Abstract
The ability of reverse transcriptase PCR (RT-PCR) to detect viable Shiga-toxin-producing Escherichia coli (STEC) was investigated. Four primer sets, each targeting a specific region in the slt-II operon, were evaluated for their stringency and specificity for slt-II mRNA. STEC were evaluated for toxin expression under various conditions, including cell growth phase, growth medium, incubation temperature, and aeration. Following primer optimization, STEC were inoculated into Trypticase soy broth and cooked ground beef enrichments. Cells were harvested and RNA or DNA was extracted at 4, 8, 12, and 24 h. RT-PCR or PCR was conducted, and the products were visualized by gel electrophoresis and by Southern blots. mRNA targets were detected in 12-h cooked ground meat enrichments with an initial inoculum of 1 CFU/g. These results indicate that RT-PCR of E. coli slt-II mRNA is useful for detection of viable STEC in ground beef.
Escherichia coli O157:H7 was first isolated as a human pathogen in 1982 and has since become a well-known agent of food-borne illness. Cases of hemorrhagic colitis, hemolytic uremic syndrome, and death have been reported following the consumption of raw or undercooked ground beef (8). Although most commonly associated with foods of animal origin, enterohemorrhagic E. coli (EHEC) may also be isolated from contaminated drinking water and acidic foods such as mayonnaise, salad dressings, and buttermilk (6, 16, 29). The infectious dose for E. coli O157:H7 and other serotypes of EHEC varies greatly between populations. For susceptible individuals such as the elderly and infants, the infectious dose may be as low as 10 cells (8).
Without an animal model, there is much controversy over which virulence factors of EHEC contribute to human pathogenicity (8). EHEC produce Shiga-like toxins (SLTs) that characterize this group of pathogenic E. coli. Other virulence markers include intimin, hemolysin, and the locus of enterocyte effacement (8). Food-borne illnesses have occurred with isolates that possess all or only a few of these markers (8, 9). Possession of slt genes is a consistent virulence marker (1), but without evidence of human illness, E. coli possessing slt are often referred to as Shiga toxin-producing E. coli (STEC). Despite the fact that EHEC strains containing slt-I and slt-II have been isolated from patients with hemorrhagic colitis, studies have shown that strains possessing only slt-II are more frequently associated with human disease complications (8, 25).
Molecular methods for the detection of E. coli O157:H7 are becoming more widely accepted as an alternative to traditional growth-based tests. The ability to specifically identify the presence of an organism within a fraction of the time required by standard approaches makes molecular tests ideal for use in food systems (2). A number of recent reports have applied molecular tools for specific detection of pathogenic E. coli in food systems (10, 24, 31, 32). Most of these assays have used DNA as the target molecule; however, recent literature has shown that DNA is not a good indicator of viable pathogens due to its persistence following cell death (12, 17, 18). mRNA, however, is quite transient and has been the subject of recent studies highlighting the potential of RT-PCRs that target various transcripts as a means to rapidly detect viable pathogens (12, 15, 28).
The expression of suspected bacterial virulence factors is affected and regulated by various environmental and stress conditions to the organism. Thus, in order to use RT-PCR as a means of specifically detecting bacteria, care must be taken to ensure abundant and dependable expression of target mRNA. Past research investigating the effect of aeration on SLT toxin production has shown that increased aeration results in optimal toxin production (33). Others have examined the production of Shiga toxin as a function of enrichment temperature, allowing a specific range of temperatures to be determined for SLT production (24).
While past research has examined the effects of some conditions on Shiga toxin production by STEC, a more in-depth study combining several growth conditions and their subsequent effect on mRNA expression has not been reported. The need to demonstrate dependable expression, particularly in food enrichments, is necessary to determine if slt mRNA could be used as a potential target for detection of viable STEC in food.
The objective of this research was to optimize an RT-PCR assay targeting slt-II mRNA for the detection of viable STEC in broth and ground beef enrichments. Effects of enrichment conditions on slt-II mRNA expression were determined, followed by detection of slt-II mRNA and DNA in broth and cooked ground beef enrichments.
MATERIALS AND METHODS
Culture maintenance.
E. coli O157:H7 (ATCC 43895; raw hamburger isolate from a hemorrhagic colitis outbreak) was obtained from Douglas Marshall, Department of Food Science and Technology, Mississippi State University. This strain contained both the slt-I and slt-II genes and was used as the primary test strain. Additional SLT-producing strains used included ATCC 43890 and ATCC 35150 (fecal isolates), serotypes O111:H-JB1-95 and O157:H7 (salami outbreaks), and serotypes O26:H11 and O26:H11 3359-70 (Centers for Disease Control [CDC] isolates).
Selected strains of Shigella (ATCC 9207, 9290, 12022, 25875, and 35964; S. boydii, S. sonnei, S. flexneri, S. flexneri, and S. boydii, respectively) were also used in assay and primer specificity tests along with selected gram-negative bacteria (Salmonella enterica serovar Typhi, Serratia marcescens, Proteus vulgaris, and nonpathogenic E. coli). All cultures were maintained as stocks in 30% glycerol at −20°C. Prior to use, each strain was cultivated for at least two consecutive 18-h transfers into tryptic soy broth (TSB) (Weber Scientific, Hamilton, N.J.) at 37°C.
Nucleic acid extraction.
Unless otherwise specified, all reagents used for RNA extraction, DNase treatments, and RT-PCR procedures were purchased from Invitrogen (Rockville, Md.). Cells were lysed and RNA was selectively extracted using Trizol per the manufacturer's instructions. For each trial, four 1-ml samples were extracted. To assist in the final RNA precipitation step, 5 μl of glycogen (2 mg/ml) was added. Previous research in our laboratory indicated that glycogen did not inhibit RT-PCR (18). Following precipitation, the RNA pellets were air-dried at 30°C for 30 min.
DNase treatments.
