Real-Time Reverse-Transcriptase–Polymerase Chain Reaction for Salmonella enterica Detection from Jalapeño and Serrano Peppers

Abstract

Outbreaks of Salmonella linked to fresh produce emphasize the need for rapid detection methods to curb the spread of foodborne pathogens. Reverse-transcriptase–polymerase chain reaction (RT-PCR) detects the presence of mRNA (shorter half-life than DNA), with greater potential of detecting viable pathogens. Real-time RT-PCR eliminates the need for gel electrophoresis and significantly enhances the speed of detection (<1 day) compared with traditional methods (>5 days). The objectives of this research were to apply real-time SYBR Green I-based RT-PCR to detect Salmonella from jalapeño and serrano peppers spiked with low and high inocula of Salmonella. Inoculated and uninoculated peppers were rinsed with water and dried under ultraviolet light for 10 min. Approximately 25 g peppers was inoculated with 10⁸ to 10¹ colony forming units (CFU) of Salmonella enterica serovar Typhimurium in a stomacher bag and hand massaged in sterile 0.05 M glycine–0.14 M saline buffer (0.05% Tween, 3% beef extract) for optimal recovery of bacteria. A short preenrichment step of 6 h in buffered peptone water was needed for the detection of low inocula (10⁴ CFU/25 g). One-milliliter portions of the extracts were serially diluted, plated on XLT4 agar, and used for RNA extraction with the Qiagen RNeasy^® Mini Kit. RT-PCR was carried out using SYBR Green I one-step RT-PCR with previously described invA gene primers and an internal amplification control. Detection limits were 10⁴ CFU/25 g (∼10² CFU/g) and 10⁷ CFU/25 g (∼10⁵ CFU/g) Salmonella from enriched and unenriched inoculated peppers, respectively. Even though this method included a 6-h incubation period, the results were still obtainable in 1 day. This method shows promise for applications in routine surveillance and during outbreaks.

Introduction

T he demand for chili peppers has been on the rise over the past 20 years as trends in consumption of various cuisines increase. Two types of chili peppers, jalapeño and serrano originating from a farm in northeastern Mexico, were recently implicated in an outbreak of salmonellosis caused by Salmonella Saintpaul with the implicated source of contamination being irrigation water used on the crops (CDC, 2008). According to the Centers for Disease Control and Prevention, ∼1438 people in 43 United States, the District of Columbia, and Canada became infected with Salmonella Saintpaul linked to the peppers (CDC, 2008). It is reported that in the mid 90s, the United States imported about 200 million pounds of fresh chili peppers, which increased to just under 675 million pounds in recent years (Lucier and Dettmann, 2008). In addition, from 1995 to 2000, these numbers grew by about 188%, increasing even further by an estimated 72% during 2000–2007 (Lucier and Dettmann, 2008). As most of the peppers produced are imported to the United States, safety has become more of a concern because of the various contamination sources and increased security at the farm level.

In recent years, the use of molecular technologies such as polymerase chain reaction (PCR) and real-time PCR (rt-PCR) have enabled the rapid detection of foodborne pathogens to enhance food safety. PCR uses gel electrophoresis to visualize the presence of the target product, whereas rt-PCR uses fluorescence. Previous studies involving rt-PCR or PCR for detection of Salmonella from produce reported detection of 1–10 colony forming units (CFU)/mL and 100 CFU/mL from mixed salad rinse waters (Bhagwat, 2004); 1 CFU/mL, 10 CFU/mL, and 100 CFU/mL from leaf lettuce and apples (Shearer et al., 2001); and 1 CFU/mL from mixed salad, cilantro, broccoli, cauliflower, and cabbage (Bhagwat, 2003). Much of these lower limits of detection were aided by a preenrichment process of around 24 h. Incubating poultry meat for 24 h allowed a detection limit of 10³ CFU/mL by PCR (Kanki et al., 2009; De Medici et al., 2003). A detection limit of 10³ CFU/g was demonstrated for buffalo meat trimmings using an 18-h incubation period (Biswas et al., 2008). Using rt-PCR, a detection limit of less than 5 CFU/25 g was shown for alfalfa sprouts, cantaloupe, mixed salad, and cilantro when incubated for 24 h (Liming and Bhagwat, 2004).

However, these DNA-based methods do not distinguish between live and dead cells. Reverse transcriptase–PCR (RT-PCR) is a potential method for detecting viable bacteria by targeting mRNA that has a shorter half-life than DNA (Maurer, 2006). Several researchers have investigated the suitability of using invA mRNA for the determination of Salmonella viability (Fey et al., 2004; Jacobsen and Holben, 2007; Gonzalez-Escalona et al., 2009). Real-time RT-PCR (rt-RT-PCR) could allow for the detection of recent contamination rather than cells that have already been killed because of implemented control measures to eliminate bacteria from the food product or environment. Rapid detection by rt-RT-PCR can help the producer recall contaminated fresh produce from the market at a much faster rate than traditional methods that can take ≥5 days.

