reduced charge (14 and 23%) inhibited growth of L. monocytogenes in milk as well as total bacteria and coliforms in ground

1. Introduction

International Journal of Food MicrobiolT Corresponding author. Tel.: +1 902 494 6031; fax: +1 902 4200219.beef significantly (Pb0.05) better than native protamine, demonstrating that the reduced charge peptides were more inhibitoryin these high protein food matrices. Electrophoretic analysis of the 21 bacteria revealed a statistically significant (Pb0.01)relationship with antimicrobial activity, where the most negatively charged bacteria were also the most susceptible to protamine.

In conclusion, components of food matrices interfered with the antibacterial effects of the peptides, however; these undesirable

interferences were reduced by altering the electrostatic properties of protamine.

D 2005 Elsevier B.V. All rights reserved.

Keywords: Protamine; Cationic antimicrobial peptides; Electrostatic interactions; Electrophoretic mobility; Cell surface charge; Listeria

monocytogenes; Coliforms; Minimum inhibitory concentrationInhibition of foodborne bacteria by native and modified

protamine: Importance of electrostatic interactions

Ross Potter1, Lisbeth Truelstrup Hansen, Tom A. GillT

Department of Food Science and Technology, Dalhousie University, P.O. Box 1000 Halifax, Nova Scotia, Canada B3J 2X4

Received 20 May 2004; received in revised form 16 September 2004; accepted 23 December 2004

Abstract

Protamine is a naturally occurring cationic antimicrobial peptide (CAP) that has shown some promise for control of

microorganisms in food. It was hypothesized that the antibacterial effect is partially due to protamines electrostatic affinity to

the negatively charged cell envelopes of actively growing bacteria. However, nonspecific binding of the CAPs to negatively

charged food particles may reduce the effect in food systems. To test the hypothesis, the antibacterial efficacies of native and

reduced charge protamines (chemically modified by randomly blocking 10 to 71% of the guanido groups of the arginine

residues) were compared in model and food systems. In Tryptic Soy Broth, moderate reductions of charge (b26%) resulted ineither a similar or slightly improved antimicrobial efficacy, measured as the minimum inhibitory concentration (MIC) toward 21

food-related bacteria. Further reductions in positive charge led to lower antimicrobial activity. Compared to protamine, the

affinity of reduced charge protamines (10 and 20%) for binding to Listeria monocytogenes cells was higher at pH 7 and 8. As

perhaps would be expected, L. monocytogenes is most sensitive to modified protamines in this pH range. Protamine with0168-1605/$ - see front matter D 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.ijfoodmicro.2004.12.019

E-mail addr1 Present Address: Canadian Food Inspection Agency, 1992

Baffin St., PO Box 1060, Dartmouth, Nova Scotia, B2Y 3Z7.ogy 103 (2005) 2334

www.elsevier.com/locate/ijfoodmicrothe use of chemicalThe increasing trend to limitess: tom.gill@dal.ca (T.A. Gill).food preservatives has generated considerable interest

in the use of dnaturalT alternatives. Several cationic

Fisher Scientific (Nepean, ON, Canada).

Protamine sulfate (2.5 g) was dissolved in 500 ml

l of Fantimicrobial peptides (CAPs) have been studied and

applied commercially for food applications including

protamine (Ueno et al., 1987), nisin (Montville and

Chen, 1998; Siragusa et al., 1999) and pediocin

(Montville and Chen, 1995, 1998). Protamine has

been used to preserve a wide variety of foods ranging

from confection items to fruits and rice. In some

instances it has been added to foods in combination

with a reducing agent (Ueno et al., 1989). Other CAPs

have also been incorporated into polyethylene based

films where they have been shown to serve a significant

preservative function (Siragusa et al., 1999).

Protamine (MW 4112 Da) is the most cationic (pI

1113) naturally occurring CAP described to date. It

consists of 31 amino acids with 20 residues being

arginine (Arg). The peptide may be isolated from the

spermatic cells of fish, birds and mammals (Rodman

et al., 1984) and is commercially recovered from

herring (clupeine) and salmon (salmine) milt (Suzuki

and Ando, 1972). Protamine is heat resistant and able

to withstand steam sterilization without loss of

antimicrobial activity (Islam et al., 1986). It appa-

rently kills susceptible bacteria by cell envelope lysis

and condensation of the cytoplasm (Johansen et al.,

1996). It is commonly used in medicine as a slow

release carrier for insulin (Brange, 1987) as well as

during surgical procedures to reverse the blood

thinning effects of heparin (Jaques, 1973). Early

references to the antimicrobial properties of protamine

suggested that protamine was more effective at

inhibiting Gram-positive than Gram-negative cells

(Brock, 1958; Islam et al., 1984; Johansen et al.,

1995). Recent work suggests that the Gram reaction

may not be as important as previously believed and, in

certain cases, some of the most protamine-susceptible

bacteria reported have been Gram-negative organisms

(Vaara and Vaara, 1983; Johansen et al., 1997;

Truelstrup Hansen et al., 2001). It has been postulated

that the charge of the molecule may be a more

important factor with regard to its antibacterial effects

(Uyttendaele and Debevere, 1993, 1994). However,

little has been reported on the relationship between

surface charge and susceptibility of target organisms.

