Journal of Applied Microbiology 1999, 87, 500–510

Production and utilization of polyclonal antibodies against

nisin in an ELISA and for immuno-location of nisin in
producing and sensitive bacterial strains
M. Bouksaim, C. Lacroix, R. Bazin1,2 and R.E. Simard
Centre de recherche en sciences et technologie du lait, Université Laval, 1Héma-Québec, and 2Département de
biochimie, Université Laval, Sainte-Foy, Québec, Canada
6944/10/98: received 26 October 1998, revised 19 May 1999 and accepted 25 May 1999

Specific nisin polyclonal

M . B OU K SA IM , C. LA C RO IX , R. BA Z IN AN D R. E. S IM AR D . 1999.
antibodies (PAb) were produced in rabbits using nisin Z produced by Lactococcus
lactis subsp. lactis biovar diacetylactis UL 719. Antisera were obtained from white female
New Zealand rabbits that were first immunized with a nisin Z-keyhole limpet
haemocyanin conjugate and boosted with free nisin Z. Nisin-specific PAb were purified
by affinity chromatography with a yield of 15 mg specific antinisin 100 ml−1 serum. The
detection limit of the ELISA test for nisin Z was 0·75 ng ml−1 in buffer but was 1·7
and 3·5 ng ml−1 in milk and complex media broth spiked (5, 10, 20 mg ml−1) with
nisin Z, respectively. In nisin Z-spiked samples, the average concentration was between
90 and 107% of actual added amount. In contrast, when the bioassay (microtitration method)
was used, only 50–63% of nisin Z biological activity could be detected. In addition, the
affinity-purified nisin PAb, antirabbit IgG gold conjugate and transmission
electron microscopy were successfully used to locate nisin Z on producing cells and to
observe its bactericidal effects against sensitive cells.

INTRODUCTION produced by a food grade bacterium, is non-toxic, and is

destroyed by pancreatic enzymes (Ray 1992).
Lactic acid bacteria have long been known to produce
Nisin is a 34 amino acid peptide antibiotic (3500 Da) and
antimicrobial proteins, called bacteriocins (Kaletta and Entian
is synthesized by a ribosomal mechanism; lanthionine is intro-
1989). The best known and most studied bacteriocin pro-
duced by the post-translational modification of serine, threo-
duced by lactic acid bacteria is nisin, which has been approved
nine and cysteine (Dodd et al. 1990). Two natural nisin
by the World Health Organization as a preservative and has
variants have been identified, nisin A and nisin Z (Mulders
been used in more than 40 countries for various specific food
et al. 1991). The nisin Z phenotype is apparently widely
applications (Hurst 1981; Delves-Broughton 1990). Nisin is
distributed. Among 26 nisogenic strains isolated from dif-
a large spectrum bacteriocin, exhibiting antimicrobial activity
ferent habitats, 14 Lactococcus lactis strains were shown to
against a large range of Gram-positive vegetative cells and
produce nisin Z (Mulders et al. 1991). The substitution of
spores (Hurst 1981). More recently, it has been shown that
asparagine (nisin Z) for histidine (nisin A) at the 27th position
Gram-negative and resistant Gram-positive bacteria also
confers good diffusion properties and increases solubility of
become susceptible after sub-lethal treatments such as freez-
nisin Z at pH values above 6·0 (De Vos et al. 1993). These
ing, heating, or exposure to weak acids or chelating agents
characteristics are particularly important for the efficient use
(Stevens et al. 1991; Kalchayanand et al. 1992). The
of nisin in solid foods in which diffusion is a limiting factor.
antimicrobial activity can be used in foods because nisin is
To assess accurately the potential use of nisin or nisin
analogues in foods, it is necessary to study their behaviour
and interactions in complex food matrices. Accurate deter-
Correspondence to: C. Lacroix, Centre de recherche en sciences et technologie mination of bacteriocin concentrations in products such as
du lait (STELA), Pavillon Paul Comtois, Université Laval, Sainte-Foy, cheese, milk and whey is required in order to understand the
Québec, Canada, G1K 7P4 relationship between bacteriocin activity and concentration
© 1999 The Society for Applied Microbiology

