Mode of action of diptericin A a bactericidal peptide induced in the hemolymph of Phormia terranovae larvae.код для вставкиСкачать
Archives of Insect Biochemistry and Physiology 10:229- 239(1989) Mode of Action of Diptericin A, a Bactericidal Peptide Induced in the Hemolymph of Phormia terranovae Larvae Elisabeth Keppi, Anthony P. Pugsley, Jean Lambert, Claude Wicker, Jean-Luc Dimarcq, Jules A. Hoffmann, and Daniele Hoffmann Unite' Associe'e au Centre National de la Recherche Scientifique 672, Endocrinologie et Immunologie des Insectes, Laborafoire de Biologie Ge'ne'rale de I'Universite' Louis Pasteur, Strasbourg ( E .K., J.L., C .W., 1.-L.D., J.A.H., D. H.) and Unit6 de Gtnttique Mole'culaire, Institut Pasteur, Paris (E.K., A.P.P.), France Diptericin A is a member o f a multigenic family of antibacterial peptides that are synthesized by larvae of fhormia terranovae (Diptera) in response to a bacterial injection o r to injury. The 82-residue peptide is active only against a limited range o f Gram-negative bacteria. Data presented suggest that the primary action of diptericin A is on the cytoplasmic membrane of growing bacteria. Key words: insect immunity, antibacterial protein, bacterial cytoplasmic membrane INTRODUCTION Larvae of the dipteran insect Phorrnia terranovae synthesize several peptides with potent antibacterial activity [l]in response to bacterial challenge or injury. Three of these peptides have recently been isolated, and their amino acid sequences have been fully or partially determined [2,3]. They show significant sequence homologies and represent members of a family of new inducible antibacterial peptides that we have termed diptericins. They are basic (pl 7.8-8.5) heat-stable molecules with a molecular weight of 8,600 Da and containing high levels of Asx, Pro, and Gly. They are distinct from other inducible antibacterialpeptides isolated from Lepidopterans (lysozymes , attacins , and cecropins ) and from Dipterans (the cecropin-like sarcotoxins ; for reviews, see  and ). Received October 29,1988; accepted February 15,1989. Acknowledgments: We express our gratitude to R. Barker, A. Klier, C. Schnaitrnan, C. Elmerich, W. Lubitz, H.G. Bornan, C. Campelli, Y. Piernont, J.Millet, and L. Le Minor who kindly supplied the bacterial strains used in this study. We are indebted to R. Klock, A. Meunier, and C. Heyer for skillful technical assistance. We also wish to acknowledge the many helpful discussions with Prof. J. Fothergill (University of Aberdeen). Address reprint requests to Dr. D. Hoffrnann, Laboratoire de Biologie Generale-UA CNRS 672-12, rue de I'Universite 67000 Strasbourg, France. 0 1989 Alan R. Liss, Inc. 230 Keppi et al. Preliminary data on the mode of action of diptericins against the Gramnegative bacterium Escherichia coli  indicated that they are probably bacteriolytic and active only on growing cells. The present paper provides a more detailed analysis of the antibacterial spectrum of diptericins and of their mode of action. MATERIALS AND METHODS Preparation of Semipure Diptericin A (SP-Diptericin A) Diptericin A was semipurified from the hemolymph of immunized larvae of Phorrnia terranovae by heat treatment, cation exchange chromatography and gel permeation as previously described . A HPLC profile of an aliquot of SP*-diptericin showed that it consisted of 20% pure diptericin and contained no other antibacterialpeptide. The concentration of pure diptericin A was estimated to be roughly equal to 1pg per ml of immune hemolymph. Bacterial Strains and Media The strains used for the study of the antibacterial spectrum were obtained from several laboratories (see Text and "Acknowledgments"). The mode of action of diptericin A was primarily investigated on E. coli K12 strains BZBlOll (gyrA) and D31 (a mutant with a defective lipopolysaccharide). DAP