100
Journal o f Food Protection, Vol. 77, No. 1, 2014, Pages 100-105 doi: 10.4315/0362-028X.JFP-13-274 Copyright © , International Association for Food Protection
Research Note
Rapid and Sensitive Detection of Escherichia coli 0157:H7 in Milk and Ground Beef Using Magnetic Bead-Based Immunoassay Coupled with Tyramide Signal Amplification MUHSIN AYDIN,1 GENE P. D. HERZIG,2 KWANG CHEOL JEONG,3 SAMANTHA DUNIGAN,2 PARTH SHAH,4 a n d SOOHYOUN AHN5* 1Department o f Molecular Biosciences and departm ent o f Biological Sciences, Arkansas State University, Jonesboro, Arkansas 72401; 3Emerging Pathogens Institute and Department o f Animal Sciences and 5Department o f Food Science and Human Nutrition, Institute o f Food and Agricultural Sciences, University o f Florida, Gainesville, Florida 32611; and 4Department o f Cell Biology and Neuroscience, Rutgers University, New Brunswick, New Jersey 07102, USA MS 13-274: Received 2 July 2013/Accepted 11 September 2013
ABSTRACT Escherichia coli 0157:H7 is a major foodbome pathogen that has posed serious problems for food safety and public health. Recent outbreaks and recalls associated with various foods contaminated by E. coli 0157:H7 clearly indicate its deleterious effect on food safety. A rapid and sensitive detection assay is needed for this harmful organism to prevent foodbome illnesses and control outbreaks in a timely manner. We developed a magnetic bead-based immunoassay for detection of E. coli Oi57:H7 (the most well-known Shiga toxigenic E. coli strain) with a 96-well microplate as an assay platform. Immunomagnetic separation (IMS) and tyramide signal amplification were coupled to the assay to increase its sensitivity and specificity. This immunoassay was able to detect E. coli 0157:H7 in pure culture with a detection limit of 50 CFU/ml in less than 3 h without an enrichment step. The detection limit was decreased 10-fold to 5 CFU/ml with addition of a 3-h enrichment step. When this assay was tested with other nontarget foodbome pathogens and common enteric bacteria, no cross-reactivity was found. When tested with artificially contaminated ground beef and milk samples, the assay sensitivity decreased two- to fivefold, with detection limits of 250 and 100 CFU/ml, respectively, probably because of the food matrix effect. The assay results also were compared with those of a sandwich-type enzyme-linked immunosorbent assay (ELISA) and an ELISA coupled with IMS; the developed assay was 25 times and 4 times more sensitive than the standard ELISA and the IMS-ELISA, respectively. Tyramide signal amplification combined with IMS can improve sensitivity and specificity for detection of E. coli 0157:H7. The developed assay could be easily adapted for other foodbome pathogens and will contribute to improved food safety and public health.
Shiga toxin-producing Escherichia coli (STEC) strains are major foodbome pathogens and have caused a large number of human illnesses. Among STEC strains associated with human infections, E. coli 0157:H7 is the most wellknown serotype and is responsible for 89 and 100% of STEC-related hospitalizations and deaths, respectively, in the United States (29). E. coli 0157:H7 causes diarrhea with or without blood, hemorrhagic colitis, and in more severe cases hemolytic uremic syndrome (15). The Centers for Disease Control and Prevention estimated that E. coli 0157:H7 causes approximately 63,000 illnesses, 2,100 hospitalizations, and 20 deaths annually in the United States (29). Most E. coli 0157:H7 infection outbreaks have been linked to the consumption of ground beef (9). However, human infections by other routes also have been reported, including the consumption of contaminated vegetables, drinking water, milk, and dairy products and direct contact with animals during visits to farms or petting * Author for correspondence. Tel: 352-392-1991; Fax: 352-392-9467; E-mail: [email protected].
