Project 2–3: Molecular Dynamics Models of Antimicrobial Agents

Principal Investigator: Eric Darve (Stanford University)

Antimicrobial paints and fabric treatments offer a lightweight, economical, form of personnel protection against biowarfare agents (BWAs) and other pathogens. These coatings also protect vehicles and equipment from the corrosive effects of microbial secretions, and antimicrobials retard the spoilage of latex paints.

In 2007, the U.S. Army reported annual expenditures of more than $100 million for topcoat materials to protect against chemical agents. Several coating methods, such as chemical vapor deposition, are effective for small surfaces but are not practical for large equipment or vehicle coatings. Newer technologies, such as self assembled monolayers, do not hold up well in harsh environments. Effective coatings must adhere well without compromising or damaging the surfaces to which they are applied.

Recent research has focused on peptides, the molecular building blocks of proteins, as components of antimicrobial paints and coatings. Antimicrobial peptides, which are common in nature, protect the underlying organism or structure by disrupting the cell membranes of bacteria that come into contact with the coating. Antimicrobial peptides act by forming large holes in cell membranes or by carpeting membrane surfaces in a manner similar to detergents.

Eric Darve, assistant professor of mechanical engineering at Stanford University, is developing high performance computer simulation capabilities to further the understanding of the molecular mechanisms behind this antimicrobial activity, as the focus of a project for AHPCRC that began in early 2009. His simulations will also prove useful in identifying and designing peptide mutations that work well in polymer coatings being developed by the Army. Darve is working in tandem with Jan Andzelm and Lars Piehler, researchers at the Army Research Laboratory’s Weapons and Materials Research Directorate at Aberdeen Proving Ground, Maryland (WMRD APG), who are producing experimental data in their laboratories.

Improving on Nature

Darve and his Army collaborators are especially interested in a well-characterized group of peptides called cecropins. These peptides were originally isolated in the early 1980s from the giant silk moth, Hyalophora cecropia, but their molecular relatives are found in many species of winged insects. As you might expect, cecropins protect their insect hosts against invasive micro-organisms. They do this by attaching themselves to and opening large pores in microbial cell membranes.

By varying the amino acids that form the cecropin peptide, analogs have been produced that are more potent than the parent compounds. This indicates that intentionally-designed cecropin analogs could be produced for specific applications and conditions. In order to do this, however, the mechanisms by which these peptides do their work must be understood much better than they are at present. Computer simulations are ideally suited for this type of exploratory work, in that a large number of alternatives may be explored, and patterns may be identified that point the way toward the most promising candidates.

Any useful synthetic peptide must preserve cecropin’s ability to distinguish between negatively-charged bacterial (prokaryotic) cell membranes and human or animal (eukaryotic) cell membranes, which have positively and negatively charged regions. Peptide molecules are made from electrically charged (hydrophilic) and electrically neutral (hydrophobic) sections. These complex molecules fold and unfold, depending on their thermal and chemical environments. The portions of the folded molecules, charged or uncharged, that face outward determine whether the molecule is attracted to and adheres to a particular type of cell membrane. Laboratory work has shown that helical molecules having a net positive charge are the most active forms for binding to the negatively-charged exteriors of bacterial cells. Once the molecules attach themselves to a bacterial cell membrane, a mechanism must be found for making the hydrophobic sections of the helices available to penetrate and breach the membrane.

Simulation work

Computer resource requirements are the greatest limiting factor in the simulation of such biomolecular systems. Simulations of this sort must model such molecular behaviors as helix formation, unwinding, insertion into a membrane, and self-assembly over relatively long time scales—microseconds to milliseconds. Current software and numerical methods do not allow modeling over the necessary time scales; Darve places a high priority on creating algorithms that would enable this gap to be bridged.

Darve’s main software platform is NAMD, a largescale parallel code for molecular dynamics simulation of proteins and cell membranes. This code is robust and faster on parallel computers than most competing codes. At present, the NAMD computer code can simulate 6–22 nanoseconds of molecular behavior per day of execution time, using 128 processors in parallel (38,000 atoms, 2 femtosecond time step).

Darve has determined the structure of cecropin in water and at a polar–nonpolar interface (such as the interface between water and a cell membrane). These molecular simulations made it possible to validate the force field parameters. He has found that the molecule starts to unwind when the protein is in bulk water, which is consistent with experimental findings.

Future Work

For the immediate future, work will center on the basic science aspects of the molecular models: validating the molecular force fields used in the model and investigating the behavior of cecropin in water and at nonpolar interfaces. These studies will provide a foundation for models of the interactions between cecropin and various types of cell membranes. Darve has developed some novel techniques to analyze long time-scale events, and these will be used to determine the unfolding and folding mechanisms. Later studies will evaluate the antimicrobial properties of cecropin and its variants to identify the most promising candidates for inclusion in active coatings. During the course of this research, the results will be validated against experimental work done at ARL.

Source: AHPCRC Bulletin, Vol. 1 No. 4 (2009)