Proteolytic degradation by endogenous proteases and/

or proteases from the endogenous microflora may

adversely impact the antimicrobial activity of CAPs in

R. Potter et al. / International Journa24food systems. Because protamine contains unusually

high levels of Arg, it would perhaps be expected to beof 0.05 M 1,2-cyclohexanedione in 0.2 M boric acid

buffer (pH 8.5) as described by Means and Feeney

(1971). Incubations at 37 8C, were carried out in1000-ml Erlenmeyer flasks in a covered shaking water

bath to eliminate photooxidation of 1,2-cyclohexane-

dione. The extents of modifications were controlled

by varying the incubation times from 1 to 60 min. The

chemical reaction for the formation of stable Arg

derivatives is illustrated in Fig. 1. Following incuba-

tion, samples were removed from the water bath and

the reaction quenched with 500 ml ice-cold 5% (v/v)

acetic acid. The chemically modified samples were

exhaustively dialyzed in a Prep/Scale Millipore Model

P34404 ultrafiltration apparatus (Millipore, Toronto,

ON, Canada) equipped with 900 cm2, 1000 Da TFEsusceptible to tryptic degradation in food systems,

since trypsin-like proteolysis normally takes place at

peptide bonds adjacent to Arg residues (Sigma, 2004).

In the present study, quantitative reductions in the

surface charge of protamine were accomplished

through chemical modification, to investigate the

impact on antimicrobial activity and to determine if

(a) electrostatic attraction is a major factor in cell

disruption; (b) the undesirable, non-specific electro-

static binding to negatively charged food particles

would be reduced, thereby enhancing the antimicro-

bial efficacy in food systems; and (c) the susceptibility

to trypsin-like enzymes would be reduced. Also, the

relationship between bacterial surface charge and

susceptibility to protamine was investigated.

2. Materials and methods

2.1. Chemical modification of protamine

Protamine sulfate was prepared from Atlantic

herring (Clupea harengus) and its purity compared

to commercial protamine sulfate (Sigma, Oakville,

ON, Canada) to confirm compliance with the United

States Pharmacopeia (1990) criteria. l-arginine HCl

(98%) and 8-hydroxyquinolinone were purchased

from Sigma, while bromine, 1,2-cyclohexanedione

(98%) and all other reagents were obtained from

ood Microbiology 103 (2005) 2334filters and flushed with five volumes of 1% (v/v)

acetic acid and ten volumes of distilled, deionized

water. After dialysis, the samples were frozen at 308C and subsequently freeze-dried. The degree ofmodification obtained in each batch of modified

peptide was determined by quantitative measurement

of underivatized, available Arg guanido groups using

the colorimetric Sakaguchi reaction (1950). Briefly,

samples (0.1 g) were re-dissolved in 40 ml of 1% (v/v)

10,000 Ag ml1 using the food related bacterialstrains and the Alamar Bluek broth dilution assaypreviously described (Truelstrup Hansen et al.,

2001). Alamar Bluek contains the non-toxic redoxindicator resazurin, which changes colour from

blue to pink as microbial growth reduces the

redox potential (Baker et al., 1994). Microplates

terici

sted o

Mo

(%

0,

0,

Fig. 1. Modification of protamine through the chemical reaction between 1,2-cyclohexanedione and the guanido groups of arginine moieties.

R. Potter et al. / International Journal of Food Microbiology 103 (2005) 2334 25acetic acid and diluted 1:50 with distilled deionized

water. One-milliliter samples were quantitatively

assayed by allowing the unaltered Arg moieties to

form a coloured complex with 8-hydroxyquinolone

which was measured by determining the absorbance at

494 nm. The average percent modified Arg groups of

peptides in each batch were subsequently calculated

from a standard curve constructed with Arg.

2.2. Determination of minimum inhibitory

concentrations (MICs) or minimum bactericidal

concentrations (MBCs) for protamines

The antimicrobial effects of chemically modified

protamines (048%) were tested at levels of 0

Table 1

Lowest minimum inhibitory concentration (MIC) and minimum bac

ranging from 0 (unmodified, native) to 48% of arginine residues, te

Bacteria Strain Incubation

temperature

(8C)

Bacillus cereus 13061d 37

Brochothrix thermosphacta 11509d 22Lactobacillus sake 1T18e 37 0,

Listeria innocua 51742d 37 0,

Listeria monocytogenes 19115d 37 20

Listeria monocytogenes Scott Ae 37 0,

Staphylococcus aureus 25923d 37 0

a Levels of protamine modification (048%) resulting in the lowest MIC v

the lowest MIC value as indicated in the table.b Lowest MIC values determined for protamines.c Lowest MBC values determined for protamines.d Strains obtained from the American Type Culture Collection, Manassae Strains as used by Truelstrup Hansen et al. (2001).with 96 wells were used for the analysis, with each

well containing 125 Al Tryptic Soy Broth (TSB;Oxoid, Nepean, ON, Canada), 010,000 Ag ml1

peptide (final concentration), 50 Al bacterial sus-pension in TSB (~1000 cfu ml1) and 20 AlAlamar Bluek (Aqueous solution; Medicorp, Mon-treal, PQ, Canada). Testing was performed in

duplicate for each bacterium with appropriate

controls. Protamine minimum inhibitory concentra-

tions (MICs) were determined after incubation for

24 h at 18, 22 or 37 8C, depending on the speciesof bacterium (Tables 1 and 2), as the lowest

concentration inhibitory to the target organism.

