Inactivation of Acid-Adapted and Nonadapted Escherichia coli O157:H7 during Drying and Storage of Beef Jerky Treated with Different Marinades

Transcription

1 1394 Journal of Food Protection, Vol. 65, No. 9, 2002, Pages Copyright Q, International Association for Food Protection Inactivation of Acid-Adapted and Nonadapted Escherichia coli O157:H7 during Drying and Storage of Beef Jerky Treated with Different Marinades MEHMET CALICIOGLU, 1 JOHN N. SOFOS, 1 * JOHN SAMELIS, 1 PATRICIA A. KENDALL, 2 AND GARY C. SMITH 1 1 Center for Red Meat Safety, Department of Animal Sciences, and 2 Department of Food Science and Nutrition, Colorado State University, Fort Collins, Colorado 80523, USA MS : Received 13 December 2001/Accepted 30 March 2002 ABSTRACT The inactivation of both acid-adapted and unadapted Escherichia coli O157:H7 during the processing of beef jerky was studied. Following inoculation with the pathogen, beef slices were subjected to different predrying marinade treatments, dried at 608C for 10 h, and stored at 258C for 60 d. The predrying treatments evaluated were as follows: (i) no treatment (C), (ii) traditional marinade (TM), (iii) double-strength TM modi ed with added 1.2% sodium lactate, 9% acetic acid, and 68% soy sauce with 5% ethanol (MM), (iv) dipping into 5% acetic acid for 10 min followed by application of TM (AATM), and (v) dipping into 1% Tween 20 for 15 min and then into 5% acetic acid for 10 min followed by TM (TWTM). Bacterial survivors were determined during drying and storage using tryptic soy agar with 0.1% pyruvate, modi ed eosin methylene blue agar, and sorbitol MacConkey agar. Results indicated that bacterial populations decreased during drying in the order of TWTM (4.9 to 6.7 log). AATM. MM. C $ TM (2.8 to 4.9 log) predrying treatments. Populations of acid-adapted E. coli O157:H7 decreased faster (P, 0.05) in AATM and TWTM than nonadapted cells during drying, whereas no signi cant difference was found in inactivation of acid-adapted and nonadapted inocula in C and TM samples. MM was more effective in inactivating the nonadapted than the adapted inoculum. Bacterial populations continued to decline during storage and dropped below the detection limit (20.4 log 10 CFU/cm 2 ) as early as day 0 (after drying) or as late as day 60, depending on acid adaptation, predrying treatment, and agar medium. The results indicated that acid adaptation may not increase resistance to the hurdles involved in jerky processing and that use of additional antimicrobial chemicals or preservatives in jerky marination may improve the effectiveness of drying in inactivating E. coli O157:H7. Jerky is a ready-to-eat dried meat product common in North America. For centuries, it has been produced by native Americans by smoking and sun-drying of meat strips. Today, numerous recommendations for making jerky are available to the industry and people drying meat at home. Recommendations vary by type of species used (e.g., beef, poultry, game animals), the form of meat (e.g., thick or thin slices, ground beef), marination technique (e.g., various ingredients, volume, time, and temperature), and drying process (e.g., oven, food dehydrator, smokehouse, and high versus low temperature). Association of jerky products with foodborne disease outbreaks (10, 25) has raised questions about the microbial safety of such products. Dried products, in general, have been considered as one of the safest food groups for humans since they involve hurdles to microbial survival or growth. Such hurdles include low water activity (a w ) (,0.85), drying temperature, and preservatives such as salt, organic acids, and sodium nitrite, depending on the composition of the marination mixture (15). Eidson et al. (13) reported that eight foodborne disease outbreaks occurred due to consumption of contaminated beef jerky (six Salmonella and two Staphylococcus aureus) * Author for correspondence. Tel: ; Fax: ; John.Sofos@colostate.edu. between 1966 and 1995 in New Mexico alone. A recent report (26) by the Food Safety and Inspection Service of the U.S. Department of Agriculture (FSIS/USDA) indicated that from 1990 to 1999 cumulative prevalence of Salmonella and Listeria monocytogenes in jerky produced in federally inspected plants was 0.31 and 0.52%, respectively. Inactivation of S. aureus, Salmonella, Clostridium perfringens, and Bacillus subtilis during drying of jerky for 4 h at 52.98C followed by another 4 h at 48.28C using a home-type dehydrator was studied by Holley (21). The results indicated that initial numbers (4 to 5 log 10 CFU/g) of all three pathogens were signi cantly reduced (by 2 to 4 log 10 CFU/g) after 8 h of drying. However, with the exception of C. perfringens (,2.0 log 10 CFU/g), other microorganisms survived. The pathogens survived at high relative humidity storage at 208C for 26 to 28 d and longer at low relative humidity refrigerated (2.58C) storage. Holley (21, 22) suggested that homemade jerky is ensured to be stable provided initial drying is done rapidly and at temperatures equal to or higher than 538C, preferably 55 to 608C. The recent Salmonella and E. coli O157:H7 outbreaks linked to jerky consumption (10, 25) have renewed interest in evaluating the ef cacy of jerky processing, especially when prepared in home-type dehydrators, in inactivating emerging pathogens (1, 14, 17 20, 25). As a response to

2 J. Food Prot., Vol. 65, No. 9 INACTIVATION OF E. COLI O157:H7 ON BEEF JERKY DURING DRYING AND STORAGE 1395 the problem, the USDA/FSIS (39) has suggested cooking meat to 71.18C before drying to eliminate risk of pathogens. The results of these studies indicate discrepancies as to whether current practices for home-dried jerky can achieve a 5-log reduction in pathogens as recommended by regulatory agencies (39). In addition, preheating of meat and/or drying of jerky at high temperatures for extended periods may result in a product that differs from traditional jerky; thus, it may be of reduced product acceptability. Use of chemical intervention strategies such as predrying treatments, however, has not been studied adequately (1). Such interventions can be a viable option to avoid severe heat treatments and may provide residual antimicrobial effects during product storage. These chemicals may include organic acids (e.g., acetic acid), ethanol, lactates, and foodgrade surfactants (e.g., polysorbates). There has been evidence