Investigation of plasma protein binding and nephrotoxicity of polymyxins

The use of the polymyxin antibiotics (colistin [polymyxin E] and polymyxin B) declined soon after their approval in the 1950s due to the high incidence of nephrotoxicity and neurotoxicity, and the availability of ‘less toxic’ therapeutic options that had become available at the time. In recent times, the use of polymyxins has increased particularly as a ‘last-line’ defence against infections caused by multidrug-resistant (MDR) Gram-negative bacteria. There is a paucity of pharmacological information for the rational use of polymyxins to enable maximising of antibacterial activity and minimising of toxicity, as well as for the development of safer polymyxins. Currently recommended dosage regimens for polymyxins are suboptimal in many patients. The dose-limiting nephrotoxicity of polymyxins is also a major barrier for optimising their clinical use. As a part of the endeavour to increase the current understanding of polymyxin pharmacology, particularly plasma protein binding and nephrotoxicity, this thesis examined the interaction of polymyxins with the human acute-phase plasma protein α1-acid glycoprotein (AGP), and also investigated mechanisms of polymyxin-induced nephrotoxicity. Polymyxins are approximately 50% bound to plasma proteins in humans. However, their mode of binding to important plasma proteins such as AGP remains unknown, despite the importance of plasma protein binding for polymyxin pharmacokinetics and pharmacodynamics. The binding interactions of the polymyxins with human AGP were evaluated using a combination of biophysical techniques. Colistin, polymyxin B and polymyxin B3 showed moderate binding affinities to AGP (dissociation constant for displacement Ki >200 µM). Observed binding affinities of polymyxins to human AGP suggested that both the polycationic decapeptide and the hydrophobic N-terminal fatty acyl chain of the polymyxin molecule are required for their binding to AGP. In addition, these results indicate that the formation of the polymyxin-AGP complex is facilitated by the presence of endogenous lipids bound to AGP. In acute states such as infection or inflammation, concentrations of AGP in human plasma increase and this could offer additional AGP binding sites for polymyxins. Ultimately, availability of unbound (free) drug to exert the antibacterial activity and to be available for filtration at the glomerulus could be altered. Extensive renal tubular reabsorption and accumulation of colistin and polymyxin B have been reported from several studies in animals and humans. There is no experimental evidence of tubular accumulation of polymyxins at the cellular level. This knowledge is crucial for the understanding of the renal tubular handling of polymyxins. A novel dual-modality iodine-labelled fluorescent probe of polymyxin (FADDI-096) was designed and synthesized. The accumulation and localisation of FADDI-096 in single kidney proximal tubular cells were quantitatively mapped using a correlative microscopic approach. The cytoplasmic and nuclear accumulation of the probe was revealed from the fluorescence responses following treatment. The concentrations of FADDI-096 in single rat (NRK-52E) and human (HK-2) kidney tubular cells were ~1,930- and 4,760-fold higher than the initial extracellular concentrations (5.0 to 50.0 µM). Intracellular accumulation of FADDI-096 was concentration- and time-dependent. In addition, a significant correlation between the intracellular accumulation of FADDI-096 and calcium concentration supported the subsequent studies of polymyxin-induced apoptosis in kidney tubular cells. Polymyxin-induced nephrotoxicity has been correlated with apoptosis in the kidney tubules. Hitherto the underlying characteristics and pathways of this phenomenon remain poorly understood. Polymyxin-induced apoptosis in cultured rat and human kidney proximal tubular cells was characterised using a series of biochemical methods and classical apoptosis cell signalling markers. Concentration- and time-dependent apoptosis caused by polymyxin treatment in both NRK-52E and HK-2 cells was in agreement with the concentration- and time-dependent accumulation of the polymyxin probe mentioned above. Notably, 95% confidence intervals of the polymyxin B concentrations required to induce 50% of maximal apoptosis (EC50) were 0.29 to 0.42 µM for HK-2 cells and 0.91 to 1.22 µM for NRK-52E cells. These observations reveal cell line dependent apoptosis induced by polymyxins. Finally, the pathways associated with apoptotic cell death induced by polymyxins were examined in NRK-52E cells. Apoptosis was confirmed from the observation of concentration-dependent DNA damage in polymyxin B treated cells. The number of NRK-52E cells with TUNEL-positive nuclei increased to 98.4 ± 2.1% when treated with 2.0 mM of polymyxin B for 24 h. Concentration-dependent activation of caspase-3, -8 and -9 was observed. The proportion of FasL positive cells increased to 91.7 ± 9.04% following treatment with 2.0 mM polymyxin B for 24 h. Notably, the loss of mitochondrial membrane potential (Δψm) and the generation of mitochondrial superoxide in the polymyxin-treated cells are in good agreement with the observed concentration- and time-dependent deformation of mitochondrial morphology. These data suggest the ability of polymyxins to induce mitochondrial toxicity in NRK-52E cells. For the very first time, this study revealed that polymyxin-induced apoptosis in cultured kidney tubular cells is mediated via the death receptor (extrinsic) and mitochondrial (intrinsic) pathways. In summary, this thesis provides novel information on both the plasma protein binding characteristics and polymyxin-induced nephrotoxicity. The very extensive accumulation of polymyxins in single kidney tubular cells may play a major role in the apoptotic cell death. Targeting the events involved in the extensive cellular uptake of polymyxins and those associated with the death receptor and mitochondrial mediated apoptotic cell death is crucial to further understanding of polymyxin-induced nephrotoxicity. Such mechanistic information will not only assist in optimising the clinical efficacy of these last-line therapeutic agents, but also expedite the discovery of safer, new polymyxins.