Prevalence of Escherichia coli in Apple Cider Manufactured in Connecticut

Transcription

1 Journal of Food Protection, Vol., No.,, Pages Copyright, International Association of Milk, Food and Environmental Sanitarians Prevalence of Escherichia coli in Apple Cider Manufactured in Connecticut DOUGLAS W. DINGMAN* Connecticut Agricultural Experiment Station, Huntington Street, P.O. Box, New Haven, Connecticut, USA MS -: Received September /Accepted January ABSTRACT Cider samples obtained from cider mills operating in Connecticut during the to production season were tested for the presence of Escherichia coli. Cider production began in mid August and continued through March, with peak production in September and October. Of cider samples tested, (%) were found to contain E. coli. Of the mills, (%) tested positive for E. coli in the cider at least once during the production year. E. coli was first observed in cider samples produced in mid to late October and was not detected in samples made after January. A trend was observed for cider to decrease in acidity and increase in o Brix (soluble sugars) throughout the production season. No correlation between ph and soluble sugars of cider and the presence of E. coli was detected. Eight mills used both dropped apples and tree-picked apples, whereas three mills used tree-picked apples only. The use of dropped apples in cider production began weeks before the first detection of E. coli in cider. E. coli was isolated from cider samples produced using dropped apples and from samples produced using only tree-picked apples. No direct correlation between the use of dropped apples or tree-picked apples and the presence of E. coli in the cider was observed. An association between the time of apple harvest and the appearance of E. coli in cider was noted. For mills providing adequate records, all contaminated cider was produced from apples harvested between mid October and mid November. Escherichia coli O:H was first recognized as a pathogen capable of causing hemorrhagic colitis and hemolytic uremic syndrome in the early s (, ). In the past, most outbreaks reported to be caused by this pathogen were due to the consumption of E. coli contaminated meat (i.e., ground beef) (, ). This microbe is a persistent problem in cattle (,, ). However, outbreaks of disease involving other types of food (fresh produce, fruit juices, etc.) have been reported with increasing frequency (,,, ). The first reported outbreak of disease presumably caused by this pathogen and associated with the consumption of fresh apple juice (cider) occurred in in Canada (). In the United States, three reported outbreaks of disease associated with the consumption of apple cider have occurred in the last decade (,, ). For this article, apple cider is defined as a fresh unfermented juice extracted from apples that has not been clarified or heat treated. The first of the U.S. outbreaks occurred in Massachusetts in. This outbreak resulted in reported cases of E. coli O: H infection and four children developed hemolytic uremic syndrome. Although apples from the orchard floor ( drops ) of two outside apple suppliers were used in the production of the cider and the cider press operator also raised cattle, no direct source of E. coli O:H was identified. The next two U.S. outbreaks associated with apple cider both occurred in the fall of, one in Connecticut and the other in Washington. These outbreaks resulted in a total of reported cases and one death (, ). In the Wash- * Author for correspondence. Tel: --; Fax: --. ington outbreak, the source of the pathogen was linked to the consumption of contaminated unpasteurized apple juice. E. coli O:H isolates from the contaminated apple juice had a restriction fragment length polymorphism DNA fingerprint pattern identical to the DNA fingerprint patterns of isolates obtained from infected individuals (). No E. coli isolate was obtained from the cider in regard to the Connecticut outbreak. These outbreaks, occurring so closely together, have raised considerable concern regarding the safety of fresh fruit juices in general and apple cider in particular. A public forum hosted by representatives of the Food and Drug Administration (FDA) was held in Arlington, Va., on December and, (), to address this concern and discuss ideas for eliminating this threat. From this meeting, it became apparent that there was a need for extensive research on fruit juice production methods and on fruit harvest and storage practices to provide knowledge on which to base regulatory decisions for the safe production of fresh juices. To date, to my knowledge, no study has been conducted to determine the occurrence of E. coli in apple cider throughout a production season. Therefore, it is not possible to determine with any certainty the critical factors that contribute to the introduction of this microbe and various pathogenic strains into cider during manufacture. This lack of certainty about critical factors, and the overall lack of information on microbial matters pertaining to fresh juice manufacturing in general, makes addressing the concern of contamination of apple cider by E. coli O:H a guesswork issue. The objective of this study was to profile the