uptake assays and peptidoglycan degradation assays were performed with E. coli K12 strain W7 (dap A, Lys). Bacteria were grown in L broth or minimal medium M63 containing 0.2% glucose as described by Miller [lo]. Xenorhabdus nernatophizus was grown at 30°C in 0.4% Oxoid bacteriological peptone containing 0.5% NaCl and 0.4% glucose (pH 7.4). Antibacterial Spectrum Assay conditions were essentially those described by Pugsley and Oudega [ll]. Soft agar (3 ml) was seeded with approximately lo6 indicator cells from an exponential phase culture and poured onto the surface of normal-strength nutrient agar in Petri dishes. Samples of SP-diptericin A (5 p1 containing the equivalent of approximately 1.5 pg of pure diptericin A) were deposited onto the indicator lawn. The plates were incubated for 8 h at 37"C, and the diameters of the clear zones were recorded. P-Galactosidase Assay E. coli D31 cells were grown to an absorbance of 0.1 at 600 nm in L broth medium containing 1mM IPTG (Sigma, St. Louis). SP-diptericinA was added at a final concentration equivalent to 3 pg/ml of pure substance, and samples (200 p1) were removed periodically and centrifuged at 3,500g for 5 min. The culture supernatants were kept on ice until further use. The pellets were resuspended in 200 p1 of L broth, and the cells were sonicated to release P-galactosidase. P-galactosidase in the supernatants and in the lysed cells was assayed according to Miller [101. *Abbreviations used: DAP = Diaminopimelic acid; 3H-DAP = (DL + meso) - 2,6-diamino[3,4,5-3H] pimelicacid; IPTC = isopropyl p-D-thiogalactopyranoside;LPS = lipopolysaccharide; SDS = sodium dodecyl sulfate; SP = semipurified. Mode of Action of Diptericin A 231 Degradation of 3H-DAP-LabeledPeptidoglycan Exponential phase E . coli W7 cells (5 x lo7) were prelabeled with 20 pl of 3H-DAP(36 Ci/mmol; 1mCi/ml) (CEA, Paris) and then grown in L broth containing unlabeled DAP (20 pg/ml) and SP-diptericinA at a final concentration equivalent to 24 pg/ml of pure substance. Samples (100 p1) were removed periodically, mixed with hot 10% SDS, and further heated at 100°Cfor 10 min. SDSinsoluble material was collected by filtration through Millipore membrane filters (HAWP, 0.45 pm pore diameter), washed with fresh medium, air-dried, and the radioactivity on the filter was determined by liquid scintillation counting. Measurement of Lysine and DAP Uptake Exponential-phase E. coli BZB 1011were suspended at a density of 5 x lo7 cells/ml in minimal M63 glucose medium at 37"C, and SP-diptericin A was added at a final concentration equivalent to 15 pg/ml of pure substance. Samples (400 pl) were removed at intervals and incubated at 37°C with 25 pg/ml chloramphenicol and 0.3 pCi of [14C]-lysine(336 mCi/mmol; Amersham) at a final concentration of 2 pM- Aliquots (100 pl) were removed at intervals and filtered through Millipore membrane filters. The radioactivity in the cells was measured on the filter by liquid scintillation counting. In a parallel experiment, the bacteria were killed by adding toluene, which permeabilized the bacterial envelope. The procedure used to measure 3H-DAPuptake was essentially as described above except that strain W7 was used. RESULTS Antibacterial Spectrum of Diptericin A The activity of SP diptericin A was tested against a battery of Gram-negative and Gram-positive bacteria. Among the Gram-negative strains, E . coli K12, Erwinia herbicola T, E . carotovora 113, Shigella dispar P15, Klebsiella pneumoniae UNF 5023, and Xenorhabdus nematophilus Xn21 all exhibited varying degrees of sensitivity. All other Gram-negative bacteria tested (two strains each of E . coli, Salmonella fyphimurium, and Enterobacter cloacae, and one each of Salmonella wien, S.enferitidis, S.derby, Serratia marcescens, Citrobacterfreundii, Aeromonas hydrophila, Pseudomonas