Minimum bactericidal concentrations (MBCs) were

determined as the lowest protamine concentration

dal concentration (MBC) values for protamines with modifications

n Gram-positive foodborne pathogenic and spoilage bacteria

dificationa

)

MICb

(Ag/ml)MBCc

(Ag/ml)Electrophoretic

mobility

(Am/s/v/cm109)10 625 1250 1.393F0.07510 10 10 2.784F0.553

10 5 39 1.549F0.12410 20 2500 0.984F0.193

39 313 1.340F0.10910 5 625 1.193F0.074

2500 1250 1.407F0.324alues. In some cases, more than one level of modification resulted in

s, VA, USA.

terici

sted o

Mo

(%)

0, 1

0, 1

0, 1

0, 1

0, 1

0, 1

0

0, 1

0, 1

0, 1

0, 1

0, 1

0, 1

0, 1

MIC

anass

l of FTable 2

Lowest minimum inhibitory concentration (MIC) and minimum bac

ranging from 0 (unmodified, native) to 48% of arginine residues, te

Bacteria Strain Incubation

temperature

(8C)

Aeromonas hydrophila 35654d 37

Aeromonas salmonicida 80204e 22

Aeromonas salmonicida 80204-1Se 22

Escherichia coli 25922d 37

Hafnia alvei S24e 18

Morganella morganii 25830d 22

Photobacterium phosphoreum P66e 18

Pseudomonas fluorescens 13525d 22

Salmonella typhimurium 14028d 37

Serratia liquefaciens 2R4e 37

Shewanella putrefaciens 8071d 37

Shewanella putrefaciens A2e 37

Vibrio parahaemolyticus 27519d 37

Yersinia enterocolitica 35669d 37

a Levels of protamine modification (048%) resulting in the lowest

the lowest MIC as indicated in the table.b Lowest MIC values determined for protamines.c Lowest MBC values determined for protamines.d Strains obtained from the American Type Culture Collection, M

R. Potter et al. / International Journa26which resulted in no growth on Tryptone Soy Agar

(TSA; Oxoid) plates which had been surface-

inoculated with contents of blue wells, after incuba-

tion for 48 h at 18, 22 and 37 8C, depending on thebacteria.

2.3. Binding of native and charge reduced protamines

to culture media components

Charge reduced (10% and 20%) and native

protamines (final sample concentrations of 1000 Agml1) were added to TSB at pH values (adjustedwith NaOH or HCl) that ranged from 5 to 8, and the

mixtures were incubated for 1 h at 24 8C. Binding ofprotamines to TSB components resulted in the

formation of a precipitate in the broth. Immediately

after incubation, the triplicate samples prepared for

each peptide and pH combination were centrifuged

at 2000g for 20 min followed by filtration through0.22-Am Nalgene syringe filters (Fisher Scientific).The quantitative amount of unbound peptide in each

samples supernatant was subsequently determined

by affinity chromatography on a heparin agarose

e Strains as used by Truelstrup Hansen et al. (2001).dal concentration (MBC) values for protamines with modifications

n Gram-negative foodborne pathogenic and spoilage bacteria

dificationa MICb

(Ag/ml)MBCc

(Ag/ml)Electrophoretic

mobility

(Am/s/v/cm109)0, 20, 26, 43, 48 N10000 N10000 0.102F0.0910 1250 2500 1.265F0.2990 625 5000 0.619F0.0420 313 625 1.057F0.3600, 20 1250 10000 1.095F0.0510 2500 10000 0.414F0.138

20 156 1.041F0.0950 2500 N10000 0.037F0.0600, 20 1250 2500 0.314F0.0580 2500 10000 0.914F0.1560 5 78 0.322F0.2680 39 1250 0.650F0.6120, 20, 26 156 2500 0.848F0.5280 1250 10000 0.570F0.051values. In some cases, more than one level of modification resulted in

as, VA, USA.

ood Microbiology 103 (2005) 2334column as described by Truelstrup Hansen and Gill

(2000).

2.4. Binding of native and charge reduced protamines

to resting bacterial cells

Listeria monocytogenes 19115 was grown over-

night in TSB at 37 8C and washed twice in sterilephosphate-buffered saline (PBS; 0.01 M NaH2PO4,

0.137 M NaCl, 2.68 mM KCl, pH 7.0) with

centrifugation at 2000g for 20 min. Pellets wereresuspended in PBS at pH values ranging from 5 to 8

(adjusted with NaOH or HCl) to an optical absorbance

at 600 nm of 0.8, which corresponded to bacterial

numbers of about 109 cfu ml1 in the final sample.Triplicate bacterial samples were subsequently mixed

with equal volumes of PBS adjusted to the corre-

sponding pH and containing 1000 Ag ml1 of peptide(native, 10% or 20% modified protamine) to give a

final sample with peptide at a concentration of 500 Agml1. After incubation for 1 h at 24 8C, theconcentration of unbound peptide was determined

after centrifugation and filtration as described above.

trypsin (Type IX from porcine pancreas; Sigma)

dissolved in PBS. Protamine samples (native, 10%

and 20% modified) with final concentrations of 1 mg

ml1 were incubated with trypsin at 24 8C. Aliquots(10 ml) were taken in duplicate every 4 min, and

placed in a boiling water bath for 10 min to inactivate

the enzyme. Affinity chromatography was subse-

quently used to monitor the disappearance of the

protamine peak and the appearance of other bnewQpeaks indicative of proteolytic degradation (Truelstrup

Hansen and Gill, 2000).

2.8. Statistical analyses

The effect of treatments was compared using

Arg residues of the original 20 residues per protamine

molecule), as a function of reaction time at 37 8C and

l of Food Microbiology 103 (2005) 2334 272.5. Antimicrobial effect of protamines in food systems

Homogenized and pasteurized 2% (fat) milk was

obtained from a local dairy. Ground beef was freshly

prepared from a roast using a laboratory scale meat

grinder. Protamines (native, 14 or 23% modified)

were added to both foods at a final concentration of 10

mg g1. L. monocytogenes 19115 was inoculated atinitial levels of 2.5103 cfu ml1 in triplicate milksamples that included controls with no protamines;

and was enumerated, on plates of Listeria Selective

Oxford agar (Oxoid) incubated at 35 8C for 48 h.Samples were taken at 4-day intervals, during storage

at 4 8C for 16 days. The protamine containing groundmeat was also stored at 4 8C and the total viablecounts and number of coliforms were determined on

plates of TSA incubated at 35 8C for 48 h and VioletRed Bile agar (Oxoid) incubated at 35 8C for 24 h,respectively.

