+ All Categories
Home > Documents > Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

Date post: 05-Jan-2017
Category:
Upload: ngonhan
View: 221 times
Download: 0 times
Share this document with a friend
31
:ritical Reviews in Food Science and Nutrition, 42(2):91-121 (2002) Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria Juan M. Rodriguez,l,* Maria ,. Martinez,l and Jan Kok2 1 Departamento de Nutrición y Bromatologra III, Universidad Complutense de Madrid, 28040 Madrid, Spain; 2Department of Genetics, Groningen Biomolecular Sciences and Biotechnology Institute, University of Groningen, 9751 NN Haren, The Netherlands Referee: Dr. Helen Dodd, Food Safety Science Division, Institute of Food Research, Norwich Research Park, Colney NR4 7UA, Norwich, United Kingdom * Corresponding author. ABSTRACT: Pediocin PA-l is a broad-spectrum lactic acid bacteria bacteriocin that shows a particularly strong activity against Listeria monocytogenes, a foodborne pathogen of special concern among the food industries. This antimicrobial peptide is the most extensively studied class Ila (or pediocin family) bacteriocin, and it has been sufficiently well characterized to be used as a food biopreservative. This review focuses on the progress that have been made in the elucidation of its structure, mode of action, and biosynthesis, and includes an overview of its applications in food systems. The aspects that need further research are also addressed. In the future, protein engineering, genetic engineering and/or chemical synthesis may lead to the development of new antimicrobial peptides with improved properties, based on some features of the pediocin PA-l molecule. KEY WORDS: pediocin PA. Listeria monocytogenes. bacteriocins, lactic acid bacteria, food preservatives, antimicrobials, food safety I. INTRODUCTION labile bacteriocins). Classna, one of the subgroups in which class n bacteriocins may be divided, is a1so terrnedthe "pediocin family" after the flfSt and most extensively studiedrepresentant of this class, pediocin PA-1.4 Members of this subgroup ("pediocin-like" bacteriocins) show very strong antilisteria1activity and have 40 to 60% sequence similarity (Table 1). The N-terrnina1region is par- ticularly weIl conservedand contains a conserved "pediocin box" motif (YGNGVXCXK).5 Nisin is currently the only bacteriocin licenced as a food additive in over 45 countries.6 However , severa1studies have shown that there are other LAB bacteriocins with potentia1 for use as food preservatives, particularly class na bacteriocins due to their antilisteria1 activity} In fact, the biopreservative potentia1 of pediocin PA-l has a1ready beencornrnercia11y exploited. The pediocin P A-1-containing ferrnentateAlta TM 2341 is a com- mercial food ingredient reported to extend the shelflife of a variety of foods and, particularly, to Bacteriocins of lactic acid bacteria (LAB) are attractive to the food industry becausethey may be used as natural biopreservatives and contribute to the improvement of the microbiological qua1ity of foods.l Also, the use of these ribosoma1ly syn- thesized antimicrobial peptides may allow a sig- nificant reduction in the levelof chemica1 addi- tives and/or in the intensity of the physical treatments currently employed during food pro- cessing. Therefore, they could a1sohelp to pro- vide healthier foods} Since the late 1920s,when the flfSt reports on the antimicrobia1 activity of a lactococca1 bacterio- cin (later ca1led nisin) were made, a large number of LAB bacteriocins have been identified, particu- larly in the last few years. On a scientific basis, LAB bacteriocins can be divided into three main classes:3,4 classI (lantibiotics), classII (sma1l heat- stable nonlantibiotics), and class rn (large heat- 1040:~398/02/$.50 @ 2002 by CRC Press LLC 91
Transcript
Page 1: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

:ritical Reviews in Food Science and Nutrition, 42(2):91-121 (2002)

Pediocin PA-1, a Wide-Spectrum Bacteriocin tromLactic Acid Bacteria

Juan M. Rodriguez,l,* Maria ,. Martinez,l and Jan Kok2

1 Departamento de Nutrición y Bromatologra III, Universidad Complutense de Madrid, 28040 Madrid, Spain;

2Department of Genetics, Groningen Biomolecular Sciences and Biotechnology Institute, University of Groningen,9751 NN Haren, The Netherlands

Referee: Dr. Helen Dodd, Food Safety Science Division, Institute of Food Research, Norwich Research Park, Colney NR4 7UA,

Norwich, United Kingdom

* Corresponding author.

ABSTRACT: Pediocin PA-l is a broad-spectrum lactic acid bacteria bacteriocin that shows a particularly strongactivity against Listeria monocytogenes, a foodborne pathogen of special concern among the food industries. Thisantimicrobial peptide is the most extensively studied class Ila (or pediocin family) bacteriocin, and it has beensufficiently well characterized to be used as a food biopreservative. This review focuses on the progress that havebeen made in the elucidation of its structure, mode of action, and biosynthesis, and includes an overview of itsapplications in food systems. The aspects that need further research are also addressed. In the future, proteinengineering, genetic engineering and/or chemical synthesis may lead to the development of new antimicrobialpeptides with improved properties, based on some features of the pediocin PA-l molecule.

KEY WORDS: pediocin PA.

Listeria monocytogenes.

bacteriocins, lactic acid bacteria, food preservatives, antimicrobials, food safety

I. INTRODUCTION labile bacteriocins). Class na, one of the subgroupsin which class n bacteriocins may be divided, isa1so terrned the "pediocin family" after the flfSt andmost extensively studied representant of this class,pediocin PA-1.4 Members of this subgroup("pediocin-like" bacteriocins) show very strongantilisteria1 activity and have 40 to 60% sequencesimilarity (Table 1). The N-terrnina1 region is par-ticularly weIl conserved and contains a conserved"pediocin box" motif (YGNGVXCXK).5

Nisin is currently the only bacteriocin licencedas a food additive in over 45 countries.6 However ,severa1 studies have shown that there are otherLAB bacteriocins with potentia1 for use as foodpreservatives, particularly class na bacteriocinsdue to their antilisteria1 activity} In fact, thebiopreservative potentia1 of pediocin PA-l hasa1ready been cornrnercia11y exploited. The pediocinP A-1-containing ferrnentate Alta TM 2341 is a com-

mercial food ingredient reported to extend theshelflife of a variety of foods and, particularly, to

Bacteriocins of lactic acid bacteria (LAB) areattractive to the food industry because they maybe used as natural biopreservatives and contributeto the improvement of the microbiological qua1ityof foods.l Also, the use of these ribosoma1ly syn-thesized antimicrobial peptides may allow a sig-nificant reduction in the levelof chemica1 addi-tives and/or in the intensity of the physicaltreatments currently employed during food pro-cessing. Therefore, they could a1so help to pro-vide healthier foods}

Since the late 1920s, when the flfSt reports onthe antimicrobia1 activity of a lactococca1 bacterio-cin (later ca1led nisin) were made, a large numberof LAB bacteriocins have been identified, particu-larly in the last few years. On a scientific basis,LAB bacteriocins can be divided into three mainclasses:3,4 class I (lantibiotics), class II (sma1l heat-stable nonlantibiotics), and class rn (large heat-

1040:~398/02/$.50@ 2002 by CRC Press LLC

91

Page 2: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

0c

.uo

.;:Q)-ucu

ID

$

UIUICU(3-OUIQ)UCQ)~0"Q)

U>

'0.U

...«W O.J CID .-

« EI-«

Q.~~\.:1

U \.:1 H~ ~ ~ V):I:UZ Z \.:1:I: ~:I:Uz~3 3 <\.:1~Q.\.:1~\.:1~\.:1~\.:1 \.:1~\.:1<~\.:10'Q.<\.:1< V)\.:1Q.~\.:1Q.:I:HZ« «H~<H\.:1<\.:1-.J\.:1~ ~>-.J<\.:1-.J<\.:1\.:1\.:1\.:1~ri:.ri:.\.:1V)\.:1:I:\.:1\.:11-\.:11-V)1-\.:1\.:1v)V)\.:10'V)\.:1<-.J~<I-Z Z><-.J><-.J~U.~3-.J\.:1\.:1ll.33333<\.:1Z\.:1Z\.:1\.:1V)\.:1\.:1\.:1\.:1\.:1~\.:1 ~ V)<Z ZZ\.:1 XZV)\.:1

<H H«<HH ~ >H> \.:1V) ~<--J-.J\.:1> H»Z~Z\.:1Z~~<H HHI-Z\.:1Z\.:1Z:I::I:I-\.:10'~\.:1O'H<\.:1H\.:1>H>-I-\.:1\.:1<\.:1HHHHH\.:1\.:1~HHHHHUUZV)H«UJ<\.:1UJZU1-1-\.:10'\.:1V)V)O'0'\.:1~UJ\.:1I-I-HI-Hu. <u.~<V)~V)< <{ <1- < <{ < < < Z <i < < <i~~ UJ~UJUJO'UJO'~UJO'\.:1<\.:1\.:1V)\.:1\.:1\.:1<{\.:1Z\.:1\.:1333~3333~3333QQQZQZ ZZZQOzO

~~~~C:~~~~~~~~UUUUUUUUUXUUUV)~V)~~~~~~~~~~:I:Z:I:~zV)V)I-UJ~O'V)~~~~z~~ ~~~V)~zV)\.:11-\.:1zzl-l-V)zzzzzUUUUUUUUUXUUU

~~5~~55~~~~~~~~~\.:1~~~\.:1\.:1\.:1\.:11..?\.:1zzzzzz zzzzzzz\.:1\.:1\.:1\.:1\.:1~1..?\.:1\.:1\.:1\.:11..?\.:1

~~~~~~ ~~~~c~~~~~~~~ ~>H~ï=~~

\.:1 < <1-<1-1-V) <:I:

I: MN ~ca ca

Lr\O C CM .r- .r-

1.0 >-UU MM N 00 M

I M C .r- .r- Z M< < .r- S- S- ~ C Q. vQ. < C < U QJ QJ .r- >

C Q. .r- .r- 4-J 4-J C U CC .r- C r- C S- U U .r- 0 .r- c

.~ U C .~ o.~ QJ ~ ~ U .~ U .~U 0 .r- U U U 4-J ..0 ..0 .r- S- 0 U0 S- U ~ .r- 0 C 0 0 S- QJ S- S-

.r- QJ ~ > U U QJ C C ~ 4-J QJ QJ~~~ S-I/) :3 I/) S- S- > u~ >QJ C ~ :3 .r- QJ QJ ~ ~ ~ ~ C .~Q.UJV)UQ.-.J~uucacaUJQ

.0Q.IucQ.I:J0"Q.IVI

VI:JVICQ.IVIC0U

H

.

~

3

>1

UI

~

UI

~I

~

.0(/)-Q):J~

:2~(/)(DQ)'- c:

"O.UQ) .Q~ '-

Q)2(/) <.Jc: mo.Cu o

E

~~0"0~ c:-m5j<t:(/) c:Q) .-'- Uc.mQ)~'- m(/) (/)'- OQ)-

i<c:

"0.-Q) Uc: m

=~'- :JQ) <.J

"0c: ,

::><.Q c:(Q .UN .-

.'-

-mCO >

mQ).c~ Oc:-

;~~W~

c:Q)(Q

"ONE .-.riO(/)~~ .->

"Oa.c:Q) t: .U-.-oc.U<.Jm m .-

"O~<.Jm (/)

~(/) .-

!0 c.

Page 3: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

inhibit the growth of Listeria monocytogenes inready-to-eat meat products.

In the past, several pediocin PA-1-producingLAB strains were independently isolated in dif -

ferent labs.8-19 However, in many cases the bacte-riocin produced received different names ( e.g.,pediocins PA-l, AcH, JD, Bac and 347,mesentericin 5) before identification and realiza-tion that all were the same molecule. As it isdesirable to conform to a uniform nomenclature,lthe more widespread name, pediocin PA-l, hasbeen adopted for this review.

Despite the fact that pediocin PA-1-produc-ing LAB have been inadvertly or empirically usedas starter cultures for many years, most of thecurrent knowledge on pediocin PA-l has beengenerated since 1992. In that year, the determina-tion ofthe pediocin PA-l amino acid sequence,20-21the application of improved protocols for its pu-rification,21 and the identification of the pediocinP A-l operon,22-23 greatly facilitated research onthis subject. In addition, the continuous develop-ment of new and improved biotechnological toolshave favored the increasing number of studiesdedicated to this bacteriocin. The purpose of thisarticle is to review the now enormous literatureon an antimicrobial peptide with such industrialpotential. Advances in the understanding of itsstructure, mode of action, and biosynthesis willbe high1ighted. Following that, we discuss thedevelopment of methods that allow its specificdetection and quantification. The results of thestudies involving application of pediocin PA-l infood substrates will also be presented. Finally, thelast part of this review exposes the main conclu-sions and future prospects of this interesting anti-bacterial peptide.

II. PHYSICOCHEMICAL PROPERTIES

induce toxicity on sensitive cells as syntheticpediocin is biologically active and has a specificactivity comparable to that of the natural fonn.24Other remarkable characteristics of the moleculeare the absence of phenylalanine, leucine,glutamine, or arginine residues and the presenceof four lysine and three histidine residues andonly one aspartate. Titration offree cysteine resi-dues with Ellman's reagent, tryptic cleavage pat-tems, mass spectrometry data, and pediocin re-duction in the presence of dithiothreitol (DTT)have provided evidence for the four cysteine resi-dues fonning two disulfide bonds (C9-CI4 andC24-C44),20.24-26 which are also spontaneouslyfonned in synthetic pediocin PA-1}4 As a reflec-tion of its amino acid composition and sequence,pediocin PA-l is a cationic peptide with a basicpI. Va1ues ranging from + 7 to +3 have been re-ported for the net charge at pH 6,1.27.28 and from8.6 to 10 for the pV.20.21.29 A comparison of theprimary structures of the pediocin-like bacterio-cins suggests that their polypeptide chains can bedivided into two functional modules, a conservedhydrophilic N-tennina1 13-sheet and a more di-verse hydrophobic or amphiphilic C-tenninala-helica1 domain24.30 (Table 1 ).

lnitia1ly, the secondary structural organiza-tion of the peptide was predicted by Henderson eta1.w following primary sequence ana1ysis by Chou-Fasman and Kyte-Coolite protocols. It is pro-posed that the structure would consist of randomcoils and 13-tums with a propensity of residues A21through 125 for a 13-sheet and a primarily hydro-philic profile except for the predicted 13-sheet area.In addition, the fragment T 22 to A34 that containsthe single tryptophan and methionine residueswould define a hydrophobic region in close spa-tia1 proximity with the C-tenninus due to thedisulfide bond C24-C44.

More recently, Chen et a1}7 predicted thepediocin PA-l secondary structure on the basis ofsingle and multiple sequence a1ignments and builta model for its 19 N-tennina1 amino acids. Themodel consisted of two 13-strands connected by atight 4-residues 13-hairpin. The N-tennina113-tumwould include the consensus sequence (YGNGV)of class lIa bacteriocins, and the tip of the13-hairpin would contain two positively chargedresidues (KIl and HIJ. The C24-C44 disulfide bond

A. Structure

The primary structure of pediocin PA-l hasbeen determined by Edman degradation of thepurified peptide and by sequence analysis of thestructural gene}O-23 Pediocin PA-l is a 44 arninoacid peptide with no postranslationa1 modifica-tions.20-21 The primary structure seems enough to

Page 4: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

would bring together H42' K43' and possibly K20.NMR data had previously shown an orderedf3-structure in the vicinity of the two cysteineresidues (C9 and Cl4)ofleucocin A, another classlIa bacteriocin.31 Figure 1 shows a modelof thepediocin PA-l structure.

B. Molecular Weight

The theoretical molecular weight of pediocinp A-l , calculated from the amino acid sequence,could be 4628 or 4624 Da, in the absence orpresence of two disulfide bonds, respectively }0.24Experimental determination of the molecularweights of natural and synthetic pediocin byelectrospray mass spectrometry revealed that thecysteine residues are oxidized and joined by twodisulfide bonds.16,17,24 Previously, a molecularweight of approximately 16,500 had been esti-mated by gel filtration of a partially purifiedsample. This value was nearly an exact multipleof the actual molecular weight, suggesting thepresence of a stable tetramer.8

c. Solubility and Stability

from Lb. plantarum WHE 92 showed that pediocinPA-l was perfectly stabIe after 21 days when thepH of the extract was mantained at 4, 5, or 6, butha1f of the activity dissapeared at pH 7.17 Al-though antimicrobia1 activity was unaffected byheating at 80°C for 60 min and at loo°C for 10min, the effect of a treatment at 121°C for 15 minis controversia1 as va1ues of residua1 activity of6%8 and 60%10 have been reported. Freezing(-25°C) or refrigeration (0 to 8°C) storage did notreduce the activity of pediocin samples after 6months and 12 weeks, respectively.321n contrast,activity was stabIe at room temperature (22 to25°C) for 6 weeks, but then it progressively de-creased and over 50% was lost in 12 weeks.32Recently, it has been shown that purified pediocinPA-l stored at 4 and 25°C is stabIe at pH 5 but notat pH 7, while it remains stabIe after storage at-20°C at both pH values.33

Pediocin PA-l activity is unaffected by treat-ment with phospholipase C, catalase, lysozyme,Dnases, Rnases, or lipases but is lost after incuba-tion with proteolytic enzymes such as trypsin,

papain, ficin, a-chimotrypsin, protease IV, pro-tease XIV, protease XXIV, and proteinase K.8.10.34Pediocin activity rapidly decreased when moni-tored in a culture of Lb. plantarum WHE 92, mostprobably because of hydrolysis by proteolyticenzymes produced at the end of the exponentialgrowth phase.17 As a result, the percentage ofresidua1 activity ranged from 1 to 6% after 21days of incubation at 15°C. The absence of con-taminating proteases may a1so explain why syn-thetic pediocin is more stabIe than the moleculeisolated from the natura1 source.24

After storage in 30% acetonitrile-0.1 %trifluoroacetic acid (TFA) for 1 week or more at-15°C, pediocin molecules became oxidized,showing a molecular weight increase of 16 ormultiples of 16.16 Similarly, after storage in 30%2-propanol-0,1% TFA for 8 months at 4°C, about50% of natural and synthetic pediocin had beentransfonned to a more hydrophilic and less activefonn (pediocinPA-1-ox).24 The pediocin PA-1-oxmolecular weight had increased by 16 relative tothat of pediocin PA-l and had suffered a 100- to200-fold reduction in its specific activity. Theoxidation of the sulfur atom of the methionineresidue of pediocin PA-l. and its transfonnation

The fact that many studies on pediocin PA-lsolubility and stability have been performed withpartially purified samples makes it difficult todirectly compare results obtained by differentauthors.