aeruginosa, and Acinetobacter calcoaceticus) were fully resistant to diptericin A. Other than for the specific case of E. coli K 12 discussed below, there was no obvious feature (presence of capsule or long oligosaccharideson outer membrane lipopolysaccharides) that distinguished between diptericin A-resistant and -sensitive strains. Likewise, all Gram-positive bacteria tested (five strains of Bacillus subtilis, three strains of B. thuringiensis and one each of B. megaterium, Staphylococcus aureus, and Micrococcus luteus), were also fully diptericin A-resistant. E. coli K12 strain D31 is routinely used by us 111and by others  to study the action of antibacterial agents of insect origin because of its high level of sensitivity. This strain carries a mutation, which makes it highly sensitive to other agents that cannot normally cross the E . coli outer membrane or that do so only inefficiently . This phenotype is often caused by the production of truncated outer membrane LPS core sugars and consequent insertion of phos- 232 Keppi et al. pholipids into the outer leaflet of the outer membrane. We therefore tested two pairs of isogenic E . coli K12 strains carrying, respectively, transposon TnlO insertions in rfu genes involved in LPS core oligosaccharide biosynthesis and a A prophage carrying the wild-type allele (strains CS1716WCS1717X and CS1716/CS1717, respectively; strains generously supplied by C. Schnaitman). All four strains were sensitive to diptericin A, but CS1716X and CS1717X (lacking the A prophage) exhibited considerably greater sensitivity than the corresponding strains with the A prophage (i.e., rfa+). The same strains were also more sensitive to the detergents sodium desoxycholate and SDS, to EDTA, and to the antibiotics rifampicin and chloramphenicol. This result confirms that the outer membrane is a barrier to the penetration of diptericin A. No mutants of E. coli K12 survived diptericin A treatment either on plates or in liquid culture, and a battery of colicin and bacteriophage-resistantmutants of E . coli K12 [ l l ] were all diptericin sensitive. It therefore seems unlikely that diptericin uses any of the known outer membrane transport systems to penetrate the cell envelope. Diptericin A Causes Lysis of E. coli K12 SP-diptericin A was added at concentrations ranging from 0.75 pg/ml to 24 pg/ml to growing cells of E. coli strain D31, and the effects were monitored by measuring culture absorbance at 600 nm. As shown in Figure 1, growth was affected by concentrations as low as 0.75 pg/ml, and the cells lysed markedly at a concentration of 6 pg/ml. Interestingly, the lowest active concentrations of diptericin in these tests are inferior to the actual concentration of diptericin in 0.5 a 0 1 2 Time (h) Fig. 1. Dose-dependent effect of diptericin Aon the absorbance of E. coli D31cultures: 5 X 10’ bacteridml in exponential growth phase were incubated in the presence of SP-diptericin A at the concentrations of 0.75 (I), 1.5 (2), 3 (3), 6 (4),12 (5), and 24 pg/ml (6). The absorbance of the cultures at 600 nm was measured at different time intervals and compared to that of a control culture (C)where diptericin A was replaced by an equivalent volume of distilled water. Mode of Action of Diptericin A 233 the immune hemolymph (estimated to be 1 pg of pure diptericin per ml). Increasing concentrations of diptericin A had a stronger effect; not only did the culture absorbance decrease more dramatically, but the lag period between the addition of diptericin A and the decrease of absorbance was also reduced. These observations indicate that diptericin A causes lysis of E . coZi cells. Viability (ability to form colonies on L broth agar) declined after 30 min of treatment . The addition of SP-diptericin A to €. coli also caused a sharp decrease in the level of cytoplasmic