Characterization of Antibacterial COOH-terminalProenkephalin-A-derived Peptides (PEAP) in Infectious FluidsIMPORTANCE OF ENKELYTIN, THE ANTIBACTERIAL PEAP209–237 SECRETED BY STIMULATEDCHROMAFFIN CELLS*

(Received for publication, February 5, 1998, and in revised form, June 14, 1998)

Yannick Goumon‡, Karine Lugardon‡, Bruno Kieffer§, Jean-Francois Lefevre§,Alain Van Dorsselaer¶, Dominique Aunis‡, and Marie-Helene Metz-Boutigue‡i

From ‡INSERM, Unite 338 de Biologie de la Communication Cellulaire, Strasbourg, France, §CNRS, UPR 9003,Cancerogenese et Mutagenese Moleculaire et Structurale, Illkirch Graffenstaden, France, and ¶CNRS, URA 31, Laboratoire deSpectrometrie de Masse Bioorganique, Chimie Organique des Substances Naturelles, Strasbourg, France

Proenkephalin-A (PEA) and its derived peptides(PEAP) have been described in neural, neuroendocrinetissues and immune cells. The processing of PEA hasbeen extensively studied in the adrenal medulla chro-maffin cell showing that maturation starts with the re-moval of the carboxyl-terminal PEAP209–239. In 1995, ourlaboratory has shown that antibacterial activity is pres-ent within the intragranular chromaffin granule matrixand in the extracellular medium following exocytosis.More recently, we have identified an intragranular pep-tide, named enkelytin, corresponding to the bisphospho-rylated PEAP209–237, that inhibits the growth of Micro-coccus luteus (Goumon, Y., Strub, J. M., Moniatte, M.,Nullans, G., Poteur, L., Hubert, P., Van Dorsselaer, A.,Aunis, D., and Metz-Boutigue, M. H. (1996) Eur. J. Bio-chem. 235, 516–525). As a continuation of this previousstudy, in order to characterize the biological function ofantibacterial PEAP, we have here examined whetherthis COOH-terminal fragment is released from stimu-lated chromaffin cells and whether it could be detectedin wound fluids and in polymorphonuclear secretionsfollowing cell stimulation. The antibacterial spectrumshows that enkelytin is active against several Gram-positive bacteria including Staphylococcus aureus, butit is unable to inhibit the Gram-negative bacteriagrowth. In order to relate the antibacterial activity ofenkelytin with structural features, various synthetic en-kelytin-derived peptides were tested. We also propose acomputer model of synthetic PEAP209–237 deduced from1H NMR analysis, in order to relate the antibacterialactivity of enkelytin with the three-dimensional struc-ture. Finally, we report the high phylogenetic conserva-tion of the COOH-terminal PEAP, which implies someimportant biological function and we discuss the puta-tive importance of enkelytin in the defensive processes.

Secretory granules from adrenal medullary chromaffin cellscontain a complex mixture of low molecular mass constituentssuch as catecholamines, ascorbate, nucleotides, calcium, andseveral water-soluble peptides and proteins. These componentsare released into the circulation in response to splanchnicnerve stimulation. Since relatively large amounts of proen-kephalin-A (PEA)1 and chromogranin-derived peptides arefound in adrenal medullary chromaffin granules, these or-ganelles have proven to be an excellent model to study intra-granular processing of these proteins. Recently, we have char-acterized the processing of bovine chromogranins A and B inchromaffin granules and in the extracellular medium followingtheir release from stimulated cultured chromaffin cells (1, 2).

PEA, the precursor protein of Met- and Leu-enkephalin, aswell as larger enkephalin-containing peptides, is highly con-served from Xenopus (3) to human (4). Originally, PEA mRNAwas described to be present in various brain regions, mostnotably in the striatum (5) as well as in neuroendocrine tissues,the pituitary (6), and adrenal gland (6, 7). In addition to theirexpression in neural tissues, PEA and its derived peptides(PEAP) are expressed in a variety of immune cells, includingConA-stimulated CD4 T lymphocytes (8), CD4 thymocytes (9),B lymphocytes (10), as well as T cell lines, macrophages, andmast cells (11). In adult thymocytes and T lymphocytes clones,PEA mRNA is not expressed constitutively, but is detectedfollowing cell activation. After exogenous administration, en-kephalins affect several immunologic functions, including an-tibody production (12), NK cell activity against tumors andviral infections (13), macrophage and polymorphonuclear leu-kocyte functions (14, 15), graft rejections (16), and mitogen-stimulated lymphocyte proliferation (17). Recently, it wasshown that very low concentrations of PEA and Met-enkepha-lin differentially affect IgM and IgG production by B cells (18).Thus, enkephalins can enhance or inhibit particular immunefunctions (13, 19). Moreover, in several studies, bidirectionaleffects were reported: low concentrations of enkephalins en-hance, whereas higher concentrations inhibit the same im-mune function. Thus, it is generally accepted that enkephalinsact as modulators of immune reactions, although their physio-logical function in the immune system remains unclear. Inaddition to its expression in cells of the immune system, PEAmRNA is expressed in other tissues, such as those comprising

* This work was supported by grants from INSERM, the Directiondes Recherches, Etudes et Techniques Contract number 96-099 (toD. A.), the Ligue Contre le Cancer (to M. H. M. B.), CNRS, the Univer-site Louis-Pasteur de Strasbourg (ULP, Federation de Recherche Neu-rosciences), the Region Alsace Contract number 96 901 13 619 97 (toY. G.), and the Association Recherche et Partage (to K. L). The costs ofpublication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked “advertisement”in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

i To whom correspondence should be addressed: Unite INSERMU-338, 5, rue Blaise Pascal, 67084 Strasbourg Cedex, France. Tel.:33-3-88-45-66-09; Fax: 33-3-88-60-08-06; E-mail: metz@neurochem.u-strasbg.fr.

1 The abbreviations used are: PEA, proenkephalin-A; HPLC, highperformance liquid chromatography; MALDI-TOF MS, matrix-assistedlaser desorption ionization time-of-flight mass spectrometry; NOE, nu-clear Overhauser effect; NOESY, nuclear Overhauser effect spectros-copy; PEAP, proenkephalin-A-derived peptides; PMNs, polymorphonu-clear neutrophils; PAGE, polyacrylamide gel electrophoresis.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 273, No. 45, Issue of November 6, pp. 29847–29856, 1998© 1998 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org 29847

by guest on March 31, 2016

http://ww

w.jbc.org/

Dow

nloaded from

the reproductive system (20, 21), heart (22, 23), and in manydeveloping tissues during gestation and the early postnatalperiod (24, 25). Hence, it has been postulated that PEAP mayplay a role in cell or tissue growth and differentiation. Recently,it has been reported that endogenous enkephalins induced inthymocytes, modulate their own expression and function toinhibit the proliferation of activated thymocytes (26).

Natural processing of PEA has been extensively studied.Since 1982, it has been well established that several opioidpeptides including Met-enkephalin and Leu-enkephalin in theratio 4:1, two COOH-terminal extended variants, Met-en-kephalin-Arg-Phe7 and the octapeptide Met-enkephalin-Arg-Gly-Leu8 are liberated by cleavage of the precursor at pairs ofbasic residues. In these studies, high concentrations of COOH-or NH2-terminal extended variants of these peptides have beenfound in bovine adrenal medullary chromaffin cells (27, 28).More recently, the processing of PEA has been well examinedin adrenal medulla chromaffin cells (29), as well as in stablytransfected mouse anterior pituitary tumor (AtT-20) cells (30),showing that PEA maturation proceeds through an orderlyseries of steps. Similarly to other precursors, PEA maturationappears to start with the removal of the carboxyl-terminalfragment, named peptide B (30, 31), corresponding toPEAP209–239 (32). Four peptide B variants were isolated frombovine adrenal medulla corresponding to the unmodified formand to this PEAP209–239 containing 1, 2, or 3 phosphate groups(33, 34). These three phosphorylation sites are clustered to-gether at positions Ser215, Ser221, Ser223 and the adjacentacidic residues have been highly conserved during evolution.Interestingly, immunoreactive forms of this peptide can befound in various regions of rat brain and circulating in bovineplasma (35).

