1197
Journal o f Food Protection, Vol. 78, No. 6, 2015, Pages 1197-1202 doi: 10.4315/0362-028X.JFP-14-394 Copyright © , International Association for Food Protection
Research Note
Investigation into Formation of Lipid Hydroperoxides from Membrane Lipids in Escherichia coli 0 1 57:H7 following Exposure to Hot Water THELMA F. CALIX-LARA,'f t . KATIE R. KIRSCH,+ MARGARET D. HARDIN,2§ ALEJANDRO CASTILLO,STEPHEN B. SMITH , 2 a n d THOMAS M. TAYLOR2**§ 1Department o f Nutrition and Food Science, Texas A&M University, College Station, Texas 77843-2253; and d epartm ent o f Animal Science, Texas A&M AgriLife Research, College Station, Texas 77843-2471, USA MS 14-394: Received 20 August 2014/Accepted 19 February 2015
ABSTRACT Although studies have shown antimicrobial treatments consisting of hot water sprays alone or paired with lactic acid rinses are effective for reducing Escherichia coli 0157:H7 loads on beef carcass surfaces, the mechanisms by which these interventions inactivate bacterial pathogens are still poorly understood. It was hypothesized that E. coli 0157:H7 exposure to hot water in vitro at rising temperatures for longer time periods would result in increasing deterioration of bacterial outer membrane lipids, sensitizing the pathogen to subsequent lactic acid application. Cocktails of E. coli 0 157:H7 strains were subjected to hot water at 25 (control) 65, 75, or 85°C incrementally up to 60 s, after which surviving cells were enumerated by plating. Formation of lipid hydroperoxides from bacterial membranes and cytoplasmic accumulation of L-lactic acid was quantified spectrophotometrically. Inactivation of E. coli 0157:H7 proceeded in a hot water exposure duration- and temperature-dependent manner, with populations being reduced to nondetectable numbers following heating of cells in 85°C water for 30 and 60 s (P < 0.05). Lipid hydroperoxide formation was not observed to be dependent upon increasing water temperature or exposure period. The data suggest that hot water application prior to organic acid application may function to increase the sensitivity of E. coli 0157:H7 cells by degrading membrane lipids.
In the United States, an estimated 175,905 illnesses attributed to foodborne Shiga toxin-producing Escherichia coli, including E. coli 0157:H7, are estimated to occur annually (28). Fresh beef is most frequently implicated in foodborne disease outbreaks associated with these patho gens (15). Beef carcasses may become cross-contaminated during slaughter and dressing. Hot water carcass sprays applied alone or in combination with organic acid application have been studied extensively for carcass decontamination (1, 4, 8, 12, 13). Water temperature can impact the observed antimicrobial efficacy of a carcass water spray. Ambient water washes, consisting of a 1.5-liter manual water wash (25 °C and 10 lb/in2 for 90 s) and a subsequent 5.0-liter automated cabinet water wash (35°C and 250 to 400 lb/in2 for 9 s), were reported to reduce E. coli 0157:H7 populations on beef carcass surfaces by 1.7 to 2.7 log CFU/cm2. The application of a postwash hot water spray (95°C and 24 lb/in2 for 5 s) was reported to provide an additional 0.8 to 2.2 log cycle of E. coli 0157:H7 reduction * Author for correspondence. Tel: 979-862-7678; Fax: 979-845-4992; E-mail: [email protected]. t Authors contributed equally to manuscript completion, t No current professional affiliation. § Present address: IEH Laboratories and Consulting Group, 15300 Bothell Way N.E., Lake Forest Park, WA 98155, USA.
in addition to that achieved by water washing alone (9, 10). Researchers evaluating the efficacy of hot water sprays on beef carcasses reported reductions in aerobic bacteria of 2.2 log CFU/cm2 (80°C and 30 lb/in2 for 5.6 s) and 1.3 log CFU/cm2 (95°C spray for 40 s), reductions in E. coli of 1.8 log CFU/cm2 (80°C and 30 lb/in2 for 5.6 s) (17) or 4.2 log CFU/cm2 (74°C and 300.2 lb/in2 for 12 s) (6), and in total coliforms of 1.4 log CFU/cm2 (77°C and 20 to 22 lb/in2 for 2.5 s) (16). Variability among observed microbial reductions between studies was reportedly likely due to differences in experimental design, equipment, and beef cut selection. Numerous factors, including the sequence of hurdle application, can influence the efficacy of antimicrobial hurdle technologies for pathogen inhibition on carcass surfaces. Castillo et al. (8) reported 4.9 log CFU/cm2 reductions of E. coli 0157:H7 inoculated on beef carcasses, following a water wash plus a hot water spray (95 °C and 24 lb/in2 for 5 s) and 2% L-lactic acid warmed to 55°C. However, when the sequence of interventions was reversed (lactic acid followed by hot water), E. coli 0157:H7 counts were 0.5 log cycle lower than those subjected to hot water plus lactic acid (8). Bosilevac et al. (4) described reductions in aerobic plate counts of 2.2 log CFU/100 cm2 following the sequential application of hot water (74°C and 700 lb/in2 for 5.5 s) and 2% L-lactic acid (42°C, pH 2.4) on preevisceration carcasses. Observed reductions were
1198
CALIX-LARA ET AL.
