Studies of the molecular mechanisms of antimicrobial peptides (AMPs) have mostly been performed with lipid bilayers, as a substitute for cell membrane. Hence, there is a persistent question as to whether the action of AMPs on bacterial membranes can be reproduced on lipid bilayers. Valuable information was obtained recently from observing the actions of AMPs on E. coli and Bacillus subtilis by time-lapse fluorescence microscopy. The goal of my dissertation is to study the direct action of AMPs on the cytoplasmic membranes by using E. coli spheroplasts, the cell form from which the outer membranes have been removed.

The key question is how to reveal the response of the spheroplast to AMPs. In our previous work, the aspiration method on giant unilamellar vesicle (GUV) has been demonstrated as very effective for researching membrane effects induced by peptides. This method is able to measure the membrane expansion due to peptide binding and simultaneously monitor membrane permeability by using dye indicators. In this work, a spray method was developed for introducing AMPs and customized the experimental procedures for performing the experiments of E. coli spheroplasts.

The living state of cytoplasmic membrane makes it difficult to deduce the molecular events from the response of live cells. Hence, the physical methods which have been developed for researching peptide activity on lipid bilayers were practiced first. These methods include X-ray diffraction (XRD), oriented circular dichroism (OCD) and GUV aspiration. These methods were practiced by studying the question as to why a hydrocarbon-stapled peptide NYAD-1 drug was reported to have membrane permeating property.

Melittin was selected as the representative of AMPs for the E. coli spheroplast experiments. To better understand the characteristics of melittin, the melittin activity on model lipid membrane was examined before advancing to the spehroplast experiment. The melittin transmembrane was proposed by correlating melittin binding on a lipid vesicle (aspirated GUV imaging) with structural studies in multilayers (XRD and OCD). This behavior was further determined by using fluorescence indicator to track the melittin distribution. The toroidal structure of melittin pore was also detected by grazing-angle X-ray anomalous diffraction. In the studies of NAYD-1 and melittin on a model membrane, our discovery of AMP's transmembrane enables us to clarify the pore-formation mechanism which has been disputed for decades.

In addition, our past work on the physical property of E. coli spheroplast cytoplasmic membrane has indicated the existence of a lipid reservoir. The lipid reservoir dominates the surface tension by balancing the membrane folds. Finally, we used the aspiration method to hold a spheroplast so as to measure the change of the spheroplast membrane area in response to the AMPs' binding. Further, a fluorescence indicator which is able to associate with AMPs was used to monitor the peptide distribution and another fluorescence dye to monitor the molecule leakage.

The spheroplast study shows that there are similarities and differences between the responses of spheroplasts and GUVs. The recent findings on the unique properties of spheroplast membranes are the key for understanding these results. Our work of understanding the AMPs activity on membrane is potentially applicable in improving peptide drug design and delivery for disease treatment.