Final Report Summary - STARS (Scientific Training in Antimicrobial Research Strategies)
The STARS (Scientific Training in Antimicrobial Research Strategies) Project (www.stars-itn.eu) was a Marie Curie ITN that brought together 12 (10 full + 2 associated) partners from universities, research centres and the private sector (2 small biotechnology and 1 large pharmaceutical company) within Europe. The STARS project objective comprised two sub-programmes, the first one focused on the development of novel antibacterial drugs and the second on new inhibitors of novel malarial proteinases.
To discover and develop novel antibacterial drugs, in silico, in vitro and in vivo experiments were carried out to gain insight into the 'druggability' of different two-component systems (TCS) and to screen for new designed inhibitors. In the in silico approach, novel drug design methodologies and cheminformatics databases were developed to support the search for new compounds. Cloud computing, newly emerging sources of very large-scale computational resources, were exploited and used to perform computational screening experiments that contributed to the identification of compounds for further experimental validation.
Novel inhibitors of ATP-binding in histidine kinases of multiple two-component systems were discovered following structure-based drug discovery and fragment-based drug discovery approaches. The initial hits were screened in vitro for their ability to inhibit autophosphorylation and the growth of Staphylococcus aureus, Streptococcus epidermidis, Escherichia coli and clinical isolates of multidrug resistant strains of Staphylococcus aureus, Acinetobacter baumannii, Strenotophomonas maltophilia and Streptococcus epidermidis. The derived lead molecules were co-crystallized with the ATP-binding domain of the CheA histidine kinase from Thermotoga maritima confirming that the mechanism of action of these inhibitors is the completion with the natural substrate ATP. The activity of these compounds was further analyzed in growth inhibition curves using virulent strains of Streptococcus pneumoniae, Streptococcus suis and multi-drug resistant strains of Staphylococcus aureus and Enterococcus faecium. More active inhibitors were obtained by ligand similarity searches and in vitro and in vivo activity assays on the identified compounds. These drug hits have potential to be developed into novel antibacterial drugs to combat infections form drug resistant bacteria.
Different TCSs of Streptococcus pneumoniae were characterised in vivo murine invasive disease models to identify TCS that are required for virulence. The most interesting finding was the differential expression of TCS12 and TCS05 at different infection sites suggesting distinct roles in infection. This means that the efficacy of inhibitors targeting TCS involved in virulence may depend highly on the environment within the host and stage of infection. TCS 12 was shown to be important in biofilm formation. In S. pneumoniae the essential TCS YycFG was unexpectedly found not to locate at the site of daughter cell division, which had been previously reported for Bacillus subtilis. This is most likely due to the evolution of different functions and mechanisms of YccF/G in these species. In S. pneumoniae the two-component system TCS13 was found to produce a bacteriocin-like peptide (BLP). Surprisingly the TCS13 deletion mutant was susceptible to BLP, indicating that the BLP operon, in addition to the peptides most probably encoded an immunity function to the bacteriocin-like peptide. This observation opens up possibilities for cross-inhibition by BLP on strains carrying unrelated TCS operons.
The construction of mutants in S. suis was hindered by the lack of efficient transformation and genetic tools. This was a serious obstacle for the generation of several mutants and validation of essential targets in the project so we explored the possibility that S. suis possessed an inducible state of competence for DNA transformation. We discovered in S. suis competence for DNA uptake and transformation was regulated by a peptide pheromone and that addition of the peptide in early growth phase could induce high efficiency transformation using plasmids or even linear DNA fragments. This has been a major breakthrough in the field making genetic manipulation and gene deletion in S. suis now routine. The new technique was further used in the project for a genome wide screen of essential genes using a saturated Mariner transposon insertion library. After library selection, changes in frequency of each insertion mutant were determined by sequencing the flanking regions en masse. We have identified 358 obligate essential genes (including the TCS YycF/G), which is essential in other Gram-positive bacteria.
The function of all TCS in S. suis was investigated by database homology searches and genetic approaches. Orthologues of YycFG were also found to be essential in Streptococcus suis an important pig pathogen and zoonotic pathogen of humans. Two new TCS targets were identified in S. suis that are conditionally essential for virulence in vivo. Both TCS were highly sensitive to hydrogen peroxide compared to the wild-type strain and low pH, which would explain their reduced virulence. Further in vitro studies revealed that neutrophil killing of the serum opsonised TCS mutant Δ0827/28 was significantly higher than killing of the corresponding wild type strain. Also TCS mutant Δ1930/31 showed a lower survival rate than the wild-type in neutrophil killing assays. The SSU0827/28 TCS was highly conserved in all database genome sequences of S. suis suggesting it would be a good target for structural based screening of novel antibacterial drug.