lower than those achieved using only hot water (2.7 log CFU/100 cm2; P > 0.05); the authors suggested that observed differences were likely the result of cooling effects of the acid solution on the carcass surface. Although thermal treatment has been shown to inactivate pathogenic microorganisms on foods, the mechanisms responsible are not fully elucidated. Bruskov et al. (5) suggested exposure to heat may activate dissolved atmo spheric oxygen, initiating the production of reactive oxygen species. The authors proposed that thermally mediated damage to DNA was due to the action of these newly formed reactive oxygen species. Reactive oxygen species, including superoxide anions (0 2 ), hydrogen peroxide (H20 2), and hydroxyl radicals (—OH), can cause oxidative damage to cellular components, including lipids, proteins, and nucleic acids (7,11). Oxidative stress has been cited as a key factor in the heat-induced death of Saccharomyces cerevisiae, evi denced by a 500 to >20,000-fold increase in heat tolerance of cells in anoxic versus aerobic conditions (14). The relationship between thermal stress and intracellular lipid peroxidation has also been demonstrated in prokaryotes. Bacillus subtilis cells exposed to heat stress (55 °C for 1 h) were found to have 314% greater levels of intracellular superoxide than those grown at 37°C (27). This study was conducted to determine the impact of hot water exposure at increasing temperature for longer exposure periods on the survival of E. coli 0157:H7 and the integrity of outer membrane lipids prior to the application of 5.0% L-lactic acid. Researchers hypothesized that in vitro exposure of E. coli 0157:H7 cells to water heated to increasing temperatures simulating beef carcass interventions would result in increased formation of primary lipid oxidation products from membrane lipids (namely, hydroperoxides) and provide increased internalization of L-lactic acid into the cytoplasm of cells. MATERIALS AND METHODS Bacterial cultures and revival procedures. E. coli 0157:H7 strains P8, P I8, and P41 (bovine carcass isolates) and Environ mental Diseases Laboratory 932 (human fecal isolate) were obtained from the Food Microbiology Laboratory culture collec tion in the Department of Animal Science (Texas A&M University, College Station). Each strain was individually resuscitated via duplicate overnight transfer and subsequent static incubation in tryptic soy broth (TSB; BD, Sparks, MD) at 35°C. Once revived, isolated colonies were obtained by streaking sorbitol MacConkey agar (BD); plates were incubated for 24 h at 35°C. Isolated colonies displaying correct phenotype (zone of precipitated bile salts without fermentation of sorbitol) were aseptically picked and transferred to TSB for incubation and preparation of tryptic soy agar (TSA; BD) slants. Bacterial cocktail preparation. Five milliliters of overnight cultures (~9.0 log CFU/ml) of E. coli 0157:H7 strains were aseptically dispensed in sterile 15-ml centrifuge tubes (Thermo Fisher Scientific, Inc., Pittsburg, PA) and centrifuged at 1,623 x g in a Jouan B4i centrifuge (Thermo Electron Corp., Madison, WI) for 15 min at 25°C. The resulting pellet for each strain was suspended in 5.0 ml of 0.1% peptone diluent, following aseptic removal of supernatant and washed twice by centrifugation under the same conditions as mentioned. Immediately following the third centrifu gation, each pellet was suspended in 5.0 ml of 0.1% peptone diluent.
J. Food Prot., Vol. 78, No. 6
Finally, 2.5 ml of each suspension were aseptically transferred and pooled in a flask containing 90.0 ml of sterile 0.1% peptone diluent to create a cocktail of —8.0 log CFU/ml. E. coli 0157:H7 cocktail exposure to hot water. E. coli 0157:H7 cocktail aliquots of 1.5 ml each were aseptically dispensed into sterile disposable borosilicate glass tubes (10 x 75 mm) and submerged in a flowing water bath set to heat cell suspensions to 25, 65, 75, or 85°C. The water bath contained sufficient water to cover about 1.8 ml of volume of tubes; bath temperature was constantly monitored by a submerged thermom eter. Timers were initiated once samples equilibrated to the desired temperature. Water temperature inside a test tube filled to 1.5 ml of peptone diluent, but without pathogen cells, was monitored by using a type K thermocouple (model 108A 800008, Sper Scientific Ltd., Scottsdale, AZ) and a four-channel data-logging meter (model 108A 800024; Sper Scientific Ltd.). Cell suspensions were exposed for 0, 15, 30, or 60 s; individual tubes were prepared to allow for each tube to be subjected to a single heating period and then be sampled. After hot water exposure, sample tubes were immediately submerged in ice-chilled water (0°C) for —20 s to arrest heat-induced bacterial inactivation and inhibit further membrane degradation. Surviving cells were quantified by preparation of decimal dilutions in 0.1% (wt/vol) peptone diluent and spread on TSA-containing petri dishes, followed by incubation for 48 h at 35°C. Colonies were then counted and recorded. Membrane lipid extraction from heated E. coli 0157:H7 cells. Methyl-tert-butyl ether lipid extraction method was per formed according to the procedure reported by Matyash et al. (22). After hot water treatment, 200 pi of each sample tube containing pathogen cocktail (—1.5 x 107 CFU) was placed into a glass centrifuge tube, twice centrifuged for 15 min at 1,623 x g, and washed with 0.1% ammonium acetate (Sigma-Aldrich Co., St. Louis, MO). Next, 1.5 ml of absolute methanol (Thermo Fisher Scientific, Inc.) was added and vortexed for 4 s. Then, 5.0 ml of methyl-tert-butyl ether was added, and the samples were incubated for 60 min at 25°C, with gentle shaking (Mistral 4500 MultiMixer, Colonial Scientific, Inc., Richmond, VA). After incubation, 1.25 ml of sterile distilled water was added to stimulate phase separation, followed by 10 min of incubation, and then a 10 min of centrifugation at 1,000 x g. Resulting upper phases were collected separately, dried under compressed nitrogen, and stored overnight in a desiccator at 4°C before analysis. Quantification of lipid hydroperoxides from E. coli 0157:H7 membrane lipids. Following lipid extraction, the formation of lipid hydroperoxides after hot water exposure was assayed by the