FEMS Microbiology Letters 233 (2004) 193–199

www.fems-microbiology.org

Interactions of lactoferricin-derived peptides with LPSand antimicrobial activity

Sebastien Farnaud a,*, Claire Spiller a, Laura. C. Moriarty a, Alpesh Patel a,Vanya Gant b, Edward W. Odell c, Robert. W. Evans a

a Randall Centre for Molecular Mechanisms of Cell Function, King�s College London, New Hunt�s House, Guy�s Campus,

St. Thomas Street, London SE1 1UL, UKb Department of Microbiology, University College Hospital, London, UK

c Division of Oral Medicine and Pathology, GKT Dental Institute, King�s College, London, UK

Received 16 January 2004; received in revised form 22 January 2004; accepted 24 January 2004

First published online 10 February 2004

Abstract

Synthetic peptides derived from human and bovine lactoferricin, as well as tritrpticin sequences, were assayed for antimicrobial

activity against wild-type Escherichia coli and LPS mutant strains. Antimicrobial activity was only obtained with peptides derived

from the bovine lactoferricin sequence and peptides corresponding to chimeras of human and bovine sequences. None of the

peptides corresponding to different regions of native human lactoferricin showed any antimicrobial activity. The results underline

the importance of the content of tryptophan and arginine residues, and the relative location of these residues for antimicrobial

activity. Results obtained for the same assays performed with LPS mutants suggest that lipid A is not the main binding site for

lactoferricin which interacts first with the negative charges present in the inner core. Computer modelling of the most active peptides

led to a model in which positively charged residues of the cationic peptide interact with negative charges carried by the LPS to

disorganise the structure of the outer membrane and facilitate the approach of tryptophan residues to the lipid A in order to

promote hydrophobic interactions.

� 2004 Published by Elsevier B.V. on behalf of Federation of European Microbiological Societies.

Keywords: Lactoferrin; Lactoferricin; Cationic antimicrobial peptides; LPS; MBC

1. Introduction

A wide variety of organisms produce antimicrobial

peptides as a primary innate immune strategy [1]. Typ-

ically, these peptides are relatively short (less than 100

amino acids), positively charged, amphiphilic and arereported to be active against bacteria, fungi, viruses and

protozoa [2]. Their modes of action can vary and are not

fully understood, but their main site of action is thought

to be the cell membrane. The enterobacterial OM bi-

layer consists of an inner monolayer containing phos-

* Corresponding author. Tel: +44-207848-6562; fax: +44-207848-

6485.

E-mail address: sebastian.farnaud@kcl.ac.uk (S. Farnaud).

0378-1097/$22.00 � 2004 Published by Elsevier B.V. on behalf of Federatio

doi:10.1016/j.femsle.2004.01.039

pholipids and an outer monolayer that is mainly

lipopolysaccharide (LPS) [3].

Different mechanisms have been proposed for the

uptake of antibiotics across the membrane [4,5]. Studies

on the interaction of antimicrobial peptides with model

phospholipid membranes have revealed low affinity forzwitterionic phospholipids [6]. After initial insertion into

themembrane, by adsorption of the unfolded form on the

surface of the negatively charged phospholipid mem-

brane, their positive charges are partially neutralised by

the negative charges of the phospholipids headgroups.

Hydrophobic forces promote the formation of a stable a-helical structure, driving the peptide further into the

monolayer [7]. This mechanism suggests that the differentsusceptibilities of wild-type and deep roughmutant stains

to hydrophobic antibiotics result from differences in the

n of European Microbiological Societies.

Fig. 1. LPS-mutant phenotypes. The four mutants derived from the

parental strain D21 carry a mutation in rfa. D21e19 is deficient in

component of the O-antigen, D21e7 is galactoseless, D21ef1 is a glu-

cose- and heptose-bound phosphateless LPS strain, and D21f2 is a

heptoseless LPS strain. The lipid A msbB mutant of MLK53 lacks the

myristic acid residue 2, the htrB mutant lacks the lauric acid residue 4

and the msbB htrB double mutant lacks both residues 2 and 4.

