Article critique

1637

Journal of Food Protection, Vol. 66, No. 9, 2003, Pages 1637–1641 Copyright q, International Association for Food Protection

Inactivation of Escherichia coli O157:H7, Listeria monocytogenes, and Salmonella in Cranberry, Lemon,

and Lime Juice Concentrates

MARA C. L. NOGUEIRA,1,2 OMAR A. OYARZÁBAL,3† A N D DAVID E. GOMBAS1*

1National Food Processors Association, 1350 I Street N.W., Suite 300, Washington, D.C. 20005, USA; 2Department of Microbiology, Biologic Sciences Institute, University of Minas Gerais, Belo Horizonte, Brazil; and 3Neogen Corporation, 620 Lesher Place, Lansing, Michigan

MS 02-460: Received 13 December 2002/Accepted 4 April 2003

ABSTRACT

The production of thermally concentrated fruit juices uses temperatures high enough to achieve at least a 5-log reduction of pathogenic bacteria that can occur in raw juice. However, the transportation and storage of concentrates at low temperatures prior to � nal packaging is a common practice in the juice industry and introduces a potential risk for postconcentration contamination with pathogenic bacteria. The present study was undertaken to evaluate the likelihood of Escherichia coli O157: H7, Listeria monocytogenes and Salmonella surviving in cranberry, lemon, and lime juice concentrates at or above temperatures commonly used for transportation or storage of these concentrates. This study demonstrates that cranberry, lemon, and lime juice concentrates possess intrinsic antimicrobial properties that will eliminate these bacterial pathogens in the event of postcon- centration recontamination. Bacterial inactivation was demonstrated under all conditions; at least 5-log Salmonella inactivation was consistently demonstrated at 2238C (2108F), at least 5-log E. coli O157:H7 inactivation was consistently demonstrated at 2118C (128F), and at least 5-log L. monocytogenes inactivation was consistently demonstrated at 08C (328F).

Until recently, fruit juices were not recognized as ve- hicles of foodborne illness because of their low pH and high organic acid levels (18, 22). However, several out- breaks associated with unpasteurized fruit juices have been reported in the last decade, and as a consequence, issues surrounding the safety of juice products started to be ad- dressed (1). The U.S. Food and Drug Administration (FDA) proposed a hazard analysis and critical control point (HACCP) regulation that includes a performance criterion to assure juice safety (7). The regulation, commonly known as juice HACCP, became effective January 2002 and re- quires juice processors to use methods to achieve a 5-log10 reduction of an appropriate target organism in the juice pro- cess (8).

Escherichia coli O157:H7 and Salmonella serotype Ty- phimurium (3, 5, 6) have been involved in foodborne out- breaks transmitted by unpasteurized apple cider. Salmonella (12) has been linked to outbreaks transmitted by the con- sumption of unpasteurized orange juice. Although Listeria monocytogenes has not been implicated in foodborne out- breaks associated with juice, L. monocytogenes has been isolated from unpasteurized apple juice and an apple-rasp- berry blend (21). Furthermore, acid-adapted strains of L. monocytogenes can survive in acidic foods (10) and be- come potential hazards.

Fruit juices are thermally processed into concentrate by heating to temperatures far in excess of those necessary to

* Author for correspondence. Tel: 202-639-5978; Fax: 202-639-5991; E-mail: [email protected].

† Present address: Department of Poultry Science, Auburn University, Au- burn, AL 36849, USA.

destroy pathogens reasonably likely to be associated with raw juice (15). Therefore, the presence of pathogens in juice concentrates would occur from postconcentration con- tamination, potentially during storage or transportation. In the event that postconcentration contamination with path- ogens cannot otherwise be prevented, the FDA has deter- mined that the requirement for a 5-log reduction would ap- ply to concentrates as well as single-strength juice (8). However, juice concentrate is not typically repasteurized because of detrimental effects to product quality.

Questions have been raised about how long pathogenic bacteria such as E. coli O157:H7, L. monocytogenes, and Salmonella can survive in juice concentrates. According to National Food Processors Association member companies, transportation and storage of fruit juices and juice concen- trates are usually at low temperatures, generally from 223 to 08C, to preserve the quality of the juice. Studies done by the National Food Processors Association have shown that Salmonella, E. coli O157:H7, and L. monocytogenes are able to survive in orange, apple, pineapple, and white grape concentrates and in banana puree at 2238C for at least 12 weeks (20).