ferric oxide method, as described previously (29). A linear standard curve was initially completed to provide for quantification of lipid peroxide content from hot water-treated samples (Fig. 1). Thereafter, standard curves were always completed prior to the assay of experimental replicates. Desiccated lipid from each 200-pl sample was reconstituted with 4.8 ml of solution of chloroform-methanol 7:3 (Acres Organics, Waltham, MA); within 10 min, 50.0 pi of xylenol orange (10 mM) and 50 pi of solution of ferrous (Fe2+ ) chloride, prepared as instructed by Shantha and Decker (29), were added. Unless indicated otherwise, all experimental reagents were purchased from Sigma-Aldrich Co. and were of highest purity. Samples were held in the dark at room temperature for 5 min prior to assay. Absorbance was read at 560 nm against a blank with a DU-7400 UV/Vis Spectrophotom eter (Beckman Coulter, Fullerton. CA). Milliequivalents (meq) of lipid peroxide per kilogram of sample were calculated as
J. Food Prot., Vol. 78. No. 6
E. C0L1 0157:H7 MEMBRANE OXIDATION BY HOT WATER
FIGURE 1. Standard curve o f ferric oxide detection o f lipid hydroperoxide formation from E. coli 0157:H7 membrane lipids. Standard curves were prepared new fo r each independent replication; symbols depict means from triplicate identical replicates. Limit o f detection for assay: 2.0 meq. Absorbance was read at 560 nm. Peroxide value (PV) = (As - A h) x M /55.84 x M„ x 2
(1)
where A,, and Ah are the absorbance of the sample and blank at 560 nm, respectively, M is slope of the standard curve as determined by regression analysis, 55.84 (atomic weight of iron), and M0 is the sample mass (g). The denominator was multiplied by a conversion factor (2) to allow for expression of peroxide value in meq of lipid hydroperoxides. L-Lactic acid accumulation by E. coli 0157:H7 following hot water exposure. E. coli 0157:H7 cells, previously subjected to hot water, were aseptically transferred (1.0 ml) to a sterile microcentrifuge tube, and 1.0 ml (wt/wt) of L-lactic acid (FCC 88, Purac America, Lincolnshire, IL) was added to give a final concentration of 5.0% (wt/wt) L-lactic acid. Acid concentration was calculated assuming bacterial suspensions possessed a density of —1.0 g/ml. Tubes were agitated by vortexing and allowed to incubate at room temperature for 5 min. Cells were removed from the acid by centrifugation and diluted in 1.0 ml of 0.1% (wt/vol) peptone diluent. Surviving E. coli 0157:H7 cells were identified by plating on surfaces of TSA-containing petri dishes, followed by incubation of inoculated plates for 48 h at 35°C prior to colony enumeration. Samples (0.5 ml) of hot water and lactic acid-treated E. coli 0157:H7 were loaded into microcentrifuge tubes containing 0.2 ml of 6 N perchloric acid (Sigma-Aldrich Co.) and 0.5 ml of a silicone oil blend. Perchloric acid was used to arrest bacterial metabolism and allow for quantification of accumulated lactic acid. Blended silicone oil (p = 1.055 g/ml) was layered on top of perchloric acid and allowed for the selective transmission of bacterial cells from the top phase into perchloric acid, once subjected to centrifugation. Silicone oils were blended at a ratio of 3:1 (vol/vol) AR 20 (p =
1199
1.008 g/ml; Fluka, Seelze, Germany): high-temperature silicone oil (p = 1.102 g/ml; Sigma-Aldrich, Inc.). In a separate glass tube (10 by 75 mm), a mixture of 0.5 ml of sample and 0.2 ml of either 0.1% peptone water or L-lactic acid was formulated; L-lactic acid was prepared from concentrate (FCC 88, Purac America) to deliver a final concentration of 5% (wt/wt). The mixture was added as the top layer (third layer) in the microcentrifuge tube. Lactic acid exposure to the microbial cells was approximately 5 min and was stopped by centrifugation (model 59V, Lab-Line Instruments, Melrose Park, IL) at maximum velocity for 1 min or until a microbial pellet was noticed at the bottom. The upper and lower phases were aseptically collected and stored under refrigeration (5°C) until analysis was conducted; the oil phase was discarded. The estimation of lactic acid accumulation was completed by using previously reported methods (3, 18). L-Lactate dehydrogenase (643 U/mg) was used to catalyze the conversion of lactate to NADFI to allow for determination of lactic acid content accumulation. A chemical trap (hydrazine monohydrate) was added to disallow the enzyme from catalyzing the conversion of NADH to L-lactic acid. The trap consisted of 1.25 ml of hydrazine monohydrate (0.4 M) glycine buffer (0.5 M, pH 9.0). Potassium carbonate (K2C 0 3; 20 pi, 5.0 M) was added to buffer perchloric acid used previously to arrest microbial metabolism (18). To the reaction vessel was added 0.1 ml of NAD solution (2.75 mM) and 0.1 ml of sample. AH reagents were purchased from Sigma-Aldrich Co. and were of highest purity. Two microliters of enzyme (undiluted) was added, and the catalysis of NADH allowed to proceed at room temperature for 60 min, after which the absorbance of NADH was monitored at 339 nm with a DU-7400 UV/Vis Spectrophotometer (Beckman Coulter). Samples of acid-exposed and peptone-exposed (control) cells were tested identically to account for residual lactate produced during bacterial fermentation of glucose in microbiological medium during over night growth. Nonaccumulated acid was determined by the difference of total acid exposure minus baseline-adjusted cytoplas mic lactic acid. L-Lactic acid concentration in E. coli 0157:H7 was determined as [L-lactic acid] = A339 nm x 0.344
(2)
where A339nm is the absorbance of NADH at 339 nm and 0.344 is a conversion factor providing the concentration of i.-lactate obtained in the sample (nanomole per milliliter). Experimental design and statistical analyses. Plate counts of surviving E. coli 0157:H7 were log transformed prior to analysis of differences in E. coli 0157:H7 numbers or lipid hydroperoxides from membrane lipids as a function of heating temperature, exposure period, and their interaction (temperature x exposure period). Statistical differences between means were obtained by one-way analysis of variance (P < 0.05) and Tukey’s honestly significant difference test (P < 0.05) using the generalized linear model procedure of SAS, version 9.0 (SAS Institute, Inc., Cary, NC).