Pediocin PA-l is stable in dilute aqueoussolutions, a1though the molecules can form activeaggregates and tend to do so more readily as thepediocin concentration increases and at refrigera-tion temperatures.10 Aggregation seems to involvecomponents of the growth medium rather thanbeing an autocatalytic process because in contrastto pediocin PA-l in supematants or to partia11ypurified samples, the homogeneous peptide doesnot aggregate even at high concentrations.20

Samples containing pediocin PA-l retainedtotal or partial activity after exposure for 24 h topH va1ues ranging from 2 to 10 at room tempera-ture, although the highest stability was achievedbetween pH 4 and 6.8,10 Evaluation of the residualactivity after storage at 15°C of a culture extract

Page 5: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

J!}."CQ)

.-Q) c:

"-c:=

O .-

-gc:"ri

-c:Q)c:

C:O~O

(1jUO.C

tcn.CQ)

OQ) "C

C.::1(1j:O:

E"C -

c: ::I

cn .-cn

~ Q) J:: .-

.:?;:0 a: .~ "C

Q) ~ .

> .>

r.-"C -Q)"C gc:Q) ::I=C: c. "- .-

~Q)(1j

"C-

~c:§~

(1j ::I U .-

~~cn"5

J::(1jQ)~

-cn::1c.

"C

cn c: .-Q)

Q) .u UJ >

"- O Q) .-

::I .-"- .~

-"- .~

(1j Q) .-v.

Q)->-o

-U .~ c.

Q) ~ .~ ...::.

cn-u+

O 'u (1j ~

J::= .-

-cn"ffi.2:'

C> cn .s ~ .g 2

~ U "- .-

> U U

O C>.- Q)

J::c:EC.

cno".;=cn

Q)EC:=-(1j (1j Q)

::I "- U

u"co-

Q)Q)-Q)

o?;-e>

E 9}; .~ (1j

v. --

.-c:c:C>

, oQ)c:

« U cn .-

c.. cnc:

~ Q) .-

c: o E

.- 00 Q) "-U "-Q)

O.-(1j-.-Q)"ccn~"CQ) Q)

c.::1 ~ 0(1j:g -

cn -

-Q) c:

O "- (1j

Q)Q) t

"-J::~O

::IL- c.

Ur-:Q)E

::I .2:' .s .-

"- .--Q)

-> "-

U) ..;::; "c (1j

u=

.(1jO~...cn .

w"ffi (1j I~ .Q .~ oC> .s I

~-J:: .

u: .Q .~ I

.c~~

Page 6: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

to sulfoxide in the pediocin PA-1-ox moleculemay explain the increase in molecular weight andhydrophilicity with()ut charge alteration.24 Re-cently, site-directed mutagenesis studies wereperformed to obtain methionine- free pediocin P A-lvariants.33 Although mutating Met31 to Asp nul1i-fied bacteriocin activity , other methionine-freepediocin PA-l variants (Met31 to a hydrophobicamino acid, such as Ala, Leu, or I1e) displayedhigher storage stability without a concomitant largeloss of activity.

D. Purification

Because pediocin PA-l is a secreted peptide,most of the purification protocols start with theconcentration of ce1l-free supernatants obtainedfrom the producing strains prior to the applicationof chromatography-based techniques.

lnitia1ly, Gonza1ez and Kunka8 partia1ly puri-fied pediocin PA-l by ammonium sulfate precipi-tation and two successive ion-exchange chromato-graphic steps of the reconstituted dialyzedprecipitate. However, no data were provided onthe evolution of the specific activity through thepurification process. After a similar precipitationstep, Bhunia et a1.10 employed gel filtration andanion-exchange chromatographies, both integratedin an FPLC system. The activities of the finalfractions were up to 98 times higher than those ofthe respective supernatants.

The ability of this bacteriocin to adsorb tothe ce1l envelope of the producing ce1ls was thebasis of the purification method developed byYang et al.35 These authors adjusted the pH ofa fu1ly grown culture to pH 6, a value coinci-dent with the highest adsorption efficiency.Then, ce1ls were co1lected by centrifugationand the pH of the ce1l suspension was read-justed to 2 for bacteriocin release. The suspen-sion was applied to a reverse-phase columncoupled to an HPLC system and 106% of theinitial activity was recovered Daba et aU6 fol-lowed a similar approach and the elution pro-file revealed two peaks with antimicrobial ac-tivity. The large HPLC peak corresponded topediocin PA-l, while the minor peak containedoxidized forms of the bacteriocin.

In 1992, Henderson et a1}O purified pediocinPA-l to homogeneity by sucessive gel filtration

of the supernatant, ion-exchange chromatogra-

phy, dia1ysis and, fina1ly, reverse-phase chroma-tography coupled to HPLC. Although the yieldwas only 0.6%, the fina1 sample was pure enoughto a1low pediocin PA-l amino acid sequencing.

Simultaneously, Nieto-Lozano et a1.21 also puri-fied this bacteriocin to homogeneity and sequencedit. Their purification procedure, with a signifi-

cantly higher activity yield (600%), includedammonium sulfate precipitation, followed by three

chromatographic steps ( cation-exchange, hydro-phobic interaction, and reverse-phase chromatog-

raphy, respectively).36-37 This basic protocol, withslight modifications, has been widely and suc-cessfully applied for the purification of a largenumber of LAB bacteriocins. Usually, the final

reverse-phase chromatography step is repeatedtwo or three times to ensure maximum purity. In

addition, the use of large volumes of washingbuffer for the washing steps during ion-exchangeand hydrophobic interaction chromatographies

may contribute considerably to purity ,26 while agel filtration step prior to the cation-exchange

chromatography may increase the purificationyield.38-41 Recently ,42 another chromatography-based method consisting in three steps (ion-ex-change, solid phase extraction, and reverse phaseHPLC) has been applied successfully to obtain

highly pure class Ila bacteriocins, includingpediocin PA-l.

Pediocin PA-l has a1so been purified by a pro-

tocol, including ethanol precipitation, preparativeisoelectric focusing, and ultrafi1tration.29 About 32%of the pediocin contained in the culture supematants

was precipitated with cold ethanol. A1most no activ-ity (3%) was lost during isoelectric focusing and a11

activity remained after ultrafi1tration. The majoradvantage of this method was the omission of chro-

matography steps, avoiding losses due to bacterio-cin sticking to the matrix of the columns.

Idea11y, the methods used for bacteriocin purifi-cation shou1d have high fina1 yields, low costs, anda high reproducibility .The purification protocols

cited above involve severa1 steps that make the pro-cess inefficient, laborious, and/or expensive, espe-cia11y if sca1ing -up for industria1 purposes is desired.

Construction of immunoaffmity columns using

Page 7: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

pediocin P A-l-specific-antibodies may constitute ana1ternative approach for the simple (even single-step) purification of pediocin PA-l. Nisin mono-clona1 antibodies43 were used successfu11y for theone-step purification of this lantibiotic in 5 m1irnrnunoaffinity columns, a1though scaling-up hasnot been eva1uated yet.441n this context, the recentgeneration of polyclona1 antibodies that specifica1lyrecognize pediocin PA-l is a promising step.40.41

III. BIOLOGICAL PROPERTIES

A. Antimicrobial Spectrum

Pediocin PA-l displays antimicrobia1 activityagainst a wide spectrum of Gram-positive bacteria,many of them responsible for food spoilage orfoodbome diseases. The activity ofpediocin PA-lagainst Listeria monocytogenes, a microorganismwith the highest mortality rate among foodbomebacteria in Western Europe and North America, isparticu1arly relevant. Therefore, many studies onthe biologica1 activity of this peptide have beenfocused on its effects on L. monocytogenes cells.45Interestingly, pediocin PA-l is not active againstbacteria used frequently in starter cultures, such aslactococci. Pediocin PA-l has simi1ar MIC va1uesat 20, 30, and 37°C, and its C-termina1 disu1fidebridge is a major detenninant of the antimicrobia1spectrum.46 The structure and composition of theouter membrane of Gram-negative bacteria do notal1ow pediocin access to its target, the cytoplasmicmembrane. However, many Gram-negative organ-isms (such as Salmonella typhimurium, Escheri-chia coli, Serratia liquefaciens, and Pseudomonasfluorescens strains) are pediocin PA-l-sensitiveafter inflicting subletha1 injuries (like freezing,gentle heating, exposure to lactic acid or EDT A, orhydrostatic-pressure pasteurization) to the outermembranes, rendering their cytoplasmic mem-branes accessible to pediocin molecu1es.47-51

activity. Several arbitrary units have been definedto quantify the antimicrobial activity of pediocinPA-l on solid media. Bhunia et a}.lo employed"antimicrobia1 activity units" (AAU) ca1culatedas the reciproca1 of the antimicrobia1 titre multi-plied by the di1ution factor and divided by theprotein concentration (in mg) determined by theLowry method. The titre was defined as the recip-roca1 of the highest di1ution producing a c1earzone of inhibition of 2 mm or 1arger on anLb. plantarum WSO-39 1awn.52 Later, Gonza1ezand Kunka8 defined one "arbitrary unit" (AU) ofbacteriocin as 5 ~1 of the highest di1ution of cul-ture supernatant yie1ding a defmite zone of growthinhibition on a 1awn of the indicator P. pentosaceusFBB63. This unit was a1so adopted by other au-thors.20.53-55

Turbidometric methods, such as the microtitrep1ate assay ,36.56 are a1so used to detect and quan-tify pediocin activity .19.21.38-40 In this system, onebacteriocin unit (BU) is defined as the reciproca1of the sample di1ution, which inhibits growth ofthe indicator organism by 50%, as indicated byspectrophotometric measures. This method is veryusefu1 to eva1uate the activity of the differentfractions obtained during pediocin PA-l purifica-tion.

B. Quantification of the Antimicrobial

Activity

In contrast to nisin, there is no internationa1unit refering to the activity of an internationalreference sarnple of pediocin PA-l. Itis impos-sible to compare arbitrary antimicrobia1 activityva1ues in cultures, culture supernatants, or (par-tia1ly) purified pediocin PA-l obtained by differ-ent authors, even if they use the sarne assay,because the MICs va1ues for vegetative cells andspores may significantly differ depending on theproducing or indicator strains, the sarnple prepa-ration method, and the bacteriocin assay condi-tions.57 In addition, the major drawback of theassays cited above is their unspecificity. In thiscontext, the recent development of pediocin P A-lspecific polyclonal antibodies enables the spe-cific quantification of the pediocin present in asarnple by using a wide array of immunoassayformats.40,41

Detecting a specific bacteriocin by searchingfor a compound with a determined molecular massmay be a practical approach to confirm the pres-ence of pediocin PA-l in cultures or foodstuffs.

Currently, the agar diffusion test is one of themost widely used methods to detect bacteriocin

Page 8: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS)has been used recently for rapid detection ofpediocin PA-l, nisin, brochocins A and B andenterocins A and B, from culture supematants.58Compared with other spectrometric techniques,MALDI- TOF MS offers some advantages for theroutine ana1ysis of bacteriocins, including ease ofuse, picomolar to femtomolar sensitivity, highmass range, ability to give extremely accuratemass values, and relative tolerance to sample con-taminants.58

c. Resistance ta Pediacin PA-1

The emergence of organisms highly tolerant ornatura11y resistant to class II bacteriocins has be-come quite common and is a potential obstacle totheir application as food biopreservatives.59 Al-though the ecology and mechanisms of resistanceto many traditiona1 antibiotics have been studiedprofusely, little is known on these aspects in thebacteriocin field.60,61 However, investigations onthis subject will rapidly increase in the near futurebecause of the industria1 interest in the applicationof bacteriocins and/or bacteriocinogenic LAB toimprove the rnicrobiologica1 qua1ity of foods.

Sensitivity to pediocin P A-l differs greatlybetween species and strains evaluated,18,59,62atlhough pediocin P A-1-resistant or -tolerantstrains seem to occur more frequently than strainsdisplaying resistance or tolerance to nisin.59Growth of surviving cells of L. monocytogenesstrains in pediocin PA-1-treated foods during re-frigeration storage (4 and 10°C) has been re-ported,53,62 and the population levels of some strainsreached very high numbers in a short time,62 evenin the presence of residual pediocin. As reductionin Listeria counts was greater with higher levelsof pediocin, it has been suggested that the levelofpediocin (or other bacteriocins) should be ad-justed to obtain maximum viability loss of themost resistant strain when used as a food preser-vative.62 Probably, the routine use of pediocinPA-l at concentrations not high enough to eradi-cate a1l Listeria cells may result in emergence ofpopulations resistant not only to pediocin but a1soto other closely related bacteriocins.63 The effect

of pediocin was higher when the initia1 Listeriapopulation was lower,62 a fact confmned by alater study that showed complete inhibition ofListeria cells on refrigerated chicken meat bypediocin treatment.64 These results emphasize theimportance of effective measures to prevent orminimize food contamination.

Noerlis and Ray65 described how a non-pediocin producer P. acidilactici strain ( obtainedafter loss of the plasmid harbouring the pediocinPA-l operon) became resistant to this bacteriocinwhile growing in its presence. However, this studya1so indicated that the acquired resistance was atransient trait because it was soon lost after subcul-turing (five subcultures) in a pediocin PA-1-freemedium. Similarly, Dykes and Hastings61 studiedthe fitness costs associated with class lIa bacterio-cin resistance in one L. monocytogenes strain andfound that the resistant strain had a lower growthrate than the sensitive one in monocultures, and,furthermore, that resistant cells were unable to in-vade populations of the sensitive variants whengrown in mixed cu1tures. The authors concludedthat this specific class lIa bacteriocin resistancewas un1ikely to become stable in natura1 popula-tions of that particular L. monocytogenes strain.However, this study may not reflect the genera1situation because on1y one resistant strain was tested.In contrast, the resistance to other LAB bacterio-cins has been shown to be stable during at least 10generations in the absence of contact with bacterio-cinS.66

Involvement of cell wa1l constitution and mem-brane lipid composition in acquired bacteriocinresistance has been suggested frequently .67-70 Thefact that stability of the resistance phenotype canvary and mayor may not be lost after subculturingin the absence of the bacteriocin indicates that amore genera1 adaptive response may a1so be in-volved in trained bacteriocin resistance.18 Benniket a1.18 studied the relationships between mem-brane lipid properties and the effect of pediocinPA-l and nisin on glycolysis, intracellular ATPpools, and membrane potentia1 of whole LAB cells.Their results indicate that not on1y membrane lipidcomposition, but a1so overa1l membrane constitu-tion may play an important role in determiningLAB susceptibility toward these bacteriocins. Theseauthors suggested that a1though LAB strains with~

Page 9: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

a natura1 high resistance toward a bacteriocin can-not avoid the membrane insertion of bacteriocin

monomers, the overa1l composition of the mem-branes may preclude the formation of pores withsufficient diameters and 1ifetimes to cause cell death

(see mode of action).Combination of pediocin P A-l with other bac-

teriocins may constitute one approach to avoid the

growth of resistant cells. A mixture of nisin andpediocin PA-l was bactericida1 to more cells in aninitia11y sensitive population, probably because cellsresistant to one bacteriocin were killed by the other }1The combination killed more cells of two LAB and

two L. monocytogenes strains than each bacterio-cin individua1ly. In addition, the mixture inhibitedgrowth of a11 clostridia1 strains tested for up to 28days, while pediocin or nisin a1one produced vari-able resu1ts. Recently, the additive effect of apply-ing pediocin PA-l in combination with bacterio-cins belonging to unrelated classes has beenreported,72 demonstrating that, in addition to nisinA, it can acts synergistica11y with lacticin 481,lacticin B, or lacticin F. In contrast, when pediocinPA-l was combined with other class lIa bacterio-cins, no synergistic effects were observed.Z6

The fact that pediocin PA-l is, structurally andfunctiona11y, completely unrelated to nisin, lacticin

481, lacticin B, or lacticin F may have practica1 con-sequences. Different authors59,66 have reported thata1though L. monocytogenes mutants with spontane-ously acquired resistance to one class n bacteriocina1so displayed cross-resistance to other members ofthis bacteriocin class, their sensivity to nisin had not

been affected. Sakacin P (a class na bacteriocin)sensitivity among 22 strains of L. monocytogenesdirectly co1Telated with the pediocin PA-l sensitivityof the strains. In another study ,63 Listeria cells surviv-ing nisin action were pediocin PA-l resistant, whilepediocin PA-l survivors remained susceptible to the

lantibiotic. The beneficia1 synergistic effects resu1tingfrom the combination of unrelated LAB bacteriocinscan be exploited to extend their potentia1 app1icationin the food industry .

pediocin binding to cytoplasmic membranes, in-sertion ofbacteriocin molecu1es in the membranes,and formation of the poration complex. This pro-cess finally leads to cell death that may occur withor without celllysis,probably depending on con-comitant activation of the cell autolysins. For areview on the mechanism of action of pediocinPA-l see Reference 73.

lnitia1ly, Gonza1ez and Kunka8 showed thatpediocin-sensitive or pediocin-insensitive Gram-positive cells adsorbed comparable levels of thebacteriocin and concluded that the specificity ofpediocin PA-l was at least not enterely dependenton the presence of specific receptor sites in sensitivecells, and that pediocin might have a high surfacebinding capacity responsible for the un1etha1 bind-ing to insensitive bacteria. Bhunia et a1.54 reportedthat the bacteriocin did not bind to the cell surface ofGram-negative bacteria. Although the presence oflipoteichoic acid in the cell wa1l of Gram-positivebacteria and its absence in Gram-negative bacteriacou1d explain this differentia1 adsorption ability , itdid not provide any explanation for the existance ofsensitive and insensitive Gram-positive bacteria.These authors54 a1so described how pediocin P A -1-treated sensitive cells lost intracellular K + ions and

ultraviolet absorbing materia1 became more perme-able to ONPG and in some cases lysed.