2 DINGMAN J. Food Prot., Vol., No. incidence of E. coli in cider manufactured in Connecticut. Information obtained from apple growers and cider producers on apple harvest and storage times and apple type, coupled with the incidence profile, was used to help identify possible factors that might have contributed to the contamination of Connecticut cider by E. coli. MATERIALS AND METHODS Assessment of E. coli survival in frozen media. Refrigerated apple cider ( ml) and Difco tryptic soy broth (TSB; Difco Laboratories, Detroit, Mich.) ( ml; adjusted to ph. with. N HCl) were inoculated using l of a : dilution of a culture of E. coli O:H gfp-ec () grown for h in TSB at C. Each inoculated medium was divided into three portions: one portion was stored at room temperature ( to C), one portion was refrigerated ( C), and -ml aliquots of the third portion were sealed into -ml plastic culture tubes and frozen in a frost-free freezer at C. The number of viable cells in the inoculum was determined by serial-dilution plating of the : dilution of the -h culture onto Difco tryptose broth, g/liter; Difco yeast extract,. g/liter; glucose, g/liter; Na HPO H O,. g/liter; NaCl,. g/liter; sodium pyruvate, g/liter; and Difco Bacto agar, g/liter (TPAP) (). At day, and various days following, CFU/ml determinations were made on the three differently stored portions of each inoculated medium by serial-dilution plating onto TPAP and incubation at C. For frozen samples, serial-dilution plating was done in triplicate. E. coli O:H gfp- ec CFUs were determined by counting green fluorescent colonies on plates held under a UV lamp ( nm). Collection and storage of cider samples. Cider samples from cider mills located in Connecticut were obtained throughout the August to March production season. For each cider batch produced, mill operators were asked to immediately (or within a few days) freeze ( C) a small portion of the finished product. Participants in this study were also asked to record the date of cider production and the type, cultivar, and length of storage of apples used in the production of each batch of cider. Apple type was defined as drops or tree-picks (apples harvested directly from the tree). The frozen cider samples were collected from the mills on a biweekly schedule, stored cold in transport, and on return to the laboratory stored at C until testing could be performed. Testing of the cider was done within weeks of the collection date and within to weeks of the date of production. The mills used in this study represented a spectrum of the types of cider manufacturing operations in Connecticut with regard to volume of production, source and type of fruit used, and type of operation (Table ). There were licensed mills operating in Connecticut in (Apple Juice Cider Roster, Connecticut Department of Agriculture, Hartford, Conn.). Assay of apple cider for the presence of E. coli. Each frozen cider sample was thawed to completion using a stream of warm water, shaken vigorously, and placed on ice. Then. ml was plated directly onto Levine eosin-methylene blue (LEMB; Difco Laboratories) plus pyruvate (.%) and TPAP plus -bromo--chloro--indolyl -D-galactopyranoside (X-gal; Sigma-Aldrich Corp.; g/ml) plus -methylumbelliferyl- -D-glucuronide (MUG; Amersham Pharmacia Biotech; g/ml) agar plates. Pyruvate was added to LEMB plates to improve detection of acid-stressed cells of E. coli (, ), and X-gal and MUG were used as indicators to identify potential E. coli colonies (, ). A -fold cider concentrate was produced by centrifugation of cider ( ml) at, rpm for min at C in a Sorvall SS- rotor TABLE. Profile of participating Connecticut cider mills and the incidence of E. coli in cider Mill number Volume produced (U.S. gal./ pressing),,,,,,, daily b,,,, Source of apples a Tree-picked Tree-picked Supplier Supplier (TP c ) Total samples Tested E. coli presence (%) (%) (%) (%) (%) (%) Total (%) a Apples