Faculty Advisor or Committee Member

Terri A. Camesano, Advisor

Faculty Advisor or Committee Member

Kevin G. Cornwell, Committee Member

Faculty Advisor or Committee Member

Tanja Dominko, Committee Member

Faculty Advisor or Committee Member

Amy M. Peterson, Committee Member

Co-advisor

Marsha W. Rolle

Identifier

etd-012318-110452

Abstract

As society draws closer to the post-antibiotic era and the pipeline for alternatives dries, there is an urgent need for the development of novel antimicrobial therapies that do not promote bacterial resistance, particularly for immunocompromised chronic wound patients. Antimicrobial peptides (AMPs), including human-derived LL37, show considerable promise as broad spectrum alternatives that also have wound healing properties; however, few have been clinically implemented as novel antimicrobials due to their cytotoxicity stemming from a poor understanding of their mechanisms and low stability in vivo. It has been suggested that tethering, or attaching AMPs onto surfaces, is a viable strategy of delivering bioactive AMPs to surfaces while reducing cytotoxicity and improving stability. Thus, we designed new chimeric versions of LL37 with collagen-binding domains (CBD), derived from collagenase (cCBD-LL37) and fibronectin (fCBD-LL37) for non-covalent tethering onto collagen, a prevalent biopolymer in commercially available wound dressings and scaffolds. Our overall hypothesis was that CBDs would mediate stable tethering of broadly active, non-cytotoxic CBD-LL37 onto collagen-based scaffolds. We first studied the loading, release and bioactivities (e.g. antimicrobial activity and cytotoxicity) of each CBD-LL37 on commercially available 100% collagen type I PURACOL® wound scaffolds. We found that both cCBD-LL37 and fCBD-LL37 bound highly to collagen, were active against relevant wound pathogens, demonstrated stable activity after 14 days of release, and were not cytotoxic to human fibroblasts. The addition of different CBDs onto LL37 also markedly altered their soluble bioactivities. Using similar methods, we then studied the loading, release and bioactivity of each CBD-LL37 on a commercially available FIBRACOL® wound scaffolds, comprised of 90% collagen type I and 10% calcium alginate biopolymers. We found that both cCBD-LL37 and fCBD-LL37 also bound highly to and retained on collagen for 14 days, but were only active against Gram-negative P. aeruginosa. This suggested that the presence other biopolymers in addition to collagen, which is common among commercial wound dressings, could cause significant differences in binding, retention and bioactivities of CBD-LL37. To better understand how CBD modification affected CBD-LL37 structure leading to different bioactivities, we studied the CBD sequence-, peptide structure-, concentration-, time-, and bilayer composition-dependent interactions of soluble CBD-LL37 and compared these findings with the properties of unmodified LL37. Using Molecular Dynamics (MD) simulations, circular dichroism (CD) spectroscopy, quartz crystal microbalance with dissipation (QCM-D), and fluorescent bilayer imaging we determined the structural basis behind CBD alterations in bioactivities. MD and CD, in addition to other intrinsic CBD properties (helicity, amphiphilicity, charge) we hypothesized that cCBD-LL37 utilized similar mechanisms as unmodified LL37 while fCBD-LL37 demonstrated based primarily on surface adsorption. We used QCM-D and Voigt-Kelvin viscoelastic modeling to determine the time- and concentration-dependent interactions of unmodified LL37 with model mammalian lipid bilayers, the mechanisms of which are still controversial in literature despite being widely studied. These results were used to propose a model for the interaction mechanism of LL37 with zwitterionic bilayers that aligned with its bioactive concentrations. LL37 adsorbed at concentrations where it is immunomodulatory until reaching a threshold which corresponded with its antimicrobial concentrations. The threshold was correlated to lipid bilayer saturation, after which LL37 formed transmembrane pores. We observed collapse of the bilayer into a rigid proteolipid film at concentrations higher than the reported cytotoxic threshold of LL37. The mechanistic and structural information for each CBD-LL37 and unmodified LL37 provided a baseline for QCM-D and Voigt-Kelvin viscoelastic modeling to further elucidate the time-, concentration-, lipid composition- and CBD sequence-dependent basis behind the observed bioactivities of cCBD-LL37 and fCBD-LL37. We found that similar to LL37, cCBD-LL37 demonstrated pore formation mechanisms likely due to their similar charges, structural content and amphiphilicity. fCBD-LL37 demonstrated time-dependent, adsorption-based mechanism likely due to its anchoring aromatic residues, low charge, and low amphiphilicity. Knowledge gained from this study allowed mechanistic predictions of two newly designed hypothetical CBD-LL37 peptides. Results from this study contribute to a better understanding of a new class of antimicrobial, non-cytotoxic therapies based on collagen-tethered CBD-LL37, bringing it closer to clinical implementation in chronic wound applications and demonstrate the viability of biopolymer tethering as a platform toward using AMPs to quench the resistance crisis.

Publisher

Worcester Polytechnic Institute

Degree Name

PhD

Department

Chemical Engineering

Project Type

Dissertation

Date Accepted

2018-01-23

Accessibility

Unrestricted

Subjects

Antimicrobial peptides, cCBD-LL37, chronic wounds, collagen, collagen binding domain, fCBD-LL37, LL37, mechanism, Quartz crystal microbalance with dissipation, tether, wound healing

Share

COinS