The proton motive force (PMF) constitutes akey parameter in the understanding of the modeof action of LAB bacteriocins because it is re-quired for many of the cells energy-dependentmetabolic processes. The PMF is the result of theelectrochemica1 gradient of protons across thebacteria1 cytoplasmic membrane and is composedof the membrane potential (/1",) and the pH gra-dient (/1pH)}4 The development of artificia1 mem-brane systems and techniques that allow the sepa-rate study of both components of the PMF haveenabled the investigation of the effects of differ-ent bacteriocins, including pediocin P A-l , on PMF .Christensen and Hutkinsl3 provided the first evi-dence for the involvement of PMF in the mecha-nismof action of pediocin PA-l. Pediocin causeda concentration-dependent dissipation of the PMFof L. monocytogenes Scott A cells. In addition,the peptide increased the membrane permeabilityof Listeria cells to protons, which agrees with adissipation of the pH gradient.

D. Mode of Action

The bactericidal mode of action of pediocinPA-l on sensitive ceIIs involves three basic steps:

Page 10: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

molecules would insert into the membrane, ag-gregate and form oligomeric structures, leadingto hydrophilic pores through which ions and smallmolecules would be released from sensitive cellsand, ultimately, to cell death. However, in thesame study ,25 it was found that pediocin binds andinserts in E. coli lipid vesicles with a high contentof zwitterionic phospholipids but without a pro-iein receptor .

A study on the in vivo effect ofpediocin PA-1 on L. monocytogenes confmned that, concomi-tant with cell death, the bacteriocin induced deple-tion of cytoplasmic A TP and irreversible K + and

phosphate efflux.84 Pediocin PA-l depleted 90%of cytoplasmic A TP when on1y 25% phosphateefflux had occurred, which indicated that loss ofA TP was due to attempts of the cell to maintainPMF rather than its inability to produce A TPbecause of phosphate loss.1n contrast, nisin causessimultaneous total A TP depletion and phosphateefflux.85-86

Chen et a1.28 characterized the physicochemi-cal interactions of pediocin PA-l with targetmembranes using two-lipid vesicle models, onebased on tota1 L. monocytogenes lipids and thesecond on synthetic phospholipids. The antimi-crobial peptide caused a time- and concentration-dependent release of carboxyfluorescein (CF) fromthe vesicles. The CF efflux rates were higher inacidic that in neutra1 or a1kaline conditions anddepended on both pediocin and lipid concentra-tions. Although pediocin PA-l was able topermeabilize the membrane of Listeria lipidvesicles in the absence of d'11, the presence of atransmembrane potential (inside negative) in-creased the CF efflux rate by 88%. In vivo andin vitro energy-enhanced action has been previ-ously reported for bavaricin MN, another class lIabacteriocin, using the same lipid vesicle model.87Generation of d'11 could have stimulated the for-mation of more and bigger pores on the targetmembranes.28 The fact that both lipid vesicle sys-tems were devoid of membrane protein receptors,together with the pediocin concentration-depen-dent CF leakage, strongly suggested that no pro-tein receptor was required for the recognition andbinding of pediocin PA-l to the membrane ofsensitive cells. Instead, the bacteriocin moleculescould have recognized specific membrane lipids

These results, together with those obtainedwith nisin,75-79 lactococcin A80, and lactococcinB,81led Bruno and Montvil1e82 to hypothesize thatLAB bacteriocins employ a common mechanismof action based on the dissipation of the PMF ofsensitive cel1s. To obtain further evidence, theseauthors82 also investigated the influence ofpediocin PA-l, nisin, lactacin F, and leuconocinS on the PMF of L. monocytogenes Scott A cel1s.During the assay period, the Listeria populationdecreased by 4 orders of magnitude after the ad-dition of pure pediocin (20 ~g mI-l) to the cul-tures. Additional1y, the residua1 d", (-30 m V) wasinsufficient for culture viability. Energized cel1slost their PMF in a concentration-dependent man-ner, while the dpH component was dissipated bya pediocin concentration lower than that requiredto eliminate d",. PMF col1apse led to growthinhibition and death because of the low intracel-lular A TP levels and the inability to carry outactive transport of nutrients and to maintain suf -ficient concentrations of cofactors, such as K+ andMg2+ .Pediocin p A-l , lactacin F, and leuconocin Sacted in an energy-independent manner, in con-trast with the energy-dependent mode of action ofnisin, a bacteriocin that seems to require a thresh-old membrane potentia1 for activity .82-83

The independent studies of Bruno andMontvil1e82 and Chikindas et a1}5 indicated thatpediocin p A-l fonned voltage-independent poreson the cytoplasmic membranes of target cel1s, aprocess that dissipated the transmembrane elec-trical potentia1, inhibited amino acid transport andcaused an efflux of smal1 ions. Their data alsoindicated that the pediocin concentration deter-mined the size exclusion limit of the pores sincemore pediocin was required to release compoundshaving higher molecular weights (up to 9400).Although pediocin (in vitro and at high concen-tration) worked in the absence ofPMF, the possi-bility that in vivo PMF could increase the effi-ciency of low pediocin concentrations on the targetmembranes was not excluded. On the basis ofthese and previous results, the authors proposed amodel for the mechanism of action of pediocinPA-l. Initially, pediocin molecules wouldnonspecifica1ly adhere to the cel1 surfaces, fol-lowed by specific binding to receptor componentsof the cytoplasmic membrane. Then, pediocin

Page 11: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

E. Structure-Function Relationshipsand/or established electrostatic interactions withthe phospholipid head groups.

As anionic phospholipids are the principa1components of L. monocytogenes membranes,88Chen et al.28 subsequently characterized the bind-ing of pediocin and pediocin fragments to vesiclesof phosphatidylglycerol by using tryptophan fluo-rescence as a probe. It is possible to differentiatetryptophan residues bound to membranes fromthose in peptide molecules in an aqueous solutionbecause spectral parameters of its fluorescenceemission change with environmenta1 polarity. Theresults revea1ed that at least one tryptophan resi-due of pediocin PA-l was inserted into the an-ionic membranes. This work also provided thefirst evidence fot the key role that electrostaticinteractions between putative positively chargedpediocin patches and the negatively charged phos-pholipid heads of the membranes play in the ini-tia1 binding of the bacteriocin to target mem-branes.

A further study89 revealed that affinity ofpediocin PA-l for phospholipid vesicles increasesas the content of negatively charged phospholip-ids is higher. In addition, the relative dissociationconstant for the peptide-lipid interaction washigher when the anionic lipid content of thevesicles decreased. This fact reflected the anioniclipid dependency of the initia1 binding to themembrane and supported the electrostatic inter-action modelofbinding. Similar results have beenobtained for the lantibiotics epilancin K783 andnisin.83,90 In addition to the initia1 unspecific elec-trostatic interactions, there are important specificinteractions between the C-terminal half ofpediocin P A-l and the target cells,24 particularlywithin the fragment spanning residues 20 to 3430(see structure-funcion relationships).

Bennik et a1.18 reported that the exposure ofsensitive and insensitive LAB strains to pediocinPA-l resulted in a similar concentration-depen-dant dissipation of ~'Il, but the perturbation ofthecytoplasmic membrane caused rapid A TP deple-tion only in the sensitive cells. These authorssuggested that not only membrane fluidity (re-lated to lipid composition), but overa1l membranecomposition would influence the pore-formingactivity of bacteriocins, as membrane proteinsmay affect lipid ordering.

Despite the fact that class na bacteriocinshave similar primary structures, their antimicro-bial spectra show differences greater than whatwould be expected from the interaction betweenthe cationic peptides and the membrane lipids.5Therefore, establishment of structure-functionrelationships constitutes one of the major chal-lenges in bacteriocin research. To facilitate theunderstanding of this section, Figure 1 shows aschematic representation of the pediocin PA-lstructure-function relationships that are (puta-tively) known at present.

Severa1 groups have described the loss ofpediocin PA-l antimicrobia1 activity as a conse-quence of DTT reduction of its disulfide bonds}5,73DTT -induced reduction had only a moderate effecton curvacin A and sakacin P activities}6 In thesebacteriocins only the equiva1ent of the flfSt pediocinPA-l disulfide bond (C9-CI4) is present. Thus, thesecond disulfide bond (C24-C44) in pediocin PA-lseems to be essentia1 for the biologica1 activity ofthe peptide. This has been shown for the equiva1entC25-C43 disulfide bond in the class na bacteriocindivercin V41.91 The fact that pediocin PA-l andenterocin A (both with two disulfide bonds) aremore active than curvacin A and sakacin P (withonly one bond) may be due, at least partly, to theextra disulfide bond.26 These studies suggest thatthe conserved bond is important but not crucia1 forantimicrobia1 activity. Similar fmdings have beenreported with carnobacteriocin B292 and leucocinA,93 but not with mesentericin YI05.941n fact, theactivity of a mesentericin variant, which had the C9and Cl4 residues substituted by serine residues, wasnotably lower than that of the native peptide. Someauthors claim that both disulfide bridges are essen-tia1 for pediocin PA-l activity because of theirsupposed role in the maintenance of a conforma-tion compatible with activity .83.94

The primary structure of class na bacterio-cins may be divided into two functional regions,the well-conserved and hydrophilic N-termina1ha1f and the more heterogeneous and relativelyhydrophobic C-terminal ha1f.24 Fimland et a1.24synthesized four hybrid bacteriocins in which themodules corresponding to pediocin PA-l (Ped),sakacin P (Sak), and curvacin A (Cur) had been

101

Page 12: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

dues, was only weak. In addition, no binding wasobserved at the same pH va1ue using the mutantNm8-l5, in which K11 and H12 had been replacedby uncharged residues (111 and L12' respectively).Binding of the fragments containing this putativepositive patch decreased by adjusting the pH to 8,probably because of the deprotonation of the his-tidine residue.

The finding that an in-frame fusion betweenthe functiona1 mature domain of pediocin PA-land the C-terminus of the secretory protein ma1-tose-binding protein (MBP) retained antimicro-bial activity96 suggests that the bacteriocinN-termina1 region does not span the phospholipidbilayers of the target cells. The same study re-vea1ed that the class Ila consensus sequence isessentia1 for pediocin PA-l activity because itsdeletion resulted in loss of the activity.

MilIer et al.97 generated a col1ection of 18pediocin mutants, 17 of which were obtained byrandom PCR mutagenesis of the pediocin PA-lstructural gene. They were produced and secretedas MBP-fusion proteins. Eight mutants (N5K,C9R, CI4S, CI4Y, C24S, G37E, G37R, andC44W) were completely inactive, while nine(KIN, WI8R, 126T, M31T, A34D, N41K, H24L,K43N, and K43E) retained some activity, rangingfrom <1 to 60%. However, the activity of theremaining mutant (K11E) was around 2.8-foldhigher than that of the MBP-pediocin PA-l chi-meric protein. The results showed that the fourcysteine residues were required for activity. Thedisulfide bond may be essentia1 to maintain theconformation ofthe sequence G1o-K11-H12-S13 thatconstitutes the apex of the 13-hairpin previouslyproposed by Chen et a1.27 and that exhibited weakhomology to several known types of consensus13 tums. Among the four lysine and three histidineresidues of pediocin PA-l, three residues (K1,H42' and K43) seemed to be particularly importantfor activity .96 The substitution KIN resulted in analmost complete inactivation of bacteriocin activ-ity, while the substitutions H42L, K43N, and K43reduced its activity by 40 to 80%. Therefore, K1may play an essentia1 role in membrane bindingwhile that of H42 and K43 would be secondary .These results97 show that protein engineering ofclass Ila bacteriocins could be a feasible way toimprove their activity, stability, and/or solubility.

exchanged. The four hybrids (Ped-Sak, Sak-Ped,Cur-Sak, and Sak-Cur) were biologically active.The hybrid bacteriocin had an inhibition spec-trum that was similar to that of the bacteriocinfrom which its C-terminal module was derivedand different from the bacteriocin that providedthe N-terminal part. Therefore, the C-terminalmodule of class Ila bacteriocins plays an impor-tant role in determining the specificity of targetcel1s. A later report30 indicated that a l5-mer pep-tide fragment derived from pediocin PA-l (resi-dues 20 to 34) interfered specifically with pediocinPA-l target cel1 interaction, leading to inhibitionof the pediocin antimicrobial activity. The frag-ment also significantly affected enterocin A ac-tivity but not that of other closely related bacte-riocins (sakacin P, curvacin A, and leucocin A).Among these bacteriocins, pediocin PA-l andenterocin A have the highest degree of homologyin the region spanned by the fragment and, addi-tional1y, are the only ones with a cysteine residuein this region. Apparently, inhibition of pediocinPA-l activity by the fragment was due to specificinterference of the fragment with pediocin-targetcel1 interactions and not merely to hydrophobicinteractions between these two elements. Recently,site-directed mutagenesis studies of pediocin P A-land sakacin P have been performed to elucidatethe structural basis for the differences observed intheir respective activities and target cellspecifities.46 By making the primary structure ofsakacin P more pediocin PA-1-like, it has beenshown that the extra disulfide bond C24-C44 ofpediocin PA-l contributes to both temperaturestability and to widening the target cel1 spectrum.The results also support the previous suggestion24.3othat the variabIe C-terminal part of pediocin-likebacteriocins interacts with the hydrophobic partof the membrane, being an important determinantof target cel1 specificity.

Chen et al.28 observed that K11 and H12 resi-dues were essential in the bacteriocin binding tophospholipid vesicles. At pH 6, two pediocin frag-ments, N15 (residues 1 to 15) and N8-15 (resi-dues 8 to 15), bound strongly to lipid vesicles andthe binding behavior of N15 was comparable tothat ofpediocin PA-l. However, the binding abil-ity displayed by fragment N7 (residues 1 to 7),which do not contain the positively charged resi-~

Page 13: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

MilIer et a1.97 submitted the pediocin C-termi-nal ha1f to hydrophobicity analysis in order tolocate apolar sequences that could participate inmembrane binding.1n this region, the stretch com-prising residues 125 to G37 had enough hydropho-bicity to allow its location inside a phospholipidbilayer if simultaneous membrane insertion oftwo or more molecules would occur, somethingthat probably does occur during the formation ofthe poration complex.

It has been suggested that the C-terminalmodule of pediocin PA-l would becomeamphiphilic, and therefore would permit the for-mation of a poration complex if it adopted a trans-membrane a-helica1 secondary structure.24 How-ever, adoption of such a structure seemsincompatible with the disulfide bond C24-C44formed by residues located at the begining andend of such an a-helix.26 The effect of this disul-fide bridge in the structure of the C-terminal re-gion will have to be determined before drawingany conclusion on the role of amphipathicity, ifany, in the formation of the poration complex.97

with a UAG stop codon22 (Figure 2). The fourgenes are organized in a single operon, governedby a promoter located directly upstream of pedA.The sequences of its -35 (TTGACA) and -10(TAGAAT) regions, separated by 18 bp, are inclose agreement with the consensus sequences ofconstitutive promoters of Gram-positive bacte-ria.22.l0l Two transcripts are produced trom thepediocin PA-l operon. The most abundant mes-senger has a size of approximately 1.2 kb andcovers pedABC, while the second transcript has asize of 3.5 kb and corresponds to the pedABCDgenes.55 There are rho-independent transcriptionalterminators directly downstream of the pedC and

pedD genes55.l02 (Figure 2).The production of some class II bacterio-

cins has been reported to be a quorum sensingphenomenon.lo3 The process is transcription-ally regulated through a signal transductionsystem integrating an induction factor (peptidepheromone ), a histidine protein kinase and aresponse regulator. The induction factor is abacteriocin-like peptide with a double-glycineleader but without bacteriocin activity.4 Thereis no evidence for the existence of a similarsystem for the activation of transcription of thepediocin PA-l genes.