Dried RNA pellets were each resuspended in 4 μl of molecular grade water (Geno Technologies, St. Louis, Mo.). The resuspended RNA was then mixed with a solution containing 10 μl of 0.1 M dithiothreitol, 3 μl of RNase Out RNase inhibitor, 10 μl of DNase (Promega, Madison, Wis.), and 3 μl of DNase buffer solution (Promega). The resuspended RNA samples were pooled and incubated at 40°C for 45 min to digest contaminating coextracted DNA.
Following DNase digestion, samples were mixed with 1 volume of acidic phenol chloroform and centrifuged at 9,300 × g for 10 min for phase separation. The top phase containing pure RNA was removed and precipitated with 0.1 volume of 2.5 M sodium acetate (pH 5.2) and 2.5 volumes of cold (−10°C) 100% ethanol. Samples were placed at −20°C for at least 2 h, followed by centrifugation at 21,000 × g for 30 min to pellet the RNA.
Nucleic acid amplification.
Four pairs of primers targeting specific sequences of the slt-II toxin gene or a conserved sequence of both slt-I and slt-II were used in this study. Sequences for all primers used in this study are given in Table 1. The four used were the slt A subunit (5), the slt B subunit (20), a conserved sequence from slt-I and slt-II (14), and the entire slt-II operon sequence (20). All of the published primer sequences had been previously designated as specific for slt-II or slt-I and -II.
TABLE 1.
Oligonucleotide primers used for amplification of extracted nucleic acid
| Primer | Sequence (5"-3") | Target gene | Amplicon size (bp) |
|---|---|---|---|
| TXAF | TTAAATGGGTACTGTCCT | slt-II | 401 |
| TXAR | CAGAGTGGTATAACTGCTGTC | ||
| TXBF | TGTTTATGGCGGTTTTATTTG | slt-II | 254 |
| TXBR | ATTATTAAACTGCACTTCAG | ||
| MK1 | GGATCCTTTACGATAGACCTTCTCGAC | slt-I/II | 225 |
| MK2 | GGATCCCACATATAAATTATTTCGCTC | ||
| MK5 | TGTAAGCTTAGCCGGACAGAG | slt-II | 1,423 |
| MK6 | CCACGGATCCGGTTATGCCTC | ||
| TufAF | ACTTCCCGGGCGACGACACTC | tufA | 579 |
| TufAR | CGCCCGGCATTACCATCTCTAC |
The conserved sequence primers contained sequence ambiguities that allowed amplification of a sequence from either slt-I or slt-II. Primer sets were subjected to a Blast sequence alignment and further screened against selected other gram-negative bacteria to evaluate specificity for Shiga toxin sequences. Positive mRNA controls in all trials were evaluated using primers for tufA, an elongation factor gene constitutively expressed in living bacterial cells (26).
For amplification by RT-PCR, a pooled and purified dried RNA pellet was resuspended in 30 μl of molecular grade water. One microliter was used to quantify RNA spectrophotometrically at 260 nm (Shimadzu UV-1201, Kyoto, Japan). The remaining volume was used for RT-PCR with the four primers. When comparative studies between various enrichment conditions were conducted (CFU, temperature, aeration, and medium effect) additional molecular grade water was added as necessary to dilute the extracted RNA to equivalent absorbance readings. The extracted standardized RNA (1.30 μg/μl) was serially diluted (undiluted, 10−2, and 10−4) using molecular grade water prior to RT-PCR in order to compare band patterns from equivalent RNA extracts from different conditions. RNA was amplified using the SuperScript One-Step RT-PCR with Platinum Taq system following the manufacturer's instructions (Invitrogen, Rockville, Md.). The appropriate no reverse transcriptase control was prepared to confirm the absence of contaminating DNA.
For DNA extraction, cells were pelleted by centrifugation at 9,300 × g for 5 min. The pellets were resuspended in 200 μl of TE (Tris-EDTA) buffer and incubated at 37°C for 5 min. Following incubation, 50 μl of 10% sodium dodecyl sulfate was added, and the sample was heated at 95°C for 5 min. One volume of chloroform was added, and the sample was centrifuged for 10 min at 13,400 × g to separate phases. The top phase was removed and mixed with 0.1 volume of sodium acetate (pH 5.2) and 2.5 volumes of ethanol to precipitate DNA.
For PCR, 3 μl of DNA was added to 2 μl (10 pmol/μl) of both forward and reverse primers and 43 μl of Platinum PCR Supermix (Invitrogen) in a sterile 0.2-ml PCR tube. Amplification conditions consisted of 1 cycle at 94°C for 2 min, followed by 35 cycles of 94°C for 1.5 min, 54°C for 1 min, and 72°C for 2 min. A final extension step of 7 min at 72°C was included. Amplification products were visualized by separation on a 1% agarose gel, stained with ethidium bromide, and visualized by UV transillumination. Band intensity profiles of the diluted and undiluted amplified RNA or DNA were then compared.
Effect of aeration.
Two loopfuls of an overnight culture were inoculated into 50 ml of TSB and incubated at 37°C with shaking at 0, 50, 100, 200, or 300 rpm to a constant cell density of ca. 2.5 × 109 CFU/ml. Statically grown cells were also evaluated at the same density. Cell densities corresponded to an optical density at 600 nm (OD600) of 1.300 and were confirmed with pour plates on tryptic soy agar (TSA). Static and aerated samples were then evaluated using the described RT-PCR procedure for relative differences in toxin mRNA expression.
Effect of cell density.
Two loopfuls of an overnight culture were inoculated into TSB and incubated at 37°C with aeration (200 rpm) until cell densities of 8.0 × 106 (OD600, 0.215, mid- to late log phase), 2.5 × 109 (OD600, 1.360, late log to early stationary phase), and 1.5 × 109 (OD600, 1.720, late stationary phase) CFU/ml were reached. These cell densities were achieved following 4, 6, and 8 h of incubation, respectively. Cell densities were confirmed by pour plating. RNA was extracted and RT-PCR was conducted.
Effect of enrichment medium.