194 S. Farnaud et al. / FEMS Microbiology Letters 233 (2004) 193–199

organisation of the outer membrane and in the accessi-

bility of the phospholipid bilayer regions [4]. In addition,

the decrease of protein content of the outer membrane

found in some rough mutants is compensated for by an

increase in phospholipid [8], where such exposedphospholipid bilayer regions would allow the rapid

penetration of hydrophobic molecules [9].

Hydrophobicity, cationicity and secondary structure

have been implicated in the antimicrobial effect [10].

Human and bovine lactoferricins (LfcinH and B) are

antimicrobial peptides arising from pepsin cleavage of

human and bovine lactoferrin, respectively [11]. A re-

gion of 11 residues in LfcinB has been identified as es-sential for antibacterial effect [12]. Recently, two regions

in LfcinH covering residues 1–5 and 19–31 were found

to be important for antimicrobial activity [13]. Interac-

tion between LPS and both lactoferricins has been

investigated, using a synthetic octadecapeptide corre-

sponding to residues 20–37 in LfcinH, and the whole

LfcinB. In both cases, the 28–34 loop region of LfcinH

and its homologous region in LfcinB were found to beinvolved in LPS binding [14].

In the present work, the antimicrobial activities of a

variety of peptides, including human and bovine lac-

toferricin-derived peptides and tritrpticin, towards wild-

type and outer-membranes mutant strains of Escherichia

coli are compared, to clarify the role of LPS in pre-

venting interaction with phospholipids and the mode of

action of peptides on the outer membrane.

2. Materials and methods

2.1. Bacterial strains

The peptides were assayed against several strains of

E. coli with structurally different LPS. The strains canbe grouped according to the region of their LPS af-

fected (Fig. 1). The relevant genotype of the K12

strains is as follows: W3110 (wild-type E. coli), D21

(rfa+), D21e19 (rfa-11), D21e7 (rfa-1), D21f1 (rfa-1,

rfa-21) and D21f2 (rfa-1, rfa-31). D21 parental strains

and derivative mutants were obtained from the E. coli

Genetic Stock Center (cgsc.biology.yale.edu/top.html).

These strains have been characterised and their LPScomposition deduced after gas chromatography anal-

ysis of carbohydrates [15–18]. Lipid A mutant strains

MLK43, 986 and 1067 were kindly donated by Pro-

fessor C.R. Raetz (Duke University Medical Center,

Durham, NC). These strains are defective in late acy-

lations of lipid A and are knockout mutants: MLK53

(htrB1::Tn10), MLK1067 (msbB::Xcam) and MLK986

(htrB1::Tn10 msbB::Xcam). MLK1067 lacks the msbB-encoded (Kdo)2-(lauroyl)-lipidIVA myristoyltransferase

and produces penta-acylated lipid A that is devoid of

the myristic acid residue. MLK53 and MLK986 are

thermosensitive and grow well at 30 �C but poorly at

37 �C. They lack the htrB-encoded (Kdo)2-lipidIVA

lauroyltransferase and produce penta-acylated lipid Athat is devoid of the lauric acid residue. The msbB htrB

double mutant MLK986 produces lipid A that lacks

both the lauric acid and myristic acid residues. E. coli

B strain WA834 (WA707 DompT504) is derived from

the wild-type E. coli B WA707. Cultures were grown

overnight in Mueller Hinton Broth medium (Difco) at

37 �C, except strains MLK53, MLK986 and MLK1067

which are thermo-sensitive and were therefore grown at32 �C.

2.2. Cationic peptides

The peptides used in this study (Table 1) can be

divided in three groups: group A, based on the human

lactoferricin sequence (Lfcin H), group B, based on

the bovine lactoferricin sequence (Lfcin B) and groupC, comprising the more active tritrpticin and tritrpti-

cin (2–12). Peptides were synthesised by Advanced

Biotechnology Centre (ABC) (Imperial College Lon-

don) using an automatic synthesiser and analysed by

HPLC. They were stored in their freeze-dried form at

)20 �C and dissolved on the day of use, as described

in the method of Hancock (www.cmdr.ubc.ca/bobh/

MIC.htm).

