Analysis of bacterial Mur amide ligase enzymes for the identification of inhibitory compounds by in silico methods
- Chamboko, Chiratidzo Respina
- Authors: Chamboko, Chiratidzo Respina
- Date: 2020
- Subjects: Pathogenic microorganisms -- Analysis , Drug resistance in microorganisms , Microorganisms -- Effect of drugs on , Antibiotics -- Effectiveness , Pathogenic bacteria , Drug tolerance , Enzymes -- Analysis , Peptide antibiotics
- Language: English
- Type: text , Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10962/161911 , vital:40690
- Description: An increased emergence of resistant pathogenic bacterial strains over the years has resulted in many people dying of untreatable infections. This has become one of the most critical global public health problems, as resistant strains are complicating treatment of infectious diseases, increasing human morbidity, mortality, and health care costs. A very limited amount of effective antibiotics is currently available, but the development of novel classes of antibacterial agents is becoming a priority. Mur amide ligases are enzymes that have been identified as potentially good targets for antibiotics, as they are uniquely found in bacteria. They are responsible for the formation of peptide bonds in a growing peptidoglycan structure for bacterial cell walls. The current work presented here focused on characterizing these Mur amide ligase enzymes and obtaining inhibitory compounds that could potentially be of use in drug discovery of antibacterial agents. To do this, multiple sequence alignment, motif analysis and phylogenetic tree constructions were carried out, followed by docking studies and molecular dynamic simulations. Prior to docking, homology modelling of missing residues in the MurF structure (PDB 1GG4) was performed. Characterization results revealed the Mur amide ligase enzymes contained defined conservation in limited regions, that ultimately mapped towards the central domain responsible for ATP binding (presence of a conserved GKT motif). Further analysis of results further unraveled the unique patterns observed within each group of the family of enzymes. As a result of these findings, docking studies were carried out on each Mur amide ligase structure. At most, two ligands were identified to be sufficiently inhibiting each Mur amide ligase. The ligands obtained were SANC00574 and SANC00575 for MurC, SANC00290 and SANC00438 for MurD, SANC00290 and SANC00525 for MurE and SANC00290 and SANC00434 for MurF. The two best ligands identified for each enzyme had docked in the active site of their respective proteins, passed Lipinski’s rule of five and had substantially low binding energies. Molecular dynamic simulations were then performed to analyze the behavior of the proteins and protein-ligand complexes, to confirm the lead compounds as good inhibitors of the Mur amide ligases. In the case of MurC, MurD and MurE complexes, the identified ligands clearly impacted the behavior of the protein, as the ligand bound proteins became more compact and stable, while flexibility decreased. There was however an opposite effect on MurF complexes, that resulted in identified inhibitors being discarded. As a potential next step, in vivo and in vitro experiments can be performed with identified ligands from this research, to further support the information presented.
- Full Text:
- Authors: Chamboko, Chiratidzo Respina
- Date: 2020
- Subjects: Pathogenic microorganisms -- Analysis , Drug resistance in microorganisms , Microorganisms -- Effect of drugs on , Antibiotics -- Effectiveness , Pathogenic bacteria , Drug tolerance , Enzymes -- Analysis , Peptide antibiotics
- Language: English
- Type: text , Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10962/161911 , vital:40690
- Description: An increased emergence of resistant pathogenic bacterial strains over the years has resulted in many people dying of untreatable infections. This has become one of the most critical global public health problems, as resistant strains are complicating treatment of infectious diseases, increasing human morbidity, mortality, and health care costs. A very limited amount of effective antibiotics is currently available, but the development of novel classes of antibacterial agents is becoming a priority. Mur amide ligases are enzymes that have been identified as potentially good targets for antibiotics, as they are uniquely found in bacteria. They are responsible for the formation of peptide bonds in a growing peptidoglycan structure for bacterial cell walls. The current work presented here focused on characterizing these Mur amide ligase enzymes and obtaining inhibitory compounds that could potentially be of use in drug discovery of antibacterial agents. To do this, multiple sequence alignment, motif analysis and phylogenetic tree constructions were carried out, followed by docking studies and molecular dynamic simulations. Prior to docking, homology modelling of missing residues in the MurF structure (PDB 1GG4) was performed. Characterization results revealed the Mur amide ligase enzymes contained defined conservation in limited regions, that ultimately mapped towards the central domain responsible for ATP binding (presence of a conserved GKT motif). Further analysis of results further unraveled the unique patterns observed within each group of the family of enzymes. As a result of these findings, docking studies were carried out on each Mur amide ligase structure. At most, two ligands were identified to be sufficiently inhibiting each Mur amide ligase. The ligands obtained were SANC00574 and SANC00575 for MurC, SANC00290 and SANC00438 for MurD, SANC00290 and SANC00525 for MurE and SANC00290 and SANC00434 for MurF. The two best ligands identified for each enzyme had docked in the active site of their respective proteins, passed Lipinski’s rule of five and had substantially low binding energies. Molecular dynamic simulations were then performed to analyze the behavior of the proteins and protein-ligand complexes, to confirm the lead compounds as good inhibitors of the Mur amide ligases. In the case of MurC, MurD and MurE complexes, the identified ligands clearly impacted the behavior of the protein, as the ligand bound proteins became more compact and stable, while flexibility decreased. There was however an opposite effect on MurF complexes, that resulted in identified inhibitors being discarded. As a potential next step, in vivo and in vitro experiments can be performed with identified ligands from this research, to further support the information presented.
