Moe program

We predicted secondary structure and disorder to investigate the characteristics of the protein structure of the intracellular third loop containing the Glu repeats. For the analysis of disorder prediction, we used DISOPRED2 [50 (link)], and for the analysis of the secondary structure, we predicted protein structure using PSI-PRED [51 (link)].
As we hypothesized that the Glu repeats forms a helix structure, we prepared a helical structure of the 8–12 residues in each poly-Glu repeats sequence by artificial measures. The three-dimensional structures of the amino acid variants from the 8–12 poly-Glu repeats were constructed using the MOE program with default parameters (Chemical Computing Group, Montreal, Canada). The structural models of the 8–12 poly-Glu repeats were docked into the X-ray structure of the Gi α-subunit from the co-crystal structure of the GPCR and G proteins (PDB-ID: 3sn6) [52 (link)] using the docking simulation program ClusPro (Boston University, Boston, MA) [53 (link)] with the “Electrostatic Favored” interaction energy score.

Three-dimensional structures of MG, LMG, CV, LCV, and their derivatives that were reported as haptens or analogues were built in Molecular Operation Environment (MOE) program (2012.10, Chemical Computing Group, Montreal, QC, Canada) [26 ]. The structures were optimized with a MMFF94 force field and a high convergence restriction of RMS gradient (0.001 kcal/mol). Then, the molecules were superimposed with atom-based RMS optimization, followed by flexible alignment with MOE’s Flexible Alignment module. The molecule alignment was performed by maximizing steric and feature overlap while minimizing internal ligand strain. During the alignment process, methods including rigid, flexible alignment and refining of existed alignment were used sequentially to obtain a reasonable overlap of the structures.

The crystallographic structure of DNA topoisomerase IV subunit B and thymidylate kinase (TMPK) were retrieved from the protein data bank at using 4HZ5 and 4QGG codes (Tari et al. 2013 (link); Kawatkar et al. 2014 (link)), respectively. Water molecules were removed, hydrogen atoms and partial charges were added, then a Gaussian contact surface around the binding sites was drawn using the MOE program (Molecular Operating Environment, Version 2019.01, Chemical Computing Group Inc., Montreal, Canada).

The Eribulin-tubulin co-crystal structure was retrieved from the PDB database (PDB ID: 5JH7). The docking of Microtubin-3 to the Eribulin binding site was performed using the Molecular Operating Environment (MOE) program (version 2009, Chemical Computing Group) as described previously [26 (link), 44 (link)]. To prepare the receptor for docking, the solvent and ions were computationally removed followed by protonation and tether minimization using the LigPrep protocol. Next, the Microtubin-3 and Eribulin were docked into the Eribulin site points using the triangle matching algorithm. The docked poses were first scored using the London dG scoring function, which estimated the free energy of binding from a given pose, followed by force field refinement and London dG rescoring. The top scoring docked poses of each molecule within the Eribulin site were retained.

A homology model of the PcCS tertiary structure was constructed with the MOE program (Chemical Computing Group Inc., Montreal, Canada), according to the manufacturer’s instructions and as previously described (Rupasinghe et al. 2003 (link)), using chicken CS (PDB code: 1AL6) as a template structure. The truncated amino acid sequence with a deletion of 19 amino acids of the N-terminal mitochondrial signal from the full-length sequence was applied to the homology model program. The terminal chemical structure of N-hydroxyamidocarboxymethyldethia coenzyme A in that model was replaced with that of AcCoA. After construction of the initial homology model, further energy minimization was performed using the Amber10 force field within the MOE distribution, until the final energy gradient became < 0.01 kcal/mol Å.

Unless otherwise noted, all reagents were purchased from commercial sources. Primers were purchased from Integrated DNA Technologies. Biapenem and tebipenem (>98% purity) were purchased from Sigma-Aldrich. Molecular graphics and analyses were performed with the UCSF Chimera package [26 (link)] and the Molecular Operating Environment (MOE) program (v 2014.09; Chemical Computing Group Inc., 1010 Sherbooke St. West, Suite #910, Montreal, QC, Canada, H3A 2R7, 2014).