during production, i.e. their mode of action and interactions
between bacteriocins and food components when bacteriocins
are used as preservatives in foods.
In recent studies, bacteriocin immunogenicity has been
tested and used to develop polyclonal (PAb) and monoclonal
antibodies (MAb) (Bhunia et al. 1990; Falahee et al. 1990;
Bhunia 1994; Stringer et al. 1995; Suarez et al. 1996a, 1996b;
Martinez et al. 1997; Bouksaim et al. 1998). Antibodies against
nisin A have already been produced and used to detect nisin
A in cheese in an immunoassay (Falahee et al. 1990; Falahee
and Adams 1992). A detection limit of 1·9 × 10−2 IU ml−1
(0·5 ng ml−1) corresponding to 40 ng g−1 was observed in
analysed cheese. A competitive direct ELISA (CD-ELISA)
was developed with PAb raised against nisin A-KLH in 6- to
8-week-old mice (Suarez et al. 1996b); the detection limit
ranged from 5 to 100 ng ml−1 of nisin A in phosphate-
buffered saline (PBS) or in MRS broth as vehicles.
Once produced, nisin can adsorb onto sensitive cell sur-
faces and remain bound to producer cells (Ray 1992). The
adsorption of nisin A onto sensitive cell surfaces in a complex
meat system has been studied (Stringer et al. 1995). Section-
staining with antibody-linked probes (antibody-mediated
staining) was successfully used to locate nisin-producing L.
lactis in situ. However, neither free nisin nor nisin bound to
susceptible cells or food components was detected. The
authors concluded that their method could not be used to
study the distribution and effectiveness of nisin in foods.
In the present paper, immunization strategies are described
for generating specific polyclonal antibodies against nisin Z,
produced by Lactococccus lactis subsp. lactis biovar diacetyl-
actis UL 719 (L. diacetylactis UL 719). The development of an
affinity chromatography purification method and a symmetric
enzyme-linked immunosorbent assay (ELISA) for the precise
and sensitive detection of nisin Z in complex media such as
milk, fermented whey permeate and fresh MRS (de Man,
Rogosa and Sharpe) medium (de Man et al. 1960), is also
described. In addition, the development of a procedure to
locate nisin on producing and sensitive cells, using affinity-
purified antinisin, antirabbit-IgG gold conjugate and electron
microscopy, is presented.
MATERIALS AND METHODS
Bacterial strains and media
Lactococcus lactis subsp. lactis biovar diacetylactis (L. diacetyl-
actis) UL 719 isolated from raw milk French cheese was used
to produce nisin Z as previously reported (Meghrous et al.
1997). Lactococcus lactis subsp. lactis (L. lactis) ATCC 11454
(American Type Culture Collection, Rockville, MD) was
used as the control, nisin A-producer strain. Pediococcus aci-
dilactici UL 5 from our collection was used as an indicator
strain in the bacteriocin activity assay (Daba et al. 1994), and
© 1999 The Society for Applied Microbiology, Journal of Applied Microbiology 87, 500–510
N IS IN I N B AC T ER IA L ST RA I NS 501
Lactococcus lactis subsp. cremoris (L. cremoris) 75266 from
Rosell Institute Inc. (Montréal, QC, Canada) as the non nisin-
producing control, sensitive to nisin. Bacterial stock cultures
were maintained at −80 °C in de Man, Rogosa and Sharpe
(MRS) broth (obtained from Rosell Institute Inc.) with 10%
(v/v) added glycerol. Dry whey permeate (Kraft, Montréal,
QC, Canada) was used to prepare the fermentation medium
for nisin Z production. Yeast extract was obtained from Difco
and pasteurized fluid milk (2%) was purchased from a local
store.
Bacteriocin activity assay
Bacteriocin activity was measured using a microtitration
method. Serial twofold dilutions of bacteriocin preparations
were performed in MicrotestTM polystyrene microplate
(Becton Dickinson, Franklin Lakes, NJ, USA) wells con-
taining 125 ml MRS broth. Aliquots (25 ml) of a 100-fold
diluted overnight culture obtained after 24 h incubation at
30 °C (final concentration of approximatively 107 cfu ml−1)
of P. acidilactici UL 5 were added to each well and microtitre
plates were incubated for 14–16 h at 30 °C. The bacteriocin
activity (expressed in arbitrary units per millilitre, AU ml−1)
was determined according to the formula: (1000/125) ×
(1/D), where D was the highest dilution that inhibited growth
of the indicator organism (Daba et al. 1994). The equivalence
between arbitrary units (AU ml−1) and international units
(IU ml−1) was determined as above, using HPLC-purified
nisin Z and Ambicin NTM (pure nisin A, Aplin and Barrett,
Beaminster, UK); 1 AU corresponded to 1 IU (40 IU  1 mg
pure nisin Z).
Production and purification of nisin Z
Fermentation medium (16 litres) containing 6% (w/v) whey
permeate powder and supplemented with 1% (w/v) yeast
extract (WPYE), was inoculated (1% v/v) with a culture of
L. diacetylactis UL 719 grown in the same medium for 14 h
at 30 °C. The fermentation was carried out in a 20 litre
bioreactor (New Brunswick Scientific, Edison, NJ, USA) for
12 h using the optimal conditions previously determined by
Amiali et al. (1998); pH was controlled at 6 by automatic
addition of 6 N NH4OH, the temperature was controlled
at 30 °C and agitation set to 70 rev min−1. After a 12 h
fermentation, the NH4OH addition was stopped and the pH
was adjusted to 4–4·5 using 6 N glacial acetic acid in order
to desorb the bacteriocins from the cell walls of producing
strains.
The purification of nisin Z from the broth medium was
based on an ammonium sulphate precipitation followed by
dialysis, as previously described by Meghrous et al. (1997).
The dialysate was concentrated by lyophilization using the
LYO-TECH freeze-dryer (LYO-SAN Inc., Lachute, QC,
502 M . B OU K SA IM E T A L.
