Specific Peptide-Peptide Interactions and Lateral Crowding in Membranes

1. COOH H2N S S S S S S S S S S T T T T T T T T T T I I I I I I I I I I -10 -30 -50 -70 -90 -110 -130 -150 CF3 d /ppm Results: Antimicrobial activity 0 0 K19 G1 A8 K12 K15 K5 -10 -10 -30 -30 -50 -50 -70 -70 -90 -90 -110 -110 -130 -130 -150 -150 d /ppm d /ppm S4 V16 GA G11 GS I9 L18 M2 A20 I13 G7 A14 A6 PGLa A3 A17 A10 L21 1:200 1:40 PGLa:lipid MAP MAG2 S S S S GA GS S PGLa T T MAP MAG2 I I GS: surfacial crowding PGLa control MAP: helix-helix interactions? GA: transmembrane crowding MAG2: synergism 1:200 1:200 1:200 1:200 1:200 PGLa:lipid PGLa:lipid PGLa:lipid PGLa:lipid PGLa:lipid 1:40 1:40 1:40 1:40 1:40 55°C 50°C 45°C 40°C 35°C 30°C 25°C 20°C 15°C -10 -30 -50 -70 -90 -110 -130 -150 -10 -30 -50 -70 -90 -110 -130 -150 -10 -30 -50 -70 -90 -110 -130 -150 -10 -30 -50 -70 -90 -110 -130 -150 0 -10 -30 -50 -70 -90 -110 -130 -150 -10 -30 -50 -70 -90 -110 -130 -150 -10 -30 -50 -70 -90 -110 -130 -150 -10 -30 -50 -70 -90 -110 -130 -150 d /ppm d /ppm d /ppm d /ppm d /ppm d /ppm d /ppm d /ppm Specific Peptide-Peptide Interactions and Lateral Crowding in Membranes Revealed by Solid State 19F NMR Stephan L. Grage1, Sergii Afonin1, Marco Ieronimo2, Marina Berditsch2, Parvesh Wadhwani1, Pavel K. Mikhailiuk2,3, Igor V. Komarov3, Anne S. Ulrich1,2 1Institut für Biologische Grenzflächen, Forschungszentrum Karlsruhe, Hermann-von-Helmholtz-Platz 1, 76021 Karlsruhe, Germany, 2Institut für Organische Chemie, Universität Karlsruhe, Fritz-Haber-Weg 6, 76131Karlsruhe, Germany, 3Organic Chemistry, Taras Shevchenko University, Kiev, 252033 Ukraine The Peptide PGLa 19F NMR to probe the peptide state Summary The insertion state of PGLa was monitored using 19F-solid state NMR on oriented samples. To this aim we labelled PGLa with CF3-bicyclopentylglycine, which links the CF3-label to the peptide backbone in a rigid way [4]. The activity of antimicrobial peptides relies on the interaction with the target membrane, where the interplay with the lipids, but also with proteinaceous components could be important. E.g. other antimicrobial peptides can modulate or synergisti-cally enhance activity [1]. To gain insight in the influence of lateral crowding and specific peptide interactions on the antimicrobial mechanism, we analyzed the behaviour of PGLa in the presence of other peptides mimicking a protein-rich membrane. We found a pronounced change of the PGLa activity in the presence of Magainin-2, and confirmed the synergism of these two peptides [1]. The other peptides did not induce equivalent changes in the PGLa behaviour, indicating a specific interaction of PGLa and Magainin-2. We tested the effect of proteins on the activity of antimicrobial peptides using PGLa as an example. This a-helical peptide, derived from frog skin, forms a polar side with 4 lysines, opposed by an alanine-rich hydrophobic side. Different states of insertion into DMPC membranes were found recently for PGLa [2,3]. At low protein:lipid ratios, PGLa lies flat on the bilayer surface (S-state). At high concentrations PGLa is surface aligned only at high temperatures. At temperatures just above the lipid phase transition, the peptide tilts into the membrane (T-state). Below the phase transition, PGLa inserts completely in a nearly upright position (I-state). The PGLa activity might be related to dimers formed in the T and I states. As a first step we evaluated which influence the peptides, used in this NMR study to mimick crowding, have on the antimicrobial activity of PGLa. PGLa combined with Magainin-2 (MAG2) leads to exhibition zones typical for a synergistic behaviour (see yellow arrows), whereas all other peptides do not change PGLa activity. In oriented samples, solid state NMR resonances become orientation depen-dent. Thus, position and splittings of the CF3-triplet signal reflect the orien-tation of PGLa. This way, all three states result in distinct 19F-NMR spectra. Escherichia coli DH5a (gram-) 50°C Natural membrane Micrococcus luteus ATCC 4698 (gram+) 35°C Different antimicrobial activity in the presence of proteins or peptides? Model membrane 15°C Samples were prepared with two PGLa concentrations: at low PGLa:DMPC (1:200), where only the S-state states occurred in the absence of a second peptide, and at high PGLa:DMPC (1:40), where the inserted T and I states were found. The total peptide:lipid ratio was 1:20. The insertion state of PGLa was followed by 19F-NMR in the presence of a second peptide. We used gramicidin A (GA), gramicidin S (GS), a model-amphipatic peptide (MAP) and magainin-2 (MAG2) to mimick crowding in different regions of the bilayer, and to probe specific interactions. Results: 19F-NMR Conclusion 0 The behaviour of PGLa changes only marginally in the presence of gramicidin A, gramicidin S or MAP. Magainin-2 on the other hand turns PGLa into the inserted I-state even at low PGLa concentrations. These results indicate that the presence of membrane proteins as such (“crowding”) has only little effect on the activity of antimicrobial peptides. Only when paired with a particular partner, such as magainin-2, the activity changes profoundly. The basis for the synergistic activity enhancement between PGLa and magainin-2 thus seems to be a specific interaction between these peptides. Acknowlegdements We thank Erik Strandberg for providing his data analysis software. The DFG (CFN Karlsruhe) and Helmholtz Association are acknowledged for the financial support. Literature [1] P. Tremouilhac et al (2006), J. Biol. Chem. 281, 32089-94 [2] R. Glaser et al (2005), Biophysical Journal88, 3392-3397 [3] P. Tremouilhac (2006) BBA Biomembranes1758, 1330-1342 [4] P.K. Mikhailiuk et al (2006) Angewandte Chem.45, 5659-5661 Contact: stephan.grage@ibg.fzk.de