P-galactosidaseafter 60 min, coincident with a sharp drop in culture absorbance (Fig. 2). These results indicate that diptericin A causes complete lysis of E . coli K12. To see whether lysis was caused by peptidoglycan breakdown, cells of E . coli strain W7 were prelabeled with 3H-DAP,which is incorporated exclusively in peptidoglycan. Subsequent addition of diptericin A caused the release of the label as hot SDS-soluble material after 2 h (Fig. 3). While this result confirms that peptidoglycan breakdown does occur, it does so only 30-60 min after the first detectable signs of lysis, as indicated by the release of P-galactosidase and the decline in culture absorbance. To see whether there were any detectableeffects on membrane function prior to this, we measured the rates of 3H-DAP and [14C]-lysineuptake in treated and control cells. Accumulation of [14C]-Lysineby the strain BZBlOll was almost immediately reduced upon addition of diptericin A (Fig. 4). Similar results were obtained with the strain W7 for the accumulation of 3H-DAP (data not shown). Thus, inhibition of active transport is the earliest detectable effect of diptericin A, preceding by 30 min any effect on viability and by 60 min any detectable signs of lysis. 0 1 2 Time (h) Fig. 2. Diptericin A-induced release of P-galactosidase from F. coli. E. coli D31 was grown in L broth supplemented with 1 mM IPTG to induce p-galactosidase synthesis; 3 p@nl SP-diptericin A was added to the culture, and 200 JLI aliquots were removed at time intervals and assayed for P-galactosidase activity. The figure represents the cell-associated activity expressed as percentage of the total activity present in the culture (cellsand medium). It also shows the evolution of the cell culture (increase up to 1 h, followed by a marked decrease) monitored by absorbance measurements. Keppi et at. 234 a 0 2 4 Time (h) Fig. 3. Effect of diptericin A on 3H-DAP-labeledpeptidoglycan in E. coli. Cells of E. coli W7 prelabeled with 3H-DAPwere grown to an absorbance of 0.1 at 600 nm in L broth containing 20 pg/ml of DAP SP-diptericinA (24 pg/ml) was added to the culture at time0 (-.-). A parallel experiment was run in which 20 m M MgS04were added together with diptericin A to the bacteria (-o-). Samples (100 pl) were removed at hires indicated and processed as described in “Materials and Methods.” The cell-associated radioactivity is expressed as a percentageof the radioactivity in the control without diptericin A. Mg+ Reduces the Lytic Effect of Diptericin A + The above results suggested that lysis was not the primary cause of death of the diptericin A-treated cells but occurred as a consequence of an earlier, less dramatic perturbation of membrane functions. We have previously observed that M g + + can reduce or eliminate lysis in such a situation . We therefore compared the effects of diptericin A in the presence or absence of varying concentrations of MgS04. Figure 5 shows that the addition of MgS04 to cultures of E. coli D31 reduced the lytic effect of diptericin A. This reduction was observed with concentrations of Mg++ ranging from 5 to 20 mM. The lytic effect of diptericin A was completely abolished when the concentration of Mg+ was higher than 20 mM. Mg+ did not prevent killing by diptericin, as determined by the plate count assay, nor did it affect the action of diptericin on amino acid uptake. The way in which Mg++ exerts its effect remains unkown, but it presumably stabilizes the outer membrane rather than acting as an osmoprotectant. + + Potentiation of the Lytic Effect of Diptericin A by Triton X-100 When E. coli cells were pretreated with the equivalent of 0.6 pg/ml pure diptericin A, they became sensitized to 0.1% Triton X-100. The treated cells lysed more rapidly and more extensively upon the addition of the detergent than