Salmonella Isolated from Ready-to-Eat Pasteurized Liquid Egg Products: Thermal Resistance, Biochemical Profile, and Fatty Acid Analysis Joshua B. Gurtler, Arthur Hinton Jr., Rebecca B. Bailey, William C. Cray Jr., Tony Z. Jin PII: DOI: Reference:
S0168-1605(15)00202-0 doi: 10.1016/j.ijfoodmicro.2015.04.010 FOOD 6875
To appear in:
International Journal of Food Microbiology
Received date: Revised date: Accepted date:
3 October 2014 31 March 2015 4 April 2015
Please cite this article as: Gurtler, Joshua B., Hinton Jr., Arthur, Bailey, Rebecca B., Cray Jr., William C., Jin, Tony Z., Salmonella Isolated from Ready-to-Eat Pasteurized Liquid Egg Products: Thermal Resistance, Biochemical Profile, and Fatty Acid Analysis, International Journal of Food Microbiology (2015), doi: 10.1016/j.ijfoodmicro.2015.04.010
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1
Salmonella Isolated from Ready-to-Eat Pasteurized Liquid Egg
IP T
Products: Thermal Resistance, Biochemical Profile, and Fatty Acid
SC R
Analysis
Joshua B. Gurtler*1, Arthur Hinton, Jr.3, Rebecca B. Bailey1, William C. Cray, Jr. 2,
MA
1
NU
Tony Z. Jin4
Food Safety and Intervention Technologies Research Unit, U.S. Department of Agriculture,
D
Agricultural Research Service, Eastern Regional Research Center, 600 East Mermaid Lane,
TE
Wyndmoor, PA 19038-8551
U.S. Department of Agriculture, Food Safety and Inspection Service, Athens, GA
3
U.S. Department of Agriculture, Agricultural Research Service,Richard B. Russell Research Center,
4
AC C
Athens, GA
EP
2
Residue Chemistry and Predictive Microbiology, U.S. Department of Agriculture, Agricultural
Research Service, Eastern Regional Research Center, Wyndmoor, PA
Running title: Salmonella from pasteurized egg products Key Words: Salmonella, liquid egg products, pasteurization
* Corresponding author. Tel.: 215-233-6788; fax: 215-233-6406. E-mail address: [email protected] (J.B. Gurtler).
1
ACCEPTED MANUSCRIPT 2
ABSTRACT (362 Words) _________________________________________________________________________
IP T
The Egg Products Inspection Act of 1970 requires that egg products in the U.S. must be pasteurized
SC R
prior to release into commerce. The USDA Food Safety and Inspection Service (FSIS) is responsible for regulating egg products. Salmonellae are infrequently isolated from pasteurized egg products by food manufacturers or the FSIS and may be present as a result of either pasteurization-resistant
NU
bacteria or post-processing contamination. In this study, seventeen strains of Salmonella isolated from
MA
pasteurized egg products and three heat-resistant control strains were compared for the following attributes: thermal resistance in liquid whole egg (LWE) at 60°C, enzymatic profiles, and serotyping
D
and phage typing,. fatty acid analysis and strain morphological variation evaluated by scanning
TE
electron microscopy. Isolates were serotyped as Heidelberg (4 isolates), Widemarsh, Mbandaka,
EP
Cerro, Thompson, 4,12:i:-, and Enteritidis (8 isolates). The D60 values in LWE ranged from 0.34 – 0.58 min. All 20 strains were recovered from LWE inoculated with 8.5 log CFU/ml of Salmonella and
AC C
pasteurized at 60°C for 3.5 min; however, some isolates were not recovered from pasteurized LWE that had been inoculated with only 4.5 log CFU/ml Salmonella and treated at 60°C for 3.5 min. Although some strains exhibited atypical enzymatic activity (e.g., reduction of adonitol, hydrolysis of proline nitroanilide or p-n-p-beta-glucuronide, and nonreduction of melibiose), differences in biochemical reactions could not be correlated with differences in thermal resistance. Furthermore, fatty acid analysis revealed that differences insaturate/unsaturated profiles may be correlated with differences in heat resistance, in two instances. One heat resistant strain (#13, Enteritidis) had the statistically lowest unsaturated/saturate ratio at 39%. However, one heat sensitive strain (#3, serovar 4,12:i:-) had the highest unsaturated/saturate ratio at 81%, and also the lowest concentration of stearic
2
ACCEPTED MANUSCRIPT 3 acid. This data represents the first steps in determining whether Salmonella contamination in pasteurized egg products may be the result of either thermally-resistant isolates or post-processing contamination. Contamination of LWE by Salmonella strains with higher heat resistance, (e.g., isolate
IP T
#’s 2, 6, 10 and 12) may indicate the ability of Salmonella to survive pasteurization, while
SC R
contamination of LWE strains with lower heat resistance (e.g., isolate #’s 1, 3, 5, 7, 8, 11, and 15) may indicate post-processing contamination of LWE by this foodborne pathogen.
MA
NU
1. INTRODUCTION
The Egg Products Inspection Act of 1970 requires that egg products in the U.S. must be pasteurized
D
prior to release into commerce, resulting in ca. 2.7 billion pounds of liquid egg (LE) pasteurized for
TE
human consumption in the U.S. every year (USDA, NASS, 2010). The USDA Food Safety and
EP
Inspection Service (FSIS) is responsible for regulating egg products, and requires liquid whole egg (LWE) to be pasteurized at 60°C for a minimum of 3.5 min, after which it may be served to consumers
AC C
with no further interventions to inactivate bacteria (9 CFR 590.570, Table 1 [Code of Federal Regulations, 2009]). Nevertheless, salmonellae are infrequently isolated from these pasteurized egg products by food manufacturers or the FSIS. Contamination of these egg products may be the result of either pasteurization-resistant bacteria, which can survive the heating process, or as a result of postprocessing contamination of pasteurized LWE by Salmonella (Blais et al., 1998; Suzuki and Yamamoto, 2009). Sporadic outbreaks of salmonellosis in the U.S. have been potentially associated with pasteurized egg products from the consumption of food items such as crab cakes, scrambled eggs, chile relleno, and a meringue-topped dessert (CDC, 2012; Durham County, 2010; Gurtler et al., 2013). Other known outbreaks of Salmonella from pasteurized egg products have been reported overseas
3
ACCEPTED MANUSCRIPT 4 (Hara-Kudo and Takatori, 2009; UKFSA, 2007a, 2007b). The USDA-FSIS has published a risk assessment model (FSIS, 2005, Exec. Summary) in which they estimate that there may be approximately 5,500 U.S. cases of salmonellosis per year that stem from contaminated pasteurized LE
IP T
products (Latimer et al., 2008). A series of recently published studies have reevaluated the
SC R
time/temperature pasteurization requirements for some LE products (Gurtler et al., 2011, 2013; Jordan et al., 2011). One study (Gurtler et al., 2013) predicted that, based on current FSIS pasteurization requirements for 10% salted LWE (63.3°C for 3.5 min), only an ca. 3.6 log reduction of Salmonella
NU
would be achieved. In the same study the authors determined that to achieve a 5 log reduction, the
MA
pasteurization requirements would need to be increased to 62.2°C for 7.5 min. In another study, the authors concluded that, based on current FSIS pasteurization requirements for 10% salted liquid egg
D
yolk (63.3°C for 3.5 min), only an ca. 2.7 log reduction of Salmonella would be achieved (Gurtler et
TE
al., 2011). In order to achieve a desired 5 log reduction, the same authors determined that the
EP
pasteurization temperature would need to be raised to 67.3°C for 3.5 min. The goal of the present study was to evaluate seventeen selected strains of Salmonella that had been isolated from ready-to-eat
AC C
pasteurized liquid egg products for variations in thermal resistance, biochemistry, serology, morphology, and fatty acid profiles. Additionally, three Salmonella isolates that were included in published liquid egg pasteurization studies were included in these same battery of tests for comparison purposes. The overall goal of the study was to identify determinants that might suggest whether the seventeen salmonellae strains isolated from pasteurized liquid egg products might be there as a result of heat-resistant survival of the thermal pasteurization process, or whether strains are heat-sensitive, suggesting that they might be present in the egg products as a result of post-processing contamination.