2.6. Determination of electrophoretic mobilities of

food-related bacteria

The 21 bacteria (Tables 1 and 2) were grown in

250 ml TSB in static cultures for 18 h at the same

incubation temperatures as were used in the MIC

assays. Similar to the procedure outlined by van der

Mei et al. (1993), cells were harvested by centrifuga-

tion at 2000g for 20 min, washed twice in 0.1 MTRIS-buffered saline (TRIS 1.21 g l1, NaCl 8.0 gl1, KCl 0.2 g l1, adjusted to pH 7.0 with HCl)before being resuspended in this buffer to obtain a

suspension with an A600 nm of 0.25. The determi-

nation of electrophoretic mobilities of these cell

suspensions was performed on a Pen Kem System

3000 (Bedford Hills, NY, USA) with at least three

independent determinations for each test strain. This

method determines the average surface charge of cells

suspended in an electrical field and is based on the

cells movement toward the positive or negative

electrodes.

2.7. Sensitivity of reduced charge protamines to

trypsin

Protamine was dissolved in sterile PBS of pH 8.0,

R. Potter et al. / International Journawhich is the optimum pH for trypsin, at a concen-

tration of 2 mg ml1, and was mixed with 1 U ml1pH 8.5. The maximum degree of modification

obtained after 60 min was 70%, representing 14 out

of 20 Arg residues.

Incubation time (min) with 1,2-cyclohexanedione0 10 20 30 40 50 60 70

Exte

nt o

f Arg

Mod

ificat

ion

(%)

0

20

40

60

80

Fig. 2. Chemical modification of the guanido groups of argininestatistical methods such as analysis of variance

(ANOVA) and simple correlation analysis as was

appropriate (Gomez and Gomez, 1984).

3. Results

3.1. Chemical modification of protamine

Fig. 2 shows the degree of modification of the

guanido groups of arginine (average percent modifiedresidues in protamine with 1,2-cyclohexanedione as a function of

incubation time.

formed, with about 80% of the added peptide being

lost from solution at pH levels z7.0 while only 1030% of the protamines were lost at pH 5.0.

The affinity of native or modified protamines for

Listeria cells increased with increasing pH values

(Fig. 4). Chemically modified protamine (10 and

20%) had a higher affinity for L. monocytogenes

19115 cells at pH 7.0 and 8.0 than did native

protamine. Interestingly, treatment with 20% modified

protamine in TSB at pH 7.0 also resulted in the lowest

MIC of 39 Ag/ml (Table 1), indicating a relationshipbetween the relative amount of protamine bound and

the sensitivity to protamine. At pH values from 5.0 to

6.0, 10% modified protamine had a higher degree of

binding to the cells than 20% modified protamine

(Fig. 4).

l of Food Microbiology 103 (2005) 23343.2. Effect of modification on bacterial inhibition in

TSB

In general, the lowest MIC values (greatest

inhibition) were most frequently associated with

either no charge reduction or a modest (up to

V20% or four blocked Arg residues) reduction inthe positively charged guanido groups of Arg, while

more extensive modifications decreased the antibac-

terial efficacy (Tables 1 and 2). Protamine with a

pH in TSB5 6 7 8

Boun

d pr

otam

ines

(%)

0

20

40

60

80

100

Native protamine10% modified protamine20% modified protamine

Fig. 3. The effect of pH on the binding of native, 10 and 20%

modified protamines to Tryptic Soy Broth (TSB). Protamines (1000

Ag ml1) were added to TSB at pH values from 5.0 to 8.0 andincubated for 1 h at 24 8C before centrifugation and assay of thesupernatant. Values represent meansFstandard deviations (n=3).

R. Potter et al. / International Journa2820% charge reduction resulted in the lowest MIC

value (39 Ag ml1) for L. monocytogenes 19115whereas MICs for E. coli for both native and 10% (or

two blocked Arg residues) modified protamines were

313 Ag ml1. Sensitivity to protamine(s) variedamong bacteria from the same genus (e.g., Listeria

spp.; Table 1). MICs ranged from only a few Ag ml1

for species such as Lactobacillus sake, L. mono-

cytogenes, Photobacterium phosphoreum and She-

wanella putrefaciens to 10,000 Ag ml1. The latterconcentration did not inhibit the highly resistant

Aeromonas hydrophila.

3.3. Binding of protamines to culture media

components and resting cells

Fig. 3 illustrates how, in the absence of bacterial

target cells, native and modified (10 and 20%)

protamines interacted with TSB components in a pH

dependent manner. Insoluble precipitates were3.4. Antimicrobial effect of modified protamines in

food systems

Fig. 5 shows the growth of L. monocytogenes

19115 in milk as well as in milk with protamines (10

mg ml1; native, 14 and 23% modified) duringstorage for 16 days at 4 8C. Initial cell numbers(2.5103 cfu ml1) increased to ca. 1010 cfu ml1 incontrols while samples containing native and modi-

fied protamines demonstrated significantly (Pb0.05)slower growth with final numbers being about

5105 cfu ml1. The effect of charge reduction

pH5 6 7 8

Boun

d pr

otam

ines

(%)

-20

0

20

40

60

80

100

Native protamine10% modified protamine20% modified protamine

Fig. 4. Binding of native, 10 and 20% modified protamines to

washed cell suspensions of L. monocytogenes 19115 in phosphate-

buffered saline with 2.5109 cfu ml1 reacted with 500 Ag ml1

protamine for 1 h at 24 8C. Values represent meansF1 standarddeviation (n=3).

at 48

ples

ed in

l of Fwas minor although significant (Pb0.05) on days 8and 12, when both modified protamines gave lower

L. monocytogenes counts than were found with

native protamine. However, it should be noted that

the milk began to clot at such high protamine

concentrations, presumably due to electrostatic inter-

actions between the peptides and negatively charged

milk proteins.