IV. BIOSYNTHESIS

A. Location of the Genetic Determinants

c. The pedABCD Genes and TheirProducts

The genetic deterrninants for the biosynthesisof pediocin PA-l are plasmid encoded in a11 pro-ducing strains isolated to date.8.11.15.17-19.23.34.98.991nseveral strains, the sizes, organization, and re-

striction profiles of the various pediocin-encod-ing plasmids are similar.17.100 It has been shownthat the plasmids responsible for production in

P. acidilactici H can be transferred intragenerica1ly

by conjugation.34

The pediocin PA-l structural gene (pedA) en-

codes a 62 amino acid pediocin precursor calledprepediocin PA-l. The 18 N-tenninal residues ofprepediocin constitute the leader sequence, whichis removed concomitantly with secretion, resultingin mature pediocin PA-l, a peptide of 44 aminoacids. The pediocin leader belongs to the group ofthe "double-glycine" leaders, which is found inmost nonlantibiotic and some lantibiotic LABbacteriocins and also in colicin V from

E. coli. These leaders share the consensus sequenceL-12S-11XXE-gL-7XX1-4XG-2G-1 with the two con-served glycine residues at positions -1 and -2

constituting a common processing site, a fact re-ftected in the name of the groUp.3,104,IO5 PrepediocinPA-l is also biologically active,55 approximately80% as active as the mature bacteriocin.l06

B. Organization of the Pediocin PA-1

Operon

Pediocin PA-l biosynthesis involves a DNAfragment of approximately 3.5 kb, comprising thefour genes pedA, pedB, pedC, and pedD.22 Eachgene is preceded by a ribosome binding site (RBS),starts with the AUG start codon (except pedB,which starts with a UUG codon), and finishes

103

Page 14: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

..00.

('\I('\III)

.0.0.

co(W)(W)

.c0.

(\It--.T-(\I

...:o-coc.Ë

$""ffic

.Q

E.

ucnccu~

t-

i.ïn0)c

'5c

:c

Q)Eocno

.c

.~

~

a>0Eo0..

Q.

coa>c.o.-

~c

.uo

'5Q)c.

Q)

-5

O

co

~"EQ)cnQ)

0..

~

u

~EQ)

.cu

(J)

N

w~~"'u:

Page 15: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

tion}2 The PedD sequence displays a high degreeofhomology with sequences ofmany others ABC

transporters.l0S

D. Secretion and Maturation of PediocinPA-1

Marugg et al.22 suggested that the 112 aminoacid protein encoded by pedB could be invo1vedin immunity of the producing cells, because mu-tational inactivation of this gene did not affectpediocin PA-l production in E. coli, a micro-organism naturally insensitive to this bacteriocin.Later, this hypothesis was confirmed as pediocinPA-l was not active against a previously sensi-tive P. pentosaceus strain that had been trans-formed with a plasmid carrying the pedB gene.55In fact, the levelof insensitivity of the resu1tingstrain was even higher than that of the parentalproducer, probably because of the higher copynumber of the plasmid encoding pedB in thetransformant and/or because pedB was clonedunder the control of P32, a strong lactococca1promoter .

Recently, severa1 immunity proteins belongingto pediocin-like bacteriocins were cloned behindP32.107 Introduction of the different pediocin-likeimmunity genes into a highly bacteriocin-sensitiveLb. sakei strain resulted in a selective immunityagainst the respective pediocin-like bacteriocinstested.1n genera1, low cross-immunity was observed;however, PedB and the immunity protein for sakacinP displayed complete cross-immunity despite thedegree of sequence simi1arity between these twoimmunity proteins is lower than that observed be-tween, for instance, PedB and the enterocin A im-

munity protein.l07The pedC gene encodes a 174 amino acid

protein (PedC), which is essential for pediocinsecretion in pediococci and E. coli,55 and belongsto the group of so-called "accesory proteins" re-quired for secretion processes involving A TP-binding cassettes (ABC) transporters. Accesoryproteins are also called "membrane fusion pro-teins" (MFP) because it has been postulated thatthey may connect the inner and outer membranesof Gram-negative bacteria to facilitate the pas-sage of substrates.l08 The PedC structure is simi-lar to those of HlyD, a protein required for secre-tion of the E. coli hemolysin A,I09 and LcnD, aprotein involved in lactococcin A transport.ll0

Finally, pedD encodes PedD, a 724 aminoacid ABC transporter dedicated to pediocin secre-tion through the membrane of the producing cells.Inactivation of this gene leads to total loss ofantimicrobial activity because of lack of secre-

The secretion and maturation processes ofpediocin PA-l depend on the membrane proteinsPedC and PedD, which together form a type Isecretion system.lll In this sec-independent sYs-tem, the substrate is direct1y secreted from thecytop1asm of the cel1 tothe extracellular environ-ment. As type I secretion systems are specific fora protein or family of closely related proteins,they are often ca1led "dedicated" transport sys-tems.112,113

A typica1 ABC transporter probably functionsas an homodimer, each part consisting of a hydro-phobic integra1 membrane domain and two A TP-binding domains, which can be fused in amultidomain protein.113 A11 transporters dedicatedto "double glycine" prebacteriocin secretion havean N-termina1 cytoplasmic extension that formsthe proteolytic domain responsible for remova1 ofthe leader peptide. In the N-termina1 part of thesebacteriocins transporters there are two conservedmotifs, the cysteine motif (QX4D/ECX2AX3MX4 Y /FGX4I/L) and the histidine motif (HY /FY NVX1J/LXDP).4,105 Based on the work of Havarstein eta1.,105 with lactococcin G, the following hypothesisfor the export ofpediocin PA-l can be made.4 Theproteolytic domain of PedD binds to prepediocinPA-l. Then, hydrolysis of A TP induces conforma-tiona1 changes in PedD leading to an integratedprocess in which remova1 of the pediocin leaderconcomitantly occurs with pediocin translocationacross the cytoplasmic membrane. In fact, expres-sion of on1y the N-termina1 172 amino acids ofPedD inE, coli cells producing prepediocin showedthat this domain was enough and essentia1 forprepediocin processing.55 Under these circum-stances, mature pediocin could on1y be found intra-cellularly, which suggests that the cleavage pro-cess can be uncoupled from secretion.

Apart from the extended N-terminal domainresponsible for the processing of the prebacteriocininto the mature form, the ABC transporters dedi-

Page 16: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

cated to double glycine leader bacteriocins prob-ably a1so differ from the earlier described "classi-ca1" ABC transporter proteins in their membranetopology. It has been reported that the lactococcinA-dedicated ABC transporter (LcnC) has fourtransmembrane segments (TMSs) instead of thesix typica1ly present in the classical ones.113

The accesory protein (PedC) has a membranetopology similar to LcnD, the lactococcin Aaccesory protein. Both are bitopic proteins with ashort N-termina1 cytoplasmic part, a transmem-brane segment and a large extracel1ular C-termi-nal part.113 Venerna et al.55 speculated that PedCmay be involved in channel formation betweenthe cytoplasmic membrane and the cel1 wal1 ofpediocin-producing cel1s. A similar role has beenproposed for LcnD.113 Recently, LcnD was shownto be present in two forms: a 52-kDa ful1-lengthform, total1y contained in the cytoplasmic mem-brane, and a sma1ler form (LcnD*, 38 kDa) that ispresent mainly in the cytoplasm with a smal1amount localized in the membrane.114 LcnC andLcnD are organized in a membrane-associatedcomplex in which LcnD, probably present as adimer, is able to interact with LcnD* and LcnC.114In addition, LcnC cross-links with a sma1l peptidethat could be the bacteriocin lactococcin A.114

E. Heterologous Expression of pedGenes

resulted in correct expression of the pediocin genessince the transforrnants produced zones of inhibi-tion on the indicator lawn}2 As bacteriocin activ-ity was also observed in cen-free supernatants,the authors concluded that pediocin was, mostprobably, secreted by the cens. E. coli mutantswith a mutated pedD gene did not secrete pediocin,which conflfffied the role of PedD in pediocin

transport.In the absence of pedC and pedD, mature

pediocin PA-l can be secreted in an active forrnwhen fused to the MBP C-terrninus and thereforecoupled to the E. coli sec machinery .96 The chi-meric protein was released into the culture me-dium because it was expressed in a leaky E. colihost in which the gene encoding the outer mem-brane murein lipoprotein had been disrupted.

Although there is a wealth of tools for ex-ploiting E. coli for heterologous gene expression,other bacteria can offer advantages either for cy-toplasmic production, secretion, or surface dis-play.116 Some LAB species, such as L. lactis areproving to be very versatile, and continuing basicresearch on gene expression in these organismswin increase their utility as alternative hosts forheterologous gene expression. In addition, manyLAB species or strains are food-grade organisms,making them potentiany useful for the heterolo-gous (in situ) production of commerciany impor-tant proteins or peptides.115

An attractive possibility for heterologous pro-duction ofpediocin PA-l is to exploit the signifi-cant amino acid homologies in both the leaderpeptides and dedicated transporters of most classII bacteriocins, some lantibiotics, and also colicinV produced by E. coli.55.104.105 Anison et al.117showed that both peptides of the two-componentlactacin F complex can us~ the bacteriocin secre-tion machinery of Carnobacterium piscicolaL V17 , a strain that produces carnobacteriocins A,BM1, and B2.118 The fact that the leaders ofthesecarnobacteriocins and both lactacin F peptidesshare the highest degree ofhomology among classII bacteriocins may have facilitated the secretionof the lactacin F peptides in this heterologoushost. van Belkum et al.119 recently demonstratedthe flexibility of the translocation apparatus byexchanging the leader peptides of different classII bacteriocins. Gene fusions were generated in

Heterologous expression systems are usua!lyemployed to elucidate the function of recombi-nant proteins or peptides or to achieve levels ofproduction higher than those of the native sources.They are a!so useful to produce a protein or apeptide by specific species, or strains, that arebetter adapted than the native producer to theenvironment where application of the proteina-ceous compund is desired.

Escherichia coli has long been the primaryprokaryotic host for cloning genes and expressingrecombinant proteins because it is weIl character-ized, many of its biologica! processes are under-stood and there are many genetic tools readilyavailable for its manipulation.115 Regardingpediocin PA-l, transformation of E. coli strainswith a plasmid containing the pediocin operon

Page 17: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

parental strain. It was possible to increase therelative pediocin PA-l production level to ap-proximately 50% in L. lactis LL108, a derivativeof L.lactis MG 1363, by increasing the copy num-ber of the plasmid-encoded ped operon. In thesame work,122 a recombinant P. acidilactici strainwas obtained that produced and secreted bothactive pediocin PA-l (an amount similar to that ofthe parental pediococca1 strain) and lactococcin A( twofold lower than that of the parentallactococca1

strain).Pediocin PA-l and lactoccoccin A are both

class II bacteriocins and, hence, likely candidatesfor expression and secretion in L. lactis via heter-ologous translocators.38 A hybrid gene specifyingan in-frame fusion ofthe lactococcin A leader andpediocin P A -138 was introduced in L. lactis IL 1403(LcnC'+, LcnD'+). TheL.lactis IL1403 derivativeproduced extracellular pediocin activity, but theproduction level was approximately 25% of thatof the parental pediococcus strain. Significantreductions in the yield of lactococcin A expressedin IL1403 have been described previously}6,8°These reductions may be attributed to the lowcopy number ofthe chromosoma1lcnC' D' genes123and/or to differences in the lactococcin A translo-cation apparatus,119,123 resulting in less efficientsecretion.

Enhanced heterologous production of pediocinPA-l was obtained by introducing the IcnC andIcnD genes into the lactococcin A leader-pediocinPA-l expression system previously developed.39When IcnC and IcnD were present on the sameplasmid (a pSH71 derivate, 50 to 60 copies percell124) as the hybrid gene, one of the resultingL.lactis IL1403 derivatives attainedpediocinPA-1 levels similar to that of the natura1 producerP. acidilactici 347. The yield of LcnCID-medi-ated translocation ofpediocin PA-l was stronglyinfluenced by the plasmid dosage and by the par-ticular lactococca1 strain employed as a produc-tion host. It has also been demonstrated that het-erologous production of other class II bacteriocinscan increase at least 10-fold when the dedicatedIcnC and IcnD genes are included in equiva1entlactococca1 expression systems.80,123 In the samestudy ,39 L. lactis strains with the ability to expressand secrete nisin A together with pediocin PA-lwere a1so constructed, which represents a first

which sequences encoding the leader peptides ofleucocin A, lactococcin A, and colicin V werefused to divergicin A, a bacteriocin that is se-creted through the sec pathway .120 The differentleader peptides were able to direct the secretion ofdivergicin in Leuconostoc gelidum, L. lactis, andE. coli, respectively (i.e., the homologous hosts).Furthermore, certain leader-translocator combi-nations led to heterologous production ofdivergicin. Colicin V was also secreted fromL. lactis when it was fused to the leucocin Aleader and the lactococcin A secretion machinerywas used. However, in some cases, the leaderpeptides were unable to direct divergicin secre-tion, suggesting that the various components ofthe translocation apparatus of class II bacteriocinsare not universally interchangeable. It may be thatthe hydrophobicity of the heterologous peptideplays an important role in the secretion process.119Although the system LcnC/LcnD can secrete agreat variety of bacteriocins that differ quite dras-tically (like lactococcins A, B, and M/N),111 itsinability to direct secretion of leucocin A in L.lactis has also been reported.121

Pediocin PA-l is a bacteriocin with a broadinhibitory spectrum and is particularly effectivein preventing the growth of L. monocytogenes, amajor concern in the dairy industry because it cangrow in a variety of dairy products at low tem-perature and pH. However, with the exception ofLb. plantarum WHE 92,17 all pediocin PA-l-pro-ducing bacteria so far isolated are pediococci,organisms usually associated with vegetable andmeat substrates but poorly adapted for growing inmilk and dairy products. Therefore, the produc-tion of pediocin PA-l in strains of dairy origin ishighly desirable.