Three different media were compared for their effect on toxin mRNA expression: TSB, modified TSB with 20 mg of sodium novobiocin (mTSB+novobiocin) (7), and E. coli enrichment broth (31). Two loopfuls of an overnight culture were inoculated into 50 ml of each medium and incubated at 37°C with aeration (200 rpm) until a cell density of ca. 2.5 × 109 CFU/ml was reached. Cell numbers were confirmed by pour plating. RNA was extracted and RT-PCR was conducted.
Effect of enrichment temperature.
Two loopfuls of an overnight culture were inoculated into 50 ml of TSB and incubated at 32 or 37°C with aeration (200 rpm) until a cell density of ca. 2.5 × 109 CFU/ml was reached. RNA was then extracted from the incubated sample and analyzed using RT-PCR.
Heat-killed cell extraction.
A 7-ml tube of TSB inoculated with the test strain was incubated at 37°C overnight. The culture was then heated to 85°C for 30 min to kill all bacterial cells. Previous research had defined such parameters as sufficient in destroying EHEC in broth and food enrichments (18). Cell death was confirmed by inoculating 100 μl of the heated culture into a tube of TSB and incubating at 37°C for 48 h. The tube was then examined for turbidity (indicative of growth and cell viability). RNA or DNA was extracted and RT-PCR of slt-II mRNA and PCR of slt II DNA was conducted using the conserved-region primers to compare stability of DNA and mRNA targets following cell death. Extractions were conducted at 0, 2, 4, 8, and 24 h following cell death.
Confirmation with other Shiga toxin-producing strains.
Two loopfuls of each strain were inoculated into separate flasks containing 50 ml of TSB and incubated with aeration (200 rpm) at 37°C until cell densities reached ca. 2.5 ×109 CFU/ml. Cell numbers were confirmed by dilution series pour plating onto TSA. RNA was extracted and amplified by RT-PCR using the conserved-region and B subunit primers. Resulting slt-II toxin mRNA expression was compared to that of ATCC 43895 to confirm the consistency of the assay. Subsequently, strains of Shigella were screened for the presence of slt genes by DNA amplification using both the conserved and B subunit primers to further establish the specificity of the assay.
Sample enrichment.
A culture of E. coli O157:H7 was grown statically in TSB overnight until a cell density of 2.5 × 109 CFU/ml (OD600, 1.350). For samples enriched in a food matrix, lean ground beef was purchased at local retail and sterilized by autoclaving at 121°C for 15 min. Meat was weighed into 25-g portions and aseptically transferred into a filter stomacher bag (Weber Scientific) with 225 ml of sterile TSB. E. coli was serially diluted using sterile 0.1% (wt/vol) peptone and inoculated by pipetting 1 ml of the dilution such that a final cell concentration of 10 or 1 CFU/g of meat was achieved. Sample inoculum was confirmed by dilution series pour plating (inoculum prior to dilution) with TSA or a five-tube most-probable-number series (diluted inoculum) with TSB prior to enrichment inoculation. A negative control was prepared containing the same amount of meat, TSB, and 1 ml of sterile peptone to confirm that RNA and DNA were extracted from inoculated replicates only.
The meat enrichment samples and controls were homogenized for 30 s using a stomacher lab blender 400 (Tekmar Company, Cincinnati, Ohio). Broth enrichments were prepared in the same manner, using the same inoculum levels but excluding autoclaved meat. All samples were incubated at 37°C with shaking (200 rpm) for 24 h. Duplicate aliquots were taken at 4, 8, 12, and 24 h for RNA and DNA extraction and amplification. To compensate for low densities, cells were pooled from 10 ml of enrichment at 4 and 8 h by centrifugation for 8 min at 10,000 × g and resuspended in 1 ml of sterile peptone. Cells were collected from 1-ml aliquots at 12 and 24 h. Dilution series pour plating with TSA confirmed cell numbers at all time points.
Immunomagnetic collection of cells.
Harvesting E. coli cells from meat enrichments was performed by immunomagnetic separation using Dynabeads' anti-E. coli O157:H7 system and following the instructions of the manufacturer (Dynal A.S., Oslo, Norway). Cells were washed twice with 1 ml of phosphate-buffered saline-Tween buffer solution prior to RNA and DNA extraction. Preliminary work in our laboratory indicated that the use of Dynabeads enhanced detection sensitivities by PCR and RT-PCR.
Preparation of labeled probe.
Biotinylated probes were synthesized using the random primer biotin labeling kit with streptavidin-alkaline phosphatase according to the manufacturer's protocol (NEN Life Science Products, Boston, Mass.). Briefly, PCR for the production of each probe was performed in a 50-μl volume containing 2 μl of both forward and reverse primers, 43 μl of Platinum Super Mix, and 3 μl of a 16-h broth culture of E. coli O157:H7. The reaction mix was amplified using the PCR program described above. Amplified product was purified using Concert rapid PCR purification system following the instructions of the manufacturer (Invitrogen). Yields were analyzed by absorbance at 260 nm to determine the concentration of PCR product and the volume of purified PCR product necessary for probe synthesis.
Detection by Southern blot hybridization.
Amplification products were separated on a 1% agarose gel, denatured in 1.5 N sodium hydroxide-1.5 M sodium chloride for 45 min, and then transferred to a positively charged nylon membrane (NEN Life Science Products) overnight. Following transfer, hybridization was conducted per the manufacturer's instructions. Briefly, the membrane was placed in a pouch with 10 ml of a prehybridization solution containing 2× standard sodium citrate and carrier DNA (salmon sperm, 50 μg/ml) and incubated at 65°C with agitation for 1 h. Buffer was removed, and a mixture of 10 ml of buffer plus 30 μl of the prepared probe and carrier DNA were added. The membrane and probe were hybridized overnight at 65°C with agitation. Following washing of the membrane to remove nonspecifically bound probe, membranes were immersed in buffer containing streptavidin-alkaline phosphatase conjugate (NEN Life Science Products) for 1 h at 65°C. Following washing of the membrane, products were detected by chemiluminescent detection using CDP-Star chemiluminescent reagent (NEN Life Science Products). Blots were developed using Kodak X-OMAT LS imaging film (Eastman Kodak Co., Rochester, N.Y.).