Our laboratory has recently shown that antibacterial activityis present within the intragranular chromaffin granule matrixand the extracellular medium following exocytosis. The firstpeptide was identified as secretolytin (2, 36), a peptide corre-sponding to the COOH-terminal sequence of bovine chromogra-nin/secretogranin I (CGB614–626). Further studies have re-vealed the antibacterial activity of a large natural CGAfragment (CGA79–431), named prochromacin (37), which is gen-erated by natural cleavage at the previously described site78–79 and released during exocytosis (1). Then, we have iden-tified chromacin-(G, P, and GP), the O-glycosylated and/orphosphorylated CGA-derived fragment (CGA173–194), as theshortest antibacterial CGA-derived fragment included inprochromacin. Secretolytin and chromacin inhibit the growthof Gram-positive bacteria (Micrococcus luteus and Bacillusmegaterium) in the micromolar range. In addition, antibacte-rial assays on soluble chromaffin granule material recoveredfrom HPLC indicated the presence of several other endogenouspeptides with potent antibacterial activity. Thus, among thecomplex mixture of intragranular matrix components, a pep-tide corresponding to the bisphosphorylated PEAP209–237 wasidentified (38). This new natural antibacterial peptide inhibitsthe growth of M. luteus in the 0.2 mM range, but has no effect ona Gram-negative bacteria, Escherichia coli (strain D22) at thesame concentration and does not lyse bovine erythrocytes. Cat-echolamines and glucocorticoids play key roles in stress situa-tions. Since these new antibacterial chromogranin-derived pep-tides and PEAP are stored with catecholamines, they may bereleased during stress and serve as an early additional protec-tive barrier against bacterial infection. As a continuation of ourprevious work (38), we now examine whether enkelytin is re-leased from stimulated chromaffin cells and polymorphonu-clear neutrophils (PMNs). Furthermore, a potential class ofagents that can simultaneously reduce infection and influence

the action of growth factors, matrix components, and othercellular effectors has recently been implicated in wound repair.Thus, antibacterial peptide PR-39, initially identified in pigintestine kills bacteria as a non-immune defense mechanism(39) and induces mammalian cells to express cell surface hepa-ran sulfate proteoglycans (40) which are involved in the woundrepair process (41). Since PEAP also affect cell and tissuegrowth (24, 25), we decided to analyze infectious fluids withrespect to the antibacterial potency of these peptides.

In addition, various natural and synthetic enkelytin-derivedpeptides were prepared and tested to identify the structuralfeatures necessary for a potent antibacterial activity toward M.luteus. In 1996, according to the Homolog method provided inPro-Explore, we reported comparative predictions of secondarystructure of enkelytin (38) and the homologous diazepam-bind-ing inhibitor-derived peptide (42), suggesting an amphipathichelical structure for PEAP224–237. Here, we generate a com-puter three-dimensional structure for the synthetic PEAP209–

237 on the basis of our 1H NMR study and discuss these struc-tural features in relation to the antibacterial activity ofenkelytin. Finally, the phylogenetic features of the highly con-served enkelytin are reported on the basis of the alignment ofPEA198–239 (according to bovine sequence) from several species,and discussed in terms of enkelytin biological importance.

EXPERIMENTAL PROCEDURES

Isolation of Peptides and Proteins Released from Stimulated CulturedCells—Chromaffin cells were isolated from fresh bovine adrenal glandsand cultured as described previously (1). Cells were plated at a densityof 107 cells/50-mm in plastic Petri dishes. After 3 days in culture, themedium was removed and cells were washed four times with Locke’ssolution (140 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO4, 2.5 mM CaCl2, 11mM glucose, 0.5 mM ascorbic acid, 15 mM Hepes, pH 7.5) and subse-quently stimulated for 10 min with Locke’s solution containing 10 mM

nicotine. External medium was carefully collected, completed with tri-fluoroacetic acid up to 0.1%. Extracellular medium was lyophilized andstored at 220 °C.

Isolation of Peptides and Proteins Released from PolymorphonuclearNeutrophils—Human PMNs were prepared to 98% homogeneity, asdescribed previously (43), from buffy coats of healthy donors of eithersex, kindly provided by the Center de Transfusion Sanguine de Stras-bourg (France). PMNs were suspended in a buffer solution containing140 mM NaCl, 5 mM KCl, 1.1 mM CaCl2, 0.1 mM EGTA, and 10 mM

Hepes, pH 7.3, at 5 3 106 cells per ml. Exocytosis of the content of thespecific and primary granules of PMNs was initiated at room temper-ature by application of 2.3 nM LukS-PV and 0.6 nM LukF-PV, the twocomponents of leukocidin from Staphylococcus aureus (44). The secre-tion was monitored by flow cytometry as described previously (45) and,when completed, PMNs were centrifuged (800 3 g) for 10 min. Thesupernatant was recovered for further analysis.

Isolation of Proteins from Periarthritis Abscess Fluids—Fluid col-lected from natural bovine knee periarthritis abscess was extractedwith 1 M acetic acid (v/v). After centrifugation at 12,000 rpm during 15min at 4 °C, the supernatant was collected and the soluble material wassuccessively filtered through Millex filters 0.45 mm and 0.22 mm andthen loaded on a HPLC column.

Purification of PEAP by Reverse Phase HPLC—PEAP were isolatedfrom cell secretion and abscess fluids using the Applied BiosystemsHPLC system 140 B. Reverse phase HPLC were successively performedon Macherey-Nagel Nucleosil columns. In some experiments, a finalpurification was performed on a Brownlee C18 column (0.5 3 150 mm;particle size 5 mM and pore size 300 Å). Absorbance was monitored at214 nm and the solvent system consisted of 0.1% (v/v) trifluoroaceticacid in water (solvent A) and 0.1% (v/v) trifluoroacetic acid in acetoni-trile (solvent B). Each HPLC elution was performed using a flow rateand gradient as indicated or shown on chromatogram.

Western Blot Analysis—Extracts of biological fluids were separatedby SDS-PAGE gels containing 17% acrylamide (46). In order to detectimmunologically reactive fragments, proteins were electrically trans-ferred to nitrocellulose sheets (47). Electrophoretic blots were stainedwith Ponceau red. They were first soaked in 3% bovine serum albuminin 25 mM sodium phosphate containing 0.9% NaCl at pH 7.5 (NaCl/Pi).Nitrocellulose sheets were quickly washed with NaCl/Pi and incubated2 h at room temperature with anti-PEAP224–237 antiserum diluted in

The Antibacterial Activity of COOH-terminal PEAP29848

by guest on March 31, 2016

http://ww

w.jbc.org/

Dow

nloaded from

NaCl/Pi (1/1000). The second antibody was an anti-rabbit IgG conju-gated with alkaline phosphatase (Bio-Rad). The nitrocellulose sheetswere stained for enzyme activity in 100 mM NaCl, 50 mM MgCl2, 100 mM

Tris/HCl, pH 8.5, containing 0.4 mM nitro blue tetrazolium (Boehringer)and 0.38 mM 5-bromo-4-chloro-3-indolyl phosphate (BoehringerMannheim).

Pyroglutamate Aminopeptidase Digestion—Peptidic material was di-gested for 2 h at 37 °C with pyroglutamate aminopeptidase, at anenzyme/protein weight ratio of 1/50, in 1 mM EDTA, 0.5 mM dithiothre-itol, 100 mM sodium phosphate buffer, pH 8.

Sequence Analysis of PEA-derived Peptides (PEAP)—The sequence ofpurified peptides was determined in our laboratory, by automatic Ed-man degradation on an Applied Biosystems 473A microsequencer. Sam-ples purified by HPLC were loaded on Polybrene-treated and precycledglass-fiber filters (1). Phenylthiohydantoin-derivatives were identifiedby chromatography on a PTH C18 column (2.1 mm 3 200 mm).

Mass Spectra Analysis—Determination of mass was carried out on aBrucker BIFLEXTM matrix-assisted laser desorption ionization time offlight mass spectrometer (MALDI-TOF MS) equipped with theSCOUTTM High Resolution Optics with X-Y multisample probe, a grid-less reflector and the HIMASTM linear detector. This instrument has amaximum accelerating potential of 30 kV and may be operated either inthe linear or reflector mode. Ionization was accomplished with a337-nm beam from a nitrogen laser with a repetition rate of 3 Hz. Theoutput signal from the detector was digitized at a sampling rate of 250MHz in linear mode and 500 MHz in reflector mode using a 1 GHzdigital oscilloscope (Lecroy model). The instrument control and dataprocessing were accomplished with software supplied by Brucker usinga Sun Sparc workstation. These studies were realized using as thematrix a-cyano-4-hydroxycinnamic acid (Sigma) prepared as a satu-rated solution in acetone. Aliquots (1–2 ml) of the sample-matrix solu-tion were deposited onto probe tips and air dried. After quick spreadingand fast evaporation of the solvent, a thin layer of matrix crystals wasobtained (48, 49). A micromolar analyte solution was applied to thematrix and allowed to dry under moderate vacuum. This preparationwas washed by applying 1 ml of a 0.5% trifluoroacetic acid in watersolution and then flushed after a few seconds. This cleaning procedureoften allows an increase in sensitivity and mass accuracy by removingthe remaining alkali cations.