2.3. Minimal bactericidal concentration measurement

Minimal bactericidal concentrations (MBCs)

(Table 2) were obtained using the recently modified

Table 1

Peptide investigated

Lfcin H: 1 NH2-GRRRRSVQWCAVSNPEATKCFQWQRNMRKVRGPPVSCIKRDSPIQCI-COOH 47

Lfcin H(1-5) -GRRRRS-----------------------------------------

Lfcin H(1-9) -GRRRRSVQW--------------------------------------

Lfcin H(11-20) -----------AVSNPEATKC---------------------------

Lfcin H(18-26) ------------------TKCFQWQRN---------------------

Lfcin H(18-26)19G------------------TGCFQWQRN---------------------

Lfcin H(18-26)21G------------------TKCGQWQRN---------------------

Lfcin H(18-26)22G------------------TKCFGWGRN---------------------

Lfcin H(18-26)25G------------------TKCFQWQGN---------------------

Lfcin H(21-31)5s ---------------------RWAWRLMRKVR----------------

Lfcin H(19-31)6s -------------------RPWAWPRLMRKVR----------------

Lfcin H(25-31) -------------------------RNMRKVR----------------

Lfcin H(28-31) ----------------------------RKVR----------------

Lfcin H(28-43) ----------------------------RKVRGPPVSCIKRDSP----

Lfcin H(37-43) -------------------------------------CIKRDSP----

Lfcin B(17-31) ------------------FKCRRWQWRMKKLGA---------------

Lfcin B(17-31)24A------------------FKCRRWQARMKKLGA---------------

Lfcin B(17-31)29A------------------FKCRRWQWRMKKAGA---------------

Lfcin B(20-25) ---------------------RRWQWR---------------------

Tritrpticin VRRFPWWWPFLRR

Tritrpticin(2-12) RRFPWWWPFRR

1 2 3 45 6

A1

B3

B4

A3

A2

A4

A5

A6

A7

A8

A10

A9

A11

A12

A13

A14

A15

B2

B1

C1

C2

All peptides were synthesised with L amino acids. (A) Peptides derived from human lactoferricin (LfcinH); (B) Peptides derived from bovine

lactoferricin (LfcinB). Regions (1–5) and (19–31) are framed; region (29–35), underlined, is the putative LPS binding region; peptides are numbered

according to the corresponding residues in human lactoferrin; substitutions are in bold; for a single substitution, residue number is indicated followed

by the replacement residue; for multiple substitutions, only the number of substitution (s) is indicated.

S. Farnaud et al. / FEMS Microbiology Letters 233 (2004) 193–199 195

method used in the Hancock laboratory (www.cmdr.

ubc.ca/bobh/MIC.htm) based on the classical microtitre

broth dilution recommended by the National Commit-

tee of Laboratory Safety and Standards (NCLSS).

2.4. Comparative modelling

Peptide modelling was performed using Modeller 4on a Silicon Graphic System. Tertiary structure pre-

dictions were based on sequence alignment with the

bovine lactoferricin sequence and its NMR-determined

3D structure (1LFC) available from the Protein Data

Bank (PDB).

3. Results

3.1. Antimicrobial activity of peptides against wild-typestrains

Peptides in Group A are based on four different re-

gions (1–4) of the whole LfcinH (Table 1) as well as

regions 5 and 6 overlapping partly regions 2 and 3, and 3and 4, respectively. No antimicrobial activity was de-

tected for any of the regions taken separately and based

on the non-modified sequence (Table 2).

Antimicrobial activity (Table 2) was obtained for two

peptides LfcinH(21–31)5s (A10) and LfcinH(19–31)6s

Table 2

MBCs of the peptides expressed in lM

Peptides WT

W3110

Lipid A mutants WA707 OmpT�

WA834

LPS–carbohydrate mutants

MLK53 MLK1067 MLK986 D21 D21e19 D21e7 D21f1 D21f2

LfcinH(1–5) >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560

LfcinH(1–9) >2560 >2560 >2560 2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560