The four ped genes were cloned into alactococcal vector and introduced into L. lactisIL1403, a pediocin PA-l-resistant, plasmid-freestrain that does not produce bacteriocin but har-bors chromosomal genes analogous to those en-coding the lactococcin A secretion apparatus (lcnCand lcnD).55.122.123 Secretion of pediocin PA-l,directed by its own leader sequence, was detectedonly when the ped operon was under the controlofthe strong lactococcal promoter P32. Even underthese conditions, the pediocin yield was less than1% of the production level by the Pediococcus

Page 18: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

has been reported that S. cerevisiae is often lim-ited as an expression system by low yields, whileother yeast species, such as P. pastoris orHansenula polymorpha, retain a1l the advantagesof S. cerevisiae and provide a reliable means ofachieving greatly elevated yields.129-131

The development of systems for heterologousproduction of pediocin P A-l and other bacteriocins,either using ABC-transporters or sec-dependent se-cretion systems, may contribute to overcome prob-lems that bacteriocin-producing starter cultures of -

ten face in food processes, such as poor adaptability ,low production, or genetic instability .In addition,these heterologous systems could be used for pro-duction of hybrid bacteriocins or pediocin-1ike bac-teriocins with improved properties obtained by pro-tein engineering. Overexpression of the ped genesby use of food-grade inducible systems, like thosebased on the PnisA promoterl32-135 or the sa1t-induc-ible Pgad promoter36,137 cou1d be another interest-ing approach. Recently, enhanced secretion ofpediocin PA-l was achieved in a L. lactis strainfollowing induced expression of pedC and pedDbehind the Pgad promoter.138

step in the construction of LAB strains that co-produce two or more well-characterized wide-spectrum bacteriocins. Recently, low-Ievelcoproduction of the nonsynergistic class lIa bac-teriocins pediocin PA-l and enterocin A inL. lactis ILI403125 and production of a variabIeamount of pediocin PA-l in E. coli, L. lactis,Streptococcus thermophilus, and Enterococcusfaecalis have also been achieved.126 The cells weretransformed with shuttle vectors containing thefour genes of the ped operon (without its ownpromoter) behind the S. thermophilus promoter

STP2201.Yeast, especially Saccharomyces cerevisiae

and Pichia pastoris, are major hosts employed inthe expression of heterologous proteins of highquality in the biopharmaceutical, industrial, andacademic environments because they combinewell-known techniques for the molecular manipu-lation of prokaryotes with the authenticity of eu-karyotic systems.127 Frequently, excessive amountsof sulfur dioxide and other chemical preserva-tives are used to prevent growth of undesirablebacteria during the making of wine, beer, andother beverages. This practice affects the qualityof the end-products and raises increasing con-suffier concerns. Schoeman et al.128 investigatedthe feasibility of controlling spoilage bacteriaduring yeast-based fermentations by expressingpediocin PA-l in S. cerevisiae. To achieve thisobjective, the pedA gene was inserted into a yeastexpression/secretion cassette in which expressionof pedA was under the control of the yeast alcoholdehydrogenase I gene (ADH1) promoter and ter-minator, and the secretion of the bacteriocin wasdirected by the secretion signal of the yeast mat -

ing pheromone a-factor (MFals). The constructwas introduced as a multicopy plasmid inS. cerevisiae. Northern blot analysis confmnedthat pedA was efficiently transcribed, while bio-logically active bacteriocin was detected in a bio-assay. However, the heterologous peptide waspresent in relatively low levels in the yeast culturesupernatant because most of the bacteriocin mol-ecules were attached to the yeast cells. Productionof extracellular pediocin PA-l by recombinantyeast strains may be optimized by enhancingpedAgene expression as weIl as by increasing thesecretability of the recombinant bacteriocin. It

F. Factors Affecting Pediocin PA-1

Biosynthesis

The amount of pediocin PA-l produced isvery strain-dependent, a1though the variabilityseems to be less than that observed, for instance,among different nisin- or leucocin Lcm1-pro-ducing strains.139 Under the same conditions,P. acidilactici H produced more pediocin thanstrains E, F, and M,I00 which was attributed tothe higher growth rate of strain H and/or to thehigher genetic stability of the plasmid on whichthe pediocin operon is located in that strain. Thelevelof pediocin PA-l produced byLb. plantarum WHE 92 was higher than that ofP. acidilactici H under optimum culture condi-tions, although production per cell was not re-ported.17 Recently, specific antibodies were usedto determine the concentrations of pediocinPA-l in the supematants of five independentlyisolated P. acidilactici strains.41 While four ofthe strains produced similar amounts (rangingfrom 1946 to 2200 ng ml-1), the bacteriocin con-

Page 19: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

v. TOXICITY

Pediocin-producing strains are found in avariety of natural foodstuffs, such as meat prod-ucts,8.10.19.34 dairy products,17 and vegetables.18Therefore, they (and most probably pediocin too)have been ingested by human consumers for cen-turies without apparent adverse effects, which is

a first indication of low toxicty of this bacteriocintoward eukaryotic cells. Additionally, treatmentof mouse myeloma cell cultures with pediocinPA-l did not affect cellular viability or morphol-ogy ,50 and no detrimental effects have been ob-served after subcutaneous, intradermic, intrave-

nous, or intraperitoneal administration of thisbacteriocin ( or fragments thereof) to mice and

rabbitsI4.40.41.144-146or after prolonged ingestion byhumans of meat products made with the pediocin-producing strain P. acidilactici H.147 Finally, the

intestinal microflora is probably unaffected bypediocin PA-l because of its inactivation by theproteolytic enzymes of the digestive tract.

VI. IMMUNOLOGICAL PROPERTIES

In general, generation of antibodies against

bacteriocins may provide sensitive and specific

tools for the detection of producing strains and for

the quantification of the peptides in culture media

and food extracts by immunochemical as-

says.40.41.43.145,146.148-150 Moreover, antibodies may be

used in single-step purification methods based on

imrnunoaffmity strategies.44 Although highly spe-cific imrnunochernistry-based methods have been

developed and routinely used as analytical tools in

many research areas, their impact in the bacterio-

cin research field has been marginal.149 Thus, iden-

tification and detection of bacteriocinogenic LAB

and bacteriocins has relied on the use of sensitive

but unspecific bioassay-based tests.

The two main difficulties encountered in rais-

ing antibodies against bacteriocins are the lack of

comrnercially available purified bacteriocins and

the need to couple them to proper carrier proteins

because of their low molecular masses ( which make

them poorly imrnunogenic or nonimrnunogenic ).

In addition, some characteristics of class II LAB

bacteriocins such as hydrophobicity or the pres-

centration in the supematalits of the remainingstrain was significantly lower (12 ng ml-l).

Pediocin PA-l production is higher when theproducing pediococci are grown at 30 to 37°C50 inmedia that allow the development of a high cellmass and a high acidity, such as MRS orTGE.50.139.140 The addition to TGE broth of supple-ments (tryptone, glucose, yeast extract) or comple-ments (niacinarnide, panthotenic acid, biotin) thatstimulate cellular proliferation notably increasedpediocin production, while production decreased

by incorporating buffering compounds.139A clear reduction in the quantity of pediocin

produced either by P. acidilactici strains orLb. plantarum WHE 92 was observed when thefmal pH of the medium exceeded 4.0.17.141 Whilebacteriocin production by P. acidilactici H wasneligible when the pH of the medium was 5 andceased at pH 5.5, even at high cell densities,17.140the activity of culture extracts from Lb. plantarumWHE 92 remainéd constant independently of thefmal pH value of the cu1ture broth between pH 4.0and 6.0 at pH and decreased only at pH 6.5.17 A laterstudy confinned that the amount of pediocin PA-lproduced by Lb. plantarum WHE 92 at the end oftheexponentia1 growth phase is independent of the pH ofthe culture medium between pH 4.0 and 6.0.141

It has been suggested that processing ofprepediocin to active pediocin by pediococci takesplace efficiently only if the pH of the medium isless than or equal to 5.0.139.142 Although this wouldexplain why pediocin production by P. acidilacticidecreases when the final pH of the culture ishigher than 5, it is not clear why that ofLb. plantarum WHE 92 is not reduced up to pH 6.0}7

Culture pH may have played an importantrole in the yield of pediocin PA-l by the heterolo-gous lactococca1 host L. lactis FI9043, becauselarger inhibition zones were generated from su-pematants of cultures grown in MRS broth (fmalpH 4.6) than from those in GM17 broth (final pH

5.3).38The highest pediocin yields can be obtained

by detennining the optimal conditions for pediocinproduction during pH- and temperature-controlledbatch cultures in a fermentor.141.143 Under suchconditions, batch cultures of P. acidilactici UL5-T5 at 37°C and pH 5.5 yielded approximately 190mg 1-1 of pediocin.

109

Page 20: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

pediocin P A-l could, thus, potentia11y facilitateobtaining antibodies of predetermined specificityfor the recognition of the native peptide mol-ecule.145

Recently, such polyclonal antibodies againstpediocin PA-l have been generated by immuni-zation of rabbits with a chemically synthesizedC-terminal fragment (PH2, residues 34 to 44)conjugated to the carrier protein keyhole limpethemocyanin (KLH).40 The limits of detection ofpediocin PA-l in MRS medium using these anti-bodies were approximately 2.5, 1, 0.025 and<0.025 ~g ml-1 by immunodotting, noncompeti-tive indirect ELISA (NCI-ELISA), competitiveindirect ELISA (CI-ELISA) and competitive di-rect ELISA (CD-ELISA), respectively. All im-munoassays and the slot-dot assay detected thebacteriocin in the supernatant of the producerP. acidilactici 347. The antibodies did not cross-react with other bacteriocins, including othermembersof class Ila LAB bacteriocins, and werevery useful in demonstrating heterologous pro-duction of pediocin PA-l in L. lactis.38.125

Pediocin P A-l-specific rabbit polyclonal an-tibodies have also been generated with a synthetic1 to 9-N-termina1 fragment (PH1) conjugated tothe same carrier protein.41 In this case, the limitsof detection of pediocin PA-l in MRS mediumwere approximately 5, 0.5, 0.01, and <0.01 ~gml-1 by protein slot-blotting, NCI-ELISA, West-ern blotting, and CI-ELISA, respectively. Theamount of free pediocin required for 50% bindinginhibition was 0.1 ~g ml-1 as determined by CI-ELISA.1n addition, results obtained with the sameassay clearly indicated that when pediocin, in-stead of ova1bumin-conjugated pediocin, was usedas solid-phase antigen, the assay was notablybetter. In contrast, pediocin did not efficientlycompete with the conjugate pediocin-horseradishperoxidase for binding to antibody-coatedmicrotiter plate wells in CD-ELISA, confirmingthe results of previous work145 and highlightingthe importance of the selection of proper immu-noassays for bacteriocin detection. The PHI-KLHpolyclona1 antibodies detected the peptide in thesupernatants of five pediocin-producingP. acidilactici strainS.41 However, it did not or on1yslighty reacted with supernatants of LAB strainsproducing enterocin A, enterocin P, sakacin P, or

ence of intrachain disulfide bonds may interferewith the development of sensitive immunoassays.

Bhunia et a1,144 eva1uated the immunogenicproperties of partially purified pediocin by immu-nizing mice and a rabbit and found that it wasunable to illicit an antibody response, even afterconjugation of the peptide to bovine serum a1bu-min. Polyclona1 and monoclona1 antibodies againstpediocin were developed by immunizing mice for12 weeks with pediocin conjugated to a polyacry-lamide matrix.148 Among 230 hybridoma clonesgenerated after two fusions, two were found toproduce antibodies that in Western immunoblotsrecognized this bacteriocin ( detection limit: 10 ~gml-l) but did not react with sakacin A, leuconocinLCM1, nisin and pediocin A. Immunization ofmice with crude precipitates of pediocin PA-l orwith pediocin bound to heat-killed Lactobacillusplantarum cells failed to produce antibodiesagainst the bacteriocin. In colony immunoblotassays, the monoclonal antibodies were usedto differentiate Ped+ and Ped- variants ofP. acidilactici RS2. Although monoclonal anti-bodies could be very useful in developing quan-titative assays for pediocin production, to date nonew developments have been reported from thisline of research.

Development of successful polyclonal andmonoclonal antibodies and immunoassay formatshave enabled the sensitive and specific detectionand quantification of nisin.43,44,149.150 Commercialpreparations ofpure pediocin PA-l, unlike nisin,are not available, and current pediocin PA-l pu-rification protocols do not allow purification ofthe large quantities of peptide needed to carry outcost-efficient immunization protocols. The use ofwhole bacteriocins, either a1one or conjugated tocarrier proteins may, if successful, lead to thegeneration of unspecific polyclona1 antibodies dueto cross-reactivity with consensus amino acidsequences found in other bacteriocins, and espe-cia1ly among class lIa LAB bacteriocins.4,26,151Short peptide fragments of proteins can be used togenerate antibodies that frequently recognize thefolded protein from which the peptide sequencewas derived,152 a fact that explains why antipeptideantibodies have become important tools in manyresearch fields.153,154 Chemica1ly synthesized frag-ments selected from the amino acid sequence of

Page 21: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

LAB bacteriocins to this approach is attractivebecause the range of minimally processed foodsavailable in the marketplace is rapidly increas-

ing}

A. Meat and Meat Products

sakacin A, bacteriocins that contain the longer(KYYGNGVxC) or shorter (YGNGVxC) con-sensus amino acid sequence present in pediocinPA-l. This result is not surprising because it hasbeen shown that, within a peptide fragment,changes in a single amino acid residue can dras-tica1ly affect protein recognition.155

The conjugate PH2-KLH has a1so been usedto immunize mice in order to obtain a source ofpediocin PA-l monoclona1 antibodies. However,among the hybridomas generated after three fu-sions, none produced antibodies.146

VII. FOOD APPLICATIONS

Although frequently in vitro studies haveshown the high antimicrobial activity of pediocinPA-l against sensitive bacteria, particularlyL. monocytogenes strains, most food applicationsof this bacteriocin have resulted in relativelymodest reductions of 1 to 3 log cycles in thepopulations of potential foodbome pathogens.45Therefore, deductions obtained from in vitro cha1-lenges should not be extrapolated to predictpediocin efficacy in a food matrix.156 Factors thatmay lead to a reduction of pediocin activity infoods include those affecting the pediocin-pro-ducing strain (such as inadequate environment,spontaneous loss of the bacteriocin-producingability, phage infection, or antagonism by othermicroorganisms) and those affecting the bacteriocinaction (such as emergence of bacteria1 resistance,bacteriocin binding to food components, or poorsolubility or stability in the food substrate).45.156.157

Although the levels of bacteria1 populationreduction achieved by pediocin PA-l in foodsmay preclude its consideration as a primary pres-ervation method, this bacteriocin and/or the pro-ducing strains could be very useful as an addi-tiona1 safety factor contributing to the "hurdle" or"barrier" approach for preservation of foods}.45In this kind of food preservation systems, thecombined effects of intrinsec ( e.g., pH, heat treat-ment, sa1t, sugar, chemica1 additives) and extrin-sic (e.g., storage conditions) inhibitory barrierswork synergistica1ly to extend storage life andensure the microbiologica1 safety of foods.158Currently, the incorporation of pediocin or other

Ka1chayanandl59 studied the action of pediocinPA-l (1,400 AU ml-]) added to beefmeat inocu-lated with food-spoilage strains of Clostridiumlaramiae, Lactobacillus spp., and Leuconostoc.Subsequently, the meat samples were vacuum-packaged and stored in refrigeration. Under theseconditions, the treated samples remained unal-tered after storage for 12 weeks and the popula-tion levels reached by the spoilage bacteria weresignificantly lower in the samples containing thebacteriocin than in the respective control samples.Simultaneously, another studyl60 eva1uated theability ofpediocin PA-l (500 to 5000 AU ml-]) toprevent the adhesion of L. monocytogenes to irra-diated beef meat and to decontaminate the meatinoculated with 102 to 104 cfu g-l. Both strategiesreduced the populations of L. monocytogenesbetween 1.2 and 2 log cycles, and the residua1bacteriocin activity was detected on the meat sur-face for at least 28 days at 5°C.

Berry et a1,l61 employed the pediocin-produc-ing strain P. acidilactici JDl-23 as starter culturefor the elaboration of meat products and foundthat it provided a 2-1og reduction in addedL. monocytogenes (106 cfu g-l) during summersausage production, which are results similar tothose obtained by Foegeding et a1.162 However,after storage for 2 weeks, 10% of the sausageswere Listeria-positive, a result contributing to thecurrent concern for the potentia1 presence of thisfoodborne pathogen in ready-to-eat foods, even ifthe initial levels are much lower in most of thecases. Later, Berry et al.163 showed thatcoinoculation of P. acidilactici JDl-23 andL. monocytogenes in frankfurters, followed byvacuum packaging and storage at 4°C, preventedthe growth of the pathogen for at least 60 days.

Yousef et al.164 compared the ability ofpediocin PA-l and a pediocin-producer to controlthe growth of three L. monocytogenes strains inwiener sausage exudates and showed that both

Page 22: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

approaches significantly decreased their growthunder refrigerated (4°C) and abusive (25°C) tem-perature conditions. Although the decrease in List -eria populations was faster when adding the bac-teriocin, lower fmal levels were achieved usingthe producing strain. Similar fmdings were ob-tained by Degnan et al.165 Surface inoculation ofwiener sausages with L. monocytogenes and apediocin-producing Pediococcus strain, packag-ing and storage at 4 and 25°C revealed that, in thiscase, the producer was unable to grow or producethe bacteriocin at 4°C, while reductions of up to3.4logs inListeria levels were observed at 25°C.12

Motlagh et al.62 examined the effectiveness ofpediocin PA-l in reducing the population of threeListeria strains in sterile ground beef, sausagemix, Cottage cheese, ice cream, and reconstituteddry milk and observed that the bacteriocin actionwas immediate and was not affected by the foodsystem tested. In addition, these authors showedthat the activity was concentration and strain de-pendent. For example, 1350 AU mI-l ofpediocinreduced the population of L. monocytogenesScottA, L. monocytogenes Ohio2, and L. ivanoviiATCC 19119 in I, 3, and 7 log cycles, respec-tively. A listericidal effect ( approximately 7 logcycles) has been observed in slurries preparedfrom vacuum-packaged ready-to-eat turkey breastmeat and challenged with L. monocytogenes inthe presence of diacetate and pediocin.166 Theincreased antilisterial activity in slurries withdiacetate in combination with the bacteriocin weredue to synergistic effects. The listericidal activityofpediocin PA-l has also been evaluated in slur-ries of beef muscle tissue, beef tallow, nonfat drymilk, and butterfat inoculated with a mixture oftwo L. monocytogenes strainS.167 For all slurriestested, the greatest decrease in Listeria counts(1.2 to 1.81og cycles) occurred within 1.5 min ofpediocin addition. In the same study, pediocinwas also encapsulated within phosphatidyl-cho-line-based liposomes before addition to the slur-ries or unencapsulated in slurries containing theemulsifier Tween 80. In both cases, the pediocinactivity recovered from the slurries was notablyhigher than that obtained after adding free pediocinPA-l. Cold storage of L. monocytogenes-inocu-lated pork in the presence of pediocin resulted ina 2 log reduction after 24 h when compared with~

untreated controls, regardless of whether thesarnples were stored under air, vacuum, or a modi -fied atmosphere.168

Goff et a1.64 bound pediocin PA-l to heat-killed producer cells by adjusting the pH of themedium to 6.0 and the preparation was added toirradiation-sterilized raw chicken breast meat. Theresu1ts suggested that pediocin-treated raw chickenexhibits antilisterial activity both before and aftercooking, and therefore may provide protection toconsumers against bacterial postprocessingrecontarninations and/or undercooking. Applica-tion of pediocin to food packaging films mayconstitute an alternative approach to controlL. monocytogenes in meats and poultry .Ming eta1.169 obtained apediocin PA-l extract from milk-based media and prepared antilisteria1 cellulosecasings by interna1 coating with the pediocinpodwer (7.75 ~g crrr2). The coated bags com-pletely inhibited growth of inoculatedL. monocytogenes through 12 weeks storage at 42C.

