Fig 1.
Workflow for homology modeling and molecular docking of the DORN1 L-type lectin domain.
The molecular modeling of DORN1 consists of two major prediction processes, (A) comparative modeling and (B) protein-ligand docking. Homology modeling of DORN1 involves fold assignment, template search, target-template alignment, model generation, and model evaluation and validation. DORN1-ligand docking includes energy minimization with molecular dynamic simulation, binding site prediction, and DORN1-ligand docking and analysis.
Table 1.
Search statistics of three selected legume lectin templates used for homology modeling of the DORN1 L-type lectin domain.
Fig 2.
Schematic representation and loop deletion information of DORN1 for site-directed mutagenesis binding assays.
(A) A schematic representation drawn into scale and sequence features of the DORN1 protein. (B) Detailed information of loop deletion mutations (white bars) on ATP binding activity of DORN1, WT (ECD): wild-type extracellular domain, Del-A: loop A deletion, Del-B: loop B deletion, Del-C: loop B deletion, Del-D: loop D deletion, Del-ExLp1: deletion of extended loop 1, Del-ExLp2: deletion of extended loop 2, WT (Lec): wild-type lectin domain.
Table 2.
List of DORN1 interacting residues with ATP obtained from prediction and target docking.
Fig 3.
Multiple sequence-structural alignment of the DORN1 L-type lectin domain and three selected templates.
Template sequence names were assigned by the Uniprot database, following by a 4-letter PDB code and chain name, separated by a dot. DBL-Dolichos biflorus (1BJQ), DLEC2-Phaseolus vulgaris (1FAT), SPL-Spatholobus parviflorus (3IPV), and DORN1-Arabidopsis thaliana (P2K1). Numbers at the beginning and the end of the alignment denote actual amino acid position in the protein sequences. Black bars-beta strands, blue bars-beta turns, dash lines-loops or coils, * predicted ATP binding residue, + conserved sugar binding residue inferred from legume lectins, red dots-Mn2+ binding residues, blue dots-Ca2+ binding residues, black dots-Mn2+ and Ca2+ binding residues. Columns with same color indicate identical amino acids or similar groups of amino acids, dashes are gap insertions. The black bars (beta strands) and blue bars (beta turns) below the alignment represent the secondary structure of the DORN1 L-type lectin domain inferred from the modeled DORN1 structure. Round dots with different colors below the aligned columns are cation binding residues inferred from the templates. Black stars denoted predicted ATP binding residues of the DORN1. Loops A, B, C and D and an extended loop inferred from the alignment with legume lectins.
Fig 4.
Ribbon style representations of the DORN1 models and superimposed models.
(A) and (B) Backbone alpha-carbon traces of the DORN1 model with 13 defined beta strands (β1 to β13), 4 defined loops (A-cyan, B-purple, C-dark blue, and D-red), 2 β-turns (t1 and t2 in blue), 1 extended loop (magenta), and other loops in front view and back view, respectively. (C) and (D) Backbone alpha-carbon traces of the superimposed models between DORN1 (tan) with the best template (green), and with other selected templates 1BJQ (purple) and 1FAT (cyan). Yellow bars next to the figures denote relative scale of the model in Angstrom unit (1Å = 10-10m).
Fig 5.
In vitro binding activities of ATP and DORN1 wild-types and deletion mutants.
(A) and (B) Graphical representation of binding activities of ATP and DORN1 wild-types and deletion mutants in vitro using a non-linear regression model (see “Materials and methods” in detail). WT (LEC): lectin domain only, WT (ECD): extracellular domain, Del-A: amino acid deletion in loop A, Del-B: deletion in loop B, Del-C: deletion in loop C, Del-D: deletion in loop D and Del-ExLp1 and Del-ExLp2: deletions in the extended loop. (C) and (D) Corresponding best-fit Kd, Bmax, and R2 values extrapolated from the non-linear regression models of A and B, respectively.
Fig 6.
Cartoon representation of the ATP binding site and interacting residues of DORN1.
(A) Close up of the DORN1 binding loops and ATP molecule in ball and stick mode, where numbered residues are those sharing H (hydrogen) bond with ATP. (B) Close up of electrostatic potential molecular surface of the DORN1 binding pocket and ATP molecule in ball and stick mode, where numbered residues are those interact with ATP via hydrogen bonds. (C) A detailed interaction map of the DORN1 and ATP docking complex, ATP resides in the middle of the map, and it is surrounded by interacting residues of 4 different loops (A to D), dashed-lines with numbers are hydrogen bonds with bond distances in angstrom sharing with atoms of ATP, where residues with a red crown denote hydrophobic interaction. Numbered residues represent their actual position of the DORN1 full-length sequence.