Statistical analysis.
All RT-PCR assays were conducted in triplicate on different days to evaluate consistency of results. One-way analysis of variance was conducted on RNA yields to determine if significant differences (P < 0.05) existed. PCR, RT-PCR, and Southern blots were performed in triplicate on different days to determine consistency of amplified products as evaluated by visual examination.
RESULTS
Selected primers were specific for slt-II or slt-I and -II by sequence alignment. Amplification products were not obtained from non-SLT-producing bacteria. Amplification products were obtained with certain Shigella strains used, including ATCC 25875, which provided a PCR product corresponding to slt-II, while ATCC 9207 and 9290 amplicons indicated both slt-I and slt-II. The results of enrichment optimization are presented in Table 2. Concentrations of RNA extracted for each treatment are presented in Table 3.
TABLE 2.
Effect of enrichment conditions on detection of slt mRNA by RT-PCRa
| Condition | cDNA band present at dilution:
|
||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| A subunit
|
B subunit
|
Conserved
|
slt operon
|
||||||||||
| 100 | 10−2 | 10−4 | 100 | 10−2 | 10−4 | 100 | 10−2 | 10−4 | 100 | 10−2 | 10−4 | ||
| Aeration (rpm) | |||||||||||||
| 0 | + | − | − | + | + | − | + | − | − | + | − | − | |
| 50 | + | − | − | + | + | − | + | + | − | + | − | − | |
| 100 | + | + | + | + | + | + | + | + | − | + | + | − | |
| 200 | + | + | + | + | + | + | + | + | − | + | + | − | |
| 300 | + | + | + | + | + | + | + | + | − | + | + | − | |
| Growth phase (CFU) | |||||||||||||
| Mid-log (106) | + | + | − | + | + | − | + | + | − | + | − | − | |
| Late log (109) | + | + | + | + | + | + | + | + | − | + | + | − | |
| Stationary (109) | + | + | − | + | + | − | + | + | − | + | − | − | |
| Mediumb | |||||||||||||
| TSB | + | + | + | + | + | + | + | + | − | + | + | − | |
| ME | + | + | + | + | + | + | + | + | − | + | + | − | |
| EEB | + | + | + | + | + | + | + | + | + | + | + | − | |
| Growth temp (°C) | |||||||||||||
| 37 | + | + | + | + | + | + | + | + | − | + | + | − | |
| 32 | + | + | − | + | + | + | + | + | − | + | − | − | |
Presence and absence of cDNA bands on gels are denoted by + and −, respectively. All trials were conducted in triplicate to verify result consistency.
TSB, tryptic soy broth; ME, modified E. coli enrichment broth with added novobiocin; EEB, E. coli enrichment broth.
TABLE 3.
Concentrations of RNA extracted for each treatmenta
| Condition | RNA yield (μg/μl) |
|---|---|
| Growth phase | |
| Mid-log (106 CFU/ml) | 1.60 ± 0.002 |
| Late log (109 CFU/ml) | 1.52 ± 0.003 |
| Stationary (109 CFU/ml) | 1.30 ± 0.0 |
| Medium | |
| TSB | 1.52 ± 0.003 |
| Modified TSB + novobiocin | 1.60 ± 0.001 |
| E. coli enrichment broth | 1.20 ± 0.002 |
| Temp (°C) | |
| 37 | 1.52 ± 0.0 |
| 32 | 1.28 ± 0.003 |
| Aeration (rpm) | |
| 0 | 1.00 ± 0.0 |
| 50 | 1.00 ± 0.002 |
| 100 | 1.56 ± 0.002 |
| 200 | 1.52 ± 0.001 |
| 300 | 1.50 ± 0.001 |
All trials were conducted in triplicate to verify result consistency. Values are means ± standard deviations of three trials.
In comparative studies, mRNA yields were adjusted to equivalency prior to amplification. Aeration during incubation enhanced toxin mRNA expression and subsequent detection by RT-PCR. Expression of slt mRNA under static conditions was less vigorous. Undiluted amplicons of all primers were visually detected except for the entire operon. The B subunit and conserved-region amplicons were also detected at the 10−2 dilution (Fig. 1).
FIG. 1.
Amplification of slt mRNA by primer sets under static enrichment conditions. RT-PCR products from dilution series RT-PCR of purified E. coli O157:H7 grown under static conditions. Lanes 1 to 3, A subunit amplicons (undiluted, diluted 10−2, and diluted 10−4, respectively); lanes 4 to 6, B subunit amplicons (undiluted, diluted 10−2, and diluted 10−4, respectively); lanes 7 to 9, conserved amplicons (undiluted, diluted 10−2, and diluted 10−4, respectively); lanes 10 to 12, entire-operon amplicons (undiluted, diluted 10−2, and diluted 10−4, respectively). Lanes 13 to 15 are the positive control and two negative controls, respectively. A 100-bp DNA ladder was used as a molecular weight marker.
Expression of slt mRNA was also influenced by bacterial growth phase. The highest amount of toxin mRNA was detected when RNA was extracted from late log- and early stationary-phase cells corresponding to a cell density of 108 to 109 CFU/ml. Expression of slt-II mRNA as measured by RT-PCR increased when the bacteria were incubated at 37°C as opposed to 32°C. At 37°C, the A subunit and the entire operon sequences supported amplification through a higher dilution than at 32°C. Figure 2 indicates RT-PCR from slt RNA under optimal enrichment conditions (37°C, 200 rpm, late log-phase extraction).
FIG. 2.
