Preliminary Studies on the Potential for Infiltration, Growth and Survival of Salmonella enterica serovar Hartford and Escherichia coli O157:H7 Within Oranges


Infiltration, survival, and growth potentials of human pathogens within oranges were investigated in a series of preliminary laboratory experiments. Infiltration potential was assessed by applying 7-log cfu of a suspension of Escherichia coli O157:H7 onto the stem scar, subjecting the oranges to a temperature decrease, juicing, and measuring numbers of E. coli O157:H7 in the resulting juice. Survival and growth studies were performed by injecting or applying pathogens to a simulated puncture wound located at various location on oranges, then incubating fruit for 5-days at either 4° or 21°C, juicing, and plating a juice sample. Bacteria were recovered on non-selective and selective media. Bacterial survival was also evaluated in orange juice at 4° and 21°C. Oranges internalized pathogens at a frequency of 3.6%, at a ratio of 0.001 to 0.0001 of the challenge levels. Growth in oranges occurred at 21°C, but not at 4°C. In fresh orange juice, at 21°C there was a 1 log decline in population in 3 d for E. coli O157:H7 and Salmonella enterica serovar Hartford. Uninoculated background flora grew 1 to 2-logs in the juice, within 3 days. These results provide evidence that internalization, survival, and growth of human bacterial pathogens may occur within oranges. There is a need to establish the likelihood and levels of human pathogens on or within oranges to provide a rationale basis for designing intervention technologies needed to assure the microbiological safety of juices.


Fresh fruit and vegetable juices are recognized as an emerging cause of foodborne illness (Parish 1997). A contributing factor is that these products are raw agricultural commodities, which may become contaminated by animal or human waste and consumed without a processing step that will kill or remove associated pathogens. While a single piece of contaminated produce may infect a single person, contaminated produce that is co-mingled, juiced, and served may infect many individuals.

Contamination, infiltration, and survival of microorganisms in fruits and vegetables have been documented, although most reported data concern plant pathogens and spoilage organisms. One potential source of entry of microorganisms into fruits and vegetables is by environmental exposure with uptake occurring through either specific morphological structures in the plant and/or through breaks in tissues that occur as a result of punctures, wounds, cuts and splits. These insults to the fruit can occur during growing or harvesting. Hill and Faville (1951) postulated that citrus on the tree could become infected by insect punctures, thorn injury, or hail damage. Subsequently, Hill and Wenzel (1963) found that 35 percent of dropped and frost-damaged oranges had microholes too small for visual detection, and half of the fruit having microholes were contaminated with microbes. Almed et al. (1973) found that citrus fruits could be damaged in numerous ways during harvesting and that the damage provides opportunities for decay organisms to gain entrance to the fruit. Other modes of entry are created during harvesting, for example tearing a portion of the peel around the stem end of the fruit during manual harvest. Mechanically harvested fruit is subject to splits, punctures and bruises. Some of this damage occurs because of the number of fruit with attached stems that are firm enough to puncture other fruit.

In addition, processing conditions may promote opportunities for microorganisms to infiltrate fruit. For example, citrus fruits may suffer peel damage and microbial penetration as a result of creasing during washing or packing, which may result in splits in the peel. When warm fruits and vegetables are put under cold ambient temperatures, the gases within fruits and vegetable contract during cooling, and there is an inward hydrostatic potential that draws in fluid from the exterior of the fruit. This inward movement of fluid has the potential to draw in microorganisms (Buchanan et al. 1999)