2. MATERIALS AND METHODS
4
ACCEPTED MANUSCRIPT 5
2.1 Determination of thermal D-values Seventeen strains of Salmonella isolated from pasteurized egg products by the USDA-FSIS were
IP T
used in this study (Table 1). Additionally, two strains of Salmonella Enteritidis phage type 8 (C398
SC R
and C405, provided by Keith Lampel, FDA, College Park, MD) and one strain of Salmonella Oranienburg DD2229 from the Dupont Salmonella collection were included for comparison purposes. These strains were used in published USDA-ARS/FSIS egg pasteurization studies (Table 1) (Gurtler et
NU
al., 2011, 2013; Jordan et al., 2011). Isolates were tested for thermal inactivation kinetics during 3
MA
repetitions in LWE at 60°C with heating times up to 5 min using the hot water-immersion capillary tube method. The come-up-time was 6 s, while the cool-down-time was 3 s. Salmonella strains were
D
individually incubated in 10mL of tryptic soy broth (TSB) (Difco, Sparks, MD) with 50 ppm nalidixic
TE
acid (NA) at 37°C for 24 h. A single 10µL loop transfer was performed and suspensions were
EP
incubated for another 24 h to reach populations of ca. 9 log CFU/mL. Suspensions were plated on tryptic soy agar (TSA) (Difco) with 0.1% sodium pyruvate and 50 ppm NA (TSAPN). Isolates were
AC C
individually centrifuged at 22°C for 10 min, the supernatant was decanted, and the bacterial pellet was resuspended in 0.4mL of 0.1% peptone water in order to produce a suspension with a population of ca. 9 log CFU of Salmonella/mL. Salmonella suspensions (0.5 mL from each of three strains) were individually added to 38mL of LWE at 21°C and mixed for 2 min. Approximately 200µL of inoculated LWE was injected into 250µL capillary tubes (Accu-Fill 90 Micropet Disposable Pipettes, Clay Adams, Inc., Parsippany, NJ) and the ends of the tubes were heat-sealed using an open flame. Capillary tubes were then placed in a stainless-steel test tube rack, by inserting the capillary tubes through the holes of attached aluminum tape. The entire test tube rack with capillary tubes was then immersed in a circulating water bath (ThermoFisher Scientific NESLAB RTE17 Digital Plus) set to
5
ACCEPTED MANUSCRIPT 6 60°C. To monitor the come-up time and temperature, Digisense Thermocouple Flexible High Temperature Wire Probes (Cole-Parmer, Vernon Hills, IL) connected to a Fluke 54 II Thermometer (Fluke Calibration, Everett, WA) were placed in the water bath and affixed inside a capillary tube
IP T
containing LWE. After the capillary tubes reached 60°C, they were sampled at 1 min intervals for up
SC R
to 5 min, then immersed in ice water for 10 s. Cooled tubes were then removed from the ice water, surface-sanitized in 70% ethanol, and rinsed twice with sterile-water. The ends of capillary tubes were then removed with ethanol-flamed wire cutters and the contents were expelled manually with a sterile
NU
4-mm pipe cleaner (Creativity Street, 4 mm Chenille Stems Pipe Cleaners, The Chenille Kraft
MA
Company, Gurnee, IL). Serial dilutions of the LWE were performed, dilutions were spiral plated in duplicate on TSAPN and incubated at 37°C for 24 h. Undiluted egg samples (50 µL) were spread
D
plated in duplicate on TSAPN, yielding a minimum detection limit of 10 CFU/ml. Colonies were
TE
counted using a Synbiosis aCOLyte Supercount (Microbiology International, Frederick, MD).
EP
Presumptive-positive Salmonella were randomly confirmed by replica-plating on XLD agar (Difco) and serological agglutination (Salmonella O Antiserum Poly A-I & Vi, Difco, Sparks, MD). D-values
AC C
were calculated from the negative inverse of the slope of the survival curve (Table 2).
2.2 Test tube heating study
Four repetitions of a second thermal study were conducted to determine the ability of various populations of Salmonella to survive pasteurization in LWE at the FSIS-mandated 60°C for 3.5 min by hot water immersion in test tubes. LWE inoculated to 3.5, 4.5, or 8.5 log CFU/mL was added (1.2 mL) to 5 mL test tubes. A thermocouple was placed in uninoculated LWE in one test tube to determine the come-up-time. After the egg in the tubes reached 60°C, all tubes were held for an additional 3.5 min, whereupon they were immediately immersed in an ice water bath. One mL of
6
ACCEPTED MANUSCRIPT 7 pasteurized egg was then removed and added to 9 mL of TSB + 35mg/L ferrous sulfate and 0.1% sodium pyruvate (TSB-FP), for a primary enrichment. Ferrous sulfate and sodium pyruvate were added to assist in the recovery of injured cells (Gurtler and Kornacki, 2009). After incubating primary
IP T
enrichments for 24 h at 37°C, selective enrichments were then made by adding 1 mL of the TSB-FP
SC R
primary enrichment to 9 mL of tetrathionate (TT) Broth (Difco) as well as to 9 mL of Rappaport Vassiliadis broth (Difco) and incubated at 42°C for 24 h. Tetrathionate Broth Base (Difco) was supplemented with 100 ppm nalidixic acid + 20 mL/L of iodine solution + 0.01 g/L of brilliant green
NU
(Sigma) which served as a selective enrichment for Salmonella along with RV broth. A 10 µL loop of
MA
both TT broth and RV broth were then streaked to separate plates of XLD agar and incubated at 37°C
D
for 24 h. Presumptive-positive Salmonella were randomly confirmed by serological agglutination.
TE
2.3 Enzymatic and serological profiles, phage typing, and fatty acid analysis
EP
Enzymatic profiles were determined by using the BBL Crystal Identification System Autoreader (Becton, Dickinson and Company, Franklin Lakes, NJ). Strains were serotyped and phage typed at the
AC C
USDA, APHIS, National Veterinary Service Laboratory in Ames, Iowa. Fatty acid analyses of salmonellae were conducted by gas chromatography and FAME MIDI Sherlock software, following the manufacturer’s directions for cell preparation (MIDI, Newark, DE) as previously published (Hinton, et al., 2004). Three replicate cultures of each Salmonella isolate were prepared, by growth on BBL Trypticase™ Soy Agar (Becton, Dickson and Company, Sparks, MD) at 30°C for 24 hr and extraction of fatty acids from cell walls of the bacteria and methyl ester generation were performed using the Sherlock Instant FAME Method (MIDI, Inc.). GC analysis of the cell extracts was performed with the Hewlett Packard HP 6890 Series GC System that was computer operated by Aligent ChemStation software (Aligent Technologies, Santa Clara, CA) and fatty acid data was analyzed by
7
ACCEPTED MANUSCRIPT 8 MIDI Microbial Identification Sherlock software (MIDI). MIDI Sherlock calibration standards were included with each GC analysis. Briefly, after three consecutivetransfers and incubation, cultures were harvested and fatty acids were extracted from the cellular membranes using the four-step
IP T
extraction procedure recommended by the manufacturer. The first step of the extraction procedure
SC R
consisted of saponification by heating the cells at 100°C for 30 min in a methanolic sodium hydroxide solution composed of 15% (wt/vol) NaOH (Spectrum Chemical Manufacturing Corp., Gardena, Calif.) dissolved in 50% (vol/vol) methanol (Sigma-Aldrich, Inc., St. Louis, Mo.). Saponification lysed the
NU
bacteria and released cellular fatty acids. Methylation of released fatty acids was accomplished by
MA
mixing the resultant suspension in a solution of 3.25 N hydrochloric acid (Ricca Chemical Co., Arlington, Tex.) in methanol (46%, vol/vol) at 80°C for 10 min. The fatty acid methyl esters (FAMEs)
D
were extracted from the aqueous phase to the organic phase of the mixture by a 10-min liquid-liquid
TE
extraction in a 1:1 solution of hexane (Sigma-Aldrich, Inc.) and methyl-tert-butyl ether (Sigma-
EP
Aldrich, Inc.). A 5-min base wash of the FAME extracts in a sodium hydroxide solution (1.08%, wt/vol) was performed to remove free fatty acids and residue from the organic extract. FAME extracts
AC C
were transferred to sample vials for identification and quantification by an Agilent 6890 gas chromatograph (GC) (Agilent Technologies.). The GC was operated by a computer containing ChemStation (Agilent Technologies) and MIDI Sherlock MIS, version 4.5, software. ChemStation software set the GC parameters of operation and controlled GC functions throughout FAME analysis. The MIS software identified and quantified FAMEs of bacterial isolates. Additionally, all isolates were plated out for isolation on XLD and XLT4, agar, to determine if they displayed typical reaction on these common salmonellae media.
2.4 Scanning electron microscopy
8
ACCEPTED MANUSCRIPT 9 Prior to scanning electron microscopy, cells were prepared by immersing glass cover slips as well as shell egg chips (ca. 2 x 2 cm) in 3 mL of TSB and individually inoculating with 20 strains of Salmonella, respectively. Glass cover slips and egg shells were then incubated for 24 h at 37°C in
IP T
order to allow for growth and attachment of salmonellae. Glass cover slips and egg shells were then
SC R
removed from the TSB suspension and rinsed with glutaraldehyde. Cover slips and eggs shells were fixed for scanning electron microscopy (SEM) by immersion in a 2.5% glutaraldehyde-0.1 M imidazole buffer (Electron Microscope Sciences, Hatfield, PA) for 1 h before washing in imidazole
NU
buffer and dehydrating in 50%, 80% and absolute ethanol, successively. Samples were critical point
MA
dried (Denton Vacuum, Cherry Hill, NJ) with carbon dioxide, mounted with Duco cement (ITW Performance Polymers, Riviera Beach, FL) and colloidal silver adhesive, and sputter-coated with a
D
thin layer of gold using a Scancoat Six Sputter Coater (BOC Edwards, Wilmington, MA). Samples
TE
were imaged with a Quanta200 FEG environmental scanning electron microscope (FEI Co., Inc.,
EP
Hillsboro, OR), with an Everhart Thornley detector, operated in the high vacuum, secondary electron imaging mode at an accelerating voltage of 5 kV.
AC C
2.5 Statistical analyses
Data were analyzed by ANOVA, and significant differences (p < 0.05) between D60°C values as well as differences in fatty acid composition were determined post hoc by Fisher’s Least Significant Difference Test (LSD) via SAS version 9.1 (SAS Institute, Inc., Cary, NC). 3. Results 3.1 Serotyping and phage typing The seventeen isolates originating from pasteurized egg products were serotyped as (Table 1) Enteritidis (8 isolates, 47%), Heidelberg (4 isolates, 24%), and one each for serovars 4,12:i:-,
9
ACCEPTED MANUSCRIPT 10 Widemarsh, Mbandaka, Cerro, and Thompson. All Enteritidis isolates phage typed out as PT-8 except for one strain that was PT-2.