B. Milk and Dairy Products

Inhibition of L. monocytogenes has been dem-onstrated in several dairy systems, includingdressed Cottage cheese, half -and-half cream, andcheese sauce.53 In all these food systems, therewas a rapid decrease in viabIe counts ofL. monocytogenes in the presence of pediocinPA-l podwer over the pR range 5.5 to 7.0 and atboth 4 and 32°C. Although a resurgence of thepathogen after 7 days at 4°C in the midly acidicand neutral dairy systems was described, decreasesin Listeria populations of at least 2 log cycleswere observed in such substrates. As cited above,Motlagh et al.62 and Degnan et aU67 also reportedthe antilisterial activity of pediocin PA-l in dairy

products.Despite the fact that pediococci are poorly

adapted to dairy substrates, it has been reportedthat different pediocin-producing P. acidilacticistrains may contribute to L. monocytogenes con-trol in milk when a very high inoculum (109 cfuml-!) is used.170 In contrast, the pediocin producerLb. plantarum WHE 92 was originally isolatedfrom a smear-surface soft cheese17 and, therefore,it must be well-adapted to this dairy environment.

Page 23: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

also notab1y decreased L. monocytogenes countsand inhibited its growth during a minumum of 16

days,172

In fact, the presence of L. monocytogenes inMunster cheese could be prevented by spraying acell suspension of Lb plantarum WHE 92 on thecheese surface at the begining of the ripeningperiod.141 Although L. monocytogenes was some-times detected at low levels «50 cfu g-l) after7 to II days of ripening, this microorganism wasunable to grow or survive in the presence of thebacteriocin-producing strain in any of the samplesexamined until the end ofripening (21 days). Thepathogen often reached counts higher than 1 Q4 cfumI-l in control samples. Lb. plantarum WHE 92did not interfere with the ripening process.

The L. lactis LL108 strain producing pediocinP A-11221acks several properties required for app1i-cation in milk and dairy products.17l Therefore, inorder to fully exploit the comrnercial potential of apediocin-producing dairy starter culture, L. lactisMM210, a strain that had been used previously inCheddar cheese manufacturing, was selected as analtemative host for the ped operon-encoding plas-mid.17l The anti1isterial abi1ity of the pediocin-producing L. lactis starter culture was tested inCheddar cheeses coinoculated with the lactococcalstrain and a rnixture of three L. monocytogenesstrains.17l In the experimental cheese, the counts ofthe pathogen decreased by 1 log cycle within1 week of ripening and then decreased by an addi-tionallog cycle within 3 months (fmal counts: 10cfu g-l). In control cheeses made with an isogenic,nonpediocin-producing L. lactis strain, Listeriacounts increased by approximately 4 log cyclesafter 2 weeks of ripening and then gradually de-creased to about 10' cfu g-l after 6 months. Thepresence of the plasmid encoding heterologouspediocin production in the lactococcal starter cul-ture did not affect its cheese-making qua1ity .

D. Legal Status

Up to date, nisin bas been tbe only bacterio-cin licenced as food preservative, althougb theantimicrobial potential of pediocin PA-l bas al-ready been commercially exploited as a pediocinPA-l-containing fermentate. Owing to their typi-cal association witb food fermentation and tbeirlong tradition as food-grade bacteria, pediocinPA-l-producing LAB are "generally recognizedas safe" (GRAS), and tberefore tbe benefits oftbis bacteriocin can already be exploited using tbeproducer strains or their fermentates, a practicefor whicb speciallabeling would not be required.173In fact, a wide use of bacteriocin-producing LABin food fermentations bas inadvertently or empiri-cally been made for centuries, and it seems clearthat bacteriocin residues are present in our dailyfood supply} Currently, tbere are some Europeanand US patents covering the use ofpediocin PA-lin dairy174 and meat products.175.176

VIII. CONCLUSION AND FUTURE

PROSPECTS

Many LAB bacteriocins have been identifiedover the yearsl but only nisin, and pediocin PA-lhave been exarnined extensively in applicationstudies in food systems. Although pediocin PA-lhas an antimicrobial spectrum not as wide as thatof nisin, it is generally more active against List-eria monocytogenes. The fact that this microor-ganism is of special concern to the food industry ,because of the high mortality and morbidity ratesassociated with listeriosis,45 may help to increasethe use of pediocin PA-l as a food biopreservative.

Pediocin PA-l has been the subject of consid-erabIe interest in recent years. However, there arestill many aspects that need further research. ltsthree-dimensional structure and mode of actionare far from completely elucidated. The establish-ment of the relationships between its structure

c. Other Foodstuffs

With respect to other foodstuffs, the additionof pediocin PA-l to liquidwhole egg containingL. monocytogenes showed an immediate reduc-tion in listeria1 counts and led to a greater reduc-tion (ofup to llog cycle) during heating whencompared with the levels found after heating un-treated liquid whole egg.45 The addition of pediocinPA-l during fermentation of a vegetable (kimchi)

113

Page 24: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

and its biological activity will be of particularinterest because this knowledgemay facilitate theenhancement of pediocin PA-l activity and agreater understanding of the mechanisms of bac-terlal resistance to the peptide. Protein engineer-ing, genetic engineering, and/or chemical synthe-sis may lead to the development of newantimicrobial peptides with improved properties,based on some features of the pediocin PA-lmolecule (pediocin PA-l variants, hybrld bacte-rlocins). The use of genetically modified organ-isms for the in situ production of pediocin PA-lmay meet with opposition fiom industries and/orconsumers. Therefore, the development of (in-ducible) "food-grade" genetic systems for pediocinPA-l production would be highly relevant. En-hancement of the pediocin P A-l productivity whenit is heterologously (co)produced by starter cul-tures may contribute to widen the applicability ofpediocin PA-l in foodstuffs.

The approval of pediocin PA-l as a foodadditive would require the existence of methodsthat allow the specific detection of this bacterlo-cin in foods. The recent development of specificpolyclonal antibodies is a promising step in thisdirection. Specific antibodies could also facilitatethe (industrial scale) purification of pediocin PA-lthrough the use of immunoaffinity columns.

Inasmuch as different laboratorles worldwideare working on the elucidation of all these points,we are confident that pediocin PA-l or pediocinPA-1-like peptides will be included in the nextgeneration of food preservatives.

ACKNOWLEDGMENTS

The authors are grateful to G. Firnland forsharing data prior to publication.

REFERENCES

1. Jack, R. W., Tagg, J. R., and Ray, B., Bacteriocinsof Gram positive bacteria. Microbiol. Rev., 59, 171-200, 1995.

2. De Vuyst, L. and Vandamme, E. J., Lactic acidbacteria and bacteriocins: their practica1 importance.In: Bacteriocins of Lactic Acid Bacteria, De Vuyst, L.and Vandamme, E. J., Eds., Chapman and Hall,

Glasgow, 1994, 1-11.

3. Klaenhammer, T. R., Genetics of bacteriocins pro-duced by lactic acid bacteria. FEMS Mjcrobjol. Rev.,

12, 39-86, 1993.4. Nes, I. F ., Diep, D. B., Havarstein, L. S., and

Brurberg, M. B., Biosynthesis ofbacteriocins in lac-tic acid bacteria. Antonje van Leeuwenhoek Int. J.Gen. Mol. Mjcrobjol., 70, 113-128, 1996.

5. Nissen-Meyer, J. and Nes, I. F., Ribosomally syn-thesized antirnicrobial peptides: their function, struc-ture, biogenesis, and mechanism of action. Arch.Mjcrobjol., 167,67-77, 1997.

6. Delves-Broughton, J., Nisin and its uses as a food

biopreservative. Food Technol. 44,100-117, 1990.7. Stiles, M. E., Biopreservation by lactic acid bacteria.

Antonje vanLeeuwenhoek Int. J. Gen. Mol. Mjcrobjol.,70,331-345, 1996.

8. Gonzalez, C. F. and Kunka, B. S., Plasmid-associ-ated bacteriocin production and sucrose fermentationinPedjococcus acjdjlactjcj. Appl. Envjron. Mjcrobjol.53, 2534-2538, 1987.

9. Bhunia, A. K., Johnson M. G., and Ray B., Directdetection of an antimicrobia1 peptide of Pedjococcusacjdjlactjcj in sodium dodecyl sulfate-polyacrilamide

gel electrophoresis. J. Industrjal Mjcrobjol., 2,319-322, 1987.

10. Bhunia, A.K., Johnson M. G., and Ray B., Purifi-

cation, characterization and antimicrobia1 spectrum ofa bacteriocin produced by Pedjococcus acjdjlactjcj. J.

Appl. Bacterjol., 65, 261-268, 1988.11. Hoover, D. G., Walsh, P. M., Kolaetis, K. M., and

Daly, M. M., A bacteriocin produced by Pedjococcusspecies associated with a 5,5 megadalton plasmid. J.Food Prot., 59,29-31, 1988.

12. Luchansky, J. B., Glass, K. A., Harsono, K. D.,Degnan, A. J., Faith, N. G., Cauvin, B., Baccus-

Taylor, G., Arihara, K., Bater, B., Maurer, A. J.,and Cassen, R. G., Genomic ana1ysis of Pedjococcusacjdjlactjcj in turkey summer sausage. Appl. Envjron.Mjcrobjol., 58, 3053-3059, 1992.

13. Christensen, D. P. and Hutkins, R. B., Collapse oftheproton motive force in Ljsterja monocytogenes causedby a bacteriocin produced by Pedjococcus acjdjlactjcj.Appl. Envjron. Mjcrobjol., 58,3312-3315, 1992.

14. Moreira, W. L., Ajslamjento y caracterjzacjón parcjalde una bacterjocjna producjda por Pedjococcus sp.347, de orjgen cárnjco. PhD Thesis, Universidad

Complutense. Madrid, Spain, 1993.15. Schved, F. Lalazar, A. Henis, Y., and Juven, B. J.,

Purification, partia1 characterization and plasmid link-age of pediocin SJ-1, a bacteriocin produced byPedjococcus acjdjlactjcj. J. Appl. Bacterjol., 74, 67-77, 1993.

16. Daba, H., Lacroix, C., Huang, J., Simard, R. E.,and Lemieux, L., Simple method of purification andsequencing of a bacteriocin produced by Pedjococcusacjdjlactjcj UL5. J. Appl. Bacterjol., 77, 682-688,1994.

17. Ennahar, S., Aoude-Werner, D., Sorokine, 0., van

Dorsselaer, A., Bringel, F., Hubert, J. C., and

Page 25: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

28. Chen, y ., Ludescher, R. D., and MontvilIe, T. J.,Electrostatic interactions, but not the YGNGV con-sensus motif, govern the binding of pediocin PA-land its fragments to phospholipid vesicles. Appl.Environ. Microbiol., 63,4770-4777, 1997.

29. Venerna, K.,Chikindas,M. L., Seegers,J.F.M. L.,Haandrikrnan, A. J., Leenhouts, K. J., Venerna,G., and Kok, J., Rapid and efficient purificationmethod for sma11, hydrophobic, cationic bacteriocins:purification oflactococcin B and pediocinPA-l. Appl.Environ. Microbiol., 63,305-309, 1997.

30. Firnland, G., Jack, R., Jung, G., Nes, I. F., andNissen-Meyer, J ., The bactericida1 activity of pediocinPA-l is specifica11y inhibited by a 15-mer fragmentthat spans the bacteriocin from the center towards theC terminus. Appl. Environ. Microbiol., 64,5057-5060,1998.

31. Sailer, M., Helrns, G. L., HenkeI, T., Niernczura,W.P.,Nakashirna, T. T.,Stiles,M.E.,andVederas,J. C., 15N- and 13C-labeled media from Anabaena sp.for universa1 isotopic labeling of bacteriocins: NMRresonance assignments of leucocin A from Leuconos-toc gelidurn and nisin A from Lactococcus lactis.Biochernistry, 32,310-318, 1993.

32. Ray, B., Pediocins of Pediococcus species. In: Bacte-riocins of Lactic Acid Bacteria, De Vuyst, L. andVandarnme, E. J. Eds., Chapman and Ha11, Glasgow,

1994,465-495.33. Fimland, G., Johnsen, L., Axelsson, L., Brurberg, M.

B., Nes, I. F ., Eijsink, V. G. H., and Nissen-Meyer, J.,A C-termina1 disulfide bridge in pediocin-like bacterio-cins renders bacteriocin activity less temperature depen-dent and is a major determinant of the antimicrobia1spect!1lm. J. Bacteriol., 182, 2643-2648, 2000.

34. Ray, S. K., Kim, W. J., Johnson, M. C., and Ray,B., Conjugal transfer of a plasmid encoding bacterio-cin production and immunity in Pediococcusacidilactici H. J. Appl. Bacteriol., 66,393-399, 1989.

35. Yang, R., Johnson, M. C., and Ray, B., Novel methodto extract large amounts of bacteriocins from lacticacid bacteria. Appl. Environ. Microbiol., 58, 3355-3359, 1992.

36. Holo, H., Nissen, 0., and Nes, I. F ., Lactococcin A,a new bacteriocin from Lactococcus lactis subsp.crernoris: isolation and characterization of the proteinand its gene. J. Bacteriol., 173,3879-3887, 1991.

37. Nissen.Meyer, J., Holo, H., Havarstein, L. S.,Sletten, K., and Nes, I. F ., A novellactoccocal bac-teriocin whose activity depends on the complemen-tary action of two peptides. J. Bacteriol., 174, 5686-5692, 1992.

38. Horn, N., Martinez, M. I., Martinez, J. M.,Hernández, P. E., Gasson, M. J., Rodriguez, J. M.,and Dodd. H. M., Production of pediocin PA-l byLactococcus lactis using the lactococcin A secretoryapparatus.Appl. Environ. Microbiol., 64,818--823, 1998.

39. Horn, N., Martinez, M. I., Martinez, J. M.,Hernández, P. E., Gasson, M. J., Rodriguez, J. M.,

Hasselmann, C., Production of pediocin AcH byLactobacillus plantarum WHE92 isolated from cheese.Appl. Environ. Microbiol., 62,4381-4387, 1996.

18. Bennik,M.H.J., VerheuI,A.,Abee, T.,Naaktgeboren-Stoffels, G., Gorris, L. G. M., and Smid, E. J., Inter-actions of nisin and pediocin PA-l with closely relatedlactic acid bacteria that manifest over l00---fold differ-ences in bacteriocin sensitivity.Appl. Environ. Microbiol.,63,3628-3636, 1997.

19. Rodriguez, J. M., Cintas, L. M. Casaus, P.,Martinez, M. I., Suárez, A., and Hernández, P. E.,Detection of pediocin PA-l producing pediococci byrapid molecular producing by rapid molecular biol-ogy techniques. Food Microbiol., 14,363-37, 1997.

20. Henderson, J. T ., Chopko, A. L., and vanWassenaar, P. D., Purification and primary structureof pediocin PA-l produced by Pediococcus acidilacticiPAC1.0. Arch. Biochem. Biophys., 295,5-12, 1992.

21. Nieto.Lozano, J. C., Nissen-Meyer, J., Sletten, K.,Peláez, C., and Nes, I. F ., Purification and arninoacid sequence of a bacteriocin produced byPediococcus acidilactici. J. Gen. Microbiol., 138, 1-6, 1992.

22. Marugg, J. D., González, C. F ., Kunka, B. S.,Ledeboer, A. M, Pucci, M. J., Toonen, M. Y.,Walker, S. A., Zoetmulder, L. C. M., andVandenbergh, P. A., Cloning, expression, and nucle-otide sequence of genes involved in production ofpediocin PA-l, a bacteriocin from Pediococcusacidilactici PAC1.0. Appl. Environ. Microbiol., 58,2360-2367, 1992.

23. Motlagh, A. M., Bhunia, A. K., Szosteck, F ., Hansen,T. R., Johnson, M. C., and Ray, B., Nucleotide andarnino acid sequence of pap-gene (pediocin AcH pro-duction) in Pediococcus acidilactici H. Lett. Appl.Microbiol., 15,45-48, 1992.

24. Firnland, G., Blingsmo, 0. R., Sletten, K., Jung, G.,Nes, I. F., and Nissen-Meyer, J., New biologica1lyactive hybrid bacteriocins constructed by combiningregions from various pediocin-like bacteriocins: theC-termina1 region is important for determining speci-ficity.Appl. Environ.Microbiol., 62,3313-3318, 1996.

25. Chikindas, M. L., Garcia.Garcerá, M. J., Driessen,A. J. M., Ledeboer, A. M., Nissen-Meyer, J., Nes, I.F., Abee, T., Konings, W. N., and Venerna, G.,Pediocin PA-l, a bacteriocin from Pediococcusacidilactici PAC1.0, forms hydrophilic pores in thecytoplasmic membrane of target cells. Appl. Environ.Microbiol., 59, 3577-3584, 1993.