Antibacterial Activity—Bacteria were grown aerobically at 37 °C inyeast extract-free Luria-Bertani medium (1% bactotryptone, and 0.5%NaCl (m/v), pH 7.5). Antimicrobial activity was based on the inhibitionof growth of M. luteus (strain A270, from Institut Pasteur), B. megate-rium (strain MA from Dr. Millet-Obert), Bacillus subtilis (strainQB935, from Dr. Klier), S. aureus (from Prof. Monteil), E. coli (strainsD22 and 363 from Dr. Bocquet, D31 from Prof. Boman, and wild strainT13773 from Prof. Monteil) in Luria-Bertani seeded medium, accordingto the method previously described (50). Peptide extract aliquots (10 ml)from HPLC fractions (200 ml of each fraction, lyophilized, and redis-solved in 50 ml of water) were incubated in microtiter plates with 100 mlof a midlogarithmic phase culture of bacteria with a starting absorb-ance of 0.001 at 620 nm. Microbial growth was assessed by the increaseof A620 nm after 16 h of incubation at 37 °C. The A620 nm value of controlcultures growing in the absence of peptide was taken as 100%.

Peptide Synthesis—Bisphosphorylated-PEAP209–237 (Ser221 andSer223 are phosphorylated) or non-phosphorylated PEAP209–237,

PEAP224–237, PEAP230–237, and PEAP209–220, were synthesized in ourlaboratory on an Applied Biosystems 432A peptide synthesizer, SYN-ERGY, using the stepwise solid-phase synthetic approach (51) with9-fluorenylmethoxycarbonyl (Fmoc chemistry). Synthesis of bisphos-phorylated peptide were performed using Fmoc-Ser[PO(OBzl)OH]-OH.Peptides were further purified by reverse-phase HPLC on a preparativeMacherey-Nagel column Nucleosil RP 300–7C18 (10 mm 3 250 mm),and finally on Macherey-Nagel Nucleosil RP 100-C18 (3 3 250 mm).After lyophilization, the synthetic peptides were analyzed by sequenc-ing and MALDI-TOF MS.

Antibody Preparation—A polyclonal rabbit serum was prepared inour laboratory against a synthetic peptide corresponding to thePEAP224–237. The first intradermal injection was performed with 500 mgof peptide coupled to hemocyanin from keyhole limpets (Megathuracrenulata) and emulsified with complete Freund’s adjuvant; a similarinjection of the peptide in incomplete Freund’s adjuvant was performed3 weeks later. Serum was collected a month later and anti-PEAP224–237

serum was purified and tested by enzyme-linked immunosorbent assay.Three-dimensional Model—The three-dimensional model of

PEAP209–237 is deduced from a 1H NMR structural analysis of 2 mg ofsynthetic peptide dissolved in 25 mM sodium acetate/d5 buffer, pH 5, inthe presence of deuterated trifluoroethanol (50% v/v). The three-dimen-

sional model was obtained using standard simulated annealing proce-dure (52) implemented in the X-PLOR program (53). The set of dis-tances used as input for the structure calculation was derived from theanalysis of a NOESY spectra recorded on a Brucker DRX 600 spectrom-eter at 283 K. The distance constraints were classified into three classeson the basis of cross-peak intensity in a 500 ms NOESY spectrum.Three types of upper limits on interproton distances, 2.7, 3.7, and 5 Å,were assigned to strong, medium, and weak NOE, respectively. Back-calculated NOESY maps were used to check the consistency of theresulting three-dimensional models with the experimental spectra andresolve the initial ambiguous NOE assignments through several runs ofstructure calculations. The program Insight II (Biosym) was used tovisualize the structures.

RESULTS

To further characterize the biological role of enkelytin, weexamined whether enkelytin is co-released with cat-echolamines from stimulated chromaffin cells and whether it ispresent in biological fluids, particularly those involved in im-mune reactions. In order to analyze its biological activity, anantibacterial spectrum was realized with the natural peptide.Several natural and synthetic enkelytin-derived peptides werealso tested to determine the structural features necessary forthe antibacterial activity. Then, the activity of these peptideswas related with the a-helical structure, obtained from recent1H NMR data (89).

Characterization of Antibacterial COOH-terminal PEAP inMaterial Released from Stimulated Cultured Bovine Chromaf-fin Cells—The complex mixture of chromogranins and PEAPrecovered in the secreted material was subjected to separationby HPLC on a reverse-phase C18 column (Fig. 1A). The differ-ent peaks were directly tested for their antibacterial activityagainst M. luteus (see “Experimental Procedures”) and se-quenced. Several peaks containing antibacterial peptides wereeluted from the column and active PEAP were detected in areas1 and 2 (including fractions 2a to 2c), eluted with acetonitrile at38 and 42%, respectively. After automatic Edman degradationof these different fractions, a unique NH2-terminal sequencewas located at position 209 of PEA. This sequence (Fig. 1B)possesses three putative phosphorylation sites (Ser215, Ser221,and Ser223) (34) and two oxidable residues (Met229 and Met237).The peptidic material present in these fractions completelyinhibited M. luteus (strain A270) growth at a concentration of0.2 mM, but was inactive against E. coli (strain D22) in a similarrange concentration. To determine the molecular differencesbetween PEAP present in fractions 1, 2a, 2b, and 2c, the se-quencing analysis was completed by a detailed study usingMALDI-TOF MS. The mass spectra analysis of the peptidespresent in peak 1 (Fig. 1C) indicated, by comparison with thecalculated molecular mass of peptide B (3658 Da), the presenceof a major fragment with a molecular mass of 3836 Da corre-sponding to the monooxidized bisphosphorylated form ofPEAP209–239 (peptide B). Three other peptides were also iden-tified as different forms of PEAP209–237/239. Thus, the molecu-lar masses of 3438 and 3516 Da are attributed to the mono- andbis-phosphorylated forms of PEAP209–237 (calculated molecularmass of 3355 Da), while the higher masses 3754 and 3931 Dacorrespond to the monooxidized monophosphorylated and tothe dioxidized triphosphorylated form of PEAP209–239 (calcu-lated molecular mass of 3658 Da). The occurrence of oxidationstates was explained by the presence of two methionine resi-dues (Met229 and Met237) in the peptide B sequence (Fig. 1B).The experimental mass values obtained for fractions 2a to 2cindicated the exclusive presence of non-oxidized mono- andbisphosphorylated forms of PEAP209–239 (data not shown).

From these studies, we can conclude that natural bisphos-phorylated forms of PEAP209–239 and PEAP209–237, named pep-tide B and enkelytin, respectively, are co-released with cat-echolamines and other neuropeptides following nicotine

The Antibacterial Activity of COOH-terminal PEAP 29849

by guest on March 31, 2016

http://ww

w.jbc.org/

Dow

nloaded from

stimulation of cultured chromaffin cells. These two peptidespossess a potent antibacterial activity against M. luteusgrowth. To further characterize the biological function of theseantibacterial PEAP, several biological fluids from injured ani-mals with infection and polymorphonuclear neutrophil secre-tions were examined.

Isolation and Characterization of Antibacterial PEAP fromInfectious Fluids—Periarthritis abscess fluid was collectedfrom cow knee, extracted by 1 M acetic acid as reported under“Experimental Procedures” and submitted to a Western blotanalysis against anti-PEA224–237 (Fig. 2D, lane 3). Two bandswere immunodetected with molecular mass of 20 and 4 kDa,respectively. The broad strongly immunoreactive band (20kDa) indicated the presence of several molecular species ofPEAP. Sequencing analysis of this material confirmed thatseveral forms of PEAP72–237/239 and PEAP80–237/239 were pres-ent within this infectious fluid. Antibacterial assays indicatedthat these 20-kDa PEAP possessed activity against M. luteus,but they were less active than enkelytin (5 versus 0.2 mM). Inconclusion, all these PEAP constitute a pool of precursorswhich have to be processed, during infection, to provide activeenkelytin. The lower 4-kDa immunodetected band is likely tobe PEAP209–237.

In order to isolate enkelytin, the acid extract was subjectedto a first HPLC on reverse-phase Macherey-Nagel Nucleosil100–5C18-HD column (4 3 250 mm) (Fig. 2A). Different frac-tions were collected, tested in antibacterial assays against M.luteus, and immunoreactivity with anti-PEAP224–237 anti-serum was screened by Western blot analysis. The immunore-active fraction (Fig. 2A, a) displayed a potent antibacterialactivity against M. luteus and sequencing indicated the pres-ence of a complex mixture of several peptides. In order toisolate the shortest antibacterial COOH-terminal PEAP, twoadditional HPLC were performed. Peptidic material containedin this fraction was first separated (Fig. 2B) on a reverse-phaseMacherey-Nagel Nucleosil 300–5C18 column (2 3 125 mm)and for complete purification of the antibacterial immunoreac-tive fraction b, a third chromatography was performed on aMacherey-Nagel Nucleosil 300–5C18 column (3 3 250 mm)(Fig. 2C). Fractions c2 and c3 were immunoreactive with anti-PEAP224–237 antiserum and after sequencing fraction c1, wedetected the NH2-terminal sequence of defensin BDO1(DFASXHTNNI; P46159) (54) and the dodecapeptide(RLXRIVVIRVXR; P2226) (55). MALDI-TOF analysis haveconfirmed the presence of these two antibacterial peptides(4273 and 1485 Da, respectively).