RESULTS AND DISCUSSION The inactivation of E. coli 0157:H7 as a result of exposure to heated water for increasing periods of time is shown in Table I. No reductions in E. coli 0157:H7 populations were observed for control samples incubated at 25°C. Inactivation of E. coli 0157:H7 by hot water increased in a temperature-dependent and exposure perioddependent manner, with significant reductions (P < 0.05) achieved after 15 s of hot water exposure (Table 1). Maximum reduction in pathogen numbers was achieved at 60 s of heating as compared with nonheated cells (Table 1).
1200
CALIX-LARA ET AL.
J. Food Prot., Vol. 78, No. 6
TABLE 1. Least-square means o f the interaction o f water temperature by exposure period on the survival ofE. coli 0157:H7a Heating period (s)
TABLE 2. Least-square means o f lipid peroxides from E. coli
0157:H7 cells following exposure to heated watera
Mean E. coli 0157:H7 survival (log CFU/ml) — ------------------------------------------------------------------25°C 65°C 75°C 85°C
Mean (meq of peroxide/kg) Heating period (s)
0 5 15 30 60
8.1 8.1 8.2 8.1 8.1
A A A A A
8.1 6.5 4.9 3.9 3.8
A AB bc
BCD cd
8.1 3.4 2.8 2.7 2.2
A cd
CDE CDEF DEF
8.1 A ND* ND ND ND
a P < 0.0001. Values presented depict means from triplicate identical replications (n = 3). Values not sharing capital letters (a through f) differ at P < 0.05, assessed by one-way analysis of variance and Tukey’s honestly significant difference test. Standard error of the mean = 0.29. * ND, nondetectable; limit of detection: 1.0 CFU/ml.
E. coli populations exposed to water set to 75°C were rapidly reduced (P < 0.05) after 5 s of treatment, with no further reductions observed following 15, 30, and 60 s of heat exposure. Significant reductions (P < 0.05) in pathogen numbers were achieved within 5 s of incubation at 85°C, with numbers decreased to below the limit of detection thereafter. Resulting trends in reductions mirror those observed previously, when hot water application for increasing lengths of time resulted in statistically greater inactivation of carcass-adhered pathogens (31). Factors intrinsic to meat, including pH, protein and fat content, and water activity, can greatly influence microbial thermal resistance (25, 33). The present thermal inactivation rates are similar to those reported in both nutrient broth (25) and deionized water (32), while the bacterial inactivation in meat matrices is generally less rapid (20, 24). Such differences are likely due to the presence of muscle and fat tissue, which exert an insulatory effect, protecting bacterial contaminants from thermal inactivation (9). Bovine tissue can also act as a buffer and may reduce the efficacy of organic acids against pathogens (2, 26). Though information regarding the mechanisms and rates of heat-induced bacterial inactivation are gained from the present study, the practical application of these data for beef producers are limited by the simplicity of the experimental system. The absence of muscle and lipid in the experimental sample, however, was required in this study to allow for direct assay of bacterial lipid deteriora tion, as beef fat would have compromised the researchers’ ability to quantify pathogen membrane lipid degradation. Degradation of E. coli 0157:H7 outer membrane lipids into lipid hydroperoxides was generally independent of hot water temperature and exposure duration (Table 2). The lack of statistical differences between samples incubated for increasing lengths of time may have resulted from the reactivity of hydroperoxides and their degradation into products not directly assayed by the experimental method used in this study, or by lack of sufficient lipids to produce degradation products in necessary quantity to be detected by the methods described. However, meq per kilogram of hydroperoxides detected by the ferric oxide method, described as being sensitive to peroxide value levels in this
0 5 15 30 60
P values
0.003
0.23
0.008 0.013 0.020 0.024 0.008
Temp (°C) 25 65 75 85
Pooled SEM
0.24 0.008 0.026 0.026 0.014
a Values presented depict means from triplicate identical replica tions (n = 3). P values assessed by one-way analysis of variance.