Amplification of slt mRNA by different primers at various dilutions. RT-PCR products from dilution series RT-PCR of purified E. coli O157:H7 grown under optimal conditions (late log phase, 37°C, 200 rpm shaking) for RT-PCR of slt toxin mRNA. Lanes 1 to 3, A subunit amplicons (undiluted, diluted 10−2, and diluted 10−4, respectively); lanes 4 to 6, B subunit amplicons (undiluted, diluted 10−2, and diluted 10−4, respectively); lanes 7 to 9, conserved amplicons (undiluted, diluted 10−2, and diluted 10−4, respectively); lanes 10 to 12, entire-operon amplicons (undiluted, diluted 10−2, and diluted 10−4, respectively). Lanes 13 to 15 are the two negative controls and the positive control, respectively. A 100-bp DNA ladder was used as a molecular weight marker.
Medium effects on toxin expression were not as obvious as changes in temperature and aeration. Detection of mRNA was identical for all media with all primers except for a higher limit of detection for the conserved primer sequence when E. coli O157:H7 was incubated in E. coli enrichment broth. All additional strains of enterohemorrhagic E. coli tested produced RT-PCR results comparable to those of ATCC 43895 (Table 4).
TABLE 4.
RT-PCR results for additional enriched EHEC strainsa
| Serotype | Origin | cDNA band present with indicated primer at dilution:
|
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| A subunit
|
B subunit
|
Conserved
|
||||||||||
| 100 | 10−2 | 10−4 | 100 | 10−2 | 10−4 | 100 | 10−2 | 10−4 | ||||
| O111:H-JB1-95 | Salami | + | + | − | + | + | − | + | + | − | ||
| O26:H11 | CDC | + | + | − | + | + | − | + | + | + | ||
| ATCC 43890 | Fecal isolate | + | + | − | + | + | − | + | + | − | ||
| ATCC 35150 | Fecal isolate | + | + | − | + | + | + | + | + | + | ||
| O26:H11 3359-70 | CDC | + | + | − | + | + | − | + | + | + | ||
| O157:H7 | Salami | + | − | − | + | + | − | + | + | + | ||
Enrichment conditions were incubation in typtic soy broth at 37°C with aeration at 200 rpm and extraction of mRNA from late-log-phase cells.
RNA extractions from heat-killed cells produced no bands from RT-PCR, while PCR of slt-II DNA consistently produced bands following cell death (Fig. 3). The B subunit and the conserved-region primers consistently supported mRNA amplification through a higher dilution than the A subunit or the entire-operon primers. Thus, the B subunit and conserved-region primers were used in subsequent studies in broth and meat enrichments.
FIG. 3.
Heat-killed EHEC nucleic acid amplified by the conserved slt region primer. Lanes 1 to 5, amplification of mRNA at 0, 2, 4, 8, and 24 h following cell death. Lanes 6 to 10, amplification of DNA at 0, 2, 4, 8, and 24 h following cell death. The figure shows amplification product separated on an electrophoresis gel and subsequent detection by Southern blot hybridization.
The time points at which amplification product was detected in broth and meat enrichments and corresponding cell numbers for each time point are listed in Tables 5 and 6, respectively. In all cases, PCR amplification products were detected at the same extraction point or a time point earlier than were RT-PCR amplification products. Detection by PCR and RT-PCR from both broth and meat enrichments was identical for both inocula at 1 or 10 CFU/g. With an initial inoculum of 1 CFU/g, detection of mRNA occurred at 12 and 24 h, whereas DNA was amplified at 8, 12, and 24 h. For an inoculum of 10 CFU/g, PCR and RT-PCR products were detected at 8, 12, and 24 h. Differences were not observed between primer sets for PCR or RT-PCR.
TABLE 5.
Detection of E. coli slt mRNA and DNA by RT-PCR and PCR in meat and broth enrichments at different inoculum levelsa
| Primer | Time (h) | Detection at inoculum:
|
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 10 CFU/g
|
1 CFU/g
|
||||||||||
| Broth
|
Meat
|
Broth
|
Meat
|
||||||||
| RNA | DNA | RNA | DNA | RNA | DNA | RNA | DNA | ||||
| B subunit | 4 | −* | −* | − | − | −* | −* | − | − | ||
| 8 | +* | +* | +* | +* | −* | +* | −* | +* | |||
| 12 | +* | +* | +* | +* | +* | +* | +* | +* | |||
| 24 | +* | +* | +* | +* | +* | +* | +* | +* | |||
| Conserved region | 4 | −* | −* | − | − | −* | −* | − | − | ||
| 8 | +* | +* | +* | +* | −* | +* | −* | +* | |||
| 12 | +* | +* | +* | +* | +* | +* | +* | +* | |||
| 24 | +* | +* | +* | +* | +* | +* | +* | +* | |||
All experiments were conducted in triplicate to verify result consistency. Presence or absence of an amplification product is denoted by + and −, respectively. *, amplification product detected by Southern blot hybridization.
TABLE 6.
CFU for meat and broth enrichments with different initial inoculum levelsa
| Time (h) | CFU after initial inoculum of:
|
||||
|---|---|---|---|---|---|
| 10 CFU/g
|
1 CFU/g
|
||||
| Broth | Meat | Broth | Meat | ||
| 0 | 250 | 250 | 25 | 25 | |
| 4 | 6.4 × 103 | 3.5 × 101 | 2.0 × 102 | 1.8 × 101 | |
| 8 | 4.1 × 106 | 1.0 × 104 | 1.5 × 105 | 4.0 × 104 | |
| 12 | 2.2 × 109 | 1.9 × 108 | 1.8 × 108 | 2.0 × 108 | |
| 24 | 8.3 × 109 | 1.3 × 1010 | 6.2 × 109 | 2.6 × 1010 | |
One milliliter of each sample was used in a dilution series pour plating with TSA. Time zero was calculated from the cell dilution of the initial inoculum used to inoculate 25 g of meat plus 225 ml of broth.
Southern blot hybridization detected amplification products at all time points in broth enrichments and at 8, 12, and 24 h for meat enrichments. Amplification products and subsequent detection by Southern blot hybridization for both broth and meat enrichments with initial inocula of 1 and 10 CFU/g are shown in Fig. 4.