LfcinH(11–20) >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560

LfcinH(18–26) >2560 2560 2560 2560 >2560 >2560 >2560 2560 2560 >2560 >2560

LfcinH(18–26)19G >2560 2560 2560 2560 >2560 >2560 >2560 2560 2560 >2560 >2560

LfcinH(18–26)21G >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560

LfcinH(18–

26)22G,24G

>2560 2560 2560 2560 >2560 >2560 >2560 2560 2560 >2560 >2560

LfcinH(18–26)25G >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 2560 >2560 >2560

LfcinH(21–31)5s 320 160 160 160 320 320 320 160 80 2560 1280

LfcinH(19–31)6s 640 640 320 320 640 320 640 320 320 640 640

LfcinH(25–31) >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560

LfcinH(28–43) >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560

LfcinH(37–43) >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560

LfcinB(17–31) 160 160 160 40 160 160 160 160 40 1280 1280

LfcinB(17–31)24A 640 320 320 320 640 320 640 320 640 640 640

LfcinB(17–31)29A 320 320 160 160 320 320 320 160 160 640 640

Lfcin(20–25) >2560 >2560 >2560 >2560 >2560 >2560 >2560 >2560 2560 >2560 >2560

Tritrpticin 10 10 10 10 10 10 10 10 10 320 10

Tritrpticin(2,10) 10 10 10 10 10 10 10 10 10 40 10

Peptides showing antimicrobial activity are in bold.

196 S. Farnaud et al. / FEMS Microbiology Letters 233 (2004) 193–199

(A11) which were based on region 4 of human lactof-

erricin. They both contain five substitutions chosen from

the bovine lactoferricin sequence. A higher MBC was

obtained for the peptide A11 which contains two addi-

tional proline residues inserted between each tryptophan

residue and their neighbouring arginine residue.

In Group B, peptides based on the bovine lactof-

erricin sequence (LfcinB), only peptides covering thewhole region from 17 to 31 displayed some antimicro-

bial activity, with an increase in MBC after substitutions

W24A and L29A were made.

Apart from tritrpticin, only peptides containing a net

positive charge of at least +5 and at least 1 tryptophan

showed some antimicrobial activity. Only native trit-

rpticin shows a net charge of only +4 but possesses three

tryptophan residues. It seems that a factor common toall the active peptides is a value of 6 for the sum of net

charge and tryptophan residues.

3.2. Comparative modelling

In order to predict the structural effect following

substitution(s) or insertion(s) of amino acid residues in

the peptides, the 3D structure of peptides A10 and B1was predicted and compared with their respective mu-

tated forms peptides A11 and B2 (Fig. 2). The predicted

structures of peptides B1 and B2 could be superimposed

with no differences in overall structure, suggesting that

the decrease in antimicrobial activity was due to the loss

of tryptophan residue. The opposite effect was observed

with peptides A10 and A11 whose models could not be

superimposed. The difference in antimicrobial activity,

despite the same number of charges and tryptophan

residues within both peptides, can only be due to

structural changes following the insertion of proline

residues. This result suggests the importance of a posi-

tion of the tryptophan relative to the arginine residues.

3.3. Antimicrobial activity of peptides against LPS-lipidA mutant strains

A slight decrease in MBC (Table 2) was observed

between the three penta- and tetra-acylated lipid A

mutant strains and the wild-type E. coli for all the ac-

tive peptides, with an inverse relationship between an-

timicrobial activity and the number of fatty acids. In

addition, a low antimicrobial activity was obtained withthree other peptides derived from the region (18–26) of

LfcinH with the three mutants. None of the native

peptides displayed any activity when tested against

these strains. This suggests that the alteration of the

very tight fatty acid packing of the hexa-acylation of

lipid A results in an increase in fluidity that could fa-

cilitate hydrophobic interactions as described with hy-

drophobic reagents [14]. Similarly, a slight increase inantimicrobial activity was observed with peptides A11

and B2 for OmpT� WA834 indicating that OmpT is

probably not directly involved in Lfcin binding, as

suggested for other Omp proteins [20]. This effect might

be the result of the increase in concentration of negative

charges on the outer membrane as suggested by Nika-

ido and Nakae [4].

