Mechanisms of adrenomedullin antimicrobial action

Robert P. Allaker a,*, Paul W. Grosvenor b, David C. McAnerney a, Barry E. Sheehan a,Bakula H. Srikanta a, Keith Pell c, Supriya Kapas a

aOral Microbiology, Institute of Cell and Molecular Science, Queen Mary, University of London, Newark Street, London E1 2AT, UKbBioinformatics Group, Centre for Infectious Disease, Institute of Cell and Molecular Science, Queen Mary, University of London,

Newark Street, London E1 2AT, UKc Imaging Unit, Institute of Cell and Molecular Science, Queen Mary, University of London, Newark Street, London E1 2AT, UK

p e p t i d e s 2 7 ( 2 0 0 6 ) 6 6 1 – 6 6 6

a r t i c l e i n f o

Article history:

Received 14 July 2005

Accepted 2 September 2005

Published on line 13 October 2005

Keywords:

Adrenomedullin

Antimicrobial peptide

Structure–function

a b s t r a c t

The mechanism of antimicrobial action of the multifunctional peptide adrenomedullin (AM)

against Escherichia coli and Staphylococcus aureus was investigated. AM (52 residues) and AM

fragments (1–12, 1–21, 13–52, 16–21, 16–52, 22–52, 26–52 and 34–52 residues) were tested for

activity. Carboxy-terminal fragments were shown to be up to 250-fold more active than the

parent molecule. Minimum inhibitory concentration values of the most active fragments

(13–52 and 16–52) and the parent molecule were 4.9 � 10�2 and 12.5 mg/ml, respectively, with

E. coli. Ultrastructural analyses of AM treated cells demonstrated marked cell wall disruption

with E. coli within 0.5 h. Abnormal septum formation with no apparent peripheral cell wall

disruption was observed with S. aureus after 2 h. Outer membrane permeabilisation assays

with E. coli confirmed that the C-terminal fragments were significantly (P < 0.05) more active.

It is suggested that postsecretory processing may generate multiple AM congeners that have

enhanced antimicrobial activities against a range of potential targets.

# 2005 Elsevier Inc. All rights reserved.

avai lab le at www.sc iencedi rect .com

journal homepage: www.elsev ier .com/ locate /pept ides

1. Introduction

The epithelium represents the first barrier against infection

with pathogenic bacteria. In health, surfaces of higher

eukaryotes, such as plants, invertebrates and vertebrates,

including humans, possess a normal microflora which causes

no damage to its host. The reason, apart from physical

barriers, includes the production of gene-encoded antimicro-

bial peptides by epithelial cells [6].

Adrenomedullin (AM) is a 52 amino acid multifunctional

peptide [8] produced by a wide variety of tissues and cells. This

peptide has many properties in common with other cationic

antimicrobial peptides including human b-defensin-2. Such

molecules are defined as peptides of 12–50 amino acids with a

net positive charge of +2 to +7 [6]. Apart from human AM being

52 amino acid residues in length, AM also has a net positive

charge owing to an excess of basic amino acids (arginine and

lysine). The molecule also has a single intramolecular

* Corresponding author. Tel.: +44 20 7882 2388; fax: +44 20 7882 2191.E-mail address: R.P.Allaker@qmul.ac.uk (R.P. Allaker).

0196-9781/$ – see front matter # 2005 Elsevier Inc. All rights reserveddoi:10.1016/j.peptides.2005.09.003

disulphide bond between residues 16 and 21, along with an

amidated tyrosine at the carboxyl terminus. In common with

other antimicrobial peptides including human b-defensin-2,

the amphipathic structure characterized by spatially sepa-

rated hydrophobic and charged regions, permits bacterial

membrane intercalation. Chemically, therefore AM does

resemble an antimicrobial peptide, however, its mode of

interaction with bacteria is unknown.

A variety of modes of action with respect to antimicrobial

activity have been proposed for cationic antimicrobial pep-

tides, including critical membrane depolarization, creation of

physical holes in membranes, induction of hydrolytic

enzymes that degrade the cell wall, disturbance of membrane

function and disruption of critical intracellular processes [16].

The aim of this study was to gain insights into the

mechanism of antimicrobial action of adrenomedullin against

the Gram-negative bacterium Escherichia coli and the Gram-

positive bacterium Staphylococcus aureus.

