Effects of Abiotic and Biotic Stresses on the Internalization and Dissemination of Human Norovirus Surrogates in Growing Romaine Lettuce

Abstract

Human norovirus (NoV) is the major causative agent of fresh-produce-related outbreaks of gastroenteritis; however, the ecology and persistence of human NoV in produce systems are poorly understood. In this study, the effects of abiotic and biotic stresses on the internalization and dissemination of two human NoV surrogates (murine norovirus 1 [MNV-1] and Tulane virus [TV]) in romaine lettuce were determined. To induce abiotic stress, romaine lettuce was grown under drought and flood conditions that mimic extreme weather events, followed by inoculation of soil with MNV-1 or TV. Independently, lettuce plants were infected with lettuce mosaic virus (LMV) to induce biotic stress, followed by inoculation with TV. Plants were grown for 14 days, and viral titers in harvested tissues were determined by plaque assays. It was found that drought stress significantly decreased the rates of both MNV-1 and TV internalization and dissemination. In contrast, neither flood stress nor biotic stress significantly impacted viral internalization or dissemination. Additionally, the rates of TV internalization and dissemination in soil-grown lettuce were significantly higher than those for MNV-1. Collectively, these results demonstrated that (i) human NoV surrogates can be internalized via roots and disseminated to shoots and leaves of romaine lettuce grown in soil, (ii) abiotic stress (drought) but not biotic stress (LMV infection) affects the rates of viral internalization and dissemination, and (iii) the type of virus affects the efficiency of internalization and dissemination. This study also highlights the need to develop effective measures to eliminate internalized viruses in fresh produce.

INTRODUCTION

Human norovirus (NoV) is the major cause of nonbacterial gastroenteritis, contributing to >95% of nonbacterial acute gastroenteritis worldwide and >60% of all food-borne illnesses reported annually in the United States (1–4). High-risk foods for human NoV contamination include fresh produce, shellfish, and ready-to-eat foods. As individuals are increasingly striving to achieve healthier diets, the consumption of fresh produce has increased in recent years, and fresh produce is now recognized as a leading cause of food-borne illness in the United States (5, 6). Human NoV alone accounts for more than 40% of the fresh-produce-related illnesses reported each year in the United States (1, 2, 4, 6–9). Outbreaks in many diverse types of produce, including fresh cut fruit, lettuce, tomatoes, melons, salads, green onions, strawberries, blueberries, raspberries, and salsa, have been attributed to human NoV (6, 7, 10–15). Human NoV is highly infectious, is resistant to common disinfectants, has a low infectious dose, and is highly stable in the environment, all features that contribute to the high prevalence of food-borne outbreaks associated with the virus and its presence and persistence in food commodities (3, 5, 7, 16–19).

Fresh produce can become contaminated with human NoV at any step from production to processing. In a survey of the point of contamination of specific food commodities with human NoV responsible for outbreaks in the United States from 2001 to 2008, nearly half of the outbreaks in leafy vegetables and fruits/nuts were associated with viral contamination that occurred during preparation or service; only a few outbreaks were traced to contamination during production and processing (2). However, in a large number of the outbreaks of human NoV in leafy vegetables and fruits/nuts, the point of contamination could not be determined, and it is possible that more contamination occurred during production and processing than could be verified in this survey (2). During production, virus-contaminated irrigation water, water for the dilution of agrochemicals and fertilizers, and water for hydroponic cultures can all introduce virus into fresh produce (20–22). Sewage-contaminated irrigation water has been theorized to account for several human NoV outbreaks in berries, although this has not been confirmed (11, 12). Producers use various water sources for irrigation, including well water, river water, and lake water; human NoV has been detected in surface waters and also well water (23–30). Infected humans can shed 10⁶ to 10¹¹ virus particles per g of feces, and it has been shown that conventional wastewater treatment practices are not sufficient to completely inactivate or remove viruses (31–35). Therefore, even treated water may harbor human NoV and, when discharged, may contaminate sources of irrigation water. The use of human-NoV-contaminated irrigation water in the production of fresh produce poses a risk of surface contamination of the food as well as the potential for internalization of the virus into the produce during growth.

Human NoV is a member of the family Caliciviridae, and all viruses within this family are nonenveloped, with a single-stranded, positive-sense RNA genome (9, 36–38). A major hindrance to human NoV research is the fact that the virus cannot be cultivated in cell culture and lacks a small-animal model. To date, much of the understanding of the molecular biology, pathogenesis, and environmental stability of human NoV has come from the study of other caliciviruses, which serve as human NoV surrogates (39–43). These surrogate viruses are cultivable in the laboratory and closely resemble human NoV with regard to size, genetic makeup, receptor binding, cell tropism, and disease manifestations. Murine norovirus (MNV) has been used extensively as a surrogate for the study of human NoV and is currently the only cultivatable member of the genus Norovirus. MNV has the closest genetic relation to human NoVs of all surrogate viruses and also has a capsid structure, viral particle size, and buoyant density similar to those of human NoVs (44). MNV has been shown to be more stable at low pHs than other surrogates, such as feline calicivirus (FCV), indicating that the environmental stability of MNV may be similar to that of human NoV (40). However, MNV does not cause gastroenteritis in mice and utilizes sialic acid for attachment to cells, a pattern different from that of human NoV, where histo-blood group antigens (HBGAs) are the cellular attachment moieties. Tulane virus (TV) is a newly recognized surrogate for human NoV and is a member of the genus Recovirus, family Caliciviridae (45). TV is closely related to human NoV in terms of genome size and organization, viral capsid structure, and viral particle size (45). TV was isolated from the stool of rhesus macaques, which implicates it as a cause of gastrointestinal disease. Further, like human NoV, TV utilizes HBGAs as cellular attachment molecules (46).