FIG. 1. Purification of PEAP209–237/239secreted from nicotine-stimulatedchromaffin cells. A, HPLC elution profileof soluble secreted peptides on a Macherey-Nagel reverse-phase Nucleosil 300–5C18column (4 3 250 mm). Absorbance was mon-itored at 214 nm and elution was performedat a flow rate of 700 ml/min, with a lineargradient as indicated in the right-handscale. Fractions numbered 1 and 2 (a-c) con-tained different forms of PEAP209–237/239. B,sequence of PEAP209–239. C, analysis byMALDI-TOF MS of the secreted pep-tides present in fraction 1. By comparisonwith the calculated molecular mass ofPEAP209–237 (3355 Da), the experimentalmasses (3438 Da and 3516 Da) correspond tothe mono- and bisphosphorylated forms ofPEAP209–237. The two other detected masses(3754 and 3931 Da) were characterized bycomparing with the calculated molecularmass of PEAP209–239 (3658 Da) and corre-spond, respectively, to the monooxidizedmonophosphorylated and dioxidized triphos-phorylated forms of PEAP209–239.

The Antibacterial Activity of COOH-terminal PEAP29850

by guest on March 31, 2016

http://ww

w.jbc.org/

Dow

nloaded from

Sequencing of immunoreactive fraction c2 indicated the NH2-terminal sequences of the defensin BDO2 (VRNHVTXRINRG-FXVPIR; P46146) (54), bactenecin-5 (RFRPPIRRPPIR;P19660) (56), bactenecin-7 (RRIRPRPPRLPR; P19661) (57),and histone H2B2 (PEPAKSAPAP; homologous to H2B2 his-tones of different species) (58–60). In addition, MALDI-TOFanalysis provided an experimental molecular mass of 4811 Da,which corresponds to the triphosphorylated form of PEAP199–237

(or the monophosphorylated form of PEAP199–238), with a pyroglu-tamic acid as the NH2-terminal end. In some experiments,mass spectra analysis have provided an experimental molecularmass of 4964 Da, corresponding to the triphosphorylated form ofPEAP199–238. As the presence of pyroglutamic acid at the NH2-terminal end (Gln199) prevents Edman degradation, the peptidicmaterial contained in fraction c2 (Fig. 2C) was treated with pyro-glutamate aminopeptidase. The resulting digest was separated ona reverse-phase Macherey-Nagel Nucleosil 300–5C18 column (3 3250 mm) and the major fraction was characterized by sequencingand MALDI-TOF analysis (data not shown). In this manner, weidentified the sequence (KRYGGFLKRFAEPLP) corresponding tothe NH2-terminal end of PEAP200–237/238, the expected fragmentgenerated after digestion with pyroglutamate aminopeptidase. Theexperimental mass of 4723 Da (with the addition of a sodium ion by

comparison with the theoretical mass of 4700 Da) confirmed thepresence of triphosphorylated PEAP200–237.

Finally, automated Edman degradation of the immunoreac-tive fraction c3 (Fig. 2D, lane 4) indicated the presence of aNH2-terminal sequence beginning at residue 209. MALDI-TOFanalysis (Fig. 3) confirmed the presence of COOH-terminalPEAP with experimental masses of 3516 Da (the bisphospho-rylated form of PEAP209–237), 3508 Da (PEAP209–238), 3523 Da(the monooxidized form of PEAP209–238), and 3805 Da (thedioxidized triphosphorylated form of PEAP209–238 with addi-tion of a sodium ion). We also detected a molecular mass of7027 Da corresponding to a dimeric form of enkelytin (3516Da). In some experiments, a narrow-bore HPLC was performedon a Brownlee C18 column. Elution was performed at a flowrate of 5 ml/min using successively 15% B over 15 min and agradient of 5% B to 80% B over 105 min. This additionalchromatography confirms the previous HPLC profile and cor-roborates the presence of PEAP199/209–237/238 (data not shown).

In order to confirm the presence of antibacterial peptidesderived from the COOH-terminal end of PEA within wounds,we examined two other infectious fluids. The first liquid wasdrained from a post-operative (post-caesarean) abscess in thesubcutaneous lining of a cow. Western blots analysis with

FIG. 2. Characterization of PEAPfrom cow knee periarthritis abscessfluid. A, HPLC elution profile on a Mach-erey-Nagel reverse-phase Nucleosil 300–5C18-HD column (4 3 250 mm) of pep-tidic material included in an acid extractof cow periarthritis abscess fluid. Absorb-ance was monitored at 214 nm and elu-tion was performed at a flow rate of 700ml/min, with a linear gradient as indi-cated in the right-hand scale. Antibacte-rial activity and immunoreactivity withanti-PEAP224–237 were detected in frac-tion a. B, HPLC elution profile on a Ma-cherey-Nagel reverse-phase Nucleosil300–5C18 column (2 3 125 mm) of pep-tidic material included in fraction a. Ab-sorbance was monitored at 214 nm andelution was performed at a flow rate of400 ml/min with a linear gradient as indi-cated in the right-hand scale. Antibacte-rial activity and immunoreactivity weredetected in fraction b. C, HPLC elutionprofile on a Macherey-Nagel reverse-phase Nucleosil 300–5C18 column (3 3250 mm) of peptidic material included infraction b. Absorbance was monitored at214 nm and elution was performed at aflow rate of 400 ml/min with a linear gra-dient as indicated in the right-hand scale.Antibacterial activity was detected infraction c1, c2, and c3 and immunoreactiv-ity in fractions c2 and c3. D, Western blotanalysis (17%, SDS-PAGE) with anti-PEAP224–237: lane 1, molecular massstandard; lane 2, intragranular chromaf-fin soluble material; lane 3, peptidic ma-terial included in acid extract of cow per-iarthritis abscess fluid; lane 4, HPLCfraction c3; lane 5, peptidic material in-cluded in acid extract of cow post-caesar-ean abcess fluid; lane 6, peptidic materialfrom induced rabbit abcesses (see “Exper-imental Procedures”); lane 7, secretionsreleased from human PMNs.

The Antibacterial Activity of COOH-terminal PEAP 29851

by guest on March 31, 2016

http://ww

w.jbc.org/

Dow

nloaded from

anti-PEAP224–237 antiserum (Fig. 2D, lane 5) indicated similarimmunoreactivity to that obtained with the periarthritis abcess(Fig. 2D, lane 3). In a second experiment, a rabbit abscessinduced by subcutaneous injection of complete Freund’s adju-vant was drained 10 days later. The material collected wastreated as for bovine knee periarthritis abscess fluid andloaded on a HPLC reverse-phase C18 column. The differentfractions were tested for antibacterial assays against M. luteusand submitted to Western blot immunodetection with anti-PEAP224–237 antiserum, sequencing, and MALDI-TOF MS. Inthe immunodetected fractions (Fig. 2D, lane 6), we identifiedthe NH2-terminal sequence of two rabbit defensins, NP1(P01376) and NP2 (P01377) (61). The experimental mass val-ues of 3892 and 3849 Da obtained for these peptides correspondto the theoretical molecular masses of defensins NP1 (3893 Da)and NP2 (3850 Da). In addition, since PEA sequences in severalspecies are highly conserved (38), the rabbit PEAP sequencesand experimental molecular masses were compared with ratPEAP (62, 63). The most likely candidates for these fragmentsare the bisphosphorylated form of PEAP202–238 and the mono-phosphorylated form of PEAP206–237 with experimental molec-ular masses of 4453 and 3851 Da, respectively, instead of 4453and 3853 Da for rat PEAP.

In conclusion, the experiments described here reveal thepresence of several peptides with antibacterial activity influids from infected wounds: defensins, bactenecins, dode-capeptide as expected, and natural PEAP, such as severalforms of PEAP72/80–237/239, PEAP199–237, PEAP209–238, andthe bisphosphorylated form of PEAP209–237 (enkelytin).Quantification of isolated enkelytin present at the inflamma-tory area could be obtained from sequencing and its concen-tration in (bovine periarthritis abscess fluid) was estimatedto be from 0.5 to 1 mM. In this concentration range, thepeptide is fully potent, indicating that enkelytin locally ex-erts genuine antibacterial activity in specific fluids. In con-trast, circulating enkelytin concentration is much less as itwas hardly detectable in plasma (data not shown). As acontinuation of this study, we have examined the presence ofantibacterial PEAP in secretions from human PMNs. Afterreverse phase HPLC on a Macherey-Nagel Nucleosil 100–

5C18-HD column (4 3 250 mm), immunoreactivity was de-tected with anti-PEAP224–237 antiserum (Fig. 2D, lane 7),indicating that PEAP are secreted for PMNs with a patternsimilar to those described for bovine periarthritis (Fig. 2D,lane 3) and rabbit abscesses (Fig. 2D, lane 5).