study, were within the linear and detectable range of the assay (Fig. 1). Nonetheless, despite the likely degradation of membrane-associated fatty acids, assay of bacterial lipid deterioration following the application of a simulated beef carcass antimicrobial hurdle process has not been previously reported to the authors’ knowledge. While others have sought to determine the impact of antimicrobial interventions on bacterial membrane integrity, past research has typically focused on the shuffling of membrane lipids by the microbe in response to the antimicrobial, rather than identify deterioration in membrane lipids (21,23). Studies reporting the degradative mechanisms of thermal processing on bacterial membranes are generally lacking in the scientific literature. Jang and Rhee (19), in studies of the effects of heating on caprylic acid inhibition of Cronobacter sakazakii in reconstituted infant formula, reported increasing heating temperature and time resulted in enhanced damage to the integrity of the bacterial membranes, as determined by differential staining. Smirnova et al. (30) reported increased superoxide formation and intracellular H20 2 in E. coli grown at 42 versus 37°C, though these reactive oxygen species can be formed from nonlipid precursors, as well as from lipid hydroperoxide-incorporating reactions. Collectively, data reported here and previous reports indicate the potential for hot water to sensitize bacterial microbes to the effects of an antimicrobial by disrupting the integrity of the membrane. Following hot water exposure, cells were exposed to 5% (wt/wt) L-lactic acid for 5 min prior to enumeration. For all treatments, numbers of E. coli 0157:H7 were reduced to nondetectable (
Journal o f Food Protection, Vol. 78, No. 6, 2015, Pages 1197-1202 doi: 10.4315/0362-028X.JFP-14-394 Copyright © , International Association for Food Protection
Research Note
Investigation into Formation of Lipid Hydroperoxides from Membrane Lipids in Escherichia coli 0 1 57:H7 following Exposure to Hot Water THELMA F. CALIX-LARA,'f t . KATIE R. KIRSCH,+ MARGARET D. HARDIN,2§ ALEJANDRO CASTILLO,STEPHEN B. SMITH , 2 a n d THOMAS M. TAYLOR2**§ 1Department o f Nutrition and Food Science, Texas A&M University, College Station, Texas 77843-2253; and d epartm ent o f Animal Science, Texas A&M AgriLife Research, College Station, Texas 77843-2471, USA MS 14-394: Received 20 August 2014/Accepted 19 February 2015
ABSTRACT Although studies have shown antimicrobial treatments consisting of hot water sprays alone or paired with lactic acid rinses are effective for reducing Escherichia coli 0157:H7 loads on beef carcass surfaces, the mechanisms by which these interventions inactivate bacterial pathogens are still poorly understood. It was hypothesized that E. coli 0157:H7 exposure to hot water in vitro at rising temperatures for longer time periods would result in increasing deterioration of bacterial outer membrane lipids, sensitizing the pathogen to subsequent lactic acid application. Cocktails of E. coli 0 157:H7 strains were subjected to hot water at 25 (control) 65, 75, or 85°C incrementally up to 60 s, after which surviving cells were enumerated by plating. Formation of lipid hydroperoxides from bacterial membranes and cytoplasmic accumulation of L-lactic acid was quantified spectrophotometrically. Inactivation of E. coli 0157:H7 proceeded in a hot water exposure duration- and temperature-dependent manner, with populations being reduced to nondetectable numbers following heating of cells in 85°C water for 30 and 60 s (P < 0.05). Lipid hydroperoxide formation was not observed to be dependent upon increasing water temperature or exposure period. The data suggest that hot water application prior to organic acid application may function to increase the sensitivity of E. coli 0157:H7 cells by degrading membrane lipids.
In the United States, an estimated 175,905 illnesses attributed to foodborne Shiga toxin-producing Escherichia coli, including E. coli 0157:H7, are estimated to occur annually (28). Fresh beef is most frequently implicated in foodborne disease outbreaks associated with these patho gens (15). Beef carcasses may become cross-contaminated during slaughter and dressing. Hot water carcass sprays applied alone or in combination with organic acid application have been studied extensively for carcass decontamination (1, 4, 8, 12, 13). Water temperature can impact the observed antimicrobial efficacy of a carcass water spray. Ambient water washes, consisting of a 1.5-liter manual water wash (25 °C and 10 lb/in2 for 90 s) and a subsequent 5.0-liter automated cabinet water wash (35°C and 250 to 400 lb/in2 for 9 s), were reported to reduce E. coli 0157:H7 populations on beef carcass surfaces by 1.7 to 2.7 log CFU/cm2. The application of a postwash hot water spray (95°C and 24 lb/in2 for 5 s) was reported to provide an additional 0.8 to 2.2 log cycle of E. coli 0157:H7 reduction * Author for correspondence. Tel: 979-862-7678; Fax: 979-845-4992; E-mail: [email protected]. t Authors contributed equally to manuscript completion, t No current professional affiliation. § Present address: IEH Laboratories and Consulting Group, 15300 Bothell Way N.E., Lake Forest Park, WA 98155, USA.
in addition to that achieved by water washing alone (9, 10). Researchers evaluating the efficacy of hot water sprays on beef carcasses reported reductions in aerobic bacteria of 2.2 log CFU/cm2 (80°C and 30 lb/in2 for 5.6 s) and 1.3 log CFU/cm2 (95°C spray for 40 s), reductions in E. coli of 1.8 log CFU/cm2 (80°C and 30 lb/in2 for 5.6 s) (17) or 4.2 log CFU/cm2 (74°C and 300.2 lb/in2 for 12 s) (6), and in total coliforms of 1.4 log CFU/cm2 (77°C and 20 to 22 lb/in2 for 2.5 s) (16). Variability among observed microbial reductions between studies was reportedly likely due to differences in experimental design, equipment, and beef cut selection. Numerous factors, including the sequence of hurdle application, can influence the efficacy of antimicrobial hurdle technologies for pathogen inhibition on carcass surfaces. Castillo et al. (8) reported 4.9 log CFU/cm2 reductions of E. coli 0157:H7 inoculated on beef carcasses, following a water wash plus a hot water spray (95 °C and 24 lb/in2 for 5 s) and 2% L-lactic acid warmed to 55°C. However, when the sequence of interventions was reversed (lactic acid followed by hot water), E. coli 0157:H7 counts were 0.5 log cycle lower than those subjected to hot water plus lactic acid (8). Bosilevac et al. (4) described reductions in aerobic plate counts of 2.2 log CFU/100 cm2 following the sequential application of hot water (74°C and 700 lb/in2 for 5.5 s) and 2% L-lactic acid (42°C, pH 2.4) on preevisceration carcasses. Observed reductions were
1198
CALIX-LARA ET AL.
