Background: Due to the increase of multidrug-resistant microorganisms, the search for
biologically active molecules does not stop. In the present study, we developed the effective QSAR
model which allows a quick search of new potential Staphylococcus aureus inhibitors in the series of
quaternary phosphonium salts. A number of the most promising 1,3-oxazol-4-yltriphenylphosphonium
derivatives with predicted activities were synthesized and examined to confirm their antibacterial
properties and the accuracy of the forecast. Furthermore, the toxicity of the investigated compounds
Methods: The predictive QSAR model was developed using Artificial Neural Network approach.
Antibacterial properties of the investigated compounds were performed using standard disk diffusion
method. The toxicity of the compounds was determined in vivo using zebrafish (Danio rerio)
and in vitro on acetylcholinesterase (AChE) enzyme as the test models.
Results: The predictive ability of the regression model was tested by cross-validation, giving the
cross-validated coefficient q2=0.82. Derivatives of 1,3-oxazol-4-yltriphenylphosphonium salts predicted
as active were synthesized and screened for their antibacterial activities. All compounds
demonstrated antibacterial activity according to the prediction. The toxicity tests indicated that all
investigated samples were less toxic than well-known cationic surfactants.
Conclusion: The most promising compound 2b exhibited strong antibacterial activity together with
low toxicity and can be considered as a new efficient biocidal agent for future investigation. In
addition, the proposed QSAR model can be used for predicting and designing novel potential
S. aureus inhibitors among ionic liquids/salts.