Canada) and stored at −20 °C until use. The lyophilized
fractions, containing bacteriocin obtained from culture super-
natant fluids, were reconstituted with 0·02 N HCl. The 30 ml
fractions were purified using a Sep-Pack C18 cartridge micro-
column (Millipore, Milford, MA, USA) and reverse-phase
preparative chromatography, as described by Meghrous et al.
(1997). The active peak was collected, concentrated by
lyophilization and stored at −20 °C under N2 atmosphere
until required. Protein concentration was determined using
the DC Protein Assay (Bio-Rad Laboratories, Mississauga,
ON, Canada) with bovine serum albumin (BSA) (Pierce
Chemical Compagny, Rockford, IL, USA) and pure nisin A
(Ambicin NTM) as standards.
Anti-nisin antisera production
Nisin Z was first coupled to the keyhole limpet haemocyanin
(KLH) obtained from Pierce Chemical Company, as
described by Bouksaim et al. (1998). Coupled-nisin Z (500
mg) in 1 ml phosphate-buffered saline (PBS, pH 7·5) was then
emulsified with complete Freund’s adjuvant (Difco) at a 1 : 1
ratio (v/v) and injected subcutaneously into white female
New Zealand rabbits at four different sites in the interscapular
region on the dorsum. Another injection of conjugate mixed
with incomplete Freund’s adjuvant was given 21 d later. Ten
days after the second injection, blood was collected and the
serum was recovered. A week later, free nisin Z mixed with
incomplete Freund’s adjuvant was injected into the rabbits.
Three additional booster injections were performed con-
secutively at 21, 14 and 21 day intervals. Ten days after the
last booster, blood was collected and the serum was recovered
and stored at −20 °C until use.
Affinity chromatography
The nisin-specific PAb were purified from serum by affinity
chromatography on nisin A (Ambicin TM) coupled to a
CNBr-activated sepharose 4B (Pharmacia Biotech, Uppsala,
Sweden) (CL-nisA) column (1·0 cm internal diameter and
20 cm length, Biorad Laboratories Ltd, Mississauga, ON,
Canada). For coupling, 3 g of freeze-dried CNBr powder
were suspended in 1 mmol l−1 HCl according to the manu-
facturer’s instructions. The gel swelled immediately, giving
a final gel volume of 10 ml after packing into the column.
The gel was washed with 1 mmol l−1 HCl (approximately
200 ml g−1 freeze-dried powder, added in several aliquots)
and stored at 4 °C until use. The ligand (KLH or nisin A)
was dissolved in coupling buffer, 100 mmol l−1 NaHCO3,
pH 8·3, containing 500 mmol l−1 NaCl (about 5 ml coupling
solution g−1 CNBr freeze-dried powder). The coupling solu-
tion containing the ligand was mixed with the gel in a stop-
pered vessel. This mixture was rotated overnight at 4 °C with
gentle stirring and was then packed into the columns. To
© 1999 The Society for Applied Microbiology, Journal of Applied Microbiology 87, 500–510
remove excess ligand after coupling, the adsorbent was
washed alternately with high and low pH buffer solutions (at
least 3 gel volumes of each buffer) at least three times. Acetate
buffer (0·1 mol l−1, pH 4·0) and coupling buffer (pH 8·3),
each containing 500 mmol l−1 NaCl, were used. Finally, the
column was washed with phosphate-buffered saline (PBS;
pH 7·5) and stored with 0·02% sodium azide (w/v) in PBS
at 4 °C until required. Coupling efficiency was determined
by measuring protein concentration at 280 nm in the washing
eluent.
To purify antinisin, 10 ml of antiserum were mixed with
PBS to decrease viscosity in a 1 : 1 ratio (v/v) and applied to
the column at a rate of 0·15 ml min−1. The column was
washed with PBS until all non-specific proteins had been
removed, when optical density at 280 nm reached zero. Speci-
fic antibodies against nisin were eluted in 1·5 ml fractions
using 100 mmol l−1 glycine–HCl (pH 2·5) and the second
elution was carried out in 1·5 ml fractions with 1 mmol l−1
HCl. Fractions were monitored by absorbance measurements
at 280 nm. The fractions with high absorbance values were
pooled, neutralized with 1 mol l−1 K2HPO4, and immediately
concentrated to at least 1 mg ml−1 of pure antinisin by chan-
ging the elution buffer to PBS 1× (pH 7·5) using cen-
trifugation at 5500 g and Centricon-30 units (Amicon Inc.,
Berverly, MA, USA). The pure antinisin was peroxidase-
labelled according to the periodate protocol of the peroxidase-
labelling kit from Boehringer Mannheim (Boehringer
Mannheim Corp., Indianapolis, IN, USA).
Using immunodot detection, the affinity-purified antinisin
was tested for its sensitivity vs. the IgG fraction of nisin Z
antisera that had been previously raised against nisin Z-KLH,
as described by Bouksaim et al. (1998). The cross-reactivity
of the crude antisera from rabbits immunized with nisin Z-
KLH against nisin Z, and that of affinity-purified antinisin
obtained when free nisin Z was injected vs. KLH, was also
tested. The labelled polyclonal antibodies were stored at
−20 °C until use.
The ELISA test
Wells of polystyrene Immulon® 2 flat bottom microtitre plates
(Dynatech Laboratories Inc., Chantilly, VI, USA) were
coated with 100 ml of affinity-purified antinisin immuno-
globulin (5 mg ml−1) in 100 mmol l−1 carbonate–bicarbonate
buffer (pH 9·6) for 1·5 h at 37 °C, followed by an overnight
incubation at 4 °C. The plates were emptied and washed four
times by filling all wells simultaneously with 150 mmol l−1
isotonic saline solution (Fisher Scientific, Nepean, ON, Can-
ada), containing 0·05% Tween-20 (S-T), using an automatic
microplate washer (Ultra-washer II, Dynatech, Chantilly, VI,
USA). To reduce non-specific binding, the wells were coated
with 300 ml 1% (w/v) Boehringer Mannheim blocking
reagent in PBS 1× (PBS) containing 0·02% (v/v) Tween-