zoos (10, 17, 19, 25, 26). Considering the impact of E. coli 0157:H7 on food safety and public health, rapid and sensitive detection methods for this pathogen in food and environmental samples are needed to monitor contamination of food products and to control outbreaks in a timely manner (14). Culture methods involving sorbitol-MacConkey agar or other selective agars such as CHROMagar and Rainbow agar are most commonly used for E. coli 0157:H7 detection (4,5, 23). These methods are also frequently combined with serological tests using antibodies to the 0157 and H7 antigens to confirm the presence of E. coli 0157:H7. However, these conventional methods are time-consuming and laborious. Most culture methods are based on screening of the non-sorbitol-fermenting 0157 serotype, and some traditional enteric media can even inhibit the growth of non0157 STEC (1, 8, 14, 16). Therefore, modification of these culture methods to simultaneously detect both E. coli 0157:H7 and other non-0157 STEC strains is difficult. Antibody and nucleic acid methods also have been developed and tested for detection of various STEC strains
J. Food Prot., Vol. 77, No. 1
101
DETECTION OF E. COU 0157:H7 IN FOODS USING IMMUNOASSAY
(2, 21, 33). These alternative methods could provide low
TABLE 1. Cross-reactivity of 1MB-TySA immunoassay with
detection limits, high specificity, and rapid processing; however, most of these methods have the disadvantages of cross-reactivity, limited multiplexing ability, and the presence of inhibitory substances in the sample. The sensitivity of these alternative methods also is highly affected by sample preparation, mostly because of interference from the sample matrix and low levels of the target pathogen mixed with high levels of background microflora. To develop a sensitive detection method, a sample preparation method is needed that can concentrate the target pathogens and isolate them from background microflora with minimal effect from the sample matrix. Immunomagnetic separation (IMS) has been used to effectively isolate and concentrate target organisms from complex matrices in food, clinical, and environmental samples (12, 31). IMS can reduce the total assay time and provide higher sensitivity for pathogen detection in complex food or environmental samples (3). Tyramide signal amplification (TySA), also known as catalyzed reporter deposition, is a signal amplification technique based on horseradish peroxidase (HRP)-catalyzed deposition of tyramide conjugates (6). This method has been used to increase detection sensitivity by amplifying chromogenic or fluorescent signals in many applications including immunohistochemistry, in situ hybridization, electron microscopy, and immunoassays (7, 18, 22, 32). Although TySA has been used in immunoassays for detection of protein markers in medical studies, its use in food safety applications for detecting foodbome pathogens has not been reported. In this study, we combined the IMS and TySA techniques to detect E. coli 0157:H7 in food and found that this approach was highly sensitive and minimally affected by other components in the food. The goal of this study was to develop a novel immunoassay based on IMS and TySA techniques for more rapid and sensitive detection of E. coli 0157:H7.
common foodbome pathogens
M A TER IA LS A N D M E TH O D S Microbial culture and growth conditions. E. coli 0157:H7 ATCC 43890 was used as the main test strain in this study. Other bacterial strains used are listed in Table 1. All non-0157:H7 STEC strains were obtained from the Citrus Research and Education Center (University of Florida, Lake Alfred), and all other strains (including E. coli 0157:H7 43890) were obtained from the University of Arkansas (Fayetteville). Culture of each bacterial strain was grown hi tryptic soy broth (BD, Franklin Lakes, NJ) at 37°C for 18 h. Cells were collected from 1 ml of culture by centrifugation (17,000 x g for 5 min) and resuspended in 1 ml of phosphate-buffered saline (PBS; 1 x , pH 7.6). The number of viable cells was determined by plate counting on MacConkey agar (Remel, Lenexa, KS). Food sample preparation. Ground beef and milk were purchased from local retailers 24 h before each experiment. Ground beef samples (25 g each) were mixed with 225 ml of 1 x PBS (pH 7.6) and homogenized with a stomacher for 1 min. The solutions were collected and then inoculated with the E. coli 0157:H7 culture. Milk samples were directly inoculated with E. coli 0157:H7 culture without dilution. The level of E. coli 0157:H7 in the food samples was adjusted to 1 to 106 CFU/ml.
Tested pathogen0
Detection results6
E. coli 0157:H7 (ATCC 43888) E. coli 0157:H7 (ATCC 43889) E. coli 0157:H7 (ATCC 43890) E. coli 0157:H7 (ATCC 43895) E. coli JM 109 (lab strain) E. coli O103:H2 E. coli O104:H4 E. coli 0145 E. coli 0111 Citrobacter freundii Enterobacter aerogenes Staphylococcus aureus S. epidermidis Listeria monocytogenes Salmonella Enteritidis Salmonella Typhimurium Salmonella Infantis Shigella dysenteriae S. flexneri
+ + + + — — —
“ All E. coli 0157:H7 strains were tested at 106 CFU/ml; the other pathogens were tested at 107 CFU/ml. b + , positive signal, i.e., greater than or equal to three times the standard deviation of the background signal; —, negative signal, i.e., less than three times the standard deviation of the background signal. The samples were used for assay immediately after inoculation without further incubation. Immunomagnetic bead-based immunofluorescence assay. The immunofluorescence assay with the immunomagnetic beads (IMBs) consisted of two reaction steps: (i) IMS and (ii) fluorescence immunoassay with TySA. For each assay, 1 x PBS (pH 7.6) and heat-killed E. coli 0157:H7 (KPL, Gaithersburg, MD) at 3 x 106 CFU/ml were added as a negative control and a positive control, respectively. For IMS, each well of a 96-well microtiter plate contained 3 pi of IMBs coated with anti-£. coli 0157 antibodies (Life Technologies, Grand Island, NY) and 100 pi of test sample. Plates were incubated at room temperature for 30 min with gentle agitation on a digital microplate shaker (IKA, Wilmington, NC). The I MB-/.’, coli 0157:H7 complexes were then separated from the solution with a magnet (DynaMag-96, Life Technologies) according to the manufacturer’s instructions, and the plates were washed three times with 150 pi of TNT buffer (0.1 M Tris-HCl, 0.15 M NaCl, and 0.05% Tween 20, pH 7.5). For the fluorescence immunoassay with TySA, 100 pi of 0.1 mg/ml HRP-labeled anti-£. coli 0157:H7 antibody (KPL) was added to the IMB-E. coli 0157:H7 complexes in each well. The plate was incubated at room temperature for 30 min with agitation and then washed three times with 150 pi of TNT buffer. TySA was performed using an ELAST ELISA (enzyme-linked immunosor bent assay) amplification system (Perkin Elmer, Waltham, MA) according to the manufacturer’s instructions. Biotinyl tyramide solution diluted to 50 pi with amplification diluent from the ELAST ELISA amplification system was added to each well. The plate was incubated for 10 min at room temperature with agitation and washed three times with 150 pi of TNT buffer, and 100 pi of 10 pg/ml streptavidin-Cy3 (Sigma-Aldrich, St. Louis, MO) was
102
AYDIN ET AL.