The Food and Drug Administration (FDA) has proposed a rule requiring implementation of a Hazard Analysis Critical Control Point (HACCP) program for processors of unpasteurized fruit and vegetable juices. A key aspect of the proposal includes a requirement for a verifiable one hundred thousand-fold (5-log) pathogen reduction program. This level of hazard reduction was considered adequate by the National Advisory Committee on Microbiological Criteria in Foods (9 April 1997) to reduce microbial hazards that can cause disease outbreaks in these commodities. Little is known about the natural occurrence and anatomical distribution of human pathogens on produce. This knowledge gap is critical because effective intervention strategies need to take into account distribution patterns of the microbial hazard. To be effective, an intervention must come in intimate contact with a pathogenic microorganism. Previous research (Merker et al., 1999), using uptake of dye, was conducted by FDA and suggested the potential for uptake and internalization of bacterial pathogens. To extend that research, the objectives of the current study were to determine the potential for human bacterial pathogens to become internalized within oranges and to determine the potential for survival and growth once present in the interior of the fruit. Parallel experiments were done to compare pathogen survival in juice. Findings were that such microorganisms can infiltrate oranges that are visually intact, and once the organism infiltrates, pathogens can survive for at least five days, and in some cases grow.

Methods and Materials

Microorganisms. Escherichia coli O157:H7 C984 (ATCC reference number 43890 isolated from the microbial flora of sheep) (Kudva et al. 1996) was obtained from Dr. Peter Feng of the Food and Drug Administration's Center for Food Safety and Applied Nutrition. This strain was transformed with plasmid pGFPuv (from Clontech) and stably expressed green fluorescent protein. Salmonella enterica serovar Hartford HO610 (isolated from a patient during an outbreak of salmonellosis associated with unpasteurized orange juice) (Cook et al. 1998) was obtained from the Centers for Disease Control and Prevention. Permanent cultures were maintained at -70° C.

Inoculum. Working cultures were maintained on brain heart infusion (BHI) slants and a loopful was transferred to a 2-ml BHI broth, grown overnight at 37ºC, and then refrigerated. This broth was used to inoculate overnight cultures and was not kept for longer than 1-week before a new culture was generated from the slant culture. Overnight cultures were started by inoculating into tryptic soy broth containing 1% dextrose (TSB + 1% G) medium using a 0.1 ml of the culture broth and were incubated at 37ºC without shaking for at least 18-hours.

Oranges. California Valencia oranges (88 count, 7.2-cm average diameter), purchased from a local produce, market were used for all experiments. Oranges were examined and those with visual defects were not used.

Media. Xylose Lysine Deoxycholate agar was from BBL. All other media were from Difco. Brilliant Blue FCF dye was from Warner-Jenkinson.

Infiltration studies. Oranges were selected and placed in a 37ºC incubator overnight to equilibrate. Concurrently, a 10-ml culture of E. coli O157:H7 was inoculated and allowed to grow overnight at 37ºC. The next morning 8-mg of Brilliant Blue FCF was added to the bacterial culture, and 0.1 ml of the dye-culture solution containing approximately 7-log cfu E. coli O157:H7 was inoculated onto the stem scar of the warm oranges, after removal of residual stem material. Oranges were then transferred to a 4ºC incubator and allowed to equilibrate for 3-hours. During this time the solution appeared to be internalized by the oranges, and the internal temperature was observed to decrease from 35ºC to 11ºC. Oranges were then juiced using the following protocol: Inoculated stem scar areas were submerged in an 80ºC water bath for one minute to sanitize the contaminated surfaces (Pao et al, 1999). Afterwards oranges were halved with a sterile knife through the midline of the orange lateral to the stem scar and juiced in a Ra-Chand Model J-210 or Model J-315 juicer. Both models are hand operated cone compression juicers that yield similar volumes of juice from orange halves. The juicers were cleaned and rinsed after each orange was juiced. A utility wipe (Kimberley-Clark, Roswell, GA) folded in quarters was placed between each orange peel and the juicer to further minimize juice contamination from the peel. Juice was collected in sterile 50-ml centrifuge tubes and the volume was recorded. Duplicate samples were plated on BHI agar and Sorbitol MacConkey plates using an Autoplate 4000 spiral plater (Spiral Biotech, Bethesda MD). Plates were incubated either at 37ºC overnight or at room temperature (21ºC) for 48-hours and counted on a laser colony scanner Model 500 A (Spiral Biotech). E. coli O157:H7 detection was by visualizing colonies under long wave ultraviolet light. The Green Fluorescent Protein causes the colonies to fluoresce. Green-glowing colonies were considered E. coli O157:H7 (fig 1). A total of 55 oranges were assayed.