Fig. 2. Structural effect of substitutions and insertions. Active peptides LfcinB(17–31) (B1) and LfcinH(21–31)5s (A10) were structurally compared to

their derivatives, LfcinB(17–31)24A (B2) and LfcinH(19–31)6s (A11), respectively. Tryptophan residues are in red, and positively charged arginine

and lysine residues are in green; inserted proline residues are in blue. Ribbon representation has been superimposed on the backbone; comparison of

(a) and (b), corresponding to peptides B1 and B2, respectively, shows that the two peptides are superimposable, implying that substitution W24A did

not create any structural alteration. Comparison of (d) and (c), corresponding to peptides A10 and A11, respectively, indicate that major structural

alterations followed the insertion of the two proline residues with a different spatial distribution of the positive charges relatively to the tryptophan

residues.

S. Farnaud et al. / FEMS Microbiology Letters 233 (2004) 193–199 197

3.4. Antimicrobial activity of peptides against LPS–carbohydrate mutant strains

The peptides were assayed for antimicrobial activity

with four E. coli strains of rough phenotype having

mutations in the LPS other than in lipid A (Table 2).

The mutants derived from the parental strain D21 have

lost progressively increasing amounts of sugar residues

which result in shorter LPS, as indicated in Fig. 1. An-

timicrobial effects were observed with the same peptides

as with the wild-type but with differing MBCs. Thechanges detected were not proportional to the decrease

in LPS chain length. For the first two mutants, D21e19

and D21e7, a decrease in MBC, indicating an increase in

sensitivity, was seen. The two other mutants with

shorter LPS presented the opposite effect. An increase in

MBC was observed with D21f1, even with tritrpticin.

For the last heptoseless mutant strain, D21f2, an MBC

higher than with the wild-type, but lower than forD21f1, was observed. This variation in MBC, non-pro-

portional to the LPS chain length, suggests a complex

mechanism.

4. Discussion

The mechanism(s) of action of antimicrobial peptideson their natural targets are poorly understood. It is

thought that most peptides bind ionically to the surface

of Gram negative bacteria and then insert into the

membrane outer leaflet [1]. Studies on LfcinB have

identified several parts of the sequence to be involved in

antimicrobial activity [19,21], emphasizing the impor-

tance of the WQW motif and identifying RRWQWR as

the minimum active sequence [20,22]. Sequence modifi-

cations have demonstrated that single amino acid sub-stitutions in longer peptides can alter the Gram positive/

negative selectivity, enhance or abolish activity [22–24].

In addition, as well as structure/activity, quantitative

relationships have been proposed [25]. Less is known of

the mode of action of LfcinH. Activity of truncated

LFcinH has been shown by several groups [11,13,26–28]

and together these data suggest that the region forming

a helix in the native protein (peptide A10) is important.Lfcin contains a short hydrophobic sequence flanked by

cationic sequence, similar to many other antimicrobial

peptides including the considerably more potent trit-

rpticin [20,27,29,30].

In the present study, we have attempted to determine

which of the components of LPS were interacting with

the cationic peptides. To this end, a group of peptides

were assayed against wild-type and a series of LPSmutants strains. We have tested peptides containing the

putative LPS binding site [14,31] (peptides A1–3 and

A10–16), peptides with variations in the FQW hydro-

phobic core (peptide A5–9), a hybrid bovine/human

sequence of predicted increased potency (peptide A12)

and its analogue with proline insertions between charge

and hydrophobic core to mimic the kinked structure of

tritrpticin (A13), and a series of charged peptides lack-ing the hydrophobic core but retaining the LPS binding

site (peptides A14–16). These peptides were compared to

bovine sequences and improved bovine sequences

(peptides B1–3) including the hydrophobic �active� corealone (peptides B4) and tritrpticin positive controls

(peptides C1–2).

198 S. Farnaud et al. / FEMS Microbiology Letters 233 (2004) 193–199

Many peptides proved unexpectedly inactive, possi-

bly because the wild-type target strain shows relative

resistance to these relatively weak peptides. Highly

charged peptides known to bind to LPS exerted no ac-

tivity (peptides A1–3). However, modified LfcinH andLfcinB and tritrpticin all exhibited sufficient activity

to allow differences in activity with the mutants to be

detected.