- Full Text:
Generation of a virtual library of terpenes using graph theory, and its application in exploration of the mechanisms of terpene biosynthesis
- Authors: Dendera, Washington
- Date: 2020
- Subjects: Terpenes , Plants -- Metabolism , Computational biology , Bioinformatics , Organic compounds -- Synthesis , Monoterpenes , Molecular biology -- Computer simulation
- Language: English
- Type: text , Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10962/123453 , vital:35439
- Description: Terpenes form a large group of organic compounds which have proven to be of use to many living organisms being used by plants for metabolism (Pichersky and Gershenzon, 1934; McGarvey and Croteau, 1995; Gershenzon and Dudareva, 2007), defence or as a means to attract pollinators and also used by humans in medical, pharmaceutical and food industry (Bicas, Dionísio and Pastore, 2009; Marmulla and Harder, 2014; Kandi et al., 2015). Following on literature methods to generate chemical libraries using graph theoretic techniques, complete libraries of all possible terpene isomers have been constructed with the goal of construction of derivative libraries of possible carbocation intermediates which are important in the elucidation of mechanisms in the biosynthesis of terpenes. Virtual library generation of monoterpenes was first achieved by generating graphs of order 7, 8, 9 and 10 using the Nauty and Traces suite. These were screened and processed with a set of collated Python scripts written to recognize the graphs in text format and translate them to molecules, minimizing through Tinker whilst discarding graphs that violate chemistry laws. As a result of the computational time required only order 7 and order 10 graphs were processed. Out of the 873 graphs generated from order seven, 353 were converted to molecules and from the 11,7 million produced from order 10 half were processed resulting in the production of 442928 compounds (repeats included). For screening, 55 366 compounds were docked in the active site of limonene synthase; of these 2355 ligands had a good Vina docking score with a binding energy of between -7.0 and -7.4 kcal.mol-1. When these best docked molecules were overlaid in the active site a map of possible ligand positions within the active site of limonene synthase was traced out.
- Full Text:
- Authors: Dendera, Washington
- Date: 2020
- Subjects: Terpenes , Plants -- Metabolism , Computational biology , Bioinformatics , Organic compounds -- Synthesis , Monoterpenes , Molecular biology -- Computer simulation
- Language: English
- Type: text , Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10962/123453 , vital:35439
- Description: Terpenes form a large group of organic compounds which have proven to be of use to many living organisms being used by plants for metabolism (Pichersky and Gershenzon, 1934; McGarvey and Croteau, 1995; Gershenzon and Dudareva, 2007), defence or as a means to attract pollinators and also used by humans in medical, pharmaceutical and food industry (Bicas, Dionísio and Pastore, 2009; Marmulla and Harder, 2014; Kandi et al., 2015). Following on literature methods to generate chemical libraries using graph theoretic techniques, complete libraries of all possible terpene isomers have been constructed with the goal of construction of derivative libraries of possible carbocation intermediates which are important in the elucidation of mechanisms in the biosynthesis of terpenes. Virtual library generation of monoterpenes was first achieved by generating graphs of order 7, 8, 9 and 10 using the Nauty and Traces suite. These were screened and processed with a set of collated Python scripts written to recognize the graphs in text format and translate them to molecules, minimizing through Tinker whilst discarding graphs that violate chemistry laws. As a result of the computational time required only order 7 and order 10 graphs were processed. Out of the 873 graphs generated from order seven, 353 were converted to molecules and from the 11,7 million produced from order 10 half were processed resulting in the production of 442928 compounds (repeats included). For screening, 55 366 compounds were docked in the active site of limonene synthase; of these 2355 ligands had a good Vina docking score with a binding energy of between -7.0 and -7.4 kcal.mol-1. When these best docked molecules were overlaid in the active site a map of possible ligand positions within the active site of limonene synthase was traced out.
- Full Text:
- «
- ‹
- 1
- ›
- »