INVESTIGATION OF ANTIBIOFILM AND ANTIVIRULENCE ACTIVITY OF 5-FLUOROCYTOSINE IN ESCHERICHIA COLI

Both acute and chronic infections by ‘opportunistic’ bacterial pathogens resistant to antibiotics are on the rise and they have been defined as a ‘slow motion catastrophe’ by the World Health Organization. Escherichia coli is an important opportunistic pathogen, and one of the most frequent causes of Urinary Tract Infections (UTIs). Due to the lack of novel promising antibiotics, novel alternative therapeutics are in demand. In particular, antivirulence agents, that selectively “disarm” pathogenic bacteria rather than killing them, are considered as a promising strategy. Several antimetabolites and nucleobase/nucleoside analogues, such as 5-fluorocytosine and 5-fluorouracil, have been shown to possess antivirulence activity in Gram negative bacteria like Pseudomonas aeruginosa and Escherichia coli. In my PhD work, based on previous observation that pyrimidine nucleotide availability can affect the production of biofilm and virulence factors in P. aeruginosa and E. coli, I evaluated fluoropyrimidines 5-fluorocytosine (5-FC), 5-fluorouracil (5-FU), and 5-fluorouridine (5-FUrd) as modulators of biofilm determinants in E. coli. I found that 5-FC showed lower antimicrobial activity to the E. coli laboratory strain MG1655 than 5-FU and 5-FUrd. However, all three fluoropyrimidines caused inhibition of curli production and general reduction of biofilm adhesion in dose-dependent manner in MG1655. This result was further confirmed by the observation that fluoropyrimidines can downregulate transcription of the curli structural gene csgA, suggesting that these compounds inhibit biofilm formation in E. coli through modulation of curli fibres expression. Since literature data indicate that 5-FC is less toxic to humans and my results proved that it possesses potent antibiofilm activity compared to other fluoropyrimidines, I focused on this molecule in further experiments. I tried to elucidate the molecular mechanisms behind the antibiofilm activity of 5-FC in E. coli and observed that curli inhibition by 5-FC requires its conversion into corresponding fluoronucleotide 5-FUMP. This led me to investigate the possible perturbation of de novo pyrimidine biosynthesis by 5-FC. Indeed, exposure to 5-FC resulted in a ca. 2-fold reduction of UMP intracellular levels, while not affecting the ATP nucleotide pool, thus suggesting a specific effect on the UMP biosynthetic pathway. Consistently, expression of the de novo pyrimidine biosynthesis genes carA and pyrB was upregulated in E. coli upon 5-FC treatment. Molecular docking studies suggested that 5-FUMP can bind the allosteric domain of carbamoyl phosphate synthetase, the first enzyme in de novo pyrimidine biosynthesis pathway, similar to its native allosteric inhibitor UMP and possibly mimic UMP’s negative feedback mechanism affecting the pyrimidine nucleotide pool. Overall, these results suggest that the antibiofilm activity of fluoropyrimidines is mediated, at least in part, by perturbation of the pyrimidine nucleotide pool. By performing E. coli MG1655 genomic library screening to identify additional determinants that counteracts the antivirulence effect of 5-FC, I identified a DNA fragment containing N-terminal domain of the penicillin-binding protein 1b (PBP1b), involved in peptidoglycan synthesis, and D8B36_18480, an unknown protein, restored curli production despite presence of 5-FC. Deleting PBP1b-encoding gene mrcB in E. coli increased curli gene expression, while overexpressing D8B36_18480 blocked it. While these two proteins do not appear to be direct targets of 5-FC in E. coli, they are potentially part of a not yet identified regulatory pathway linking peptidoglycan biosynthesis and curli production. Interestingly, along with exhibiting promising antibiofilm effect, 5-FC inhibit the expression of all major virulence factors of uropathogenic E. coli (UPEC) like curli fibers, secreted toxins, type I and P fimbriae, suggesting that the effect of 5-FC can also take place in clinically relevant isolates of E. coli. This result is further supported by the observation of reduction in UPEC cell adhesion and cytotoxicity against bladder epithelial cells upon 5-FC treatment. Overall, these results indicate that 5-FC affects the E. coli uropathogenesis effectively. Further, 5-FC was able to synergize specifically with peptidoglycan-targeting -lactam antibiotics alone and kill the biofilm-embedded UPEC bacteria. This result aligns well with the data I obtained in genetic screening that highlights a regulatory connection between peptidoglycan biosynthesis and the biofilm determinant ‘curli fibres’ production. Taken together, my results highlight that biofilm inhibitors like 5-FC can be used to potentiate conventional antibiotics to combat difficult-to-treat bacterial biofilm-associated infections, and such molecules can be used as discovery platform for identifying novel biofilm determinants and, thus, new targets for antivirulence strategies.