Only peptides containing a sum of tryptophan resi-

dues and net positive charges of at least 6 showed some

antimicrobial activity. The importance of the trypto-

phan was emphasised by the decrease in potency after

substitution of Trp8 in peptide B1 despite the predictedconservation in overall structure (Fig. 2(a) and (b)). The

decrease in potency between peptides A10 and A11

(Fig. 2(c) and (d)) suggests a link between both types of

interaction with the membrane. The insertion of proline

residues affects neither the net charge nor the trypto-

phan content of the peptide but only changes their rel-

ative position. As shown in Fig. 2(d), such insertions are

predicted to alter the tertiary structure of the peptideand further separate positively charged residues from

tryptophans. This interdependence of both kinds of in-

teractions, electrostatic and hydrophobic, for antimi-

crobial activity was confirmed by results obtained with

the LPS mutants.

Lipid A mutants did not show a major increase in

susceptibility to any peptides suggesting that in con-

trast to the effect observed with hydrophobic reagents,the tight fatty acid packing in lipid A is not the pri-

mary site of interaction for lactoferricin. The slight

decrease in MBC could be due to a less rigid penta-

and tetra-acylated lipid A, producing a more

fluid outer membrane with increased hydrophobic

interactions.

The successive increase and decrease in sensitivity

observed with successive shortening of LPS suggestselectrostatic interactions between the peptide basic

residues and the negative charges of LPS. Shortening of

O-antigen and part of the outer core, strains D21e19

and D21e7, respectively, allows unmasking of the neg-

ative charges carried by the phosphate groups in the

inner core increasing the ionic interaction. It is inter-

esting to note that the increase in susceptibility to

cationic peptides for the first two mutant strainsD21e19 and D21e7 is contrary to the increase in am-

picillin resistance previously described [17], which sug-

gests a different mechanism of action. Further

shortening of LPS in strains D21f1 and D21f2 results in

the loss of most of the negative charges and conse-

quently less interaction with the cationic peptides. The

compensation of the loss of further electrostatic inter-

action, with the heptoseless mutant D21f2, by a facili-tated access to the lipid A and stronger hydrophobic

interactions, could be responsible for the slight increase

in susceptibility compared to D21f1.

These results support a two-step mechanism where

positive residues of the cationic peptide first interact

with negative charges carried by the LPS; the resulting

disorganisation of the structure of the outer membrane

then enables the tryptophan residues to approach thelipid-A for hydrophobic interactions, leading to further

penetration of the outer membrane. This mechanism can

also account for the difference in MBC between peptides

B1 and B2 (Fig. 2(a) and (b)), since only the removal of

the tryptophan residue, and therefore the decrease in

hydrophobic interaction with no structural alteration,

can be responsible for the increase in MBC.

As suggested by Ulvatne and Vorland [24], the dis-ruption of the outer membrane structure is unlikely to

be the main factor leading to cell death. The suscepti-

bility of Gram negative bacteria to cationic peptides has

been proposed to be associated with factors that facili-

tate the transport of the peptide across the outer mem-

brane, such as the magnitude and location of the LPS

charge, the concentration of LPS in the outer mem-

brane, outer membrane molecular architecture and thepresence or absence of the O-antigen side chain [10].

Based on the mechanism suggested above, cationic

peptides should not be taken as antimicrobial agents but

probably more as disorganizing or permeabilizing

agents that could increase susceptibility to more efficient

antibiotics. To this aim more studies should be

done with peptides in conjunction with hydrophobic

antibiotics.

Acknowledgements

We are grateful to Professor C.R. Raetz from Duke

University Medical Center, Durham, for providing the

lipid A mutants strains. Sebastien Farnaud was funded

by a Wellcome Trust Project Grant (Ref. 059414); Claire

Spiller was funded by a grant from King�s College

London School of Biomedical Sciences Summer Stu-

dentship Scheme; Laura Moriarty was funded by a grantfrom the Nutricia Research Foundation; Alpesh Patel

was funded by a studentship from the BBSRC.