that acid-adaptation of E. coli O157:H7 and other pathogens may enhance their survival in acidic foods and increase cross-protection to other types of stresses (3, 16, 27 29, 32 34). To date, it is not known whether acid-adapted pathogenic bacterial cells may survive better than non acid-adapted cells during the jerkymaking process. Therefore, the objective of the present study was to evaluate the effectiveness of various chemicalbased predrying treatments (modi ed marinades) on acidadapted or nonadapted E. coli O157:H7 cells during preparation, drying, and storage of whole muscle beef jerky. MATERIALS AND METHODS Bacterial strains used and preparation of inoculum. A ve-strain composite of E. coli O157:H7 was used for inoculating beef slices. These strains were EO139 (a venison jerky isolate, provided by Dr. M. P. Doyle, University of Georgia, Grif n, Ga.), ATCC (hamburger isolate), ATCC (human isolate), ATCC (human isolate), and ATCC (human isolate). Each strain was propagated (358C, 24 h) and maintained on tryptic soy agar (TSA; Difco Laboratories, Sparks, Md.) slants at 48C. Strains were subcultured monthly. The cultures were activated by transferring a loopful of each strain into 9-ml of tryptic soy broth (TSB; Difco) and incubating at 358C for 24 h. A 0.1-ml portion of each culture was then transferred into 9-ml tubes of glucosefree TSB for non acid-adapted cells or glucose-free TSB with 1% added glucose (Sigma Chemical Company, St. Louis, Mo.) for acid-adapted cells (6). After incubation at 358C for 22 to 24 h (the ph of the broths was 5.5 and 7.1 for acid-adapted and nonadapted cultures, respectively), cultures of individual strains were combined in a sterile test tube before centrifuging at 6,000 rpm (2,900 3 g) for 15 min at 218C. The resulting pellet was washed once with 0.1% phosphate-buffered saline (Sigma) to remove residual organic material, recentrifuged, and then resuspended in phosphate-buffered saline to a nal volume of 100 ml. The average level of inoculum was 8.0 log CFU/ml. Preparation of meat slices. Vacuum packaged and frozen (2188C) beef inside rounds (,3 months) were purchased from Colorado State University Meat Science Laboratory (Fort Collins, Colo.). Following thawing at 48C overnight, inside rounds were sliced at 0.6-cm thickness using a food slicer (model 610, Hobart Corp., Troy, Ohio) and cut into pieces of 8.7 by 4.0 cm using a plastic template and knife. Approximately 100 slices (2.2 kg) of meat were vacuum packaged and kept frozen at 2188C until used (1 to 3 weeks). Inoculation procedure. Frozen beef slices were thawed at 48C for 24 h and placed on plastic trays covered with aluminum foil. Under a laminar ow hood, 0.5 ml of the E. coli O157:H7 inoculum was placed on the upper surface of each slice and spread onto the entire surface area using a sterile bent glass rod. Bacteria were allowed to attach to the meat surface for 15 min at ambient temperature. Beef slices were then inverted and the other side was inoculated following the same procedure. The resulting level of inoculum was approximately 6.5 log 10 CFU/cm 2. Predrying marinade treatments. The predrying treatments included the following: (i) control, no treatment (C), (ii) marination with traditional marinade (TM) (ph 4.3), (iii) marination with modi ed marinade (MM) (ph 3.0), (iv) dipping into 5% acetic acid solution (ph 2.5) for 10 min followed by TM (AATM), and (v) sequential dipping into 1% Tween 20 solution (ph 6.6) for 15 min and then 5% acetic acid solution for 10 min followed by marination with TM (TWTM). The TM was prepared for 1.0 kg of meat (2) as follows: 60 ml of soy sauce (Kikkoman Foods, Walworth, Wis.), 15 ml of Worcestershire sauce (Heinz, Pittsburgh, Pa.), 0.6 g of black pepper (Heller Seasoning and Ingredients Inc., Chicago, Ill.), 1.25 g of garlic powder (Excalibur Seasoning Co. Ltd., Pekin, Ill.), 1.5 g of onion powder (Excalibur), and 4.35 g of hickory-smoked salt (Tone Brothers Inc., Ankeny, Iowa). In the present study, 34 ml of this marinade was spread manually onto 450-g inoculated beef slices, and then the slices were manually mixed to cover the entire surface area using amesterilized forceps. Approximately 30 ml of the marinade remained on the meat slices. The modi ed marinade was prepared for 1.0 kg of meat as follows: 120 ml of milder soy sauce (Kikkoman) containing approximately 4.7 to 5.0% ethanol as preservative, 30 ml of Worcestershire sauce, 0.6 g of black pepper, 1.25 g of garlic powder, 1.5 g of onion powder, 4.35 g of smoke- avored salt, 3.6 ml of food-grade sodium-l-lactate of a 60% preparation (Purac Inc., Lincolnshire, Ill.), and 16 ml of glacial acetic acid to adjust the ph to 3.0. A 77-ml portion of this marinade solution was spread onto 450 g of beef and mixed to cover the surfaces of the meat slices. Approximately 60 ml of the MM remained on the meat. This was twice the volume of TM. For treatment AATM, meat slices were dipped at ambient temperature for 10 min into 5% (vol/vol) acetic acid solution prepared using glacial acetic acid (Mallinckrodt Baker) (450 ml per 450 g of meat) in a glass container. These slices were drained for 2 min to remove excessive uid using an empty dehydrator tray and placed on an aluminum foil covered tray followed by marination with traditional marinade as described for TM. For treatment TWTM, meat slices were dippped for 15 min at ambient temperature into 1% (vol/vol) Tween 20 (Fisher Scienti c Inc., Fair Lawn, N.J.) solution (1 liter per 1 kg of meat). These slices were drained for 2 min, and then steps for treatment AATM were followed. Tween 20 (polysorbate 20 or polyoxyethylene-20-sorbitan monolaurate) is permitted for use in food as an emulsi er (3). The black pepper and garlic and onion powders used in the studies had been irradiated by their manufacturers. Use of nonirradiated spices in preliminary studies resulted in product contamination with spore-forming bacteria, which developed on agar plates at later stages of drying. Following application of marination treatments, the slices on the trays were covered with aluminum foil and held at 48C for 24 h before drying. Drying. Treated and refrigerated meat slices were dried at 608C for 10 h in American Harvest Gardenmaster dehydrators (model FD-1000, Nesco, Chaska, Minn.). The dehydrators were cylindrical in shape and were composed of a base unit and three drying trays. The dehydrator base unit generated hot air, which