IP T
3.2 Determination of thermal D60 values
SC R
The thermal D60 values in LWE at 60°C over a 5 min heating period, as determined by three replicate experiments, ranged from 0.34 – 0.58 min (Table 2). The seven least heat-resistant serovars with thermal D60 values of from 0.34 - 0.36 min were four strains of S. Heidelberg, and one strain each
NU
of serovars 4,12:i:-, Widemarsh and Mbandaka. These results are not surprising as S. Heidelberg is
MA
known to not be especially heat-resistant. The three most heat-resistant serovars with thermal D60 values of from 0.55 – 0.58 min were two strains of Enteritidis (isolated from pasteurized frozen salt
D
yolk and from pasteurized frozen egg whites) as well as the Dupont Oranienberg DD2229 isolate.
TE
These results support the prior inclusion of the Oranienberg DD2229 isolate in previous liquid egg
EP
thermal inactivation studies (Gurtler et al., 2011; 2013; Jordan et al., 2011;); however, the two FDA Enteritidis strains, which were also included in the pasteurization studies, only had thermal D60 values
AC C
of 0.42 and 0.45 min. Thermal inactivation curves are graphically displayed in figures 1 and 2. Figure 1 contains results from the Salmonella isolates with the highest D-values, where all strains were viable after 3 min of pasteurization. In contrast, Figure 2 displays strains with the lowest D-values , in which 5 isolates reached non-detectable levels at 3 min of pasteurization.
3.3 Test tube heating studies (60°C for 3.5 min) All 20 strains survived pasteurization at 60°C for 3.5 min during two repetitions, when inoculated at 8.5 CFU/mL, as determined by enrichment (Table 3). Conversely, none of the 20 strains survived the same pasteurization regimen when LWE was inoculated at 3.5 log CFU/mL (minimum detection
10
ACCEPTED MANUSCRIPT 11 limit = 1 CFU/mL). However, when inoculated at 4.5 log CFU/mL and heated at 60°C for 3.5 min, over 4 repetitions, the previously determined three most heat-resistant serovars (Enteritidis strain #’s 2, 6, and Oranienberg DD2229) survived at rates of 3/4, 2/4, and 3/4, respectively. On the other hand,
IP T
the seven most heat-sensitive isolates (as determined by D-values) only survived pasteurization at
SC R
60°C for 3.5 min at rates of 1/4, 2/4, 1/4, 1/4, 1/4, 1/4, and 2/4, respectively.
3.4 Biochemical profile
NU
Data on biochemical profiles of all 20 strains of Salmonella were attained by use of the BBL
MA
Crystal Identification System Autoreader, which tests for 32 separate biochemical reactions (Table 4). Although some strains exhibited atypical enzymatic activity (e.g., reduction of adonitol, hydrolosis of
D
proline nitroanilide and p-n-p-beta-glucuronide, and nonreduction of melibiose) no differences in
TE
biochemical reactions could be correlated with similarity in thermal resistance. When plated on XLD
EP
and XLT4 agars, all strains displayed typical black colonies on XLD agar; however, ten strains out of twenty did not blacken the colonies on XLT4 agars, and remained white. These were strain #2
AC C
(Enteritidis PT-8), #6 (Enteritidis PT-2), #9 (Enteritidis PT-8), #10 (Enteritidis PT-8), #12 (Enteritidis PT-8), #13 (Enteritidis PT-8), #16 (Enteritidis PT-8), #17 (Enteritidis PT-8), #18 (Enteritidis PT-8), and #20 (Oranienburg). XLT4 and XLD agars both contain an H2S indicator system in which reduction of sodium thiosulfate produces hydrogen sulfide which is detected by the addition of ferric ions to form black colonies. Both XLD and XLT4 agars contained equivalent amounds of L-lysine, ferric ammonium citrate and sodium thiosulfate. Thus, it is not apparent why ten Salmonella isolates (nine (9) Enteritidis and one Oraniengburg) colonies failed to turn the characteristic blackened center on XLT4, while all strains produced blackened centers on XLD. Hence, XLT4 agar should be used with caution when used as a cultural confirmation based on blackening center of Salmonella colonies.
11
ACCEPTED MANUSCRIPT 12 Of colonies that did not blacken on XLT4, 9 of ten were S. Enteritidis, while only one S. Enteritidis blackened colonies on the medium; namely Enteritidis, FDA C398, Phage type 8, from a hen’s ovary,
IP T
University of Pennsylvania.
SC R
3.5 Fatty acid methyl ester analysis
While it has been hypothesized that greater concentrations of saturated fatty acids within the bacterial cell membrane may contribute to increased heat resistance (Alvarez-Ordonez et al., 2008;
NU
2009; Yang et al., 2014a; 2014b), our analysis suggests this may be true for at least two isolates (Table
MA
5). One example may be the heat-resistant S. Enteritidis #12, as assessed by 3 of 4 samples inoculated at 4.5 CFU/ml surviving pasteurization for 3.5 min at 60°C (Table 3). In this case, strain #12 had the
D
highest levels of saturated fatty acids (73%) and the lowest levels of unsaturates (27%), yielding the
TE
statistically lowest unsaturate/saturate ratio of 39%. On the other hand, heat-sensitive strain #3
EP
(serovar 4,12:i:-) contained the highest levels of unsaturates (43.5%) as well as the highest unsaturate/saturate ratio at 81%. Strain #3 also contained the lowest concentration (0.38%) of stearic
AC C
acid (18:0), which is longest saturate detected, and is a fatty acid that is solid at room temperature. For other strains, no correlations between fatty acids and thermal resistance could be found. Other factors, which might explain this lack of correlation include the presence of heat shock proteins and the production of EPS or cell clumping.
3.6 Scanning electron microscopy Scanning electron microscopy revealed that some heat-resistant strains (strain #’s 2, 10 and 20) had a greater propensity for attachment to glass slides or egg shells and cell clumping, while other heat-sensitive strains (strain #’s 1, 3 and 5) did not have increased attachment or cell clumping. The
12
ACCEPTED MANUSCRIPT 13 morphological characteristic of cell clumping could potentially aid in increasing the heat resistance of a bacterial strain. Note that we classified strain #10 as a heat resistant strain in this case based on its ability to be the only strain out of 20 strains capable of surviving pasteurization at 60°C for 3.5 min for
SC R
IP T
each of 4 replicate experiments when inoculated at 4.5 log CFU/ml (see results displayed in Table 3).
4. Discussion
NU
This data represents the first steps in determining whether Salmonella contamination in
MA
pasteurized egg products may be the result of either thermally-resistant isolates or of post-processing contamination from heat-sensitive isolates. The significance of this study lies in the fact that the 2.7
D
billion pounds of liquid egg produced in the U.S. every year are considered ready-to-eat products,
TE
often served without further heating in food items such as smoothies, milk shakes, protein drinks, ice
EP
cream, eggnog, cookie dough, Hollandaise sauce, traditional mousse, mayonnaise, and salad dressings. Further, pasteurized liquid egg products occasionally test positive for Salmonella and have been
AC C
associated with known foodborne outbreaks of salmonellosis (Blais et al., 1998; CDC, 2012; Durham County, 2010; Gurtler et al., 2013; Hara-Kudo and Takatori, 2009; Suzuki and Yamamoto, 2009; UKFSA, 2007a, 2007b). Additionally, the FSIS Microbiological Testing Program for Salmonella in Pasteurized Egg Products has detected Salmonella in liquid pasteurized egg at rates of 0.59-0.66% in liquid albumen from 2011 – 2013 (FSIS, 2014). Also, from 1995 – 2013 the agency detected Salmonella in 0.51% of liquid whole egg, liquid egg yolk or dried yellow egg products (FSIS, 2014). Further, the USDA-FSIS estimates that up to 5,500 cases of salmonellosis may occur in the U.S. each year from liquid pasteurized egg products (Latimer et al., 2008).
13
ACCEPTED MANUSCRIPT 14 In the present study, strains of Salmonella with high heat resistance, (isolate #’s 2, 6, 10 and 12) may indicate the ability of Salmonella to survive pasteurization, while strains which were comparatively heat-senstive (i.e., isolate #’s 1, 3, 5, 7, 8, 11, and 15) may likely be present in the
IP T
liquid egg products as the result of post-processing contamination.
SC R
Two of the isolates used in the present study were provided by the U.S. FDA, and were included in the FDA’s published pasteurization studies (Shah et al., 1991). These strains are SE PT8 C398 and SE PT8 C405, originally isolated from a New England egg-related salmonellosis outbreak, and from
NU
the ovary of a laying hen, respectively. Shah et al. (1991) reported that these two isolates were highly
MA
heat resistant when compared with all 17 Salmonella isolates heated in liquid whole egg at 57.2°C. Salmonella Oranienberg DD2229 was included in our present study as a heat-resistant isolate of
D
Salmonella that had been used in our previously-published egg pasteurization studies (Gurtler et al.,
TE
2011, 2013; Jordan et al., 2011). Historically, serovar Oranienberg has been isolated from eggs and
EP
has been included in published egg pasteurization studies going back 70 years (Anellis et al., 1954; Cantor and McFarlane, 1949; Cotterill, 1968; Cotterill and Glauert, 1969, 1972; Cotterill et al., 1973;
AC C
Gibbons and Moore, 1944; McBee and Cotterill, 1971; Schneider, 1946; Solowey et al., 1946, 1949; Winter et al., 1946). Our strain of Oranienberg was originally provided to us by Dupont from their Salmonella library (Jensen and Hubner, 1996). It is interesting to note that in the FSIS (2014) data (on Salmonella in Pasteurized Egg Products Testing Program: Serotypes, CY 1995 – 2013) one of our serovars (Widemarsh, isolated from pasteurized dried egg whites) isolated from liquid egg doesn’t show up on the list out of the 19 most common serovars isolated from liquid egg products. ). From a literature search of SCOPUS and Pubmed, Salmonella Widemarsh appears to be a very rare serovar.
