In most ecosystems, bacteria primarily exist as structured surface-associated biofilms that can be highly tolerant to antibiotics and thus represent an important health issue. Here we explored drug repurposing as a strategy to identify new antibiofilm compounds, screening over 1000 compounds from the Prestwick Chemical Library of approved drugs for specific activities that prevent biofilm formation by Escherichia coli. Most growth-inhibiting compounds, which include known antibacterial but also antiviral and other drugs, had also reduced biofilm formation. However, we also identified several drugs that were biofilm-inhibitory at doses where only weak or no effect on planktonic growth could be observed. Activities of the most specific antibiofilm compounds were further characterized using gene expression analysis, proteomics and microscopy. We observed that most of these drugs acted by repressing genes responsible for production of curli, major component of E. coli biofilm matrix. This repression apparently occurred through induction of several different stress responses, including DNA and cell-wall damage, and homeostasis of divalent cations, demonstrating that biofilm formation can be inhibited through a variety of molecular mechanisms. One tested drug, tyloxapol, inhibited biofilm formation independent of curli expression or growth, by suppressing bacterial attachment at the surface.
Importance Prevention of bacterial biofilm formation is one of the major current challenges in microbiology. By systematically screening a large number of approved drugs for their ability to suppress biofilm formation by Escherichia coli, here we identified a number of prospective antibiofilm compounds. We further demonstrated different mechanisms of action for individual compounds, from induction of replicative stress to disbalance of cation homeostasis to inhibition of bacterial attachment to the surface. Our work demonstrates the potential of drug repurposing for the prevention of bacterial biofilm formation and suggests that also for other bacteria the activity spectrum of antibiofilm compounds is likely to be broad.