lower than those achieved using only hot water (2.7 log CFU/100 cm2; P > 0.05); the authors suggested that observed differences were likely the result of cooling effects of the acid solution on the carcass surface. Although thermal treatment has been shown to inactivate pathogenic microorganisms on foods, the mechanisms responsible are not fully elucidated. Bruskov et al. (5) suggested exposure to heat may activate dissolved atmo spheric oxygen, initiating the production of reactive oxygen species. The authors proposed that thermally mediated damage to DNA was due to the action of these newly formed reactive oxygen species. Reactive oxygen species, including superoxide anions (0 2 ), hydrogen peroxide (H20 2), and hydroxyl radicals (—OH), can cause oxidative damage to cellular components, including lipids, proteins, and nucleic acids (7,11). Oxidative stress has been cited as a key factor in the heat-induced death of Saccharomyces cerevisiae, evi denced by a 500 to >20,000-fold increase in heat tolerance of cells in anoxic versus aerobic conditions (14). The relationship between thermal stress and intracellular lipid peroxidation has also been demonstrated in prokaryotes. Bacillus subtilis cells exposed to heat stress (55 °C for 1 h) were found to have 314% greater levels of intracellular superoxide than those grown at 37°C (27). This study was conducted to determine the impact of hot water exposure at increasing temperature for longer exposure periods on the survival of E. coli 0157:H7 and the integrity of outer membrane lipids prior to the application of 5.0% L-lactic acid. Researchers hypothesized that in vitro exposure of E. coli 0157:H7 cells to water heated to increasing temperatures simulating beef carcass interventions would result in increased formation of primary lipid oxidation products from membrane lipids (namely, hydroperoxides) and provide increased internalization of L-lactic acid into the cytoplasm of cells. MATERIALS AND METHODS Bacterial cultures and revival procedures. E. coli 0157:H7 strains P8, P I8, and P41 (bovine carcass isolates) and Environ mental Diseases Laboratory 932 (human fecal isolate) were obtained from the Food Microbiology Laboratory culture collec tion in the Department of Animal Science (Texas A&M University, College Station). Each strain was individually resuscitated via duplicate overnight transfer and subsequent static incubation in tryptic soy broth (TSB; BD, Sparks, MD) at 35°C. Once revived, isolated colonies were obtained by streaking sorbitol MacConkey agar (BD); plates were incubated for 24 h at 35°C. Isolated colonies displaying correct phenotype (zone of precipitated bile salts without fermentation of sorbitol) were aseptically picked and transferred to TSB for incubation and preparation of tryptic soy agar (TSA; BD) slants. Bacterial cocktail preparation. Five milliliters of overnight cultures (~9.0 log CFU/ml) of E. coli 0157:H7 strains were aseptically dispensed in sterile 15-ml centrifuge tubes (Thermo Fisher Scientific, Inc., Pittsburg, PA) and centrifuged at 1,623 x g in a Jouan B4i centrifuge (Thermo Electron Corp., Madison, WI) for 15 min at 25°C. The resulting pellet for each strain was suspended in 5.0 ml of 0.1% peptone diluent, following aseptic removal of supernatant and washed twice by centrifugation under the same conditions as mentioned. Immediately following the third centrifu gation, each pellet was suspended in 5.0 ml of 0.1% peptone diluent.
J. Food Prot., Vol. 78, No. 6
Finally, 2.5 ml of each suspension were aseptically transferred and pooled in a flask containing 90.0 ml of sterile 0.1% peptone diluent to create a cocktail of —8.0 log CFU/ml. E. coli 0157:H7 cocktail exposure to hot water. E. coli 0157:H7 cocktail aliquots of 1.5 ml each were aseptically dispensed into sterile disposable borosilicate glass tubes (10 x 75 mm) and submerged in a flowing water bath set to heat cell suspensions to 25, 65, 75, or 85°C. The water bath contained sufficient water to cover about 1.8 ml of volume of tubes; bath temperature was constantly monitored by a submerged thermom eter. Timers were initiated once samples equilibrated to the desired temperature. Water temperature inside a test tube filled to 1.5 ml of peptone diluent, but without pathogen cells, was monitored by using a type K thermocouple (model 108A 800008, Sper Scientific Ltd., Scottsdale, AZ) and a four-channel data-logging meter (model 108A 800024; Sper Scientific Ltd.). Cell suspensions were exposed for 0, 15, 30, or 60 s; individual tubes were prepared to allow for each tube to be subjected to a single heating period and then be sampled. After hot water exposure, sample tubes were immediately submerged in ice-chilled water (0°C) for —20 s to arrest heat-induced bacterial inactivation and inhibit further membrane degradation. Surviving cells were quantified by preparation of decimal dilutions in 0.1% (wt/vol) peptone diluent and spread on TSA-containing petri dishes, followed by incubation for 48 h at 35°C. Colonies were then counted and recorded. Membrane lipid extraction from heated E. coli 0157:H7 cells. Methyl-tert-butyl ether lipid extraction method was per formed according to the procedure reported by Matyash et al. (22). After hot water treatment, 200 pi of each sample tube containing pathogen cocktail (—1.5 x 107 CFU) was placed into a glass centrifuge tube, twice centrifuged for 15 min at 1,623 x g, and washed with 0.1% ammonium acetate (Sigma-Aldrich Co., St. Louis, MO). Next, 1.5 ml of absolute methanol (Thermo Fisher Scientific, Inc.) was added and vortexed for 4 s. Then, 5.0 ml of methyl-tert-butyl ether was added, and the samples were incubated for 60 min at 25°C, with gentle shaking (Mistral 4500 MultiMixer, Colonial Scientific, Inc., Richmond, VA). After incubation, 1.25 ml of sterile distilled water was added to stimulate phase separation, followed by 10 min of incubation, and then a 10 min of centrifugation at 1,000 x g. Resulting upper phases were collected separately, dried under compressed nitrogen, and stored overnight in a desiccator at 4°C before analysis. Quantification of lipid hydroperoxides from E. coli 0157:H7 membrane lipids. Following lipid extraction, the formation of lipid hydroperoxides after hot water exposure was assayed by the