14
ACCEPTED MANUSCRIPT 15 Regarding fatty acid analysis, Alvarez-Ordonez et al. (2008), in testing Salmonella Typhimurium for thermal resistance at 58°C, found that D-values were maximum for cells with low UFA/SFA ratios, and, consequently, with low membrane fluidity. Other studies have reported similar findings
IP T
(Alvarez-Ordonez et al., 2009; Yang et al., 2014A, 2014B). We found a similar trend with one heat
SC R
resistant strain (#13) that had the statistically lowest unsaturated/saturate ratio at 39%. Further, one heat sensitive strain (#3) had the highest unsaturated/saturate ratio at 81%, and also the lowest concentration of stearic acid. For the other 18 strains, the proportion of saturates to unsaturates in the
NU
bacterial cell showed no correlation with heat resistance for the Salmonella isolates tested in this
MA
study. Other cellular characteristics, which might account for thermal resistance include the presence of heat shock proteins, or the production of EPS and cell clumping. Regarding electron microscopy,
D
we found that three of the most heat-resistant strains (strain #’s 2, 10 and 20) were also three with the
TE
highest levels of attachment and cell clumping (Figures 2 and 3). It is known that the thermal
EP
resistance of bacterial spores increases commensurate to spore clumping (Aiba and Toda, 1966; Cerf, 1977; Furukawa et al., 2005; Pflug et al., 2001; Sakaguchi and Amaha, 1951; Stumbo, 1965; Toda and
AC C
Aiba, 1966). A similar mechanism may also come into play with regard to the heat resistance of vegetative cells. While this is a preliminary study to determine if cells clump when grown under laboratory conditions, we recognize that results may be different if cells were grown in LWE. Future studies should evaluate these salmonellae for clumping when grown in LWE. Further, Davey et al. (2001) have called into question the validity of assuming that cell clumping may affect the thermal death time of bacterial cells. The authors state that the Davey model shows that the thermal come-uptime for a clump mass is negligible when compared to typical exposure times during a thermal treatment. Nevertheless, the authors also state that clumping will lead to clumps of cells producing only single colonies on an agar plate, thus potentially underestimating the actual CFU/ml of a
15
ACCEPTED MANUSCRIPT 16 menstruum. Future studies should evaluate these salmonellae for thermal resistance when grown in LWE. The goal of this study was to evaluate twenty strains of Salmonella for variations in thermal
IP T
resistance, biochemistry, serology, and fatty acid profiles. Isolates were serotyped as Heidelberg (4
SC R
isolates), Widemarsh, Mbandaka, Cerro, Thompson, 4,12:i:-, and Enteritidis (8 isolates). These results are not surprising as Enteritidis and Heidelberg are the two most commonly isolated Salmonella serovars from poultry laying flocks (FSIS, 2014), and in this case, comprised 71% of serovars isolated
NU
from pasteurized egg products in this study. It is also not surprising that 7 of 8 Enterititis isolates were
MA
phage type 8, as PT-8 is the most common phage type among Enterititidis isolates in laying flocks in the U.S.
D
The D60 values of the 20 strains in LWE ranged from 0.34 – 0.58 min. All 17 strains survived
TE
pasteurization at 60°C for 3.5 min when inoculated at 8.5 log CFU/mL as determined by enrichment;
EP
however, when inoculated at 4.5 log CFU/mL some strains did not survive pasteurization at 60°C for 3.5 min. Although some strains exhibited atypical enzymatic activity (e.g., reduction of adonitol,
AC C
hydrolosis of proline nitroanilide or p-n-p-beta-glucuronide, and nonreduction of melibiose) these differences in biochemical reactions could not be correlated with similarity in thermal resistance. Additionally FAME analysis revealed that fatty acid profiles may be correlated with heat resistance of some Salmonella strains. It is to be noted that the growth conditions for salmonellae may also play a role in thermal resistance. Using cells that have previously been grown in TSB may not be fully representative of the heat resistance of cells grown in liquid egg. Nevertheless, previously published studies have used liquid media to grow salmonellae prior to determining its thermal resistance in liquid egg (Froning et al., 2002). The data we present here represents the first steps in determining whether Salmonella contamination in pasteurized egg products may be the result of either thermally-
16
ACCEPTED MANUSCRIPT 17 resistant isolates or post-processing contamination. Strains with higher heat resistance, (e.g., isolate #’s 2, 6, 10 and 12) may indicate the ability of Salmonella to survive pasteurization, while strains with lower heat resistance (e.g., isolate #’s 1, 3, 5, 7, 8, 11, and 15) may be present in egg products as the
IP T
result of post-processing contamination. These findings may be useful in the production of LWE that
SC R
are not contaminated by this foodborne pathogen.
ACKNOWLEDGEMENTS
NU
The authors would like to thank Drs. David Geveke and Brendan Niemira for kindly reviewing
MA
this manuscript prior to submission for publication. Special thanks to Ms. Kimberly Ingram for
AC C
EP
TE
D
providing technical support in the completion of the MIDI fatty acid analyses.
17
ACCEPTED MANUSCRIPT 18
TABLE 1. Salmonella strains used in this study and results of USDA, APHIS, NVSL serotyping and phage typing.
‡
4,12:i:-
FSIS 4 FSIS 5 FSIS 6
*
Pasteurized Frozen Salt Yolk Pasteurized Salt Yolk
Cerro
‡
Liquid Pasteurized Sugar Yolk
Heidelberg
Pasteurized Frozen Salt Yolk
Enteritidis, Phage type 2 Heidelberg
‡
Mbandaka
MA
FSIS 7‡ FSIS 8
Pasteurized Frozen Salt Yolk
Enteritidis, Phage type 8
Pasteurized Frozen Whites
NU
FSIS 3
Source of Isolate
IP T
FSIS 2*
Strain Information Heidelberg
SC R
Strain # FSIS 1‡
Pasteurized Liquid Whites Pasteurized Whole Egg
FSIS 9
Enteritidis, Phage type 8
Pasteurized Egg Whites
FSIS 10
Enteritidis, Phage type 8
Pasteurized Liquid Salted Yolk
D
FSIS 11
‡
Heidelberg
Scramble Egg Mix
Enteritidis, Phage type 8
Pasteurized Frozen Whole Egg
FSIS 13
Enteritidis, Phage type 8
Pasteurized Egg Whites
FSIS 16 FSIS 17 FDA 18
†
EP
FSIS 15
‡
Thompson
Pasteurized Egg Whites
Widemarsh
Pasteurized Dried Egg Whites
AC C
FSIS 14
TE
FSIS 12
Enteritidis, Phage type 8
Pasteurized Egg Whites
Enteritidis, Phage type 8
Pasteurized Frozen Salt Yolk Enteritidis, FDA C405, Phage type 8 Egg yolk, New England salmonellosis outbreak
FDA 19† DuPont 20†*
Enteritidis, FDA C398, Phage type 8 Hen’s Ovary, University of Pennsylvania Oranienberg, Strain DD2229
DuPont culture collection
†
Isolates not isolated from pasteurized egg but were included in published liquid egg pasteurization studies (Gurtler et al., 2011; Jordan et al., 2011; Gurtler et al., 2013). * One of the three most heat-resistant of Salmonella isolates. ‡ One of the seven most heat-sensitive of Salmonella isolates.
18
ACCEPTED MANUSCRIPT 19
1
0.34†‡ E
0.958
-2.904
2
0.55* A
0.955
-1.810
3
0.34‡ E
0.993
-2.926
4
0.45 CD
0.9834
-2.222
8.441
5
0.35 ‡ E
0.976
-2.975
8.704
6
0.58* A
0.964
-1.733
8.079
7
0.34 ‡ E
0.976
-2.930
8.565
8
0.36 ‡ E
0.995
-2.795
8.234
9
0.44 CD
0.989
-2.272
8.757
10
0.40 DE
0.981
-2.496
8.9049
11
0.35‡ E
0.970
-2.856
8.904
12
0.43 CD
0.984
-2.315
9.835
13
0.46 BCD
TE
TABLE 2. Thermal inactivation kinetics of salmonellae in liquid whole egg, inoculated at 8.5 log CFU/mL and pasteurized at 60°C by hot water-immersion capillary tube method. Strain # D-value (min)§ R2 Slope y-intercept
0.988
-2.195
9.031
14
0.40 DE
0.988
-2.525
8.681
15
0.35 ‡ E
0.994
-2.872
8.587
0.47 BC
0.997
-2.144
8.630
0.44 CD
0.985
-2. 244
8.565
0.42 DE
0.973
-2.439
8.837
19
0.45 CD
0.979
-2.233
9.215
20
0.55 * AB
0.961
-1.833
8.674
Ave.