J. Food Prot., Vol. 77, No. 1
then washed three times with 150 j.il of TNT buffer. After washing, 100 (il o f HRP-labeled antibody (0.1 mg/ml) against E. coli 0157:H 7 (KPL) was added to each well. The plate was incubated at room temperature for 1 h with agitation and then washed with 150 pi o f TNT buffer. After washing, 100 p! of 3,3',5,5'tetramethylbenzidine (TMB) substrate solution (0.4 mg/ml; KPL) was added to each well, and the plate was incubated at room temperature for 10 min according to the manufacturer’s instruc tions. To stop the color development, 100 pi o f 1 M H2S 0 4 solution or TMB stop solution (KPL) was added to each well, and the color development was measured at 450 nm. PBS (1 x , pH 7.6) and heat-killed E. coli 0157:H 7 (KPL) were used as a negative and a positive controls, respectively, for each assay.
Y
antibody
Streptavidin-Cy3
FIGURE 1. IMB-based sandwich immunoassay with tyramide signal amplification (TySA). TySA can increase the sensitivity o f the assay by providing more binding sites fo r streptavidin-Cy3. then added to each well. The plate was incubated again for 30 min at room temperature with gentle agitation, and the Cy3-labeled IM B-E. coli 0157:H 7 complexes were then washed three times with 150 pi of TNT buffer and resuspended in 150 pi o f TNT buffer for detection of the fluorescent signal. Figure 1 illustrates the protocol used for the IMB-based sandwich assay with TySA.
Fluorescence signal measurement and data analysis. Fluorescence from Cy3 was measured with a multidetection microplate reader (POLARstar OPTIMA, BMG Labtech, Durham, NC) or a multimode microplate reader (Synergy HT, Biotek, Winooski, VT) with excitation and emission wavelengths of 544 and 590 nm, respectively. The data were recorded in the attached computer and exported to MS Excel (Microsoft, Redmond, WA) for further analysis. The average fluorescent signal from the negative control was used as a background signal, and a net signal for each sample was calculated by subtracting the background signal from its fluorescent signal. In this study, a positive signal was defined as any net fluorescent signal greater than three times the standard deviation of the background signal. Detection limits were estimated from the equation obtained from the dose-response curves drawn for sensitivity tests. The detection limit was defined as the cell level when the net fluorescent signal was equal to the positive signal threshold.