ferric oxide method, as described previously (29). A linear standard curve was initially completed to provide for quantification of lipid peroxide content from hot water-treated samples (Fig. 1). Thereafter, standard curves were always completed prior to the assay of experimental replicates. Desiccated lipid from each 200-pl sample was reconstituted with 4.8 ml of solution of chloroform-methanol 7:3 (Acres Organics, Waltham, MA); within 10 min, 50.0 pi of xylenol orange (10 mM) and 50 pi of solution of ferrous (Fe2+ ) chloride, prepared as instructed by Shantha and Decker (29), were added. Unless indicated otherwise, all experimental reagents were purchased from Sigma-Aldrich Co. and were of highest purity. Samples were held in the dark at room temperature for 5 min prior to assay. Absorbance was read at 560 nm against a blank with a DU-7400 UV/Vis Spectrophotom eter (Beckman Coulter, Fullerton. CA). Milliequivalents (meq) of lipid peroxide per kilogram of sample were calculated as
J. Food Prot., Vol. 78. No. 6
E. C0L1 0157:H7 MEMBRANE OXIDATION BY HOT WATER
FIGURE 1. Standard curve o f ferric oxide detection o f lipid hydroperoxide formation from E. coli 0157:H7 membrane lipids. Standard curves were prepared new fo r each independent replication; symbols depict means from triplicate identical replicates. Limit o f detection for assay: 2.0 meq. Absorbance was read at 560 nm. Peroxide value (PV) = (As - A h) x M /55.84 x M„ x 2
(1)
where A,, and Ah are the absorbance of the sample and blank at 560 nm, respectively, M is slope of the standard curve as determined by regression analysis, 55.84 (atomic weight of iron), and M0 is the sample mass (g). The denominator was multiplied by a conversion factor (2) to allow for expression of peroxide value in meq of lipid hydroperoxides. L-Lactic acid accumulation by E. coli 0157:H7 following hot water exposure. E. coli 0157:H7 cells, previously subjected to hot water, were aseptically transferred (1.0 ml) to a sterile microcentrifuge tube, and 1.0 ml (wt/wt) of L-lactic acid (FCC 88, Purac America, Lincolnshire, IL) was added to give a final concentration of 5.0% (wt/wt) L-lactic acid. Acid concentration was calculated assuming bacterial suspensions possessed a density of —1.0 g/ml. Tubes were agitated by vortexing and allowed to incubate at room temperature for 5 min. Cells were removed from the acid by centrifugation and diluted in 1.0 ml of 0.1% (wt/vol) peptone diluent. Surviving E. coli 0157:H7 cells were identified by plating on surfaces of TSA-containing petri dishes, followed by incubation of inoculated plates for 48 h at 35°C prior to colony enumeration. Samples (0.5 ml) of hot water and lactic acid-treated E. coli 0157:H7 were loaded into microcentrifuge tubes containing 0.2 ml of 6 N perchloric acid (Sigma-Aldrich Co.) and 0.5 ml of a silicone oil blend. Perchloric acid was used to arrest bacterial metabolism and allow for quantification of accumulated lactic acid. Blended silicone oil (p = 1.055 g/ml) was layered on top of perchloric acid and allowed for the selective transmission of bacterial cells from the top phase into perchloric acid, once subjected to centrifugation. Silicone oils were blended at a ratio of 3:1 (vol/vol) AR 20 (p =
1199
1.008 g/ml; Fluka, Seelze, Germany): high-temperature silicone oil (p = 1.102 g/ml; Sigma-Aldrich, Inc.). In a separate glass tube (10 by 75 mm), a mixture of 0.5 ml of sample and 0.2 ml of either 0.1% peptone water or L-lactic acid was formulated; L-lactic acid was prepared from concentrate (FCC 88, Purac America) to deliver a final concentration of 5% (wt/wt). The mixture was added as the top layer (third layer) in the microcentrifuge tube. Lactic acid exposure to the microbial cells was approximately 5 min and was stopped by centrifugation (model 59V, Lab-Line Instruments, Melrose Park, IL) at maximum velocity for 1 min or until a microbial pellet was noticed at the bottom. The upper and lower phases were aseptically collected and stored under refrigeration (5°C) until analysis was conducted; the oil phase was discarded. The estimation of lactic acid accumulation was completed by using previously reported methods (3, 18). L-Lactate dehydrogenase (643 U/mg) was used to catalyze the conversion of lactate to NADFI to allow for determination of lactic acid content accumulation. A chemical trap (hydrazine monohydrate) was added to disallow the enzyme from catalyzing the conversion of NADH to L-lactic acid. The trap consisted of 1.25 ml of hydrazine monohydrate (0.4 M) glycine buffer (0.5 M, pH 9.0). Potassium carbonate (K2C 0 3; 20 pi, 5.0 M) was added to buffer perchloric acid used previously to arrest microbial metabolism (18). To the reaction vessel was added 0.1 ml of NAD solution (2.75 mM) and 0.1 ml of sample. AH reagents were purchased from Sigma-Aldrich Co. and were of highest purity. Two microliters of enzyme (undiluted) was added, and the catalysis of NADH allowed to proceed at room temperature for 60 min, after which the absorbance of NADH was monitored at 339 nm with a DU-7400 UV/Vis Spectrophotometer (Beckman Coulter). Samples of acid-exposed and peptone-exposed (control) cells were tested identically to account for residual lactate produced during bacterial fermentation of glucose in microbiological medium during over night growth. Nonaccumulated acid was determined by the difference of total acid exposure minus baseline-adjusted cytoplas mic lactic acid. L-Lactic acid concentration in E. coli 0157:H7 was determined as [L-lactic acid] = A339 nm x 0.344
(2)
where A339nm is the absorbance of NADH at 339 nm and 0.344 is a conversion factor providing the concentration of i.-lactate obtained in the sample (nanomole per milliliter). Experimental design and statistical analyses. Plate counts of surviving E. coli 0157:H7 were log transformed prior to analysis of differences in E. coli 0157:H7 numbers or lipid hydroperoxides from membrane lipids as a function of heating temperature, exposure period, and their interaction (temperature x exposure period). Statistical differences between means were obtained by one-way analysis of variance (P < 0.05) and Tukey’s honestly significant difference test (P < 0.05) using the generalized linear model procedure of SAS, version 9.0 (SAS Institute, Inc., Cary, NC).
