In recent years, tree fruits have emerged as a new concern for Listeria monocytogenes contamination. The objective of the current study was to evaluate the potential internalization of L. monocytogenes from the surface of avocados into the edible portions of the fruit during certain postharvest practices simulated in a laboratory setting. One set of intact avocados was spot inoculated with L. monocytogenes on the stem scar, and the second set was hydrocooled in water contaminated with L. monocytogenes. Under these experimental conditions, L. monocytogenes internalized into the avocado pulp through the stem or stem scar after both spot inoculation and hydrocooling. In avocados spot inoculated with 50, 130, 500, and 1,300 CFU per fruit, bacteria were detected in the edible portion adjacent to the stem scar within 15 days postinoculation during storage at 4°C. In avocados hydrocooled in water containing L. monocytogenes at 106 and 108 CFU/ml, bacteria reached the bottom end of the fruit, and the populations in the edible portion adjacent to the stem scar reached up to 5.90 to 7.19 log CFU/g within 10 to 15 days during storage at 4°C. Dye mixed with inoculum was useful for guiding subsequent sampling, but dye penetration patterns were not always consistent with bacterial penetration.
Listeria monocytogenes is an intracellular foodborne pathogen that can survive and grow in a wide variety of reservoirs such as animals, the natural environment, food, and food processing environments (22). The disease it causes, listeriosis, has a high case fatality rate, especially among immunocompromised and elderly individuals and pregnant women (22). A variety of foods have been associated with listeriosis outbreaks: milk, cheese, fruits, vegetables, ready-to-eat deli meat, and hot dogs (5, 8, 9, 28). In recent years, significant outbreaks and sporadic cases of listeriosis associated with fresh produce commodities have occurred, representing an emerging public health concern. In 2011, cantaloupes contaminated with L. monocytogenes caused a multistate outbreak in the United States resulting in 33 deaths, making it the worst listeriosis outbreak in history (5, 28). In 2014, stone fruits, which do not usually support the growth of L. monocytogenes, were implicated in a multistate outbreak in the United States (20). In 2015, a multistate outbreak in the United States that was associated with contaminated caramel apples caused 35 illnesses and three deaths (6). Understanding the circumstances that lead to contamination of these foods with L. monocytogenes requires careful attention to how these foods are harvested, transported, and prepared.
Although cross-contamination from the fruit surface to the pulp during cutting by the consumer can pose a significant risk, internalization of bacterial pathogens during postharvest processing could also be an important contributing factor to foodborne diseases. Postharvest internalization of Salmonella in whole tomatoes was observed during the wet-dump stage, when a large amount of water was absorbed into the apoplastic free space of the fruit tissues. Apertures in tomato surfaces such as wounds, the stem scar, and lenticels located at the rim of the stem attachment area became infiltrated with water due to the temperature differential, dwell time, hydrostatic pressure, and the presence of residual surfactants (1, 2, 10, 26, 30). Internalization of Salmonella through the tomato stem scar can also occur when the fruit is placed on a contaminated moist surface, even without immersion in water. Water infiltration into hydrocooled strawberries enabled passive internalization of phytopathogens suspended in water (13). Peppers also take up water during hydrocooling (10). Penteado et al. (29) found that water intake during hydrocooling of mangos resulted in Salmonella internalization into the fruit pulp. Similarly, E. coli O157:H7 was internalized in apples submerged in water containing this pathogen (3). Eblen et al. (12) found that spot contamination of stem scars of oranges with a water suspension of E. coli O157:H7 and Salmonella led to internalization of these bacteria into the fruit.
Avocado is one of the tree fruit commodities that have emerged as a new concern for L. monocytogenes contamination. L. monocytogenes has been isolated from avocado pulp (14) and guacamole products (15), and the contamination probably occurred in the production environment. However, the incidence of L. monocytogenes on fresh whole avocados has not been reported, possibly because previous surveillance work did not involve fresh whole avocados. Although postprocessing contamination can be a major contributor to the occurrence of L. monocytogenes in avocado pulp and guacamole products, research is needed to study the risk of L. monocytogenes contamination in fresh whole avocados.