ELISA. In this study, the ELISA was conducted with and without modification. Assay conditions for both standard and modified ELISAs including incubation time and temperature, and concentrations of antibodies and substrates were optimized before running the assay. For the standard ELISA, each well of a microtiter plate was coated with 50 pi of 0.5 pg/ml anti-E. coli 0 1 5 7 antibodies (KPL) overnight at 4°C or for 2 h at room temperature. For the modified ELISA, which was performed in a manner similar to that of the IMB-based fluorescent assay, 3 pi of IMBs coated with anti—E. coli 0 1 5 7 antibodies (Life Technolo gies) was added to each well o f a microtiter plate for antibody immobilization. After antibody immobilization by either direct coating (standard ELISA) or employing IMBs (modified ELISA), 100 pi of test sample was added to each well, and the plate was incubated at room temperature for 45 min with gentle agitation. The IMB-bacteria complexes were then washed three times for 3 min each with 150 pi of TNT, and 100 pi o f 1% bovine serum albumin (10 mg/ml in 1 x PBS) was added to each well. The plate was incubated at room temperature for 30 min with agitation and
RESULTS AND DISCUSSION Specificity of IMB-based fluorescence assay. The specificity of the developed assay (IMB-TySA immunoas say) was validated by testing the assay with four strains of E. coli 0157:H7 and other foodbome pathogens, including non0157:H7 STEC strains. All samples were prepared as described above, and the number of viable cells was determined by plate counting. All E. coli 0157:H7 strains were tested at 106 CFU/ml, and all other pathogens were tested at a higher level of 107 CFU/ml. PBS (1 x , pH 7.6) and heat-killed E. coli 0157:H7 were also used in the assay as negative and positive controls, respectively. The results are shown in Table 1. The IMB-TySA immunoassay also was tested with 74 E. coli 0157:H7 strains isolated from cattle farms, and all strains were successfully identified (data not shown). All E. coli 0157:H7 strains tested in the present study were detected with a positive signal; in contrast, none of the 15 nontarget strains cross-reacted with the IMB-TySA immunoassay. The positive threshold was defined as a net fluorescent signal equal to three times the standard deviation of the background. The background signal was obtained from the fluorescence signal of the negative control. The results indicate that the IMB-TySA immunoassay can identify the target organism E. coli 0157:H7 without cross-reactivity with other common foodbome pathogens, including non0157:H7 STEC or common enteric bacteria. In this study, commercial IMBs with anti-E. coli 0157 antibodies were used for capturing the targets, and polyclonal antibodies against E. coli 0157:H7 were used as detection antibodies. The results of the specificity tests suggest that the pairing of anti-E. coli 0157 and anti-E. coli 0157:H7 antibodies allows differentiation of the 0157:H7 serotype from other STEC serotypes. The specificity of the IMB-TySA immuno assay might be enhanced even further by employing different types of capture antibodies such as anti-H7 or anti-E. coli 0157:H7 antibodies. Sensitivity of IMB-based fluorescence assay. The IMB-TySA immunoassay was validated with bacterial cultures to evaluate the feasibility of the assay in food safety testing. The sensitivity of the IMB-TySA immunoassay also was compared with that of an ELISA, which is the most commonly used immunoassay for pathogen detection. In this study, the ELISA was mn with and without modification. For the modified ELISA (IMS-ELISA), IMS was employed instead of directly immobilizing antibodies on the microplate
DETECTION OF E. COL1 0157:H7 IN FOODS USING IMMUNOASSAY
J. Food Prot., Vol. 77, No. 1
TABLE 2. Sensitivity of IMB-TySA immunoassay compared with ELlSAa Cell level (CFU/ml)
10° 101 102 103 104 105 106
Net fluorescence signal for IMB-TySA immunoassay
[87 [145 430 1,800 4,300 7,900 10,050
± 60] ± 100] ± 150 ± 340 + 580 ± 1,031 + 1,200
Net absorbance signal ELISA
[0.01 [0.03 [0.05 [0.11 0.24 0.52 0.81
IMS-ELISA
+ 0.01] [0.01 ± 0.01] ± 0.03] [0.04 + 0.03] + 0.04] ± + + ±
0.05] 0.07 0.10 0.12
[0.09 0.21 0.47 0.79 0.98
0.02] 0.05 0.09 0.10 + 0.13
+ ± + ±
° Values in brackets are below the cutoff for a positive signal, which was defined as greater than or equal to three times the standard deviation of the background signal. The standard deviation of the background signal was 78 for the IMB-TySA immunoassay and 0.04 for both ELISAs, and the threshold limits for a positive signal were calculated as 234 (3 x 78) and 0.12 (3 x 0.04), respectively.