0.42
17 18
IP T
SC R
NU MA
D
EP
AC C
16
9.205 8.220 8.619
§
Values in the same column not followed by the same letter are significantly different (p
S0168-1605(15)00202-0 doi: 10.1016/j.ijfoodmicro.2015.04.010 FOOD 6875
To appear in:
International Journal of Food Microbiology
Received date: Revised date: Accepted date:
3 October 2014 31 March 2015 4 April 2015
Please cite this article as: Gurtler, Joshua B., Hinton Jr., Arthur, Bailey, Rebecca B., Cray Jr., William C., Jin, Tony Z., Salmonella Isolated from Ready-to-Eat Pasteurized Liquid Egg Products: Thermal Resistance, Biochemical Profile, and Fatty Acid Analysis, International Journal of Food Microbiology (2015), doi: 10.1016/j.ijfoodmicro.2015.04.010
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1
Salmonella Isolated from Ready-to-Eat Pasteurized Liquid Egg
IP T
Products: Thermal Resistance, Biochemical Profile, and Fatty Acid
SC R
Analysis
Joshua B. Gurtler*1, Arthur Hinton, Jr.3, Rebecca B. Bailey1, William C. Cray, Jr. 2,
MA
1
NU
Tony Z. Jin4
Food Safety and Intervention Technologies Research Unit, U.S. Department of Agriculture,
D
Agricultural Research Service, Eastern Regional Research Center, 600 East Mermaid Lane,
TE
Wyndmoor, PA 19038-8551
U.S. Department of Agriculture, Food Safety and Inspection Service, Athens, GA
3
U.S. Department of Agriculture, Agricultural Research Service,Richard B. Russell Research Center,
4
AC C
Athens, GA
EP
2
Residue Chemistry and Predictive Microbiology, U.S. Department of Agriculture, Agricultural
Research Service, Eastern Regional Research Center, Wyndmoor, PA
Running title: Salmonella from pasteurized egg products Key Words: Salmonella, liquid egg products, pasteurization
* Corresponding author. Tel.: 215-233-6788; fax: 215-233-6406. E-mail address: [email protected] (J.B. Gurtler).
1
ACCEPTED MANUSCRIPT 2
ABSTRACT (362 Words) _________________________________________________________________________
IP T
The Egg Products Inspection Act of 1970 requires that egg products in the U.S. must be pasteurized
SC R
prior to release into commerce. The USDA Food Safety and Inspection Service (FSIS) is responsible for regulating egg products. Salmonellae are infrequently isolated from pasteurized egg products by food manufacturers or the FSIS and may be present as a result of either pasteurization-resistant
NU
bacteria or post-processing contamination. In this study, seventeen strains of Salmonella isolated from
MA
pasteurized egg products and three heat-resistant control strains were compared for the following attributes: thermal resistance in liquid whole egg (LWE) at 60°C, enzymatic profiles, and serotyping
D
and phage typing,. fatty acid analysis and strain morphological variation evaluated by scanning
TE
electron microscopy. Isolates were serotyped as Heidelberg (4 isolates), Widemarsh, Mbandaka,
EP
Cerro, Thompson, 4,12:i:-, and Enteritidis (8 isolates). The D60 values in LWE ranged from 0.34 – 0.58 min. All 20 strains were recovered from LWE inoculated with 8.5 log CFU/ml of Salmonella and
AC C
pasteurized at 60°C for 3.5 min; however, some isolates were not recovered from pasteurized LWE that had been inoculated with only 4.5 log CFU/ml Salmonella and treated at 60°C for 3.5 min. Although some strains exhibited atypical enzymatic activity (e.g., reduction of adonitol, hydrolysis of proline nitroanilide or p-n-p-beta-glucuronide, and nonreduction of melibiose), differences in biochemical reactions could not be correlated with differences in thermal resistance. Furthermore, fatty acid analysis revealed that differences insaturate/unsaturated profiles may be correlated with differences in heat resistance, in two instances. One heat resistant strain (#13, Enteritidis) had the statistically lowest unsaturated/saturate ratio at 39%. However, one heat sensitive strain (#3, serovar 4,12:i:-) had the highest unsaturated/saturate ratio at 81%, and also the lowest concentration of stearic
2
ACCEPTED MANUSCRIPT 3 acid. This data represents the first steps in determining whether Salmonella contamination in pasteurized egg products may be the result of either thermally-resistant isolates or post-processing contamination. Contamination of LWE by Salmonella strains with higher heat resistance, (e.g., isolate
IP T
#’s 2, 6, 10 and 12) may indicate the ability of Salmonella to survive pasteurization, while
SC R
contamination of LWE strains with lower heat resistance (e.g., isolate #’s 1, 3, 5, 7, 8, 11, and 15) may indicate post-processing contamination of LWE by this foodborne pathogen.
MA
NU
1. INTRODUCTION
The Egg Products Inspection Act of 1970 requires that egg products in the U.S. must be pasteurized
D
prior to release into commerce, resulting in ca. 2.7 billion pounds of liquid egg (LE) pasteurized for
TE
human consumption in the U.S. every year (USDA, NASS, 2010). The USDA Food Safety and
EP
Inspection Service (FSIS) is responsible for regulating egg products, and requires liquid whole egg (LWE) to be pasteurized at 60°C for a minimum of 3.5 min, after which it may be served to consumers
AC C
with no further interventions to inactivate bacteria (9 CFR 590.570, Table 1 [Code of Federal Regulations, 2009]). Nevertheless, salmonellae are infrequently isolated from these pasteurized egg products by food manufacturers or the FSIS. Contamination of these egg products may be the result of either pasteurization-resistant bacteria, which can survive the heating process, or as a result of postprocessing contamination of pasteurized LWE by Salmonella (Blais et al., 1998; Suzuki and Yamamoto, 2009). Sporadic outbreaks of salmonellosis in the U.S. have been potentially associated with pasteurized egg products from the consumption of food items such as crab cakes, scrambled eggs, chile relleno, and a meringue-topped dessert (CDC, 2012; Durham County, 2010; Gurtler et al., 2013). Other known outbreaks of Salmonella from pasteurized egg products have been reported overseas
3
ACCEPTED MANUSCRIPT 4 (Hara-Kudo and Takatori, 2009; UKFSA, 2007a, 2007b). The USDA-FSIS has published a risk assessment model (FSIS, 2005, Exec. Summary) in which they estimate that there may be approximately 5,500 U.S. cases of salmonellosis per year that stem from contaminated pasteurized LE
IP T
products (Latimer et al., 2008). A series of recently published studies have reevaluated the
SC R
time/temperature pasteurization requirements for some LE products (Gurtler et al., 2011, 2013; Jordan et al., 2011). One study (Gurtler et al., 2013) predicted that, based on current FSIS pasteurization requirements for 10% salted LWE (63.3°C for 3.5 min), only an ca. 3.6 log reduction of Salmonella
NU
would be achieved. In the same study the authors determined that to achieve a 5 log reduction, the
MA
pasteurization requirements would need to be increased to 62.2°C for 7.5 min. In another study, the authors concluded that, based on current FSIS pasteurization requirements for 10% salted liquid egg
D
yolk (63.3°C for 3.5 min), only an ca. 2.7 log reduction of Salmonella would be achieved (Gurtler et
TE
al., 2011). In order to achieve a desired 5 log reduction, the same authors determined that the
EP
pasteurization temperature would need to be raised to 67.3°C for 3.5 min. The goal of the present study was to evaluate seventeen selected strains of Salmonella that had been isolated from ready-to-eat
AC C
pasteurized liquid egg products for variations in thermal resistance, biochemistry, serology, morphology, and fatty acid profiles. Additionally, three Salmonella isolates that were included in published liquid egg pasteurization studies were included in these same battery of tests for comparison purposes. The overall goal of the study was to identify determinants that might suggest whether the seventeen salmonellae strains isolated from pasteurized liquid egg products might be there as a result of heat-resistant survival of the thermal pasteurization process, or whether strains are heat-sensitive, suggesting that they might be present in the egg products as a result of post-processing contamination.