Several studies have evaluated whether enteric viruses and their surrogates can be internalized in growing produce, with various results depending on the virus used, plant growth conditions, and plant varieties (47–54). However, no studies have systematically evaluated whether abiotic and biotic stresses on growing produce affect the levels of enteric virus internalization and dissemination. In this study, the levels of internalization and dissemination of human norovirus surrogates MNV-1 and TV in romaine lettuce grown in soil were determined. The plants were maintained under normal conditions or under drought or flood conditions, mimicking extreme weather events, to determine whether abiotic stresses affect virus internalization. In addition, infection with a phytopathogen (lettuce mosaic virus [LMV]) was introduced into lettuce in order to determine whether biotic stress affected enteric virus internalization and dissemination.

MATERIALS AND METHODS

Viruses and cell culture.

Murine norovirus 1 (MNV-1) was generously provided by Herbert W. Virgin IV at the Washington University School of Medicine, and Tulane virus (TV) was a generous gift from Xi Jiang at the Cincinnati Children's Hospital Medical Center. MNV-1 and TV were propagated in confluent monolayers of the murine macrophage cell line RAW 264.7 and the monkey kidney cell line MK2-LLC (ATCC, Manassas, VA), respectively (47). RAW 264.7 cells were cultured in high-glucose Dulbecco's modified Eagle medium (DMEM) (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (FBS) (Invitrogen), at 37°C under a 5% CO₂ atmosphere. For growing MNV-1 stock, confluent RAW 264.7 cells in T-150 flasks were infected with MNV-1 at a multiplicity of infection (MOI) of 0.1. After 1 h of incubation at 37°C, 15 ml of DMEM with 2% FBS was added. The virus was harvested 2 days postinoculation (p.i.) by three freeze-thaw cycles and low-speed centrifugation at 1,000 × g for 30 min. MK2-LLC cells were cultured in low-serum Eagle's minimum essential medium (Opti-MEM; Invitrogen), supplemented with 2% FBS, at 37°C under a 5% CO₂ atmosphere. For growing TV stock, MK2-LLC cells in T-150 flasks were washed with Hanks' balanced salt solution (HBSS) and were subsequently infected with TV at an MOI of 0.1. After 1 h of incubation at 37°C, 15 ml of Opti-MEM with 2% FBS was added. The virus was harvested on day 2 p.i. and was subjected to three freeze-thaw cycles, followed by centrifugation at 1,000 × g for 30 min to remove cellular debris.

Virus enumeration by plaque assays.

MNV-1 and TV were quantified by plaque assays in RAW 264.7 and LLC-MK2 cells, respectively (47, 55). Briefly, cells were seeded into six-well plates (Corning Life Sciences, Wilkes-Barre, PA) at a density of 2 × 10⁶ per well. After 24 h of incubation, RAW 264.7 or MK2-LLC cell monolayers were infected with 400 μl of a 10-fold dilution series of MNV-1 or TV, respectively, and the plates were incubated for 1 h at 37°C with gentle agitation every 10 min. The cells were overlaid with 3 ml of Eagle minimum essential medium (MEM) containing 1% agarose, 2% FBS, 1% sodium bicarbonate, 0.1 mg of kanamycin/ml, 0.05 mg of gentamicin/ml, 15 mM HEPES (pH 7.7), and 2 mM l-glutamine. After incubation at 37°C under 5% CO₂ for 2 days, the plates were fixed in 10% formaldehyde. The plaques were visualized by staining with 0.05% (wt/vol) crystal violet. The limit of detection for the viral plaque assay was determined to be 0.5 log₁₀ PFU/ml. Viral titers were expressed as mean log₁₀ PFU per milliliter ± standard deviation.

Determination of viral survival in a conventional potting mix.

One gram of Fafard 3B potting mix was weighed and placed in a sterile petri dish. For the virus-inoculated soils, 1 ml of 1 × 10⁷ PFU/ml of virus (MNV-1 or TV) was pipetted onto the soil. (Control soils received no virus.) The plates were sealed with Parafilm and were maintained in the lab under conditions of 12 h of light and 12 h of darkness, at 20°C and 50% relative humidity (RH). On day 0 (immediately after inoculation), day 7, and day 14, three samples from each group (MNV-1 inoculated, TV inoculated, or uninoculated control) were processed for the determination of viral survival. One gram of soil contained in each petri dish was transferred to a sterile stomacher bag, followed by the addition of 10 ml sterile distilled water (diH₂O) to each bag. Each soil sample was blended for 2 min, and then the liquid was collected from each stomacher bag and was transferred to a sterile 15-ml tube. The samples were centrifuged for 30 min at 1,000 × g to pellet soil particles; the supernatant was transferred to a new 15-ml collection tube; and the volume of the collected supernatant was recorded. Virus survival was determined using a plaque assay, and results were reported as log₁₀ PFU per gram of soil.