as occurs in the standard ELISA. This way, we were able to separately determine the effects on sensitivity of two important assay components: (i) IMS of the target organism and (ii) TySA. Detection limits of E. coli 0157:H 7 in these assays were determined based on analysis of serial dilutions (1 to 106 CFU/ml) of overnight cultures of E. coli 0157:H7. Results from the IMB-TySA immunoassay and both the standard and modified ELISAs are presented in Table 2. Dose-response curves and dynamic ranges for detection of E. coli 0157:H 7 from the IMB-TySA immunoassay with and without a 3-h enrichment step are shown in Figure 2. The positive threshold was defined as the net fluorescent signal equal to three times the standard deviation of the background as described above. Detection limits of E. coli 0157:H 7 in the IMB-TySA immunoassay with and without enrichment were
Cell Numbers (logN CFU/ml)
103
estimated from dose-response curves (Fig. 2) to be 5 and 50 CFU/ml, respectively. This 10-fold increase in sensitivity with the addition of a 3-h enrichment step suggests that the IMB-TySA immunoassay can detect a lower number of cells after a longer enrichment period. When enrichment was employed, the intensity of the signal reached saturation at high levels of E. coli 0157:H 7 (> 1 0 5 CFU/ml). This hook effect commonly occurs in sandwich assays in the presence of high levels of analyte (13). The detection limit of the IMBTySA immunoassay for E. coli 0157:H7 (50 CFU/ml without enrichment) also was compared with the limits of the standard ELISA and IMS-ELISA. These limits were also estimated from dose-response curves (not shown) plotted using data from Table 2 and were 1,260 and 200 CFU/ml, respectively. These results indicate that the IMB-TySA immunoassay is 25 times and 4 times more sensitive than the standard ELISA and the IMS-ELISA, respectively. The detection limit of the ELISA decreased more than six times when IMS was incorporated into the standard ELISA protocol. These data indicate that both TySA and IMS contributed to the higher sensitivity of the IMB-TySA immunoassay. In the dose-response curves obtained from the assay, we observed a dynamic range of three orders of magnitude, between 102 and 105 CFU/ml and between 103 and 106 CFU/ml with and without enrichment, respectively. This finding is in agreement with the finding reported previously that well-performing sandwich assays with a multiwell plate format or other array format should exhibit a dynamic range of two to three orders of magnitude (24). The linear regression equations obtained were Y = 2955.2x — 5121.2 (R2 = 0.996) for enriched samples and Y = 2835x — 6745 (R2 = 0.991) for nonenriched samples. R 2 values of dynamic ranges, which are close to 1.0, indicate that the IMBTySA immunoassay is highly reliable and could be used for quantification of a target pathogen in a limited range of levels. Various antibody-based methods for detection of E. coli 0157:H 7 have been reported in the past few years. In one
Cell Numbers (logN CFU/ml)
FIGURE 2. Dose-response curves (A) and dynamic ranges (B) of the developed E. coli 0157:H7 detection assay with and without enrichment. (A) Solid and dotted lines represent the polynomial curve fits for dose-responses of the assay with and without enrichment, respectively. Dashed horizontal line is the positive threshold of the assay, defined as a net signal equal to three times the standard deviation of the background signal. (B) Solid and dotted lines represent the linear fit for the signal intensity with and without enrichment with R2 values of 0.996 and 0.991, respectively. Each data point represents the average of triplicate measurements.
104
AYDIN ET AL.
Cell Numbers (logN CFU/ml) FIGURE 3. Dose-response cuiyes from the IMB-TySA immuno assay fo r E. coli 0157:H7 in pure culture (9 , dashed line), ground beef (M, solid line), and milk ( 4 , dotted line). Solid horizontal line is the positive threshold o f the assay, defined as a net signal equal to three times the standard deviation o f the background signal. Each data point represents the average o f triplicate measurements.
recent study of a fluorescence immunoassay combined with IMS, the detection limit was 10 CFU/ml for E. coli 0157:H7 (33). In another study employing an amperometric immunosensor, the detection limit for E. coli 0157:H7 was approximately 250 CFU/ml (20). Two recently developed bead-based immunoassays had E. coli 0157:H7 detection limits of 10 to 50 CFU/ml (11, 31). Qi et al. (27) reported that an immunochromatography assay combined with nanoparticles could detect E. coli 0157:H7 at as low as 1,000 CFU/ml. Based on these previous reports, we suggested that our IMB-TySA immunoassay has higher sensitivity than most immunoassays for detection of E. coli 0157:H7 and the detection limit is comparable to that of the most sensitive immunoassay reported. In addition to high sensitivity, the IMB-TySA immunoassay has additional advantages over other immunoassays: it is simple to run without expensive equipment or reagents, and it can facilitate sample preparation and target concentration by employing IMBs. Detection of E. coli 0157:H7 in food samples. The performance of the IMB-TySA immunoassay was evaluated with ground beef and milk samples spiked with E. coli 0157:H7 at 1 to 106 CFU/ml. All food samples were tested without enrichment. Dose-response curves obtained from ground beef and milk samples spiked with E. coli 0157:H7 are shown in Figure 3 and compared with the dose-response curve from pure E. coli 0157:H7 culture. Detection limits of E. coli 0157:H7 in ground beef and milk samples were estimated from dose-response curves to be 250 and 100 CFU/ml, respectively, i.e., five- and twofold increases from the 50 CFU/ml achieved with pure culture of E. coli 0157:H7. Signal intensities decreased by as much as 52 and