2. MATERIALS AND METHODS
4
ACCEPTED MANUSCRIPT 5
2.1 Determination of thermal D-values Seventeen strains of Salmonella isolated from pasteurized egg products by the USDA-FSIS were
IP T
used in this study (Table 1). Additionally, two strains of Salmonella Enteritidis phage type 8 (C398
SC R
and C405, provided by Keith Lampel, FDA, College Park, MD) and one strain of Salmonella Oranienburg DD2229 from the Dupont Salmonella collection were included for comparison purposes. These strains were used in published USDA-ARS/FSIS egg pasteurization studies (Table 1) (Gurtler et
NU
al., 2011, 2013; Jordan et al., 2011). Isolates were tested for thermal inactivation kinetics during 3
MA
repetitions in LWE at 60°C with heating times up to 5 min using the hot water-immersion capillary tube method. The come-up-time was 6 s, while the cool-down-time was 3 s. Salmonella strains were
D
individually incubated in 10mL of tryptic soy broth (TSB) (Difco, Sparks, MD) with 50 ppm nalidixic
TE
acid (NA) at 37°C for 24 h. A single 10µL loop transfer was performed and suspensions were
EP
incubated for another 24 h to reach populations of ca. 9 log CFU/mL. Suspensions were plated on tryptic soy agar (TSA) (Difco) with 0.1% sodium pyruvate and 50 ppm NA (TSAPN). Isolates were
AC C
individually centrifuged at 22°C for 10 min, the supernatant was decanted, and the bacterial pellet was resuspended in 0.4mL of 0.1% peptone water in order to produce a suspension with a population of ca. 9 log CFU of Salmonella/mL. Salmonella suspensions (0.5 mL from each of three strains) were individually added to 38mL of LWE at 21°C and mixed for 2 min. Approximately 200µL of inoculated LWE was injected into 250µL capillary tubes (Accu-Fill 90 Micropet Disposable Pipettes, Clay Adams, Inc., Parsippany, NJ) and the ends of the tubes were heat-sealed using an open flame. Capillary tubes were then placed in a stainless-steel test tube rack, by inserting the capillary tubes through the holes of attached aluminum tape. The entire test tube rack with capillary tubes was then immersed in a circulating water bath (ThermoFisher Scientific NESLAB RTE17 Digital Plus) set to
5
ACCEPTED MANUSCRIPT 6 60°C. To monitor the come-up time and temperature, Digisense Thermocouple Flexible High Temperature Wire Probes (Cole-Parmer, Vernon Hills, IL) connected to a Fluke 54 II Thermometer (Fluke Calibration, Everett, WA) were placed in the water bath and affixed inside a capillary tube
IP T
containing LWE. After the capillary tubes reached 60°C, they were sampled at 1 min intervals for up
SC R
to 5 min, then immersed in ice water for 10 s. Cooled tubes were then removed from the ice water, surface-sanitized in 70% ethanol, and rinsed twice with sterile-water. The ends of capillary tubes were then removed with ethanol-flamed wire cutters and the contents were expelled manually with a sterile
NU
4-mm pipe cleaner (Creativity Street, 4 mm Chenille Stems Pipe Cleaners, The Chenille Kraft
MA
Company, Gurnee, IL). Serial dilutions of the LWE were performed, dilutions were spiral plated in duplicate on TSAPN and incubated at 37°C for 24 h. Undiluted egg samples (50 µL) were spread
D
plated in duplicate on TSAPN, yielding a minimum detection limit of 10 CFU/ml. Colonies were
TE
counted using a Synbiosis aCOLyte Supercount (Microbiology International, Frederick, MD).
EP
Presumptive-positive Salmonella were randomly confirmed by replica-plating on XLD agar (Difco) and serological agglutination (Salmonella O Antiserum Poly A-I & Vi, Difco, Sparks, MD). D-values
AC C
were calculated from the negative inverse of the slope of the survival curve (Table 2).
2.2 Test tube heating study
Four repetitions of a second thermal study were conducted to determine the ability of various populations of Salmonella to survive pasteurization in LWE at the FSIS-mandated 60°C for 3.5 min by hot water immersion in test tubes. LWE inoculated to 3.5, 4.5, or 8.5 log CFU/mL was added (1.2 mL) to 5 mL test tubes. A thermocouple was placed in uninoculated LWE in one test tube to determine the come-up-time. After the egg in the tubes reached 60°C, all tubes were held for an additional 3.5 min, whereupon they were immediately immersed in an ice water bath. One mL of
6
ACCEPTED MANUSCRIPT 7 pasteurized egg was then removed and added to 9 mL of TSB + 35mg/L ferrous sulfate and 0.1% sodium pyruvate (TSB-FP), for a primary enrichment. Ferrous sulfate and sodium pyruvate were added to assist in the recovery of injured cells (Gurtler and Kornacki, 2009). After incubating primary
IP T
enrichments for 24 h at 37°C, selective enrichments were then made by adding 1 mL of the TSB-FP
SC R
primary enrichment to 9 mL of tetrathionate (TT) Broth (Difco) as well as to 9 mL of Rappaport Vassiliadis broth (Difco) and incubated at 42°C for 24 h. Tetrathionate Broth Base (Difco) was supplemented with 100 ppm nalidixic acid + 20 mL/L of iodine solution + 0.01 g/L of brilliant green
NU
(Sigma) which served as a selective enrichment for Salmonella along with RV broth. A 10 µL loop of
MA
both TT broth and RV broth were then streaked to separate plates of XLD agar and incubated at 37°C
D
for 24 h. Presumptive-positive Salmonella were randomly confirmed by serological agglutination.
TE
2.3 Enzymatic and serological profiles, phage typing, and fatty acid analysis
EP
Enzymatic profiles were determined by using the BBL Crystal Identification System Autoreader (Becton, Dickinson and Company, Franklin Lakes, NJ). Strains were serotyped and phage typed at the
AC C
USDA, APHIS, National Veterinary Service Laboratory in Ames, Iowa. Fatty acid analyses of salmonellae were conducted by gas chromatography and FAME MIDI Sherlock software, following the manufacturer’s directions for cell preparation (MIDI, Newark, DE) as previously published (Hinton, et al., 2004). Three replicate cultures of each Salmonella isolate were prepared, by growth on BBL Trypticase™ Soy Agar (Becton, Dickson and Company, Sparks, MD) at 30°C for 24 hr and extraction of fatty acids from cell walls of the bacteria and methyl ester generation were performed using the Sherlock Instant FAME Method (MIDI, Inc.). GC analysis of the cell extracts was performed with the Hewlett Packard HP 6890 Series GC System that was computer operated by Aligent ChemStation software (Aligent Technologies, Santa Clara, CA) and fatty acid data was analyzed by
7
ACCEPTED MANUSCRIPT 8 MIDI Microbial Identification Sherlock software (MIDI). MIDI Sherlock calibration standards were included with each GC analysis. Briefly, after three consecutivetransfers and incubation, cultures were harvested and fatty acids were extracted from the cellular membranes using the four-step
IP T
extraction procedure recommended by the manufacturer. The first step of the extraction procedure
SC R
consisted of saponification by heating the cells at 100°C for 30 min in a methanolic sodium hydroxide solution composed of 15% (wt/vol) NaOH (Spectrum Chemical Manufacturing Corp., Gardena, Calif.) dissolved in 50% (vol/vol) methanol (Sigma-Aldrich, Inc., St. Louis, Mo.). Saponification lysed the
NU
bacteria and released cellular fatty acids. Methylation of released fatty acids was accomplished by
MA
mixing the resultant suspension in a solution of 3.25 N hydrochloric acid (Ricca Chemical Co., Arlington, Tex.) in methanol (46%, vol/vol) at 80°C for 10 min. The fatty acid methyl esters (FAMEs)
D
were extracted from the aqueous phase to the organic phase of the mixture by a 10-min liquid-liquid
TE
extraction in a 1:1 solution of hexane (Sigma-Aldrich, Inc.) and methyl-tert-butyl ether (Sigma-
EP
Aldrich, Inc.). A 5-min base wash of the FAME extracts in a sodium hydroxide solution (1.08%, wt/vol) was performed to remove free fatty acids and residue from the organic extract. FAME extracts
AC C
were transferred to sample vials for identification and quantification by an Agilent 6890 gas chromatograph (GC) (Agilent Technologies.). The GC was operated by a computer containing ChemStation (Agilent Technologies) and MIDI Sherlock MIS, version 4.5, software. ChemStation software set the GC parameters of operation and controlled GC functions throughout FAME analysis. The MIS software identified and quantified FAMEs of bacterial isolates. Additionally, all isolates were plated out for isolation on XLD and XLT4, agar, to determine if they displayed typical reaction on these common salmonellae media.
2.4 Scanning electron microscopy
8
ACCEPTED MANUSCRIPT 9 Prior to scanning electron microscopy, cells were prepared by immersing glass cover slips as well as shell egg chips (ca. 2 x 2 cm) in 3 mL of TSB and individually inoculating with 20 strains of Salmonella, respectively. Glass cover slips and egg shells were then incubated for 24 h at 37°C in
IP T
order to allow for growth and attachment of salmonellae. Glass cover slips and egg shells were then
SC R
removed from the TSB suspension and rinsed with glutaraldehyde. Cover slips and eggs shells were fixed for scanning electron microscopy (SEM) by immersion in a 2.5% glutaraldehyde-0.1 M imidazole buffer (Electron Microscope Sciences, Hatfield, PA) for 1 h before washing in imidazole
NU
buffer and dehydrating in 50%, 80% and absolute ethanol, successively. Samples were critical point
MA
dried (Denton Vacuum, Cherry Hill, NJ) with carbon dioxide, mounted with Duco cement (ITW Performance Polymers, Riviera Beach, FL) and colloidal silver adhesive, and sputter-coated with a
D
thin layer of gold using a Scancoat Six Sputter Coater (BOC Edwards, Wilmington, MA). Samples
TE
were imaged with a Quanta200 FEG environmental scanning electron microscope (FEI Co., Inc.,
EP
Hillsboro, OR), with an Everhart Thornley detector, operated in the high vacuum, secondary electron imaging mode at an accelerating voltage of 5 kV.
AC C
2.5 Statistical analyses
Data were analyzed by ANOVA, and significant differences (p < 0.05) between D60°C values as well as differences in fatty acid composition were determined post hoc by Fisher’s Least Significant Difference Test (LSD) via SAS version 9.1 (SAS Institute, Inc., Cary, NC). 3. Results 3.1 Serotyping and phage typing The seventeen isolates originating from pasteurized egg products were serotyped as (Table 1) Enteritidis (8 isolates, 47%), Heidelberg (4 isolates, 24%), and one each for serovars 4,12:i:-,
9
ACCEPTED MANUSCRIPT 10 Widemarsh, Mbandaka, Cerro, and Thompson. All Enteritidis isolates phage typed out as PT-8 except for one strain that was PT-2.