Plant cultivation under normal or extreme weather conditions.

Seeds of romaine lettuce (Lactuca sativa) were planted in plug trays and were grown under greenhouse conditions. Ten days after germination, plugs were transplanted into 4-in pots containing Fafard 3B potting mix. Thirty days after germination, plants were maintained in the lab under conditions of 12 h of light and 12 h of darkness, at 20°C and 50 to 65% RH. The plants were separated into 4 groups: normal conditions with no viral inoculation (control), normal conditions with viral inoculation, drought conditions with viral inoculation, and flood conditions with viral inoculation. The pots were kept in tray flats, and trays were filled with water for delivery to lettuce. For normal conditions, the trays were filled every 5 days with 1,000 ml water. For drought conditions, the trays were filled every 14 days with 1,000 ml water. For flood conditions, the trays were filled every 2 days with 2,000 ml. Following lettuce harvest, the percentage of moisture was determined as described below.

Virus inoculation, sample collection, and virus internalization.

Plants were inoculated at the base of the shoot with 20 ml of 1 × 10⁷ PFU/ml virus, in two separate 10-ml treatments 2 h apart on the same day. Following inoculation, the surface of the soil was covered with Parafilm to limit viral contamination of the aerial portions of the plant. At days 0 (before viral inoculation), 1, 3, 7, and 14, the leaves, shoots, and roots were harvested. All tissues were submerged in 2,000 ppm chlorine following harvest to remove any virus present on the surface of the plant. Following chlorine treatment, lettuce tissues were rinsed for 5 min in diH₂O, and following the diH₂O rinse, residual chlorine was inactivated by the submersion of tissues in 0.25 M sodium thiosulfate. The shoot sample consisted of the aerial portion of the lettuce 3 in above the soil line, and 10-g samples (total, 30 to 40 g) of leaves were randomly selected from each plant for homogenization. The samples were homogenized using liquid nitrogen and mortars and pestles. Homogenized tissue was resuspended in 5 ml phosphate-buffered saline (PBS), and homogenates were centrifuged at 1,000 × g for 30 min to remove cellular debris. The virus-containing supernatant was transferred to a new collection tube for viral enumeration by plaque assay.

Soil measurements.

Soil samples from each time point were collected for the measurement of pH, salinity, conductivity, total dissolved solids (TDS), and percentage of moisture. To determine the percentage of moisture in the soil samples, 1 g of soil was weighed, and the wet weight was recorded. The soil sample was then dried overnight in a 60°C drying oven, followed by 1 h of incubation in a desiccator. The percentage of moisture was determined by dividing the dry weight by the initial wet weight and then multiplying by 100. To determine the pH, salinity, conductivity, and TDS, 1-g samples of soil were suspended in 10 ml of diH₂O. The samples were shaken vigorously for 2 min to disperse the soil in the water. Samples were allowed to settle at room temperature for 1 h. Measurements were taken using standardized meters for pH (accumet basic pH meter; Fisher Scientific), salinity (Fisher Traceable salinity meter), conductivity (Fisher Traceable conductivity meter), and TDS (Fisher Traceable conductivity/TDS meter) (56).

Inoculation of lettuce with LMV.

Seeds of romaine lettuce (Lactuca sativa) were planted in plug trays and were grown under greenhouse conditions. Ten days after germination, plugs were transplanted into 4-in pots containing Fafard 3B potting mix. Thirty days after germination, plants were maintained in the lab under conditions of 12 h of light and 12 h of darkness, at 20°C and 50% RH. Thirty days postgermination, lettuce was infected with lettuce mosaic virus (LMV) (ATCC PV-63) (ATCC, Manassas, VA, USA). The virus was inoculated into the lettuce using the plant virus revival method provided by ATCC (57). Briefly, 0.1 g of LMV-infected lettuce tissue was homogenized with a mortar and pestle and was resuspended in 3 ml phosphate-buffered saline. Approximately 25 μl of the LMV virus suspension was inoculated into two leaves of each lettuce plant using a cotton swab and 1% carborundum powder (400 grit; Fisher Science Education, Hanover Park, IL, USA) to damage the leaf surface and allow for viral entry. Following inoculation, the leaves were washed with diH₂O to remove residual carborundum powder and salts. The plants were allowed to grow for another 10 days until symptoms of LMV infection (yellow circular spots) were observed. At day 40 postgermination, lettuce plants were inoculated with TV, and plants were harvested using the methods described above. LMV infection was confirmed using a commercially available diagnostic LMV enzyme-linked immunosorbent assay (ELISA) kit (Agdia, Elkhart, IN) on leaf homogenates.

Statistical analysis.

All experiments were performed in triplicate. Statistical analysis was performed by one-way analysis of variance (ANOVA) using Tukey's family comparisons using Minitab 16 statistical analysis software (Minitab Inc., State College, PA). A P value of <0.05 was considered statistically significant.

RESULTS

Stability of MNV-1 and TV in a commercial potting mix.