References

[1] Hancock, R.E.W. and Lehrer, R. (1998) Cationic peptides: a new

source of antibiotics. TIB. TECH. 16, 82–88.

[2] Martin, E., Gantz, T. and Lehrer, R. (1995) Defensins and other

endogenous peptide antibiotics of vertebrates. J. Leukoc. Biol. 58,

128–136.

[3] Labischinski, H., Barnickel, G., Bradaczek, H., Naumann, D.,

Rietschel, E.T. and Giesbrecht, P. (1985) High state of order of

isolated bacterial lipopolysaccharide and its possible contribution

to the permeability barrier properties of the outer membrane. J.

Bacteriol. 162, 9–20.

[4] Nikaido, H. and Nakae, T. (1979) The outer membrane of gram-

negative bacteria. Adv. Microb. Physiol. 20, 163–250.

S. Farnaud et al. / FEMS Microbiology Letters 233 (2004) 193–199 199

[5] Hancock, R.E.W., Raffle, V.J. and Nicas, T.I. (1981) Involvement

of the outer membrane in gentamicin and streptomycin uptake

and killing in Pseudomonas aeruginosa. Antimicrob. Agents

Chemother. 19, 777–785.

[6] Gazit, E., Lee, W.J, Brey, P.T. and Shai, Y. (1994) Mode of action

of the antimicrobial cecropin B2: a spectrofluorometric study.

Biochemistry 33, 10681–10692.

[7] Vorland, L.H., Ulvatne, H., Rekdal, O. and Svendsen, J.S. (1999)

Initial binding sites of antimicrobial peptides in Staphylococcus

aureus and Escherichia coli. Scand. J. Infect. Dis. 31, 467–473.

[8] Smit, J., Kamio, Y. and Nikaido, H. (1975) Outer membrane of

Salmonella typhimurium: chemical analysis and freeze-fracture

studies with lipopolysaccharide mutants. J. Bacteriol. 124, 942–

958.

[9] Ames, G.F., Spudich, E.N. and Nikaido, H. (1974) Protein

composition of the outer membrane of Salmonella typhimurium:

effect of lipopolysaccharide mutations. J. Bacteriol. 117, 406–416.

[10] Shai, Y. (1999) Mechanism of the binding, insertion and desta-

bilization of the phospholipid bilayer membrane by a-helicalantimicrobial and cell non-selective membrane-lytic peptides.

Biochim. Biophys. Acta 1492, 55–70.

[11] Bellamy, W., Takase, M., Yamauchi, K., Wakabayashi, H.,

Kawase, K. and Tomita, M. (1992) Identification of the bacteri-

cidal domain of lactoferrin. Biochim. Biophys. Acta 1121, 130–

136.

[12] Kang, J.H., Lee, M.K., Kim, K.L. and Hahm, K.S. (1996)

Structure–biological activity relationship of 11-residue highly

basic peptide segment of bovine lactoferrin. Int. J. Pept. Protein

Res. 48, 357–363.

[13] Moriarty, L.C. (2002). Ph.D. thesis, University of London.

[14] Elass-Rochard, E., Roseanu, A., Legrand, D., Trif, M., Salmon,

V., Motas, C., Montreuil, J. and Splik, G. (1995) Lactoferrin–

lipopolysaccharide interaction: involvement of the 28–34 loop

region of human lactoferrin in the high-affinity binding to

Escherichia coli 055B5 lipopolysaccharide. J. Biochem. 312, 839–

845.

[15] Boman, H., Eriksson-Grennberg, K.G., Normark, S. and Mats-

son, E. (1968) Resistance of Escherichia coli to penicillins. IV.

Genetic study of mutants resistant to D, LL-ampicillin concentra-

tions of 100 lg/ml. Genet. Res. 12, 169–185.

[16] Boman, H. and Monner, D.A. (1975) Characterisation of

lipopolysaccharides from Escherichia coli K-12 mutants. J. Bac-

teriol. 121, 455–463.

[17] Eriksson-Grennberg, K.G., Nordstrom, K. and Englund, P.