3 1396 CALICIOGLU ET AL. J. Food Prot., Vol. 65, No. 9 TABLE 1. Mean (log CFU/cm 2 ) populations (standard deviation) of bacteria determined on tryptic soy agar with 0.1% pyruvate (TSAP) from beef jerky inoculated with previously acid-adapted or nonadapted Escherichia coli O157:H7, exposed (24 h, 48C) to various marination predrying treatments, and dried at 608C for 10 h (n 5 4) a Time Acid-adapted E. coli O157:H7 Non acid-adapted E. coli O157:H7 After inoculation b,c After marination 0 h 6.4 AZ 4 h 4.4 BZ 7 h 3.2 BZY 10 h 3.2 BZY (0.0) 4.5 BCZ 3.8 BZ 3.5 BZ 6.2 AZ 3.8 BZY 2.9 BZY 2.8 BZY 6.1 AZ 3.3 BZY 1.9 BCY 1.1 CY 6.0 AZ 2.1 BY 1.1 BCY 0.8 CY 4.1 BZ 3.3 BZY 3.1 BZY 6.4 AZ 4.5 BZ 3.7 BZ 3.0 BZY 3.2 BZY (0.8) 1.7 CY 1.9 CY 5.9 AZ 3.3 BZY (1.2) 1.6 CY (1.1) 1.8 CY (1.0) 5.8 AZ 2.5 BY (0.9) 1.4 CY 1.6 CY a C, inoculated beef slices received no marinade predrying treatment before refrigeration at 48C for 24 h and drying; TM, inoculated beef slices were marinated with traditional marinade (ph 4.3), held at 48C for 24 h, and then dried; MM, inoculated beef slices were marinated with modi ed marinade (double amount, added 1.2% lactate, 9% acetic acid, and soy sauce with 5% ethanol; ph 3.0), held at 48C for 24 h, and then dried; AATM, inoculated beef slices were dipped into 5% acetic acid solution (ph 2.5) for 10 min at ambient temperature, drained for 2 min and then marinated with traditional marinade before holding at 48C for 24 h, and then dried; TWTM, inoculated beef slices were dipped into 1% Tween 20 (ph 6.6) for 15 min and then into 5% acetic acid solution for 10 min at ambient temperature, marinated with traditional marinade before holding at 48C for 24 h, and then dried. b ABCD, means within a column lacking a common letter are different (P, 0.05). c ZYX, means within a row lacking a common letter are different (P, 0.05). was ventilated upward through the sides and a hole in the middle of the trays. The target drying temperature was based on the air temperature measured with inserted thermocouples (Type K beaded probes, Pico Technology Ltd., Cambridge, England) from the middle hole of the dehydrator. The dehydrators with empty trays were preheated to 608C (1408F) for approximately 20 min, and then the trays were loaded with meat slices and placed in the dehydrator. During drying, the temperature of the dehydrators and surface temperature of meat slices on the bottom, middle, and top trays were monitored using thermocouples (Pico Technology) and recorded with real-time data recording software (Pico Technology). After drying, the jerky strips were held in the dehydrators overnight and then placed into 24-oz Whirl-Pak sterile plastic bags (Nasco, Fort Atkinson, Wis.) for storage at ambient temperature ( C). Analysis. Two samples (1 slice per sample) per treatment were aseptically transferred into 18-oz sterile plastic stomaching bags (Nasco, Modesto, Calif.) at each sampling interval. These intervals included after inoculation and 0, 4, 7, and 10 h of drying for each treatment and days 15, 30, and 60 of storage at 258C. A 25-ml portion of 0.1% sterile buffered peptone water (BPW; Difco) was added to each sample bag before pummeling for 2 min (120 strikes per min). Serial decimal dilutions were made using 9-ml BPW tubes, and 0.1-ml portions were surface plated onto each of duplicate plates of each agar medium. Bacterial populations were enumerated using TSA (Difco) with 0.1% sodium pyruvate (Fisher Scienti c) (TSAP) (27), sorbitol MacConkey agar (SMAC; Difco), and modi ed eosin methylene blue agar (MEMB) prepared as described by Clavero and Beuchat (11). All plates were incubated at 358C for 48 h. The enumeration detection limit was 20.4 log 10 CFU/cm 2. When numbers of bacteria dropped below the detection limit, enrichment of samples was done using modi ed EC broth (Difco) containing novobiocin (Sigma) (20 mg/ ml). After incubation at 358C for 24 h, samples were streak-plated onto SMAC agar plates and a latex agglutination test (Remel Diagnostic Reagents, Lenexa, Kans.) was conducted on colonies to con rm presence of the O157 antigen. In addition to microbiological analyses, ph and a w of beef jerky slices were also determined at the same sampling intervals. The ph was measured from the same samples used for microbiological analysis (25 ml of BPW added and pummeled for 2 min) using a digital ph meter (Accumet 50, Fisher Scienti c, Houston, Tex.) with a glass ph electrode (Hanna Instruments, Ann Arbor, Mich.). The a w values of beef slices were determined by method of AOAC International (30). One jerky slice was cut into small pieces to t into the plastic container and measurement was done using an a w meter (model D2100, Rotronic Instrument Corp., Huntington, N.Y.). Statistical analysis. Two independent replicates of the study were conducted. Microbiological data were converted to log 10 CFU/cm 2 and evaluated using a (acid adaptation 3 number of replicates 3 predrying treatments 3 drying [sampling] times 3 numbers of samples 3 agar media, respectively) factorial design. Data were analyzed by analysis of variance for main ( xed) effects (acid adaptation, predrying treatment, drying time, and agar media) and four-way interactions among acid adaptation, predrying treatment, drying time, and agar media using the SAS statistical software (version 6.1, SAS Institute Inc., Cary, N.C.). Least squares means were separated by Fisher s least signi cance test using the general linear models procedure of SAS. A signi cance level of 0.05 was used for all statistical analyses. Storage data were analyzed separately. Data from each agar medium are presented separately (Tables 1 through 4) because the large volume of data does not allow presentation in a single table. Mean counts of surviving bacterial populations between 0 and 7 h during drying were used to determine D-values by calculating the inverse of the slope of the linear regression

