A study on the mode of action of clinically relevant antimicrobial peptides

• Author(s): Melo, Manuel Nuno, 1982-
• Date: 2010
• Persistent ID: http://hdl.handle.net/10451/1881
• Origin: Repositório da Universidade de Lisboa
• Subject(s): Péptidos; Membrana celular; Saturação; Biofísica molecular; Teses de doutoramento - 2010
• Document Type: Doctoral Thesis
• Language: English
• Advisor(s): Castanho, Miguel Augusto Rico Botas, 1967-; Coutinho, Ana Isabel Abrantes, 1965-

Description

Tese de doutoramento, Bioquímica (Biofísica Molecular), Universidade de Lisboa, Faculdade de Ciências, 2010 The antimicrobial peptides (AMPs) omiganan (ILRWPWWPWRRK– NH2) and BP100 (KKLFKKILKYL–NH2) were biophysically studied with bacterial and mammalian cell membrane models, essentially using optical spectroscopy techniques. Peptide-membrane binding was interpreted under a Nernst partition formalism. Both peptides strongly prefer the anionic bacterial membrane models over the zwitterionic mammalian ones, justifying, at least in part, higher antibacterial than hemolytic activities. Deviations to the expected binding behavior were observed at high bound peptide-to-lipid (P:L) ratios in the membrane whenever anionic models were used. These deviations could be ascribed to membrane saturation and occurred with both peptides. The saturation threshold could be identified for both peptides; the obtained critical P:L ratios were also consistent with membrane surface charge neutralization. Disruptive events were observed above the saturation threshold: internalization into the membrane (omiganan), leakage, membrane aggregation, membrane surface charge neutralization, and structural reorganization (BP100). The plausibility of high membrane coverage was evaluated using a mathematical model devised to estimate the extent of binding under physiological conditions. Not only is saturation a possible phenomenon but it was also shown to be a potential requirement for peptide activity. This hypothesis could be verified using the model with published data on several AMPs. The model could be further adapted to provide a means to predict, from simple biophysical parameters (a binding constant and a critical P:L ratio), the peptide concentration at which antibacterial activity is triggered. Different methods to implement this prediction are presented.In vivo measurements using BP100 with Escherichia coli were carried out to further test the correlation between membrane saturation and bacterial death. A suitable method to determine the occurrence of binding saturation with bacteria could not be devised. Nonetheless, bacterial surface charge did become neutralized by the peptide at the same concentrations that caused loss of viability, supporting a connection between the phenomena.