.

p e p t i d e s 2 7 ( 2 0 0 6 ) 6 6 1 – 6 6 6662

2. Materials and methods

2.1. Organisms and culture conditions

E. coli BUE55 was used as an indicator organism to assess the

structure–function relationship of AM and AM fragments. This

strain was originally isolated because of its increased

sensitivity to polymyxin B and has been used in other studies

of antimicrobial peptides [13]. E. coli (NCTC 9001), S. aureus

(Oxford antibiotic sensitive strain; NCTC 6571) and C. albicans

(ATCC 24433) were also used. Microorganisms were grown in

Tryptone Soy Broth (TSB; Lab M) at 37 8C in air with 5% CO2.

2.2. Peptides

Synthetic AM and AM fragments (residues 1–12, 1–21, 13–52,

16–21, 16–52, 22–52, 26–52 and 34–52) were obtained from

Phoenix Pharmaceuticals (Karlsruhe, Germany). Fragmenta-

tion positions in the AM molecule are shown as follows:

YRQSMNNFQGLR12S13FGC16RFGTC21T22VQKL26AHQIYQFT34

DKDKDNVAPRSKISPQGY52.

2.3. Minimum inhibitory concentration (MIC) andminimum bactericidal concentration (MBC) determinations

MIC determinations of peptides and fragments were deter-

mined in a broth micro-dilution assay against E. coli BUE55 [3].

Serial 1 in 2 dilutions of peptides were made in TSB medium.

Wells were inoculated with a 1 in 10 dilution of an overnight

bacterial culture (final concentration approximately 5 � 105

colony forming units/ml). After incubation at 37 8C for 24 h,

MICs were recorded as the lowest concentration of peptide

inhibiting growth as measured by optical density (540 nm).

MBCs of peptides were determined by inoculating samples

from MIC determination wells onto blood agar base (Oxoid

CM271) with defibrinated horse blood (5%, v/v). Plates were

then incubated at 37 8C in air with 5% CO2 for 24 h.

2.4. Transmission electron microscopy of bacteria

S. aureus and E. coli cultures were centrifuged (10,000 � g for

5 min) to pellet cells and phosphate buffered saline added to

provide a final concentration of approximately 108 colony

forming units/ml. These suspensions were then treated with

AM (final concentration of 500 mg/ml) for various lengths of

time (30–120 min) at 37 8C. Peptide concentration and expo-

sure time were based upon the characterisation of human b-

defensin-3 [7]. The bacteria were then centrifuged (10,000 � g

for 5 min), cold glutaraldehyde (0.5%, v/v) in 0.1 M sodium

cacodylate buffer (pH 7.4) was added, and samples were held

for 2 h at 4 8C. After removal of fixative, 4% (w/v) molten

agarose was added to provide a solid pellet. Cacodylate buffer

was then added and samples kept on ice. Each pellet was cut

into 1 mm cubes and post-fixed for a further 2 h in 1% osmium

tetroxide. Dehydration was carried out in ascending concen-

trations of ethanol. Samples were immersed in a mixture of

propylene oxide and TAAB resin. Samples were then finally

embedded in TAAB resin and polymerized at 60 8C for 48 h.

Sections were stained for 20 min with 2% (w/v) aqueous uranyl

acetate followed by 6 min with Reynolds lead citrate. Sections

were examined in a Jeol JEM 1200 transmission electron

microscope operating at 80 kV.

2.5. Cell-bound 1-N-phenylnaphthylamine measurements

The mode of interaction of full length AM and AM fragments

with the surface of E. coli was studied with the hydrophobic

fluorescent probe 1-N-phenylnaphthylamine (NPN) [10]. Mid-

log-phase cells were centrifuged (10,000 � g for 5 min) and

suspended in 5 mM HEPES buffer (pH 7.2) at an optical density

of 0.5 at 540 nm. Cells were then exposed to peptide/peptide

fragments (100 mg/ml) for 1 h at 37 8C. NPN (dissolved in

acetone at a concentration of 500 mM) was added at a final

concentration of 10 mM. Samples were centrifuged (10,000 � g

for 1 min) and the NPN concentration of the supernatant, i.e.

dye not bound to permeabilized cells was determined by

measuring the fluorescence in a Fluostar Optima plate reader

(BMG Lab Tech). Excitation and emission wavelengths for NPN

were set at 350 and 420 nm, respectively. Controls without

added cells or AM were included.

