PMA efficiently enters cells with compromised but not intact membranes, and, upon photo activation, PMA is irreversibly cross linked to DNA and prevents PCR amplification of target sequences. In this study, we developed a PMA real-time PCR assay to quantify both viable and total E. coli O157:H7 recovered from lettuce plants in the laboratory and field; these results were compared with quantities of this organism determined by culture methods. We also used this approach to design plant inoculation and incubation conditions in the laboratory that more closely mimic field environments and could be used to study the plant associated behaviors of virulent outbreak-associated strains of the pathogen.Seeds of Romaine lettuce cv. Parris Island were planted in Sunshine mix potting soil and grown in an environmental chamber with a light intensity of 230 mmol m22 s 22 . The chamber was maintained at a constant relative humidity and daily temperatures of 22uC and 18uC . Starting 2 weeks after emergence, the plants were fertilized weekly with Hoagland nutrient water. Four weeks after plant emergence, 10 2-ml drops of E. coli O157:H7 ATCC 700728 or EC4045 were inoculated onto the leaves. The bacterial suspension was dispensed with an Eppendorf repeater pipette onto the interveinal spaces of two leaves per plant to yield approximately 106 CFU/leaf. E. coli O157:H7 ATCC 700728 was also inoculated onto lettuce plants using a spray bottle as described previously for field studies. Spray bottles were calibrated to deliver a dose adjusted to 106 CFU of E. coli O157:H7 ATCC700728 per spray, and a single spray was applied to each plant from a distance of 20 cm and with a dispersal area of approximately 0.3 m2 . Immediately after inoculation, plants were transferred to a growth chamber maintained with a 12-h photoperiod , constant relative humidity , and temperatures of 18uC and 23uC . Within 24 h after application, free water was no longer visible on the surface of the plants. Cross contamination of strains EC4045 and ATCC 700728 was prevented by inoculating one strain per plant and keeping plants spatially separated. Five leaves or five plants were collected for measurement of E. coli O157:H7 amounts at the time of inoculation and then 1, 2 and 7 days after application of the pathogen onto the plants.Field trials were conducted in the Salinas Valley during spring 2010 and summer 2010 . Permits and approvals for use of United States-owned land were granted by the United States Department of Agriculture. All beds were direct seeded according to standard commercial practice with two rows of Romaine lettuce cv. Green Towers per bed. The field trials were conducted exactly as described in Moyne et al.on 4-week old lettuce plants after thinning. Inoculation was carried out in the morning by delivering one spray per 4-week-old lettuce plant just after thinning. For E. coli O157:H7 enumeration on lettuce,vertical grow shelf entire lettuce heads were removed by cutting at the base with a sterile scalpel approximately 3 cm above the ground; each head was bagged individually at the time of collection.

Twelve lettuce heads were sampled immediately after inoculation, and 24 lettuce plants were then randomly sampled among the inoculated beds at 2 h and 2 and 7 days.The lettuce was brought to the laboratory from the field in a cooler on ice and analyzed within 24 h.One leaf or the entire lettuce head was placed into a sterile Whirl-Pak bag with 50 ml of 0.1% peptone water. The lettuce was homogenized in a Stomacher 400 laboratory blender at medium speed for 2 min. A total of 500 ml of the resulting cell suspension was used for PMA real time PCR and 1 ml was serial diluted for CFU enumerations on TSA supplemented with 50 mg/ml rifampicin. To improve the limit of detection of culturable cells, the remaining cell suspension was filtered through disposable analytical filter units . Filter membranes were then laid onto CHROMagar O157 and incubated at 37uC overnight. The detection limit by this approach was 1 CFU per leaf or plant in the laboratory experiments and 10 CFU per plant in the field experiments. In addition to plating for CFU enumeration, the lettuce heads were mixed at a ratio of 1:2 with TSB supplemented with rifampicin at 50 mg/mland incubated at 42uC for 18 h. When CFU estimates were low, the enrichment broth culture was spiral plated onto CHROMagar O157 supplemented rifampicin at 50 mg/ml and incubated at 37uC overnight. The enrichment was considered positive if mauve colonies typical of E. coli O157:H7 on this medium were observed.PMA was dissolved in water to obtain a 20 mM stock solution and stored at 220uC in the dark. For comparisons of different PMA concentrations, 40, 50, or 100 mM PMA was added to 500 ml cells of E. coli O157:H7 in clear transparent 2-ml micro-tubes and incubated, with shaking at 400 rpm, at room temperature in the dark for 15, 30, or 60 min. After incubation with PMA, samples were exposed to a 500W halogen light at a distance of 20 cm for 3 min with occasional mixing. To avoid overheating, the cultures were kept on ice during light exposure. Assessment of PMA for distinguishing live and dead cells were performed by incubating one half of a 12-h E. coli O157:H7 culture in 0.1% peptone and suspending the other cell fraction in 70% isopropanol for 10 min . Both live and dead fractions were then washed twice in 0.1% peptone and suspended in 0.1% peptone for exposure to PMA. DNA was extracted from cells recovered from laboratory cultures and lettuce with or without prior exposure to PMA. First, 500 ml of the cell suspension was concentrated by centrifugation at 10,0006g for 2 min. The resulting cell pellet was then washed once with 0.1% peptone, suspended in 100 ml of Prepman solution , and boiled for 5 min. Debris was removed by centrifugation at 10,0006g for 2 min, and 1 ml of the supernatant was used for real-time PCR. High quality E. coli O157:H7 genomic DNA for use in the real-time PCR assay development was obtained using the Wizard Genomic Purification kit as described by the manufacturer. Real-time PCR was performed on an ABI 7500 Fast Real-Time PCR system using 0.25 mM of forward and reverse primers, 1X Ssofast Evagreen Supermix and 1 ml of prepared DNA template. Cycling conditions included an initial activation step at 95uC for 10 s,vertical hydroponic followed by 40 cycles of denaturation at 95uC for 5 s, and annealing/extension temperatures between 60 to 68uC for 30 s for the primers shown in Table S1, supporting information.