Comparison of the Antibacterial Activities of Natural andSynthetic Enkelytin- and Peptide B-derived Fragments—In or-der to further extend the bacterial spectrum initially reportedfor enkelytin (38), we decided to test the antibacterial activityof this natural peptide against several Gram-positive and neg-ative bacteria. The data reported on Table I show that enkely-tin entirely inhibits M. luteus and B. megaterium growth at 0.2mM; it also inhibits the growth of S. aureus, being fully active ata concentration of 4.5 mM. Enkelytin was inactive toward B.subtilis under similar experimental conditions. Four differentstrains of E. coli (D22, D31, 663, and a wild strain, T13773)were tested with a peptide concentration of 3 mM but no anti-bacterial activity was detectable. These tested concentrationswere in accordance with the amount of enkelytin found withinphysiological fluids. To conclude, this analysis spectrum indi-cates that the antibacterial activity of natural enkelytin isselective for several Gram-positive bacteria strains. In addi-tion, it is important to point out that this new antibacterialpeptide is able to inhibit the growth of S. aureus.

In order to characterize the structural features necessary forthe antibacterial activity of enkelytin, we have tested severalnatural and synthetic PEAP against the growth of Gram-pos-itive (M. luteus, strain A270) and Gram-negative (E. coli strainD22) bacteria. Natural enkelytin, PEAP209–237 (peptide 1, Fig.4A) and natural bisphosphorylated PEAP209–239, known aspeptide B (peptide 2, Fig. 4A) completely inhibit the growth ofM. luteus at a concentration of 0.2 mM (Fig. 4B), but wereunable to inhibit that of E. coli in the concentration range from0.2 to 3 mM.

After preparation of synthetic enkelytin (bisphosphorylatedPEAP209–237), the peptide was loaded on a reverse phase chro-matography. The HPLC profile and the MALDI-TOF MS indi-cated the presence of different molecular forms. Therefore,synthetic active enkelytin was further purified on a Macherey-Nagel Nucleosil 300–5C18 column (125 3 3 mm) and analyzedby sequencing and MALDI-TOF. After purification and se-quencing of the active synthetic form, we evaluated that only10% of the synthetic peptide adopts a conformation with theeffective antibacterial activity (peptide 3, Fig. 4A). At thisstage, its activity was closer to that of the natural peptide(100% of bacteria growth inhibition at 3 mM), in contrast withour previous work where we did not consider that only a lowpercentage of synthetic peptide adopts the active conformation

FIG. 3. MALDI-TOF MS of the peptidic material included infraction c3. The four experimental molecular mass values 3508, 3516,3523, and 3805 Da correspond to PEAP209–238, the bisphosphorylatedform of PEAP209–237, the monooxidized form of PEAP209–238 and thedioxidized triphosphorylated form of PEAP209–238 with addition of asodium ion, respectively.

TABLE IActivity spectrum of natural enkelytin

(bisphosphorylated PEAP209–237)MIC, the minimal inhibitory concentration is expressed in micromo-

lar; .3, means that no antibacterial activity was found with peptideconcentration lower than 3 mM. For S. aureus, 100% inhibition is ob-tained at 4.5 mM but may be reached with a lower value.

Bacteria MIC

mM

Gram-positive bacteriaM. luteus 0.2B. megaterium 0.2B. subtilis .3S. aureus 4.5

Gram-negative bacteriaE. coli D22 .3E. coli D31 .3E. coli 663 .3E. coli T13773 .3

The Antibacterial Activity of COOH-terminal PEAP29852

by guest on March 31, 2016

http://ww

w.jbc.org/

Dow

nloaded from

(38). These results suggest important conformational differ-ences between the different synthetic isoforms. In parallel ex-periments, we were able to show from the three-dimensional1H NMR analysis of PEAP209–237 (89) that proline residues areresponsible for conformational changes (cis-trans isomeriza-tion). In contrast with the natural and synthetic bisphospho-rylated peptide, the low antibacterial activity of the non-mod-ified synthetic peptide (peptide 4, Fig. 4A) suggests animportant role of the two phosphorylated serine residues inactive structure. Thus, at 3 mM the synthetic non-phosphory-lated PEAP209–237 (peptide 4, Fig. 4A) was inactive against M.luteus; the concentration has to be raised to 100 mM to induce a20% inhibition of bacterial growth.

Finally, in order to correlate the antibacterial activity withthe NH2 and COOH domain and the length of the peptidicchain of enkelytin, we have tested four shorter peptides (Fig.4A): PEAP209–220 (peptide 5, Fig. 4A), PEAP224–237 (peptide 6),PEAP230–237 (peptide 7), and PEAP233–237; this later fragmentcorresponds to Met-enkephalin (peptide 8). As shown in Fig.4B, the antibacterial assay of the NH2- and COOH-terminaldomains (peptides 5 and 6) at a concentration of 500 mM indi-cates a 25 and 20% inhibition of growth, respectively, whereasshort COOH-terminal peptides 7 and 8 were inactive at theconcentration range of 500 mM.

These studies were completed with antibacterial assaysagainst E. coli (strain D22) growth and erythrocyte lysis. In theconcentration range from 0.2 to 500 mM, none of the peptides

listed in Fig. 4A showed neither any detectable antibacterialactivity against this Gram-negative bacterium nor any hemo-lytic activity. In conclusion, the antibacterial activity of enk-elytin toward M. luteus is directly related to three structuralparameters: (i) the length of the peptidic chain, (ii) the naturalconformational constraints induced by the three proline resi-dues Pro212, Pro214, Pro227, and (iii) the phosphorylation ofSer221 and Ser223.

Computer Model of PEAP209–237—An extensive study usingbiophysical techniques has been carried out on PEAP209–237 inour laboratory (89). We refer to some of these data to draw upthe computer model of PEAP209–237 fitting with the biologicalactivity. Circular dichroism (CD) spectra recorded with in-creasing percentage of trifluoroethanol showed that syntheticPEAP209–237 folds progressively into an helical structure, asthe percentage of rifluoroethanol is increased up to 50%, asshown by the appearance of a negative band at 220 nm. The CDspectra displayed no change with trifluoroethanol concentra-tion above 50%.

The presence of helical structure was confirmed in the 1HNMR spectra of synthetic PEAP209–237 by the presence of reg-ular Ha(i), HN(i13) NOE for residues from Ser215 to Gly219 andfrom Glu228 to Phe236. PEAP209–237 sequence contains threeproline residues which are able to adopt either the cis or transconformation of the peptide bond. The two isomers are charac-terized by distinctive NOE patterns between the protons of theproline and those of the preceeding residue. Thus, each threeproline residues showed different behavior: (i) Pro212 has atrans conformation, (ii) cis and trans NOE patterns wereclearly identified from Pro214 and Leu213, as indicated by tworesonance frequency values, and (iii) both cis and trans NOEpatterns were found for Pro227, but no different chemical shiftswere observed for the two isomers. Therefore, two models werecalculated with Pro227 either in the cis or trans conformation.

In both models (Pro227 cis and Pro227 trans) presented as aribbon diagram (Fig. 5, A and B, respectively), the conforma-tions of Pro212 and Pro214 were set to be trans. The Pro227

FIG. 4. Antibacterial activity of natural and synthetic enkely-tin-derived peptides. A, identification of the 8 different peptidestested. B, peptides at different concentrations were incubated 16 h at37 °C with M. luteus (strain A 270) in yeast extract-free Luria-Bertanimedium as described under “Experimental Procedures.” Microbialgrowth was assessed by measuring the increase at A620 nm. Values foundwith control cultures grown in the absence of peptide were taken as 0%.Numbers in each column indicate the peptide concentration inhibitingbacterial growth. Experimental values are given 6 5%.

FIG. 5. Three-dimensional structure of enkelytin correspond-ing to PEAP209–237. Ribbon representation of the three-dimensionalstructure of PEAP209–237 in a 50% trifluoroethanol/water solution ac-cording to X-PLOR program (53). Both cis (A) and trans (B) conforma-tions of Pro227 are deduced from the 1H NMR data (89). The NH2-terminal parts of the two models (Phe209 to Pro227) have the sameorientation. E, glutamic acid residue; P, proline residue; S, serineresidue.