used in the production of cider were either obtained from the ground (drops) or hand-picked directly from the tree (treepicked). All mills used apples obtained from an orchard owned and operated by the mill owner or from an independent supplier of apples (supplier). b Cider was manufactured on a daily basis. c The independent apple supplier was required by the cider mill operator to provide only tree-picked apples. (Sorvall Products, L. P., Newtown, Conn.) followed by suspension of the resulting pellet into TSB (. ml). This concentrate (. ml) was plated onto LEMB plus pyruvate and TPAP plus X-gal plus MUG agar plates, and inoculated plates were incubated for to days at C and then observed for potential E. coli colonies. Metallic green colonies growing on LEMB plates plus blue pigmented and/or fluorescent white colonies (viewed under UV light; nm) appearing on TPAP plus X-gal plus MUG agar plates were picked as potential E. coli isolates and spot tested onto LEMB plates and sorbitol-macconkey agar plates (SMAC; D-sorbitol, g/liter; MacConkey agar base, g/liter) () to aid in identification. Potential E. coli cultures were also streak isolated onto TPAP plates to obtain pure cultures and then tested using API E System kits (biomérieux Vitek, Inc.) to confirm or negate the isolate as E. coli. Measurement of cider ph and o Brix. The ph of cider was measured using an Orion Research Model analog ph meter (Orion Research Inc., Boston, Mass.) standardized to ph. and cider samples equilibrated to room temperature. The measurement of cider soluble sugars ( o Brix) was performed on room temperature cider samples using a Fisherbrand (Fisher Scientific, Pittsburgh, Pa.) hand-held Brix refractometer standardized using H O at room temperature. RESULTS AND DISCUSSION Effect of freezing on the survivability of E. coli in cider. Because of the irregularity with which cider was produced at the cider mills and the distances that the mills were located from the laboratory, it was not possible to collect and assay the samples immediately after production. Therefore, a cider storage method that would not interfere with the sensitivity of the assay was necessary before this

3 J. Food Prot., Vol., No. E. COLI IN APPLE CIDER temperature cider). TPAP (used in this experiment) is reported to revive acid-stressed cells (, ) and possibly accounted for this increased measurement in the survivability of E. coli in cider in relationship to the survivability shown in the earlier reports. Subsequent work on the isolation of E. coli from frozen cider samples used TPAP medium. Freezing and refrigeration of cider did not hinder the survival of E. coli in cider within a -month and -month period, respectively. Therefore, freezing was used in this study to preserve cider samples before assay for E. coli presence. FIGURE. Survival of E. coli O:H gfp-ec in (A) apple cider and (B) TSB (ph.). Symbols indicate storage at different temperatures as follows:, to C;, C; and, C. Data points for the to C and C samples are the average obtained from counting a minimum of two plates. Data points for the C samples are the average obtained from counting three plates from triplicate platings. study could be performed. Frozen storage of cider was tested as a way to meet this necessity. The effect of storage conditions on the survival of E. coli O:H in cider (ph.) and the comparative control, TSB (ph.), is shown in Figure. Storage of cider at room temperature resulted in an approximate log reduction in E. coli O:H gfp-ec CFU within days. In comparison, storage of this microbe at room temperature in TSB (ph.) resulted in negligible loss in CFU for about days followed by alog decline in CFU during the next days. Refrigeration ( to C) and freezing ( C) of cider preserved E. coli O:H gfp-ec during a -day period. By days, refrigerated cider had exhibited a log decline in CFU, whereas frozen cider exhibited only an approximate % drop. Storage in TSB (ph.) under refrigeration and freezing conditions resulted in stable count and rapid decline to the E. coli cell number, respectively. A likely explanation for the differences in E. coli survival observed between the freezing of cider and TSB (ph.) is that the soluble sugars in cider acted as cell stabilizers under freezing conditions (, ). The results reported herein are in general agreement with earlier reports on storage of E. coli in cider under refrigerated and nonrefrigerated conditions (, ). However, the survival times measured in this study (Figure ) were longer than in the earlier reports (i.e., between and days for