Hydrophobic effects on antibacterial and channel-forming properties of cecropin A-melittin hybrids.

The design of cecropin-melittin hybrid analogues is of interest due to the similarities in the structure of the antimicrobial peptides cecropin and melittin but differences in their lytic properties. We suspected that a hydrophobic residue in position 2 of milittin (Ile8 in the hybrid) plays an important role in the activity of the 15-residue hybrid, KWKLFKKIGAVLKVL-NH2, [CA(1-7)M(2-9)NH2] and have now examined its role in the analogue toward five test bacteria. Deletion of Ile8 reduced activity, and it was not restored by lengthening to 15 residues by addition of another threonine at the C-terminus. Replacement of Ile8 by a hydrophobic leucine maintained good activity and Ala8 was equally active for four organisms, although less active against Staphylococcus aureus. Replacement by the hydrophilic Ser8 strongly reduced potency against all five organisms. Deletion of Leu15 decreased activity, but addition of Thr16 maintained good activity. The presence of hydrophobic residues appears to have a significant effect on the process of antibacterial activity. These peptide analogues showed voltage-dependent conductance changes and are capable of forming ion-pores in planar lipid bilayers. The antibacterial action of the peptides is thought to be first an ionic interaction with the anionic phosphate groups of the membrane followed by interaction with the hydrocarbon core of the membrane and subsequent reorientation into amphipathic alpha-helical peptides that form pores (ion-channels), which span the membrane. The analogue also showed an increase in alpha-helicity with an increase in hexafluoro 2-propanol concentration.

D-enantiomers of 15-residue cecropin A-melittin hybrids.

The all-D enantiomers of six 15-residue hybrids of cecropin A and melittin were synthesized. They contained the seven N-terminal residues of cecropin A, followed by eight residues from the N-terminal region of melittin. They were pure and of the correct composition and structure. The peptides were compared with their all-L enantiomers. The L and D isomer pairs were each exact mirror images by circular dichroism at several concentrations of hexafluoroisopropanol, and at 12 or 20% were highly helical. The L analogs were rapidly hydrolyzed by trypsin but the D analogs were very resistant, making them suitable candidates for orally active drugs. These 15-mers did not form ion channels in normal lipid bilayers made in decane, but those bilayers made in squalene were thinner and the peptides did form ion-conducting channels. The D/L pairs of peptides were very active antibiotics against five representative Gram-negative and Gram-positive bacteria. In each case the D and L isomers were essentially equally active within experimental error. This is interpreted to mean that the peptides do not act by tight interactions with chiral receptors, enzymes or lipids. The action of these peptides against these organisms is best explained by self-aggregation and the formation of ion-conducting pores across bacterial membranes.

Design and synthesis of antimicrobial peptides.

The cecropins are a group of potent antimicrobial peptides, initially discovered in insects but later found in other animals including mammals. Synthetic peptide chemistry has played an important role in establishing their primary sequences, as well as the steps in the processing of the biosynthetic preprocecropins. Solid-phase peptide synthesis has been the method of choice. Synthetic chimeric peptides have led to more active products and a better understanding of their mode of action. The structural requirements for high activity include a basic amphipathic N-terminus, a short central flexible sequence and a hydrophobic helical C-terminus. Cecropin-melittin hybrids as small as 15 residues are highly active. In planar lipid bilayers the cecropins form pores which pass ions and carry a current under a voltage gradient. Synthetic D-enantiomers of several antibacterial peptides carry the same current as the natural all-L-peptides and are equally active against several test bacteria. Therefore, the activity is not dependent on chiral interactions between the peptides and the lipid bilayers or the bacterial membranes. Recent examination of retro and retroenantio peptides has further defined the limits of the structural requirements of these peptides. Some of the hybrid peptides are active against Plasmodium falciparum and Mycobacterium smegmatis.