3. Results

The pattern of E. coli growth inhibition by the complete AM

peptide is illustrated in Fig. 1. The lowest concentration of AM

to prevent bacterial growth was found to be 12.5 mg/ml

(2.07 mmol l�1). Carboxy-terminal AM fragments demon-

strated increased activity (up to �250) when compared to

whole AM peptide against E. coli BUE55 (Fig. 2). Fragments

demonstrated varying degrees of inhibition. MIC values of

0.049, 0.049, 0.098 and 0.39 mg/ml were demonstrated with

fragments 13–52, 16–52, 22–52 and 34–52, respectively (molar

values of 0.011, 0.012, 0.029 and 0.184 mmol l�1, respectively).

No inhibition was observed with AM fragments 1–12, 1–21, 16–

21 or 26–52. Corresponding MBC values of 0.049, 0.049, 0.19 and

1.56 mg/ml were demonstrated with the active fragments.

With the Oxford strain of S. aureus MIC values of 12.5, 0.098,

0.098, 0.19 and 0.19 mg/ml were demonstrated with whole AM

and AM fragments 13–52, 16–52, 22–52 and 34–52, respectively

(molar values of 2.07, 0.022, 0.023, 0.057 and 0.09 mmol l�1,

respectively). No inhibition was observed with the other four

fragments tested. Both the whole peptide and eight fragments

were shown to be inactive against the yeast C. albicans.

Morphological changes occurring upon exposure of E. coli

and S. aureus to 500 mg/ml AM were examined by transmission

electron microscopy. Marked disruption to the integrity of the

outer and cytoplasmic membranes with E. coli was observed in

50/100 of exposed bacteria examined within 0.5 h (Fig. 3). In the

corresponding control (0.5 h incubation with no peptide) only

10/100 demonstrated any evidence of membrane disruption.

With S. aureus no apparent lysis or gross leakage of cellular

cytoplasmic contents was observed within 2 h. However,

evidence of abnormal septum (cross wall) formation was

observed in 20/100 of exposed bacteria examined (Fig. 4). In

the corresponding control (2 h incubation with no peptide)

only 2/100 demonstrated any evidence of abnormal septum

formation.

With the fluorescent probe 1-N-phenylnaphthylamine, the

mean amount (% �S.E.M.) of dye remaining in the super-

p e p t i d e s 2 7 ( 2 0 0 6 ) 6 6 1 – 6 6 6 663

Fig. 1 – Growth inhibition of E. coli BUE55, as measured by optical density in a broth micro-dilution assay, with decreasing

concentrations (mg/ml) of adrenomedullin (mean W S.E.; n = 9).

natant, as an indication of outer membrane permeabilisation,

for the parent molecule, inactive fragments (1–12, 1–21, 16–21

and 26–52) and active C-terminal fragments (13–52, 16–52, 22–

52 and 34–52) was 89.89 � 4.80, 99.65 � 6.92 and 77.71 � 4.02,

respectively. A significant difference was found between the

active C-terminal and other fragments (P < 0.05; Student’s

independent t-tests).

4. Discussion

AM has 30% homology at the nucleotide level to the cecropin

group of antimicrobial peptides, and therefore, might have a

similar mechanism of action. The mechanism of action of

cecropins is thought to involve channel formation in mem-

Fig. 2 – Growth inhibition of E. coli BUE55, as measured by optic

concentrations (mg/ml) of adrenomedullin fragments (mean W S

branes and subsequent lysis [2]. It has been shown that AM has

high affinity with E. coli lipopolysaccharide (LPS) [1], and

therefore, it might be expected that Gram-negative bacteria

would be more susceptible than Gram-positive organisms.

However, as clearly shown by the minimum inhibitory

concentrations, this was not the case. While, C. albicans was

shown to be resistant to both the whole AM molecule and

fragments. It is suggested that the binding capacity of AM may

be absent or reduced in this yeast thus making it resistant to

the antimicrobial action of AM.

To help elucidate how E. coli and S. aureus are possibly

killed, morphological changes were examined by transmis-

sion electron microscopy following exposure to AM at a

concentration above the MIC for a short period of time. Marked

changes were observed in the cell wall of E. coli within 0.5 h,

al density in a broth micro-dilution assay, with decreasing

.E.; n = 6).

p e p t i d e s 2 7 ( 2 0 0 6 ) 6 6 1 – 6 6 6664

Fig. 3 – Electron micrographs of untreated (A) and AM

treated (B) E. coli. Treated bacteria exposed to 500 mg/ml

adrenomedullin for 0.5 h. Note marked cell wall disruption

in treated cells.