4 J. Food Prot., Vol. 65, No. 9 INACTIVATION OF E. COLI O157:H7 ON BEEF JERKY DURING DRYING AND STORAGE 1397 TABLE 2. Mean (log CFU/cm 2 ) populations (standard deviation) of bacteria determined on modi ed methylene eosin blue agar (MEMB) from beef jerky inoculated with previously acid-adapted or nonadapted Escherichia coli O157:H7, exposed (24 h, 48C) to various marination predrying treatments, and dried at 608C for 10 h (n 5 4) a Time Acid-adapted E. coli O157:H7 Non acid-adapted E. coli O157:H7 After inoculation b,c After marination 0 h 6.4 AZ (0.0) 4 h 3.7 BZ 7 h 2.1 CY 10 h 2.0 CZY 6.2 AZ 3.7 BZ 3.0 BCZ 2.5 CZY 6.2 AZ 3.4 BZY 2.1 CY 2.2 CZY 5.9 AZ 2.4 BY (1.1) 0.9 CX 0.0 DX 5.6 AZ 1.3 BX 0.0 CW 0.0 CX 4.0 BZ (0.9) 2.9 CZY 2.7 CZ 4.0 BZ 3.4 BCZ 2.7 CZ 6.2 AZ 2.4 BY 1.8 BYX 1.8 BY 5.9 AZ 2.5 BY (0.9) 1.4 CYX (1.0) 1.6 CY (0.9) 5.6 AZ 1.8 BYX 1.2 BX 1.4 BY a C, inoculated beef slices received no marinade predrying treatment before refrigeration at 48C for 24 h and drying; TM, inoculated beef slices were marinated with traditional marinade (ph 4.3), held at 48C for 24 h, and then dried; MM, inoculated beef slices were marinated with modi ed marinade (double amount, added 1.2% lactate, 9% acetic acid, and soy sauce with 5% ethanol; ph 3.0), held at 48C for 24 h, and then dried; AATM, inoculated beef slices were dipped into 5% acetic acid solution (ph 2.5) for 10 min at ambient temperature, drained for 2 min and then marinated with traditional marinade before holding at 48C for 24 h, and then dried; TWTM, inoculated beef slices were dipped into 1% Tween 20 (ph 6.6) for 15 min and then into 5% acetic acid solution for 10 min at ambient temperature, marinated with traditional marinade before holding at 48C for 24 h, and then dried. b ABCD, means within a column lacking a common letter are different (P, 0.05). c ZYX, means within a row lacking a common letter are different (P, 0.05). line. Means and standard deviations of ph and a w data were determined. RESULTS Tables 1 through 4 show populations of E. coli O157: H7 during drying and storage of beef slices. Statistical analysis of the microbial data revealed that the main effects of acid adaptation, predrying treatment, drying time, and agar media and the interactions of acid adaptation 3 predrying treatment 3 drying time 3 agar media were signi cant (P, 0.05). Temperature. Changes in dehydrator air temperature and meat surface temperatures during drying are shown in Figure 1. After loading the preheated dehydrators with meat slices, air temperature decreased from 60 to 408C at the beginning of drying and then gradually increased back to 608C within approximately 4 h. Surface temperatures of meat from all trays reached C within approximately 6 h. At the end of drying, temperatures of meat surfaces decreased to room temperatures (approximately 25 to 308C) within approximately 1 h. Effect of agar media. There was no signi cant difference (P $ 0.05) in numbers of bacteria recovered by TSAP, MEMB, and SMAC within each treatment after inoculation and marination (0 h). In products inoculated with non acidadapted culture, there was no signi cant difference in bacterial populations recovered by the different media within each treatment at 4 h of drying, although smaller populations of bacteria were recovered with MEMB and SMAC compared with TSAP. However, at 7 and 10 h of drying, counts on TSAP and MEMB were signi cantly higher than those on SMAC, irrespective of predrying treatment. In products inoculated with acid-adapted cultures, counts on TSAP were signi cantly higher than those on SMAC at 4, 7, and 10 h of drying. Counts on MEMB at 4 h, however, were signi cantly higher than those on SMAC in MM, AATM, and TWTM but not in C and TM samples. At 7 and 10 h, populations recovered by each of the three agar media were signi cantly different from each other in the C and TM treatments. In general, bacterial populations recovered were the highest on TSAP, followed by MEMB, and nally by SMAC. These results suggest that level of bacterial injury, as estimated by differences in populations between nonselective (TSAP), medium selective (MEMB), and most selective (SMAC) agar media, may have been affected by acid adaptation, predrying treatment, and drying time. Effect of predrying marinade treatments. Initial bacterial numbers did not change (#1.0 log 10 CFU/cm 2 ) signi cantly (P $ 0.05) after application of predrying treatments and holding of meat slices at 48C for 24 h before drying, regardless of acid adaptation, marination treatment, or agar medium (Tables 1 through 3). However, it is noteworthy that MM, AATM, and TWTM treatments resulted in relatively higher SMAC agar bacterial population reductions compared with C and TM treatments. Although not signi cant, these differences may indicate possible bacterial injury due to exposure to marinade components during refrigerated storage for 24 h. Effect of drying. Regardless of acid adaptation or recovery media used, initial bacterial counts were signi -

5 1398 CALICIOGLU ET AL. J. Food Prot., Vol. 65, No. 9 TABLE 3. Mean (log CFU/cm 2 ) populations (standard deviation) of bacteria determined on sorbitol MacConkey agar (SMAC) from beef jerky inoculated with previously acid-adapted or nonadapted Escherichia coli O157:H7, exposed (24 h, 48C) to various marination predrying treatments, and dried at 608C for 10 h (n 5 4) a Acid-adapted E. coli O157:H7 Non acid-adapted E. coli O157:H7 Time After inoculation After marination 0 h 6.2 AZ 4 h 2.8 BZY 7 h 1.1 CY 10 h 1.0 CZY 6.2 AZ (0.0) 3.0 BZY 2.1 CZY 1.4 CZY 5.9 AZ 2.2 BZY 0.9 BY 0.8 BZY 5.4 AZ 1.4 BY 20.1 BCY 5.2 AZ 0.5 BX 3.6 BZ (1.1),20.4 CX 2.3 BZY,20.4 CX,20.4 CX 2.3 BZ 3.8 BZ 3.0 BZ 1.8 CZ 5.9 AZ 1.6 BY 0.5 BY 0.5 BY 5.6 AZ 1.8 BZY 0.4 BY (0.9) 0.2 BY 5.5 AZ (0.8) 1.5 BY 0.3 BCY 0.1 CY a C, inoculated beef slices received no marinade predrying treatment before refrigeration at 48C for 24 h and drying; TM, inoculated beef slices were marinated with traditional marinade (ph 4.3), held at 48C for 24 h, and then dried; MM, inoculated beef slices were marinated with modi ed marinade (double amount, added 1.2% lactate, 9% acetic acid, and soy sauce with 5% ethanol; ph 3.0), held at 48C for 24 h, and then dried; AATM, inoculated beef slices were dipped into 5% acetic acid solution (ph 2.5) for 10 min at ambient temperature, drained for 2 min and then marinated with traditional marinade before holding at 48C for 24 h, and then dried; TWTM, inoculated beef slices were dipped into 1% Tween 20 (ph 6.6) for 15 min and then into 5% acetic acid solution for 10 min at ambient temperature, marinated with traditional marinade before holding at 48C for 24 h, and then dried. b ABCD, means within a column lacking a common letter are different (P, 0.05). c ZYX, means within a row lacking a common letter are different (P, 0.05).