Certain postharvest processing practices can affect the microbial contamination of fresh whole fruits and vegetables. After avocados are received at the packing facility, orchard litter can be removed using water alone or in combination with brushing. In some instances, high-pressure water blasting is necessary to remove insect pests or pollen deposits from the fruit rind (21, 25). Washed fruits, occasionally still wet, are then packed in boxes with or without sheet inserts or liners and immediately transferred to cooling rooms. To reduce fruit water loss that usually causes patches on the skin, moisture is added to the cooling room by spraying water mist into the cooling airstream (18). These procedures will increase the amount of free water on fruit surfaces, which can facilitate the survival and spread of L. monocytogenes.
To reduce avocado susceptibility to mechanical damage, physiological disorders, and phytopathogens, avocado fruits are commonly harvested in midafternoon when fruit turgidity is not high (18). On the trees, avocado flesh temperatures well above 40°C have been recorded in direct sunlight even when ambient temperatures are 15 to 25°C (32). Heat from fruit respiration also contributes to a considerable temperature increase in harvested avocados. Hence, rapid fruit precooling after harvest to near final storage temperatures is critical and particularly important for maintaining freshness during long-term storage (23, 24, 33). The rate of cooling has a significant effect on the long-term storage potential, and vacuum cooling and hydrocooling are among the fastest cooling techniques. Avocados can be precooled to storage or transport temperatures in about 8 h by forced-air cooling (31) or in about 30 min by hydro-cooling (16). Hydrocooling was introduced in avocado packing houses in the 1970s (4) and produces rapid, inexpensive, and uniform cooling. Unlike other precooling techniques such as a cooling room, forced air, and vacuum cooling, hydrocooling causes no moisture loss, thereby improving the organoleptic properties of the fruit. Because of these advantages, hydrocooling has been adopted by some domestic and international avocado growers and packing houses, especially for the Hass variety (23, 33), which accounts for about 90% of the world avocado market (11). The ability of L. monocytogenes to internalize into avocado pulp during various postharvest practices, especially those that provide opportunities for pathogens in water to come in contact with the fruit, has not been explored. Avocado pulp supports the growth of L. monocytogenes (19); therefore, any L. monocytogenes that internalizes into avocados could pose a significant health risk. The objective of the present study was to evaluate the impact of certain postharvest practices on the internalization of L. monocytogenes into the edible portions of fresh whole avocados.
MATERIALS AND METHODS
Avocado fruits. Bagged Hass avocados (Mission Avocados, Oxnard, CA) were obtained from a local supermarket, and physiologically mature but unripe fruits with no mechanical damage were selected for the study. Each avocado weighed 150 to 180 g.
Bacterial culture. L. monocytogenes serotype 1/2b strain AW001, previously isolated from fresh whole avocados in the United States in July 2014, was used for inoculation. The culture was grown at 37°C in brain heart infusion broth for 18 h and subsequently diluted in water to the desired levels.
Spot inoculation of L. monocytogenes. The stem scar areas were inoculated with 100 μl of five different levels of L. monocytogenes (four low levels of 50, 130, 500, and 1,300 CFU per fruit and one high level of 106 CFU per fruit), and the inoculated avocados were subsequently stored at 4°C for up to 15 days to simulate storage conditions following postharvest processing and packing (23). For two of the inoculation levels, 50 and 500 CFU per fruit, avocados were inoculated with both L. monocytogenes and 1% acid blue 9 (Chem-Impex International, Wood Dale, IL). The dye does not interfere with growth and colony formation of L. monocytogenes (data not shown).
Hydrocooling in the L. monocytogenes–dye solution. To simulate field temperatures of avocados exposed to direct sunlight, fruits were heated to 40°C by incubation in an Innova 42R programmable incubator (New Brunswick Scientific, Enfield, CT) with a relative humidity at 92%. Fruit temperature was measured as described elsewhere (32). Prewarmed avocados were then submerged for 20 min in 6°C water containing L. monocytogenes and 1% acid blue 9. L. monocytogenes was added to the hydrocooling solution to achieve final levels of 106 and 108 CFU/ml. The high level of inoculum was used to allow enumeration of L. monocytogenes populations internalized into avocado flesh. During the hydrocooling process, water temperature was monitored and maintained at 6°C with ANOVA 40 refrigerated circulators (ANOVA Industries, Houston, TX). The avocados were subsequently removed from the water, dried in a biosafety hood for 1 h, and stored at 4°C for up to 15 days.