(1971) Resistance of Escherichia coli to Penicillins. IX. Genetics

and physiology of class II ampicillin-resistant mutants that are

galactose negative or sensitive to bacteriophage C21, or both. J.

Bacteriol. 108, 1210–1223.

[18] Monner, D.A., Jonsson, S. and Boman, H.G. (1971) Ampicillin-

resistant mutants of Escherichia coli K-12 with lipopolysaccharide

alterations affecting mating ability and susceptibility to sex-specific

bacteriophages. J. Bacteriol. 107, 420–432.

[19] Vaara, M. and Nurminen, M. (1999) Outer membrane perme-

ability in Escherichia coli mutants that are defective in the late

acyltransferases of lipid A biosynthesis. Antimicrob. Agents

Chemother. 43, 1459–1462.

[20] Sallmann, F.R, Baveye-Descamps, S., Pattus, F., Salmon, V.,

Branza, N., Spik, G. and Legrand, L. (1999) Porins OmpC and

PhoE of Escherichia coli as specific cell-surface targets of human

lactoferrin. J. Biol. Chem. 274, 16107–16114.

[21] Ulvatne, H., Haukland, H.H., Olsvik, O. and Vorland,

L.H. (2001) Lactoferricin B causes depolarization of the

cytoplasmic membrane of Escherichia coli ATCC 25922 and

fusion of negatively charged liposomes. FEBS Lett. 492,

62–65.

[22] Kuwata, H., Yamauchi, K., Teraguchi, S., Ushida, Y., Shimok-

awa, Y., Toida, T. and Hayasawa, H. (2001) Functional fragments

of ingested lactoferrin are resistant to proteolytic degradation in

the gastrointestinal tract of adult rats. J. Nutr. 131, 2121–

2127.

[23] Hoek, K.S, Milne, J.M, Grieve, P.M., Dionysius, D.A. and Smith,

R. (1997) Antibacterial activity in bovine lactoferrin-derived

peptides. Antimicrob. Agents Chemother. 41, 54–59.

[24] Ulvatne, H. and Vorland, L.H. (2001) Bactericidal kinetics of 3

lactoferricins against Staphylococcus aureus and Escherichia coli.

Scand. J. Infect. Dis. 33, 507–511.

[25] Strom, M.B., Haug, B.E., Rekdal, O., Skar, M.L., Stensen, W.

and Svendsen, J.S. (2002) Important structural features of 15-

residue lactoferricin derivatives and methods for improvement of

antimicrobial activity. Biochem. Cell. Biol. 80, 65–74.

[26] Azuma, M., Kojima, T., Yokoyama, I., Tajiri, H., Yoshikawa, K.,

Saga, S. and Del Carpio, C.A. (1999) Antibacterial activity of

multiple antigen peptides homologous to a loop region in human

lactoferrin. J. Pept. Res. 54, 237–241.

[27] Chapple, D.S., Joannou, C.L., Mason, D.L., Shergill, J.K., Odell,

E.D., Gant, V. and Evans, R.W. (1998) A helical region on human

lactoferrin. Its role in antibacterial pathogenesis. Adv. Exp. Med.

Biol. 443, 215–220.

[28] Chapple, D.S., Mason, D.L., Joannou, C.L., Mason, D.L.,

Odell, E.D., Gant, V. and Evans, R.W. (1998) Structure–

function relationship of antibacterial synthetic peptides ho-

mologous to a helical surface region on human lactoferrin

against Escherichia coli Serotype 0111. Infect. Immun. 66,

2434–2440.

[29] Odell, E.W., Sarra, R., Foxworthy, M., Chapple, D.S. and Evans,

R.W. (1996) Antibacterial activity of peptides homologous to a

loop region in human lactoferrin. FEBS Lett. 382, 175–178.

[30] Schibli, D.J., Epand, R.F., Vogel, H.J. and Epand, R.M. (2001)

Tryptophan-rich antimicrobial peptides: comparative properties

and membrane interactions. Biochem. Cell. Biol. 80, 667–

677.

[31] Wimley, W.C. and White, S.H. (1996) Experimental determined

hydrophobicity scale for proteins at membrane interfaces. Nat.

Struct. Biol. 3, 842–848.