refrigerated cider and and days for room Presence of E. coli in cider. Of the apple cider samples produced and collected throughout the production season, (%) contained E. coli (Table ). All the isolates fermented sorbitol, suggesting that they were not E. coli O:H (, ). E. coli was detected in samples from (%) of the mills. Except for mill, the mills in which E. coli had been found in the cider had a contamination level of to % of the seasonal production. It is unknown why mill had such a high level of contamination. Reliable comparison studies could not be made between the presence of E. coli in the cider and the total microbial count of the cider. Although freezing of cider did not likely lower E. coli cell counts during months of storage time (Figure A), freezing was observed to lower the total microbial count of cider (data not shown). Maximum cider production occurred in late September and early October for all mills (Fig. ). Samples collected from cider produced in September and October accounted for % (% for the six mills in which E. coli was found in the cider) of the total samples collected during the -month study and the seasonal pattern of production of the six contaminated mills was similar to the pattern of production for the mills. E. coli was first isolated from cider manufactured in late October and last found in cider produced at the end of January. This produced an apparent window in the time of appearance of E. coli within the cider production season. It is interesting that the three documented U.S. outbreaks of E. coli O:H associated with contaminated cider occurred in mid October to mid November (,, ). One interpretation for this window of E. coli appearance in Connecticut-manufactured cider is that it is due to a sampling effect (i.e., higher number of samples tested providing a higher likelihood of isolating E. coli). This is not believed to have happened because the span of the window (weeks to ) was about week later than the span of maximum production (weeks to ) (Fig. ) and occurred at a time during the decline in the production of cider. Also, during API E testing of potential E. coli isolates obtained from the cider samples, Pantoea (formerly Enterobacter) agglomerans (a coliform that naturally inhabits soil and plant material) (,, ) was randomly identified throughout the study period (data not shown). This window is also not likely due to an effect of a change in the apparent susceptibility of the cider. Cider used in the freezing study was manufactured in early September. I suggest that this apparent window in the presence of E. coli in

4 DINGMAN J. Food Prot., Vol., No. FIGURE. Weekly cider production pattern for (A) all participating cider mills, (B) the cider mills (mills,,,,, and ) in which E. coli contaminated cider had been found, and (C) the cider samples identified as containing E. coli. Connecticut-manufactured cider was attributed to a common unidentified factor(s) that occurred in the production of cider during this period. Discussions on sources of possible fecal contamination of cider at the FDA meeting in Arlington, Va. (), focused on the use of dropped apples, contamination from apple handling by workers (during harvesting and pressing), and unsanitary pressing conditions in cider mills. For Connecticut-manufactured cider, use of dropped fruit occurred primarily at the beginning of the production season but also extended through most of the season (Table ). Apple handling by workers and cider mill operating conditions were consistent throughout the production season. If any of these three possible sources for fecal contamination were a major factor contributing to the appearance of E. coli in cider, then E. coli cider contamination would occur randomly throughout the production season and not as shown in Figure. Therefore, the three potential sources of fecal contamination of cider presented at the FDA meeting were apparently not key factors contributing to contamination of Connecticut-manufactured cider in this study. Correlation of E. coli presence in cider with apple type and date of harvest. Only records obtained from three of the mills in which E. coli contaminated cider was produced (mills,, and ) were adequate for constructing a profile of apple use based on the cider production date (Table ). These mills generally used apples from their own orchards and accounted for of the incidences in which E. coli was found in cider. These mills primarily used the dropped apples in the first weeks of cider production, while tree-picked apples were used throughout the to cider