Comparison of three methods to study biofilm formation by clinical strains of Escherichia coli

Biofilm formation seems to be a key factor in many bacterial infections, particularly those involving prosthetic implants or urinary catheters, where Escherichia coli is frequently involved. We have determined the ability to form biofilm in vitro of 34 E. coli isolates by 3 different methods (crystal violet staining, BioFilm Ring Test®, and resazurin assay) and tried to correlate biofilm production with phylogenetic background and with the presence of different genes involved in biofilm synthesis. Only 3 isolates (including positive control E. coli ATCC 25922) were classified as strong biofilm producers (1B1, 1D, and 1B2 = control) by the 3 methods, 2 isolates by 2 different methods, and 5 additional isolates by only 1 method. All isolates possessed the csgA gene belonging to the csgABC operon encoding curli, and its regulator csgD. By contrast, only 76% possessed pgaA gene which is part of the pgaABCD operon encoding a polysaccharide adhesin. Interestingly, one of the strong biofilm producers did not harbor pgaA. In the second part, we have selected 5 specific isolates to study the impact of various experimental conditions on biofilm formation. For all these isolates, biofilm production was decreased in anaerobiosis and increased in LB medium compared with brain heart infusion medium, but at various degrees for the different isolates. These results underline the problems encountered in comparing the different published studies using various methods to study biofilm formation in vitro and the great need of standardization.