The purpose of this study was to apply a previously described SYBR Green I-based one-step RT-PCR assay for the detection of Salmonella from jalapeño and serrano peppers within 1 day. Previously reported invA primers that target the chromosomally located invA gene (Galan and Curtiss, 1989) specific for Salmonella species were used in this one-step assay (D'Souza et al., 2009). Peppers were inoculated with high (10⁸–10⁴ CFU/mL) and low (10³–10¹ CFU/mL) inocula and, for the latter, short enrichment times of 6 h were applied.

Materials and Methods

Growth of Salmonella enterica serovar Typhimurium

Salmonella enterica serovar Typhimurium (DT104 strain 2486) was grown in 10 mL of trypticase soy broth (Difco–Becton Dickinson, Sparks, MD) at 35°C for 24 h. The cultures were transferred for two consecutive times at 24-h intervals before use. A 10-fold serial dilution of an overnight culture (1–9 mL) in peptone buffer (Oxoid, Basingstoke Hampshire, England) per tube, was spread plated on trypticase soy agar (Difco–Accumedia, Lansing, MI) and xylose lactose tergitol 4 agar (XLT4), and incubated at 37°C for 24–48 h to determine counts.

Preparation of peppers

Chili peppers (jalapeño and serrano peppers) were purchased at a local grocery store. Before inoculation, the chili peppers were washed for 1 min using tap water, followed by rinsing with sterile deionized water, and placed in an open sterile Petri dish in a BSL-2 hood to dry under ultraviolet light for 10 min at room temperature. The chili peppers (25 g) were inoculated with 0.1 mL Salmonella enterica serovar Typhimurium high inocula (10⁸ to 10⁴) and low inocula (10³ to 10¹) and allowed to dry for 10 min at room temperature under aseptic conditions in a BSL-2 hood. Washed and dried uninoculated peppers were used as controls. Each experiment was run in duplicate and replicated at least twice.

Bacterial elution from peppers

Bacteria were eluted from 25 g peppers by washing with 10 mL of 0.05 M glycine–0.14 M saline buffer (pH 9.0) (Fisher Scientific, Fair Lawn, NJ) and 0.05% Tween-20 (Promega, Madison, WI) to aid in removal of bacterial cells. To improve bacterial recovery, glycine–saline buffer with 3% beef extract and 0.05% Tween-20 was also tested and added to the chili peppers in a sterile stomacher bag. The samples were hand rubbed for 1 min and a 10 mL sample was taken and centrifuged at 8000 g for 10 min at 4°C (Sorvall Legend 23 R; Thermo Electron, Asheville, NC). The supernatant was decanted and the cell pellet was resuspended in 1 mL of 0.1% peptone buffer (Oxoid) before plating and isolation of RNA.

Preenrichment and bacterial enumeration

For low inocula levels, short-enrichment times were needed. Twenty-five grams of peppers (inoculated and uninoculated controls) was placed in 225 mL of sterile buffered peptone water (Oxoid) (Andrews and Hammack, 2007) and incubated for 6 h at 37°C. After the 6-h incubation period, a 10 mL sample was removed, placed in a 15-mL centrifuge tube, and spun at 8000 g for 10 min at 4°C (Sorvall Legend 23 R; Thermo Electron). The supernatant was decanted and the cell pellet was resuspended in 1 mL of 0.1% peptone buffer (Oxoid). Serial dilutions of enriched and unenriched samples were surface plated (0.1 mL) on XLT4 agar and incubated at 37°C for 24–48 h to enumerate typical black colonies for Salmonella detection.

Addition of stressed and autoclaved cells

Portions of 24-h grown Salmonella enterica serovar Typhimurium were also cold stressed at −20°C and 4°C and heat-stressed at ∼45°C for 24 h prior to use for testing the ability of the assay to detect stressed cells. Another portion was heated in an autoclave at 121°C, 15 pounds per square inch (psi), for 15 min. These stressed cells or autoclaved cultures were inoculated on the peppers as described earlier and either directly detected/assayed or preenriched and plated on XLT4 or assayed by real-time RT-PCR.