while no similar disruption of the peripheral cell wall was

observed with S. aureus. However, when compared to cells not

exposed to AM, disruption of normal cross wall formation was

evident at 2 h with S. aureus. No disruption was evident at

earlier time points. Staphylococci division occurs with the

peripheral cell wall growing inwardly and eventually forming

a transverse cross wall (septum), the newly synthesized

peptidoglycan then undergoes localized hydrolysis to com-

plete cell separation [5]. Changes to the mechanism by which

the septum is deformed into the peripheral wall, caused by

antimicrobial molecules, have been shown to cause lethality

during cell division [4]. Further studies to fully elucidate such

disruption of this critical process with respect to AM are

required. It is possible that there are multiple potential anionic

targets of peptide action against bacteria, including nucleic

acid, cell wall, cytoplasmic membrane and enzymes, and that

these targets are differentially accessed according to the

bacterium and peptide concerned [4].

Shorter more active peptides may well yield important new

information relevant to the mechanism of action of AM. A

significant structure–function relationship was found

between the antimicrobial effect of AM, against both Gram-

positive and Gram-negative bacteria, and the carboxy-term-

inal portion of the peptide. Sequence analysis, using various

secondary structure prediction severs [17] all indicated a

conserved a-helical region in AM from residues 21 to 33, while

the remaining residues show no consistently ordered regions

(Grosvenor and Allaker, unpublished observations). Results

with fragments indicate the proposed a-helix is required for

optimum antimicrobial activity. Fragments (13–52 and 16–52)

demonstrated 250-fold greater activity when compared to the

whole AM peptide against E. coli. Cleavage within the proposed

a-helix resulted in lack of antimicrobial activity as demon-

strated by fragment 26–52. Smaller peptides may be able to

penetrate the bacterial cell more effectively and act against

intracellular targets including nucleic acid. Implementation of

COMPASS [15] allowed a sensitive profile–profile alignment of

the AM sequence against known structural sequences in PDB

(Protein DataBase); this gave a significant E value (1.1e�03) to

part of the DNA dependant RNA polymerase (RpoB) of

Sacchromyces spp. corresponding to residues 14–52 (Grosvenor

and Allaker, unpublished observations). Although RpoBs show

a low sequence homology to each other these enzymes are

structurally conserved amongst eukaryotes and prokaryotes.

The comparison indicates another potential mode of action for

AM, namely interference with bacterial holoenzyme tran-

scription complexes. Thus, disruption of DNA transcription

may well result in abnormal cross wall formation as

demonstrated in this study.

Interaction of certain antimicrobials with Gram-negative

bacteria is well known to cause enhancement of outer

membrane permeability to the large lysozyme molecule

(14 kD). Growth of E. coli was shown to be significantly reduced

in the presence of AM and lysozyme, in comparison to AM

alone. While, growth of S. aureus was not significantly reduced

in the presence of AM and lysozyme, in comparison to AM

alone (Allaker, unpublished observations). The fluorescent

probe 1-N-phenylnaphthylamine was also used in the current

study to confirm that outer membrane disruption of E. coli

occurs in the presence of AM and that carboxy-terminal

fragments of AM are significantly more active.

It is generally agreed that the AM concentration to inhibit

bacterial growth is in excess of the levels normally found in

the systemic circulation [8]. However, it is not inconceivable

under certain circumstances, such as chronic inflammatory

conditions, that elevated plasma levels may contribute to an

p e p t i d e s 2 7 ( 2 0 0 6 ) 6 6 1 – 6 6 6 665

Fig. 4 – Electron micrographs of untreated (A) and AM treated (B and C) S. aureus. Treated bacteria exposed to 500 mg/ml

adrenomedullin for 2 h. Note abnormal cleavage furrow in treated cells.

antimicrobial response. Indeed, in periodontitis, a chronic

inflammatory disease of the supporting structures of the teeth

AM concentrations found in gingival crevicular fluid are in the

1–2 mg/ml range [11]. All MIC values of the fragments of AM

tested in the current study were found to be below this

concentration.

Recently, it has been shown that the cationic antimicrobial

peptide cathelicidin LL-37, found on the skin surface, is

processed by a serine protease-dependant mechanism into

novel antimicrobial peptides with enhanced antimicrobial

action and loss of pro-inflammatory activity [14]. It is

suggested that the single human cathelicidin gene generates

multiple products with a range of biological activities, each

relevant to the local environment in which they are released.