Survival Studies in Oranges. Oranges used were either at room temperature (21ºC) or at 4ºC after equilibrating overnight under refrigeration. The overnight cultures of test organisms were diluted to approximately 4-log cfu/ml in 0.1% peptone water. Oranges were inoculated with a total of 0.05-ml (approximately 500 cells/orange) at one of five locations (fig 2). A single 0.05-ml volume was injected into the stem area using a tuberculin syringe with a 95-mm needle to introduce the test organism into the core region. Also, five different 0.01-ml injections (approximately 100 cells) were made in the albedo (white part of the orange) and the orange section portion. The inoculation protocols also included simulated wounding by using the tip of a sterile 1-ml plastic pipette tip (Rainin Instrument company, Woburn MA) to bore into the orange at five locations per orange (fig 3). Two depths were examined: 1) shallow (4 to 5-mm) into the albedo level and 2) deep (10 to 11-mm) into the section portion. Each wound was then inoculated with 0.01-ml of the diluted overnight culture. Inoculated oranges were then incubated either at room temperature or at 4ºC for 5-days. A zero day observation was taken for inoculated oranges both at 4º and 21ºC. The oranges were then quartered with a sterile knife and juiced in a Colworth Stomacher 400 for 1 minute. An aliquot of the expressed juice was plated onto either BHI media or selective agar (Sorbitol MacConkey for E. coli O157:H7 or xylose lysine desoxycholate (XLD) for S. Hartford) using a Spiral Biotech Autoplate 4000. The plates were either incubated at 37ºC overnight or at room temperature for two days. The plates were counted on a laser colony scanner (Spiral Biotech).

Survival Studies in Juice. Juice was obtained from uninoculated oranges and was subsequently inoculated with an overnight culture of either E. coli O157:H7 or S. Hartford at approximately 4-log cfu. Aliquots were removed daily for 5-days and plated onto BHI plates and onto selective media (Sorbitol MacConkey for E. coli O157:H7 and XLD for S. Hartford) using a spiral plater. E. coli O157:H7 counts were confirmed by checking for fluorescence with a long wave ultraviolet lamp. After overnight incubation at 37ºC, the plates were counted and colony forming units per ml were calculated.

Statistical Analysis. Means and standard errors of the means were calculated using commercial spreadsheet software (Lotus 123, release 5.0 Lotus Development Corp.). In the survival studies for E. coli O157:H7 and S. Hartford ten replicate oranges were evaluated. All results reported as log units are log (base 10) units.


Infiltration Studies of E. coli O157:H7

Two of the 55 oranges (3.6%) that were challenged with external contamination internalized E. coli O157:H7, as measured in the expressed juice. Observed counts were 33,300 and 2,360 cfu/orange. This represents an uptake rate of 0.001 to 0.0001 of the number of cfu applied. No E. coli O157:H7 were observed in uninoculated control samples or in juice from the inoculated orange half opposite the stem scar orange half.

Survival Studies in the Orange

Table I indicates that at after a 5-day incubation at 4°C E. coli O157:H7 populations declined at all anatomical sites (maximum 0.6-log cfu/ml) when injected or applied to simulated wounds. After incubation at 21°C, however, there was a 2-log increase of bacteria in section portions that were injected or inoculated within simulated wounds. There was nearly a 1-log increase in population at 21°C when bacteria were applied to the albedo through a simulated wound or when injected into the core. No growth was observed in oranges where cultures were injected into the albedo, nor was there evidence of E. coli O157:H7 in any of the control samples. Table II demonstrates the growth potential of S. Hartford in oranges. At 4°C there was a 0.3-log cfu/ml increase in microbial population in the albedo portion for the simulated wound samples. In contrast, bacterial populations declined 0.7-log cfu/ml in section portions, in the simulated wound treatment. Bacterial populations remained constant for all injected samples that were incubated at 4°C. However, at 21°C all experimental treatment samples exhibited increased bacterial populations (range 0.2 to 3.9-log cfu). Largest growth increases were observed in the section samples, confirming the results seen with E. coli O157:H7. No Salmonella spp. were observed in control samples.