20 (BMb-PBS-T) and incubated for 1 h at 37 °C. Next, 100 ml
of pure nisin Z standard solution (1 mg ml−1) (positive con-
trol) or nisin Z-containing samples were serially diluted in
BMb-PBS-T, and plates were incubated for 1 h at 37 °C.
Unbound material was removed by washing five times with
S-T, and 100 ml of antinisin peroxidase conjugate (1/2000)
were added to each well. Plates were then incubated for 1 h
at 37 °C then washed at least 12 times with S-T; bound
peroxidase was determined by adding 50 ml to each well from a
solution of 1 mg ml−1 of o-phenylenediamine (OPD) substrate
and hydrogen peroxide, prepared as described by the manu-
facturer (Gibco-BRL, Life Technologies, Burlington, ON,
Canada). The reaction was stopped after 25 min incubation
at 37 °C by the addition of 50 ml 1 mol l−1 H2SO4. The
absorbance at 490 nm (A490) was measured using a MR 5000
microplate reader (Dynatech). The end-point nisin Z con-
centration for each sample was arbitrarily assigned, after
removing the average blank absorbance value, as the
maximum dilution that yielded 50% or more of the absorb-
ance of the nisin-free control. In all cases, the test was per-
formed with at least four parallel wells on one plate when
developing optimal conditions. For testing samples, the
experiments were carried out in quadruplicate.
Determination of nisin Z concentration in milk,
fermented whey permeate and MRS
The ELISA method was tested for its ability to quantify nisin
Z in complex systems. Nisin Z recovery by the ELISA test
and the microtitration method was compared after the
addition of a known concentration (5, 10 and 20 mg ml−1) of
pure nisin Z to milk, fermented whey permeate (FWP) and
fresh MRS. The FWP was obtained from continuous mixed
strain immobilized cell fermentation of whey permeate (6%
w/v) supplemented with 1·5% yeast extract and 25 mmol l−1
KCl. The fermentation was carried out using three strains of
Lactococcus (two of L. lactis subsp. lactis and one of L. lactis
subsp. lactis biovar diacetylactis) grown at pH 6·2 and 30 °C
(Lamboley et al. 1997). Before the test, milk and FWP were
treated as described by Fowler et al. (1975). A 10 ml aliquot
of 0·02 N HCl was added to 2·5 ml samples of milk or FWP.
The pH was adjusted to 2·0 2 0·1 with 6 N HCl and the
samples were heated to 100 °C for 5 min. After cooling to
20 °C, the volume was adjusted to 12·5 ml with 0·02 N HCl,
and samples were centrifuged for 20 min at 4000 g at 4 °C.
The supernatant fluid was held at 4–7 °C for 30 min, and
filtered through a 0·22 mm Sartorius MinisartR sterile filter
(Sartorius AG, Göttingen, Germany) to give a clear extract.
Before the ELISA test, all samples were adjusted to pH 6
using 6 N NaOH. The extracts of milk, FWP and MRS,
without added nisin Z, were prepared as above and used as
nisin-free controls in the ELISA test. The MRS samples
were directly filtered through a 0·22 mm Sartorius filter,
© 1999 The Society for Applied Microbiology, Journal of Applied Microbiology 87, 500–510
N IS IN I N B AC T ER IA L ST RA I NS 503
adjusted to pH 6 and analysed after being spiked with nisin
Z as above. The ELISA test and the microtitration method
were performed in parallel, and nisin Z recovery from the
samples was compared between the two techniques. The
experiments were carried out in quadruplicate.
Locating nisin Z on producing cells of L.
diacetylactis UL 719
To locate nisin Z on L. diacetylactis UL 719 producer cells, an
overnight culture was grown in the whey permeate medium at
30 °C, as described above. A culture of L. cremoris 75266 was
used as the non-nisin producing control. Cells were collected
from the fermented medium by centrifuging at 10 000 g for
20 min at 4 °C. The pellet was washed twice with PBS, fixed
with 3% para-formaldehyde in PBS for 6 h at 4 °C, and then
washed three times with PBS. The pellet was dehydrated in
ethyl alcohol solution at concentrations of 30, 50, 70, 95 and
99% for 2 min at each concentration, consecutively. The
dehydrated pellet was incubated overnight in LR-White resin
(Marivac Ltd, Halifax, NS, Canada). Then, the resin was
removed and replaced by fresh resin, which was polymerized
for 7–8 h at 4 °C and sectioned using a Reichert OMU2
ultamicrotome (Reichert-Jung AG, Wien, Austria). The
resulting thin sections were automatically fixed on formvan
film-coated grids (stabilized with nickel) (J. B. EM. Service
Inc., Dorval, QC, Canada). The grids were incubated with
1/500 (v/v) affinity-purified antinisin diluted with 1% BMb
reagent in PBS (BMb-PBS) for 1 h at 37 °C. After washing
with PBS containing 0·02% (v/v) Tween-20 (PBS-T), the
grids were incubated for one more hour with 1/20 (v/v)
diluted antirabbit IgG gold conjugate in BMb-PBS buffer
(Sigma), and washed extensively with PBS-T followed by
sterile water. The grids were then treated for 90 min under
an osmium tetroxide (1% in water) atmosphere, stained suc-
cessively for 7 min using a 3% solution of uranyl acetate
followed by Reynold’s Pb-oxide solution (Bozzola and Russel
1992), and washed with distilled water. Ten grids made from
each strain pellet were analysed using the MET Joel 1200EX
electronic microscope (JOEL Inc., Peabogy, MA, USA).
Determination of bactericidal effect of nisin on
sensitive cells of L. cremoris 75266
In order to observe the bactericidal effects of nisin on L.
cremoris 75266 cells, an overnight culture was grown in WPYE
at 30 °C, as described above. The minimum inhibitory con-
centration (MIC) was determined by the microtitration tech-
nique as described by Daba et al. (1994), using 100-fold
diluted overnight culture in each well of the microplate (final
concentration of approximately 106 cfu ml−1) of L. cremoris
as indicator organism. Then, a 5 ml sample of the culture
was treated for 3 min at 24 °C with nisin A at 1 and 10 times
504 M . B OU K SA IM E T A L.
the MIC. Cells were collected by centrifugation at 10 000 g
for 20 min at 4 °C. The pellet was washed twice with PBS,
fixed with 3% para-formaldehyde in PBS for 6 h at 4 °C, and
then washed three times with PBS. The sample for electron
microscopy was prepared as described above.
RESULTS
Nisin Z was purified from the culture supernatant fluids of
L. diacetylactis UL 719 by ammonium sulphate precipitation
followed by preparative reversed-phase chromatography.
After HPLC analysis, two active peaks were collected separ-
ately. The peak representing 70% of the total bacteriocin, as
reported by Meghrous et al. (1997) and Bouksaim et al.
(1998), was retained for this study. The specific activity was
measured by the microtitration method using Ped. acidilactici
UL 5 as the sensitive organism, and estimated to be identical
to that of Ambicin NTM (pure nisin A), equal to 51 200 AU
mg−1 (51 200 IU mg−1).
Nisin antibody production
In order to develop an Enzyme-Linked Immunosorbent
Assay (ELISA) test for nisin Z, PAb were raised in white
female New Zealand rabbits against pure nisin Z from L.
diacetylactis UL 719. To produce the antibodies, first, nisin
Z coupled to KLH (nis Z-KLH) was injected twice, as
described by Bouksaim et al. (1998); the animals then received
four other injections of unconjugated nisin Z. Ten days after
the last injection, high crude antisera antibody titres were
measured (1/128 000) with the dot blot assay. After puri-
fication by affinity chromatography, a yield of 15 mg antinisin
antibody was obtained from a starting volume of 100 ml
immune rabbit serum.
ELISA test
To develop the ELISA test for nisin Z detection, affinity
purified antinisin was conjugated to peroxidase. The detec-
tion limit of the peroxidase-labelled antinisin for nisin Z was
determined by the dot blot assay, as described previously
(Bouksaim et al. 1998), and was equal to 0·75 ng ml−1 pure
nisin Z (Fig. 1B). The cross-reactivity of the crude antisera
and of purified antinisin conjugated to peroxidase with KLH
was also studied using the dot blot assay. Results show that
the crude antisera exhibited significant anti-KLH reactivity
(Fig. 1C), while this reactivity was almost zero in the purified
antinisin preparation (Fig. 1D). This result confirms the
efficiency of the purification procedure.
In this study, fresh MRS, milk and FWP were spiked with
three different known concentrations of nisin Z (5, 10 and
20 mg ml−1). Before ELISA analysis, the samples were diluted
with BMb-PBS-T to a final concentration of 1 mg ml−1 nisin
© 1999 The Society for Applied Microbiology, Journal of Applied Microbiology 87, 500–510
Fig. 1 Detection of nisin Z using dot blot. Serially-diluted nisin
Z, starting at 100 ng ml−1, was used. Lane A shows the
detection profile of nisin Z in carbonate–bicarbonate buffer
(pH 9·6) using total IgG fractions from immunized rabbits with
nisin Z-KLH (Bouksaim et al. 1998) and the commercial
antirabbit IgG peroxidase conjugate as secondary antibody.
Lane B shows the detection of nisin Z in carbonate–bicarbonate
buffer using the affinity-purified antinisin peroxidase conjugate.
Lanes C and D show a cross-reaction of serially-diluted KLH,
starting at 100 ng ml−1, with crude antisera from immunized rabbits
with nisin Z-KLH (lane C) or with affinity-purified antinisin
peroxidase conjugate as detector antibody (lane D)
Z and then tested against controls without added nisin Z.
The standard curve was linear in the concentration range
from 0·4 and 7·8 ng ml−1 (Fig. 2). In the developed symmetric
ELISA test, the detection limits for nisin Z were from 0·4 to
0·8 ng ml−1 in BMb-PBS-T buffer, 1·7 ng ml−1 in FWP and
milk, and 3·5 ng ml−1 in MRS medium.
ELISA was used to determine the specificities of the pure
polyclonal antinisin obtained against nisin Z in supernatant
fluids from 16 h MRS cultures of Ped. acidilactici UL5, L.
lactis ATCC 11454, L. cremoris 75266, and in a KLH solution
(1 mg ml−1). A cross-reaction was only observed in the super-
natant fluid from L. lactis ATCC 11454, and no cross-reac-
tions were observed in the other supernatant fluids, nor in
the KLH solution (data not shown).
In the ELISA test, the optimal dilutions for the affinity-
purified polyclonal antinisin capture antibody and purified
antinisin conjugated to peroxidase were 5 mg ml−1 and
1/2000, respectively; these were used for determinations of
sensitivity and specificity. Non-specific binding of antinisin
was efficiently prevented by incubating microplates for 1 h at
37 °C with blocking reagent in PBS buffer, supplemented
with 0·02% Tween-20 (BMb-PBST). The total time for
 N IS IN I N B AC T ER IA L ST RA I NS 505