We first determined the stability of MNV-1 and TV in soils. One milliliter of 1 × 10⁷ PFU/ml of virus (MNV-1 or TV) was pipetted onto the soils and was mixed uniformly. On day 0 (immediately after inoculation), day 7, and day 14, three samples from each group (MNV-1 inoculated, TV inoculated, or the uninoculated control) were processed for the determination of viral survival by a plaque assay (Fig. 1). On day 0, the titer of MNV-1 in soil was 6.47 ± 0.54 log₁₀ PFU/g, and the titer of TV was 6.31 ± 0.10 log₁₀ PFU/g. The titer of MNV-1 decreased by approximately 2 log units on day 7, to 4.50 ± 0.16 log₁₀ PFU/g, and remained stable at this level until day 14, with a final titer of 4.17 ± 0.34 log₁₀ PFU/g. The titer of TV was reduced by approximately 1 log unit at days 7 and 14, to 5.37 ± 0.37 and 5.44 ± 0.22 log₁₀ PFU/g, respectively. All control soils were negative for infectious virus. This result suggests that both MNV-1 and TV were relatively stable in soils for at least 14 days.

FIG 1.

Stability of MNV-1 and TV in a commercial potting mix. One gram of Fafard 3B potting mix was weighed and was placed in a sterile petri dish. For virus-inoculated soils, 1 ml of 1 × 10⁷ PFU/ml of virus (MNV-1 or TV) was pipetted onto the soil. (Control soils received no virus.) On day 0 (immediately after inoculation), day 7, and day 14, three samples from each group (MNV-1 inoculated, TV inoculated, or the uninoculated control) were processed for the determination of viral survival by a standard plaque assay. Each point represents results for an average of 3 soil samples ± standard deviations.

Effects of abiotic stress on the internalization and dissemination of MNV-1 in romaine lettuce grown in soil.

As shown in Fig. 2, under normal growth conditions, MNV-1 was disseminated to the leaves, and low levels of infectious virus were detected. On day 1 postinoculation, 2 of 3 roots were positive for MNV-1, with an average titer of 2.57 ± 2.23 log₁₀ PFU/g (Fig. 2A). On days 3, 7, and 14 postinoculation, 3 of 3 roots were positive for MNV-1, with titers of 2.69 ± 0.42, 2.74 ± 1.03, and 3.83 ± 0.37 log₁₀ PFU/g, respectively (Fig. 2A). Similarly, on day 1 postinoculation, 2 of 3 shoots were positive for MNV-1, with an average titer of 1.25 ± 1.09 log₁₀ PFU/g (Fig. 2B). All shoot tissues were positive for MNV-1 on days 3, 7, and 14 postinoculation, with titers of 1.96 ± 0.34, 1.71 ± 0.58, and 2.96 ± 0.81 log₁₀ PFU/g, respectively (Fig. 2B). On day 1 postinoculation, 1 of 3 leaf samples was positive for MNV-1, with an average titer of 0.55 ± 0.96 log₁₀ PFU/g (Fig. 2C). All leaf tissues were positive for MNV-1 on days 3, 7, and 14 postinoculation, with average titers of 1.67 ± 0.39, 1.95 ± 0.67, and 2.52 ± 0.42 log₁₀ PFU/g, respectively (Fig. 2C). The average percentage of moisture, pH, conductivity, salinity, and TDS of the MNV-1-inoculated soil under normal conditions were 75.95%, pH 6.6320, 603.8 μS, 0.3260 ppt, and 335.0 ppm, respectively (Table 1).

FIG 2.

MNV-1 internalization and dissemination in romaine lettuce grown in soil under normal, drought, and flood conditions. (A) Root; (B) shoot; (C) leaf. Each bar represents the average viral titer (in log PFU per gram) for 3 plants; error bars, standard deviations. Groupings (a, b, c, d, e) are based on one-way ANOVA using Tukey's comparisons, with a 95% confidence interval, within each tissue type (root, shoot, leaf) under all experimental conditions. The detection limit for the plaque assay is 0.5 log PFU/g tissue. The number of virus-positive tissues among total plant tissues is given above each bar. Values with different lowercase letters (a, b, c, d, e) are significantly different (P < 0.05). Values with overlapping lowercase letters are not significantly different (P > 0.05).

TABLE 1.

Soil measurements for MNV-inoculated soil maintained under normal, drought, and flood conditions^a

Conditions	Conductivity (μS)	Salinity (ppt)	TDS (ppm)	pH	Moisture (%)
Normal	603.8 ± 231.2 A	0.3260 ± 0.13 A	355.0 ± 135.6 A	6.63 ± 0.34 A	75.95 ± 10.22 A
Drought	1,613.2 ± 708.0 B	0.8987 ± 0.37 B	736.0 ± 305.2 B	6.37 ± 0.13 B	38.16 ± 16.81 B
Flood	364.9 ± 137.6 A	0.1979 ± 0.09 A	215.0 ± 81.7 A	6.77 ± 0.17 A	78.78 ± 8.22 A

^aEach value represents the average of 3 measurements ± standard deviation. Groupings (A, B) are based on one-way ANOVA using Tukey's comparisons, with a 95% confidence interval. Values followed by different letters (A, B) within a column are significantly different (P < 0.05).