FIG. 4.
Amplification of slt mRNA and DNA following broth and meat enrichments. Broth images are on the left, and meat images are on the right. Top left: broth, initial cell number of 10 CFU/ml; bottom left, broth, initial cell number of 1 CFU/ml; top right, ground beef, initial cell number of 10 CFU/g; bottom right, ground meat, initial cell number of 1 CFU/g. For each image, the lane numbers are as follows: 1 to 4, amplification of mRNA by the B subunit primer set at 4, 8, 12, and 24 h, respectively; 5 to 8, amplification of DNA by the B subunit primer set at 4, 8, 12, and 24 h, respectively; 9 to 12, amplification of mRNA by the conserved primer set at 4, 8, 12, and 24 h, respectively; 13 to 16, amplification of DNA by the conserved primer set at 4, 8, 12, and 24 h, respectively.
DISCUSSION
As the number of studies involving rapid detection of food pathogens has increased over the past several years, many investigations have examined detection of E. coli O157:H7 by PCR. Several of these studies targeted the slt-II toxin gene when detecting pathogenic E. coli from enrichment broth and artificially contaminated foods (5, 11, 19). The use of bacterial mRNA as a means of pathogen detection in foods, particularly that of slt mRNA of E. coli O157:H7, has not been widely studied.
The role of specific environmental conditions that influence the regulation of slt-II by enterohemorrhagic E. coli has not been well defined. Shiga toxin production is apparently repressed by high levels of iron and reduced temperature (23). Muhldorfer et al. (22) used a phoA gene fusion to the slt-II promoter to study regulation of this element. Results from their study indicated that slt-II was unaffected by temperature. The present study provides evidence that temperature does impact slt-II toxin mRNA expression. The differences between our study and the previous study may be due to the methods of measurement.
The mature SLT II toxin consists of one A subunit surrounded by five B subunits, and the genes exist in an operon with one promoter in front of the A subunit (23). Some have suggested that expression of the B subunit region may be initiated by the existence of a second promoter upstream of the B subunit sequence, thereby accounting for five B subunits to one A subunit in the mature toxin (3, 23). However, further study has been unable to validate the presence of a second promoter (27). The B subunit was consistently a more robust amplicon than either the A subunit or the entire operon. These differences may reflect differences in primer binding abilities or mRNA stability and could be addressed in future studies using additional primer sets for this target.
Previous studies have examined the effect of oxygen on toxin production. James and Keevil (13) found that both SLT-I and SLT-II genes were constitutively expressed and not dependent on the availability of oxygen. Toxin expression was measured indirectly by a reversed-phase latex agglutination assay. Muhldorfer et al. (22) also found no effect of oxygen on toxin expression using a phoA-slt-II promoter gene fusion. The current study found that toxin mRNA expression increased upon aeration when extracted from equivalent cell densities. Again, differences in methods used to measure toxin expression in these two systems likely account for the results of these two studies.
Another study investigated the instability of mRNA following cell death. Sheridan et al. (27) developed an RT-PCR procedure to detect mRNA from selected genes of E. coli that had been killed with various combinations of heat, ethanol, and time treatments. Results indicated that certain transcripts may be detectable by RT-PCR for over 2 h but not at 16 h following cell death by heat treatments (100°C for 5 min or 80°C for 10 min) (27). mRNA of slt was not detected by RT-PCR immediately following cell death, while DNA was consistently amplified through 24 h following cell death.
Klein and Juneja (15) conducted a study to examine mRNA as an indicator of cell viability for Listeria monocytogenes. Amplification was conducted using three genes as targets: iap, coding for the p60 extracellular protein; hly, coding for virulence factor listeriolysin O; and prfA, coding for virulence factor regulation. Amplification of both mRNA and DNA from the targeted genes was compared following cell death from heating for 15 min at 121°C. Their results indicated that target DNA was extremely stable and capable of being amplified following cell death, unlike mRNA, which was not amplified by RT-PCR following cell death (15). McKillip et al. (18) also studied the stability of 16S rDNA compared to 16S rRNA following cell death and found comparable results. rDNA degraded very little through 48 h following cell death by mild heat treatment with only a 10-fold loss in PCR amplification efficiency, while rRNA was more unstable following cell death at 85°C.
Transcript pools in bacteria vary widely in their relative abundance throughout cell growth. Szabo and Mackey (29) reported large differences in sefA expression by Salmonella in growth phase, growth medium, and growth temperature. In the present study, toxin expression was weaker in log-phase cells as opposed to late log- and early stationary-phase cells. Ideally, an mRNA target is abundantly and constitutively expressed while exhibiting species or strain specificity. Finding such a target for EHEC or STEC would be difficult. A transcript that is constitutively expressed but at high levels is extremely rare, leading to a trade-off between assay sensitivity and reliability. Such a target that also affords species specificity further complicates the issue. Under conditions of enrichment, slt-II or slt-I and -II toxin expression was consistent and was associated with cell viability.
The current study determined that slt DNA was stable following heat treatment and supported amplification by PCR 24 h post-cell death, whereas slt mRNA did not amplify by RT-PCR at any time point following cell death. These findings support the hypothesis that detection of slt mRNA was associated with cell viability. Production of slt-II mRNA was influenced by aeration, temperature, and growth phase. For optimal toxin production, these findings suggest enrichment parameters of inoculating the cells in enrichment broth and incubating at 37°C with aeration in excess of 100 rpm. However, aeration at 200 rpm is recommended because of the increased rate of cell growth, and RNA extraction during late log/early stationary phase for optimal yields. The most robust target for RT-PCR was the B subunit or the slt-I and -II conserved region.