The DosRST TCS of Mycobaterium tuberculosis (Mtb), responsible for the entrance of Mtb into latency, was structurally and functionally studied through different technological approaches, i.e. electrophoretic mobility shift assays (EMSA), auto-phosphorylation assays, trans-phosphorylation assays and isothermal titration calorimetry. The phosphorylation events leading to DosR activation were studied at the molecular level and new insights into the amino acid residues responsible for DosR activation were gained. The implemented techniques performed with wild-type and mutant proteins allowed screening for DosR inhibitors, a starting point of a new rational-based drug discovery project directed against latent tuberculosis. Mtb causes around 1.5 million deaths annually and new drugs are need to treat infections with the growing number of highly drug-resistant TB strains.
Eukaryotic cationic peptides were studied as an alternative approach to synthetic antibiotics. A plasmid for detection of gene expression in lactic acid bacteria (LAB), based on the gene encoding the mCherry fluorescent protein was developed. Additional expression vectors were generated for the production of eukaryotic antimicrobial peptides (AMPs) in lactic acid bacteria. A screen was carried to select antimicrobial peptides that are most effective against drug resistant pathogens while having a low impact on commensal bacteria and low cytotoxicity for mammalian cells. For this task we screened alpha-helical AMPs from different species of frogs, which are known to have evolved effective AMPs to protect the skin. We identified eukaryotic peptides with a high antimicrobial activity against several pathogens but low inhibitory effects on the commensals used in this study. These promising candidates could serve as good candidates for drug- resistant bacteria causing difficult wound infections in animals and humans.
Regarding the identification of novel anti-malarial compounds, intensive studies using molecular simulation techniques of P. falciparum SUB1 (PfSUB1), an essential protease implicated in the parasite invasiveness, contributed to identifying the key “hotspots” associated with inhibitor binding. Virtual screening hits (i.e. small peptide-based ketoamide compounds and diamino-quinoxaline derivatives) were synthesized and their biological activity screened against SUB1. They showed significant inhibitory activity against SUB1 in the micromolar range (IC50 1-20 micromolar). Assays for SUB1 inhibitors were improved, and recombinant forms of SUB1 orthologues from the other human malaria pathogens P. vivax and P. knowlesi were produced. These assays with synthesized molecules gave new insights into the structural requirements of inhibitors and helped to setup a structure activity relationships (SAR) model. The knowledge generated in this project will have a high societal impact through its contribution to knowledge and strategies for treating malaria, a disease causing extensive mortality and morbidity.
The research results from the STARS project have been widely disseminated to academic and industrial audiences both locally in the countries of respective STARS partners and internationally through peer-review publications in top-ranked journals, posters or presentations in scientific conferences, in the annual STARS meetings and the International Antimicrobial Drug Discovery Symposium.
To discover and develop novel antibacterial drugs, in silico, in vitro and in vivo experiments were carried out to gain insight into the 'druggability' of different two-component systems (TCS) and to screen for new designed inhibitors. In the in silico approach, novel drug design methodologies and cheminformatics databases were developed to support the search for new compounds. Cloud computing, newly emerging sources of very large-scale computational resources, were exploited and used to perform computational screening experiments that contributed to the identification of compounds for further experimental validation.
Novel inhibitors of ATP-binding in histidine kinases of multiple two-component systems were discovered following structure-based drug discovery and fragment-based drug discovery approaches. The initial hits were screened in vitro for their ability to inhibit autophosphorylation and the growth of Staphylococcus aureus, Streptococcus epidermidis, Escherichia coli and clinical isolates of multidrug resistant strains of Staphylococcus aureus, Acinetobacter baumannii, Strenotophomonas maltophilia and Streptococcus epidermidis. The derived lead molecules were co-crystallized with the ATP-binding domain of the CheA histidine kinase from Thermotoga maritima confirming that the mechanism of action of these inhibitors is the completion with the natural substrate ATP. The activity of these compounds was further analyzed in growth inhibition curves using virulent strains of Streptococcus pneumoniae, Streptococcus suis and multi-drug resistant strains of Staphylococcus aureus and Enterococcus faecium. More active inhibitors were obtained by ligand similarity searches and in vitro and in vivo activity assays on the identified compounds. These drug hits have potential to be developed into novel antibacterial drugs to combat infections form drug resistant bacteria.
Different TCSs of Streptococcus pneumoniae were characterised in vivo murine invasive disease models to identify TCS that are required for virulence. The most interesting finding was the differential expression of TCS12 and TCS05 at different infection sites suggesting distinct roles in infection. This means that the efficacy of inhibitors targeting TCS involved in virulence may depend highly on the environment within the host and stage of infection. TCS 12 was shown to be important in biofilm formation. In S. pneumoniae the essential TCS YycFG was unexpectedly found not to locate at the site of daughter cell division, which had been previously reported for Bacillus subtilis. This is most likely due to the evolution of different functions and mechanisms of YccF/G in these species. In S. pneumoniae the two-component system TCS13 was found to produce a bacteriocin-like peptide (BLP). Surprisingly the TCS13 deletion mutant was susceptible to BLP, indicating that the BLP operon, in addition to the peptides most probably encoded an immunity function to the bacteriocin-like peptide. This observation opens up possibilities for cross-inhibition by BLP on strains carrying unrelated TCS operons.