Under drought conditions, the levels of internalization and dissemination of MNV-1 were lower than those under normal conditions. On day 1 postinoculation, 3 of 3 leaf samples were positive for infectious MNV-1, with an average titer of 1.33 ± 0.65 log₁₀ PFU/g; however, the level of MNV-1 in leaves decreased over the course of the study (Fig. 2C). Specifically, on days 3 and 7 postinoculation, only 1 of 3 leaf samples were positive for infectious MNV-1, and on day 14 postinoculation, no infectious MNV-1 was detected in any of the replicates. On day 1 postinoculation, all shoot samples tested positive for infectious MNV-1, with a titer of 3.14 ± 1.72 log₁₀ PFU/g, but again, the level of MNV-1 detected in the shoots decreased over the 14-day study period (Fig. 2B). Only 1 of 3 shoot samples was positive on day 3 postinoculation; however, on day 7 postinoculation, 3 of 3 shoot tissues were positive for MNV-1, with an average titer of 2.22 ± 0.80 log₁₀ PFU/g (Fig. 2B). On day 14 postinoculation, no shoot samples were positive for MNV-1 (Fig. 2B). On day 1 postinoculation, 3 of 3 root samples were positive for MNV-1, with an average titer of 3.68 ± 0.27 log₁₀ PFU/g (Fig. 2A). The average percentage of moisture in the soil under drought conditions was 38.16%, which was significantly different (P < 0.05) from that under normal conditions (Table 1). Also, the pH was 6.3707, the conductivity was 1,613.2 μS, the salinity was 0.8987 ppt, and the TDS was 736.0 ppm (Table 1). All of the soil measurements were also significantly different from those under normal soil conditions (P < 0.05).

The rates of MNV-1 internalization and dissemination under flood conditions were similar to those under normal conditions. On day 3 postinoculation, all leaf samples were positive for MNV-1, with a titer of 1.45 ± 0.27 log₁₀ PFU/g (Fig. 2C). On days 7 and 14 postinoculation, 3 of 3 leaf samples were positive for MNV-1, with average titers of 2.80 ± 0.49 and 2.02 ± 0.18 log₁₀ PFU/g, respectively (Fig. 2C). On days 1, 3, and 7 postinoculation, 2 of 3 shoot samples were positive for MNV-1, and on day 14 postinoculation, 3 of 3 shoot samples were positive (Fig. 2B). MNV-1 was detected in all root samples on all study days except for day 14, when none of the 3 samples tested positive for MNV-1 (Fig. 2A). The percentage of moisture under flood conditions was 78.78% (Table 1). The pH, conductivity, salinity, and TDS were pH 6.7657, 364.9 μS, 0.1979 ppt, and 215.0 ppm, respectively (Table 1). These soil measurements were also significantly different from those taken under normal soil conditions (P < 0.05), a finding consistent with the goal of mimicking flood conditions.

Effects of abiotic stress on the internalization and dissemination of TV in romaine lettuce grown in soil.

Next, we determined whether abiotic stress had similar effects on the internalization of TV, another animal calicivirus. In order to directly compare internalization efficiencies, experiments were carried out side-by-side with those described above for MNV-1. On all study days, under normal conditions, the roots of all plants were positive for TV (Fig. 3A). The average titer detected in the roots on day 14 postinoculation was 3.78 ± 1.25 log₁₀ PFU/g (Fig. 3A). The day 14 shoots of all lettuce plants were also positive, with an average titer of 4.30 ± 0.85 log₁₀ PFU/g detected (Fig. 3B). TV-inoculated lettuce grown under normal conditions contained infectious virus in the leaves of all replicates on day 14 postinoculation, with an average titer of 4.63 ± 0.69 log₁₀ PFU/g (Fig. 3C). The average percentage of moisture, pH, conductivity, salinity, and TDS under normal growth conditions were 68.8%, pH 6.6313, 865.4 μS, 0.4880 ppt, and 523.1 ppm, respectively (Table 2). This result also demonstrated that the efficiencies of internalization and dissemination of TV at days 7 and 14 under normal conditions were significantly higher than those for MNV-1 (compare Fig. 2 and 3, normal conditions) (P < 0.05).

FIG 3.

TV-1 internalization and dissemination in romaine lettuce grown in soil under normal, drought, and flood conditions. (A) Root; (B) shoot; (C) leaf. Each bar represents the average viral titer (in log PFU per gram) for 3 plants; error bars, standard deviations. Groupings (a, b, c, d, e) are based on one-way ANOVA using Tukey's comparisons, with a 95% confidence interval, within each tissue type (root, shoot, leaf) under all experimental conditions. The detection limit for the plaque assay is 0.5 log PFU/g tissue. The number of virus-positive tissues among total plant tissues is given above each bar. Values with different lowercase letters (a, b, c, d, e) are significantly different (P < 0.05). Values with overlapping lowercase letters are not significantly different (P > 0.05).

TABLE 2.

Soil measurements for TV-inoculated soil maintained under normal, drought, and flood conditions^a

Conditions	Conductivity (μS)	Salinity (ppt)	TDS (ppm)	pH	Moisture (%)
Normal	865.4 ± 412.7 A	0.4880 ± 0.23 A	523.1 ± 288.0 A	6.63 ± 0.14 A	68.84 ± 7.51 A
Drought	2,179.7 ± 1,328.5 B	1.2453 ± 0.69 B	741.2 ± 124.6 B	6.46 ± 0.29 A	46.77 ± 18.55 B
Flood	282.3 ± 131.0 A	0.1400 ± 0.09 A	165.0 ± 76.6 C	6.67 ± 0.33 A	80.43 ± 6.93 C

^aEach value represents the average of 3 measurements ± standard deviation. Groupings (A, B, C) are based on one-way ANOVA using Tukey's comparisons, with a 95% confidence interval. Values followed by different letters (A, B, C) within a column are significantly different (P < 0.05).