the biological activity of added nisin Z was detected in spiked

milk and FWP, and 50% in spiked MRS (Table 1).

Fig. 2 Standard curve obtained with serially-diluted nisin Z,
starting at 1 mg ml−1 in Mb-PBS-T solution, using 5 mg ml−1 of
affinity-purified antinisin as capture antibody and 1/2000 of the
peroxidase-labelled antinisin. The assay was carried out in
quadruplicate and the equation of the regression line was:
Y  0·0408x ¦ 0·0351, r2  0·98, in the concentration range from
0·4 to 7·8 ng ml1
performing the assay was approximately 5 h when the micro-
plates were pre-coated with capture antibodies. The coated
plates could be stored at 4 °C for at least 8 weeks without loss
of activity (Märtlbauer et al. 1988).
The assessment of the ELISA test was carried out using
the supernatant fluid from L. diacetylactis UL 719, FWP, and
in milk, MRS and FWP samples each containing nisin Z at
three concentrations (5, 10 and 20 mg ml−1). The mean nisin
Z concentrations tested by ELISA corresponded to 107, 96
and 97% of the added nisin Z concentration in spiked milk,
FWP and MRS samples (Table 1). In contrast, in the micro-
titration method using Ped. acidilactici UL 5 as indicator
strain and the correspondance between arbitrary units and
nisin Z concentrations (40 AU  1 mg), only about 63% of
Immunogold localization
To locate nisin Z on producing cells of L. diacetylactis UL719,
the pellet of an overnight culture in WPYE grown at 30 °C
was used. A clear location and distribution image of nisin Z
was observed in cell preparations (Fig. 3a) from the analysis
using the affinity-purified antinisin, antirabbit IgG gold con-
jugate and electron microscopy. Nisin Z was detected on cell
walls, as well as on the inside and outside of cells. Cells of
L. cremoris 75266, which were used as non-nisin-producing
controls, did not exhibit an immunoreaction under these
analytical conditions (Fig. 3b). However, when cells of L.
cremoris were treated with 500 ng ml−1 nisin (10 × MIC),
bactericidal effects were clearly observed. The adsorption
of nisin to L. cremoris cells, and its incorporation into cell
membranes, caused membrane disruption, subsequent mem-
brane pore formation leading to cell death (Fig. 4c). At
1 × MIC (50 ng ml−1 pure nisin), nisin is localized around
the cells without affecting their state, as shown in Fig. 4a.
DISCUSSION
Previous attempts to generate specific antibodies against nisin
A (Stringer et al. 1995), pediocin AcH (Moreira 1993) and
pediocin RS2 (Bhunia 1994) have failed, or given unsatis-
factory results, when these bacteriocins were injected alone
into animals. The production of antibodies against bac-
teriocins requires the conjugation of bacteriocins to large
carrier proteins, which render them immunogenic (Suarez
et al. 1996a). However, no immunoresponse was observed
for pediocin AcH conjugated to bovine serum albumin, or
pediocin RS2 bound to heat-killed Lactobacillus plantarum
(Bhunia 1994). To avoid purifying the relatively large quan-

Table 1 Titration of nisin Z in spiked (5, 10 and 20 mg ml−1) milk, FWP and MRS by the ELISA test using 5 mg ml−1 of the affinity-
purified antinisin as capture and 1/2000 of the peroxidase-labelled antinisin as detector, and the microtitration method using Ped. acidilactici
UL 5 as indicator organism. The milk and whey were treated with 0·02 N HCl and adjusted to pH 6, and the MRS was directly filtered
and tested. Results are the averages of four experiments
––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
—
Nisin Z concentration detected Activity titres detected by the
by ELISA (mg ml−1) microtitration method (AU ml−1)
—
–––––––––––––––––––––––––––––––––––––––– —––––––––––––––––––––––––––––––
Spiked nisinZ 0·02N
(mg ml−1) Milk FWP MRS Milk FWP MRS HCl buffer
––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
—
5 5·2 2 0·7 4·55 2 0·7 5·1 2 0·25 160 160 128 256
10 10·7 2 1·3 9·8 2 0·9 9·8 2 0·8 320 320 256 512
20 20·8 2 3 20·4 2 1·0 20·4 2 2·2 640 640 512 1024
––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
—

© 1999 The Society for Applied Microbiology, Journal of Applied Microbiology 87, 500–510
506 M . B OU K SA IM E T A L.

Fig. 3 Localization of nisin Z on

bacterial cells using 1/500 dilution of
antinisin as the primary antibody and
1/20 antirabbit IgG gold conjugate
as the detector by transmission
electron microscopy. (a)
Distribution of nisin Z inside, on the
cell wall and outside the nisin Z-
producing cell, L. diacetylactis UL 719
(719: photo A); (b) L. cremoris 75266
(R  Rosell: photo b) was used as a
negative control

Fig. 4 Determination of nisin A bactericidal effects on L. cremoris 75266 (Rosell) sensitive cells, using 1/500 dilution of affinity-purified
antinisin as the primary antibody and 1/20 antirabbit IgG gold conjugate as the detector, by transmission electron microscopy. (a) Effect of
1 × MIC on L. cremoris cells; (b) effect of nisin A (10 × MIC) integration by susceptible cells; (c) disrupted cell (10 × MIC). The
untreated L. cremoris 75266 (R  Rosell: photo b) cells were used as controls (Fig. 3b).

© 1999 The Society for Applied Microbiology, Journal of Applied Microbiology 87, 500–510