26. Eijsink, V. G. H., Skeie, M., Middelhoven, P. H.,Brurberg, M. B., and Nes, I. F ., Comparative studiesof class IIa bacteriocins of lactic acid bacteria. Appl.Environ. Microbiol., 64, 3275-3281, 1998.

27. Chen, y ., Shapira, R., Eisenstein, M., and Montville,T. J., Functiona1 characterization of pediocin PA-lbinding to liposomes in the absence of a protein re-ceptor and its relationship to a predicted tertiary struc-ture. Appl. Environ. Microbiol., 63,524-531, 1997.

115

Page 26: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

52. Tagg, J. R., Dajani, A. S., and Wannamaker, L.W ., Bacteriocins of Gram-positive bacteria. Bacteriol.Rev., 40, 722-756, 1976.

53. Pucci, M. J., Vedamuthu, R. E. R., Kunka, B. S.,and Vandenbergh, P. A., Inhibition of Listeriamonocytogenes by using bacteriocin PA-l producedby Pediococcus acidilactici PAC1.0. Appl. Environ.Microbiol., 54, 2349-2353, 1988.

54. Bhunia, A. K., Johnson, M. C., Ray, B., andKalchayanand, N., Mode of action of pediocin AcHfrom P ediococcus acidilactici H on sensitive bacterialstrains. J. Appl. Bacteriol., 70, 25-33, 1991.

55. Venema, K., Kok, J., Marugg, J. D., Toonen, M.Y., Ledeboer, A. M., Venema, G., and Chikindas,M. L., Functional ana1ysis of the pediocin operon ofPediococcus acidilactici PAC1.0: PedB is the immu-nity protein and PedD is the precursor processingenzyme. Mol. Microbiol., 17,515-522, 1995.

56. Geis, A., Singh, J., and Teuber, M., Potential oflactic streptococci to produce bacteriocin. Appl.Environ. Microbiol., 45, 205-211, 1983.

57. Blom, H., Katla, T ., Hagen, B. F ., and Axelsson, L.,A model assay to demonstrate how intrinsic factorsaffectdiffusion ofbacteriocins. Int. J. FoodMicrobiol.,38, 103-109, 1997.

58. Rose, N. L., Sporns, p ., and McMuIlen, L., Detec-tion of bacteriocins by matrix-assisted laser desorp-tion/ionization time-on-flight mass spectrometry .Appl.Environ. Microbiol., 65, 2238-2242, 1999.

59. Rasch, M. and Knochel, S., Variations in toleranceof Listeria monocytogenes to nisin, pediocin PA-land bavaricin A. Lett. Appl. Microbiol., 27, 275-278,1998.

60. Dykes, G. A., Bacteriocins: ecological and evolution-

ary significance. Trends Ecol. Evol., 10, 186-189,1995.

61. Dykes, G. A. and Hastings, J. W., Fitness costsassociated with class lIa bacteriocin resistance in List-eria monocytogenes B73. Lett. Appl. Microbiol., 26,5-8, 1998.

62. Motlagh, A. M., HoIla, S., Johnson, M. C., Ray, B.,and Field, R. A., Inhibition of Listeria spp. in sterilefood systems by pediocin AcH, a bacteriocin pro-duced by Pediococcus acidilactici H. J. Food Prot.,55, 337-343, 1992.

63. Murray, M. and Richard, J. A., Comparative studyof the antilisteria1 activity of nisin A and pediocinAcH in fresh ground pork stored aerobicaIly at 5°C. J.Food Prot., 60, 1534-1540, 1997.

64. Goff, J. H., Bhunia, A. K., and Johnson, M. G.,Complete inhibition of low levels of Listeriamonocytogenes on refrigerated chicken meat withpediocin AcH bound to heat-killed Pediococcusacidilactici ceIls. J. Food Prot., 59, 1187-1192, 1996.

65. Noerlis, Y. and Ray, B., Factors influencing immu-nity and resistance of Pediococcus acidilactici to bac-teriocin pediocin AcH. Lett. Appl. Microbiol., 18,138-143, 1994.

and Dodd. H. M., Enhanced production of pediocinPA-l, and coproduction of nisin and pediocin PA-l,by Lactococcus lactis. Appl. Environ. Microbiol., 65,

4443-4450,1999.40. Martinez, J. M., Martinez, M. I., Suárez, A. M.,

Herranz, C., Casaus, P ., Cintas, L. M., Rodriguez,J. M., and Hernández, P. E., Generation of polyclona1antibodies of predetermined specificity againstpediocin PA-l. Appl. Environ. Microbiol., 64,4536-4545, 1998.

41. Martinez, J. M., Martinez, M. I., Herranz, C.,Cintas, L. M., Rodriguez, J. M., and Hernández, P .E., Polyclona1 antibodies ofpredetermined specificityagainst a synthetic 1-9 N-termina1 fragment of pediocinPA-l : specificity and cross-reactivity with other classIla bacteriocins.Microbiology, 145,2777-2787 ,1999.

42. Guyonnet, D., Fremaux, C., Cenatiempo, Y. andBerjeaud, J. M., Method for rapid purification ofclass Ila bacteriocins and comparison of their activi-ties. Appl. Environ. Microbiol., 66, 1744-1748.

43. Suárez, A. M., Rodriguez, J. M., Morales, P.,Hernández, P. E., and Azcona-Olivera, J. I., Devel-opment of monoclonal antibodies to the lantibioticnisin A. J. Agric. Food Chem., 44,2936-2940,1996.

44. Suárez, A. M., Azcona, J. I., Rodriguez, J. M.,Sanz, B., and Hernández, P. E., One-step purifica-tion of nisin A by immunoaffmity chromatography.Appl. Environ. Microbiol., 63,4990--4992,1997.

45. Muriana, P. M., Bacteriocins for control of Listeriaspp. in food. J. Food Prot., Suppl., 54-63, 1996.

46. Firnland, G., Johnsen, L., Brurberg, M. B.,Axelsson, L., Nes, I. F. Eijsink, V ., and Nissen-Meyer, J., Structure-function analysis of sakacinP and pediocin PA-l by site-directed mutagen-esis. Abstracts of the Sixth Symposium on LacticAcid Bacteria: Genetics, Metabolism and Appli-cations, FEMS, Veldhoven, The Netherlands,C83,1999.

47. Kalchayanand, N., Hanlin, M. B., and Ray, B.,Subletha1 injury makes Gram-negative and resistantGram-positive bacteria sensitive to the bacteriocins,pediocin AcH and nisin. Lett. Appl. Microbiol., 15,239-243, 1992.

48. Kalchayanand, N., Sikes, A., Dunne, C. P., andRay, B., Factors influencing death and injury offoodborne pathogens by hydrostatic pressure-pasteur-ization. Food Microbiol., 15, 207-214, 1998.

49. Kalchayanand, N., Sikes, A., Dunne, C. P., andRay, B., Interaction of hydrostatic pressure, time andtemperature of pressurization and pediocin AcH oninactivation of foodborne bacteria. J. Food Prot., 61,425-431, 1998.

50. Ray, B., Pediocin(s) of Pediococcus acidilactici as afood biopreservative. In: Food Biopreservatives ofMicrobial Origin, Ray, B. and Daeschel, M., Eds.,CRC Press, Boca Raton, 1992,265-322.

51. Ray, B., Sublethal injury, bacteriocins and food mi-crobiology. ASM News, 45, 34-36, 1993.

Page 27: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

80. van Belkum, M. J., Hayema, B. J., Jeeninga, R. E.,Kok, J., and Venerna, G., Organization and nucle-otide sequence of two lactococcal bacteriocin oper-ons. Appl. Environ. Microbiol., 57,492-498,1991.

81. Venerna, K., Abee, T., Haandrikman, A. J.,Leenhouts, K. J., Kok, J., Konings, W. N., andVenerna, G., Mode of action of1actococcin B, a thiol-activated bacteriocin from Lactococcus lactis. Appl.Environ. Microbiol., 59, 1041-1048. 1993.

82. Bruno, M. E. C. and Montville, T. J., Comrnonmechanistic action of bacteriocins from lactic acidbacteria. Appl. Environ. Microbiol., 59, 3003-3010,1993.

83. Driessen, A. J. M., van den Hooven, H. W ., Kuiper,W ., van den Kamp, M., Sahl, H. G., Konings, R. N.H., and Konings, W. N., Mechanistic studies oflantibiotic-induced permeabilization of phospholipidvesicles. Biochemistry, 34: 1606-1614, 1995.

84. Chen, Y. and Montville, T. J., Efflux of ions andATP depletion induced by pediocin PA-l are con-comitant with cell death in Listeria monocytogenesScott A. J. Appl. Bacteriol., 79, 684-{i90, 1995.

85. Abee, T.,Rombouts,F.M.,Hugenholtz,J.,Guihard,G., and Letellier, L., Mode of action of nisin Zagainst Listeria monocytogenes Scott A grown at highand low temperatures. Appl. Environ. Microbiol., 60,1962-1968, 1994.

86. Winkowski, K., Bruno, M. E.C., and Montville, T .J., Correlation of bioenergetic parameters with celldeath in Listeria monocytogenes cells exposed to ni-sin. Appl. Environ. Microbiol., 60, 4186-4188, 1994.

87. Kaiser, A. L. and Montville, T. J ., Purification of thebacteriocin bavaricin MN and characterization of itsmode of action against Listeria monocytogenes ScottA cells and lipid vesicles. Appl. Environ. Microbiol.,62, 4529-4535, 1996.

88. Winkowski, K., Ludescher, R. D., and Montville,T. J., Physicochemical characterization of the nisin-membrane interaction with liposomes derived fromListeria monocytogenes. Appl. Environ. Microbiol.,

62,323-327,1996.89. Chen, Y., Ludescher, R. D., and Montville, T. J.,

Influence of lipid composition of pediocin PA-l bind-ing to phospholipid vesic1es. Appl. Environ. Microbiol.,64, 3530-3532, 1998.

90. EI-Jastimi, R. and Lafleur, M., Structural character-ization of free and membrane-bound nisin by infraredspectroscopy. Biochim. Biophys. Acta, 1324, 151-

158,1997.91. Bhugaloo- Vial, p ., Douliez, J .P ., Mollé, D., Dousset,

X., Boyaval, p ., and Marion, D., Delineation of keyamino acid side chains and peptide domains for anti-microbial properties of divercin V41, a pediocin-Iikebacteriocin secreted by Carnobacterium divergensV41. Appl. Environ. Microbiol., 65,2895-2900,1999.

92. Quadri L. E. N., Sailer, M., Roy, K. L., Vederas, J.C., and Stiles, M. E., Chemical and genetic charac-terization ofbacteriocins produced by Carnobacterium

66. Rekhif, N., Atrih, A., and Levefre, G., Selection ofspontaneous mutants of Listeria monocytogenes A TCC15313 resistant to different bacteriocins produced bylactic acid bacteria strains. Curr. Microbiol., 28,237-241, 1994.

67. Ming, X. and Daeschel, M. A., Nisin resistance offoodbome bacteria and the specific resistance responsesof Listeria monocytogenes Scott-A. J. Food Prot., 56,944-948, 1993.

68. Davies, E. A., Falahee, M. B., and Adams, M. R.,Involvement of the cell envelope of Listeriamonocytogenes in the acquisition of nisin resistance.J. Appl. Bacteriol., 81, 139-146, 1996.

69. Maisnier-Patin, S. and Richard, J., Cell wall changesin nisin-resistant variants of Listeria innocua grownin the presence of high nisin concentrations. FEMSMicrobiol. Lett., 140,29-35, 1996.

70. Goulhen, F., Meghrous, J., and Lacroix, C., Char-acterization of nisin-resistant strains of Pediococcusacidilactici UL5, a producer of pediocin. J. Appl.Microbiol., 85, 387-397, 1998.

71. Hanlin, M.B., Kalchayanand, N., Ray, P ., and Ray .,B., Bacteriocins of lactic acid bacteria in combinationhave greater antibacterial activiy. J. Food Prot., 56,252-255, 1993.

72. Mulet.Powell, N., Lacoste.Armynot, A. M., VilÏas,M., and De Buochberg, M. S., Interactions betweenpairs ofbacteriocins from lactic acid bacteria. J. FoodProt., 61, 1210-1212, 1998.

73. Montville, T. J. and Chen, y ., Mechanistic action ofpediocin and nisin: recent progress and unresolvedquestions. Appl. Microbiol. Biotechnol., 50,511-519,1998.

74. Mitchell, P ., Chemiosmotic coupling in oxidative andphotosynthetic phosphorylation. Biol. Rev. CambridgePhilos. Soc., 41,445-502, 1966.

75. Bruno, M. E. C., Kaiser, A., and Montville, T. J.,Depletion of proton motive force by nisin in Listeriamonocytogenes cells. Appl. Environ. Microbiol., 58,2255-2259, 1992.

76. Gao, F. H. , Abee, T ., and Konings, W. N., Mecha-nism of action of the peptide antibiotic nisin in 1ipo-somes and cytochrome C oxidase-containingproteoliposomes. Appl. Environ. M icrobiol., 57, 2164-

2170,1991.77. Kordel, M., Schiiller, F., and Sahl, H. G., Interac-

tion of the pore-forming peptide antibiotics Pep5, ni-sin and subti1in with non-energized liposomes. FEBSLett., 224, 99-102, 1989.

78. Okereke, A. and Montville, T. J., Nisin dissipatesthe proton motive force of the obligate anaerobeClostridium sporogenes PA3679. Appl. Environ.

Microbiol., 58,2463-2467,1992.79. Ruhr, E. and Sahl, H. G., Mode of action of the

peptide antibiotic nisin and influence of the mem-brane potential of whole ce11s and on cytop1asmic andartificial membrane vesicles. Antimicrob. Agents

Chemother., 27,841-845,1985.

117

Page 28: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

piscicola LV 17B.J. Biol. Chern., 269,12204-12211,1994.

93. Hastings, J. W., Sailer, M., Johnson, K., Roy, K.L., Vederas, J. C., and Stiles, M. E., Characteriza-tion of leucocin A-UAL187 and cloning of the bacte-riocin gene from Leuconostoc gelidurn. J. Bacteriol.,173,7491-7500, 1991.

94. Fleury, y ., Dayem, M. A., Montagne, J. J.,Chaboisseau, E., Caer, J. P ., Nicolas, P ., and Delfour ,A., Cova1ent structure, synthesis, and structure-functionstudies ofmesentericin Y105, a defensive peptide fromGrarn-positive bacteria Leuconostoc rnesenteroides. J.Biol. Chern., 271, 14421-14429, 1996.

95. Chen, y ., Physicochernica1 Characterization and Struc-ture Function Relationship for the Interaction ofPediocin PA-l with Listeria monocytogenes Cell andLipid Vesicles. PhD thesis, The State University ofNew Jersey, New J'ersey, 1998.

96. Mi1Ier, K. W., Schamber, R., Chen, Y., and Ray,B., Production of active chimeric pediocin PA-l inEscherichia coli in the absence of processing andsecretion genes from Pediococcus pap operon. Appl.Environ. Microbiol., 64, 14-20, 1998.

97. Mi1Ier, K. W ., Schamber, R., Chen, Osmanagaoglu,0., and Ray, B., Isolation and characterization ofpediocin AcH chimeric protein mutants with alteredbactericidal activity. Appl. Environ. Microbiol., 64,1997-2005, 1998.

98. Bhunia, A. K., Bhowmik, T. K., and Johnson, M.G., Detennination of bacteriocin-encoding plasrnidsof Pediococcus acidilactici strains by Southern hy-

bridization.Lett.Appl.Microbiol., 18,168-170,1994.99. Jager, K. and Harlander, S., Characterization of a

bacteriocin from Pediococcus acidilactici PC andcomparison of bacteriocin-producing strains usingmolecular typing procedures. Appl. Microbiol.Biotechnol., 37, 631-637, 1992.

100. Ray, B., Motlagh, A. M., Johnson, M. C., andBozoglu, F ., Mapping of pSMB74, a plasmid encod-ing bacteriocin AcH production (Pap+ ) trait inPediococcus acidilactici H. Lett. Appl. Microbiol.,15, 35-37, 1992.

101. van der Vossen, J. M. B. M., van der Lelie, D., andVenema, G., Isolation and characterization of Strep-tococcus crernoris Wg2-specific promoters. Appl.Environ. Microbiol., 53,2452-2457,1987.

102. Motlagh, A. M., Bukhtiyarova, M., and Ray, B.,Complete nucleotide sequence of pSMB74, a plasmidencoding the production of pediocin AcH inPediococcus acidilactici. Lett. Appl. Microbiol., 18,305-312, 1994.

103. Kleerebezem, M., Quadri, L. E. N., Kuipers, 0. P .,and de Vos, W. M., Quorum sensing by peptidepheromones and two-component signal-transductionsystems in Gram-positive bacteria. Mol. Microbiol.,24,895-904, 1997.

104. Havarstein, L. S., Holo, H., and Nes, I. F., The

leaderpeptideofcolicinVsharesconsensussequences

with the leader peptides that are common among pep-tide bacteriocins produced by Gram-positive bacteria.