J. Food Prot., Vol. 77, No. 1
48% in ground beef and milk samples, respectively, relative to those obtained from the pure E. coli 0157:H7 culture. The results from the IMB-TySA immunoassay also were compared with those from the IMS-ELISA. The IMSELISA detection limits were 1.4 x 103 CFU/g in ground beef and 1.1 x 103 CFU/ml in milk, respectively (data not shown). The decrease in sensitivity observed in both assays can be explained by the effects of the complex matrices of ground beef and milk. A similar decrease in sensitivity or signal intensity in food samples has been observed by others. Reinders et al. (28) detected E. coli 0157:H7 in milk samples using IMS methods with a sensitivity of 105 CFU/ml. They found that the sensitivity did not change regardless of the presence of milk in the samples, but the signal was lower in the presence of milk. In another study using immunochromatography with immunomagnetic nanoparticles to detect E. coli 0157:H7, Shen et al. (30) noted that the sensitivity of the assay was reduced by 100fold in food samples, including milk and beef, to a detection limit of 105 CFU/ml, whereas the limit was 103 CFU/ml in culture. When Shen et al. (30) used quartz crystal microbalance immunosensors the detection limits were 23 and 53 CFU/ml for E. coli 0157:H7 in PBS and milk, respectively, which indicates the sensitivity of the assay decreased by more than twofold in food samples. From the comparison with other studies, our results indicate that even in complex food samples the IMBTySA immunoassay is able to detect E. coli 0157:H7 with only moderate loss of sensitivity, and the sensitivity achieved is still higher or comparable to that of the most sensitive immunoassays previously reported. In this study, an IMB-based immunoassay combined with TySA was developed for rapid and sensitive detection of E. coli 0157:H7. E. coli 0157:H7 is a primary concern for the food industry, and rapid and sensitive detection of this pathogen is very important to ensure food safety. The IMB-TySA immunoassay was able to detect E. coli 0157:H7 with a detection limit of 50 CFU/ml within 3 h without enrichment with a dynamic range of at least three orders of magnitude. The sensitivity of the assay was further enhanced 10-fold after a 3-h enrichment step. This immunoassay could also be applied to food samples with detection limits of 250 and 100 CFU/ml in ground beef and milk, respectively, suggesting only a moderate loss of sensitivity. The IMB-TySA immunoassay was very specific for E. coli 0157:H7, and no cross-reactivity was observed for other non-0157:H7 STEC or common foodborne pathogens. This assay still has limitation of not being able to discriminate live cells from dead cells; however, it will be a valuable tool for highly sensitive and specific screening of a large number of samples before the additional techniques are employed to verify the viability of the target pathogen. The results from this study indicate this IMB-based assay combined with TySA can be a simple, sensitive, and rapid system for detection of foodborne pathogens. The IMBTySA immunoassay also could be used to simultaneously detect multiple pathogens when used with beads internally coded with addressing dyes. This assay is a rapid and sensitive alternative to current pathogen screening systems.
J. Food Prot., Vol. 77, No. 1
DETECTION OF E. COU 0157:H7 IN FOODS USING IMMUNOASSAY
ACKNOWLEDGMENTS The authors are grateful to Dr. Michael Johnson (Department of Food Science, University of Arkansas, Fayetteville) and Dr. Michelle Danyluk (Department of Food Science and Human Nutrition, University of Florida, Gainesville) for providing all STEC and other pathogen strains used in this study.
17.
18.
REFERENCES 1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15. 16.