Survival Studies in Orange Juice

Table III indicates that at 4°C there was no change, and at 21°C there was a 1-log decrease in E. coli O157:H7 and S. Hartford cfu levels in orange juice that was stored for three days. The average juice pH was 3.65. Uninoculated background microflora increased about 1-log cfu in orange juice stored at 4°C and 2-log for orange juice maintained at 21°C for 3-days.


Key findings of the current study were the demonstration that infiltration of human pathogens into oranges can occur and that survival and growth can occur under certain conditions. The demonstration in this study that a human pathogen can be internalized by oranges at an uptake frequency of 3% and at a level of 0.001 to 0.0001 of the challenge levels is in agreement with the earlier dye-uptake studies (Merker et al., 1999). In that study oranges slightly or moderately internalized dye at a 3% uptake frequency. This close association suggests that dye uptake may serve as a useful surrogate for microbial pathogens, and may be a useful tool when pathogen challenge studies would be unsuitable. It is noteworthy that the observed infiltration frequency and level in the present study may be conservative since experiments were performed with intact fruit, without any obvious defects, and included an external heat decontamination step. Oranges having visually or non-visually obvious peel defects may allow for higher infiltration frequencies and levels. These findings are consistent with pathogen uptake studies using other fruit and vegetable commodities, including apples and tomatoes (Buchanan et al, 1999; Bartz and Showalter, 1981).

Cold storage for 5-days reduced the survivability of E. coli O157:H7 compared to time zero observations, but the was generally not the case for S. Hartford. Thus, while refrigeration was shown to prevent growth, these findings indicate that refrigeration cannot be used to ensure reduction in microbial pathogen populations. The current results support research on other commodities that demonstrate the survival of E. coli O157:H7 and Salmonella spp. in low pH environments (Buchanan et al. 1999, Ryu et al. 1998, Uljas et al. 1998; Zhuang et al. 1995).

Growth of S. Hartford and E. coli O157:H7 within oranges held at room temperature may result from compartmentalization and micro-environments within intact oranges. Segregated vesicles confine the acidic juice within oranges and provided these vesicles are not disturbed may create regions outside the vesicles where bacterial growth could occur. In contrast, if these compartments are destroyed, such as during juicing operations, the acidic pH would prevail throughout the juice and provide a uniformly low pH environment for a microorganism. This hypothesis was tested and is supported by this study. These results suggest that pH micro-environments do exist in the intact fruit, and permit pathogen growth. This research also confirms previous reports demonstrating growth of human bacterial pathogens in internal fruit tissues from tomatoes, apples, and melons (Asplund, K. and Nurmi, 1991; Janisiewicz et al., 1999; Golden et al., 1993; Zhuang et al. 1995)

While the current study has provided evidence for the potential for human bacterial pathogens to enter, survive, and grow within intact oranges, their natural occurrence on or within oranges is unknown. Further research is required to better characterize the factors that lead to contamination and infiltration of human pathogens within oranges. Most critically however, there is a need for data to establish the likelihood of these events occurring during cultivation, harvesting, transport, storage, or processing. In addition, data are needed to establish the natural levels of human pathogens on or within oranges. This information would provide a rationale basis for designing intervention technologies that are needed to assure the microbiological safety of juices.


Almed, E.M., Martin, F.G., and Fluck, R.C. 1973. Damaging stresses to fresh and irradiated citrus fruit. J. Food Sci. 38:230-233.