Isolation of RNA from bacteria

RNAlater Stabilization Reagent (Qiagen, Valencia, CA) (2 mL) was added to 1 mL of overnight Salmonella enterica serovar Typhimurium, enriched and unenriched inoculated pepper samples, uninoculated pepper controls, and peptone water. The combination of stabilization reagent and bacterial culture/extract was incubated at room temperature (25°C) for 10 min and mixed with a vortex for several times. The cells were centrifuged (5417C; Eppendorf, New York, NY) for 10 min at 8000 g at 4°C and the pellet was stored at −80°C or used directly for extraction.

The RNeasy Mini Purification Kit (Qiagen) was used to extract RNA from all samples described above and the manufacturer's instructions were followed. Briefly, 100 μL of Tris-EDTA (10 mM Tris-Cl, 1 mM ethylenediaminetetraacetic acid (EDTA), pH 8.0; Promega) containing 3 mg/mL lysozyme (Sigma–Aldrich, St. Louis, MO) was added to each pellet, mixed with a vortex, and incubated at room temperature for 10 min with mixing. RLT buffer (350 μL) was added, followed by 250 μL of 100% ethanol (Aaper Alcohol, Shelbyville, KY), and transferred to the RNeasy mini column. The column was centrifuged at 8000 g for 15 s and the collection tube was discarded. RW1 buffer (700 μL) was added to the column and centrifuged for 15 s at 8000 g to wash the column. RPE buffer (500 μL) was added and centrifuged at 8000 g for 15 s and repeated once. The RNA was then eluted off the column with 40 μL of RNase-free water and centrifuged at 8000 g for 1 min. The collection tubes containing RNA were capped and stored at −80°C until further analyses.

DNase I treatment

DNase I treatment was carried out using the TURBO DNA-free® Kit (Ambion^®; Applied Biosystems, Austin, TX). The procedure consisted of 5 μL of 10 × TURBO DNase buffer, 1 μL of rDNase I DNA-free, and 34 μL of nuclease-free water. The reagents were then mixed with 10 μL of RNA extracts and incubated at 37°C for 30 min. Five microliters of DNase Inactivation Reagent was added and incubated for 2 min. The sample was centrifuged at 10,000 g for 2 min. The supernatant was removed and placed in a new tube.

Nanodrop quantification of RNA

After extraction of RNA (before and after DNAse I treatment), RNA quantification was carried out using a NanoDrop^® Spectrophotometer (NanoDrop ND-1000 Spectrophotometer; Thermo Scientific, Wilmington, DE). Nucleic acid concentration was determined using ultraviolet spectrophotometry at 260 nm and purity was determined using the absorbance ratios of 260/280 and 260/230 to detect protein and organic carryover, respectively. Samples with both ratios ≥1.8 were used to provide a good quality of RNA with little to no carryover.

rt-RT-PCR procedure

The SuperScript^TM III Platinum^® SYBR^® Green One-Step qRT-PCR Kit (Invitrogen, Carlsbad, CA) was used in the amplification of mRNA to cDNA, followed by DNA amplification as described by D'Souza et al. (2009). Briefly, 25 μL reactions contained 5 μL RNA extracts in RNAse–DNAse free water, Superscript III (SSIII) one-step RT-PCR kit reagents (Invitrogen) with SYBR Green I, 0.02 μM of each invA primer (F: 5′CACGCTCTTTCGTCTGGCA3′; R2: 5′TACGGTTCCTTTGACGGTGCGA3′), and 1.6 μL bovine serum albumin (Sigma–GenoSys, St. Louis, MO) and an internal amplification control (IAC) (see below). The thermocycler conditions were as follows: RT at 50°C/30 min, denaturation at 95°C/5 min, followed by 45 cycles at 95°C/45 s, 58°C/45 s, 72°C/45 s, and final extension at 72°C/7 min in an iCycler (Bio-Rad, Hercules, CA). After amplification, melt temperature (T _m) analysis from 50°C to 95°C with 0.5°C increments was carried out. The software provided by iCycler was used to obtain threshold cycle values, T _m, and the standard curve. The expected T _m was ∼86°C for the Salmonella-specific invA product and 82°C for the IAC product.

Agarose gel electrophoresis

Products were separated by agarose gel electrophoresis (2% agarose gels) (in 1 × Tris-acetate–EDTA buffer; Fisher Scientific, Pittsburgh, PA) and visualized under ultraviolet light after staining with ethidium bromide using the Gel Doc System/Station (Bio-Rad). A 100-bp DNA marker (Promega) was also run to confirm the presence of the 347-bp target amplified product and the 154-bp IAC product.

Preparation of the IAC

The IAC was added to the SuperScript III Platinum SYBR Green One-Step qRT-PCR mix to eliminate the possibility of false-negatives due to PCR inhibitors, enzyme inactivation, or instrument malfunction as described earlier (D'Souza et al., 2009). The concentration of the amplified IAC RNA product was diluted to reach an optimal concentration of 1.9 fg/μL in the rt-RT-PCR mix.