tities of bacteriocins that are required for the production of
immunological reactions in animals, Martinez et al. (1997)
studied a chemically synthesized N-terminal fragment (PH1)
of pediocin PA-1 for its ability to generate polyclonal anti-
bodies. The resulting serum and ascite antibodies were speci-
fic for PH1, but did not recognize the whole pediocin PA-1
molecule.
The selection of carrier proteins, conjugation procedures
and immunoassay formats depends on each situation, accord-
ing to the characteristics of each bacteriocin. In an earlier
study, the production of nisin Z-KLH polyclonal antibodies
was demonstrated in white female New Zealand rabbits
(Bouksaim et al. 1998). The detection limit of the polyclonal
antinisin Z antiserum was as low as 3 ng ml−1 pure nisin Z
in carbonate–bicarbonate buffer, and 155 ng ml−1 in milk and
whey samples, by the dot blot test (Bouksaim et al. 1998). In
this study, an improved strategy was used to produce anti-
nisin. First, HPLC-purified nisin Z of L. diacetylactis UL
719 was coupled to KLH (nis-Z-KLH) and injected into
rabbits, twice, at 21 day intervals. The immunization pro-
cedure was continued by injecting HPLC-pure free nisin Z
four times. Ten days after the last injection of nisin Z alone,
the crude antiserum was analysed and a specific titre of
1/128 000 was observed. In a recent study, Suarez et al.
(1996a) immunized mice with nisin A-KLH. However, the
animals showed an apparent serum titre ranging from
1/12 800 to 1/25 000, which is lower than the 1/128 000
obtained when nisin Z was injected alone after injecting nisin
Z-KLH into rabbits. In view of these results, the immunizing
protocol used in the present study appears promising and has
advantages over other immunization procedures reported in
the literature. According to our immunizing protocol, free
nisin Z can induce the production of specific antibodies in
rabbits that were previously immunized twice with nisin Z-
KLH. Free nisin Z can be immunogenic, probably due to the
initial immunoresponse to nisin Z-KLH which led to the
production of nisin Z-reactive immune cells activated sub-
sequently by injections of free nisin Z. However, as already
reported in the literature, free nisin cannot induce primary
immunoresponses in animals.
As far as is known, the only other three nisin A polyclonal
or monoclonal antibodies that have been reported were gen-
erated by the immunization of sheep with nisin A–egg albu-
min conjugate (Falahee et al. 1990), the immunization of mice
with nisin A-KLH conjugate (Suarez et al. 1996a), and the
generation of hybridoma-secreting specific MAb in mice
immunized with nisin A-KLH conjugate (Suarez et al.
1996b). In comparison, a detection limit of 40 ng g−1 was
reported by Falahee et al. (1990) for an enzyme immunoassay,
and detection limits ranging from 5 to 100 ng ml−1 were
reported by Suarez et al. (1996a) and 10 ng ml−1 by Suarez
et al. (1996b) for a competitive direct ELISA. A very low
detection limit of approximately 10 mg ml−1 (corresponding
© 1999 The Society for Applied Microbiology, Journal of Applied Microbiology 87, 500–510
N IS IN I N B AC T ER IA L ST RA I NS 507
to 32 000 AU ml−1) was obtained with the monoclonal anti-
body-based immunoassay for pediocin RS2, developed by
Bhunia (1994).
Apart from the difference in immunoassay format, there
are several differences between the immunological and puri-
fication procedures that were devised here and those
described by Suarez et al. (1996a) and Falahee et al. (1990).
First, the latter authors obtained specific antiserum by using
a long immunization process (9 months). In contrast, our
procedure generated specific antibodies in less than 15 weeks.
For antinisin Ab purification, Falahee et al. (1990) used a
single affinity chromatography column with nisin as a ligand,
while Suarez et al. (1996a,b) used, directly, the crude antisera
and hybridoma supernatant fluids respectively, in their CD-
ELISA. In the present work, we used an affinity chro-
matography column with nisin A as ligand to bind specific
the antinisin Ab and to remove the IgG not specific to nisin.
The detection limit of our ELISA test in the range 1·7 to
3·5 ng ml−1 with complex media is lower than that of Falahee
et al. (1990) at 0·5 ng ml−1 of diluted cheese sample, cor-
responding to 40 ng g−1 of cheese.
The yield obtained of 15 mg pure antinisin 100 ml−1 anti-
serum is very low compared with 4 mg ml−1 total IgG fraction
when antisera were purified using A/G protein, as described
by Bouksaim et al. (1998). This indicates that nisin-specific
antibodies represent only a small fraction of the total rabbit
antibodies, even for hyperimmunized animals. The puri-
fication procedure used in this study was able to remove all
non-specific rabbit IgG present in serum as natural immuno-
globulins that could be retained if A/G protein was used.
The pure antinisin was horseradish peroxidase-labelled and
tested for its sensitivity. In the dot blot test, the lowest
concentration of nisin Z that produced a positive signal, using
a dilution ratio of 1/500 of antinisin peroxidase, was 0·75 ng
ml−1 (Fig. 1), which is approximately 4 times lower than a
previously reported detection limit of 3 ng ml−1 obtained for
the total IgG fraction from immunized rabbits by immunodot
blot and chemiluminescence (Bouksaim et al. 1998). The anti-
nisin peroxidase obtained cross-reacted with Ambicin NTM
and with nisin A, which was produced in supernatant fluids
of L. lactis ATCC 11454 culture. However, the antibodies
did not recognize KLH or pediocin contained in the super-
natant fluid of Ped. acidilactici UL 5 culture grown overnight
at 30 °C in MRS.
Using the microtitration assay, MIC for Ped. acidilactici
UL 5 of pure nisin Z was 310 ng ml−1 (Bouksaim et al. 1998),
which is much higher than the lowest concentration of nisin
Z that produced a positive signal with ELISA (0·75 ng ml−1).
In contrast, Bhunia (1994) developed a dot immunoblotting
test that failed to detect pediocin at concentrations of 1600
AU ml−1 (400 ng ml−1) or less in culture supernatant fluids
of Ped. acidilactici RS2 or AcH.
Polyclonal-Ab produced against nisin A reacted extensively
508 M . B OU K SA IM E T A L.
with supernatant fluids of a nisin Z-producer strain and