The construction of mutants in S. suis was hindered by the lack of efficient transformation and genetic tools. This was a serious obstacle for the generation of several mutants and validation of essential targets in the project so we explored the possibility that S. suis possessed an inducible state of competence for DNA transformation. We discovered in S. suis competence for DNA uptake and transformation was regulated by a peptide pheromone and that addition of the peptide in early growth phase could induce high efficiency transformation using plasmids or even linear DNA fragments. This has been a major breakthrough in the field making genetic manipulation and gene deletion in S. suis now routine. The new technique was further used in the project for a genome wide screen of essential genes using a saturated Mariner transposon insertion library. After library selection, changes in frequency of each insertion mutant were determined by sequencing the flanking regions en masse. We have identified 358 obligate essential genes (including the TCS YycF/G), which is essential in other Gram-positive bacteria.
The function of all TCS in S. suis was investigated by database homology searches and genetic approaches. Orthologues of YycFG were also found to be essential in Streptococcus suis an important pig pathogen and zoonotic pathogen of humans. Two new TCS targets were identified in S. suis that are conditionally essential for virulence in vivo. Both TCS were highly sensitive to hydrogen peroxide compared to the wild-type strain and low pH, which would explain their reduced virulence. Further in vitro studies revealed that neutrophil killing of the serum opsonised TCS mutant Δ0827/28 was significantly higher than killing of the corresponding wild type strain. Also TCS mutant Δ1930/31 showed a lower survival rate than the wild-type in neutrophil killing assays. The SSU0827/28 TCS was highly conserved in all database genome sequences of S. suis suggesting it would be a good target for structural based screening of novel antibacterial drug.
The DosRST TCS of Mycobaterium tuberculosis (Mtb), responsible for the entrance of Mtb into latency, was structurally and functionally studied through different technological approaches, i.e. electrophoretic mobility shift assays (EMSA), auto-phosphorylation assays, trans-phosphorylation assays and isothermal titration calorimetry. The phosphorylation events leading to DosR activation were studied at the molecular level and new insights into the amino acid residues responsible for DosR activation were gained. The implemented techniques performed with wild-type and mutant proteins allowed screening for DosR inhibitors, a starting point of a new rational-based drug discovery project directed against latent tuberculosis. Mtb causes around 1.5 million deaths annually and new drugs are need to treat infections with the growing number of highly drug-resistant TB strains.
Eukaryotic cationic peptides were studied as an alternative approach to synthetic antibiotics. A plasmid for detection of gene expression in lactic acid bacteria (LAB), based on the gene encoding the mCherry fluorescent protein was developed. Additional expression vectors were generated for the production of eukaryotic antimicrobial peptides (AMPs) in lactic acid bacteria. A screen was carried to select antimicrobial peptides that are most effective against drug resistant pathogens while having a low impact on commensal bacteria and low cytotoxicity for mammalian cells. For this task we screened alpha-helical AMPs from different species of frogs, which are known to have evolved effective AMPs to protect the skin. We identified eukaryotic peptides with a high antimicrobial activity against several pathogens but low inhibitory effects on the commensals used in this study. These promising candidates could serve as good candidates for drug- resistant bacteria causing difficult wound infections in animals and humans.
Regarding the identification of novel anti-malarial compounds, intensive studies using molecular simulation techniques of P. falciparum SUB1 (PfSUB1), an essential protease implicated in the parasite invasiveness, contributed to identifying the key “hotspots” associated with inhibitor binding. Virtual screening hits (i.e. small peptide-based ketoamide compounds and diamino-quinoxaline derivatives) were synthesized and their biological activity screened against SUB1. They showed significant inhibitory activity against SUB1 in the micromolar range (IC50 1-20 micromolar). Assays for SUB1 inhibitors were improved, and recombinant forms of SUB1 orthologues from the other human malaria pathogens P. vivax and P. knowlesi were produced. These assays with synthesized molecules gave new insights into the structural requirements of inhibitors and helped to setup a structure activity relationships (SAR) model. The knowledge generated in this project will have a high societal impact through its contribution to knowledge and strategies for treating malaria, a disease causing extensive mortality and morbidity.
The research results from the STARS project have been widely disseminated to academic and industrial audiences both locally in the countries of respective STARS partners and internationally through peer-review publications in top-ranked journals, posters or presentations in scientific conferences, in the annual STARS meetings and the International Antimicrobial Drug Discovery Symposium.