Under drought conditions, levels of viral internalization and dissemination into TV-inoculated lettuce were lower than those under normal conditions (P < 0.05) (Fig. 3), a finding similar to that for MNV-1 (Fig. 2). The titer detected in the roots on day 1 postinoculation was 4.15 ± 0.46 log₁₀ PFU/g but decreased to 2.44 ± 0.98 log₁₀ PFU/g on day 14 (Fig. 3A). Similarly, low levels of TV were recovered from the shoots, and on day 14 postinoculation, the average titer of TV in shoot tissues was 1.39 ± 1.30 log₁₀ PFU/g (Fig. 3B). Day 7 postinoculation was the only study day on which all leaf samples were positive for TV, with a titer of 1.93 ± 0.13 log₁₀ PFU/g (Fig. 3C). On day 14 postinoculation, 2 of 3 leaf samples were positive for TV, with an average titer of 1.06 ± 0.99 log₁₀ PFU/g (Fig. 3C). The average percentage of moisture under drought conditions was 46.77%, significantly different from that under normal conditions (Table 2) (P < 0.05). The average pH, conductivity, salinity, and TDS were pH 6.4587, 2,179.7 μS, 1.2453 ppt, and 741.2 ppm, respectively, and all soil measurements, except for pH, were significantly different from those under normal conditions (Table 2) (P < 0.05).

TV internalization and dissemination patterns in lettuce grown under flood conditions showed no significant difference from those under normal conditions (P > 0.05) (Fig. 3), a finding that was also similar to that for MNV-1 (Fig. 2). All leaf samples tested positive for virus on day 14, with an average titer of 3.29 ± 0.76 log₁₀ PFU/g (Fig. 3C). Additionally, all shoot samples tested positive for virus on day 14, with a titer of 3.35 ± 0.18 log₁₀ PFU/g (Fig. 3B). Again, all root samples tested positive for infectious virus on all study days. The average viral titer in the lettuce roots on day 14 was 2.43 ± 0.98 log₁₀ PFU/g (Fig. 3A). The average percentage of moisture in the soil under flood conditions was 80.43%, statistically different from that under normal conditions (Table 2). The average pH, conductivity, salinity, and TDS were pH 6.6667, 282.3 μS, 0.1400 ppt, and 165.00 ppm, respectively (Table 2).

Effects of biotic stress on the internalization and dissemination of TV in romaine lettuce grown in soil.

At day 40 postgermination, lettuce plants (LMV infected and uninfected) were inoculated with 20 ml of 1 × 10⁷ PFU/ml of TV. Plants were allowed to grow for 14 days, and leaves, shoots, and roots were harvested at day 0, day 1, day 3, day 7, and day 14 postinoculation. Harvested plant tissues were homogenized, and viral titers were determined using plaque assays.

All of the lettuce infected with LMV tested positive for the virus in leaf samples by ELISA; uninfected lettuce plants were ELISA negative. In LMV-infected plants, infectious TV was detected in 2 of 3 roots on day 1 postinoculation, with an average titer of 3.59 ± 3.11 log₁₀ PFU/g (Fig. 4A). Similarly, 2 of 3 shoot samples tested positive for infectious TV in LMV-infected plants, with an average titer of 3.15 ± 2.92 log₁₀ PFU/g, and 3 of 3 leaf samples were positive for TV, with a higher titer of 4.41 ± 0.41 log₁₀ PFU/g (Fig. 4B and C). On all other study days, infectious TV was recovered in 3 of 3 samples of all lettuce tissues from plants infected with LMV, except for day 7 root tissues, where 2 of 3 samples were positive (Fig. 4). The highest titer in the leaf tissue of lettuce plants infected with LMV—an average titer of 4.79 ± 0.71 log₁₀ PFU/g—was detected on day 14 (Fig. 4C).

FIG 4.

Effects of biotic stress on the internalization and dissemination of TV in romaine lettuce grown in soil. (A) Root; (B) shoot; (C) leaf. Each bar represents the average viral titer (log PFU per gram) for 3 plants; error bars, standard deviations. Groupings (a, b, c) are based on one-way ANOVA using Tukey's comparisons, with a 95% confidence interval, within each tissue type (root, shoot, leaf) under each experimental condition (uninfected, LMV infected). The detection limit for the plaque assay is 0.5 log PFU/g tissue. The number of virus-positive tissues among total plant tissues is given above each bar. Values with different lowercase letters (a, b, c) are significantly different (P < 0.05). Values with overlapping lowercase letters are not significantly different (P > 0.05).