The Antibacterial Activity of COOH-terminal PEAP 29853

by guest on March 31, 2016

http://ww

w.jbc.org/

Dow

nloaded from

residue induces a bend in the three-dimensional structure,which adopts a L shape and breaks the helical structure ob-served on either side of Pro227. It is striking that both isomersof Pro227 lead to the same kind of spatial proximity between aglutamic acid and a serine side chain (Ser223/Glu230, in the cisconformation and Ser221/Glu228 in the trans one). In enkelytin,when the two serine residues (Ser221 and Ser223) are phospho-rylated, the negatively charged phosphate groups probably in-duce conformational change by electrostatic interactions (64).In contrast to the COOH-terminal fragment 227–237 whichadopts a helical conformation, the structure of the NH2-termi-nal end (fragment 209–214) is poorly defined, due to the lack ofmedium range NOE, partially explained by an averaging overa broad range of conformations resulting in the cis-trans isom-erism of Pro214.

DISCUSSION

Despite intensive research to counter the development ofnew bacterial resistance, no novel classes of antibacterialagents have been discovered in the past 30 years. Currently,there is a great interest in antibacterial peptides as an attemptto resolve this challenge. Thousands of such molecules havebeen synthesized, but just a few, such as magainins, are cur-rently being tested in clinical trials. Thus, the structural andbiological characterization of new natural antibacterial pep-tides, derived from naturally processed precursors is a topic ofgrowing interest in relation to their therapeutic use. The in-tracellular proteolytic processing of protein precursors occursin storage compartments in nervous and endocrine systems. Ithas been established that the processing takes place at dibasicsites (65), at single basic residues, at peptide bonds involvinghydrophobic amino acid (1), and at sites marked by the consen-sus RX(K/R)R sequence (66). The tertiary structure, in part dueto post-translational modifications (phosphorylation, glycosyla-tion . . . ) must play an important role in cleavage site accessi-bility. As large amounts of enkephalins and PEAP are presentin adrenal medullary chromaffin granules, these vesicles havebeen extensively used as a source for studying the naturalprocessing of PEA (28). This protein, which is widely distrib-uted in many cell types, shows cell-specific processing patterns.

Recently, we have characterized enkelytin, a new antibacte-rial peptide which corresponds to the bisphosphorylatedPEAP209–237 (38), derived from peptide B (PEAP209–239) (32).As shown here, this natural peptide displays a potent antibac-terial activity against Gram-positive bacteria M. luteus, B.megaterium, and S. aureus, but was unable to inhibit thegrowth of Gram-negative bacteria such as the tested E. colistrains. This COOH-terminal domain of PEA has been wellconserved during evolution, and proteolytic processing of PEAin the adrenal medulla starts at this COOH-terminal region(31). Recently, it has been demonstrated that AtT-20 cellstransfected with rat recombinant PEA gene released peptide B

20 min after PEA synthesis (30), indicating that this peptide israpidly generated.

In the present study, two antibacterial PEAP, the bisphos-phorylated peptide B (PEAP209–239) and enkelytin (PEAP209–

237) are shown to be secreted from cultured chromaffin cellsfollowing stimulation. This result suggests that these two pep-tides that are co-released with catecholamines in stress situa-tions may play an important role in defense mechanisms. Fur-thermore, we have established the presence in bovineinfectious fluids of several antibacterial fragments includingPEAP209–237, PEAP199–237/238, and major larger precursor frag-ments, PEAP72/80–237/239. After extensive extracellular process-ing, these 20-kDa fragments generate enkelytin and its derivedpeptides. Interestingly, in these fluids the concentration ofenkelytin (0.5–1 mM) is in accordance with the antibacterialactivity (Table I).

In a previous paper (38), according to the Homolog methodprovided in Pro-Explore, we reported comparative structuralpredictions of enkelytin and the homologous antibacterial di-azepam-binding inhibitor-derived peptide, showing an am-phipathic helical structure for PEAP224–237. In the two modelsA and B reported here (Fig. 5), it is important to note thatPro227, which is highly conserved in PEA sequence from severalspecies (Fig. 6), is breaking a regular helical conformation witha bend formation. This bending brings the glutamic acid resi-dues (Glu228 and Glu230) close to the phosphorylated serineresidues (Ser221 and Ser223). The repulsive electrostatic inter-actions resulting from the phosphorylation of Ser221/223 may actas molecular switch for the antibacterial activity. Thus, thephosphorylation of Ser221 and Ser223 by addition of negativecharges could open the “boomerang angle” (38) or increase theability of this peptide to bind divalent ions and thus induce theantimicrobial activity of enkelytin as described previously for apoly(Asp) antibacterial peptide (67, 68). The confirmation ofthis model will be provided by 1H NMR studies of the bisphos-phorylated synthetic PEAP209–237 and the bisubstituted glu-tamic at sites of phosphorylated serine residues, in aqueoussolution including divalent ions and in membrane environ-ment. Moreover, it is interesting to point out that the helicalstructure for the COOH-terminal Met-enkephalin as we reporthere differs significantly from 1H NMR structures previouslydescribed for Met- or Leu-enkephalin (69, 70). This is probablydue to the extension of the NH2-terminal region.

Antibacterial peptides have to be positively charged in orderto bind to bacterial surfaces, which are usually negativelycharged. Curiously, the net charges of the most active peptidesnumbered 1 to 3 (Fig. 4), were calculated to be 27, 26, and 27,respectively. However, enkelytin and peptide B, although neg-atively charged, may act by a pore-forming or carpet-like mech-anism, as recently described (71). However, other mechanismscan also be considered such as peptide membrane receptors on

FIG. 6. Sequence comparison of bo-vine PEA198–239 with correspondingfragments from several species. PEAsequences were retrieved from the Swiss-Prot or GenBank data base: bovine(P01211) (7), human (P012100) (3, 81), pig(JL0067) (82), rat (P04094) (62, 63),mouse (P22005) (83), Mesocricetus aura-tus (Syrian golden hamster) (MAU09941)(84), guinea pig (P47969) (85), Xenopuslaevis (P01012) (86), Mytilus edulis andTheromyzon tessulatum (leech) (87). Leu-and Met-enkephalins were underlined.S*, phosphorylated serine residues in bo-vine sequence. 2, deletion. Arrows indi-cate natural proteolytic cleavage sites.

The Antibacterial Activity of COOH-terminal PEAP29854

by guest on March 31, 2016

http://ww

w.jbc.org/

Dow

nloaded from

bacterial membranes, the possibility for the peptides to act as“oblique-oriental” peptides (72) or the ability for these anionicpeptides to bind divalent ions (67, 68). At this stage, however,the mechanism by which enkelytin and peptide B inhibitsbacteria growth remains to be determined. The presence ininfectious fluids of antibacterial COOH-terminal PEAP to-gether with other antibacterial peptides supports their poten-tial role in host defense. Defensins and bactenecins are thoughtto be released at infection and inflammation sites. In the pres-ent study, several purification steps were necessary (3 succes-sive HPLC) to isolate the different forms of active PEAP fromperiarthritis abscess fluid, suggesting that interactions occurbetween these acidic fragments and the cationic antibacterialpeptides, such as defensins or bactenecins. The formation ofmolecular complexes including several peptides may be impor-tant to obtain a synergistic antibacterial efficiency. The com-puter model obtained for the synthetic PEAP209–237 (Fig. 5)indicated a long amphipathic a-helical structure. This struc-ture completes our previous predicted model concerning thea-helical structure of PEAP224–237 (38).

PEAP209–239 is the most highly conserved domain of theprotein precursor with a yield of homology around 90% (Fig. 6).Proline residue located in position 212 in bovine sequence ischanged to Ala, Ser, Glu, or Phe residues in other species.Because of the high conservation of the COOH-terminal do-main of PEA, the antibacterial activity appears to have oc-curred early in evolution.