Time0 2 64

L. m

onoc

ytog

enes

(cf

u/ml)

10010110210310410510610710810910101011

Control (no protamine)Native protamine14% Mod. protamine23% Mod. protamine

Fig. 5. Growth and inhibition of L. monocytogenes 19115 in milk sam

16-day incubation period. No Listeria spp. (b100 cfu/ml) were detectdeviation (n=3).

R. Potter et al. / International JournaThe addition of protamine to ground meat (Fig.

6a and b) had no significant effect on the numbers

of total bacteria or coliforms compared to controls

during storage at 4 8C for 10 days. Controlsamples contained initial coliform numbers of 103

cfu ml1. However, the addition of charge reducedprotamines significantly reduced total viable counts

and coliforms, by 2 and 4 log units, respectively

(Fig. 6a and b).

3.5. Effect of bacterial surface charge on resistance/

susceptibility to protamine

The bacterial surface charge, measured as electro-

phoretic mobility, was found to be important for the

antimicrobial efficacy of native protamine. That is,

when the relationship between the electrophoretic

mobility of the 21 test strains and their sensitivity to

protamine as determined by MBC was analysed, the

correlation coefficient (r) relating log10 MBC and

electrophoretic mobility was 0.6624 (r2=0.4388,n=21, Pb0.01), indicating the more negativelycharged food-related bacteria were generally more

sensitive to protamine (Tables 1 and 2). Brochothrix

thermosphacta was the most negatively charged strain

and also had the lowest MBC value whereas

Pseudomonas fluorescens, the most neutral strain,

was not killed by the maximum assay concentration of

10,000 Ag ml1. The correlation between electro-

oC (Days)10 12 14 16 18

containing native and modified protamines (10,000 Ag ml1) over anon-inoculated control samples. Values represent meansF1 standard

ood Microbiology 103 (2005) 2334 29phoretic mobility and sensitivity to protamine was

even more pronounced for Gram-positive test strains

considered alone (r2=0.7864, n=7, Pb0.01). Whenstrains previously observed to be proteolytically

degrading protamine (Truelstrup Hansen et al.,

2001), i.e., A. hydrophila, Aeromonas salmonicida

80504, Bacillus cereus, Hafnia alvei and Serratia

liquefaciens, (Tables 1 and 2) were excluded from the

statistical analysis, the correlation between electro-

phoretic mobility and sensitivity to protamine became

stronger (r2=0.7771, n=16, Pb0.01). The correlationbetween log10 MIC and EM was also significant

(r2=0.3933, n=21, Pb0.01).

3.6. Effect of modification on the tryptic digestablity

of protamine

Chromatographic analysis of the digests showed

that all three protamines (native, 10 and 20%

modified) were rapidly degraded in the presence of

trypsin. There were no significant differences in the

g c 10 23% Mod. protamine

l of Funts

( Lo

789fu/g) 11

121314

Control (No protamine)Native protamine 14% Mod. protamine

a

R. Potter et al. / International Journa30rate or extent of proteolytic degradation among the

different protamines.

4. Discussion

Modified protamine with a modest charge reduc-

tion gave MICs that were the same, or in one case

lower than MICs observed with unmodified prot-

amine, while modification of more than four Arg

Time at 40 2 4

0 2 4

Tota

l Via

ble

Co

0123456

Time at 4oC

Co

liform

s (Lo

g cfu/

g)

0

1

2

3

4

5

6

7

8

9

10

Control (No protamine)Native protamine 14% Mod. protamine23% Mod. protamine

b

Fig. 6. Growth and inhibition of endogenous (a) total viable bacteria and

protamines (10,000 Ag g1) during storage at 4 8C. Values represent meaood Microbiology 103 (2005) 2334moieties resulted in loss of antimicrobial efficacy.

The finding that loss of antimicrobial activity was

related to reduction of positive charge is consistent

with other studies on antimicrobial efficacy of CAPs

with varying overall charge. Matsuzaki et al. (1997)

found that the antimicrobial effects of synthetic

magainins with charges ranging from 0 (no net

charge) to +7 increased as charge increased. Sim-

ilarly, Falla and Hancock (1997) reported that an

increase in the positive charge of indolicidin

oC (Days)6 8 10

6 8 10

12

(Days)(b) coliforms in ground beef containing added native and modified

nsF1 standard deviation (n=3).

l of Fanalogues enhanced the inhibition of E. coli by 2 to

4-fold.

Protamines in this study had 16 (20% modifica-

tion), 18 (10%) and 20 (native) positive charges per

average molecule, making these peptides significantly

more positively charged than any of the other CAPs

referred to above. The Arg modifications with 1,2-

cyclohexanedione resulted in molecular weight

increases of 5 and 9% for protamines with 10 and

20% modified arginine moieties, respectively.

Changes in molecular weight and/or molecular

architecture for the more extensively modified prot-

amines may have contributed to the lack of activity in

spite of high positive charges, e.g., 10 positive charges

for 48% modified protamine (molecular weight

increase of 22%). It was not possible to chemically

modify all 20 arginine residues and this may perhaps

have been due to steric hindrance.