in its absence. When the detergent was added 15min after the diptericin, there was a considerable delay before lysis occurred, whereas there was no delay when Triton was added 45-60 min after the diptericin. Detergent alone Mode of Action of Diptericin A - 0 2 5 10 235 15 Time (min) Fig. 4. Effects of diptericin A on [14C]-lysinetransport in E. coli. E. coli BZB 1011 was grown in M63 medium containing 0.4% glucose to an absorbance of 0.1 at 600 nm. SP-diptericin A (15 pg/ml) was added, and 400 pI samples were removed at times indicated, supplemented with chloramphenicol (25 pg/ml) and 2 pM [14C]-lysine(0.3 pCi/sample). After 1 min, the samples were filtered through nitrocellulose filters, which were dried and counted. Hatched columns: control culture without diptericin: dotted columns: diptericin treated culture; open columns: killed bacteria. did not cause lysis. This result indicates that diptericin and Triton X-100 act synergistically (Fig. 6). DISCUSSION Diptericin A is an antibacterial peptide that is effective only against a limited range of Gram-negative bacteria. Gram-positive bacteria and eukaryotic cells (e.g., sheep red blood cells) are not affected by diptericin A. Mutants of E. coZi with modified LPS exhibit the highest sensitivity, while two "smooth" strains with long LPS oligosaccharide side chains (E. coZi serotypes 06 and 010) were completely resistant, possibly because diptericin A could not penetrate the cell surface (see ). These data indicate that diptericin A alone would not protect insect larvae against general microbial attack. However, Phormia larvae produce other agents that might act together with diptericin A to provide fuller protection. it is interesting to note that other antibacterial agents isolated from the hemolymph of immunized P. terranova larvae have different activity spectra and that one of them is related to cecropin [3,16]. Figure 7 shows the kinetics of the various effects observed in this study. Amino acid transport is affected within a very short time. The viability decreases after 30 min (the effects of diptericin A on membrane functions are presumably reversible up to this time), and lysis and the release of cytoplasmic P-galactosidase follow after a further 30 min. Eventually, the cell wall 236 Keppi et al. 2.8 C' C 4 3 2 1 A I 0 1 2 Time (h) Fig. 5. Effect of diptericin A on €. coli in the presence of varying concentrations of M g + + . E. coli D31 in exponential growth phase was incubated with SP-diptericin A (3 @ml) in L broth medium at 37°C; the following concentrations of MgS04were added: 2.5 m M (1); 5 m M (2); 10 m M (3); 20 m M (4); diptericin Alone (A); control (C); incubationwith 20 m M MgS04(C'). disintegrates as indicated by the release of DAP. Diptericin A therefore appears to cause a succession of events resulting most probably from a primary effect on the functioning of the cytoplasmic membrane. This would be consistent with the synergistic effects of diptericin A and Triton X-100. Previous studies showed that Triton X-100 potentiated the effect of another antimicrobial agent, microcin E492, which also seems to affect the E . coli cytoplasmic membrane 1171. Among the antibacterial peptides isolated so far from insects, some information on the mode of action is available for cecropins and the cecropin-like sarcotoxins I, two groups of related molecules isolated, respectively, from Hyulophoru cecropia  and Surcophugu peregrinu [18,19], the latter species being relatively close to P. terrunovue. Cecropins and sarcotoxins act on both Grampositive and Gram-negative bacteria, which is in contrast to the narrow activity spectrum of diptericin A. Moreover, diptericin A acts only on growing bacteria , which is not the case for cecropins or sarcotoxins I. Studies with cecropin analogues indicate the importance of