6 J. Food Prot., Vol. 65, No. 9 INACTIVATION OF E. COLI O157:H7 ON BEEF JERKY DURING DRYING AND STORAGE 1399 TABLE 4. Mean (log 10 CFU/cm 2 ) populations of surviving bacteria determined on different agar media during storage at 258C of beef jerky that was inoculated with previously acid-adapted or nonadapted Escherichia coli O157:H7, exposed (24 h, 48C) to various marination predrying treatments, and dried at 608C for 10 h before aerobic storage (n 5 4) for 60 d a Acid-adapted E. coli O157:H7 Non acid-adapted E. coli O157:H7 Agar media Days b Tryptic soy agar and 0.1% pyruvate (TSAP) Modi ed eosin methylene blue agar (MEMB) Sorbitol MacConkey agar (SMAC) AZY c,d 1.4 BZY 0.3 BCX 2.0 AZY 0.0 BY 1.0 AZY 3.5 AZ 2.0 BZ 1.2 BCZ 2.5 AZY 0.8 BZY 0.2 BCZY 1.4 AZY 0.2 BZY 2.8 AZY 0.7 BY,20.4 CX,20.4 e CZ 2.2 AZY 0.0 BY 0.8 AZY 1.1 AY 0.2 ABYX,20.4 BX,20.4 e BZ 0.0 AX,20.4 AZ,20.4 AZ,20.4 AZ 0.8 AY,20.4 BX, 0.4 e BX,20.4 e BZ 0.0 AX,20.4 AZ,20.4 AZ,20.4 AZ 3.1 AZY 1.7 BZY 0.6 CY 2.7 AZ 1.1 BZ 0.7 BZ 2.3 AZ 0.9 BZ 30 AZY 1.7 BZY 0.1 CX 2.7 AZ 1.2 BZ 0.0 BCZY 1.8 AZY 0.5 BZ 20.2 BZ 1.9 AY 20.3 BX,20.4 BX,20.4 e BZ 1.8 AY 0.5 AY 1.8 AY 201 BYX 20.1 BX,20.4 e BZ 1.6 AY 0.2 AZY 1.6 AY BYX,20.4 e BX,20.4 e BZ 1.4 AY 0.1 ACY a C, inoculated beef slices received no marinade predrying treatment before refrigeration at 48C for 24 h and drying; TM, inoculated beef slices were marinated with traditional marinade (ph 4.3), held at 48C for 24 h, and then dried; MM, inoculated beef slices were marinated with modi ed marinade (double amount, added 1.2% lactate, 9% acetic acid, and soy sauce with 5% ethanol; ph 3.0), held at 48C for 24 h, and then dried; AATM, inoculated beef slices were dipped into 5% acetic acid solution (ph 2.5) for 10 min at ambient temperature, drained for 2 min and then marinated with traditional marinade before holding at 48C for 24 h, and then dried; TWTM, inoculated beef slices were dipped into 1% Tween 20 (ph 6.6) for 15 min and then into 5% acetic acid solution for 10 min at ambient temperature, marinated with traditional marinade before holding at 48C for 24 h, and then dried. b Bacterial counts at 10 h of drying (Tables 1 through 3) are included as day 0 counts for statistical analysis of the storage data. c ABCD, means within a column of each agar medium lacking a common letter are different (P, 0.05). d ZYX, means within a row lacking a common letter are different (P, 0.05). e Negative by enrichment. Observations without this superscript on day 60 from TSAP are enrichment positive.