Protamine binding to TSB components increased

as pH increased from 5 to 8, as in previous studies

(Truelstrup Hansen and Gill, 2000). However, con-

trary to the initial hypothesis, the decrease in surface

charge on the modified protamines did not reduce

binding to TSB components throughout the exper-

imental pH range, indicating that the binding is not

controlled solely by electrostatics. The modified

protamines were more hydrophobic [measured using

the 1-anilino-8-naphthalenesulfonate (ANS) probe]

throughout the pH range than native protamine (data

not shown) and the differences in ANS hydro-

phobicity may explain some of the TSB/protamine

interactions. The importance of cell surface charge

and initial electrostatic attraction between the CAP

and cell surface for the subsequent antimicrobial

efficacy of CAPs was evidenced by more negatively

charged bacterial cells being significantly more

sensitive to the cationic protamines. Abachin et al.

(2002) showed that a L. monocytogenes DltA-mutant

with increased negative surface charge, due to the

absence of d-alanine-lipoteichoic acid in its outer

cell wall, was more sensitive to the cationic

antimicrobials colistin, polymyxin B and nisin than

the wild type. The same relationship between surface

charge and sensitivity to CAPs was found for a

Staphylococcus aureus DltA-mutant and its parent

(Peschel et al., 1999) and may be general for Gram-

R. Potter et al. / International Journapositive bacteria (Neuhaus and Baddiley, 2003).

Likewise, the susceptibility of Gram-negative bac-teria to cationic antimicrobials has been related to the

overall charge of outer cell wall components,

including lipopolysaccharide and core region phos-

phorylated polysaccharides (Nikaido, 2003). Further

evidence indicating a relationship between protamine

binding to the bacterial cell and inhibition comes

from the observation that L. monocytogenes bound

more 20% modified protamine than 10% or native

protamine at pH 7.0, and that the 20% modified

protamine also gave the lowest MIC. Earlier work

also demonstrated that increased binding of prot-

amine to the cell surfaces of L. monocytogenes and

E. coli in the pH range of 58 correlated with a

lowering of MICs measured in the same pH range

(Truelstrup Hansen and Gill, 2000).

Initially, it was hypothesized that electrostatic

interactions controlled the degree of binding of

protamine to cell surfaces and would be inversely

related to the degree of chemical modification at pH

levels where the cell surfaces retained a net negative

charge. However, this was clearly not the case, as

more modified than native protamine was bound to

Listeria cells at pHz7. Further work will be requiredto explain this phenomenon.

Chemical modification of up to 20% of prota-

mines arginine residues did not significantly

decrease degradation of protamine. This is perhaps

not surprising, since with 20% blocking there would

still be 16 unmodified arginine residues per molecule

on average. In fact, structural alterations may not be

an effective means of protecting CAPs from proteo-

lytic digestion because of the multitude of bacterial

proteases with varying specificities. At least two or

three such protamine degrading proteases have been

found (Obata et al., 1997; Stumpe et al., 1998;

Truelstrup Hansen and Gill, 2000). Other strategies

to protect the CAPs may be more practical. To give

examples, Siragusa et al. (1999) applied nisin as a

meat preservative in a polyethylene packaging

material; and Ruissen et al. (1999) utilized xanthan

gum to improve the delivery and efficacy of an

antifungal CAP in an oral rinse. A synthetic hybrid

of cecropin-melittin with in vitro activity against

Vibrio anguillarum, a salmon pathogen, was only

able inhibit the bacteria in infected fish if adminis-

tered continuously via an intraperitoneal miniosmotic

ood Microbiology 103 (2005) 2334 31pump (Jia et al., 2000). Therefore, design of smart

CAP delivery methods to ensure continuous release,

mg ml1 in various protein-rich food products

protamine.

In conclusion, moderately charge reduced prot-

l of Freduced exposure to tissue and bacterial proteases

and/or non-specific binding to anionic proteins may

be used to increase the range of applications for

these peptides.

A small charge reduction on protamine was

initially hypothesized to reduce the undesirable

interactions between the CAP and components of

the food matrix. The present work demonstrated that a

moderate charge reduction improved the efficacy of

(modified) protamine in high protein food systems

such as milk (pH 6.9) and ground beef (pH 5.7). The

combination of electrostatic forces as well as the small

increase in hydrophobicity are likely to be responsible

for the improved antimicrobial activity of the more

amphiphilic modified protamine analogues in food

matrices. It has been pointed out that both the

amphipathic nature of CAPs and the high positive

charge contribute to the lysis of the bacterial envelope

(Hwang and Vogel, 1998; Shai, 2002). For the more

amphipathic (chemically modified) protamine ana-

logues in the present study, there was no increase in

antimicrobial efficacy in artificial culture media while

the antibacterial efficacy increased in the tested food

matrices.

More work is needed to determine which factors in

complex foods reduce the antibacterial efficacy of

protamine and other CAPs. In milk the modified (14

and 23%) protamines were only slightly more

inhibitory to L. monocytogenes 19115 than native

protamine, but the chemically modified peptides

inhibited the total microflora as well as coliforms

significantly better than the native protamine in

ground beef. This difference may be related to the

different types of food matrices (e.g., liquid versus

particulate, pH and macro- and micro-nutrient com-

position). The protein levels in ground beef (21%;

Anonymous, 2003) are substantially higher than those

in milk (3.3%; Anonymous, 2003). However, the

lower pH of ground beef also decreases the negative

charge on proteins, thereby reducing the electrostatic

interactions between muscle proteins and the added

CAP. Also, the content of divalent metal ions known

to reduce the antimicrobial effect of protamine (Pink

et al., 2003), is higher in milk than in beef with Ca2+

being 1.11 mg g1 and 0.09 mg g1 and Mg2+ being0.11 mg g1 and 0.22 mg g1 in milk and beef,

R. Potter et al. / International Journa32respectively (Anonymous, 2003). To date, the appli-

cation of protamine to food preservation has beenamines inhibited L. monocytogenes in milk and Gram-

negative spoilage bacteria (Enterobacteriaceae) in

ground beef significantly better than native protamine,

whereas the antimicrobial efficacies of both types of

protamine in a TSB model system were similar.