production season (Table ). E. coli was found in three samples in which dropped apples were used for cider production (weeks,, and ). However, for weeks to the mills used dropped apples extensively without the presence of E. coli. E. coli was also isolated from cider samples in which only tree-picked fruit was used (weeks,, and ). In a fourth mill, E. coli was also detected in cider where only tree-picked apples were used (mill ; Table ). Additionally, apples used for production of the cider that resulted in the outbreak of E. coli O:H in Washington were reportedly tree picked (). In general, dropped fruit would have a higher probability of contamination with animal feces than tree-picked fruit. High occurrences of E. coli have been reported for dropped fruit obtained from orchards fertilized with animal manure before use in the production of cider (). However, the present study did not demonstrate that the use of dropped apples (especially those obtained early in the production season) were any more likely to have caused E. coli contamination of Connecticut-manufactured cider than were use of tree-picked apples (Table ). Also, animal manure is not generally used to fertilize apple orchards in Connecticut. Cider production records were also used to construct a profile of apple use based on the date of apple harvest (Table ). A sharp window of appearance of E. coli in cider produced from apples harvested from weeks to (mid

5 J. Food Prot., Vol., No. E. COLI IN APPLE CIDER TABLE. Number of apple cultivars used in cider production based on cider production date Source of apples b TABLE. Number of apple cultivars used in cider production based on apple harvest date Source of apples b Production week a Mill Mill Mill Week of harvest a Mill Mill Mill () c () () () () () () a Week is the first week in September starting Sunday, August,. b Apples used in the production of cider were either harvested from the ground (drops) or from the tree (tree-picked). c Numbers in parentheses indicate the number of apple cultivars used that week in the production of cider contaminated with E. coli. October to mid November) was observed. No specific apple cultivar(s) harvested between weeks and appeared to be associated with E. coli contamination of cider. However, the cultivars McIntosh, Red Delicious, Golden Delicious, Northern Spy, Jonagold, Mutzu, Melrose, and Cortland were predominantly used in cider found to be contaminated by E. coli (data not shown). Use of apple cultivars harvested as drops or tree picked before week did not result in the production of E. coli contaminated cider. Maximum apple harvest for the three mills was late September to early October (weeks to ), with final harvest in early November (weeks to ) (personal communication). These observations suggest that factors associated with harvest practices common to the three mills were the primary source for E. coli contaminated cider in this study. The observation that other mills in which E. coli contaminated cider was produced also had E. coli contamination that occurred within this period (Fig. ) suggests that common harvest factors occurred for them as well. () () c () () () () () () () () () a Week is the first week in September starting Sunday, August,. b Apples used in the production of cider were either harvested from the ground (drops) or from the tree (tree-picked). c Numbers in parentheses indicate the number of apple cultivars harvested that were later used in the production of cider contaminated with E. coli. One factor common to apple harvest is apple storage practices in relation to the weather. Following picking in late September to early October, apples used for cider production at mills,, and were sometimes temporarily stored in warehouses or outside. Apples at mill were routinely stored outside the mill for several days after delivery (personal observation). Increasingly cold weather and scarcity of food sources likely induced rodents and other small animals to migrate from the field into the warehouses or to outside storage areas. A killing frost occurred in Mt. Carmel, Conn., on October (week ) and a cold rain followed by snow occurred on November and and November through (weeks and ; Connecticut Agricultural Experiement Station weather data), respectively. These weather conditions, coupled with exposed apple storage practices, could have been critical factors for determining the harvest-limited time span for E. coli contamination of cider presented in Table. Duration of apple storage may also have been a contributing critical factor. Of the E. coli contaminated cider samples collected from mill, all cider was produced using apples that had been in storage for a minimum of week. For mills and, apples used in cider found to be contaminated with E. coli had been stored a minimum of and weeks, respectively. Correlation of E. coli presence in cider to cider ph and o Brix. To determine whether various physical properties of cider might contribute to the presence of E. coli in cider, samples collected throughout the study were mea-