Results

Bacterial cell recovery from peppers

Extracting Salmonella from the pepper skin proved to be much easier than removal from the stem of the pepper. The hard, smooth surface of the pepper without any crevices left no room for trapping target bacteria (Rusin et al., 2002). Recovery of Salmonella enterica serovar Typhimurium cells, as determined by plating from the surface of the jalapeño and serrano peppers, was about 10³ CFU from the initial inocula of 10⁷ CFU/25 g using the optimal glycine–saline buffer containing Tween-20 and beef extract (Table 1). The addition of Tween-20 acted as a surfactant to help lift the cells off the outer surface of peppers and into suspension. When a lower inoculum of 10⁴ CFU/25 g was used, recovery remained at ∼1.08 × 10³ CFU/25 g after enrichment as determined by plating on XLT4 agar and observance of typical black colonies when compared with Salmonella controls.

Table 1.

Recovery of Salmonella enterica Serovar Typhimurium Without Enrichment from Peppers by Plating on XLT4 Agar and Nanodrop Readings of RNA Extracts

Buffer	Level of inocula log count	Cells recovered by plating on XLT4	RNA extracted (ng/μL)
Glycine–saline	10⁷	(5.5 ± 0.7) × 10¹	7.8 ± 0.7
Glycine–saline with 0.05% Tween-20	10⁷	(8.20 ± 11) × 10²	7.0 ± 3.18
Glycine–saline with 0.1% Tween-20	10⁷	(7.5 ± 3) × 10²	4.15 ± 0.07
Glycine–saline with 0.05% Tween-20 and 3% beef extract	10⁷	(6.45 ± 16) × 10³	5.85 ± 0.77

RNA extraction yields from peppers

For detection of Salmonella without enrichment, the 10⁷ CFU inoculum yielded an average of 6 ng/μL nucleic acid from both types of peppers. Purity of the RNA extracts was assessed, which showed little to no protein or organic carryover using the ratios of A260/A280 and A260/A230 that were >1.8. After an enrichment process, with an improved 10⁴ detection limit, 263.34 ± 18.39 ng/μL of RNA was extracted with little or no protein or organic carryover.

rt-RT-PCR detection and specificity

Detection of 10⁷ CFU/25 g from both inoculated jalapeño and serrano peppers without enrichment was achieved using RT-PCR (Fig. 1A). After 6-h preenrichment, an increase in detection by 3 logs to 10⁴ CFU/25 g (Fig. 2A) for both jalapeño and serrano peppers was obtained. Even with preenrichment, this method could still yield results in 1 day. To determine the specificity of the amplified products, T _m analysis was carried out. T _m curves showed peaks at ∼86°C for the target-amplified invA product and the IAC showed peaks at 82°C. This showed that false-positives were not obtained and that indeed the detection limit was 10⁴ CFU/25 g (Figs. 1B and 2B). Also, the possibility of false-negatives was eliminated, because the IAC peak at 82°C showed that no inhibitors were present and that there was no equipment malfunction or enzyme failure. Further confirmation by gel electrophoresis revealed the 347-bp target product as expected, with similar detection limit to RT-PCR Ct values of 10⁴ CFU/25 g. Gel electrophoresis is generally not required during routine analysis using rt-PCR, but was carried out during this research to confirm the presence of the target products of appropriate size. Negative controls did not show any T _m at ∼86°C or any product at 347 bp as expected. Negative controls consisted of water, trypticase soy broth, peptone, or uninoculated peppers, which did show the T _m at 82°C for the IAC and the 154-bp IAC product by agarose gel electrophoresis (Figs. 1C and 2C). Again, this showed that false-negatives were not obtained.

FIG. 1.

(A) Detection limits of Salmonella enterica showing cycle threshold values (Ct) from spiked jalapeno peppers by real-time reverse-transcriptase–polymerase chain reaction (rt-RT-PCR) without enrichment. 1: Positive control of S. Typhimurium RNA extracts; 2: peppers inoculated with 10⁸ CFU/25 g; 3: peppers inoculated with 10⁷ CFU/25 g; 4: uninoculated pepper control; and 5: negative control (water). (B) Melt temperature curve depicting specific S. enterica amplification products at ∼86°C with detection limits of 10⁷ CFU/25 g without enrichment from jalapeno peppers. IAC shows peaks at 82°C for the negative controls. (C) Agarose gel electrophoresis of rt-RT-PCR amplified products (347 bp) from jalapeno peppers spiked with Salmonella enterica serovar Typhimurium without enrichment. Lane M: 100-bp ladder marker; lanes 1–6: products from jalapeno peppers spiked with 10⁸–10³ CFU/25 g of Salmonella enterica serovar Typhimurium; lane 7: uninoculated pepper control; lane 8: positive control of Salmonella Typhimurium RNA extracts; lane 9: negative control (water) with IAC at 154 bp.