Future work will focus on the mechanisms through

which protamine gains entry into the cell envelope

and on the design of appropriate systems for deliver-

ing protamine and other CAPs into food matrices.

Future work will also involve the construction of

protamine analogues with site specific amino acid

substitutions to improve the antimicrobial efficacy.

Acknowledgements

This work was funded by a Natural Sciences and

Engineering Research Council of Canada (NSERC)

discovery grant to author Gill as well as an NSERC

Multidisciplinary Network Grant on the structure and

function of food-related polymers of which Dr. Gill

was a co-applicant. Krista M. Terry and John W.

Thompson are thanked for their excellent technical

assistance.

References

Abachin, E., Poyart, C., Pellegrini, E., Milohanic, E., Fielder, F.,

Berche, P., Tireu-Cuot, P., 2002. Formation of d-alanyl-lip-including whiting (pH 6.9), chicken (pH 6) and

ground beef (pH 5.6) was found to be negligible with

respect to inhibition of the total microflora and

coliforms (Uyttendaele and Debevere, 1993). The

same researchers found that addition of 5 mg ml1

protamine to pasteurized whiting, curry sauce (pH

6.3) and ground beef retarded the growth of inocu-

lated L. monocytogenes and B. cereus in all products

except the ground beef, indicating an effect of pH and

the food matrix on the antimicrobial efficacy oflimited mainly to carbohydrate-rich foods (Ueno et

al., 1987), presumably because of lower levels of non-

specific binding of protamine in such foods. The

effect of native protamine added at levels of up to 5

ood Microbiology 103 (2005) 2334oteichoic acid is required for adhesion and virulence of Listeria

monocytogenes. Molecular Microbiology 43, 114.

R. Potter et al. / International Journal of Food Microbiology 103 (2005) 2334 33Anonymous, 2003. USDA National Nutrient Database Standard

Reference. Released 16, July 2003. www.nal.usda.gov/fnic/

cgi-bin/nut_search.pl (Accessed February 4, 2004).

Baker, C.N., Banerjee, S.N., Tenover, F.C., 1994. Evaluation of

Alamar colorimetric MIC method for antimicrobial suscepti-

bility testing of Gram-negative bacteria. Journal of Clinical

Microbiology 32, 12611267.

Brange, J., 1987. Galenics of Insulin. Springer-Verlag, Berlin,

pp. 3436. Chapter 3.

Brock, T.G., 1958. The effect of salmine on bacteria. Canadian

Journal of Microbiology 4, 6571.

Falla, T.J., Hancock, R.E.W., 1997. Improved activity of a synthetic

indolicidin analogue. Antimicrobial Agents and Chemotherapy

41, 771775.

Gomez, K.A., Gomez, A.A., 1984. Statistical Procedures for

Agricultural Research, 2nd ed. J. Wiley & Sons, Toronto, ON,

Canada.

Hwang, P.M., Vogel, H.J., 1998. Structurefunction relationships of

antimicrobial peptides. Biochemistry and Cell Biology 76,

235246.

Islam, N.M.D., Itakura, T., Motohiro, T., 1984. Antibacterial spectra

and minimum inhibition concentration of clupeine and salmine.

Bulletin of the Japanese Society of Scientific Fisheries 50,

17051708.

Islam, N., Itakura, T., Hakura, T., 1986. Inhibitory effect of prota-

mine on the growth from the spores of Bacillus species. Bulletin

of the Japanese Society of Scientific Fisheries 52, 913917.

Jaques, L.B., 1973. Protamineantagonist to heparin. Canadian

Medical Association Journal 108, 12911297.

Jia, X., Patrzykat, A., Devlin, R.H., Ackerman, P.A., Iwama, G.K.,

Hancock, R.E.W., 2000. Antimicrobial peptides protect coho

salmon from Vibrio anguillarum infections. Applied and

Environmental Microbiology 66, 19281932.

Johansen, C., Gill, T., Gram, L., 1995. Antibacterial effect of

protamine assayed by impedimetry. Journal of Applied Bacteri-

ology 78, 297303.

Johansen, C., Gill, T., Gram, L., 1996. Changes in the cell

morphology of Listeria monocytogenes and Shewanella putre-

faciens resulting from the action of protamine. Applied and

Environmental Microbiology 62, 10581064.

Johansen, C., Verheul, A., Gram, L., Gill, T., Abee, T., 1997.

Protamine-induced permeabilization of cell envelopes of Gram-

positive and Gram-negative bacteria. Applied and Environ-

mental Microbiology 63, 11551159.

Matsuzaki, K., Sugishita, K., Harada, M., Fujii, N., Miyajima, K.,

1997. Interactions of an antimicrobial peptide, magainin 2 with

outer and inner membranes of Gram-negative bacteria. Biochi-

mica et Biophysica Acta 1327, 119130.

Means, G.E., Feeney, R.E., 1971. Chemical Modification of

Proteins. Holden-Day, San Francisco. 254 pp.

Montville, T.J., Chen, Y., 1995. Efflux of ions and ATP depletion

induced by pediocin PA-1 are concomitant with cell death in

Listeria monocytogenes Scott A. Journal of Applied Micro-

biology 70, 684690.