Cranberry, lemon, and lime concentrates have intrinsic characteristics (low pH values and high titratable acidities) that create a hostile environment for bacterial growth and survival. Cranberry, lemon, and lime concentrates also con- tain compounds described as having antimicrobial effects (4, 16, 23). The present study is focused on evaluating the likelihood of E. coli O157:H7, Salmonella, or L. monocy- togenes surviving in these concentrates at or above tem- peratures commonly used for transportation or storage of juice concentrates.

J. Food Prot., Vol. 66, No. 91638 NOGUEIRA ET AL.

TABLE 1. Properties of cranberry, lemon, and lime juice con- centrates used in this study

Concen- trate pH

Titratable aciditya 8Brixb

Cranberry Lemon Lime

2.0–2.2 1.8–2.0

2.2

10.5–13.2 30.0–36.0

32.8

45–52.6 48–55

50

a Units are g citric acid/g juice concentrate. b Corrected at 238C.

T A

B L

E 2 .

In a ct

iv a ti

o n

o f

E .

co li

O 15

7 :H

7 ,

S al

m o n el

la ,

an d

L is

te ri

a m

o n oc

y to

g en

es a t

2 2 3 8C

C ra

nb er

ry

E .

co li

S al

m on

el la

L is

te ri

a

L em

o n

E .

co li

Sa lm

o n el

la L is

te ri

a

L im

e

E .

co li

S al

m on

el la

L is

te ri

a

T ri

al :

1 2

1 2

1 2

1 2

1 2

1 2

1 2

1 2

1 2

In oc

u lu

m (l

o g

C F

U /1

0 g )

A Ia 1

h

8 .1 1 N T

b

8 .0 2 2

7 .8 2 N T

7 .7 2 2

8 .0 1 N T

7 .7 1 2

8 .1 2 N T

8 .0 1 1

7 .8 2 N T

7. 9

2 2

8 .0 1 N T

7 .9 1 1

8. 1

2 N T

8 .0 2 1

7 .8 2 N T

7 .7 2 2

8 .0 1 N T

8 .9 1 2

6 h

2 4

h 1

w ee

k 2

w ee

k s

2 2 2 2

2 1 1 2

2 2 2 2

2 2 2 2

2 2 2 2

1 2 1 1

2 2 2 2

2 2 2 2

2 2 2 2

2 2 2 2

2 2 2 2

1 2 1 2

2 2 2 2

2 1 2 2

2 2 2 2

2 2 2 2

2 2 2 2

2 1 2 2

a S

am p le

d w

it hi

n 1 5

m in

af te

r in

o cu

la ti

o n

(A I)

. b

N ot

te st

ed .

MATERIALS AND METHODS

Juice concentrates. The concentrates were obtained from National Food Processors Association member companies at 8Brix consistent with levels used during transportation. The 8Brix levels of the concentrates were determined by using a refractometer (Bausch and Lomb, Rochester, N.Y.). The pH was determined with an Orion 620 pH meter (Orion Research Inc., Boston, Mass.), and the titratable acidity (wt/wt as citric acid) was calculated by NaOH titration to a pH 8.1 endpoint.

Bacterial strains and culture conditions. Composites of � ve strains were prepared for each pathogen used in this study. The E. coli O157:H7 composite included the following National Food Processors Association strains: N-4064, isolated from apple cider involved in a foodborne outbreak; N-4070, isolated from apple juice from a 1996 juice outbreak; N-4072, isolated from apple cider; N-4073, isolated from apple cider from a 1996 Con- necticut outbreak; and N-4087, isolated from an outbreak involv- ing salami. The Salmonella composite included the following juice-isolated serotypes, the � rst three originally obtained from M. E. Parish, University of Florida: N-4019 (Salmonella Rubislaw), N-4020 (Salmonella Gaminara), N-4021 (Salmonella Hartford), N-4088 (Salmonella Enteriditis) isolated from orange juice, and N-4089 (Salmonella Muenchen) isolated from orange juice from a 1999 orange juice outbreak. The L. monocytogenes composite included strains N-7003 (isolated from raw milk), N-7016 (iso- lated from meat), N-7175 (serotype 1/2b, isolated from a meat plant environment), ATCC 19113 (serotype 3, human isolate), and ATCC 7644 (human isolate).

