The LP values for E. co/i 0157:H7 determined in the presence of aerobic mesophilic bacteria of apple cider stored at 23 C for 1611 was 1.48 h, while populations ...
FOODBORNE PATHOGENS AND DISEASE Volume 6, Number 4, 2009 © Mary Ann Liebert, Inc. 001: 10.1089/tpd.2008.0233
Growth Parameters of Escherichia coil 0157: H7, Salmonella spp., Listeria monocytogenes, and Aerobic Mesophilic Bacteria of Apple Cider Amended with Nisin—EDTA Dike 0. Ukuku, Howard Zhang, and Lihan Huang
Abstract The effect of nisin (0 or 300 IU/mL), ethylenediamine tetraacetic acid (EDTA, 20mM), and nisin (300 IU)—EDTA (20 mM) on growth parameters, including lag period (LP) and generation time, of Esclu'ricliia coli 0157:H7, Listerm mwwct/foçenes, and Salmonella spp. in the presence or absence of aerobic mesophilic bacteria of apple cider during storage at 5 C for up to 16 days or 23 C for 16h was investigated. The growth data were analyzed and fitted to the modified Gompertz model. The LP values for aerobic mesophilic bacteria of apple cider (control) and those amended with EDTA and nisin during storage at 5 C were 1.61, 1.76, and 5.45 days, respectively. In apple cider stored at 23 C for 16 h, the LP values for the same bacteria and treatment were 3.24, 3.56, and 5.85 h, respectively. The LP values for E. co/i 0157:H7 determined in the presence of aerobic mesophilic bacteria of apple cider stored at 23 C for 1611 was 1.48 h, while populations for L. monocyfogenes and Salmonella in the same cider declined. In sterile apple cider left at 23C for 16h, the LP values for E. coli 0157:1-17, Salmonella, and L. monocijtogenes averaged 2.74, 2.37, and 3.16 h, respectively. The generation time for these pathogens were 0.402, 0.260, and 0.187 log (CFU/mL)/h, respectively. Addition of nisin and EDTA combination caused a decline in lag phase duration and the populations for all pathogens tested, suggesting possible addition of this additive to freshly prepared apple cider to enhance its microbial safety and prevent costly recalls.
Introduction are not free from natcontaminants; they are in frequent contact with soil, insects, animals, and humans during growing and harvesting and in the processing plant (Murdock and Brokaw, 1957). By the time they reach the packinghouse, most fresh produce retains populations of 104 to i' microorganisms/g (Brackett, 1994; Beuchat, 1995). Microorganisms (bacteria, yeast, and mold) that contaminate juice or cider mostly conic from the surface of the fruits, especially if fruits were damaged, rotten, or decayed. Unpasteurized fruit juices have different classes of native microorganisms whose populations may vary significantly, depending on the type, origin, and surface structure of the fruits used in making the juice. Juice contamination with pathogens can also conic from contaminated water (Besser et al., 1993), human handling during processing, the facility, air, or equipment used. RUITS ' AND VEGETABLES' SURFACES
F ural
Reports of enteric diseases through consumption of u11pasteuri7ed apple cider contaminated with human pathogens have been reported (McLellan and Splittstoesser, 1996). The most frequently reported foodborne pathogen in apple cider is Escheric/ria co/i 0157:H7 (Madden, 1992; Besser i't al., 1993; CDC, 1996, 1997; Cody et al., 1999), as well as in orange juices (CDC, 1999). Recently, an outbreak of human listenosis was associated with the consumption of contaminated cider (Bock, 2008), and the potential outbreak in the future remains a concern. L'steria monocijtogenes is a food safety concern because it is widespread in the environment, can attach to a variety of surfaces (Cox et al., 1989; Mafu et al., 1990; Beuchat and Brackett, 1991; Carlin et al., 1996), grows under refrigerated conditions (Farber and Peterkin, 1991; Marston, 1995), and is a frequent resident in certain foodprocessing establishments (Brackett, 1988; Heisick et al., 1995). The presence of spoilage microflora native to food surfaces and any human bacterial pathogens is of practical