7 1400 CALICIOGLU ET AL. J. Food Prot., Vol. 65, No. 9 TABLE 5. Decimal reduction times (D-values) of bacteria determined with different agar media from beef jerky that was inoculated with previously acid-adapted or nonadapted Escherichia coli O157:H7, exposed (24 h, 48C) to various marination predrying treatments, and dried at 608C for 7 h a Jerky marination treatment Acid-adapted E. coli O157:H7 D-value (h) r 2 Non acid-adapted E. coli O157:H7 D-value (h) r 2 FIGURE 1. Mean (n 5 4) temperatures of dehydrator air (middle hole air temperature) and the surface of beef slices during drying at 608C for 10 h using a home-type food dehydrator. cantly (P, 0.05) reduced in all treatments after 4 h of drying (Tables 1 through 3). Bacterial populations further declined between 4 and 7 h of drying, but these declines were not signi cant. No signi cant reduction in bacterial populations was found in any treatment inoculated with either acid-adapted or nonadapted cultures between 7 and 10 h of drying. In general, AATM and TWTM resulted in signi cantly lower bacterial populations as determined with selective media than the C and TM treatments at 4, 7, and 10 h of drying for both culture types. However, AATM and TWTM resulted in lower bacterial counts in products inoculated with acid-adapted cells than nonadapted cells, indicating increased susceptibility of acid-adapted E. coli O157:H7 to these treatments during drying. Populations of bacteria in TWTM and AATM products inoculated with acid-adapted cultures decreased below the detection limit (20.4 log CFU/cm 2 ) on SMAC by 7 and 10 h of drying, respectively. However, viable cells of E. coli O157:H7 were recovered from these treatments by enrichment (Table 1). With the exception of bacterial counts determined by SMAC at 7 h from products inoculated with acid-adapted culture, bacterial counts between the C and TM products were not signi cantly different at each drying time, irrespective of acid adaptation and agar media. In general, populations of nonadapted bacteria in MM products in all three media were signi cantly lower compared with acid-adapted bacteria. This may indicate that MM was more effective in reducing counts of nonadapted than acid-adapted E. coli O157:H7 cells. Total log reductions of bacteria in the products inoculated with nonadapted E. coli O157:H7 at the end of 10 h of drying, as determined on TSAP and SMAC, were 3.2 and 4.3 for C, 2.6 and 4.8 for TM, 4.6 and 6.1 for MM, 4.7 and 6.2 for AATM, and 4.9 and 6.3 log CFU/cm 2 for TWTM, respectively. Corresponding reductions in products inoculated with acid-adapted cells were 3.1 and 5.3 for C, 2.8 and 4.9 for TM, 3.5 and 5.4 for MM, 5.2 and 6.7 for AATM, and 5.5 and 6.7 for TWTM log CFU/cm 2. Tryptic soy agar with 0.1% pyruvate (TSAP) C TM MM AATM TWTM Modi ed eosin methylene blue agar (MEMB) C TM MM AATM TWTM Sorbitol MacConkey agar (SMAC) C TM MM AATM TWTM a C, inoculated beef slices received no marinade predrying treatment before refrigeration at 48C for 24 h and drying; TM, inoculated beef slices were marinated with traditional marinade (ph 4.3), held at 48C for 24 h, and then dried; MM, inoculated beef slices were marinated with modi ed marinade (double amount, added 1.2% lactate, 9% acetic acid, and soy sauce with 5% ethanol; ph 3.0), held at 48C for 24 h, and then dried; AATM, inoculated beef slices were dipped into 5% acetic acid solution (ph 2.5) for 10 min at ambient temperature, drained for 2 min and then marinated with traditional marinade before holding at 48C for 24 h, and then dried; TWTM, inoculated beef slices were dipped into 1% Tween 20 (ph 6.6) for 15 min and then into 5% acetic acid solution for 10 min at ambient temperature, marinated with traditional marinade before holding at 48C for 24 h, and then dried. D-values calculated based on bacterial counts determined on each of the three agar media between 0 and 7 h of drying are shown in Table 5. Irrespective of acid adaptation, AATM and TWTM treatments resulted in smaller D-values than the C and TM in all agar media, indicating faster declines in the numbers of bacteria. However, drying of MM products inoculated with acid-adapted culture resulted in higher D-values than when inoculated with nonadapted culture. Other than this particular treatment, D-values for products inoculated with acid-adapted cells were relatively smaller than for products inoculated with nonadapted cells, indicating possible increase in susceptibility of acid-adapted cells to hot-air drying for 7 h. In general, D-values decreased from highest to lowest within each treatment as calculated from counts of survivors determined with TSAP, MEMB, and SMAC agar. The difference be-

8 J. Food Prot., Vol. 65, No. 9 INACTIVATION OF E. COLI O157:H7 ON BEEF JERKY DURING DRYING AND STORAGE 1401 TABLE 6. Mean (standard deviation) ph values (n 5 4) of beef jerky slices inoculated with acid-adapted or nonadapted Escherichia coli O157:H7, exposed (24 h, 48C) to various marination predrying treatments, dried at 608C for 10 h, and stored at 258C for 60 d a Steps and time Inoculated with acid-adapted E. coli O157:H7 Inoculated with non acid-adapted E. coli O157:H7 Processing 0 h h 5.61 (0.01) 7 h h 5.61 (0.09) 5.45 (0.15) 5.62 (0.02) 5.60 (0.03) 5.56 (0.07) 4.74 (0.16) 4.81 (0.08) 4.84 (0.09) (0.11) 4.65 (0.02) (0.02) (0.03) 4.66 (0.04) (0.09) 5.71 (0.08) 5.76 (0.07) (0.04) (0.02) 4.64 (0.07) 4.73 (0.14) 4.75 (0.03) (0.17) 4.80 (0.13) 4.82 (0.15) 4.82 (0.21) 4.60 (0.30) 4.83 (0.15) 4.82 (0.15) 4.78 (0.24) Storage Day (0.04) Day Day (0.08) 5.54 (0.02) (0.09) (0.04) (0.13) (0.07) 5.63 (0.03) 5.70 (0.04) 5.54 (0.11) (0.07) (0.14) (0.04) 4.66 (0.30) 4.70 (0.14) 4.74 (0.13) 4.61 (0.30) 4.83 (0.12) a C, inoculated beef slices received no marinade predrying treatment before refrigeration at 48C for 24 h and drying; TM, inoculated beef slices were marinated with traditional marinade (ph 4.3), held at 48C for 24 h, and then dried; MM, inoculated beef slices were marinated with modi ed marinade (double amount, added 1.2% lactate, 9% acetic acid, and soy sauce with 5% ethanol; ph 3.0), held at 48C for 24 h, and then dried; AATM, inoculated beef slices were dipped into 5% acetic acid solution (ph 2.5) for 10 min at ambient temperature, drained for 2 min and then marinated with traditional marinade before holding at 48C for 24 h, and then dried; TWTM, inoculated beef slices were dipped into 1% Tween 20 (ph 6.6) for 15 min and then into 5% acetic acid solution for 10 min at ambient temperature, marinated with traditional marinade before holding at 48C for 24 h, and then dried. tween the D-values of individual treatments became less notable as the media became more selective, indicating bacterial injury in all treatments. Effect of storage. Populations of bacteria surviving the drying process continued to decrease in all treatments during storage at ambient temperature (258C) (Table 4). In general, bacterial counts on MM, AATM, and TWTM treatments declined faster than those on C and TM products. The differences in surviving populations between nonselective (TSAP) and selective (MEMB and SMAC) agar media were signi cant in C and TM products on day 15 but not on day 30 and day 60. No signi cant difference in counts between