FIG. 2.

(A) Detection limits of S. enterica showing cycle threshold values (Ct) from spiked jalapeño peppers by rt-RT-PCR after enrichment. 1: Positive control; 2–4: peppers inoculated with 10⁶–10⁴ CFU/25 g Salmonella; 5: negative control (water); and 6: peppers inoculated with 10³ CFU/25 g Salmonella. (B) Melt temperature curves depicting specific S. enterica amplification products at ∼86°C, with detection limits of 10⁴ CFU/25 g after enrichment from jalapeno peppers; internal amplification control shows peak at 82°C. (C) Agarose gel electrophoresis of rt-RT-PCR amplified products (347 bp) from jalapeno peppers spiked with Salmonella enterica serovar Typhimurium after enrichment. Lane M: 100-bp ladder marker; lanes 1–8: products from jalapeno peppers spiked with 10⁸–10¹ CFU/25 g of Salmonella enterica serovar Typhimurium; lane 9: uninoculated pepper control; lane 12: positive control of RNA extracted from Salmonella Typhimurium; lane 13: negative control (water).

Detection of stressed cells

Cold-stressed Salmonella enterica serovar Typhimurium cells at 4°C and −20°C as well as heat-stressed cells at 45°C overnight (at high and low inoculum levels) from both peppers showed similar detection to nonstressed cells using high and low inocula levels. Preenrichment for 6 h aided in the recovery and detection of the low inoculums of stressed cells from peppers in a manner similar to nonstressed cells (data not shown). Peppers inoculated with autoclaved cells did not result in detection by this real-time RT-PCR method.

DNase I treatment

DNase I treatment was performed to determine that amplification was from RNA without any carryover of DNA. When traditional PCR on the DNase I-treated extracts using the described invA gene primers was carried out, all treated samples did not result in any positive PCR products as expected, demonstrating the absence of any DNA carryover. However, the ability to detect the target RNA from jalapeño and serrano peppers was reduced by 1 log in unenriched peppers and also by 1 log on enriched peppers (data not shown).

Discussion

Salmonella from jalapeño and serrano peppers were recovered by gently massaging the outer surface of the pepper in an optimal glycine–saline buffer solution containing 3% beef extract and 0.05% Tween-20. Previous research suggested that the attachment of bacterial cells to the surface of produce could be due to the roughness or hydrophobic properties of the produce surface (Iturriaga et al., 2003). This study used a buffer solution that contained a surfactant to aid in removal of the Salmonella cells from the surface of peppers to improve recoveries. The similar amounts of RNA recovered by all tested buffers could be attributed to the lower end detection limit of the instrument, even though increased recovery in bacterial count was obtained. Further optimization of elution buffers is needed to improve recovery of bacteria from produce surfaces. Several factors may have played a role in the low recovery, as attachment of bacteria to produce surfaces depends on the serovar of Salmonella (Shi et al., 2007), as some microorganisms bind or attach more strongly than others. In addition, depending on the drying time on the surface of produce, cell injury or death may have occurred (Lang et al., 2004). Preenrichment or enrichment steps can help aid in the recovery/resuscitation of these injured cells as well as increase cell numbers to improve detection probability. Therefore, this study used short preenrichment times of 6 h to overcome these limitations of recovery and improve detection of target bacteria to obtain rapid and sensitive detection within two 8-h working shifts.

Several studies have shown that an 18–24-h enrichment procedure is necessary to determine the presence of low numbers of bacteria (Perelle et al., 2004; Cheng et al., 2008). An 18-h enrichment process was needed to obtain a detection limit of 10³ CFU/mL by PCR for Salmonella in raw milk (Perelle et al., 2004). Cheng et al. (2008) showed that a 24-h enrichment process prior to RT-PCR was needed to obtain a detection limit of 10³ CFU/mL for Salmonella in chili powder. Bhagwat (2004) showed a lower Salmonella detection limit of 100 CFU/mL from mixed salad rinse after enrichment for 5 h. Others showed detection of 1 CFU/mL from leaf lettuce and apples (Shearer et al., 2001) for Salmonella after a 24-h enrichment; and 1 CFU/mL from mixed salad, cilantro, broccoli, cauliflower, and cabbage for Salmonella enterica serovar Typhimurium using rt-PCR (Bhagwat, 2003), after a 20-h enrichment. In this study, a shorter 6-h enrichment process allowed the detection of lower inoculum levels at ∼10⁴ CFU/25 g within two 8-h working shifts.