Strains were transferred from tryptic soy agar (Difco Labo- ratories, Detroit, Mich.) slants into individual tryptic soy broth (Difco) tubes and incubated at 358C overnight. Cells were then transferred to tryptic soy broth adjusted to pH 5 with 1 N HCl and incubated at 358C for an additional 24 h. Final acid adaptation was achieved by resuspension of the cells in cold 10 mM citrate buffer, pH 4, and holding overnight at 48C. Cells were composited to obtain approximately equal numbers of each strain. Each strain and composite was enumerated by serial dilution in 0.1% peptone water and spread plating on tryptic soy agar. Plates were incubated at 358C for 24 h.

Inoculation protocol. For all trials at 2238C; all trials in- vestigating E. coli and Salmonella in cranberry, lemon, and lime juice concentrates at 2118C; and the � rst trial for L. monocyto- genes in lemon juice concentrate at 2118C, the juice samples were dispensed in 10-g aliquots into sterile Whirl-Pak bags (Nasco, Fort Atkinson, Wisc.) and brought to or below the target temperature (223 or 2118C) by holding in a temperature-controlled freezer overnight prior to inoculation. In the second trial for L. monocy- togenes in lemon juice concentrate at 2118C, in both trials in lime and cranberry at 2118C, and in the trial for L. monocytogenes in cranberry concentrate at 08C, the concentrate samples were dis-

J. Food Prot., Vol. 66, No. 9 INACTIVATION OF PATHOGENS IN JUICE CONCENTRATES 1639

TABLE 3. Inactivation of E. coli O157:H7 and Salmonella at 2118C; all samples in two trials were inoculated at 6.9 log CFU per 10 g of sample

Trial:

Cranberry

E. coli

1 2

Salmonella

1 2

Lemon

E. coli

1 2

Salmonella

1 2

Lime

E. coli

1 2

Salmonella

1 2

AIa

1 h 6 h

24 h 1 week 2 weeks

1 1 2 2 2 2

1 NTb

2 2 2 2

2 2 2 2 2 2

2 NT 2 2 2 2

1 2 2 2 2 2

1 2 2 2 2 2

2 2 2 2 2 2

2 2 2 2 2 2

1 1 2 2 2 2

1 NT 2 2 2 2

2 2 2 2 2 2

2 NT 2 2 2 2

a Within 15 min after inoculation (AI). b Not tested.

TABLE 4. Inactivation of Listeria monocytogenes in cranberry, lemon, and lime concentrates at 2118C

Concen- trate Sample inactivation

Trial 1 Trial 2 Lemon Inoculuma 6.9 5.9 4.9 3.9 2.9 6.5 5.5 4.5 3.5

AIb

3 h 6 h

2/2c

0/2 0/5

2/2 0/2 0/5

2/2 0/2 0/5

2/2 0/2 0/5

2/2 0/2 0/5

1/2 0/2 0/5

2/2 1/2 0/5

0/2 1/2 0/5

0/2 0/2 0/5

Lime Inoculum 6.6 5.6 4.6 3.6 6.6 5.6 4.6 3.6 AI 3 h 6 h

2/2 0/2 0/5

2/2 0/2 0/5

2/2 0/2 0/5

1/2 0/2 0/5

2/2 0/2 0/5

2/2 0/2 0/5

2/2 0/2 0/5

1/2 0/2 0/5

Cranberry Inoculum 6.5 5.5 4.5 3.5 6.5 5.5 4.5 3.5 AI 3 h 6 h

2/2 0/2 0/5

2/2 0/2 0/5

2/2 0/2 1/5

2/2 0/2 0/5

2/2 0/2 0/5

2/2 1/2 0/5

2/2 0/2 0/5

2/2 0/2 0/5

a Inocula (log CFU/sample) calculated from enumeration of inoculum strains and decimal dilution prior to enrichment. b Sampled within 15 min after inoculation. c Number of positive samples per number of samples tested.

pensed in 200-ml aliquots into sterile glass bottles and brought to the target temperatures prior to inoculation. All samples were in- oculated with 0.1 to 0.85 ml of a composite, to achieve a mini- mum of 106 CFU of each strain per sample, and then quickly returned to the target temperature. For all experiments, Dickson SX 100 temperature recorders (Dickson, Addison, Ill.) were in- terspersed among the samples to monitor concentrate exposure temperatures. The samples were inoculated quickly to avoid high- er temperatures in the concentrates. For all trials run at 2118C or higher, samples were stomached or agitated brie� y after inocula- tion to achieve an even distribution of the cells. At 2238C, con- centrates were solid, preventing mixing, so they were surface in- oculated. Negative controls were prepared as above using sterile citrate buffer in place of the inoculum.