agar media was observed in other treatments during the entire storage period. This indicates death of bacteria was more extensive after 15 d of storage on C and TM jerky, whereas in other treatments pathogen inactivation was found during the entire storage period. The earliest complete elimination (enrichment negative) of the pathogen during product storage was achieved in TWTM-treated jerky; this occurred on day 15 in product inoculated with acidadapted E. coli O157:H7. By day 60, counts on products from all treatments were below the detection limit by direct plating on all three media. However, viable cells were recovered from C and TM jerky by enrichment, irrespective of acid adaptation. ph and a w. Overall, ph values of meat slices after exposure to the predrying marination treatments of MM (4.64 to 4.74), AATM (4.40 to 4.64), and TWTM (4.34 to 4.60) were lower than C (5.51 to 5.68) and TM (5.45 to 5.65). The ph values of products from all treatments increased slightly during drying probably due to increasing buffering capacity resulting from meat protein denaturation. No appreciable change in ph of the products was observed during storage (Table 6). The a w of the products decreased remarkably, as expected, and there was no clear pattern for the effect of individual treatments on a w decreases of meat slices during drying. Also, no notable change in a w was found among treatments during storage. The nal a w of the nished products varied between and Slight uctuations between sampling times were probably due to variation among slices (Table 7). DISCUSSION The results of this study indicated that inactivation of E. coli O157:H7 during drying of beef jerky may be affected by the composition of the predrying marination treatment. Although drying reduced bacterial populations in all treatments, MM, AATM, and TWTM resulted in greater levels of reduction compared with C and TM treatments. Survival of the pathogen during drying was not signi - cantly different between C and TM jerky, indicating no additional antimicrobial effect from TM. Although TM contained Worcestershire sauce, the ph values of control and TM jerky were not appreciably different, probably due to the use of a limited volume of marinade per beef slice. The potential contribution of TM in enhancing reduction of E. coli O157:H7 and other pathogens has also been studied by other researchers. Keene et al. (25) reported that drying

9 1402 CALICIOGLU ET AL. J. Food Prot., Vol. 65, No. 9 TABLE 7. Mean (standard deviation) a w values (n 5 4) of beef jerky slices inoculated with acid-adapted or nonadapted Escherichia coli O157:H7, exposed (24 h, 48C) to various marination predrying treatments, dried at 608C for 10 h, and stored at 258C for 60 d a Steps and time Inoculated with acid-adapted E. coli O157:H7 Inoculated with non acid-adapted E. coli O157:H7) Processing 0 h (0.002) 4 h (0.004) 7 h (0.002) 10 h (0.149) (0.015) (0.047) (0.098) (0.021) (0.041) (0.033) (0.158) (0.118) (0.025) (0.009) (0.072) (0.016) (0.015) (0.043) (0.141) (0.031) (0.001) (0.019) (0.040) (0.286) (0.008) (0.056) (0.042) (0.028) (0.004) (0.142) (0.168) (0.091) (0.007) (0.057) (0.071) (0.004) (0.011) (0.106) (0.008) (0.004) Storage Day (0.020) Day (0.022) Day (0.002) (0.013) (0.053) (0.031) (0.014) (0.007) (0.021) (0.012) (0.003) (0.018) (0.024) (0.027) (0.027) (0.009) (0.001) (0.009) (0.002) (0.045) (0.047) (0.054) (0.008) (0.035) (0.021) (0.023) (0.006) (0.008) (0.009) (0.008) a C, inoculated beef slices received no predrying treatment or marinade before refrigeration at 48C for 24 h and drying; TM, inoculated beef slices were marinated with traditional marinade (ph 4.3), held at 48C for 24 h, and then dried; MM, inoculated beef slices were marinated with modi ed marinade (double amount, added 1.2% lactate, 9% acetic acid, and soy sauce with 5% ethanol; ph 3.0), held at 48C for 24 h, and then dried; AATM, inoculated beef slices were dipped into 5% acetic acid solution (ph 2.5) for 10 min at ambient temperature, drained for 2 min and then marinated with traditional marinade before holding at 48C for 24 h, and then dried; TWTM, inoculated beef slices were dipped into 1% Tween 20 (ph 6.6) for 15 min and then into 5% acetic acid solution for 10 min at ambient temperature, marinated with traditional marinade before holding at 48C for 24 h, and then dried. whole muscle jerky after marination (ph 4.2) at temperatures lower than 62.88C was not adequate for extensive reduction of E. coli O157:H7. In contrast, Harrison and Harrison (17) reported that drying TM whole muscle jerky at 608C for 10 h was suf cient to deliver.5.0 log reduction of E. coli O157:H7, L. monocytogenes, and Salmonella Typhimurium, as determined with selective agar media. In another study, Harrison et al. (20) compared the effectiveness of TM, oven heating at 718C before drying, boiling in marinade before drying, and postdrying heating of TM jerky at 57.38C for 10 min in reducing numbers of E. coli O157: H7, L. monocytogenes, and Salmonella on whole muscle beef jerky during drying at 608C for 10 h. Their results showed that heating and boiling before drying reduced the numbers of all three pathogens before drying below the detection limit on selective agar media (SMAC for E. coli O157:H7), indicating an immediate action. At the end of drying, total reductions from all treatments were equal to or greater than 5.8, 3.9, and 4.6 logs for E. coli O157:H7, L. monocytogenes, and Salmonella, respectively, even with TM. The authors also evaluated the sensory attributes of the products and concluded that oven heating jerky after drying resulted in a safer, yet acceptable jerky product. In the present study, TM resulted in lower bacterial reductions compared with MM, AATM, and TWTM during drying. In addition, there was a 1.6- to 2.0-log difference in bacterial populations in TM product between nonselective (TSAP) and the most selective agar (SMAC), indicating presence of injured cells. Although not expected to repair their injury on jerky, sublethally injured pathogenic cells may contaminate the environment and other foods, and if they repair their injury, they may become a health risk. Jordan et al. (24) reported that combinations of lactate (50 mmol), ethanol (5%), and high acidity (ph 3.0 by HCl) signi cantly (P, 0.05) reduced the viability of exponential-phase, stationary-phase, and acid-habituated cells (grown to midexponential phase at ph 5.8) of E. coli O157: H7 in TSB. However, stationary-phase cells and habituated cells were more resistant to the combinations compared with exponential-phase cells or nonhabituated cells. Their results further indicated that such combinations of lactate, ethanol, and high acidity killed the cells by disrupting ph homeostasis and leading to changes in gene expression and enzyme activity. This phenomenon was modi ed and adopted as a predrying treatment (MM) in the present study by using commercially available soy sauce containing approximately 5% ethanol and adding 2% of a 60% sodium lactate preparation (resulting in 1.2% nal concentration) and 9% acetic acid to TM. The volume of marinade solution was also increased by using double the amount of soy sauce and Worcestershire sauce to deliver a more highly concentrated marinade solution than TM. The MM was shown to be signi cantly more effective in reducing the populations of nonadapted E. coli O157:H7 cells on beef slices during drying compared with TM and C. These ndings indicated that combinations of lactate, ethanol, and acetic acid may be effective in inactivating pathogenic bacteria in processed foods. Dipping meat slices in a 5% acetic acid solution before TM (AATM) and subsequent cold storage (48C, 24 h) be-