specific residues in the molecule . The bactericidal effect of the cecropin-like sarcotoxin Mode of Action of Diptericin A 237 3 h E c A 0 0 c(3 v 15 35 45 60 0 1 2 Time (h) Fig. 6. Effect of Triton X-I00 o n the absorbance of E. coli in the presence of diptericin A: 1.5 pg of SP-diptericin A was added to 500 pl of an exponentially growing culture of E. coli BZB 1011 containing 5 x lo7 bacteria. Triton X-100 at a final concentration of 0.1%was added to the culture after either 15,35,45, or 60 min, and the absorbance at 600 nm was recorded. C, C’: control in the presence or in the absence of Triton; A: diptericin A alone. I is probably due to its ionophore activity and to a block of ATP generation caused by the dissipation of the proton gradient [18,21]. The bacteriostatic properties of the attacins  isolated from immune hemolymph of H.cecropia rule out a possible analogy with the mode of action of diptericin A, which.is initially bacteriostatic but becomes bactericidal after 30 min. Likewise, a similarity with lysozyme, which has been isolated from many insect species, was ruled out by the specific action of this enzyme, which hydrolyses p-(1-4)glycosidicbonds between N-acetylglucosamine and N-acetyl-muramicacid units of the bacterial wall. Diptericin A does not exhibit this activity (not shown). Thus, we conclude that diptericin A probably acts at the level of the cytoplasmic membrane of susceptible bacteria, leading eventually to death and cell lysis. Interaction with a component involved in active amino acid uptake would be consistent with the absence of action of diptericin A on resting cells. The effects of the diptericin appear to be reversible during the first 30 min, implying that diptericin is diluted out when the cells are plated out. We note that high concentrations of diptericin A (relative to other characterized antibacterial agents of insect origin) are required to kill even the highly sensitive 238 Keppi et at. 'c 0 c 0 a v) cn.(I) % - m W i 0 W n 02 30 60 120 Time (min) Fig. 7. Kinetics of the effects of diptericin A o n f. coli cultures. E. coIi K12 strains such as D31. Other antibacterial agents with similar models of action (e.g., colicin A, E l and I , magainins , cecropins , and the cecropin-like sarcotoxins [191have either hydrophobic a-helices or amphiphilic helices in their interactive domains. These helices are proposed to span the cytoplasmic membrane and in some cases are known to form channels through which ions diffuse across the membrane. Structure predictions for diptericin A, the primary sequence of which has been determined , show the complete absence of segments of significant hydrophobicity. Helix-breaking proline residues are present along the entire length of the molecule, making it unlikely that any part of it could span the cytoplasmic membrane as an a-helix (data not shown). Thus, the mode of action of diptericin A may be different from that of any previously characterized, membrane-active antibacterial agent. LITERATURE CITED 1. Keppi E, Zachary D, Robertson M, Hoffmann D and Hoffmann J: Induced antibacterial proteins in the hemolymph of Pkormia terranovae (Diptera). Insect Biochem 26,395 (1986). 2. Keppi E, Dimarcq JL, Lambert J, Zachary D, Reichhart JM, Hoffmann D, Keller R, Hoffmann J: Recherches sur les mecanismes de defense antibacterienne chez les Insectes: Isolement de peptides antibacteriens dans I'hemolymphe du diptPre Pkormia terranovae. CR Acad Sci (Paris) ser. 111303. 155 (1986). 3. Dimarcq JL, Keppi E, Dunbar B, Lambert J, Reichhart JM, Hoffmann D, Rankine SM, Fothergd JE, Hoffmann JA: Insect immunity. Purification and characterization of a family of novel inducible antibacterial proteins from immunized larvae of the dipteran Pkormia terranovae Mode of Action of Diptericin A 239 and complete amino-acid sequence of the predominant member, diptericin A. Eur J Biochem 171,17 (1988). 