Monitoring pathogen survival. For trials of concentrates in 10-g bags, two samples were randomly selected after inoculation (within 15 min) and at designated times up to 2 weeks. At each sampling time, 90 ml of universal preenrichment broth (UPB, Dif- co) was added to the sample. For trials of concentrates stored in 200-ml bottles, a 10-ml sample was drawn from the bottle and added to 90 ml UPB. For all trials investigating L. monocytogenes inactivation in concentrates at 211 and 08C, decimal dilutions of the samples were prepared in UPB at each sampling time; this

approach was designed to estimate log reduction achieved in the event that survivors were detected after enrichment.

In all trials, the primary dilution of concentrate into UPB was neutralized to pH 7 by aseptic addition of a predetermined volume of 3 N NaOH prior to incubation or further dilution. This neutralization step was added to avoid any artifact of technique that would result in negative samples because of a low pH during enrichment. Samples were then incubated at 358C for 72 h. After enrichment, samples were streaked onto selective plate media. UPB is recognized for its ability to recover sublethally injured bacteria (2, 12), so the cells recovered by enrichment would be expected to grow on selective media.

Sorbitol MacConkey (Difco) and EMB (Difco) agar plates were used for detection of E. coli O157:H7. Xylose-lysine-des- oxycholate agar (Difco) plates were used for detection of Sal- monella, and Palcam (Difco) plates were used for L. monocyto- genes. All plates were incubated at 358C for 24 h. Samples were scored as positive if typical colonies of the target organism were recovered on the selective agar plate. Atypical colonies were iden- ti� ed by Vitek 32 (BioMèrieux, Hazelwood, Mich.).

Detection limit. The ability to detect low numbers of E. coli O157:H7, L. monocytogenes, and Salmonella was veri� ed in each experiment by adding UPB to three uninoculated samples, neu-

J. Food Prot., Vol. 66, No. 91640 NOGUEIRA ET AL.

TABLE 5. Inactivation of Listeria monocytogenes in cranberry juice concentrate at 08C

Inocu- luma

Trial 1

7.0 6.0 5.0

Trial 2

7.0 6.0 5.0

AIb

6 h 24 h

4/4c

4/10 0/4

4/4 1/10 0/4

4/4 0/10 0/4

4/4 0/10 0/4

4/4 0/10 0/4

4/4 0/10 0/4

a Inocula (log CFU/sample) calculated from enumeration of in- oculum strains and decimal dilution prior to enrichment.

b Sampled within 15 min after inoculation (AI). c Number of positive samples per number of samples tested.

tralizing with NaOH as described above, and inoculating with the appropriate dilutions of the composites to achieve 2, 1, and 0 log CFU/sample. Samples were incubated and then plated as de- scribed above. The sample inoculated with the lowest level of cells that demonstrated growth determined the limit of detection for the method.

RESULTS AND DISCUSSION

Juice concentrates from different sources were used in our studies. The ranges of 8Brix, pH, and titratable acidity are shown in Table 1.

For each trial, the cells were adapted to pH 4 before inoculation to trigger the acid tolerance response known to enhance cell survival in acidic foods (9, 11, 13, 14, 17). In this way, we tried to reproduce a worst-case scenario, in which the product is contaminated with acid-tolerant cells.

Temperature has been shown to in� uence bacterial pathogen survival in juice, with cells surviving longer at lower temperatures (19). Table 2 shows the inactivation of E. coli O157:H7, Salmonella, and L. monocytogenes at 2238C, which could be considered a worst-case storage or transportation temperature. Salmonella was not detectable in any of the three concentrates held at 2238C within 15 min after inoculation.

As high as 8 log CFU E. coli O157:H7 and L. mono- cytogenes were generally unrecoverable within 15 min and 1 h after inoculation, respectively. However, positives were detected sporadically up to 2 weeks after inoculation. Pos- itive controls performed with these trials con� rmed a limit of detection of #40 CFU/sample.

In the trials at 2238C, the concentrates were solid at the time of inoculation and throughout storage. Therefore, in the positive samples, it is possible that the inoculum froze at the surface of the concentrate in locations of pure ice and did not come in contact with the concentrate and its antimicrobial properties during storage. However, the majority of samples exhibited .7-log CFU inactivation of pathogen, particularly for Salmonella, indicating that the antimicrobial properties of the concentrates can be effective at temperatures as low as 2238C.