After 40 PCR cycles, melting curves were generated by increasing the temperature from 60 to 95uC at 0.2uC/10 s and recording the fluorescence. Threshold cycle values were automatically generated by the 7500 Fast Real-Time PCR software. To quantify the number of E. coli O157:H7 cells in the samples, a standard curve was included for each plate with three replicates for each DNA concentration. Standard curves were generated by amplifying a 5-log DNA dilution series from a known concentration of E. coli O157:H7 cells.The stx1 or stx2 genes coding for Shiga-toxin 1 and 2 are typically used as genetic targets for detection or quantification of E. coli O157:H7. Because these genes are absent from the nontoxigenic strain ATCC 700728 and a genome sequence for this strain was not available, real-time PCR primers designed for other genes in E. coli O157:H7 were compared including enterocyte effacement loci genes csgA , eae , espA , fliC , ler , and lpfA. Our goal was to develop a method capable of detecting small amounts of E. coli O157:H7 ATCC 700728 cells among high numbers of endogenous bacteria present on lettuce leaves . Therefore, primers were selected based on minimum Ct for amplifying 1 ng of E. coli O157:H7 ATCC 700728 genomic DNA and maximum Ct for amplifying 1 ng of DNA isolated from the closely related bacterial strain E. coli K12. These comparisons were performed because non-pathogenic E. coli were detected on some of the lettuce heads sampled in previous field studies,vertial hydroponic albeit in much lower amounts than evaluated in this assay. E. coli K12 DNA was amplified in low quantities with all primer sets when the annealing temperature was less than 65uC . Although increasing the annealing temperature reduced the non-specific amplification of E. coli K12 DNA, the sensitivity of E. coli ATCC 700728 DNA detection also decreased and lowered the dynamic range of the assay . Based on these comparisons, we selected primers targeting the lpfA gene and an annealing temperature of 60uC because this combination resulted in the earliest detectable amplification of target E. coli O157:H7 ATCC 700728 genomic DNA and yielded the lowest levels of non-specific amplification of E. coli K12 genomic DNA. The efficiency of the PCR with lpfA primers was between 92 to 100% and the correlation coefficient was between 0.98 and 0.99.To accurately quantify live and dead cells in a mixture, PMA should completely inhibit amplification of DNA from dead cells without affecting the enumeration of live cells. Real-time PCR amplification of nucleic acids from dead cells of E. coli O157:H7 ATCC 700728 was significantly reduced, but not completely inhibited, after treatment with PMA at concentrations of 50 and 100 mM, as indicated by an increase in the Ct values . These concentrations of PMA were high enough to influence the accurate detection of viable cells in the real-time PCR assay by increasing the Ct values for the viable cells. In contrast, PMA at a concentration of 40 mM did not interfere with DNA amplification from viable cells in the real-time PCR assay . Amplification of DNA from dead cells was also significantly inhibited by over 9 Ct values at that concentration of PMA . Increasing the exposure to PMA from 15 to 60 min yielded only modest effects , and therefore, we used an incubation time of 30 min for subsequent studies. To fully investigate the factors that resulted in albeit low, but still measurable, amounts of DNA amplification from dead cells after incubating in PMA, we examined the effects of long versus short PCR products on distinguishing cell viability. In contrast to the 165 bp PCR product resulting from the lfpA primers, PCR amplification was completely inhibited from ATCC 700728 cells using hlyA primers . However, DNA amplification using the hlyA primers was 10-fold less sensitive for detecting low quantities of E. coli O157:H7 cells . The limit of detection for both E. coli O157:H7 ATCC 700728 and EC4045 with the lpfA primers was 100 CFU/ml or 5000 CFU per plant or leaf . Therefore, the lpfA primers were used in further studies for quantifying E. coli O157:H7 ATCC 700728 and EC4045 recovered from plants.Previous studies have shown that PMA combined with real-time PCR is most accurate for viable cell enumeration when viable cells greatly outnumber dead cells [28–30]. In our field studies, E. coli O157:H7 ATCC 700728 populations recovered on laboratory culture media declined drastically shortly after inoculation of 106 cells per lettuce head [14], establishing the potential for high numbers of dead cells that may interfere with quantification of the viable cells remaining on the plants. Indeed, suspensions of 106 dead E. coli 0157:H7 ATCC 700728 cells were partially detected by PMA real-time PCR . In a mixture of 106 dead and 104 viable E. coli O157:H7, the PMA real-time PCR assay accurately estimated viable cell amounts . However, the assay overestimated the numbers of E. coli O157:H7 when 103 viable cells were measured in the presence of 106 dead cells of the same strain .E. coli O157:H7 ATCC 700728 was applied to lettuce plants in the growth chamber either by spray or application of 2-ml drops directly onto the leaves. The latter was used specifically to prevent the hazards associated with aerosolization of the virulent strain EC4045 that can occur during spray application. The inoculum level was adjusted to deliver approximately 106 CFU per plant by spray or per leaf by drop, and after inoculation, the plants were immediately incubated at low relative humidity and moderate daytime/nighttime temperatures of 22 and 18uC. With both inoculation methods, culturable E. coli O157:H7 ATCC 700728 cell populations declined to an average of 2.660.2 log CFU per leaf or plant within the first 24 h after inoculation .