IP T
3.2 Determination of thermal D60 values
SC R
The thermal D60 values in LWE at 60°C over a 5 min heating period, as determined by three replicate experiments, ranged from 0.34 – 0.58 min (Table 2). The seven least heat-resistant serovars with thermal D60 values of from 0.34 - 0.36 min were four strains of S. Heidelberg, and one strain each
NU
of serovars 4,12:i:-, Widemarsh and Mbandaka. These results are not surprising as S. Heidelberg is
MA
known to not be especially heat-resistant. The three most heat-resistant serovars with thermal D60 values of from 0.55 – 0.58 min were two strains of Enteritidis (isolated from pasteurized frozen salt
D
yolk and from pasteurized frozen egg whites) as well as the Dupont Oranienberg DD2229 isolate.
TE
These results support the prior inclusion of the Oranienberg DD2229 isolate in previous liquid egg
EP
thermal inactivation studies (Gurtler et al., 2011; 2013; Jordan et al., 2011;); however, the two FDA Enteritidis strains, which were also included in the pasteurization studies, only had thermal D60 values
AC C
of 0.42 and 0.45 min. Thermal inactivation curves are graphically displayed in figures 1 and 2. Figure 1 contains results from the Salmonella isolates with the highest D-values, where all strains were viable after 3 min of pasteurization. In contrast, Figure 2 displays strains with the lowest D-values , in which 5 isolates reached non-detectable levels at 3 min of pasteurization.
3.3 Test tube heating studies (60°C for 3.5 min) All 20 strains survived pasteurization at 60°C for 3.5 min during two repetitions, when inoculated at 8.5 CFU/mL, as determined by enrichment (Table 3). Conversely, none of the 20 strains survived the same pasteurization regimen when LWE was inoculated at 3.5 log CFU/mL (minimum detection
10
ACCEPTED MANUSCRIPT 11 limit = 1 CFU/mL). However, when inoculated at 4.5 log CFU/mL and heated at 60°C for 3.5 min, over 4 repetitions, the previously determined three most heat-resistant serovars (Enteritidis strain #’s 2, 6, and Oranienberg DD2229) survived at rates of 3/4, 2/4, and 3/4, respectively. On the other hand,
IP T
the seven most heat-sensitive isolates (as determined by D-values) only survived pasteurization at
SC R
60°C for 3.5 min at rates of 1/4, 2/4, 1/4, 1/4, 1/4, 1/4, and 2/4, respectively.
3.4 Biochemical profile
NU
Data on biochemical profiles of all 20 strains of Salmonella were attained by use of the BBL
MA
Crystal Identification System Autoreader, which tests for 32 separate biochemical reactions (Table 4). Although some strains exhibited atypical enzymatic activity (e.g., reduction of adonitol, hydrolosis of
D
proline nitroanilide and p-n-p-beta-glucuronide, and nonreduction of melibiose) no differences in
TE
biochemical reactions could be correlated with similarity in thermal resistance. When plated on XLD
EP
and XLT4 agars, all strains displayed typical black colonies on XLD agar; however, ten strains out of twenty did not blacken the colonies on XLT4 agars, and remained white. These were strain #2
AC C
(Enteritidis PT-8), #6 (Enteritidis PT-2), #9 (Enteritidis PT-8), #10 (Enteritidis PT-8), #12 (Enteritidis PT-8), #13 (Enteritidis PT-8), #16 (Enteritidis PT-8), #17 (Enteritidis PT-8), #18 (Enteritidis PT-8), and #20 (Oranienburg). XLT4 and XLD agars both contain an H2S indicator system in which reduction of sodium thiosulfate produces hydrogen sulfide which is detected by the addition of ferric ions to form black colonies. Both XLD and XLT4 agars contained equivalent amounds of L-lysine, ferric ammonium citrate and sodium thiosulfate. Thus, it is not apparent why ten Salmonella isolates (nine (9) Enteritidis and one Oraniengburg) colonies failed to turn the characteristic blackened center on XLT4, while all strains produced blackened centers on XLD. Hence, XLT4 agar should be used with caution when used as a cultural confirmation based on blackening center of Salmonella colonies.
11
ACCEPTED MANUSCRIPT 12 Of colonies that did not blacken on XLT4, 9 of ten were S. Enteritidis, while only one S. Enteritidis blackened colonies on the medium; namely Enteritidis, FDA C398, Phage type 8, from a hen’s ovary,
IP T
University of Pennsylvania.
SC R
3.5 Fatty acid methyl ester analysis
While it has been hypothesized that greater concentrations of saturated fatty acids within the bacterial cell membrane may contribute to increased heat resistance (Alvarez-Ordonez et al., 2008;
NU
2009; Yang et al., 2014a; 2014b), our analysis suggests this may be true for at least two isolates (Table
MA
5). One example may be the heat-resistant S. Enteritidis #12, as assessed by 3 of 4 samples inoculated at 4.5 CFU/ml surviving pasteurization for 3.5 min at 60°C (Table 3). In this case, strain #12 had the
D
highest levels of saturated fatty acids (73%) and the lowest levels of unsaturates (27%), yielding the
TE
statistically lowest unsaturate/saturate ratio of 39%. On the other hand, heat-sensitive strain #3
EP
(serovar 4,12:i:-) contained the highest levels of unsaturates (43.5%) as well as the highest unsaturate/saturate ratio at 81%. Strain #3 also contained the lowest concentration (0.38%) of stearic
AC C
acid (18:0), which is longest saturate detected, and is a fatty acid that is solid at room temperature. For other strains, no correlations between fatty acids and thermal resistance could be found. Other factors, which might explain this lack of correlation include the presence of heat shock proteins and the production of EPS or cell clumping.
3.6 Scanning electron microscopy Scanning electron microscopy revealed that some heat-resistant strains (strain #’s 2, 10 and 20) had a greater propensity for attachment to glass slides or egg shells and cell clumping, while other heat-sensitive strains (strain #’s 1, 3 and 5) did not have increased attachment or cell clumping. The
12
ACCEPTED MANUSCRIPT 13 morphological characteristic of cell clumping could potentially aid in increasing the heat resistance of a bacterial strain. Note that we classified strain #10 as a heat resistant strain in this case based on its ability to be the only strain out of 20 strains capable of surviving pasteurization at 60°C for 3.5 min for
SC R
IP T
each of 4 replicate experiments when inoculated at 4.5 log CFU/ml (see results displayed in Table 3).
4. Discussion
NU
This data represents the first steps in determining whether Salmonella contamination in
MA
pasteurized egg products may be the result of either thermally-resistant isolates or of post-processing contamination from heat-sensitive isolates. The significance of this study lies in the fact that the 2.7
D
billion pounds of liquid egg produced in the U.S. every year are considered ready-to-eat products,
TE
often served without further heating in food items such as smoothies, milk shakes, protein drinks, ice
EP
cream, eggnog, cookie dough, Hollandaise sauce, traditional mousse, mayonnaise, and salad dressings. Further, pasteurized liquid egg products occasionally test positive for Salmonella and have been
AC C
associated with known foodborne outbreaks of salmonellosis (Blais et al., 1998; CDC, 2012; Durham County, 2010; Gurtler et al., 2013; Hara-Kudo and Takatori, 2009; Suzuki and Yamamoto, 2009; UKFSA, 2007a, 2007b). Additionally, the FSIS Microbiological Testing Program for Salmonella in Pasteurized Egg Products has detected Salmonella in liquid pasteurized egg at rates of 0.59-0.66% in liquid albumen from 2011 – 2013 (FSIS, 2014). Also, from 1995 – 2013 the agency detected Salmonella in 0.51% of liquid whole egg, liquid egg yolk or dried yellow egg products (FSIS, 2014). Further, the USDA-FSIS estimates that up to 5,500 cases of salmonellosis may occur in the U.S. each year from liquid pasteurized egg products (Latimer et al., 2008).
13
ACCEPTED MANUSCRIPT 14 In the present study, strains of Salmonella with high heat resistance, (isolate #’s 2, 6, 10 and 12) may indicate the ability of Salmonella to survive pasteurization, while strains which were comparatively heat-senstive (i.e., isolate #’s 1, 3, 5, 7, 8, 11, and 15) may likely be present in the
IP T
liquid egg products as the result of post-processing contamination.
SC R
Two of the isolates used in the present study were provided by the U.S. FDA, and were included in the FDA’s published pasteurization studies (Shah et al., 1991). These strains are SE PT8 C398 and SE PT8 C405, originally isolated from a New England egg-related salmonellosis outbreak, and from
NU
the ovary of a laying hen, respectively. Shah et al. (1991) reported that these two isolates were highly
MA
heat resistant when compared with all 17 Salmonella isolates heated in liquid whole egg at 57.2°C. Salmonella Oranienberg DD2229 was included in our present study as a heat-resistant isolate of
D
Salmonella that had been used in our previously-published egg pasteurization studies (Gurtler et al.,
TE
2011, 2013; Jordan et al., 2011). Historically, serovar Oranienberg has been isolated from eggs and
EP
has been included in published egg pasteurization studies going back 70 years (Anellis et al., 1954; Cantor and McFarlane, 1949; Cotterill, 1968; Cotterill and Glauert, 1969, 1972; Cotterill et al., 1973;
AC C
Gibbons and Moore, 1944; McBee and Cotterill, 1971; Schneider, 1946; Solowey et al., 1946, 1949; Winter et al., 1946). Our strain of Oranienberg was originally provided to us by Dupont from their Salmonella library (Jensen and Hubner, 1996). It is interesting to note that in the FSIS (2014) data (on Salmonella in Pasteurized Egg Products Testing Program: Serotypes, CY 1995 – 2013) one of our serovars (Widemarsh, isolated from pasteurized dried egg whites) isolated from liquid egg doesn’t show up on the list out of the 19 most common serovars isolated from liquid egg products. ). From a literature search of SCOPUS and Pubmed, Salmonella Widemarsh appears to be a very rare serovar.
