In silico Screening, Docking, and Redesigning of Traditional Chinese Medicinal Compounds Against Streptococcus pneumoniae Glycosyl Hydrolase GHIP and Peptidoglycan Hydrolase LytB

Streptococcus pneumoniae remains a major public health concern as it is the most common causative agent of community-acquired pneumonia. GHIP and LytB, hydrolases that function in adhesion and colonization, are promising targets for drug design to combat this pathogen. The enzymes were modeled using SWISS-MODEL to refine disordered loop regions. 6,699 curated molecules from the Traditional Chinese Medicine Systems Pharmacology (TCMSP) database were docked with the two enzymes separately. The top-binding ligands for GHIP were MOL011124.1, MOL003432.1, and MOL006638.1 with binding free energies (ΔG) of -7.7, -7.7, and -7.6 kcal/mol, respectively. For LytB, the two molecules with highest binding affinity were MOL011273.1 and MOL008461.1, with ΔG of -8.5 and -8.3 kcal/mol, respectively. All top-binding TCMSP compounds for each enzyme target satisfied Lipinski’s rules. They were then redesigned to increase binding affinity. The two redesigned ligands with highest binding affinity to GHIP are MOL006638x8571_5 and MOL006638x8571_1, with ΔG of -10.4 and -10.2 kcal/mol, respectively. For LytB, the two redesigned ligands with highest binding affinity were MOL001810x008596_1 and MOL001810x008596_2 with ΔG of -12.4 and -10.7 kcal/mol, respectively. The top-binding redesigned ligands for each enzyme had moderate to low solubility and only one ligand (for LytB) satisfied Lipinski’s rules. Nonetheless, these ligands fall under the extended oral druggable space and may be further developed as intravenous drugs or emulsions.