Detection of pathogenic microorganisms in food by RT-PCR is dependent not only on the abundance of the selected target gene but also on the presence of inhibitors in the food matrix that may prevent sufficient extraction of template for amplification or may inhibit the amplification process. Inhibitors have been reported in a variety of products, including blood, feces, and foods (4, 34). The amplification of nucleic acid may be inhibited or prevented in three ways: interference with cell lysis and subsequent nucleic acid extraction, nucleic acid degradation or capture, and blocking of primer annealing by inhibitory substances (34). Enrichment procedures can overcome many of these obstacles by increasing cell numbers. Immunomagnetic separation was also used to increase assay sensitivity. Our sensitivities with immunomagnetic separation correspond with those of previous studies (30).
In the present study, matrix effects were observed for PCR products only by Southern blot detection. DNA amplification products were observed by blots following 4 h of enrichment in broth, but were not visible after 4 h in meat enrichments. Gel visualization was consistent for PCR products. With RT-PCR, matrix effects between broth and meat enrichments were more pronounced and more variable. Decreases in cell and/or nucleic extraction efficiency from the meat enrichment compared to broth are a likely source of the decreased sensitivity by both DNA and mRNA from meat enrichments (18, 19). The differences in sensitivity of detection of mRNA compared to DNA may represent the inherent instability of mRNA compared to DNA or inhibition of the reverse transcriptase step of RT-PCR. These issues need to be addressed to optimize RT-PCR detection techniques and should be addressed in future studies through the use of enhanced RT enzymes and/or various reaction component additives.
In the current study, similarities in amplification efficiencies between the B subunit and conserved primer sequence under enrichment conditions were observed. This is not surprising, given that the B subunit primer targets a sequence that exists in five locations per whole toxin along with the fact that the conserved primer by definition may amplify a sequence found in both slt-I and slt-II. The existence of more template may account for the increased efficiency of the primers.
As a related matter, PCR primer specificity is a critical factor in the detection of a particular bacterium. The current study included a number of corollary investigations to determine the ability of the B subunit and conserved-region primers in detecting bacteria other than EHEC. Of the Shigella strains screened, ATCC 25875 amplified with the B subunit primers, while ATCC 9207 and 9290 amplified with both the B subunit and conserved sequence primers. These results are not surprising, since Shigella and E. coli are closely related and the slt genes are thought to have come from Shigella. The ability to detect Shigella in addition to STEC is not a detriment to the assay.
It is crucial to validate molecular detection assays through application to food enrichments to evaluate their practical use. EHEC are shed in the feces of warm-blooded animals, and both clinical and laboratory studies have shown ground beef to be a major vehicle for E. coli O157:H7 (1, 8). The current study was able to detect the presence of 1 CFU of viable Shiga toxin-producing E. coli per g in 25 g of ground beef following 12 h of enrichment, for a complete assay time of less than 24 h. Assay confirmation by Southern blot was achieved in less than 36 h.
Application of RT-PCR to other emerging technologies such as molecular beacons and real-time PCR will further enhance this technique. RT-PCR of mRNA shows promise as a tool for the rapid detection of viable food-borne pathogens.
Acknowledgments
Funding was provided by the USDA NRI (award 99035201-8125) and Dairy Management, Inc.
We thank John McKillip for helpful reviews of the manuscript.
Footnotes
This is paper number FSR 01-44 in the Journal Series of the Department of Food Science, North Carolina State University, Raleigh, N.C.
REFERENCES
- 1.Acheson, D. W. K. 2000. How does Escherichia coli O157:H7 testing in meat compare with what we are seeing clinically? J. Food Prot. 63:819-821. [DOI] [PubMed] [Google Scholar]
- 2.Barbour, W., and G. Tice. 1997. Genetic and immunological techniques for detecting foodborne pathogens and toxins, p. 710-727. In M. P. Doyle (ed.), Fundamentals and frontiers in food microbiology. ASM Press, Washington, D.C.
- 3.Beltrametti, F., A. Kresse, and C. Guzman. 1999. Transcriptional regulation of the esp genes of enterohemorrhagic Escherichia coli. J. Bacteriol. 181:3409-3418. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Bickley, J., J. Short, D. McDowell, and H. Parkes. 1996. Polymerase chain reaction (PCR) detection of Lysteria monocytogenes in diluted milk and reversal of PCR inhibition caused by calcium ions. Lett. Appl. Microbiol. 22:153-158. [DOI] [PubMed] [Google Scholar]
- 5.Chen, J., R. Johnson, and M. Griffiths. 1998. Detection of verotoxigenic Escherichia coli by magnetic capture-hybridization PCR. Appl. Environ. Microbiol. 64:147-152. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Dineen, S., K. Takeuchi, J. Soudah, and K. Boor. 1998. Persistence of Escherichia coli O157:H7 in dairy fermentation systems. J. Food Prot. 61:1602-1608. [DOI] [PubMed] [Google Scholar]
- 7.Doyle, M. P., and J. Schoeni. 1987. Isolation of Escherichia coli O157:H7 from retail fresh meats and poultry. Appl. Environ. Microbiol. 53:2394-2396. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Doyle, M., T. Zhao, J. Meng, and S. Zhao. 1997. Escherichia coli O157:H7, p. 171-191. In M. P. Doyle (ed.), Food microbiology: fundamentals and frontiers, 5th ed. ASM Press, Washington, D.C.