The antibacterial COOH-terminal PEAP may originate fromchromaffin cells, since these cells contain high levels of PEAP,or from immune cells (e.g. PMNs). PEA has been reported to besignificantly expressed in the immune system and may providea basis for neuroimmune interactions (8–11). The local inflam-matory response initiates the synthesis and the secretion ofopioid peptides by immune cells. When Freund’s adjuvant isused to induce unilateral hindpaw inflammation in rats, PEAmRNA are abundant in cells of the inflammed tissue, butabsent in non-inflammed tissue. Numerous cells infiltratingthe inflammed subcutaneous tissue are stained intensivelywith Met-enkephalin, suggesting that PEAP are synthesizedand processed within various types of immune cells at the siteof inflammation (73). Moreover, exposure of rats to lipopolysac-charide endotoxin leads to PEA mRNA and protein expressionin macrophages within lymph nodes and in chromaffin cellswithin adrenal glands (74). One physiological effect of PEAP isto up-regulate or enhance the immune response at low concen-trations, but this effect is abolished at high concentrations.Other studies performed in invertebrates suggest a potentialdual role of PEA in defensive processes (75, 76). Thus, enkely-tin degradation at the infection site by two endopeptidases,neuropeptide-degrading endopeptidase and angiotensin-con-verting enzyme present in granulocytes, generate Met-en-kephalin and its derived peptides (76). Met-enkephalin en-hances the immune reaction in patients with cancer or AIDS(77). With regard to this immune modulating property, Met-enkephalin has been proposed to be classified as a cytokine(78). Moreover, this pentapeptide can bind opioid receptorspresent in peripheral inflammed tissues to mediate an analge-sic effect (79). The involvement of opioids in neuroimmunoregu-latory events appears to have a long evolutionary history. Al-though the relationship between the immune and nervoussystems was discovered in vertebrates, it also exists in inver-tebrates (80) and the co-release of enkelytin and Met-enkepha-lin represents an unified neuroimmune protective response tostress situations that may be accompanied with infectious dis-eases. Taken together, these two peptides would provide ahighly beneficial survival strategy at the very beginning of a

proinflammatory process.Our studies provide new data concerning the biological char-

acterization of the COOH-terminal antibacterial PEAP namedenkelytin, first isolated from chromaffin granules and nowrecovered as secretory products from stimulated chromaffincells and in wound fluids. In view of the widespread distribu-tion of PEA, these peptides may also be present and secretedfrom other endocrine, neuroendocrine, and immune cells. Dueto their nonspecific activity on membranes, the antibacterialpeptides possess cytotoxic activities and may not only play arole in antimicrobial defense, but also in inflammatory pro-cesses. Since antibacterial PEAP are released with cat-echolamines and chromogranins, the latter being precursors toother peptides with antibacterial activities (87), they may playa role in stress situations and act as one immediate protectivebarrier against infection. The identification of different classesof antibacterial peptides in a diverse range of organisms, in-cluding prokaryotes, insects, frogs, and mammals, suggeststhat they play a potentially important role in host defenseagainst microbial infections.

Acknowledgments—We thank Dr. P. Haas, J. Knobloch (CNRS, UPS840), Dr. D. Colin (Laboratoire de toxicologie bacterienne, Faculte deMedecine Strasbourg) for help in collecting biological fluids; Dr. P.Bulet, M. Schneider (CNRS UPR 9022) and Dr. B. Jaulhac (Laboratoirede toxicologie bacterienne, Faculte de Medecine Strasbourg) for thegenerous gift of bacteria. We are indebted to Drs. O. Sorokine and J. M.Strub (CNRS, URA 31, Strasbourg France) for mass spectrometry anal-ysis of different peptides and G. Nullans (INSERM U 338) for synthesisof peptides. We are grateful to Dr. N. Grant (INSERM U 338) forimproving the English version of the manuscript. Finally, we expressour sincere gratitude to the two anonymous reviewers for their sugges-tions on the first version of this manuscript that helped us to furthercharacterize enkelytin.

REFERENCES

1. Metz-Boutigue, M. H., Garcia-Sablone, P., Hogue-Angeletti, R. & Aunis, D.(1993) Eur. J. Biochem. 145, 659–676

2. Strub, J. M., Garcia-Sablone, P., Lonning, K., Taupenot, L., Hubert, P., VanDorsselaer, A., Aunis, D. & Metz-Boutigue, M. H. (1995) Eur. J. Biochem.229, 356–368

3. Comb, M., Seeburg, P. H., Adelman, J., Eiden, L. & Herbert, E. (1982) Nature295, 663–666

4. Martens, G. J. M. & Herbert, E. (1984) Nature 310, 251–2545. Uhl, G. R., Navia, B. & Douglas, J. (1988) J. Neurosci. 8, 4755–47646. Pittius, C. W., Kley, N., Loeffler, J. P. & Hollt, V. (1985) EMBO J. 4, 1257–12607. Noda, M., Furutani, Y., Takahashi, H., Toyosato, M., Hirose, T., Inayama, S.,

Nakanishi, S. & Numa, S. (1982) Nature 295, 202–2068. Zurawski, P., Benedik, G. M., Kamb, B. J., Abrams, J. S., Zurawski, S. M. &

Lee, F. D. (1986) Science 232, 772–7759. Linner, K. M., Beyer, H. S. & Sharp, B. M. (1991) Endocrinology 128, 717–724

10. Rosen, H., Behar, O., Abramsky, O. & Ovadia, H. (1989) J. Immunol. 143,3703–3707

11. Martin, J., Prystowsky, M. B. & Hogue-Angeletti, R. (1987) J. Neurosci. Res.18, 82–87

12. Jankovic, B. D. & Maric, D. (1987) Ann. N. Y. Acad. Sci. 496, 115–12513. Faith, R. E., Murgo, A. J., Clinkscales, C. W. & Plotnikoff, N. P. (1987) Ann.

N. Y. Acad. Sci. 496, 137–14514. Foris, G., Medgyesi, G. A. & Hauck, M. (1986) Mol. Cell. Biochem. 69, 127–13715. Foris, G., Medgyesi, G. A., Nagy, J. T. & Varga, Z. (1987) Ann. N. Y. Acad. Sci.

496, 151–15716. Maric, D. & Jankovic, B. D. (1987) Ann. N. Y. Acad. Sci. 496, 126–13617. Plotnikoff, N. P. & Miller, G. C. (1983) Int. J. Immunopharmacol. 5, 437–44118. Das, K. P., Hong, J. S. & Sanders, V. M. (1997) J. Neuroimmunol. 73, 37–4619. Oleson, D. R. & Johnson, D. R. (1988) Brain Behav. Immun. 1, 171–18620. Kilpatrick, D. L., Borland, K. & Jin, D. F. (1987) Proc. Natl. Acad. Sci. U. S. A.

84, 5695–569921. Muffly, K. E., Jin, D. F., Okulicz, W. C. & Kilpatrick, D. L. (1988) Mol.

Endocrinol. 2, 979–98522. Howells, R. D., Kilpatrick, D. L., Bailey, L. C., Noe, M. & Udenfriend, S. (1986)

Proc. Natl. Acad. Sci. U. S. A. 83, 1960–196323. Springhorn, J. P. & Claycom, B. W. C. (1989) Biochem. J. 258, 73–7824. Keshet, E., Polakiewicz, R. D., Itin, A., Ornoy, A. & Rosen, H. (1989) EMBO J.

8, 2917–292325. Kew, D. & Kilpatrick, D. L. (1990) Mol. Endocrinol. 4, 337–34026. Linner, K. M., Quist, H. E. & Sharp, B. M. (1995) J. Immunol. 154, 5049–506027. Kojima, K., Kilpatrick, D. L., Stern, A. S., Jones, B. N. & Udenfriend, S. (1982)

Arch. Biochem. Biophys. 215, 638–64328. Dillen, L., Miserez, B., Claeys, M., Aunis, D. & De Potter, W. (1993)

Neurochem. Int. 22, 315–32229. Rostovtsev, A. P. & Wilson, S. P. (1994) Mol. Cell. Endocrinol. 101, 277–28530. Mathis, J. P. & Lindberg, I. (1992) Endocrinology 131, 2287–229631. Liston, D., Patey, G., Rossier, J., Verbanck, P. & Vanderhaeghen, J. J. (1984)

The Antibacterial Activity of COOH-terminal PEAP 29855

by guest on March 31, 2016

http://ww

w.jbc.org/

Dow

nloaded from

Science 225, 734–73732. Stern, A. S., Jones, B. N., Shively, J. E. & Udenfriend, S. (1981) Proc. Natl.