6 DINGMAN J. Food Prot., Vol., No. FIGURE. ph and o Brix profile of cider samples obtained from the mills (mills,,,,, and ) in which E. coli contaminated cider had been found. (A) ph profile of the cider samples. (B) o Brix profile of the cider samples. Open squares ( ) represent ph and o Brix values of the cider samples containing E. coli. sured for ph and o Brix. For the mills, cider ph ranged between. and., and the o Brix was between and throughout the production season (data not shown). Samples at or above ph. were randomly observed throughout the study, and several samples had little soluble sugar, possibly due to the addition of water to the cider. A trend of increased ph and o Brix was observed in cider as the production season progressed. Figure (A and B) shows the ph and o Brix, respectively, of the six mills that were detected positive for E. coli in the cider. Individual cider samples containing E. coli (marked as squares) were apparently not limited by ph or o Brix. Based on Figure, no correlation between cider ph or between cider o Brix and the presence of E. coli in the cider was apparent. Although changes in ph and o Brix of the cider during the production season were not found to be correlated with the presence of E. coli, changes in ph or sugar content of an individual apple during ripening may increase the susceptibility of the apple to growth of E. coli. Apple harvesting late in the season (late October to early November) may have included dropped or possibly overripe tree-picked apples that could better support the growth of E. coli. ph measurement of bruises in ripe McIntosh apples were sometimes found to be near ph. (data not shown). This change in the physiology of an apple (which may be associated with apple cultivar) could be another common critical factor that contributes to the window for the appearance of the E. coli contamination of cider. These common factors point to the quality or the nature of the fruit being used in cider production as contributing to the problem of cider contamination. The four outbreaks caused by E. coli O:H (,,, ) and the outbreaks caused by Cryptosporidium () and Salmonella () have all been attributed to the fruit. In the future, we will perform research to understand the apple E. coli association (ground and apple bruise temperature and ph over the season, growth in apple bruises, and rodent exposure) to help clarify the problem of E. coli contamination of cider. ACKNOWLEDGMENT I thank S. M. Douglas for helpful suggestions in the preparation of the manuscript and Cindy Musante for technical assistance. I also thank Susan Fratamico for kindly providing the gfp-containing E. coli strain and the owners/operators of the various Connecticut cider mills for agreeing to participate in this study.

7 J. Food Prot., Vol., No. E. COLI IN APPLE CIDER REFERENCES. Beji, A., J. Mergaert, F. Gavini, D. Izard, K. Kersters, H. Leclerc, and J. DeLay.. Subjective synonymy of Erwinia herbicola, Erwinia milletiae and Enterobacter agglomerans and redefinition of the taxon by genotypic and phenotypic data. Int. J. Syst. Bacteriol. :.. Besser, R. E., S. M. Lett, J. T. Weber, M. P. Doyle, T. J. Barret, J. G. Wells, and P. M. Griffin.. An outbreak of diarrhea and hemolytic uremic syndrome from Escherichia coli O:H in freshpressed apple cider. JAMA :.. Brenner, K. P., C. C. Rankin, Y. R. Roybal, G. N. J. Stelma, P. V. Scarpino, and A. P. Dufour.. New medium for the simultaneous detection of total coliforms and Escherichia coli in water. Appl. Environ. Microbiol. :.. Centers for Disease Control and Prevention.. Salmonella typhimurium outbreak traced to a commercial apple cider-new Jersey. Morbid. Mortal. Weekly Rep. :.. Calabrese, J. P., and G. K. Bissonnette.. Improved detection of acid mine water stressed coliform bacteria on media containing catalase and sodium pyruvate. Can J. Microbiol. :.. Centers for Disease Control and Prevention.. Foodborne outbreaks of enterotoxigenic Escherichia coli Rhode Island and New Hampshire,. Morbid. Mortal. Weekly Rep. :.. Centers for Disease Control and