Although the major advantage to rt-PCR is the speed of detection (Feng, 2007), both PCR and rt-PCR target DNA that does not distinguish between dead and viable cells. An advantage of using RT-PCR is that it targets mRNA with a shorter half life than DNA, and therefore, has greater potential for detecting primarily viable cells that could cause infection (Maurer, 2006). As most quality control analyses are done on finished products, this is an excellent way to determine if the critical control points are working and that no viable cells are present.

Studies have been carried out to determine the suitability of using RNA to detect the viability of cells. Researchers have shown that invA mRNA from Salmonella enterica serovar Typhimurium can degrade rapidly after 50 h in drinking and pond water (Fey et al., 2004). These researchers quantified S. Typhimurium mRNA in drinking water using invA and 16s RNA genes that detected 2–20 and 100–1000 copies, respectively. Real-time RT-PCR with invA gene primers showed a detection limit of 5 × 10⁴ seeded S. Typhimurium cells per gram of soil (Jacobsen and Holben, 2007), and pasteurizing the samples at 65°C for 30 min resulted in the inability to detect invA mRNA. These results indicate that invA mRNA appears to be a suitable target for cell viability determination for Salmonella.

Recently, using the same target invA mRNA, a TaqMan-based RT-PCR method has shown a detection limit of 5 CFU/25 g for Salmonella Saintpaul and Salmonella Newport from spinach, tomatoes, and jalapeno and serrano peppers; however, a longer preenrichment of 24 ± 2 h was used (Gonzalez-Escalona et al., 2009). Even though our study is similar as it is based on the invA mRNA for RT-PCR detection, it differs from the work described by these researchers in that the less-expensive SYBR Green I fluorescent dye was used in the real-time assay when compared with TaqMan^® fluorescent probes. Also, our study successfully used a shorter preenrichment time of 6 h for Salmonella enterica serovar Typhimurium detection from jalapeno and serrano peppers.

An IAC was applied to determine if there were any false-negatives that may have been caused by PCR inhibitors (Moreira et al., 2009), enzyme failure, or instrument malfunction. PCR inhibitors from food matrices include lipids, proteins, enzymes, chemical additives, fiber, other bacteria, and acids (Lee and Fairchild, 2006). The target and IAC were both amplified using the same primer sets to enable the use of only one set of PCR reaction conditions as has been suggested by other researchers (Hoorfar et al., 2004).

A DNAse I treatment was performed to eliminate genomic DNA that may have been present (Werbrouck et al., 2007). Harper et al. (2003) suggested that all genomic DNA may not be completely removed by the treatment process. In this study, traditional PCR run on DNAse I-treated samples did not show any product on agarose gels, indicating that DNA was indeed eliminated (data not shown). Importance must be placed on removal of any carryover DNA that could lead to erroneous results. The lower yield of RT-PCR after DNAse I treatment showed that caution must be used while extracting RNA. It is also crucial to ensure that the DNAse I does not contain any RNAses that could destroy or degrade the total RNA. The RNAses, if any present, could result in decreased yield (Werbrouck et al., 2007). When DNAse I was applied to our study, a 1 log reduction in detection was observed. It is possible that the loss of 1 log could be due to the DNAse I containing some carryover RNAses. This could have resulted in the degradation of some of the RNA present, leading to lower yields.

In nonhost environments, Salmonella more than likely persists in starved or stressed states (Gonzalez-Escalona et al., 2009); therefore, stressed cells were used to inoculate the peppers for detection by this newly developed SYBR Green I-based assay. Salmonella enterica serovar Typhimurium cells that were cold-stressed at 4°C and − 20°C as well as heat-stressed at ∼45°C overnight (at high and low inoculum levels) were used to inoculate peppers and also to simulate conditions during transport or storage of peppers. Preenrichment for 6 h aided in the recovery and detection of these stressed cells, yielding similar detection limits compared with nonstressed cells. Also, previous studies using inoculation of a mixture of microorganisms (Escherichia coli, Enterococcus faecalis, Lactobacillus, etc.) associated with background flora on food commodities followed by enrichment did not show any cross-reactivity by this rt-RT-PCR assay (Techathuvanan et al., 2009a). This assay also did not show cross-reactivity against a panel of various non-Salmonella bacteria in pure culture as previously reported (D'Souza et al., 2009). Amplified products were not obtained/detected either by T _m analysis or gel electrophoresis from the extracts of the background flora (either pure culture or when inoculated on food matrices), indicating the specificity of this assay.