Montville, T.J., Chen, Y., 1998. Mechanistic action of pediocin and

nisin: recent progress and unresolved questions. AppliedMicrobiology and Biotechnology 50, 511519.Neuhaus, F., Baddiley, J., 2003. A continuum of anionic charge:

structures and functions of d-alanyl-teichoic acid in Gram-

positive bacteria. Microbiology and Molecular Biology Reviews

67, 686723.

Nikaido, H., 2003. Molecular basis of bacterial outer membrane

permeability revisited. Microbiology and Molecular Biology

Reviews 67, 593656.

Obata, H., Sugiyama, A., Kawahara, H., Muramatsu, T., 1997.

Purification and properties of an aminopeptidase from a

protamine-degrading marine bacterium. Bioscience, Biotechnol-

ogy, and Biochemistry 61, 11021108.

Peschel, A., Otto, M., Jack, R.W., Kalbacher, H., Jung, G., Gotz, F.,

1999. Inactivation of the dlt operon in Staphylococcus aureus

confers sensitivity to defensins, protegrins and other antimicro-

bial peptides. Journal of Biological Chemistry 274, 84058410.

Pink, D.A., Truelstrup Hansen, L., Gill, T.A., Quinn, B.E.,

Beveridge, T.J., 2003. Divalent calcium ions interfere with the

penetration of protamine through the polysaccharide brushes of

the outer membrane of Gram negative bacteria. Langmuir 19,

88528858.

Rodman, T.C., Pruslin, F.H., Allfrey, V.G., 1984. Protamine-DNA

association in mammalian spermatazoa. Experimental Cell

Research 150, 269281.

Ruissen, A.L.A., van der Reijden, W.A., van tHof, W., Veerman,

E.C.I., Amerongen, A.V.N., 1999. Evaluation of the use of

xanthan as a vehicle for cationic antifungal peptides. Journal of

Controlled Release 60 (1), 4956.

Sakaguchi, S., 1950. A new method for the colorimetric determi-

nation of arginine. Journal of Biochemistry 37, 231236.

Shai, Y., 2002. Mode of action of membrane active antimicrobial

peptides. Biopolymers 66, 236248.

Sigma, 2004. http://www.sigma-origins.co.uk/pdfs/articles/

1074167259.pdf (Accessed February 25, 2004).

Siragusa, G.R., Cutter, C.N., Willett, J.L., 1999. Incorporation of

bacteriocin in plastic retains activity and inhibits surface growth

of bacteria on meat. Food Microbiology 16, 229235.

Stumpe, S., Schmid, R., Stephens, D.L., Georgiou, G., Bakker, E.P.,

1998. Identification of OmpT as the protease that hydrolyzes the

antimicrobial peptide protamine before it enters growing cells of

Escherichia coli. Journal of Bacteriology 180, 40024006.

Suzuki, K., Ando, T., 1972. Studies on protamine: XVII. The

complete amino acid sequence of clupeine YI. Journal of

Biochemistry 72, 14331446.

Truelstrup Hansen, L., Gill, T.A., 2000. Solubility and antimicrobial

efficacy of protamine on Listeria monocytogenes and Escher-

ichia coli as influenced by pH. Journal of Applied Microbiology

88, 10491055.

Truelstrup Hansen, L., Austin, J.W., Gill, T.A., 2001. Anti-

bacterial effect of protamine in combination with EDTA and

refrigeration. International Journal of Food Microbiology 66,

149161.

Ueno, R., Fujita, Y., Yamamoto, M., Kozakai, H., 1987. Multi-

plication inhibitor for Bacillus cereus. European Patent Appli-

cation 0273 606.

Ueno, R., Fujita, Y., Nagamura, Y., Kamino, Y., Tabata, A., 1989.

Method for preserving food. European Patent Application0372 091.

USP XXII, 1990. The United States pharmacopeia, twenty second

revision. The United States Pharmacopeial Convention Inc.,

Rockville, MD.

Uyttendaele, M., Debevere, J., 1993. Evaluatie van protamine als

conserveermiddel. Voedingsmiddeltechnologie 20, 5355.

Uyttendaele, M., Debevere, J., 1994. Evaluation of the antimicrobial

activity of protamine. Food Microbiology 11, 417427.

Vaara, M., Vaara, T., 1983. Polycations as outer membrane-

disorganizing agents. Antimicrobial Agents and Chemotherapy

24, 114122.

van der Mei, H., de Vries, J., Busscher, H.J., 1993. Hydrophobic

and electrostatic cell surface properties of thermophilic dairy

streptococci. Applied and Environmental Microbiology 59,

43054312.

R. Potter et al. / International Journal of Food Microbiology 103 (2005) 233434

Inhibition of foodborne bacteria by native and modified protamine: Importance of electrostatic interactionsIntroductionMaterials and methodsChemical modification of protamineDetermination of minimum inhibitory concentrations (MICs) or minimum bactericidal concentrations (MBCs) for protaminesBinding of native and charge reduced protamines to culture media componentsBinding of native and charge reduced protamines to resting bacterial cellsAntimicrobial effect of protamines in food systemsDetermination of electrophoretic mobilities of food-related bacteriaSensitivity of reduced charge protamines to trypsinStatistical analyses

ResultsChemical modification of protamineEffect of modification on bacterial inhibition in TSBBinding of protamines to culture media components and resting cellsAntimicrobial effect of modified protamines in food systemsEffect of bacterial surface charge on resistance/susceptibility to protamineEffect of modification on the tryptic digestablity of protamine

DiscussionAcknowledgementsReferences