Eastern Regional Research Center, U.S. Department of Agriculture, Agricultural Research Service, Wyndmoor, Pennsylvania. Mention of trade names or commercial products in this article is solel y for the purpose of providing specific information and does not imply recommendation or endorsement b y the U.S. Department of Agriculture. 487
488
importance with regard to quality and food safety. Antagonism of native microflora of fruits and vegetables to the survival of human bacterial pathogens has been investigated (Francis and O'Beirne, 1998a, 1998b; Ukuku et al., 2004). In these studies, native microflora on fruit surfaces was found to he antagonistic to the survival of L. inonocytogem's population. Unpasteurized fresh juices are major commodities in the United States and can easily be contaminated with potential life-threatening foodborne pathogens. However, consumers' demand for product freshness and quality is driving the food industry in the search for nonthermal pasteurization methods, including the use of effective, generally recognized as safe (GRAS) antimicrobials that would not impact on the sensorial quality of the juice. There are several reports that nisin used in combination with a chelating agent exhibits a bactericidal effect toward both Gram-positive and Gram-negative bacteria (Blackburn et al., 1989; Stevens et al., 1991, 1992a, 1992b; Cutter and Siragusa, 1995). Nisin is a pentacyclic heterodetic subtype A lantibiotic peptide synthesized by Lee! acoccus lactis subsp. lactis (Gross and Morell, 1971; Jung, 1991; Shiba at al., 1991). When used alone, nisin is an effective inhibitor of Gram-positive bacteria (Benkerroum and Sandine, 1988; Harris et al., 1991; Ray, 1992; Ukuku and Shelef, 1997) and bacterial spores (Harris et al., 1991; Ray, 1992). In the United States, nisin has received GRAS status and is approved for use in some processed cheese spreads to prevent the outgrowth of clostridial spores and toxin production (Benkerroum and Sandine, 1988; Okereke and Montville, 1992). The data from these studies reported inactivation of bacteria in different media by nisin. However, the effect of nisin treatment on bacterial growth parameters was not reported. Survival of L. monoci/togenes Scott A in buffer system was determined by a modified version of Gompertz equation (Linton at al., 1995). Similarly, kinetics of growth and inhibition of L. m000ci/toçenes in the presence of antioxidant food additives in broth has been reported (Yousef etal., 1991). In the later study, growth curves fitted by different logistic models (Garrett et al., 1978; Gibson et al., 1987; Bahk et al., 1990) were used to determine growth parameters. In our current study, we used the modified version of Gompertz equation model as defined by Gibson etal. (1987) to estimate growth parameters for aerobic mesophilic bacteria of apple cider amended with or without nisin, eth ylenediarnine tetraacetic acid (EDTA), and a combination of the two additives. Also, growth parameters of E. co/i 0157:H7, L. nionocl/togenes, and Salmonella spp. inoculated in apple cider with the presence of or absence of native microflora were investigated. Finally, the effects of nisin, EDTA, and nisin-EDTA combination on growth parameters of E. co/i 0157:H7, L. nlonoci/togenes, and Salmonella spp. inoculated in sterile and nonsterile apple cider were also monitored. In all these studies, the effects of storage temperatures (5 C for 16 days or room temperature [approximately 23 C] for up to 16h) on growth parameters, including lag period (LP) and growth rate, were investigated. The data generated from this study may help the regulatory agencies in its decision process for estimating behavioral pattern for these pathogens in apple cider, which may help apple cider production facilities on the benefits of using nisin-EDTA combination to enhance the microbial safety of freshly prepared unpasteurized apple cider.
UKUKU ET AL.