Prevention.. Outbreak of Escherichia coli O:H infections associated with drinking unpasteurized commercial apple juice British Columbia, California, Colorado, and Washington, October. Morbid. Mortal. Weekly Rep. :.. Centers for Disease Control and Prevention.. Outbreaks of Escherichia coli O:H infection and Crytosporidiosis associated with drinking unpasteurized apple cider Connecticut and New York, October. Morbid. Mortal. Weekly Rep. :.. Centers for Disease Control and Prevention.. Outbreaks of Escherichia coli O:H infection associated with eating alfalfa sprouts Michigan and Virginia, June July. Morbid. Mortal. Weekly Rep. :.. Doyle, M. P., and J. L. Schoeni.. Isolation of Escherichia coli O:H from retail fresh meats and poultry. Appl. Environ. Microbiol. :.. Ewing, W. H., and M. A. Fife.. Enterobacter agglomerans (Beijernick) comb. nov. (the Hericola-Lathyri bacteria). Int. J. Syst. Bacteriol. :.. Feng, P.. Escherichia coli serotype O:H: novel vehicles of infection and emergence of phenotypic variants. Emerging Infect. Dis. :.. Food and Drug Administration.. Current science and technology on fresh juices; notice of meeting. Fed. Register :.. Fratamico, P. M., M. Y. Deng, T. P. Strobaugh, and S. A. Palumbo.. Construction and characterization of Escherichia coli O: H strains expressing firefly luciferase and green fluorescent protein and their use in survival studies. J. Food Prot. :.. Gaudet, I. D., L. Z. Florence, and R. N. Coleman.. Evaluation of test media for routine monitoring of Escherichia coli in nonpotable waters. Appl. Environ. Microbiol. :.. Gavini, F., J. Mergaert, A. Beji, C. Mielcarek, D. Izard, K. Kersters, and J. DeLay.. Transfer of Enterobacter agglomerans (Beijerinck ) Ewing and Fife to Pantoea gen. nov. as Pantoea agglomerans comb. nov. and description of Pantoea dispersa sp. nov. Int. J. Syst. Bacteriol. :.. Gherna, R. L.. Preservation, p.. In P. Gerhardt, R. G. E. Murray, R. N. Costilow, E. W. Nester, W. A. Wood, N. R. Krieg, and G. B. Phillips (ed.), Manual of methods for general microbiology. American Society for Microbiology, Washington, D.C.. Goverd, K. A., and F. W. Beech.. The occurrence and survival of coliforms and Salmonellas in apple juice and cider. J. Appl. Bacteriol. :.. Griffin, P. M., and R. V. Tauxe.. The epidemiology of infections caused by Escherichia coli O:H, other enterohemorrhagic E. coli, and the associated hemolytic uremic syndrome. Epidemiol. Rev. :.. Hancock, D. D., T. E. Besser, M. L. Kinsel, P. I. Tarr, D. H. Rice, and M. G. Paros.. The prevalence of Escherichia coli O: H in dairy and beef cattle in Washington state. Epidemiol. Infect. :.. Karmali, M. A., B. T. Steele, M. Petric, and C. Lim.. Sporadic cases of haemolytic-uraemic syndrome associated with faecal cytotoxin and cytotoxin-producing Escherichia coli in stools. Lancet :.. Leyer, G. J., L. Wang, and E. A. Johnson.. Acid adaptation of Escherichia coli O:H increases survival in acidic foods. Appl. Environ. Microbiol. :.. March, S. B., and S. Ratnam.. Sorbitol-MacConkey medium for detection of Escherichia coli O:H associated with hemorrhagic colitis. J. Clin. Microbiol. :.. Ojeda, A., V. Prado, J. Martinez, C. Arellano, A. Borczyk, W. Johnson, H. Lior, and M. M. Levine.. Sorbitol-negative phenotype among enterohemorrhagic Escherichia coli strains of different serotypes and from different sources. J. Clin. Microbiol. :.. Riley, L. W., R. S. Remis, S. D. Helgerson, H. B. McGee, J. G. Wells, B. R. Davis, R. J. Hebert, E. S. Olcott, L. M. Johnson, N. T. Hargrett, P. A. Blake, and M. L. Cohen.. Hemorrhagic colitis associated with a rare Escherichia coli serotype. N. Engl. J. Med. :.. Shere, J. A., K. J. Bartlett, and C. W. Kaspar.. Longitudinal study of Escherichia coli O:H dissemination on four dairy farms in Wisconsin. Appl. Environ. Microbiol. :.. Steele, B. T., N. Murphy, G. S. Arbus, and C. P. Rance.. An outbreak of hemolytic uremic syndrome associated with ingestion of fresh apple juice. Clin. Lab. Observations :.. Teague, O., and A. W. Clurman.. A method of preserving typhoid stools for delayed examination and a comparative study of the efficacy of eosin brilliant-green agar, eosin methylene-blue agar, and endo agar for the isolation of typhoid bacilli from stools. J. Infect. Dis. :.. Zhao, T., M. P. Doyle, and R. E. Besser.. Fate of enterohemorrhagic Escherichia coli O:H in apple cider with and without preservatives. Appl. Environ. Microbiol. :.. Zhao, T., M. P. Doyle, J. Shere, and L. Garber.. Prevalence of enterohemorrhagic Escherichia coli O:H in a survey of dairy herds. Appl. Environ. Microbiol. :.