The concentrates were all semiliquid at 2118C, so this temperature was selected for the next round of trials. Table 3 shows the inactivation of E. coli O157:H7 and Salmonella in cranberry, lemon, and lime concentrates at 2118C. Sal- monella inoculated at levels as high as 6.9 log CFU was undetectable in concentrates 15 min after inoculation. E.

coli O157:H7 was undetectable in lemon concentrate by 1 h after inoculation and by 6 h after inoculation in cranberry and lime concentrates. Positive controls performed concur- rently with these trials indicated that the limit of detection was #5 CFU/sample.

During early experiments at 2118C using similarly high inoculum levels, L. monocytogenes survivors were de- tected sporadically in cranberry, lemon, and lime concen- trate samples more than 24 h after inoculation (data not shown). To quantify the level of inactivation of L. mono- cytogenes in these concentrates, samples were decimally di- luted prior to enrichment. Results (Table 4) show a 5-log CFU reduction in lemon and lime concentrates by 6 h after inoculation in all cases. However, L. monocytogenes was recovered from one cranberry concentrate sample, even though L. monocytogenes were not recoverable from other samples with higher levels of inocula. The positive sample detected in cranberry concentrate after 6 h in trial 1 (Table 4) was the only positive among the 20 samples tested, in- cluding samples with 10 and 100 times higher levels of inocula. Positive controls run concurrently with these trials exhibited a limit of detection of 2 CFU/sample. This spo- radic positive was not reproducible, and the reason for it is unknown. At 2118C, the cranberry concentrate was semi- liquid; therefore, it is possible that some ice crystals re- mained, encapsulating and protecting the inoculum. Con- sequently, a trial was performed at a higher temperature, still consistent with commercial transportation conditions.

Table 5 shows the inactivation of L. monocytogenes in cranberry juice concentrate at 08C. Samples taken 6 h after inoculation showed that a 5-log CFU reduction was already achieved, despite some positive samples at inoculum levels higher than 5 log CFU/sample. In all samples taken 24 h after inoculation, Listeria inocula as high as 7.0 log CFU were undetectable. The limit of detection in these trials was approximately 1 CFU/sample. Whether the increased stor- age temperature was entirely responsible for the complete inactivation of the inoculum is unknown and worthy of fur- ther research. However, under the conditions of this study, L. monocytogenes were consistently and rapidly inactivated in cranberry concentrate at 08C.

CONCLUSIONS

This study demonstrates that cranberry, lemon, and lime juice concentrates possess intrinsic antimicrobial prop- erties, which will eliminate these bacterial pathogens in the event of postconcentration recontamination. The sporadic positives obtained in this study at 223 and 2118C might warrant further research, but at least a 5-log reduction of Salmonella was consistently demonstrated at 223 and 2118C, at least a 5-log reduction of E. coli O157:H7 was consistently demonstrated at 2118C, and at least a 5-log reduction of L. monocytogenes was consistently demon- strated in cranberry concentrate at 08C. Consequently, cran- berry, lemon, and lime juice concentrates, with properties consistent with the concentrates used in this study and held at the above temperatures or greater for more than a few hours, should not need to be repasteurized to control the potential presence of these pathogens.

J. Food Prot., Vol. 66, No. 9 INACTIVATION OF PATHOGENS IN JUICE CONCENTRATES 1641

ACKNOWLEDGMENTS

The authors thank Dr. Yuhuan Chen, Dr. Jaheon Koo, and Amy Hu for valuable help with the experiments and Sandra Arze and Cynthia Tho- masson for their technical assistance.

REFERENCES

1. Anderson, S., and K. Bailey. 2001. Recent FDA juice HACCP reg- ulations. Food Saf. Mag. N15:18–25.

2. Bailey, J. S., and N. A. Cox. 1992. Universal preenrichment broth for the simultaneous detection of Salmonella and Listeria in foods. J. Food Prot. 55:256–259.

3. Besser, R. E., S. M. Lett, J. T. Weber, M. P. Doyle, T. J. Barret, J. G. Wells, and P. M. Grif� n. 1993. An outbreak of diarrhea and he- molytic uremic syndrome from Escherichia coli O157:H7 in fresh pressed apple cider. JAMA 269:2217–2220.

