Search

ABSTRACTWe recently described a family of cationic antimicrobial peptides (CAMPs) selected from a combinatorial library that exhibited potent, broad-spectrum activity at neutral pH and low ionic strength. To further delimit the utility and activity profiles of these peptides, we investigated the effects of solution conditions, such as pH and ionic strength, on the efficacy of the peptide antimicrobials against a panel of microorganisms. Peptide minimum sterilizing concentrations (MSCs) varied linearly with pH for each subtype within our family of CAMPs for all organisms tested. The peptides were much less effective against Gram-negative bacteria at high pH, consistent with a decrease in net positive charge on the peptides. A similar trend was observed for the fungus Candida albicans. Surprisingly, the opposite pH trend was observed with the Gram-positive Staphylococcus aureus. In addition, an additive ionic strength effect was observed with increasing buffer strengths at identical pH values. The extreme difference in the observed pH behavior between Gram-negative and Gram-positive organisms is attributed to the presence of native charged molecules in the much thicker peptidoglycan layer of the Gram-positive organism. The novel species-specific effects of pH observed here have important implications for applications using CAMPs and for the design of novel CAMPs.

The interactions that drive the folding of β-barrel membrane proteins have not been well studied because there have been few available model systems for membrane β-sheets. In this work, we expand on a recently described model system to explore the contributions of interstrand hydrogen bonds, side-chain/side-chain interactions and side-chain/membrane interactions to β-sheet formation in membranes. These experiments are based on the observation that the hydrophobic hexapeptide acetyl-Trp-Leu-Leu-Leu-Leu-Leu-OH (AcWLLLLL) folds, cooperatively and reversibly, into oligomeric, antiparallel β-sheets in phosphatidylcholine membranes. To systematically characterize the important interactions that drive β-sheet formation in membranes, we have used circular dichroism spectroscopy to determine the membrane secondary structure of each member of a complete host-guest family of related peptides of the form AcWLL-X-LL, where X is one of the natural amino acids. Peptides with hydrophobic X-residues of any size or character (X = Ala, Val, Ile, Leu, Cys, Met, Phe and Trp) form similar β-sheets in membranes, while peptides with any polar X-residue or Gly or Pro at the X-position are random-coils, even when bound to membranes at high concentrations. The observed membrane sheet preferences correlate poorly with intrinsic sheet propensity scales measured in soluble proteins, but they correlate well with several membrane hydrophobicity scales. These results support the idea that the predominant interactions of the side-chains in membrane-bound β-sheets are with the membrane lipids, and that backbone hydrogen bonding is the major driving force for the stabilization of β-sheets in membranes.

The solution structure of reactive-site hydrolyzed turkey ovomucoid third domain (OMTKY3*) was determined by n.m.r. methods. A total of 655 distance constraints was applied in a distance geometry/simulated annealing approach to calculate a family of structures consistent with the n.m.r. data. The input data included 24 torsion angle constraints, 14 hydrogen bonds, 611 constraints derived from two-dimensional nuclear Overhauser enhancement spectroscopy data, and three disulfide bridges. Stereospecific assignments were included for the hydrogens of 26 β-methylene groups and for seven isopropyl methyl groups (46% chiral assignments). OMTKY3*in solution retains the global fold and overall secondary structure of the intact inhibitor (OMTKY3) but exhibits local structural differences at and adjacent to the clip site. In particular, the hydrogen-bonding network observed at the reactive-site of the intact inhibitor is disrupted, and the position of Tyr20 is altered in the modified inhibitor. No evidence was found for ion pairing between the oppositely charged termini at the clip site. Surprisingly, in light of numerous changes indicating that OMTKY3*is less stable than OMTKY3, rotation of the Tyr31 ring was found to be slow in OMTKY3*at 30°C. In OMTKY3, slow rotation of the Tyr31 ring was observed only at temperatures below 15°C. The n.m.r. structures of OMTKY3*are compared here with the similarly calculated structures of OMTKY3. This represents the first comparison of an intact and modified (reactive-site clipped) proteinase inhibitor under identical conditions. On comparison with published X-ray structures of modified avian ovomucoid third domains from two other species, the present structure of OMTKY3*in solution was found to resemble that of the Japanese quail protein (OMJPQ3*) more closely than that of the more closely homologous silver pheasant protein (OMSVP3*).