RESULTS AND DISCUSSION The inactivation of E. coli 0157:H7 as a result of exposure to heated water for increasing periods of time is shown in Table I. No reductions in E. coli 0157:H7 populations were observed for control samples incubated at 25°C. Inactivation of E. coli 0157:H7 by hot water increased in a temperature-dependent and exposure perioddependent manner, with significant reductions (P < 0.05) achieved after 15 s of hot water exposure (Table 1). Maximum reduction in pathogen numbers was achieved at 60 s of heating as compared with nonheated cells (Table 1).
1200
CALIX-LARA ET AL.
J. Food Prot., Vol. 78, No. 6
TABLE 1. Least-square means o f the interaction o f water temperature by exposure period on the survival ofE. coli 0157:H7a Heating period (s)
TABLE 2. Least-square means o f lipid peroxides from E. coli
0157:H7 cells following exposure to heated watera
Mean E. coli 0157:H7 survival (log CFU/ml) — ------------------------------------------------------------------25°C 65°C 75°C 85°C
Mean (meq of peroxide/kg) Heating period (s)
0 5 15 30 60
8.1 8.1 8.2 8.1 8.1
A A A A A
8.1 6.5 4.9 3.9 3.8
A AB bc
BCD cd
8.1 3.4 2.8 2.7 2.2
A cd
CDE CDEF DEF
8.1 A ND* ND ND ND
a P < 0.0001. Values presented depict means from triplicate identical replications (n = 3). Values not sharing capital letters (a through f) differ at P < 0.05, assessed by one-way analysis of variance and Tukey’s honestly significant difference test. Standard error of the mean = 0.29. * ND, nondetectable; limit of detection: 1.0 CFU/ml.
E. coli populations exposed to water set to 75°C were rapidly reduced (P < 0.05) after 5 s of treatment, with no further reductions observed following 15, 30, and 60 s of heat exposure. Significant reductions (P < 0.05) in pathogen numbers were achieved within 5 s of incubation at 85°C, with numbers decreased to below the limit of detection thereafter. Resulting trends in reductions mirror those observed previously, when hot water application for increasing lengths of time resulted in statistically greater inactivation of carcass-adhered pathogens (31). Factors intrinsic to meat, including pH, protein and fat content, and water activity, can greatly influence microbial thermal resistance (25, 33). The present thermal inactivation rates are similar to those reported in both nutrient broth (25) and deionized water (32), while the bacterial inactivation in meat matrices is generally less rapid (20, 24). Such differences are likely due to the presence of muscle and fat tissue, which exert an insulatory effect, protecting bacterial contaminants from thermal inactivation (9). Bovine tissue can also act as a buffer and may reduce the efficacy of organic acids against pathogens (2, 26). Though information regarding the mechanisms and rates of heat-induced bacterial inactivation are gained from the present study, the practical application of these data for beef producers are limited by the simplicity of the experimental system. The absence of muscle and lipid in the experimental sample, however, was required in this study to allow for direct assay of bacterial lipid deteriora tion, as beef fat would have compromised the researchers’ ability to quantify pathogen membrane lipid degradation. Degradation of E. coli 0157:H7 outer membrane lipids into lipid hydroperoxides was generally independent of hot water temperature and exposure duration (Table 2). The lack of statistical differences between samples incubated for increasing lengths of time may have resulted from the reactivity of hydroperoxides and their degradation into products not directly assayed by the experimental method used in this study, or by lack of sufficient lipids to produce degradation products in necessary quantity to be detected by the methods described. However, meq per kilogram of hydroperoxides detected by the ferric oxide method, described as being sensitive to peroxide value levels in this
0 5 15 30 60
P values
0.003
0.23
0.008 0.013 0.020 0.024 0.008
Temp (°C) 25 65 75 85
Pooled SEM
0.24 0.008 0.026 0.026 0.014
a Values presented depict means from triplicate identical replica tions (n = 3). P values assessed by one-way analysis of variance.
study, were within the linear and detectable range of the assay (Fig. 1). Nonetheless, despite the likely degradation of membrane-associated fatty acids, assay of bacterial lipid deterioration following the application of a simulated beef carcass antimicrobial hurdle process has not been previously reported to the authors’ knowledge. While others have sought to determine the impact of antimicrobial interventions on bacterial membrane integrity, past research has typically focused on the shuffling of membrane lipids by the microbe in response to the antimicrobial, rather than identify deterioration in membrane lipids (21,23). Studies reporting the degradative mechanisms of thermal processing on bacterial membranes are generally lacking in the scientific literature. Jang and Rhee (19), in studies of the effects of heating on caprylic acid inhibition of Cronobacter sakazakii in reconstituted infant formula, reported increasing heating temperature and time resulted in enhanced damage to the integrity of the bacterial membranes, as determined by differential staining. Smirnova et al. (30) reported increased superoxide formation and intracellular H20 2 in E. coli grown at 42 versus 37°C, though these reactive oxygen species can be formed from nonlipid precursors, as well as from lipid hydroperoxide-incorporating reactions. Collectively, data reported here and previous reports indicate the potential for hot water to sensitize bacterial microbes to the effects of an antimicrobial by disrupting the integrity of the membrane. Following hot water exposure, cells were exposed to 5% (wt/wt) L-lactic acid for 5 min prior to enumeration. For all treatments, numbers of E. coli 0157:H7 were reduced to nondetectable (