Fruit sampling and detection and enumeration of L. monocytogenes in avocado mesocarp. Fruits were analyzed after spot inoculation and hydrocooling. Before analysis, avocados were submerged in 1.5 liters of 0.525% sodium hypochlorite solution for 5 min to sterilize the fruit surface, washed in sterile deionized water twice, and then dried. The rinsates from the second wash were centrifuged at 3,500 × g, and the resulting pellets were resuspended in Butterfield's phosphate buffer (BPB) and plated onto both agar Listeria Ottavani & Agosti (ALOA; AEB520080, bioMérieux, St Louis, MO) and RAPID'L.mono agar (3563694, Bio-Rad Laboratories, Hercules, CA) to verify that the fruit surfaces were free of L. monocytogenes after the hypochlorite treatment. To further verify the efficiency of the surface disinfection method, the total aerobic plate count was determined for pellets resuspended in BPB as described in the Bacteriological Analytical Manual (BAM) (27). For qualitative analysis, pulp was enriched in buffered Listeria enrichment broth (BLEB) as described in the BAM (17). For quantitative analysis, pulp was homogenized and serially diluted in BLEB, and 150 μl of the pulp-broth mixture was spread plated onto four ALOA plates. This direct plating scheme had a lower limit of detection (LOD) of 16.6 CFU/g. The remainder of the pulp-broth mixture was incubated as described in the BAM (17) to determine whether any of the samples that were negative by direct plating still had detectable levels of the pathogen.
For spot inoculation at the four low levels (50, 130, 500, and 1,300 CFU per fruit), the pulp was analyzed on days 5 and 15. For spot inoculation at the high level (106 CFU per fruit), fruits were sampled on days 5 and 10. For hydrocooling at both levels of inoculum, fruits were sampled on days 5, 10, and 15. L. monocytogenes from areas 1, 2, 5, and 6 (Fig. 1A) was also enumerated for each fruit collected on days 10 and 15 after hydrocooling. Three biological replicates were performed for each inoculation level at each time point. The stem scar (nonedible) was removed using a sterile knife, and the avocado was cut transversally starting from the bottom end of the fruit. Edible pulp was collected from six areas at each time point (Fig. 1A).
Statistical analysis. For each time point (days 10 and 15) and each inoculation level, a one-way analysis of variance (ANOVA) was performed on the log CFU values to compare the levels of L. monocytogenes in different areas of the avocado after hydrocooling. For each area at each inoculation level, a t test was performed on the log CFU values to compare the levels of L. monocytogenes between day 10 and day 15 after hydrocooling. Some samples in areas 5 and 6 were positive for L. monocytogenes as confirmed after enrichment, but the levels were below the LOD (16.6 CFU/g) for direct plating. For statistical analysis, the level of L. monocytogenes in these samples was assumed to be half of the LOD (8.3 CFU/g).
Complete surface disinfection of the avocados that had been inoculated with L. monocytogenes by either spot inoculation or hydrocooling in bacterial suspension was critical to prevent pathogen transfer from the rind to the pulp during subsequent sampling. A 5-min wash in 0.525% sodium hypochlorite solution followed by thorough water rinsing provided complete disinfection of the fruit surface, and no L. monocytogenes or aerobic bacteria were recovered from the water rinsates. The L. monocytogenes strain used in the present study was isolated from a naturally contaminated avocado and was able to grow in avocado pulp but not on the skin, as demonstrated previously (7).
Dye was mixed with the inoculum to assist in the subsequent sampling. Spot inoculation of the stem scar area led to the uptake and spread of the dye solution into the avocado mesocarp in areas 1, 2, and 3 (Fig. 1B); therefore, these three areas were analyzed for L. monocytogenes. Two of the three fruits inoculated at the 106 CFU per fruit level were positive for L. monocytogenes on day 5 in area 1, but no pulp areas were positive in avocados inoculated at lower levels (Table 1). On day 10, all three fruits inoculated at 106 CFU per fruit were positive in area 1, and on day 15 two of the three fruits inoculated with lower levels (50, 130, 500, and 1,300 CFU per fruit) were positive in area 1 (Table 1). Even though dye permeated the spot-inoculated fruits up to area 3, L. monocytogenes infiltrated only area 1 of the mesocarp.