- 9.Feng, P., S. Weagant, and S. Monday. 2001. Genetic analysis for virulence factors in Escherichia coli O104:H21 that was implicated in an outbreak of hemorrhagic colitis. J. Clin. Microbiol. 39:24-28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fratamico, P., and T. Strobaugh. 1998. Simultaneous detection of Salmonella spp. and Escherichia coli O157:H7 by multiplex PCR. J. Indust. Microbiol. Biotechnol. 21:92-98. [Google Scholar]
- 11.Gannon, V., S. D'Souza, T. Graham, R. King, K. Rahn, and S. Read. 1997. Use of flagellar H7 gene as a target in multiplex PCR assays and improved specificity in identification of enterohemorrhagic Escherichia coli strains. J. Clin. Microbiol. 35:656-662. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Herman, L. 1997. Detection of viable and dead Listeria monocytogenes by PCR. Food Microbiol. 14:103-110. [Google Scholar]
- 13.James, B., and C. Keevil. 1999. Influence of oxygen availability on physiology, verocytotoxin expression and adherence of Escherichia coli O157. J. Appl. Microbiol. 86:117-124. [DOI] [PubMed] [Google Scholar]
- 14.Karch, H., and T. Meyer. 1989. PCR amplification of Shiga-like toxin genes. J. Clin. Microbiol. 27:2751-2757. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Klein, P., and V. Juneja. 1997. Sensitive detection of viable Listeria monocytogenes by reverse transcription-PCR. Appl. Environ. Microbiol. 63:4441-4448. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.McIngvale, S., X. Chen, J. McKillip, and M. A. Drake. 2000. Survival of Escherichia coli O157:H7 in buttermilk as affected by contamination point and storage temperature. J. Food Prot. 63:441-444. [DOI] [PubMed] [Google Scholar]
- 17.McKillip, J., L. Jaykus, and M. A. Drake. 1998. rRNA stability in heat-killed and UV-irradiated enterotoxigenic Staphylococcus aureus and Escherichia coli O157:H7. Appl. Environ. Microbiol. 64:4264-4268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.McKillip, J., L. Jaykus, and M. A. Drake. 1999. Nucleic acid persistence in heat-killed Escherichia coli O157:H7 from contaminated skim milk. J. Food Prot. 62:839-844. [DOI] [PubMed] [Google Scholar]
- 19.McKillip, J. L., L. Jaykus, and M. A. Drake. 2000. A comparison of methods for the detection of Escherichia coli O157:H7 from artificially contaminated dairy products using PCR. J. Appl. Microbiol. 89:49-55. [DOI] [PubMed] [Google Scholar]
- 20.Meng, J., S. Zhao, M. Doyle, S. Mitchell, and S. Kresovich. 1997. A multiplex PCR for identifying Shiga-like toxin-producing Escherichia coli O157:H7. Lett. Appl. Microbiol. 24:172-176. [DOI] [PubMed] [Google Scholar]
- 21.Meyer, T., H. Karch, J. Hacker, H. Bocklage, and J. Heesemann. 1992. Cloning and sequencing of a shiga-like toxin II-related gene from Escherichia coli O157:H7 strain 7279. Zentralbl. Bakteriol. 276:176-188. [DOI] [PubMed] [Google Scholar]
- 22.Muhldorfer, I., J. Hacker, G. Keusch, D. Achenson, H. Tschape, A. Kane, A. Ritter, T. Olschlager, and A. Donohue-Rolfe. 1996. Regulation of the Shiga-like toxin II operon in Escherichia coli. Infect. Immun. 64:495-502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.O'Brien, A., and R. Holmes. 1987. Shiga and Shiga-like toxins. Microbiol. Rev. 51:206-220. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Palumbo, S., J. Call, F. Schultz, and A. Williams. 1995. Minimum and maximum temperatures for growth and verotoxin production by hemorrhagic strains of Escherichia coli. J. Food Prot. 58:352-356. [DOI] [PubMed] [Google Scholar]
- 25.Restaino, L., H. Castillo, D. Stewart, and M. Tortorello. 1996. Antibody-direct epifluorescent filter technique and immunomagnetic separation of 10-h screening and 24-h confirmation of Escherichia coli O157:H7 in beef. J. Food Prot. 59:1072-1075. [DOI] [PubMed] [Google Scholar]
- 26.Russmann, H., H. Schmidt, J. Heesemann, A. Caprioli, and H. Karch. 1994. Variants of shiga-like toxin II constitute a major toxin component in Escherichia coli O157 strains from patients with haemolytic uremic syndrome. J. Med. Microbiol. 40:338-343. [DOI] [PubMed] [Google Scholar]
- 27.Sheridan, G., C. Masters, J. Shallcross, and B. Mackey. 1998. Detection of mRNA by reverse transcription-PCR as an indicator of viability in Escherichia coli cells. Appl. Environ. Microbiol. 64:1313-1318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Sung, L., M. Jackson, A. O'Brien, and R. Holmes. 1990. Transcription of the Shiga-like toxin type II and Shiga-like toxin type II variant operons of Escherichia coli. J. Bacteriol. 172:6386-6395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Szabo, R., E. Todd, and A. Jean. 1986. Method to isolate Escherichia coli O157:H7 from food. J. Food Prot. 49:768-772. [DOI] [PubMed] [Google Scholar]
- 30.Tortorello, M. L., K. Reineke, D. Stewart, and R. Raybourne. 1998. Comparison of methods for determining the presence of Escherichia coli O157:H7 in apple juice. J. Food Prot. 61:1425-1430. [DOI] [PubMed] [Google Scholar]
- 31.Tsai, Y., and S. Ingham. 1997. Survival of Escherichia coli O157:H7 and Salmonella ssp. in acidic condiments. J. Food Prot. 60:751-755. [DOI] [PubMed] [Google Scholar]
- 32.Vernozy-Rozand, C. 1997. Detection of Escherichia coli O157:H7 and other verocytotoxin-producing E. coli (VTEC) in food. J. Appl. Microbiol. 82:537-551. [DOI] [PubMed] [Google Scholar]
- 33.Weagant, S., J. Bryant, and K. Jinneman. 1994. An improved rapid technique for isolation of Escherichia coli O157:H7 from foods. J. Food Prot. 58:7-12. [DOI] [PubMed] [Google Scholar]
- 34.Weeratna, R. D., and M. Doyle. 1991. Detection and production of verotoxin I of Escherichia coli O157:H7 in food. Appl. Environ. Microbiol. 57:2951-2955. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Wilson, I. 1997. Inhibition and facilitation of nucleic acid amplification. Appl. Environ. Microbiol. 63:3741-3751. [DOI] [PMC free article] [PubMed] [Google Scholar]