with pure nisin Z (Suarez et al. 1996a), and the reverse was
observed in this study for polyclonal antibodies against nisin
Z. These data might be expected as nisins A and Z differ by
only a single amino acid, at position 27 (Mulders et al. 1991).
In this study, the detection limit for both nisin A and Z were
the same when tested by ELISA in BMb-PBST. This result
suggests that the difference in amino acid at position 27 in
the nisin structure (A and Z) does not result in a significant
modification of the antigenic structure of the molecule.
The performance of the nisin A CD-ELISA (monoclonal
Ab) was studied by analysing nisin in spiked (0·25, 1·25 and
5 mg g−1) cheese spread; average recovery was found to be
116% (Suarez et al. 1996b). In the ELISA test developed by
Falahee et al. (1990), the recovery of nisin from cheese sam-
ples spiked with commercially used concentrations of nisin
averaged 85%. In this study, milk, MRS and FWP were
spiked with three different concentrations of nisin Z (5, 10
and 20 mg ml−1) and analysed by the ELISA test. The recov-
ery of nisin Z from MRS, milk and FWP samples averaged
between 91 and 107% (Table 1) by the ELISA test, which is
better than the recovery of nisin from cheese samples (85%)
(Falahee et al. 1990) and the average recovery obtained by
the microtitration method, approximately 50–63% (Table 1).
According to the absolute value, the immunoassay is more
sensitive, and more easily adapted to automation and the
analysis of multiple samples, than the labour-intensive bio-
assay. Furthermore, while the bioassay only detects active
nisin, the immunoassay can detect three forms of nisin: active,
inactive and prepronisin (non-mature bacteriocin) which con-
tains conserved epitopes for the antibodies. In this study, the
observed detection limits in spiked (5, 10 and 20 mg ml−1)
MRS, milk and FWP were 3·5, 1·7 and 1·7 ng ml−1, respec-
tively.
Recently, antibody-linked probes were used to locate nisin-
producing L. lactis and Bacillus cereus (sensitive strains) as
spoilage organisms in situ in a fermented meat system (Strin-
ger et al. 1995). Lactococcus lactis was observed in all meat
sections to which it had been added as starter culture at all
times during growth. The authors also concluded that free
nisin A, or nisin A bound to sensitive vegetative cells or food
components, were not detected. As this method failed to
detect free nisin or nisin-sensitive organisms, it cannot be
used to study the distribution and effectiveness of nisin in
foods.
Antibody-linked probes have also been used for the
immuno-location of micro-organisms and their products in
natural food systems to demonstrate the sites at which growth
occurs, and thereby to suggest methods of reducing microbial
spoilage (Hurst and Peterson 1971). In this study, the
immuno-location of nisin Z on nisin Z-producing L. diacetyl-
actis UL 719 cells was successfully carried out using affinity-
purified antinisin, antirabbit IgG gold conjugate and electron
© 1999 The Society for Applied Microbiology, Journal of Applied Microbiology 87, 500–510
microscopy. This technique was able to locate free nisin Z on
the outside and the inside of L. diacetylactis UL 719 cells,
respectively (Fig. 3a). In addition, nisin Z bound to cell walls
could also be observed, which may correspond to a mixture
of prepronisin Z in the process of maturation and the active
bacteriocin. Studies have reported that enzymatic processing
of nisin-precursor molecules begins just before the start of
detectable nisin synthesis (Hurst 1981). One hypothesis is
that the enzymatic modifications in molecular structure occur
before the cleavage of the inactive prepronisin. Therefore,
the active nisin molecules that may be toxic to producer-cell
membranes do not reach a high concentration but are secreted
outside the cells immediately after cleavage (Kaletta and
Entian 1989). In contrast, the non-nisin-producing control,
L. cremoris 75266, was not stained. These data, obtained with
a strain from the same L. lactis sp. as the nisin-producing
culture L. lactis subsp. lactis biovar diacetylactis UL 719,
confirmed the specificity and sensitivity of the affinity-puri-
fied antinisin produced. In addition, a control experiment
performed with the antirabbit IgG gold conjugate alone, and
L. diacetylactis UL 719 cells, resulted in no detectable signals.
When an overnight culture of L. cremoris 75266 was nisin-
treated with 1 × MIC, the cells were not affected (Fig. 4a).
At high concentrations (10 × MIC) under the same exper-
imental conditions, the bactericidal effects of nisin were very
clear, as shown by the disruption of the cytoplasmic mem-
brane and its subsequent poration (Fig. 4b, c). Among the
affected cells, some cells were destroyed and disorganized
(Fig. 4c). The primary target of nisin in vegetative cells is the
cytoplasmic membrane. Ramseier (1960) reported a strong
adsorption of nisin to Clostridium butyricum vegetative cells.
Nisin-treated cells leaked ultraviolet-absorbing material and
subsequently lysed. The author concluded that nisin acts as
a cationic surface-active detergent. As shown in Fig. 4(c), the
membrane disruption in nisin-treated cells resulted in a loss
of cytoplasmic content. This membrane disruption is now
believed to be caused by the incorporation of nisin into the
membrane, and by the subsequent formation of ion channels
or pores (Gao et al. 1991). An efflux of K¦, amino acids and
ATP occurs through the membrane pores and results in the
dissipation of membrane potential in sensitive cells. Recently,
Demel et al. (1996) studied membrane interactions of lan-
tibiotic nisin on monomolecular layers of lipids at the air/
water interface as a model membrane. The natural lantibiotics
nisin A and nisin Z proved to have a high affinity for the
anionic lipids phosphatidylglycerol and bis(phosphatidyl)
glycerol (cardiolipin). Nisin might also have affinity for
anionic lipids in target membranes of Gram-positive bacteria.
In summary, the present paper describes a reliable immun-
ization procedure for the efficient production of serum-spec-
ific polyclonal antibodies to nisin Z. Direct injection of free
nisin Z, after induction of an immunoresponse to nisin Z-
KLH conjugate, allowed the generation of polyclonal anti-