4. Engstrom A, Xanthopoulos KG, Boman HG, Bennich H: Amino acid cDNA sequences of lysozyme from Hyalophora cecropia. EMBO J 4,2119 (1985). 5. Hultmark D, Engstrom A, Andersson K, Steiner H, Bennich H, Boman HG: Insect immunity. Attacins, a family of antibacterial proteins fromHyaZophoracecropia.EMBO J 2,571 (1983). 6. Hultmark D. Steiner H, Rasmuson T, Boman HG: Insect immunity. Purification and properties of three inducible bactericidal proteins from hemolymph of immunized pupae of Hyalophora cecropia. Eur J Biochem 106,7 (1980). 7. Okada M, Natori S Purification and characterizationof a bactericidal protein from hemolymph of Sarcophagu peregrinu (flesh-fly)larvae. Biochem J 211, 727 (1983). 8. Dunn PE: Biochemical aspects of insect immunology. Annu Rev Entomol31,321(1986). 9. Boman HG, Hultmark D: Cell-free immunity in insects. Annu Rev Microbiol41, 103 (1987). 10. Miller JH: Experiments in Molecular Genetics. Cold Spring Harbor Laboratory Cold Spring Harbor, NY, (1972). 11. Pugsley AP, Oudega B: Methods for studying colicins and their plasmids. In: Plasmids, Practical approach. Hardy K, ed. IRL Press, Oxford, UK, p 105 (1987). 12. Boman HG, Steiner H: Humoral immunity in Cecropia pupae. Curr Top Microbiol Immunol 94-95,75 (1981). 13. Boman HG, Jonsson S, Monner D, Nomark S, Bloom GD: Cell-surface alterations in Escherichia coli K-12 with chromosomal mutations changing ampicillin resistance. Ann NY Acad Sci 182, 342 (1971). 14. van der Ley P, de Graaf P, Tommassen J: Shielding of Escherichiu coli outer membrane proteins as receptors for bacteriophages and colicins by 0-antigenic chains of lipopolysaccharide. J Bacterioll668,449 (1986). 15. Pugsley AP, Schwartz M: Colicin E2 release: Lysis, leakage or secretion? Possible role of a phospholipase. EMBO J 3,2393 (1984). 16. Lambert J, Keppi E, Dimarcq JL, Wicker C, Reichhart JM, Dunbar B, Lepage P, Van Dorsselaer A, Hoffmann J, Fothergill J, Hoffmann D: Insect immunity: Isolation from immune blood of the Dipteran Phormia terranovae of two insect antibacterial peptides with sequence homology to rabbit lung macrophage bactericidal peptides. Proc Natl Acad Sci USA 86,262 (1989). 17. de Lorenzo V, Pugsley AP: Microcin E492, a low molecular-weight peptide antibiotic which causes depolarization of the Escherichiu coli cytoplasmic membrane. Antimicrob Agents Chemother 27,666 (1985). 18. Okada M, Natori S: Mode of action of a bactericidal protein induced in the hemolymph of Sarcophaga peregrinu flesh-fly) larvae. Biochem J 222,119 (1984). 19. Okada M, Natori S: Primary sturcture of sarcotoxin I, an antibacterial protein induced in hemolymph of Sarcophugu peregrinu. Biochem J 260,7174 (1985). 20. Andreu D, Merrifield RB, Steiner H, Boman HG: N-terminal analogues of cecropin A: Synthesis, antibacterial activity, and conformational properties. Biochemistry 24,1683 (1985). 21. Okada M, Natori S: Ionophore activity of a sarcotoxin I, a bactericidal protein of Sarcophaga peregrinu. Biochem J 229,453 (1985). 22. Engstrom P, Carlsson A, Engstrom A, Tao ZJ, Bennich H: The antibacterial effect of attacins from the silk moth Hyalophoru cecropia is directed against the outer membrane of Escherichiu coli. EMBO J 3,3347 (1984). 23. Pattus F, Hertz F, Martinez C, Provencher SW, Lazdunski C: Secondary structure of the poreforming colicin A and the C-terminal fragment. Experimental fact and structure prediction. Eur J Biochem 152,681 (1985). 24. Zazloff M: Magainins, a class of antimicrobial peptides from Xenopus skin: Isolation, characterization of two active forms and partial cDNA sequence of a precursor. Proc Natl Acad Sci USA 84, 5449 (1987). 25. Steiner H: Secondary structure of cecropins: antibacterial peptides from the moth Hyalophoru cecropia. FEBS Lett 137,283 (1982).