Examining the antimicrobial activity and toxicity to animal cells of different types of CO-releasing molecules.

Transition metal carbonyl complexes used as CO-releasing molecules (CORMs) for biological and therapeutic applications may exhibit interesting antimicrobial activity. However, understanding the chemical traits and mechanisms of action that rule this activity is required to establish a rationale for the development of CORMs into useful antibiotics. In this work the bactericidal activity, the toxicity to eukaryotic cells, and the ability of CORMs to deliver CO to bacterial and eukaryotic cells were analysed for a set of seven CORMs that differ in the transition metal, ancillary ligands and the CO release profile. Most of these CORMs exhibited bactericidal properties that decrease in the following order: CORM-2 > CORM-3 > ALF062 > ALF850 > ALF186 > ALF153 > [Fe(SBPy3)(CO)](BF4)2. A similar yet not entirely coincident decreasing order was found for their induction of intracellular reactive oxygen species (ROS) in E. coli. In contrast, studies in model animal cells showed that for any given CORM, the level of intracellular ROS generated was negligible when compared with that measured inside bacteria. Importantly, these CORMs were in general not toxic to eukaryotic cells, namely murine macrophages, kidney LLC-PK1 epithelial cells, and liver cell line HepG2. CORM-2 and CORM-3 delivered CO to the intracellular space of both E. coli and the two types of tested eukaryotic cells, yet toxicity was only elicited in the case of E. coli. CO delivered by ALF186 into the intercellular space did not enter E. coli cells and the compound was not toxic to either bacteria or to eukaryotic cells. The Fe(ii) carbonyl complex [Fe(SBPy3)(CO)](2+) had the reverse, undesirable toxicity profile, being unexpectedly toxic to eukaryotic cells and non-toxic to E. coli. ALF153, the most stable complex in the whole set, was essentially devoid of toxicity or ROS induction ability in all cells. These results suggest that CORMs have a relevant therapeutic potential as antimicrobial drugs since (i) they can show opposite toxicity profiles towards bacteria and eukaryotic cells; (ii) their activity can be modulated through manipulation of the ancillary ligands, as shown with the three {Ru(CO)3}(2+) and two zerovalent Mo based CORMs; and (iii) their toxicity to eukaryotic cells can be made acceptably low. With this new approach, this work contributes to the understanding of the roots of the bactericidal action of CORMs and helps in establishing strategies for their development into a new class of antibiotics.