Virtual Screening, Lead Optimization And Biological Activity Study Of Novel LpxC Inhibitors As Potential Antibacterial Drugs

Having experienced the periods of random discovery and development, the drug discovery process has stepped into the rational drug design stage nowadays. With years of development, Computer Aided Drug Design (CADD) methods now play a key role in aiding rational drug design for the discovery of lead compound and its optimization. Achieving so many successful precedents in recent years, CADD receives more and more attentions and starts to be widely adopted by most medicinal chemistry researchers. One of its branches, the computer based virtual screening, a computational technique used in drug discovery area frequently, can deal with large compound libraries structure searching in a short time for efficiently identifying potential structures which are most likely to bind to a drug target, such as a protein receptor or enzyme. In fact, the application of virtual screening technology finally enables a panel of famous pharmaceutical companies and research institutions to save lots of time and screening cost, greatly speeding up the pace of new drug research and development in a desirable hit rate.To date, a myriad of Gram-negative bacterial infection cases have led to significant human morbidity and mortality. Since the available of therapeutic effective antibiotics dwindles and the more common appearance of multi-drug-resistant bacteria, strategies for confronting Gram-negative bacterial diseases such as identification of novel antibacterial targets, new lead compounds and handy treatment are always urgently needed. Among various enzymes involved in bacterial growth, metabolism and virulence related processes, UDP-3-O-(R-3-hydroxymyristol)-N-acetylglucosamine deacetylase (LpxC), a zinc-dependent hydrolase residing in the second step, or the so called first committed step of lipid A’s biosynthetic pathway has recently risen to an extremely attractive antibacterial target owing to its indispensible role in controlling lipid A and the follow-up lipopolysaccharide (LPS) and bacterial outer membrane’s formation. The effective inhibition of LpxC has turned out to not only obviously reduce bacterial survival rate and toxicity but to prevent the formal bacterial outer membrane production which serves as a formidable permeability barrier to multiple mechanisms based antibiotics. By gene sequence alignment, high LpxC homology is observed among different types of Gram-negative bacteria while no homology is found to any known animal and human’s protein. In short, facing the more and more commonplace of drug-resistant bacteria emerged at the present and the fact that no clinical drug is designed to aim at LpxC enzyme previously, potent inhibitors targeting this prospective antibacterial target are highly desired for high effective, low toxic, and broad spectrum antibiotics development to complement current antibacterial agents in the near future.In this paper, a variety of CADD approaches were employed towards LpxC inhibitor lead compounds’discovery and structural modification. Firstly, a virtueal screening of SPECS commodity compound library, Comprehensive Medicinal Chemistry (CMC) library as well as our lab’s previously synthesized compounds for other drug targets was performed to discover novel lead structures. The SPECS database was hierarchically in silico filtered by "Rule of Lipinski’s Five", solubility restriction, molecular docking, structure-based pharmacophore and molecular structure cluster analysis successively. Finally,56SPECS compounds were picked up and purchased for further Minimum Inhibitory Concentration (MIC) test, together with46available small molecules in our laboratory. Afterwards, in vitro antibacterial activity experiment of these102screened compounds was conducted, and their MICs were determined according to the NCCLS protocol adapted to96-well plates and LB media using double dilution method. At last,3molecules with MIC value of64μg/ml and8molecules with MIC value of128μg/ml against Pseudomonsa aeruginosa while2molecules with MIC value of64μg/ml,4molecules with MIC value of128μg/ml and1molecule with MIC value of256μg/ml against Escherichia coli were found. Although none of these obtained compounds holds a comparable MIC value to clinical drug levofloxacin, gentamicin or cefotaxime sodium, some of their MICs are still within some other drugs’CLSI recommended bacterial sensitive MIC standard range. Therefore, these identified compounds might be promising hits for our further investigations of LpxC structural modification and chemical synthesis. Besides, through molecular docking and binding pattern analysis,76drug structures from CMC library with untested MIC values also caught our attention and provided us useful assistance for future potent LpxC inhibitor design and synthesis job.In addition to LpxC inhibitors’virtual screening, in this thesis, we tried to perform some structural optimization research for previously reported typical LpxC lead compounds as well. A combination application of diversified rational drug design strategies such as ligand common feature based pharmacophore model, ChemMapper scaffold hopping, Topomer Search leading compound hopping, and de novo drug design was carried out for LpxC inhibitors’structural modification. Once these new fragments and scaffolds were designed out, they are reasonably pieced together and then validated by molecular docking and specific binding mode analysis. Notably, apart from the routinely modification of the hydrophobic domain and zinc ion binding groups, we also would like to reinforce the ligand-protein interaction by focusing fragment based drug design targeting LpxC’s UDP binding pocket, which has little report previously. In summary, both virtual screening result and lead compound optimization study will facilitate our LpxC inhibitor development, which will promote novel potent antibacterial drug discovery in future.