Detection limits of around 400 CFU/g (10⁴ CFU/25 g) after a 6-h enrichment step using RT-PCR was similar to results obtained by other researchers using molecular detection methods for produce. Yuk et al. (2006) evaluated Salmonella on the surface of tomatoes with a detection limit of 10² CFU/tomato, with an enrichment period of 4 h. Even though Yuk et al. (2006) used only a 4-h enrichment process, the additional cost of using the flow-through immunocapture system may be a limiting factor in using this method. This newly developed rt-RT-PCR assay requires approximately similar time for detection when compared with nucleic acid sequence-based amplification assays (D'Souza and Jaykus, 2003) or novel loop-mediated isothermal assays (Techathuvanan et al., 2009b) to obtain similar results. However, as suggested by Gonzalez-Escalona et al. (2009), collaborative studies to determine interlaboratory reproducibility are needed before the method can be used for routine diagnostic purposes, without excluding the need for the physical isolation of bacteria for regulatory purposes.

In conclusion, this method rapidly detected Salmonella within two 8-h workshifts and was much faster than traditional plating. Rapid Salmonella detection methods will ultimately save the food industry time and money by avoiding costly recalls and protecting public health. Future work should focus on improving bacterial recovery, RNA yields, and purity to increase detection sensitivity to ≤1 log CFU/g of peppers within 1 work day or at the most, two workshifts, to allow for timely product recall or prevention of release of contaminated products for sale.

Footnotes

Acknowledgments

The authors gratefully acknowledge the support of the Tennessee Agricultural Experiment Station for this research. This research was carried out by Nathan Miller as part of the requirements for the fulfillment of his Master's Degree. The authors are grateful to Dr. David A. Golden for critically reviewing and editing this manuscript. The use of trade names in this manuscript does not imply endorsement by the University of Tennessee nor criticism of similar ones not mentioned.

Disclosure Statement

No competing financial interests exist.

References

Andrews

, Hammack

. Salmonella. Bacteriological Analytical Manual Online. Jackson

. Gaithersburg, MD: Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration www.cfsan.fda.gov/∼ebam/bam-5.html December 2007.

Bhagwat

. Simultaneous detection of Escherichia coli O157: H7, Listeria monocytogenes and Salmonella strains by real-time PCR. Int J Food Microbiol, 2003; 84:217–224.

Bhagwat

. Rapid detection of Salmonella from vegetable rinse-water using real-time PCR. Food Microbiol, 2004; 21:73–78.

Biswas

, Kondaiah

, Bheilegaonkar

et al. Microbial profiles of frozen trimmings and silver sides prepared at Indian buffalo meat packing plants. Meat Sci, 2008; 80:418–422.

[CDC] Centers for Disease Control and Prevention. Investigation of Outbreak of Infections Caused by Salmonella Saintpaul. CDC: Atlanta, GA, 2008.

Cheng

, Lin

, Van

et al. Rapid detection of Salmonella in foods using real-time PCR. J Food Protect, 2008; 71:2436–2441.

De Medici

, Croci

, Delibato

et al. Evaluation of DNA extraction methods for use in combination with SYBR Green I real-time PCR to detect Salmonella enterica serotype enteritidis in poultry. Appl Environ Microbiol, 2003; 69:3456–3461.

D'Souza

, Critzer

, Golden

. Real-time reverse-transcriptase-PCR (RT-PCR) for the rapid detection of Salmonella using invA primers. Foodborne Pathog Dis, 2009; 6:1–10.

D'Souza

, Jaykus

. Nucleic acid sequence based amplification for the rapid and sensitive detection of Salmonella enterica from foods. J Appl Microbiol, 2003; 95:1343–1350.

10.

Feng

. Rapid methods for the detection of foodborne pathogens: current and next-generation technologies. Food Microbiology: Fundamentals and Frontiers, 3 ^rd. Doyle

, Beuchat

. Washington, DC: ASM Press, 2007; 911–934.

11.

Fey

, Eichler

, Flavier

et al. Establishment of a real-time PCR-based approach for accurate quantification of bacterial RNA targets in water, using Salmonella as a model organism. Appl Environ Microbiol, 2004; 70:3618–3623.

12.

Galan

, Curtiss

. Cloning and molecular characterization of genes whose products allow Salmonella Typhimurium to penetrate tissue-culture cells. Proc Natl Acad Sci USA, 1989; 86:6383–6387.

13.