For uninfected lettuce plants, there was no study day on which the titer of internalized TV detected was significantly different from that in lettuce plants infected with LMV (P > 0.05). On day 1, 3 of 3 root tissues and 3 of 3 leaf tissues tested positive for infectious TV, with average titers of 4.29 ± 0.26 log₁₀ PFU/g and 3.43 ± 0.78 log₁₀ PFU/g, respectively (Fig. 4A and C). Among shoot tissues on day 1 postinoculation, 2 of 3 tissues tested positive for TV, with an average titer of 2.62 ± 2.51 log₁₀ PFU/g (Fig. 4B). On all other study days tested, 3 of 3 lettuce tissues tested positive for TV, except for day 3 leaves and day 7 roots, for which 2 of 3 samples tested positive (Fig. 4). The highest TV titer in the leaf tissues—an average of 4.62 ± 1.06 log₁₀ PFU/g—was detected on day 14 postinoculation (Fig. 4C). Therefore, these results demonstrated that biotic stress did not significantly alter the internalization and dissemination of TV grown in soil (P > 0.05).

DISCUSSION

Human NoV is the leading causative agent of fresh-produce-associated outbreaks of gastroenteritis. Several outbreaks of human NoV have been associated with the consumption of lettuce or mixed salads, and fresh produce remains a high-risk food for human NoV contamination, because it can become contaminated at both the preharvest and postharvest stages. Previous research utilizing hydroponic cultivation of lettuce has shown that viruses can become internalized via the root and be disseminated to the aerial portions of the plant (47, 48, 50). The plant growth medium has been shown to play the most significant role in pathogen internalization via uptake by the root system (51). The factors that contribute the most significantly to viral transport in the soil include the soil type, water saturation, pH, conductivity, and organic matter (58, 59). Based on the variability that soil growth medium adds to pathogen internalization, the detection of virus internalization and dissemination in fresh produce grown in soil has been less well established than in hydroponic systems (51, 52).

Effects of abiotic stress on the internalization and dissemination of viruses in soil-grown lettuce.

In the field, produce may be subjected to extreme weather events, such as drought or flooding, and therefore, the effects of these events on viral internalization and dissemination in soil-grown lettuce were evaluated. In this study, it was found that both TV- and MNV-1-inoculated lettuce grown in soil under normal conditions internalized the virus in the leaves of all replicates on day 14 postinoculation, with average titers of 4.63 log₁₀ PFU/g and 2.52 log₁₀ PFU/g for TV and MNV-1, respectively. Under normal soil conditions, TV and MNV-1 internalization was observed at a higher level than under other soil conditions. Viral internalization and dissemination under flood conditions were similar to those under normal conditions. However, it was observed that viral internalization and dissemination were significantly impaired in water-depleted soil (drought conditions). Only on 1 study day (day 7 postinoculation) were all the leaves and all the shoots positive for infectious TV, indicating a decrease in dissemination efficiency under drought stress. Under drought conditions, the efficiency of MNV-1 internalization was also markedly lower than that under normal conditions; on day 14 postinoculation, none of the lettuce leaves were positive for infectious MNV-1. It is possible that the increased salinity, conductivity, and TDS of the drought soil conditions caused the viruses to bind to the soil matrix, decreasing their dissemination efficiency. It is also possible that the lower abundance of water in soil under drought conditions led to less passive uptake of the viruses by the lettuce. The fact that both TV and MNV-1 were found to be internalized and disseminated in lettuce grown under normal and flood conditions and were relatively stable in a potting mix indicates that the lower dissemination rate under drought conditions is not due to viral inactivation in soil.

A previous study investigating the impact of extreme weather events on bacterial internalization observed that when a high inoculum was applied to soil (10⁸ to 10⁹ CFU/ml), the level of internalization of Salmonella enterica serovar Typhimurium was higher in drought-affected or flooded soils than under normal conditions (60). Bacteria can respond to taxis and environmental changes, and these factors are likely to influence their entry into fresh produce. During drought stress, bacteria may move into plant tissues in order to access water, and under flood conditions, they may move into plants, because plants have higher nutrient concentrations than the surrounding saturated soils. However, viruses would be more influenced by factors that impact their passive movement in soils. In this study, viral dissemination was found to be highest under normal conditions, with minimal differences observed under flood conditions. The average percentage of moisture in TV-inoculated soil under flood conditions was 80.43%, a significant increase over that under normal conditions, and the percentage of moisture in MNV-1-inoculated soil under flood conditions was 78.78%. A highly water saturated soil allows for more viral movement, since all the pores in the soil are open and the virus has less interaction with the soil particles (61–65). However, in this study, it was found that high water saturation did not have a significant influence on virus internalization, perhaps because of the decrease in plant transpiration or the dilution of virus due to the overabundance of water.

It was observed that the level of viral internalization was significantly reduced when the lettuce was grown under drought conditions. The average conductivity, salinity, and TDS in soils under drought conditions with TV or MNV-1 inoculation were significantly different from those under normal conditions. It has been demonstrated previously that the presence of cations in the soil favors viral absorption by soil particles, since the presence of cations limits the amount of repulsive forces between the virus and the soil (63, 66). The high conductivity, salinity, and TDS of the drought-affected soil may have caused the virus to persist in the soil, since the virus bound to charged particles and was unavailable for passive uptake by the lettuce. Under acidic conditions, viruses normally possess a negative charge. Thus, the virus may adhere to positively charged materials found in the soil and be tightly bound in the soil matrix. In terms of soil, pH 6.1 to 6.5 is considered slightly acidic, and pH 6.6 to 7.3 is considered neutral (58, 59). Under drought conditions, the soil pH was 6.4 with TV and 6.3 with MNV-1, which may account for the decrease in viral dissemination for TV, since the virus was bound to soil. In neutral or alkaline soils, the virus will not bind to materials in the soil and can easily be disseminated throughout the soil matrix (64, 65). For both TV and MNV-1, under both normal and flood conditions, the soil was in the neutral range, which would have eased viral movement from the soil to the plant.