10 J. Food Prot., Vol. 65, No. 9 INACTIVATION OF E. COLI O157:H7 ON BEEF JERKY DURING DRYING AND STORAGE 1403 fore drying at 608C was also signi cantly more effective in reducing the numbers of the pathogen than the C and TM treatments. It is well documented that E. coli O157:H7 may survive in acidic foods for extended periods, particularly at low temperatures (e.g., 48C). However, such survival has been shown to decrease at elevated temperatures (7, 8, 11, 34, 38), as has been documented with E. coli O157:H7 in fermented sausage (e.g., ph 4.8), which decreased dramatically during storage at ambient (e.g., 258C) compared with lower temperatures (48C) (7, 8). Similarly, treatments TWTM, AATM, and MM, which had ph (4.7) values similar to those of fermented sausage, resulted in greater inactivation during storage at 258 compared with the higher ph of C and TM treatments. The effectiveness of an acetic acid dip as a predrying treatment can also be explained by the results of a recent report (37) that demonstrated that an initial low ph shock (in TSB, ph 3.5 with HCl, for 24 h) followed by exposure to low a w (0.90) was signi cantly more effective in reducing numbers of bacteria compared with an initial exposure to low a w followed by exposure to low ph. These authors (37) hypothesized that initial acid stress may lead to large energy depletion of the cells and subsequently may sensitize cells to other environmental stresses such as low a w or heat. On the other hand, initial exposure to a low a w environment may cause cells to synthesize osmoprotective compounds and, thus, remain in a state of low metabolic activity with a lower expenditure of energy, which may lead to increased resistance to secondary stresses such as acidity. In the present study, bacterial cells were exposed to acetic acid (shock) rst and then to reduction of a w (drying), simulating the aforementioned study in a food processing system (37). This may provide an additional explanation as to why marination with acid (MM, AATM, and TWTM) enhanced bacterial inactivation during drying. It is known that bacterial cells attached to the surface of a product such as meat become more resistant to stress factors such as heat than nonattached cells (23). Another study has shown that spraying beef carcasses inoculated with high levels of E. coli O157:H7 with 5% Tween 20 before spraying with 2% lactic acid resulted in signi cantly higher reductions of the pathogen compared with spraying with lactic acid alone or water (9). It was speculated that Tween 20 may loosen or prevent cellular attachment on the meat surface through its surfactant and hydrophobic effects, thus making cells more vulnerable to the effect of subsequent acid exposure (9). This proposed carcass decontamination method was applied as a predrying treatment in the present study (TWTM) before acetic acid dip and marinating with a traditional recipe. The results indicated that this treatment resulted in higher reductions of bacterial populations in the product than dipping into acetic acid alone and marination, particularly when the product was inoculated with acid-adapted culture. For example, MEMB and SMAC agar populations of bacteria from TWTM jerky inoculated with acid-adapted E. coli O157:H7 were signi - cantly lower than AATM at 4 and 7 h but not at 10 h during drying. In general, populations of bacteria were reduced linearly in all treatments until 7 h of drying, which was followed by a slower reduction or no change. This decelerated reduction appeared as a tailing effect. Shadbolt et al. (36) studied the mechanism of this phenomenon in broth with E. coli M23 D2:H2 at low a w levels (0.75 to 0.9) and at incubation temperatures permitting growth (15 to 458C). Their results indicated that bacterial numbers declined under these conditions and a tailing effect was observed. These investigators also showed that addition of chloramphenicol to broth reduced the occurrence of the tailing effect and concluded that bacteria may synthesize an intracellular de novo protein under low a w conditions that has an osmoprotective effect. This mechanism may not apply to the tailing effect observed in the present study largely because bacteria were not kept under growth-permitting conditions for extended periods. In addition, E. coli O157: H7 cells were exposed to multiple stresses during drying, which were continually changing (i.e., a w and temperature). In another study, Datta and Benjamin (12) compared acid resistance response of stationary-phase E. coli O157:H7 cells in high-density (10 9 CFU/ml) and low-density (10 7 CFU/ml) populations in ph 2.5 Luria broth at 378C for 2 h. The results indicated that cells were more resistant to acid conditions in low-density compared with high-density populations. These researchers suggested that stationaryphase cells may produce a novel diffusible substance that at high concentrations sensitizes cells to low ph. Low-density cells may become more resistant to acid stress simply because of a lower concentration of this novel substance compared with high-density cell suspensions. It has been shown that population density is monitored in several gramnegative bacterial populations by a group of diffusible compounds known as homoserine lactones (40). Elevated levels of these compounds in high-density bacterial populations may play a role in response to stress conditions. To our knowledge, it is unknown if this phenomenon can occur in solid food environments. However, this may provide a potential explanation of the tailing effect observed in the present study. Bacterial reductions may slow after the populations are reduced to a certain level on jerky during drying, potentially due to reduced concentration of potential novel diffusible substances in the environment. This potential scenario may also explain why the tailing effect was less evident in TM jerky on which bacterial reduction was slower compared with other treatments. The tailing effect, however, may have also been due, at least in part, to product case hardening, any potential effects of high altitude, the low humidity conditions of Colorado (1), or other factors. It has been established that acid adaptation of E. coli O157:H7 enhances survival of the pathogen in acidic environments and causes increased cross-protection to other stress factors, such as heat, low a w, and preservatives (4, 5, 16, 29, 32 34). However, contradictory data also exist. For example, Ryu et al. (35) compared the survival of acidadapted, nonadapted, and acid-shocked cells of E. coli O157:H7 in dried beef powder during storage. They reported that there was no signi cant difference between survival of these cell types of the pathogen. The authors concluded that acid adaptation did not result in cross-protection