14
ACCEPTED MANUSCRIPT 15 Regarding fatty acid analysis, Alvarez-Ordonez et al. (2008), in testing Salmonella Typhimurium for thermal resistance at 58°C, found that D-values were maximum for cells with low UFA/SFA ratios, and, consequently, with low membrane fluidity. Other studies have reported similar findings
IP T
(Alvarez-Ordonez et al., 2009; Yang et al., 2014A, 2014B). We found a similar trend with one heat
SC R
resistant strain (#13) that had the statistically lowest unsaturated/saturate ratio at 39%. Further, one heat sensitive strain (#3) had the highest unsaturated/saturate ratio at 81%, and also the lowest concentration of stearic acid. For the other 18 strains, the proportion of saturates to unsaturates in the
NU
bacterial cell showed no correlation with heat resistance for the Salmonella isolates tested in this
MA
study. Other cellular characteristics, which might account for thermal resistance include the presence of heat shock proteins, or the production of EPS and cell clumping. Regarding electron microscopy,
D
we found that three of the most heat-resistant strains (strain #’s 2, 10 and 20) were also three with the
TE
highest levels of attachment and cell clumping (Figures 2 and 3). It is known that the thermal
EP
resistance of bacterial spores increases commensurate to spore clumping (Aiba and Toda, 1966; Cerf, 1977; Furukawa et al., 2005; Pflug et al., 2001; Sakaguchi and Amaha, 1951; Stumbo, 1965; Toda and
AC C
Aiba, 1966). A similar mechanism may also come into play with regard to the heat resistance of vegetative cells. While this is a preliminary study to determine if cells clump when grown under laboratory conditions, we recognize that results may be different if cells were grown in LWE. Future studies should evaluate these salmonellae for clumping when grown in LWE. Further, Davey et al. (2001) have called into question the validity of assuming that cell clumping may affect the thermal death time of bacterial cells. The authors state that the Davey model shows that the thermal come-uptime for a clump mass is negligible when compared to typical exposure times during a thermal treatment. Nevertheless, the authors also state that clumping will lead to clumps of cells producing only single colonies on an agar plate, thus potentially underestimating the actual CFU/ml of a
15
ACCEPTED MANUSCRIPT 16 menstruum. Future studies should evaluate these salmonellae for thermal resistance when grown in LWE. The goal of this study was to evaluate twenty strains of Salmonella for variations in thermal
IP T
resistance, biochemistry, serology, and fatty acid profiles. Isolates were serotyped as Heidelberg (4
SC R
isolates), Widemarsh, Mbandaka, Cerro, Thompson, 4,12:i:-, and Enteritidis (8 isolates). These results are not surprising as Enteritidis and Heidelberg are the two most commonly isolated Salmonella serovars from poultry laying flocks (FSIS, 2014), and in this case, comprised 71% of serovars isolated
NU
from pasteurized egg products in this study. It is also not surprising that 7 of 8 Enterititis isolates were
MA
phage type 8, as PT-8 is the most common phage type among Enterititidis isolates in laying flocks in the U.S.
D
The D60 values of the 20 strains in LWE ranged from 0.34 – 0.58 min. All 17 strains survived
TE
pasteurization at 60°C for 3.5 min when inoculated at 8.5 log CFU/mL as determined by enrichment;
EP
however, when inoculated at 4.5 log CFU/mL some strains did not survive pasteurization at 60°C for 3.5 min. Although some strains exhibited atypical enzymatic activity (e.g., reduction of adonitol,
AC C
hydrolosis of proline nitroanilide or p-n-p-beta-glucuronide, and nonreduction of melibiose) these differences in biochemical reactions could not be correlated with similarity in thermal resistance. Additionally FAME analysis revealed that fatty acid profiles may be correlated with heat resistance of some Salmonella strains. It is to be noted that the growth conditions for salmonellae may also play a role in thermal resistance. Using cells that have previously been grown in TSB may not be fully representative of the heat resistance of cells grown in liquid egg. Nevertheless, previously published studies have used liquid media to grow salmonellae prior to determining its thermal resistance in liquid egg (Froning et al., 2002). The data we present here represents the first steps in determining whether Salmonella contamination in pasteurized egg products may be the result of either thermally-
16
ACCEPTED MANUSCRIPT 17 resistant isolates or post-processing contamination. Strains with higher heat resistance, (e.g., isolate #’s 2, 6, 10 and 12) may indicate the ability of Salmonella to survive pasteurization, while strains with lower heat resistance (e.g., isolate #’s 1, 3, 5, 7, 8, 11, and 15) may be present in egg products as the
IP T
result of post-processing contamination. These findings may be useful in the production of LWE that
SC R
are not contaminated by this foodborne pathogen.
ACKNOWLEDGEMENTS
NU
The authors would like to thank Drs. David Geveke and Brendan Niemira for kindly reviewing
MA
this manuscript prior to submission for publication. Special thanks to Ms. Kimberly Ingram for
AC C
EP
TE
D
providing technical support in the completion of the MIDI fatty acid analyses.
17
ACCEPTED MANUSCRIPT 18
TABLE 1. Salmonella strains used in this study and results of USDA, APHIS, NVSL serotyping and phage typing.
‡
4,12:i:-
FSIS 4 FSIS 5 FSIS 6
*
Pasteurized Frozen Salt Yolk Pasteurized Salt Yolk
Cerro
‡
Liquid Pasteurized Sugar Yolk
Heidelberg
Pasteurized Frozen Salt Yolk
Enteritidis, Phage type 2 Heidelberg
‡
Mbandaka
MA
FSIS 7‡ FSIS 8
Pasteurized Frozen Salt Yolk
Enteritidis, Phage type 8
Pasteurized Frozen Whites
NU
FSIS 3
Source of Isolate
IP T
FSIS 2*
Strain Information Heidelberg
SC R
Strain # FSIS 1‡
Pasteurized Liquid Whites Pasteurized Whole Egg
FSIS 9
Enteritidis, Phage type 8
Pasteurized Egg Whites
FSIS 10
Enteritidis, Phage type 8
Pasteurized Liquid Salted Yolk
D
FSIS 11
‡
Heidelberg
Scramble Egg Mix
Enteritidis, Phage type 8
Pasteurized Frozen Whole Egg
FSIS 13
Enteritidis, Phage type 8
Pasteurized Egg Whites
FSIS 16 FSIS 17 FDA 18
†
EP
FSIS 15
‡
Thompson
Pasteurized Egg Whites
Widemarsh
Pasteurized Dried Egg Whites
AC C
FSIS 14
TE
FSIS 12
Enteritidis, Phage type 8
Pasteurized Egg Whites
Enteritidis, Phage type 8
Pasteurized Frozen Salt Yolk Enteritidis, FDA C405, Phage type 8 Egg yolk, New England salmonellosis outbreak
FDA 19† DuPont 20†*
Enteritidis, FDA C398, Phage type 8 Hen’s Ovary, University of Pennsylvania Oranienberg, Strain DD2229
DuPont culture collection
†
Isolates not isolated from pasteurized egg but were included in published liquid egg pasteurization studies (Gurtler et al., 2011; Jordan et al., 2011; Gurtler et al., 2013). * One of the three most heat-resistant of Salmonella isolates. ‡ One of the seven most heat-sensitive of Salmonella isolates.
18
ACCEPTED MANUSCRIPT 19
1
0.34†‡ E
0.958
-2.904
2
0.55* A
0.955
-1.810
3
0.34‡ E
0.993
-2.926
4
0.45 CD
0.9834
-2.222
8.441
5
0.35 ‡ E
0.976
-2.975
8.704
6
0.58* A
0.964
-1.733
8.079
7
0.34 ‡ E
0.976
-2.930
8.565
8
0.36 ‡ E
0.995
-2.795
8.234
9
0.44 CD
0.989
-2.272
8.757
10
0.40 DE
0.981
-2.496
8.9049
11
0.35‡ E
0.970
-2.856
8.904
12
0.43 CD
0.984
-2.315
9.835
13
0.46 BCD
TE
TABLE 2. Thermal inactivation kinetics of salmonellae in liquid whole egg, inoculated at 8.5 log CFU/mL and pasteurized at 60°C by hot water-immersion capillary tube method. Strain # D-value (min)§ R2 Slope y-intercept
0.988
-2.195
9.031
14
0.40 DE
0.988
-2.525
8.681
15
0.35 ‡ E
0.994
-2.872
8.587
0.47 BC
0.997
-2.144
8.630
0.44 CD
0.985
-2. 244
8.565
0.42 DE
0.973
-2.439
8.837
19
0.45 CD
0.979
-2.233
9.215
20
0.55 * AB
0.961
-1.833
8.674
Ave.
0.42
17 18
IP T
SC R
NU MA
D
EP
AC C
16
9.205 8.220 8.619
§
Values in the same column not followed by the same letter are significantly different (p
