Methylene blue analogues : in vitro antimicrobial activity and in silico pharmacophore modelling

Abstract

Antibiotics have drastically reduced the morbidity and mortality associated with fatal infections and have made previously impossible procedures, e.g. heart surgery, possible by allowing infections to be prevented or treated. Unfortunately, the misuse of antibiotics has resulted in the development of numerous antimicrobial resistant bacterial strains. It is estimated that by the year 2050, deaths caused by antimicrobial resistant infections will surpass deaths caused by cancer. Lipopeptides and oxazolidinones are the only newly developed antibiotic classes that have been registered in the past two decades, both are indicated for the treatment of Gram-positive bacteria. The rapid spread of antimicrobial resistance as well as the lack of new antibiotics underlines the need for novel antimicrobial agents. Methylene blue was developed in 1876 as a dye in the textile industry and has since found numerous applications in different fields, e.g. biology, chemistry and medicine. Methylene blue has also been used as a lead compound in the development of quinolines (antimalarial agents) and phenothiazines (neuroleptic agents). Methylene blue is known to have antimicrobial properties although few studies have determined its minimum inhibitory concentration, which is the gold standard used to measure antimicrobial activity. Methylene blue’s precise antimicrobial mode of action is still unclear and to our knowledge no pharmacophore model has been created to investigate its mode of action. Computer-aided drug design is used in various aspects of drug discovery. When the drug target is unknown, ligand based drug design can be used to create a pharmacophore model by using the activity of known compounds. The pharmacophore model can be used to identify important chemical features that are necessary for activity and can also be used to predict the activity of other compounds. However, to obtain reliable results the pharmacophore model has to be properly validated. The aim of this study was to firstly screen methylene blue, methylene green, neutral red, new methylene blue, azure B, dimethyl methylene blue, methylene violet, cresyl violet, acriflavine, nile blue and ethylthioninium chloride for activity against Staphylococcus aureus, Staphylococcus epidermidis, Escherichia coli, Klebsiella pneumonia, Salmonella enterica and Candida albicans. Thereafter, the minimum inhibitory concentration (MIC) as well as the minimum bactericidal concentration (MBC) of the analogues with activity was determined against the relevant pathogens. The MIC data was used to develop and validate a common feature pharmacophore model. The pharmacophore model was validated with three statistical metrics, i.e. the enrichment factor (EF), hit rate (HR) and the area under the curve of a receiver operator characteristic (ROC-

AUC), which determined the pharmacophore model’s ability to differentiate between active/decoy compounds. The validated pharmacophore model was used to map similar antibacterial compounds in an attempt to elucidate methylene blue’s antimicrobial mode of action. Most analogues only had activity against S. aureus, S. epidermidis and K. pneumonia. Dimethyl methylene blue, new methylene blue and acriflavine proved to be the most active compounds against S. aureus [1 μg/ml (dimethyl methylene blue); 4 μg/ml (new methylene blue); 8 μg/ml (acriflavine)], S. epidermidis [1 μg/ml (dimethyl methylene blue); 4 μg/ml (new methylene blue); 2 μg/ml (acriflavine)] and K. pneumoniae [8 μg/ml (dimethyl methylene blue); 0.5 μg/ml (new methylene blue); 2 μg/ml (acriflavine)]. During method development, both the standard- and a less common bactericidal test were unable to deliver reliable MBC results. Comparing these two methods to one another it was found that acriflavine is bactericidal against S. aureus, and both methylene blue and azure B are bactericidal against K. pneumoniae. A common feature pharmacophore model was created (rank score: 26.664, max. fit value: 4), which was able to accurately identify active methylene blue analogous out of the test set (EF2%: 51, HR2%: 100%, ROC-AUC: 1.00 ± 0.00). Six phenothiazine derivatives with known antibacterial activity had high