Asplund, K. and Nurmi, E. 1991. The growth of salmonellae in tomatoes. Int. J. Food Microbiol. 13:177-182.

Bartz, J.A. and Showalter, R.K. 1981. Infiltration of tomotoes by aqueous bacterial suspensions. Phytopathology 71: 515-518.

Buchanan, R.L., Edelson, S.G. 1999. pH-dependent stationary-phase acid resistance response of enterohemorrhagic Escherichia coli in the presence of various acidulants. J. Food Prot. 62:211-218.

Buchanan R.L., Edelson, S.G., Miller, R.L., and Sapers, G.M. 1999. Contamination of intact apples after immersion in an aqueous environment containing Escherichia coli O157:H7. J. Food Prot. 62:444-450.

Cook, K.A., Dobbs, T.E., Hlady, W.G., Wells, J.G., Barrett, T.J., Puhr, N.D., Lancette, G.A., Bodager, D.W., Toth, B.L., Genese, C.A., Highsmith, A.K., Pilot, K.E., Finelli, L., and Swerdlow, D.L. 1998. Outbreak of Salmonella serotype Hartford infections associated with unpasteurized orange juice. J. Am. Med. Assn. 280:1504-1509.

Merker, R.I., Edelson-Mammel, Davis, V., and Buchanan, R.L. 1999. Preliminary experiments on the effect of temperature differences on dye uptake by oranges and grapefruit. U.S. Food and Drug Administration, Center for Food Safety and Applied Nutrition, 200 C ST SW, Washington, DC 20204.

Golden, D.A., Rhodehamel, E.J., and Kautter, D.A. 1993. Growth of Salmonella spp. in cantaloupe, watermelon, and honeydew melons. J. Food Prot. 194:194-196.

Hill, E.C. and Faville, L.W. 1951. Studies on the artificial infection of oranges with acid tolerant bacteria. Proc. Fla. State Hort. Soc. 64:174-177.

Hill, E.C. and Wenzel, F.W. 1963. Microbial populations of frozen oranges. Proc. Fla. State Hort. Soc. 76:276-281.

Janisiewicz, W.J., Conway, W.S., Brown, M.W., Sapers, G.M., Fratamico, P., and Buchanan, R.L. 1999. Fate of Escherichia coli O157:H7 on fresh-cut apple tissue and its potential for transmission by fruit flies. Appl. Environ. Microbiol. 65:1-5.

Kudva I.T., Hatfield P.G., and Hovde C.J. 1996. Escherichia coli O157:H7 in microbial flora of sheep.

J. Clin. Microbiol. 34: 431-433.

Pao, S., and Davis, C.L. 1999. Enhancing microbiological safety of fresh orange juice by fruit immersion in hot water and chemical sanitizers. J. Food Prot. 62:756-760.

Parish M.E. 1997. Public health and nonpasteurized fruit juices. Crit. Rev. Microbiol. 23:109-119.

Ryu J.H., and Beuchat L.R. 1998. Influence of acid tolerance responses on survival, growth, and thermal cross-protection of Escherichia coli O157:H7 in acidified media and fruit juices. Int. J. Food Microbiol. 45:185-193.

Uljas, H.E., and Ingham, S.C. 1998. Survival of Escherichia coli O157:H7 in synthetic gastric fluid after cold and acid habituation in apple juice or trypticase soy broth acidified with hydrochloric acid or organic acids. J. Food Prot. 61:939-947.

Zhuang, R.Y., Beuchat, L.R., and Angulo, F.J. 1995. Fate of Salmonella montevideo on and in raw tomatoes as affected by temperature and treatment with chlorine. Appl. Environ. Microbiol. 61:2127-2131.