Acad. Sci. U. S. A. 78, 1962–196633. D’Souza, N. B. & Lindbergh, I. (1988) J. Biol. Chem. 263, 2548–255234. Watkinson, A., Young, J., Varro, A. & Dokray, G. J. (1989) J. Biol. Chem. 264,

3061–306535. Lindberg, I. & White, L. (1986) Biochem. Biophys. Res. Commun. 139,

1024–103236. Strub, J. M., Hubert, P., Nullans, G., Aunis, D. & Metz-Boutigue, M. H. (1996)

FEBS Lett. 379, 273–27837. Strub, J. M., Goumon, Y., Lugardon, K., Capon, C., Lopez, M., Moniatte, M.,

Van Dorsselaer, A., Aunis, D. & Metz-Boutigue, M. H. (1996) J. Biol. Chem.271, 28533–28540

38. Goumon, Y., Strub, J. M., Moniatte, M., Nullans, G., Poteur, L., Hubert, P.,Van Dorsselaer, A., Aunis, D. & Metz-Boutigue, M. H. (1996) Eur. J. Bio-chem. 235, 516–525

39. Agerberth, B., Gunne, H., Odeberg, J., Kogner, P., Boman, H. G. &Gudmundsson, G. H. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 195–199

40. Gallo, R. L., Ono, M., Povsic, T., Page, C., Eriksson, E., Klagsbrun, M. &Bernfield, M. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 11035–11039

41. Frohm, M., Gunne, H., Bergman, A. C., Agerberth, B., Bergman, T., Boman,A., Liden, S., Jornvall, H. & Boman, H. G. (1996) Eur. J. Biochem. 237,86–92

42. Knudsen, J., Mandrup, S., Rasmussen, J. T., Andreasen, P. H., Poulsen, F. &Kristiansen, K. (1993) Mol. Cell. Biochem. 123, 129–138

43. Finck-Barbancon, V., Duportail, G., Meunier, O. & Colin, D. A. (1993) Biochim.Biophys. Acta 1182, 275–282

44. Colin, D. A., Mazurier, I., Sire, S. & Finck-Barbancon, V. (1994) Infection andImmunity 62, 3184–3188

45. Meunier, O., Falkenrodt, A., Monteil, H. & Colin, D. A. (1995) Cytometry 21,241–247

46. Laemmli, U. K. (1970) Nature 227, 680–68547. Towbin, H., Staehelin, T. & Gordon, J. (1979) Proc. Natl. Acad. Sci. U. S. A. 76,

4350–435448. Vorm, O. & Mann, M. (1994) J. Am. Soc. Mass. Spec. 5, 955–95849. Vorm, O., Roepstorff, P. & Mann, M. (1994) Anal. Chem. 66, 3281–328750. Bulet, P., Dimarcq, J. L., Hetru, C., Lagueux, M., Charlet, M., Hegy, G., Van

Dorsselaer, A. & Hoffman, J. A. (1993) J. Biol. Chem. 268, 14893–1489751. Merrifield, R. B. (1963) J. Am. Chem. Soc. 85, 2149–215452. Nilges, M., Gronenborn, A. M., Brunger, A. T. & Clore, G. M. (1988) Protein

Eng. 2, 27–3853. Brunger, A. T. (1992) X-PLOR Manual, Version 3.0, Yale University, New

Haven, CT54. Selsted, M. E., Tang, Y. Q., Morris, W. L., McGuire, P. A., Novotny, M. J.,

Smith, W., Henschen, A. H. & Cullor, J. S. (1993) J. Biol. Chem. 268,6641–6648

55. Romeo, D., Skerlavaj, B., Bolognesi, M. & Gennaro, R. (1988) J. Biol Chem.263, 9573–9575

56. Zanetti, M., Del Sal, G., Storici, P., Schneider, C. & Romeo, D. (1993) J. Biol.Chem. 268, 522–526

57. Frank, R. W., Gennaro, R., Schneider, K., Przybylski, M. & Romeo, D. (1990)J. Biol. Chem. 265, 18871–18874

58. Liu, T. J., Liu, L. & Marzluff, W. F. (1987) Nucleic Acids Res. 15, 3023–303959. Dobner, T., Wolf, I., Mai, B. & Lipp, M. (1991) DNA Seq. 1, 409–41360. Perry, M., Thomsen, G. H., Roeder, R. G. (1985) J. Mol. Biol. 185, 479–499

61. Selsted, M. E., Brown, D. M., DeLange, R. J., Harwig, S. S. L. & Lehrer, R. I.(1985) J. Biol. Chem. 260, 4579–4584

62. Howells, R. D., Kilpatrick, D. L., Bhatt, R., Monahan, J. J., Pooman, M. &Udenfriend, S. (1984) Proc. Natl. Acad. Sci. U. S. A. 81, 7651–7655

63. Yoshikawa, K., Williams, C. & Sabol, S. L. (1984) J. Biol. Chem. 259,14301–14308

64. Terzi, E., Poteur, L. & Trifilieff, E. (1992) FEBS Lett. 309, 413–41665. Rholam, M., Nicolas, P. & Cohen, P. (1986) FEBS Lett. 207, 1–666. Nakayama, K. (1997) Biochem. J. 327, 625–63567. Brogden, K. A., Ackermann, M. & Huttner, K. M. (1997) Antimicrob. Agents

Ch. 41, 1615–161768. Brogden, K. A., De Lucca, A. J., Bland, J. & Elliott, S. (1996) Proc. Natl. Acad.

Sci. U. S. A. 93, 412–41669. D’Alagni, M., Delfini, M., Di Nola, A., Eisenberg, M., Paci, M., Roda, G. &

Veglia, G. (1996) Eur. J. Biochem. 240, 540–54970. Milon, A., Miyazawa, T. & Higashijima, T. (1990) Biochemistry 29, 65–7571. Gazit, E., Boman, A., Boman, H. G. & Shai, Y. (1995) Biochemistry 34,

11479–1148872. Brasseur, R., Pillot, T., Lins, L., Vandekerckhove, J. & Rosseneu, M. (1997)

Trends Biochem. Sci. 22, 167–17173. Przewlocki, R., Hassan, A. H., Lason, W., Epplen, C., Herz, A. & Stein, C.

(1992) Neuroscience 48, 491–50074. Behar, O., Ovadia, H., Polakiewicz, R. D. & Rosen, H. (1994) Endocrinology

134, 475–48175. Stefano, G. B. Salzet, B. & Fricchione, G. L. (1998) Immunol. Today 19,

265–26876. Salzet, M., Salzet, B., Tamieski, A., Verger-Bocquet, M., Goumon, Y., Aunis,

D., Metz-Boutigue, M. H., Cadet, M. & Stefano, G. B. (1998) J. Immunol., inpress

77. Schafer, M., Mousa, S. A., Zhang, Q., Carter, L. & Stein, C. (1996) Proc. Natl.Acad. Sci. U. S. A. 93, 6096–6100

78. Plotnikoff, N. P., Faith, R. E., Murgo, A. J., Herberman, R. B. & Good, R. A.(1997) Clin. Immunol. Immunopathol. 82, 93–101

79. Stein, C., Hassan, A. H., Lehrberger, K., Grefing, J. & Yassouridis, A. (1993)Lancet 343, 321–324

80. Stefano, G. B., Cadet, P., Dokun, A. & Scharrer, B. (1990) Brain Behav.Immun. 4, 323–329

81. Noda, N., Teranishi, Y., Takahashi, H., Toyosato, M., Notake, M., Nakanishi,S. & Numa, S. (1982) Nature 297, 431–434

82. Watkinson, A., Dockray, G. J., Young, J. & Gregory, H. (1988) J. Neurochem.51, 1252–1257

83. Kilpatrick, D. L., Zinn, S. A., Fitzgerald, M., Higuchi, H., Sabol, S. L. &Meyerhardt, J. (1990) Mol. Cell. Biol. 10, 3717–3726

84. Zhu, Y. S., Branch, A. D., Robertson, H. D. & Intuirrisi, C. E. (1994) DNA CellBiol. 13, 25–35

85. Laforge, K. S., Unterwald, E. M. & Kreek, M. J. (1995) Mol. Cell. Biol. 15,2080–2089

86. Wong, M., Rius, R. A. & Loh, Y. P. (1991) Brain Res. Mol. Brain Res. 11,197–205

87. Salzet, M. & Stefano, G. B. (1997) Brain Res. 768, 224–23288. Metz-Boutigue, M. H., Goumon, Y., Lugardon, K., Strub, J. M. & Aunis, D.

(1998) Cell. Mol. Neurobiol. 18, 249–26689. Kieffer, B., Goumon, Y., Diffmann, B., Lefevre, J. F., Aunis, D. & Metz-

Boutigue, M. H. (1998) J. Biol. Chem. 273, in press

The Antibacterial Activity of COOH-terminal PEAP29856

by guest on March 31, 2016

http://ww

w.jbc.org/

Dow

nloaded from

Dorsselaer, Dominique Aunis and Marie-Hélène Metz-BoutigueYannick Goumon, Karine Lugardon, Bruno Kieffer, Jean-François Lefèvre, Alain Van

CELLS 237 SECRETED BY STIMULATED CHROMAFFIN−ANTIBACTERIAL PEAP209

Peptides (PEAP) in Infectious Fluids: IMPORTANCE OF ENKELYTIN, THE Characterization of Antibacterial COOH-terminal Proenkephalin-A-derived

doi: 10.1074/jbc.273.45.298471998, 273:29847-29856.J. Biol. Chem. 

  http://www.jbc.org/content/273/45/29847Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/273/45/29847.full.html#ref-list-1

This article cites 86 references, 32 of which can be accessed free at

by guest on March 31, 2016

http://ww

w.jbc.org/

Dow

nloaded from