Andreoli, S. P., H. Trachtman, D. W. Acheson, R. L. Siegler, and T. G. Obrig. 2002. Hemolytic uremic syndrome: epidemiology, pathophysiology, and therapy. Pediatr. Nephrol. 17:293-298. Anklam, K. S., K. S. Kanankege, T. K. Gonzales, C. W. Kaspar, and D. Dopfer. 2012. Rapid and reliable detection of Shiga toxinproducing Escherichia coli by real-time multiplex PCR. / . Food Prot. 12:643-650. Benoit, P. W., and D. W. Donahue. 2003. Methods for rapid separation and concentration of bacteria in food that bypass timeconsuming cultural enrichment. J. Food Prot. 66:1935—1948. Bettleheim, K. A. 1998. Reliability of CHROMagar 0157 for the detection of enterohaemorrhagic Escherichia coli (EHEC) but not EHEC belonging to other serogroups. J. Appl. Microbiol. 85:425428. Bettleheim, K. A. 1998. Studies of Escherichia coli cultured on rainbow agar 0157 with particular reference to enterohaemorrhagic Escherichia coli (EHEC). Microbiol. Immunol. 42:265-269. Bobrow, M. N„ K. J. Shaughnessy, and G. J. Lift. 1991. Catalyzed reported deposition, a novel method of signal amplification. II. Application to membrane immunoassays. J. Immunol. Methods 137: 103-112. Brouns, I., L. Van Nassauw, J. Van Genechten, M. Majewski, D. W. Scheuermann, J. Timmermans, and D. Adriaensen. 2002. Triple immunofluorescence staining with antibodies raised in the sample species to study the complex innervation pattern of intrapulmonary chemoreceptors. J. Histochem. Cytochem. 50:575-582. Centers for Disease Control and Prevention. 2007. Laboratoryconfirmed non-0157 Shiga toxin-producing Escherichia coli— Connecticut, 2000-2005. Morb. Mortal. Wkly. Rep. 56:29-31. Centers for Disease Control and Prevention. 2009. OutbreakNet foodbome outbreak online database. Available at: http://wwwn.cdc. gov/foodbomeoutbreaks/. Accessed 25 May 2013. Centers for Disease Control and Prevention. 2009. Outbreak of Shiga toxin-producing Escherichia coli 0157 infection associated with a day camp petting zoo—Pinellas Country, Florida, May-June 2007. Morb. Mortal. Wkly. Rep. 58:426-428. Clotilde, L. M., C. Bernard IV, G. L. Hartman, D. K. Lau, and J. M. Carter. 2011. Microbead-based immunoassay for simultaneous detection of Shiga toxins and isolation of Escherichia coli 0157 in foods. J. Food Prot. 74:373-379. Favrin, S. J., S. A. Jassim, and M. W. Griffiths. 2003. Application of a novel immunomagnetic separation-bacterophage assay for the detection of Salmonella Enteritidis and Escherichia coli 0157:H7 in food. Int. J. Food Microbiol. 85:63—71. Fernando, S. A., and G. S. Wilson. 1992. Studies of the “ hook” effect in the one-step sandwich immunoassay. J. Immunol. Methods 151:47-66. Gould, L. H., C. Bopp, N. Strockbine, R. Atkinson, V. Baselki, B. Body, R. Carey, C. Crandall, S. Hurd, R. Kaplan, M. Neill, S. Shea, P. Somsei, M. Tobin-D’Angelo, P. M. Griffin, P. Gemer-Smidt, and Centers for Disease Control and Prevention. 2009. Recommendations for diagnosis of Shiga toxin-producing Escherichia coli infections by clinical laboratories. Morb. Mortal. Wkly. Rep. Recomm. Rep. 58:114. Hunt, J. M. 2010. Shiga toxin-producing Escherichia coli (STEC). Clin. Lab. Med. 30:21-45. Hussein, H. S., and L. M. Bollinger. 2008. Influence of selective media on successful detection of Shiga toxin-producing Escherichia coli in food, fecal, and environmental samples. Foodbome Pathog. Dis. 5:227-244.
19.
20.
21.
22.
23.
24. 25.
26.
27.
28.
29.
30.
31.
32.
33.
105
Islam, M., M. P. Doyle, S. C. Phatak, P. Millner, and X. Jiang. 2004. Persistence of enterohemorrhagic Escherichia coli 0157:H7 in soil and on leaf lettuce and parsley grown in fields treated with contaminated manure composts or irrigation water. J. Food Prot. 67:1365-1370. Karsten, S. L., V. M. D. Van Deerlin, C. Sabatti, L. H. Gill, and D. H. Geschwind. 2002. An evaluation of tyramide signal amplification and archived fixed and frozen tissue in microarray gene expression analysis. Nucleic Acids Res. 30:E4. Lahti, E. M., M. Eklund, P. Ruutu, A. Siitonen, L. Rantala, P. Nuorti, and T. Honkanen-Buzalski. 2002. Use of phenotyprng and genotyping to verify transmission of Escherichia coli 0157:H7 from dairy farms. Eur. J. Clin. Microbiol. Infect. Dis. 21:189-195. Li, Y., P. Cheng, J. Gong, L. Fang, J. Deng, W. Liang, and J. Zheng. 2012. Amperometric immunosensor for the detection of Escherichia coli 0157:H7 in food specimens. Anal. Biochem. 421:227-233. Lorusso, V., A. Dambrosio, N. C. Quaglia, A. Parisi, G. Lasalandra, G. Mula, S. Virgilio, G. Lucifora, M. Dario, and G. Normanno. 2011. Development of a multiplex PCR for rapid detection of verocytotoxin-producing Escherichia coli 026 in raw milk and ground beef. J. Food Prot. 74:13-17. Luk, C., G. Giovannoni, D. R. Williams, A. J. Lees, and R. de Silva. 2009. Development of a sensitive ELISA for quantification of threeand four-repeat tau isoforms in tauopathies. J. Neurosci. Methods 180:34-42. March, S. B., and S. Ratnam. 1986. Sorbitol-MacConkey medium for detection of Escherichia coli 0 157 :H7 associated with hemorrhagic colitis. J. Clin. Microbiol. 23:869-872. Nielsen, U. B., and B. H. Geierstanger. 2004. Multiplexed sandwich assay in microarray format. J. Immunol. Methods 290:107-120. O ’Brien, S. J., G. K. Adak, and C. Gilham. 