4. Bodel, P. T, R. Cotran, and E. H. Kass. 1959. Cranberry juice and antimicrobial action of hippuric acid. J. Lab. Clin. Med. 51:881– 888.

5. Centers for Disease Control and Prevention. 1975. Salmonella Ty- phimurium outbreak traced to a commercial apple cider—New Jer- sey. Morb. Mortal. Wkly. Rep. 24:87–92.

6. Centers for Disease Control and Prevention. 1996. Outbreak of Esch- erichia coli O157:H7 infections associated with drinking unpasteur- ized commercial apple juice—British Columbia, California, Colo- rado, and Washington. Morb. Mortal. Wkly. Rep. 45:975.

7. Food and Drug Administration. 2001. Hazard analysis and critical control point (HAACP); procedures for the safe and sanitary pro- cessing and importing of juice. Fed. Regist. 63:20450–20486.

8. Food and Drug Administration. 2001. Hazard analysis and critical control point (HAACP); procedures for the safe and sanitary pro- cessing and importing of juice, � nal rule. Fed. Regist. 66:6137–6202.

9. Foster, J. W., and H. K. Hall. 1990. Adaptive acidi� cation tolerance response of Salmonella Typhimurium. J. Bacteriol. 172:771–778.

10. Gaham, C. G. M., B. O’Driscoll, and C. Hill. 1996. Acid adaptation of Listeria monocytogenes can enhance survival in acidic foods and during milk fermentation. Appl. Environ. Microbiol. 62:3128–3132.

11. Garren, D. M., M. A. Harrison, and S. M. Russell. 1997. Retention

of acid tolerance and acid shock responses of Escherichia coli O157: H7 and non-O157:H7 isolates. J. Food Prot. 60:1478–1482.

12. Hammack, T. S., R. M. Amaguana, and W. H. Andrews. 2001. An improved method for the recovery of Salmonella serovars from or- ange juice using universal preenrichment broth. J. Food Prot. 64: 659–663.

13. Hill, C., B. O’Driscoll, and I. Booth. 1995. Acid adaptation and food poisoning microorganisms. Int. J. Food Microbiol. 28:245–254.

14. Kroll, R. G., and R. A. Patchett. 1992. Induced acid tolerance in Listeria monocytogenes. Lett. Appl. Microbiol. 14:224–227.

15. Larkin, J. W. FDA memorandum to OPDFB, microbiological critical control point for certain shelf-stable and concentrated juice products, September 29, 2000. Food and Drug Administration, Washington, D.C.

16. Lee, Y., J. Owens, L. Thrupp, and T. C. Cesario. 2000. Does cran- berry juice have antimicrobial activity? JAMA 283:1691.

17. Leyer, G. J., L. L. Wang, and E. A. Johnson. 1995. Acid adaptation of Escherichia coli O157:H7 increases survival in acidic foods. Appl. Environ. Microbiol. 61:3752–3755.

18. Mazzotta, A. S. 2000. Thermal inactivation of stationary-phase and acid-adapted Escherichia coli O157:H7, Salmonella, and Listeria monocytogenes in fruit juices. J. Food Prot. 64:315–320.

19. Miller, L. G., and C. W. Kaspar. 1994. Escherichia coli O157:H7 acid tolerance and survival in apple cider. J. Food Prot. 57:460– 464.

20. Oyarzábal, O. A., M. C. L. Nogueira, and D. E. Gombas. 2003. Survival of Escherichia coli O157:H7, Listeria monocytogenes, and Salmonella in juice concentrates. J. Food Prot. 66:1595–1598.

21. Sado, P. N., K. C. Jinneman, G. J. Husby, S. M. Sorg, and C. J. Omiecinski. 1998. Identi� cation of Listeria monocytogenes from un- pasteurized apple juice using rapid test kits. J. Food Prot. 61:1199– 1202.

22. Sheryl, A., and L. J. Harris. 2001. The effects of freezing and thaw- ing on the survival of Escherichia coli O157:H7 in apple juice. Int. J. Food Microbiol. 67:89–96.

23. Ulate-Rodriguez, J., H. W. Schafer, E. A. Zotolla, and P. M. David- son. 1997. Inhibition of Listeria monocytogenes, Escherichia coli O157:H7, and Micrococcus luteus by linear furanocoumarins in cul- ture media. J. Food Prot. 60:1046–1049.