and 1 mM calcium acetate. The deconvoluted mass of the complex of 2 S100A6 and 2 SIP(189–219) molecules bound to 4 Ca²⁺ is calculated from the +8, +10 and +12 charge states to be 28,081 ±1 Da, which is consistent with the measured masses for each molecule.

Figure 5

Structure of Ca²⁺-loaded S100A6 with SIP(189–219). The two S100A6 subunits are colored in blue and green (a–c). The two SIP(189–219) molecules are colored in red and orange (a–c). a) Stereo view of the backbone atoms in the final ensemble of the 20 representative conformers. b) Ribbon diagram of the representative structure. Helices are marked by H-I∼IV for one S100A6 subunit and H-I’~IV’ for the other S100A6 subunit. H-A and H-B indicates the two helices of the SIP(189–219). c) Surface model of S100A6 bound to SIP(189–219) in ribbon diagram.

Figure 6

Analysis of contacts between S100A6 and SIP(189–219). Residues from the SIP peptide are linked by black lines with the two helices in rectangles. Blue lines represent hydrophobic contacts with a cutoff of 4.5 Å. Red lines represent electrostatic interactions such as hydrogen bonds or salt bridges. S100A6 residues marked by a prime represent the other subunit of S100A6. Each contact is labeled with the number of occurrences in the ensemble of 20 conformers.

Figure 7

SIP Leu196 is a key residue for S100A6 binding. The binding interfaces around SIP Leu196 and Met207 are shown in a) ribbon diagram and b) surface representation. Hydrophobic residues in b) are colored in yellow and pale yellow for strong (Leu, Ile, Val, Phe) and weak (His, Ala) hydrophobic residues. c) Isothermal calorimetry studies of S100A6 and SIP mutants.

Figure 8

Comparison of S100 protein-target structures. Target peptides are highlighted in red. The structures shown are: S100A10-Annexin II(1–10), PDB code: 1BT6; S100B-p53(367–388), 1DT7; S100B-Ndr kinase(62–87), 1PSB; S100B(human)-TRTK12, 1MQ1; S100B(rat)-TRTK12, 1MWN; S100A6-SIP(189–219), 2JTT.