PMC

The three-dimensional structure of TmCBM61. The overall fold of TmCBM61 is shown from the side (A) and top (B) as a color-ramped ribbon. The tryptophan residues comprising the binding site are shown as sticks and labeled. A calcium binding site is also shown, with the metal shown as a blue sphere. C, the non-crystallographic dimer of TmCBM61 in complex with β-1,4-galactotriose determined at 1.4 Å resolution reveals the general location of the carbohydrate binding site. The carbohydrate is shown as green sticks, and the binding site tryptophans are shown as gray sticks.

Recognition of a β-1,4-galactan helix. A, construction of a β-1,4-galactooligosaccharide using the Φ and Ψ angles observed for the oligosaccharide in the TmCBM61 structures. B, an idealized β-1,4-galactooligosaccharide using previously reported Φ and Ψ angles. Relevant dimensions of the helices in A and B are indicated. C and D, the β-1,4-galactan helix in A modeled into the active site of TmCBM61, which was based on the position of the β-1,4-galactooliosaccharide position determined in the crystal structures. E and F, the idealized β-1,4-galactan helix in B modeled into the active site of TmCBM61. In panels C–F, the β-1,4-galactooligosaccharide representing the sugars observed in the crystal structures is shown as green sticks. Relevant resides in the TmCBM61 active site are shown as sticks and labeled. In D and F, the protein is shown as a solvent-accessible surface.

Similarities between starch recognition and β-1,4-galactan recognition. A, a top view of the TmCBM61 binding site (tryptophan residues shown as yellow sticks and the protein backbone shown in gray ribbon) showing a bound β-1,4-galactotetraose molecule (green sticks) created based on the overlay shown in D. Overlaid with this is the aromatic platform (blue sticks) from the maltooligosaccharide binding protein MalX from S. pneumoniae with the maltoheptaose molecule shown as purple sticks (). B, shows a 90° rotation of the view in A and reveals the similar dimensions of the amylose helix and the galactan helix. MalX was chosen as a representative α-1,4-glucan binding protein because it was crystallized with the largest maltooligosaccharide ligand to date, revealing the helical nature of amylose fragments; it also represents the generally conserved manner of maltooligosaccharide binding by most non-catalytic proteins with this specificity.

Details of the TmCBM61 binding site revealed by ligand complexes. A, the binding site of the β-1,4-galactotriose complex of TmCBM61 determined at 0.95 Å resolution. The electron density (blue mesh) for the sugar is shown as a maximum likelihood ()/σ_a-weighted () 2F_o − F_c map (contoured at 2σ; 1.24 electrons/ų). The binding sites for monomer 1 (B) and monomer 2 (C) of the β-1,4-galactotriose complex of TmCBM61 were determined at 1.4 Å resolution. The electron densities (blue mesh) for the sugars in B and C are shown as maximum likelihood ()/σ_a-weighted () 2F_o − F_c maps (contoured at 1σ; 0.45 electrons/ų). D, an overlay of the sugars in TmCBM61 binding sites of the monomer in the 0.95 Å resolution structure (green) and monomer 1 of the 1.4 Å resolution structure (blue). The protein is shown as a solvent-accessible surface with the relevant tryptophan residues colored purple. In all panels, the sugars and relevant binding residues are shown as sticks. The subsites occupied by the sugar residues are numbered in red. Hydrogen bonds are shown as dashed lines, with their lengths indicated in Å. Relevant water molecules are shown as red spheres.

Labeling of A. thaliana stem with TmCBM61 and LM5. A–C, low magnification images showing labeling of whole stem cross-sections with TmCBM61 (A) and LM5 (B). C, control image produced by labeling with anti-His/FITC-conjugated secondary antibody only. D–I, high magnification images showing labeling of vascular tissues. Sections were probed with TmCBM61 (D) or LM5 (E) without any prior pretreatment or following pretreatment of sections with GH53 prior to labeling with TmCBM61 (F) or LM5 (G). Control sections probed with anti-His/FITC or anti-rat/FITC only (H and I, respectively) are also shown. J–M, high magnification images showing TmCBM61 labeling of different tissues. J, protoxylem; K, pith parenchyma; L, interfasicular tissue; M, epidermis with stomatal opening (s). Note that TmCBM61 labeling appeared as two parallel lines (double arrowheads in J and K) but not to intercellular spaces (is). The positions of non-labeled portions of stomatal cell walls are indicated by asterisks in M. c, cortical parenchyma. ph, phloem; x, xylem; if, interfasicular tissue; p, pith parenchyma. Scale bars, 250 μm (A–C) or 25 μm (D–M).

Comparative glycan microarray analysis of the binding profiles of TmCBM61 and LM5. Microarrays of oligosaccharides and polysaccharides derived from plant cell walls were probed with TmCBM61 (A) or LM5 (B). Control arrays were probed with anti-His/alkaline phosphatase-conjugated secondary antibody only (C). As shown by the boxed region in A, each sample was represented on the arrays by four spots (two concentrations, 1 and 0.2 mg/ml, and in duplicate). A further set of arrays was treated with GH53 prior to probing with TmCBM61 (D) or LM5 (E). Control arrays were treated with GH53 prior to probing with anti-His/alkaline phosphate-conjugated secondary antibody only (F). G–J, heat maps showing quantification of arrays. The data shown are mean, background-adjusted spot signals from three independent probing experiments. The data shown in G and H are for arrays probed with TmCBM61 and LM5, respectively, without any pretreatment. The data shown in I and J are from arrays probed with TmCBM61 and LM5, respectively, pretreated with GH53 prior to probing. For each probe, the maximal value in the whole data set (i.e. G and I for TmCBM61, and H and J for LM5) was set to 100, and all other values were adjusted accordingly. K, sample list with array coordinates for the samples that produced mean background-adjusted signals of >0. A full sample list is provided in supplemental Table 2.