Additionally, drought stress can lead to osmotic stress within the plant, leading to loss of turgor of the plant as well as a decreased root mass (67, 68). The decrease in root mass would have a direct effect on the amount of water uptake and viral internalization by the plant. During drought stress, plant cell membranes can dissociate, and plant proteins may lose function, which can lead to an overabundance of reactive oxygen species (ROS) in the plant (68). These ROS have a high oxidation potential, and it is possible that the ROS can react with the viral protein capsid and render the virus noninfectious. Although not directly evaluated in this study, these other effects of drought stress on plants may have altered the rates of viral internalization and dissemination.

Effects of biotic stress on the internalization and dissemination of TV in soil-grown lettuce.

It has been suggested that phytopathogen infections, which occur frequently during field cultivation of crops, can affect the interaction between human bacterial pathogens and plants (69). Therefore, the effects of LMV infection on the rates of TV internalization and dissemination in lettuce were investigated. No difference in the level of TV internalization was found between lettuce plants infected with LMV and uninfected plants. Similarly, a study evaluating Salmonella Typhimurium internalization in iceberg lettuce infected with LMV found that following foliar contamination with S. Typhimurium, the level of bacterial internalization in LMV-infected lettuce was not significantly different from that in uninfected lettuce (57).

An innate plant immune response to pathogen invasion is stomatal closure, which would inhibit human pathogen invasion from aerial portions of the plant but would not affect internalization via the root, which was tested in this study (57). Also, specific alleles in lettuce that help the plant combat or tolerate LMV infection have been identified. LMV infection induces the expression of these genes that block the entry of the virus into the vascular tissue, thereby inhibiting the systematic spread of LMV throughout the plant (57, 70). However, this response would not inhibit human enteric viruses from moving through the vascular tissues when the point of entry of the human pathogen was distinct from the LMV-infected tissues, which was the case in this study. LMV infection did not have a significant impact on the levels of TV internalization and dissemination in lettuce; however, the effects of other biotic stressors (such as other viruses, bacteria, fungi, and nematodes) and additional routes of enteric virus contamination during biotic stress warrant further investigation.

Effects of virus type on the rates of internalization and dissemination in soil-grown lettuce.

Overall, MNV-1 exhibited much less efficient internalization and dissemination than TV under normal conditions and abiotic stress conditions. This could be due to differences between the two viruses, such as environmental stability and cellular receptors. MNV-1 has been reported to be very stable under laboratory conditions; however, its stability in soil has not been evaluated. The stability of MNV-1 in a commercial potting mix was evaluated in this study, and it was found that the titer of MNV-1 decreased by 2 log units after 14 days, compared to a 1-log-unit reduction in the TV titer over the same period, suggesting either that TV was more stable in soil than MNV-1 or that MNV-1 adhered to components of the soil matrix more than TV. Additionally, MNV-1 utilizes sialic acid as its cellular receptor, while TV utilizes the histo-blood group antigens (HBGAs). Sialic acid is ubiquitous in nature, while HBGAs are less abundant. It is possible that more sialic acid residues were present in the soil and bound MNV-1, making it less available for uptake than TV.

It should be noted that a high virus inoculum (2 × 10⁸ PFU) was applied to each lettuce plant in this study and that the level of viral recovery from lettuce tissues was relatively low (1 to 4 log₁₀ PFU/g). However, given the fact that human NoV can be shed at a level of 10¹¹ particles per g of feces and that fewer than 10 virus particles were sufficient to cause illness, this low level of internalization may pose risks for food safety and public health (9, 19). In fact, human NoV has been detected in sources of surface water and groundwater, which may provide irrigation water for produce. However, limited quantitative data are available for the levels of human NoV present in surface water and groundwater. Analysis of river water in France and Norway showed that human NoV GI and GII were present at levels of 10² to 10³ RNA copies/liter (71, 72). Further investigation into the levels of human NoV in irrigation water, as well as determination of the minimum level of human NoV leading to internalization in fresh produce, can help to elucidate the occurrence of this phenomenon during field production of fresh produce.

In summary, these results demonstrate that (i) abiotic stress induced by drought conditions significantly decreased the rates of TV and MNV-1 internalization and dissemination in romaine lettuce grown in soil, (ii) biotic stress induced by LMV infection did not significantly affect the rates of TV internalization and dissemination, and (iii) different viruses (MNV-1 and TV) have differing rates of internalization and dissemination in soil-grown lettuce. Human NoV is highly stable in the environment, and there is potential for irrigation water to become contaminated with the virus and then be distributed in soils and on crops, which may lead to viral internalization in fresh produce. Internalized viruses would be protected from all surface decontamination practices and hence would pose a significant risk to public health. Since climate change is leading to more severe weather events, understanding the effect of these stresses on the internalization of enteric viruses in produce can help to predict and prevent future outbreaks.