Figure 1.
Overall structures of binary and ternary complexes of ERGIC-53–CRD.
(A) Schematic representation of Glc1Man9GlcNAc2 showing the nomenclature of oligosaccharide residues and branches. Glucose residue is shown in red. Ribbon models of binary and ternary complexes of ERGIC-53–CRD are shown in (B) and (C), respectively. Ca2+-free ERGIC-53–CRD and Ca2+/α2-Man3-bound ERGIC-53–CRD are colored pink and cyan, respectively. MCFD2 is shown in wheat. The bound α2-Man3 and Ca2+ ions are shown as green stick and magenta sphere models, respectively. In the trimannosyl ligand, the two mannose residues giving unambiguous electron densities were drawn. Ca2+-binding loops (residues 155–161 and 176–185) are highlighted in orange in (C).
Table 1.
Data collection and refinement statistics for binary and ternary complexes of ERGIC-53–CRD.
Figure 2.
Sugar-binding site of ERGIC-53.
Omit Fo–Fc electron density map of α2-Man2 contoured at 2.0 σ. Additional electron densities flank the Manα1,2-Man moiety occupying the primary binding sites (sites 2 and 3), suggesting two alternative binding modes: (A) mode I and (B) mode II. The mannose residue occupying site 1 was modeled by superimposing the reducing-terminal mannose residue of the α2-Man2 ligand in the VIP36 crystal structure (PDB code: 2DUR) [30], which has a highest resolution (1.65 Å) among the L-type lectin-sugar complexes so far reported, while the modeling of the mannose residue accommodated in site 4 was based on the crystal structure of the non-reducing-terminal mannose residue of α2-Man2 bound in ERGIC-53 (PDB code: 4GKX) [31].
Figure 3.
Comparison of sugar-binding sites of ERGIC-53 and VIP36.
(A) Structure of the central α2-Man2-binding pocket (sites 2 and 3) of ERGIC-53-CRD (cyan, this study). (B) Structure of α2-Man2-binding site (3 and 4) of ERGIC-53-CRD (slate, PDB code: 4GKX) [31]. (C) Superposition between the two complexes of ERGIC-53–CRD. Assignments of the moieties of Man9GlcNAc2 are indicated in red characters. (D) VIP36–CRD (orange) complexed with α2-Man2 corresponding to Man(D1)-Man(C) (PDB code: 2DUR) [30]. (E) VIP36–CRD complexed with Man-α1,2-Man-α1,3-Man corresponding to Man(C)-Man(4)-Man(3) (PDB code: 2E6V) [30]. (F) Superposition between the Man(D1)-Man(C) bound complexes formed with ERGIC-53–CRD (cyan) and VIP36–CRD (orange). The variable residues among L-type lectins are indicated by boxes.
Figure 4.
Structural models of L-type lectins with monoglucosylated high-mannose-type oligosaccharides.
Surface models of (A) ERGIC-53–CRD/MCFD2 and (B) VIP36–CRD are shown with the glucose, mannose, and N-acetylglucosamine residues displayed in magenta, green, and slate stick models, respectively. To model Glc1Man8GlcNAc2, we selected Glc-α1,3–Man coordinates from an insect arylphorin glycoprotein [PDB code: 3GWJ (molecule D)] based on the torsion angle energy estimated using the PDB-CARE program [40]. Other typical Glc-α1,3–Man structures (PDB codes: 3GWJ (molecule A), 3O0W, 3OGV, and 3OG2) with energetically acceptable torsion angles were also superimposed on the ERGIC-53 (A, bottom) and VIP36 (B, bottom) complexes. These coordinates are indicated by thin stick models (glucose: red, mannose: green). The nomenclature of oligosaccharide residues of Glc1Man8GlcNAc2 are shown as in Figure 1A.