bodies, which could be used to develop a sensitive symmetric
ELISA immunoassay. Potential applications of the antinisin
antibodies include analysis of nisin in complex systems by
ELISA, rapid screening of nisin-producing strains (A and Z)
and study of bacteriocin distribution, bactericidal effects of
nisin on sensitive bacteria, interactions and bacteriocin mode
of action in complex food matrices. These antinisin antibodies
could also be used in immunoaffinity chromatography for
the development of a single-step bacteriocin purification, as
described by Suarez et al. (1997). Finally, the ELISA used
with the bioassay could estimate the specific activity of bac-
teriocins; this would make available a tool for studying the
production of bacteriocins and their changes in antimicrobial
activity in food during shelf-life.
ACKNOWLEDGEMENTS
This research was carried out within the program of the
Canadian Research Network on Lactic Acid Bacteria, sup-
ported by the National Science and Engineering Research
Council of Canada, Agriculture and Agri-Food Canada, Nov-
alait Inc., The Dairy Farmers of Canada and the Institute
Rosell Inc., and was also supported by the Fond pour les
Chercheurs et l’Avancement de la Recherche from the Prov-
ince of Québec. The technical assistance of Céline Paquin
and Lise Lemieux is especially acknowledged. The authors
wish to thank Alpin and Barrett Ltd for kindly supplying
AmbicinTM.
REFERENCES
Amiali, N., Lacroix, C. and Simard, R.E. (1998) High nisin Z
production by Lactococcus lactis ssp. lactis biovar. diacetylactis
UL719 in whey permeate with aeration. World Journal of Micro-
biology and Biotechnology 14, 887–894.
Bhunia, A.K. (1994) Monoclonal antibody-based enzyme immuno-
assay for pediocins of Pediococcus acidilactici. Applied and Environ-
mental Microbiology 60, 2692–2696.
Bhunia, A.K., Johnson, M.C., Ray, B. and Belden, E.L. (1990)
Antigenic property of pediocin AcH produced by Pediococcus
acidilactici H. Journal of Applied Bacteriology 69, 211–215.
Bouksaim, M., Fliss, I., Meghrous, J., Simard, R.E. and Lacroix,
C. (1998) Immunodot detection of nisin Z in milk and whey using
enhanced chemiluminescence. Journal of Applied Microbiology 84,
176–184.
Bozzola, J.J. and Russel, L.D. (1992) Electron Microscopy. Boston,
MA: Jones and Bartlett Publishers Inc., 542 p.
Daba, H., Lacroix, C.J., Huang, J., Simard, R.E. and Lemieux,
L. (1994) Simple method of purification and sequencing of a
bacteriocin produced by Pediococcus acidilactici UL 5. Journal of
Applied Bacteriology 77, 682–688.
De Vos, W.M., Muldlers, J.W.M., Siezen, R.J., Hugenholtz, J. and
Kuipers, O.P. (1993) Properties of nisin Z and distribution of
© 1999 The Society for Applied Microbiology, Journal of Applied Microbiology 87, 500–510
N IS IN I N B AC T ER IA L ST RA I NS 509
its gene, nisZ, in Lactococcus lactis. Applied and Environmental
Microbiology 59, 213–218.
Delves-Broughton, J. (1990) Nisin and its uses as a food preserv-
ative. Food Technology 44(11), 100–107.
de Man, J.C., Rogosa, M. and Sharpe, M.E. (1960) A medium for
cultivation of lactobacilli. Journal of Applied Bacteriology 23, 130–
135.
Demel, R.A., Peelen, T., Siezen, R.J., De Kruijff, B. and Kuipers,
O.P. (1996) Nisin Z, mutant nisin Z and lacticin 481 interactions
with anionic lipids correlate with antimicrobial activity. A mono-
layer study. European Journal of Biochemistry 235, 267–274.
Dodd, H.M., Horn, N. and Gasson, M.J. (1990) Analysis of the
genetic determinant for production of the peptide antibiotic nisin.
Journal of General Microbiology 136, 556–566.
Falahee, M.B. and Adams, M.R. (1992) Cross-reactivity of bac-
teriocins from lactic acid bacteria and lantibiotics in a nisin bio-
assay and ELISA. Letters in Applied Microbiology 15, 214–216.
Falahee, M.B., Adams, M.R., Dale, J.W. and Morris, B.A. (1990)
An enzyme immunoassay for nisin. International Journal of Food
Science and Technology 25, 590–595.
Fowler, G.G., Jarvis, B. and Tramer, J. (1975) The assay of nisin
in foods. Society of Applied Bacteriology, Technical Series 8, 91–
105.
Gao, F.H., Abee, T. and Konings, W.N. (1991) Mechanism of
action of the peptide antibiotic nisin in liposomes and cytochrome
C oxidase-containing proteoliposomes. Applied and Environmental
Microbiology 57, 2164–2170.
Hurst, A. (1981) Nisin. Advances in Applied Microbiology 27, 85–
123.
Hurst, A. and Peterson, G.M. (1971) Observations on the con-
version of an inactive precursor protein to the antibiotic nisin.
Canadian Journal of Microbiology 17, 1379–1384.
Kalchayanand, N., Hanlin, M.B. and Ray, B. (1992) Sublethal injury
makes Gram-negative and resistant Gram-positive bacteria sen-
sitive to the bacteriocins, pediocin AcH and nisin. Letters in
Applied Microbiology 15, 121–243.
Kaletta, C. and Entian, K. (1989) Nisin, a peptide antibiotic: cloning
and sequencing of the nisA gene and posttranslation processing
of its peptide product. Journal of Bacteriology 171, 1597–1601.
Lamboley, L., Lacroix, C., Champagne, C.P. and Villemard, J.C.
(1997) Continuous mixed strain mesophilic lactic starter pro-
duction in supplemented whey permeate medium using immo-
bilized cell technology. Biotechnology and Bioengineering 56, 502–
516.
Martinez, M.I., Rodriguez, J.M.A., Suarez, A., Martinez, J.M.,
Azcona, J.I. and Hernandez, P.E. (1997) Generation of polyclonal
antibodies against chemically synthetized N-terminal fragment of
the bacteriocin pediocin PA-1. Letters in Applied Microbiology 24,
488–492.
Märtlbauer, E., Gares, M. and Terplan, G. (1988) Enzyme immuno-
assay for the macrocyclic trichothecene roridin A: Production,
properties, and use of rabbit antibodies. Applied and Environ-
mental Microbiology 54, 225–230.
Meghrous, J., Lacroix, C., Bouksaim, M., Lapointe, G. and Simard,
R.E. (1997) Genetic and biochemical characterization of nisin Z
produced by Lactococcus lactis ssp. lactis biovar. diacetylactis UL
719. Journal of Applied Microbiology 83, 133–138.
510 M . B OU K SA IM E T A L.
Moreira, W.L. (1994) PhD Thesis. Universidad Complutense de
Madrid, Madrid, Spain.
Mulders, J.W.M., Boerrigter, I.J., Rollema, H.S., Siezen, R.J. and
De Vos, W.M. (1991) Identification and characterization of the
lantibiotic nisin Z, a natural nisin variant. European Journal of
Biochemistry 201, 581–584.
Ramseier, H.R. (1960) Die wirkung von nisin auf Clostridium
butyricum. Archive in Mikrobiologie 37, 57–94.
Ray, B. (1992) Nisin of Lactococcus lactis ssp. lactis as a food bio-
preservative. In Food Biopreservatives of Microbial Origin, ed.
Ray, B. and Daeschel, M. pp. 207–264. Boca Raton: CRC Press
Inc.
Stevens, K.A., Shelden, B.W., Klapes, N.A. and Klaenhammer,
T.R. (1991) Nisin treatment for inactivation of salmonella species
and other Gram-negative bacteria. Applied and Environmental
Microbiology 57, 3613–3615.
© 1999 The Society for Applied Microbiology, Journal of Applied Microbiology 87, 500–510
Stringer, S.C., Dodd, C.E.R., Morgan, M.R.A. and Waites, W.M.
(1995) Locating nisin-producing Lactococcus lactis in a fermented
meat system. Journal of Applied Bacteriology 78, 341–348.
Suarez, A.M., Azcona-Olivera, J.I., Rodriguez, J.M., Sanz, B. and
Hernandez, P.E. (1997) One-step purification of nisin A by
immunoaffinity chromatography. Applied and Environmental
Microbiology 63, 4990–4992.
Suarez, A.M., Rodriguez, J.M., Hernandez, P.E. and Azcona-
Olivera, J.I. (1996a) Generation of polyclonal antibodies against
nisin: Immunization strategies and immunoassay development.
Applied and Environmental Microbiology 62, 2117–2121.
Suarez, A.M., Rodriguez, J.M.P., Morales, P., Hernandez, P.E.
and Azcona-Olivera, J.I. (1996b) Development of monoclonal
antibodies to the lantibiotic nisin A. Journal of Agriculture and
Food Chemistry 44, 2936–2940.