Gonzalez-Escalona

, Hammack

, Russell

et al. Detection of live Salmonella sp. cells in produce by a TaqMan-based quantitative reverse-transcriptase real-time PCR targeting invA mRNA. Appl Environ Microbiol, 2009; 75:3714–3720.

14.

Harper

, Hilton

, Jones

. RT-PCR for the pseudogene-free amplification of the glyceraldehyde-3-phosphate dehydrogenase gene (gapd) Mol Cell Probes, 2003; 17:261–265.

15.

Hoorfar

, Malorny

, Abdulmawjood

et al. Practical considerations in design of internal amplification controls for diagnostic PCR assays. J Clin Microbiol, 2004; 42:1863–1868.

16.

Iturriaga

, Escartin

, Beuchat

, Martinez-Peniche

. Effect of inoculum size, relative humidity, storage temperature, and ripening stage on the attachment of Salmonella Montevideo to tomatoes and tomatillos. J Food Protect, 2003; 66:1756–1761.

17.

Jacobsen

, Holben

. Quantification of mRNA in Salmonella sp. Seeded soil and chicken manure using magnetic capture hybridization RT-PCR. J Microbiol Meth, 2007; 69:315–321.

18.

Kanki

, Sakata

, Taguchi

et al. Effect of sample preparation and bacterial concentration on Salmonella enterica detection in poultry meat using culture methods and PCR assaying of preenrichment broths. Food Microbiol, 2009; 26:1–3.

19.

Lang

, Harris

, Beuchat

. Evaluation of inoculation method and inoculum drying time for their effects on survival and efficiency of recovery of Escherichia coli O157:H7, Salmonella, and Listeria monocytogenes inoculated on the surface of tomatoes. J Food Protect, 2004; 67:732–741.

20.

Lee

, Fairchild

. Sample preparation for PCR. PCR Methods in Foods. Maurer

. New York: Springer Science +Business Media, Inc., 2006; 41–50.

21.

Liming

, Bhagwat

. Application of a molecular beacon—real-time PCR technology to detect Salmonella species contaminating fruits and vegetables. Int J Food Microbiol, 2004; 95:177–187.

22.

Lucier

, Dettmann

. Potato acreage down, prices up. Vegetables and Melons Outlook. Economic Research Service/USDA: Washington, DC, 2008.

23.

Maurer

. The mythology of PCR: a warning to the wise. PCR Methods in Foods. Maurer

. New York: Springer Science + Business Media, Inc., 2006; 27–40.

24.

Moreira

, Conceicao

RDS

et al. IMS using in-house monoclonal antibody-coated magnetic beads associated to PCR assay for detection of Salmonella Typhimurium in raw meats. J Food Safety, 2009; 29:59–72.

25.

Perelle

, Dilasser

, Malorny

et al. Comparison of PCR-ELISA and LightCycler real-time PCR assays for detecting Salmonella spp. in milk and meat samples. Mol Cell Probes, 2004; 18:409–420.

26.

Rusin

, Maxwell

, Gerba

. Comparative surface-to-hand and fingertip-to-mouth transfer efficiency of gram-positive bacteria, gram-negative bacteria, and phage. J Appl Microbiol, 2002; 93:585–592.

27.

Shearer

AEH

, Strapp

, Joerger

. Evaluation of a polymerase chain reaction-based system for detection of Salmonella enteritidis, Escherichia coli O157: H7, Listeria spp., and Listeria monocytogenes on fresh fruits and vegetables. J Food Protect, 2001; 64:788–795.

28.

Shi

, Namvar

, Kostrzynska

et al. Persistence and growth of different Salmonella serovars on pre- and postharvest tomatoes. J Food Protect, 2007; 70:2725–2731.

29.

Techathuvanan

, Draughon

, D'Souza

. Real-time reverse-transcriptase-PCR for the rapid and sensitive detection of Salmonella from pork. Presented at the Annual IFT Meeting: Anaheim, CA, 2009aposter no. P-152-37.

30.

Techathuvanan

, Draughon

, D'Souza

. Reverse-transcriptase loop-mediated isothermal amplification for the rapid and sensitive detection of Salmonella from pork. Presented at the Annual ASM Meeting: Philadelphia, PA, May 18, 2009bposter no. P-065.

31.

Werbrouck

, Botteldoorn

, Uyttendaele

et al. Quantification of gene expression of Listeria monocytogenes by real-time reverse transcription PCR: optimization, evaluation and pitfalls. J Microbiol Methods, 2007; 69:306–314.

32.

Yuk

, Warren

, Schneider

. Preliminary evaluation of flow-through immunocapture followed by real-time PCR for the detection of Salmonella serovars on tomato surfaces within 8 hours. J Food Protect, 2006; 69:2253–2257.