Materials and Methods
Bacterial strains and inoculum preparation A mixed bacterial cocktail containing two strains of E. call 0157:H7, strain SEAI31388 and a strain involved in Oklahoma juice outbreaks; L. monoci/togene.c F8027 (Serntype 4b) and F8385 (Serotype 1/2b) received from Dr. Larry Beuchat, University of Georgia; and Salmonella Stanley H0558 and Salmonella Newport 111275 (all associated with alfalfa sproutrelated outbreaks, obtained from Dr. Patricia Griffin, CDC) were used in this study. Bacterial strains were prepared by two successive loop transfers of individual strains incubated at 36 C for 18 h in 5 mL Brain Heart Infusion Broth (BBL/Difco, Sparks, MD) for E. co/i 0157:H7and Salmonella spp. and Trypticase Soy Broth (BBI./Difco) supplemented with 0.69% yeast extract (BBL/Difco) (TSBY) for L. monocit togenes. A final transfer of 0.2 mL was made into 20 mL Brain Heart Infusion Broth or TSBY with incubation at 36 C for 18 under static conditions. The bacterial cells were harvested by centrifugation (3000g. 5 min) at 4 C. The cell pellets were washed twice in 0.1% peptone water (BBL/Difco), and the cell pellets were re-suspended in 20 mL of 0.1% peptone water to make the inoculum cocktail. The final bacterial concentration for each pathogen in the inoculum suspension averaged 107 CFU/mL, and 0.1 ml, (approximately 3.62 x 10'CFU/mL) of the inoculum was added to all cider samples (10 m[. each). Preparation of antimicrobial solutions and treatments A stock solution of nisin (10' IU; Sigma, St. Louis, MO) was prepared at a concentration of 2500 pg/mL (10' lU) ill N HCI. The stock solution was sterilized by filtration through 0.22 tm filters (Millipore, Bedford, MA). Similarly, a stock solution of 2 M disodium EDTA (Fisher Scientific, Pittsburgh, PA) was prepared in distilled deionized water (ddH2O), autoclaved at 121 C for 15 mm, and then stored at room temperature until used. All test solutions were prepared by dilution with sterile ddH 2O as required. Concentrations of the chemicals in the solutions used alone or in combination were EDTA, 20mM; nisin, 3001U/mL; and nisin-EDTA, 3001U/ mL and 20 mM, respectively. The pH of the apple cider tested averaged 3.6 ± 0.2 (Orion; Fisher Scientific). Microbiological analyses Apple cider containing inoculated bacteria and varying concentrations of antimicrobial agents as previously stated was stored at 5 C for up to 16 days and at room temperature (approximately 23C) for up to 16h. Periodically, 0.1 mL sample was taken out every 2 h from the sample stored at 23 C and every 3 days from the samples at 5 C, and were plated on plate count agar and a range of selective media or were serially diluted (samples >8 h) before plating to monitor survival and growth parameters of all pathogens. Selective agar media and Bismuth Sulfite agar (CM201; BBL/Difco) were used for Salmonella, and Cefixume Tellurite Sorhitol MacConkey agar (BBL/Difco) for E. coli 0157:117 with incubation at 37C for 48h. The colonies determined were confirmed according to the FDA Bacteriological Analytical Manual followed by conventional biochemical methods for each pathogen. For L. mouocytogenes, Listerin identification
INACTIVATION OF HUMAN BACTERIAL PATHOGENS IN APPLE CIDER BY NISIN—EDTA agar (l'ALCAM; Sigma) containing L/SIL'iia selective supplement (L-4660; Sigma) was used with incubation at 37 C for 48 ii (Lovett and Flitchins, 1988).
489
A9
Data analysis
06 U-
The parameters (A, B, It, and D) in the modified Gompertz model were determined by nonlinear regression procedure (Hintze, 1999). Plate counts for all bacteria were converted to log values and then fitted to this model. The modified Gompertz model as described by Gibson ci al. (1987) was used as a primary model to describe the growth parameters of aerobic bacteria, E. co/i 01 57:H7, Salmonella, and L. Inoiloci/toge;les in apple cider amended with or without nisin, EDTA, or a combination of the two additives. The modified Cornpertz model as described by Gibson ci al. (1987) is shown in Equation I below. log 10 (CFIJmL) = A+(B -A) x EXP{ -EXP(--Cx (Cl - D))} (1) where variables A, B, C, and Dare the model parameters. A is the initial CFU/mL of aerobic bacteria, E. co/i 01571-1:7, Salinane/la, and L. i0000ci/togencs or the asymptotic log ]() count as time approaches to zero in the model; B is the maximum CFLJ/mL in the asymptotic log i count as time approaches to infinity; the parameter C is It and is the relative generation time (GT) at time D point. Parameter D is the inflex point of the curve [hour (Ii) or day (d)] where absolute growth rate is at maximum. The experimental growth data were fitted to the modified Gornpertz model using NCSS 2000, a Windowbased statistical package (Hintze, 1999). A secondary growth parameters [the lag phase duration () and the specific growth rate (K)] for each growth curve were derived under isothermal condition as described by Huang (2003a, 2003h) using the following Equations 2 and 3.