Table I. Inoculation and recovery of Escherichia coli O157:H7 C984 in oranges before and after incubation at 4° and 21°C for 5 days

Incubation Time/Temp°C

0 Day

5 Day

Inoculation method/site




0.84 (0.22) a

1.03 (0.27)

0.26 (0.16)

1.70 (0.64)






0.26 (0.16)

0.42 (0.20)

0.13 (0.12)

0.31 (0.19)






0.39 (0.19)

0.55 (0.21)

0.13 (0.12)

2.43 (0.67)





Simulated wound/albedo

0.42 (0.20)

0.13 (0.12)

None Detected

0.91 (0.52)





Simulated wound/section

0.44 (0.22)

0.57 (0.22)

None Detected

2.85 (0.65)






None Detected

None Detected

None Detected

None Detected





a Escherichia coli O157:H7, recovered by direct plating onto BHI agar; inoculum level was 500 cfu/orange. Values represent mean log cfu/ml (n=10) from expressed juice of inoculated oranges with standard error of the mean in parenthesis. Minimum detection level was 10 cfu/ml per orange. Thus, minimum mean detection level for the 10 oranges was log(cfu/ml)=0.1 when oranges without detectable E. coli O157:H7 were assigned a value of log(cfu/ml)=0.00.

b Number of oranges that tested positive per number of oranges tested

Table II. Inoculation and recovery of Salmonella enterica serovar Hartford HO610 in oranges before and after 5 days of incubation at 4° and 21°C
 Incubation Time/Temp°C
0 Day5 Day
Inoculation method/site21°21°
Inject/core0.13 (0.12)a0.52 (0.20)0.18 (0.17)1.50 (0.60)
Inject/albedoNone Detected0.14 (0.13)None Detected0.34 (0.20)
Inject/sectionNone Detected0.29 (0.17)None Detected3.65 (0.45)
Simulated wound/albedoNone Detected0.39 (0.19)0.32 (0.19)1.34 (0.51)
Simulated wound/section0.81 (0.24)0.61 (0.22)0.14 (0.13)4.51 (0.35)
ControlNone DetectedNone DetectedNone DetectedNone Detected
a Salmonella enterica serovar Hartford, recovered by direct plating onto XLD agar; inoculum level was 500 cfu/orange. Values represent mean log cfu/ml (n=10) from expressed juice of inoculated oranges with standard error of the mean in parenthesis. Minimum detection level was 10 cfu/ml per orange. Thus, minimum mean detection level for the ten oranges was log(cfu/ml)=0.1 when oranges without detectable S. Hartford were assigned a value of log(cfu/ml)=0.00. b Number of oranges that tested positive per number of oranges tested
Table III. Survival of Escherichia coli O157:H7 and Salmonella enterica serotype Hartford in orange juicea at 4° and 21°C
E. coliSalmonellaControlE. coliSalmonellaControl

aTarget inoculum level was 4-log/ml; mean juice pH was 3.65

b log cfu/ml juice


2 petri dishes containing green fluorescent colonies of Eschericha coli O157:H7 (left) and control (right).

Figure 1 Green fluorescent colonies of Eschericha coli O157:H7 (left) and control (right).

Injection of culture into orange core using a 1 ml syringe.

Figure 2 Injection of culture into orange core using a 1 ml syringe.

Using a 1 ml pipette tip to puncture orange surface.

Figure 3 Simulated wound technique using a 1 ml pipette tip to puncture orange surface. This was followed by inoculation with human bacterial pathogens

Mark O. Walderhaug, Sharon G. Edelson-Mammel, Antonio J. DeJesus, B. Shawn Eblen, Arthur J. Miller, and Robert L. Buchanan

Page Last Updated: 12/19/2017
Note: If you need help accessing information in different file formats, see Instructions for Downloading Viewers and Players.
Language Assistance Available: Español | 繁體中文 | Tiếng Việt | 한국어 | Tagalog | Русский | العربية | Kreyòl Ayisyen | Français | Polski | Português | Italiano | Deutsch | 日本語 | فارسی | English