Microbiology, 140, 2383-2389, 1994.105. Havarstein, L. S., Diep, D. B., and Nes, I. F., A

family of bacteriocin ABC transporters carry out pro-teolytic processing oftheir substrates concomitant withexport. Mol. Microbiol., 16, 229-240, 1995.

106. Ray, B., Schamber, R., and MilIer, K. W., The

pediocin AcH precursor is biologically active. Appl.Environ. Microbiol., 65, 2281-2286, 1999.

107. Fimland, G., Nes, I. F., Eijsink, V., and Nissen.Meyer, J., Structure function studies of immunityproteins from the pediocin-family. Abstracts of theSixth Symposium on Lactic Acid Bacteria: Genetics,Metabolism and Applications, FEMS, Veldhoven, The

Netherlands, C81,1999.108. Dinh, T., Paulsen, I. T., and Saier, J., A family of

extracytoplasmic proteins that allow transport oflargemolecules across the outer membranes of Gram-nega-tive bacteria. J. Bacteriol., 176, 3825-3831, 1994.

109. Wandersman, C., Secretion across the bacteria1 outermembrane. Trends Genet., 8,317-322,1992.

110. Schiilein, R., Gentschev, I., Monenkopf, H. J., and

Goebel, W ., A topological model for the haemolysintranslocator protein HlyD. Mol. Gen. Genet., 234,155-163, 1992.

111. Stoddard, G. W., Petzei, J. P., Van Belkum, M. J.,Kok, J., and McKay, L. L., Molecular ana1yses ofthe 1actococcin A gene cluster from Lactococcus lactissubsp.lactis biovar diacetylactis WM4. Appl. Environ.Microbiol., 58, 1952-1961, 1992.

112. Salmond, G. P. C. and Reeves, P. J., Membranetraffic wardens and protein secretion in Gram-nega-tive bacteria. Trends Biochem. Sci., 18, 7-12, 1993.

113. Franke, C. M., Topology of a Type I Secretion Sys-tem for Bacteriocins of Lactococcus lactis. PhD The-sis. University of Groningen, Honand, 1998.

114. Varcamonti, M., Nicastro, G., Venema, G., andKok, J., Lactococcin A export in Lactococcus lactis.Abstracts of the Sixth Symposium on Lactic AcidBacteria: Genetics, Metabolism and Applications,FEMS, Veldhoven, The Netherlands, C93, 1999.

115. Billman.Jacobe, H., Expression in bacteria other thanEscherichia coli. Curr. Opin.Biotech., 7,499-504,1996.

116. 0lins, P .0., Expression systems. Quantity versus au-thenticity of heterologously produced proteins. Curr.Opin. Biotech., 7,487-488, 1996.

117. Allison, G. E., Worobo, R. W., Stiles, M. E., and

Klaenhammer, T. R., Heterologous expression ofthe lactacin F peptides by Carnobacterium piscicolaLVI7. Appl. Environ. Microbiol., 61, 1371-1377,1995.

118. Ahn, C. and Stiles, M. E., Plamid-associated bacte-riocin production by a strain of Carnobacteriumpiscicola from meat. Appl. Environ. Microbiol., 56,

2503-2510,1990.119. van Belkum, M. J., Worobo, R. W., and Stiles, M.

E., Double-glycine-type leader peptides direct secre-

Page 29: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

tion of bacteriocins by ABC transporters: colicin Vsecretion in Lactococcus lactis. Mol. Microbiol., 23,1293-1301, 1997.

120. Worobo, R. W., van Belkum, M. J., Sailer, M.,Roy, K. L., Vederas,J. C.,and Stiles M. E.,A signalpeptide secretion-dependent bacteriocin fromCarnobacteriurn divergens. J. Bacteriol., 177, 3143-3149, 1995.

121. van Belkum, M. J. and Stiles, M. E., Molecularcharacterization of genes invo1ved in the productionof the bacteriocin leucocin A from Leuconostoc

gelidurn. Appl. Environ. Microbiol., 61, 3573-3579,1995.

122. Chikindas, M.L., Venema, K., Ledeboer, A. M.,Venema, G., and Kok, J., Expression of lactococcinA and pediocin PA-l in heterologous hosts. Lett. Appl.Microbiol., 21, 183-189, 1995.

123. Venema, K., Dost, M. H. R., Beun, P. A. H.,

Haandrikman, A. J., Venema, G., and Kok. J., Thegenes for secretion and maturation of lactococcins arelocated on the chromosome of Lactococcus lactis IL 1403.Appl. Environ. Microbiol., 62, 1689-1692, 1996.

124. Gasson, M.J. and Anderson, P .H., High copy num-ber plasmid vectors for use in lactic streptococci.FEMS Microbiol. Letts., 30, 193-196, 1985.

125. Martinez, J.M., Kok, J., Sanders, J. WandHernández, P .E., Hetero1ogous co-production ofenterocin A and pediocin PA-l by Lactococcus lactis:detection by specific peptide-directed antibodies. Appl.Environ. Microbiol., 66, 3543-3549, 2000.

126. Coderre, P. E. and Somkuti, G. A., C1oning andexpression of the pediocin operon in Streptococcustherrnophilus and other lactic feffi1entation bacteria.Curr. Microbiol., 39, 295-301, 1999.

127. Eckart, M. R. and Bussineau, C. M., Quality andauthenticity of heterologous proteins synthesized inyeast. Curr. Opin. Biotech., 7, 525-530, 1996.

128. Schoeman, H., Vivier, M. A., du Toit, M., Dicks, L.M. T ., and Pretorius, I. S., The development ofbactericidal yeast strains by expressing the Pediococcusacidilactici pediocin gene (pedA) in Saccharornycescerevisiae. Yeast, 15,647-656, 1999.

129. Cregg, J. M., Vedvick, T. S., and Raschke, W. C.,Recent advances in the expression of foreign genes inPichia pastoris. Bio-Technology, 11,905-909, 1993.

130. Faber, K. N., Harder, W., Ab, G., and Veenhuis,

M., Methylotropic yeasts as factories for the produc-tion of foreign proteins. Yeast, 11, 1331-1344, 1995.

131. Sudbery, P. E., The expression of recombinant pro-teins in yeasts. Curr. Opin. Biotech., 7, 517-524,1996.

132. Kuipers, 0. P., Beerthuyzen, M. M., de Ruyter, P.G. G. A., Luesink, E. J., and de Vos, W. M., Auto-regu1ation of nisin biosythesis in Lactococcus lactisby signal transduction. J. Biol. Chern., 270, 27299-27304, 1995.

133. de Ruyter, P. G. G. A., Kuipers, 0. P ., Beerthuyzen,M. M., van Alenboerrigter, I. J., and de Vos, W.

M., Functional analysis of promoters in the nisin genecluster of Lactococcus lactis. J. Bacteriol., 178,3434-3439, 1996.

134. de Ruyter, P. G. G. A., Kuipers, 0. P., and de Vos,W. M., Controlled gene expression systems forLactococcus lactis with the food-grade inducer nisin.Appl. Environ. Microbiol., 62,3662-3667,1996.

135. Kuipers, O.P.,deRuyter,P. G.G.A.,Kleerebezem,M., and de Vos, W. M., Controlled overproduction ofproteins by lactic acid bacteria. Trends Biotechnol.

15,135-140,1997.136. Sanders, J. W., Venema, G., and Kok, J., A chlo-

ride-inducible gene expression cassette and its use ininduced lysis of Lactococcus lactis. Appl. Environ.

Microbiol., 63,4877-4882,1997.137. Sanders,J. W.,Leenhouts,K.,Burghoorn,J.,Brands,

J. R., Venema, G., and Kok, J., A chloride-inducibleacid resistance mechanism in Lactococcus lactis and itsregulation. Mol. Microbiol., 27,299-310, 1998.

138. Kemperman, R., Sanders, J.W., Venema, G., andKok, J., Enhanced secretion of pediocin PA-l byoverexpression of the accessory protein PedC. Ab-stracts of the Sixth Symposium on Lactic Acid Bacte-ria: Genetics, Metabolism and Applications, FEMS,Veldhoven, The Netherlands, C80, 1999.

139. Yang, R. and Ray, B., Factors influencing produc-tion of bacteriocins by lactic acid bacteria. F ood

Microbiol., 11,281-291,1994.140. Biswas, S. R., Purbita, R., Johnson, M. C., and

Ray, B., Influence of growth conditions on the pro-duction of a bacteriocin, pediocin AcH, by P ediococcusacidilactici H. Appl. Environ. Microbiol., 57, 1265-

1267,1991.141. Ennahar, S., Assobhel, 0., and Hasselmann, C.,

Inhibition of Listeria monocytogenes in a smear-sur-face soft cheese by Lactobacillus plantarum WHE92,a pediocin AcH producer. J. Food Prot., 61,186-191,1998.

142. Ray, B., Motlagh, A. M., and Johnson, M. C., Pro-cessing of prepediocin in Pediococcus acidilactici.FEMS Microbiol. Rev., 12,119,1993.

143. Goulhen, F., Meghrous, J., and Lacroix, C., Pro-duction of a nisin Z pediocin mixture by pH-con-trolled mixed-strain batch cultures in supplementedwhey permeate. J. Appl. Microbiol., 86, 399-406,1999.

144. Bhunia, A. K., Johnson, M. C., Ray, B., and Belden,E. L., Antigenic property of pediocin AcH producedby Pediococcus acidilactici H. J. Appl. Bacteriol., 69,

211-215,1990.145. Martinez, M. I., Rodriguez, J. M., Suárez, A.,

Martinez, J. M., Azcona, J. I., and Hernández, P.E., Generation of polyclonal antibodies against achemically synthesized N-terminal fragment of thebacteriocin pediocin PA-l. Lett. Appl. Microbiol., 24,488-492, 1997.

146. Martinez, J. M., Martinez, M. ., Herranz, C., Mo-rales, P ., Fernández, M. F ., Cintas, L. M.,

Page 30: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

Rodriguez, J. M., and Hernández, P. E., Use andlimitations of antipeptide antibodies for detection andimmunopurification of pediocin PA-l. Abstracts ofthe Sixth Symposium on Lactic Acid Bacteria: Genet-ics, Metabolism and Applications, FEMS, Veldhoven,The Netherlands, C39, 1999.

147. Wu, W. H., Rule, D. C., Busboom, J. R., Field, R.A., and Ray, B., Starter culture and time/temperaturestorage influences on quality of fermented muttonsausages. J. Food Sci., 56,916-919, 925, 1991.

148. Suárez, A. M., Rodriguez, J. M., Hernández P. E.,and Azcona.Olivera, J. I., Generation of polyclonalantibodies against nisin: inmunization strategies andimmunoassays development. Appl. Environ. Microbiol.,

62,2117-2121,1996.149. Bhunia, A. K., Monoclonal antibody-based enzyme

immunoassay for pediocins of Pediococcus acidilactici.Appl. Environ. Microbiol., 60,2692-2696, 1994.

150. Bouksaim, M., Fliss, I., Meghrous, J., Simard, R.,and Lacroix, C., Immunodot detection of nisin Z inmi1k and whey using enhanced chemiluminiscence. J.Appl. Microbiol., 84, 176-184, 1998.

151. Aymerich, T., Holo, H., Havarstein, L. S., Hugas,M., Garriga, M., and Nes, I. F ., Biochemical andgenetic characterization of enterocin A from Entero-coccus faecium, a new antilisterial bacteriocin in thepediocin family of bacteriocins. Appl. Environ.Microbiol., 62, 1676-1682.

152. Dyson, H. J. and Wright, P. E., Antigenic peptides.FASEB J., 9,37-42,1995.

153. Walter, G., Production and use of antibodies againstsynthetic peptides. J. lmmunol. Methods, 88, 149-161.

154. Tam, J. P ., Recent advances in multiple antigen pep-tides. J. lmmunol. Methods, 196, 17-32, 1996.

155. Rolland, M. P., Bitri, L., and Besancon, P., Mono-specificity of the antibodies to bovine as1-casein frag-ment 140-149: application to the detection ofbovinemi1k in caprine dairy products. J. Dairy Res., 62,83-

88,1995.156. Schillinger, U., Geisen, R., and Holzapfel, W. H.,

Potential of antagonistic microorganisms and bacte-riocins for the biological preservation of foods. TrendsFood Sci. Technol., 7, 158-164, 1996.

157. Daeschel, M. A., Applications and interactions ofbacteriocins from lactic acid bacteria in food and bev-erages. In: Bacteriocins of Lactic Acid Bacteria,Hoover, D.G. and Steenson, L. R., Eds., AcademicPress, San Diego, 1993, 63-91.

158. Leistner, L., Food preservation by combined meth-ods. Food Res. Int., 25, 151-158, 1992.

159. Kalchayanand, N., Extension of Shelf-Life ofVacuum-Packaged Refrigerated Fresh Beefby Bacte-riocins of Lactic Acid Bacteria. PhD Thesis, Univer-sity of Wyoming, 1990.

160. Nielsen, J. W., Dickson, J. S., and Crouse, J. D.,Use of a bacteriocin produced by Pediococcusacidilactici to inhibit Listeria monocytogenes associ-

ated with fresh meat. Appl. Environ. Microbiol., 56,2142-2145, 1990.

161. Berry, E. D., Liewen, M. B., Mandigo, R. W., andHutkins, R. W ., Inhibition of Listeria monocytogenesby bacteriocin producing Pediococcus during themanufacture of fermented semidry sausage. J. FoodProt., 53, 194-199, 1990.

162. Foegeding, P. M., Thomas, A. B., Pilkington, D. H.,and Klaenhammer, T. R., Enhanced control of List-eria monocytogenes by in situ-produced pediocinduring dry fermented sausage production. Appl.Environ. Microbiol., 58, 884-890, 1992.

163. Berry, E. D., Hutkins, R. W ., and Mandigo, R. W .,The use of bacteriocin producing Pediococcusacidilactici to control postprocessing Listeriamonocytogenes contamination in frankfurters. J. F oodProt., 54, 681-686, 1991.

164. Yousef, A. E., Luchansky, J. B., Degnan, A. J., andDoyle, M. D., Behavior of Listeria monocytogenes inwiener exudates in the presence of Pediococcusacidilactici H or pediocin AcH during storage at 4 or

25°C.Appl.Environ.Microbiol.,57,1461-1467,1991.165. Degnan, A. J., Yousef, A. E., and Luchansky, J. B.,

Use of Pediococcus acidilactici to control Listeriamonocytogenes in temperature-abused vacuum-pack-aged wieners. J. Food Prot., 55, 98-103, 1992.

166. Schlyter, J. H., Glass, K. A., LoetTelholz, J., Degnan,A. J., and Luchansky, J. B., The effects of diacetatewith nitrite, lactate, or pediocin on the viability ofListeria monocytogenes in turkey slurries. Int. J. F oodMicrobiol., 19, 271-281, 1993.

167. Degnan, A. J., Buyong, N., and Luchansky, J. B.,Antilisterial activity of pediocin AcH in model foodsystems in the presence of an emulsifier or encapsu-lated within liposomes. Int. J. Food Microbiol., 18,127-138, 1993.

168. Khojasteh, A. and Murano, E.A., Inability of heatstress to affect sensitivity of Listeria monocytogenesto pediocin in pork. J. Food Safety, 16, 201-208,1996.

169. Ming,X.,Weber, G. H.,Ayres,J. W.,and Sandine,W. E., Bacteriocins applied to food packaging mate-rials to inhibit Listeria monocytogenes on meats. J.Food Sci., 62, 413-415, 1997.

170. Raccach, M. and GesheIl, D. J., The inhibition ofListeria monocytogenes in milk by pediococci. FoodMicrobiol., 10, 181-186, 1993.

171. Buyong, N., Kok, J., and Luchansky, J. B., Use ofa genetically enhanced, pediocin-producing starter cul-ture, Lactococcus lactis subsp. lactis MM217, to con-trol Listeria monocytogenes in Cheddar cheese. Appl.Environ. Microbiol., 64, 4842-4845, 1998.

l72. Choi, S. Y. and Beuchat, L. R., Growth inhibition ofListeria monocytogenes by a bacteriocin of Pediococcusacidilactici M during fermentation of kirnchi. F oodMicrobiol., 11, 30l-307, 1994.

173. Holzapfel, W. H., Geisen, R., and Schillinger, U.,Biological preservation of foods with reference to

Page 31: Pediocin PA-1, a Wide-Spectrum Bacteriocin trom Lactic Acid Bacteria

eria monocytogenes Using a Bacteriocin. EuropeanPatent Application, 89101125.6, 1989.

176. Bourdreaux, D. P. and Matrozza, M. A.,Method and Composition for Extending theShelf-Life of Processed Meats. U .S. Patent,

5137319,1992.

protective cultures, bacteriocins and food-grade en-zymes. Int. J. Food Microbiol.~ 24, 343-362, 1995.

174. Gonzalez, C. F ., Method for Inhibiting Bacteria1 SPOil-age. US Patent,4883673, 1989.

175. Vandenbergh, P. A., Pucci, M. J., Kunka, B. S.,and Vedamuthu, E. R., Method for Inhibiting List-


Recommended