0
0
0
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 Time (Day) -#-Control Nisin -. EDTA
B E U-
0 0 CM 0
0
J 0. 0
a.
o in in in N- ,C 0
C'O
in in c in in in cc coo — Time (h) —Control Nisin EDTA
CO
in in in in
co
in
LO -C'J
n in in in in in - -
CN i NC\J-
K = (B - A)1t/2.718(2)
FIG. 1. Growth kinetics of mesophilic bacteria in apple
(3)
cider containing nisin (0 and 3001U/mT..) and ethylenediamine tetraacetic acid (20n-LM) during storage at 5 C for 14 days (A) and 23 C for 16h (B).
All experiments were performed in triplicate with duplicate samples analyzed at each sampling time. Data were analyzed by anal sis of variance (ANOVA) to determine statistically significant differences among the treatments. Bonferroni mean separation test was used to determine significant differences (p < 0.05) among means (Miller, 1981). y
Results and Discussion Growth kinetics of aerobic mesophi/ic bacteria of apple cider
The population of aerobic mesophilic bacteria determined from fresh apple cider averaged 3.9 log 1 CFU/mL. The growth curves of aerobic mesophilic bacteria in apple cider amended with nisin and EDTA were sigmoidal during storage at 5 C for 14 days (Fig. 1A) and at room temperature (23 C) for 16h (Fig. 113), and the data were accurately fitted to the modified Gompertz model (Gibson ci al., 1987). The lag phase and the growth rate (ji) determined using this model are shown in Table 1. The maximLim cell population determined in apple cider stored at 5 C for 14 days or left at 23 C for 16 h averaged 7.01 = 1.16log 10 CFU/mL and 7.53 1 1.23 logio
CFU/mL, respectively. Similarly, these populations in apple cider amended with EDTA and nisin averaged 6.43 ± 0.14log Ic) CFU/mL and 5.73±0.23log ii CFU/mL, respectively, and were lower than the numbers determined in control cider. There was an initial decline in cell population of aerobic mesophilic bacteria of apple cider amended with nisin-EDTA during the lag phase (Fig. 2A, B). The populations of aerobic mesophilic bacteria in samples treated with nisin—EDTA declined from 3.9logo CFU/mL to 1.95log1ç> CFU/mL at day 2 of storage at 5 C (Fig. 2A), and 1.62logi CFU/mL at 2h of storage at 23 C (Fig. 213). The effect of nisin (3001U), EDTA (20mM), or nisin (300 ILJ)-EDTA (20 mM) treatment affected growth parameters of the aerobic mesophilic bacteria in apple cider during storage at 5 C for up to 16 days (Table 1). The growth parameters for aerobic mesophilic bacteria, including LP and CT, averaged 1.61 and 0.702 days, respectively. Apple cider containing EDTA alone had a slightly higher LI' (1.76 days) and a lower GT (0.515 da s). However, in cider amended with nisin alone, the LP and CT determined were 5.45 and 0.956 y
490
UKUKU ET AL.
TABLE 1. GROWTH PARAMETERS OF NATIVE MICROFLORA OF APPLE CIDER AMENDED wi-rii NIsIN, EDTA, AND NIsIN-EDTA
Generation film? (day)
Maxim urn growth (logy, CFU/mL)
Additives
Lag period (day)
At 5 C 0 0.02 M EDTA 3001U Nisin Nisin-EDTA
1.61 + 0.52 0.702 10.153 1.76+0.35 0.515 ± 0.134 5.45 ± 0.32 0.956 ± 0.121 ND ND
7.01 ± 1.16 6.43--0.14 5.73--0.23 ND
At 23 C 0 0.02 M EDTA 300 IU Nisin Nisin-EDTA
3.24 ± 0.22 0.677+ 0.124 3.56 ± 0.35 0.800 + 0.127 5.85 ± 0.38 1.755 ± 0.115 ND ND
7.53-1.23 6.40±0.18 6.39--0.29 ND
Values represent mean SD of three studies with duplicate determination. EDTA, ethylenediamine tetraacetic acid; ND, not determined as growth was not evident, but pathogens survived.
determined were significantly (p