The degradation of AM by host plasma membrane enzymes to

major degradation products of 2–52 and 8–52, with smaller

amounts of 26–52, 27–52, 28–52 and 33–52 [9] has been

demonstrated. In a similar study [12] degradation of AM by

matrix metalloproteinase-2 to fragments of 8–52, 11–52, 23–52,

29–52, 11–28 and 11–22 has also been shown. Thus, post-

secretory processing may generate multiple AM congeners

with multiple potential targets that have enhanced antimi-

crobial activities and host immunoregulatory activities.

Acknowledgements

This work was funded by St. Bartholomew’s and the Royal

London Charitable Foundation.

r e f e r e n c e s

[1] Allaker RP, Kapas S. Adrenomedullin expression by gastricepithelial cells in response to infection. Clin Diagn LabImmunol 2003;10:546–51.

p e p t i d e s 2 7 ( 2 0 0 6 ) 6 6 1 – 6 6 6666

[2] Christensen B, Fink J, Merrifield RB, Mauzerall D. Channel-forming properties of cecropins and related modelcompounds incorporated into planar lipid membranes.Proc Natl Acad Sci USA 1988;85:5072–6.

[3] Dykhuizen RS, Frazer R, Duncan C, Smith CC, Golden M,Benjamin N, et al. Antimicrobial effect of acidified nitrite ongut pathogens: importance of dietary nitrate inhost defence. Antimicrob Agents Chemother1996;40:1422–5.

[4] Friedrich CL, Moyles D, Beveridge TJ, Hancock REW.Antibacterial action of structurally diverse cationicpeptides on Gram-positive bacteria. Antimicrob AgentsChemother 2000;44:2086–92.

[5] Giesbrecht P, Kersten T, Maidhof H, Wecke J.Staphylococcal cell wall: morphogenesis and fatalvariations in the presence of penicillin. Microbiol Mol BiolRev 1998;62:1371–414.

[6] Hancock REW, Diamond G. The role of cationic peptides ininnate host defences. Trends Microbiol 2000;8:402–10.

[7] Harder J, Bartels J, Christophers E, Schroder JM. Isolationand characterisation of human b-defensin-3, a novelhuman inducible peptide antibiotic. J Biol Chem2001;276:5707–13.

[8] Hinson JP, Kapas S, Smith DM, Adrenomedullin, amultifactorial paracrine regulator. Endocrinol Rev2000;21:138–67.

[9] Lewis LK, Smith MW, Brennan SO, Yandle TG, Richards AM,Nicholls MG. Degradation of human adrenomedullin (1–52)by plasma membrane enzymes and identification ofmetabolites. Peptides 1997;18:733–9.

[10] Loh B, Grant C, Hancock RE. Use of the fluorescent probe 1-N-phenylnaphthylamine to study the interactions ofaminoglycoside antibiotics with the outer membrane ofPseudomonas aeruginosa. Antimicrob Agents Chemother1984;26:546–51.

[11] Lundy FT, O’Hare MMT, Briggs JE, McKibben BM, Linden GJ.Adrenomedullin in gingival crevicular fluid in periodontalhealth and disease. J Dent Res 2001;80:1176.

[12] Martinez A, Oh H-R, Unsworth EJ, Bregonzio C, Saaverda JM,Stetler-Stevenson WG, et al. Matrix metalloproteinase-2cleavage of adrenomedullin produces a vasoconstrictor outof a vasodilator. Biochem J 2004;383:413–8.

[13] Moore AJ, Beazley WD, Bibby MC, Devine DA. Antimicrobialactivity of cecropins. J Antimicrob Chemother1996;37:1077–89.

[14] Murakami M, Lopez-Garcia B, Braff M, Dorschner RA, GalloRL. Postsecretory processing generates multiplecathelicidins for enhanced topical antimicrobial defense. JImmunol 2004;172:3070–7.

[15] Sadreyev RI, Baker D, Grishin NV. Profile–profilecomparisons by COMPASS predict intricate homologiesbetween protein families. Protein Sci 2003;12:2262–72.

[16] Yeaman MR, Yount NY. Mechanisms of antimicrobialpeptide action and resistance. Pharmacol Rev2003;55:27–55.

[17] Secondary structure analysis web references:;PSIPRED, http://bioinf.cs.ucl.ac.uk/psipredJNET,http://www.compbio.dundee.ac.uk/�www-jpred/jnetSAM,http://www.cse.ucsc.edu/research/compbio/HMM-apps/T02-query.htmlJUFO, http://www.jens-meiler.de/jufo.html.