Hydrocooling caused the dye solution to permeate much more extensively into the avocado pulp than did spot inoculation. Dye solution permeated the entire fruit (Fig. 1C through 1F) via the stem scar but did not penetrate the fruit skin. All mesocarp areas (areas 1 to 6) were thus analyzed for L. monocytogenes internalization as described above. On day 5 after hydrocooling in water containing 106 CFU/ml L. monocytogenes, area 1 samples from all three replicate fruits were pathogen positive, and samples from all other areas were positive in at least one of the three replicates. On day 5 after hydrocooling in water containing 108 CFU/ml L. monocytogenes, samples from all areas were pathogen positive. Analysis of the avocado pulp on days 10 and 15 after hydrocooling revealed internalized L. monocytogenes in all areas of the mesocarp in all replicates. Enumeration of L. monocytogenes in various areas of the mesocarp revealed a gradient in bacterial populations. For each inoculation level at each time point, areas 1 and 2 (adjacent to the stem) were colonized at much higher levels than were areas 5 and 6 (bottom end of the fruit) (Table 2). No significant differences in internalization of L. monocytogenes or population growth were observed on days 10 and 15 for fruits hydrocooled in water containing 106 or 108 CFU/ml L. monocytogenes.
We hypothesized that water uptake by the fruit and internalization of bacteria occur via the stems or stem scars. To simulate contamination during the washing phase or during postpacking exposure to water condensate, avocados were spot inoculated in the stem scar areas with 100 μl of water containing various levels of L. monocytogenes. This study is the first to reveal that spot contamination of stem scar areas of whole avocados can lead to the internalization of L. monocytogenes into the edible portions of the fruit.
Experimental contamination of avocados during hydrocooling was conducted with two inoculum levels, 6 and 8 log CFU/ml, of L. monocytogenes in the hydrocooling solution. These high inoculum levels were chosen to evaluate the potential for bacterial internalization during hydrocooling and to quantitatively assess the populations of internalized L. monocytogenes. All six areas of the mesocarp were permeated by the dye solution (Fig. 1C through 1F) and were positive for L. monocytogenes (Table 2). This finding suggests that water influx during hydrocooling, possibly driven by the temperature differential, was strong enough to cause the infiltration of L. monocytogenes from the stem end to the bottom end of the avocado. Significantly higher (P ≤0.05) levels of L. monocytogenes were recovered from the stem end (areas 1 and 2) than from the bottom end (areas 5 and 6), indicating that pathogen internalization took place through the stem scar and then spread into other areas of the fruit. Bacterial populations reached 6.16 to 7.19 log CFU/g at the stem end and 1.56 to 2.27 log CFU/g at the bottom end of the mesocarp on day 15 after hydrocooling. For fruits at both inoculation levels, in general the levels of L. monocytogenes in each area were slightly higher at day 15 than at day 10, even though the difference was not significantly different probably because of the limited number of replicates and high individual fruit variability. Iturriaga et al. (19) found that L. monocytogenes could grow in avocado pulp under refrigeration conditions, with an approximately 1-log increase from day 10 to day 15 in avocado pulp stored at 4 to 7°C. This finding is consistent with our findings.
In the present study, the purpose of using the dye was to track the water uptake and distribution in the avocado mesocarp and to assist with subsequent sampling for microbiological analysis. The dye permeation and bacterial penetration into the mesocarp did not always occur to the same extent, which was not surprising because dye particles were much smaller than the L. monocytogenes cells. In spot-inoculated fruits, dye traveled further into the pulp than did L. monocytogenes cells (Tables 1 and 2 and Fig. 1); however, during hydrocooling, L. monocytogenes travelled as far as did the dye (Table 2 and Fig. 1). All areas that were positive for L. monocytogenes also contained dye, which indicates the usefulness of this dye for future studies on bacterial internalization into vegetal matrices.
Herein, we demonstrated for the first time the potential of L. monocytogenes to internalize into the mesocarp of avocados following experimental contamination by spot inoculation in stem scar area and by hydrocooling in contaminated water. Water uptake and L. monocytogenes infiltration took place through the stem scar and then spread within the mesocarp. As a result, the stem end of the pulp was more heavily colonized by L. monocytogenes than was the bottom end. These results also suggest that L. monocytogenes is not likely to penetrate the avocado skin. Because the results of the present study were obtained in a laboratory setting, a follow-up study of L. monocytogenes internalization in a pilot processing plant would allow more accurate determination of the risk of contamination associated with various postharvest practices used for avocado processing. Enhanced sanitation and control of water quality used for washing and precooling of avocados may prevent avocado colonization by L. monocytogenes under certain postharvest conditions.
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the U.S. Food and Drug Administration.