2001. Contact with farming environment as a major risk factor for Shiga toxin (verocytotoxin)-producing Escherichia coli 0157 infection in humans. Emerg. Infect. Dis. 7:1049-1051. Olsen, S. J., G. Miller, M. Kennedy, C. Higgins, J. Walford, G. McKee, K. Fox, W. Bibb, and P. Mead. 2002. A waterborne outbreak of Escherichia coli 0157:H7 infections and hemolytic uremic syndrome: implications for rural water systems. Emerg. Infect. Dis. 8:370-375. Qi, H., Z. Zhong, H. X. Zhou, C. Y. Deng, H. Zhu, J. F. Li, X, L. Wang, and F. R. Li. 2011. A rapid and highly sensitive protocol for the detection of Escherichia coli 0157:H7 based on immunochromatography assay combined with the enrichment technique of immunomagnetic nanoparticles. Int. J. Nanomed. 6:3033-3039. Reinders, R. D., A. Bama, L. J. A. Lipman, and P. G. H. Bijker. 2002. Comparison of the sensitivity of manual and automated immuno magnetic separation methods for detection of Shiga toxin-producing Escherichia coli 0157:H7 in milk. J. Appl. Microbiol. 92:1015—1020. Scallan, E., R. M. Hoekstra, F. J. Angulo, R. V. Tauxe, M. A. Widdowson, S. L. Roy, J. L. Jones, and P. M. Griffin. 2011. Foodbome illness acquired in the United States— major pathogens. Emerg. Infect. Dis. 17:7-15. Shen, Z. Q„ J. F. Wang, Z. G. Qiu, M. Jin, X. W. Wang, Z. L. Chen, J. W. Li, and F. H. Cao. 2011. QCM immunosensor detection of Escherichia coli 0157:H7 based on beacon immunomagnetic nanoparticles and catalytic growth of colloidal gold. Biosens. Bioelectron. 26:3376-3381. Wang, H., Y. Li, A. Wang, and M. Slavik. 2011. Rapid, sensitive, and simultaneous detection of three foodbome pathogens using magnetic nanobead-based immunoseparation and quantum dot—based multiplex immunoassay. J. Food Prot. 74:2039-2047. Zaidi, A. U., H. Enomoto, J. Milbrandt, and K. A. Roth. 2000. Dual florescent in situ hybridization and immunohistochemical detection with tyramide signal amplification. J. Histochem. Cytochem. 48: 1369-1375. Zhu, P„ D. R. Shelton, S. Li, D. L. Adams, J. S. Kams, P. Amstutz, and C. M. Tang. 2011. Detection of E. coli 0157:H7 by immunomagnetic separation coupled with fluorescence immunoas say. Biosens. Bioelectron. 30:337-341.
Copyright of Journal of Food Protection is the property of Allen Press Publishing Services Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Journal o f Food Protection, Vol. 77, No. 1, 2014, Pages 100-105 doi: 10.4315/0362-028X.JFP-13-274 Copyright © , International Association for Food Protection
Research Note
Rapid and Sensitive Detection of Escherichia coli 0157:H7 in Milk and Ground Beef Using Magnetic Bead-Based Immunoassay Coupled with Tyramide Signal Amplification MUHSIN AYDIN,1 GENE P. D. HERZIG,2 KWANG CHEOL JEONG,3 SAMANTHA DUNIGAN,2 PARTH SHAH,4 a n d SOOHYOUN AHN5* 1Department o f Molecular Biosciences and departm ent o f Biological Sciences, Arkansas State University, Jonesboro, Arkansas 72401; 3Emerging Pathogens Institute and Department o f Animal Sciences and 5Department o f Food Science and Human Nutrition, Institute o f Food and Agricultural Sciences, University o f Florida, Gainesville, Florida 32611; and 4Department o f Cell Biology and Neuroscience, Rutgers University, New Brunswick, New Jersey 07102, USA MS 13-274: Received 2 July 2013/Accepted 11 September 2013
ABSTRACT Escherichia coli 0157:H7 is a major foodbome pathogen that has posed serious problems for food safety and public health. Recent outbreaks and recalls associated with various foods contaminated by E. coli 0157:H7 clearly indicate its deleterious effect on food safety. A rapid and sensitive detection assay is needed for this harmful organism to prevent foodbome illnesses and control outbreaks in a timely manner. We developed a magnetic bead-based immunoassay for detection of E. coli Oi57:H7 (the most well-known Shiga toxigenic E. coli strain) with a 96-well microplate as an assay platform. Immunomagnetic separation (IMS) and tyramide signal amplification were coupled to the assay to increase its sensitivity and specificity. This immunoassay was able to detect E. coli 0157:H7 in pure culture with a detection limit of 50 CFU/ml in less than 3 h without an enrichment step. The detection limit was decreased 10-fold to 5 CFU/ml with addition of a 3-h enrichment step. When this assay was tested with other nontarget foodbome pathogens and common enteric bacteria, no cross-reactivity was found. When tested with artificially contaminated ground beef and milk samples, the assay sensitivity decreased two- to fivefold, with detection limits of 250 and 100 CFU/ml, respectively, probably because of the food matrix effect. The assay results also were compared with those of a sandwich-type enzyme-linked immunosorbent assay (ELISA) and an ELISA coupled with IMS; the developed assay was 25 times and 4 times more sensitive than the standard ELISA and the IMS-ELISA, respectively. Tyramide signal amplification combined with IMS can improve sensitivity and specificity for detection of E. coli 0157:H7. The developed assay could be easily adapted for other foodbome pathogens and will contribute to improved food safety and public health.
Shiga toxin-producing Escherichia coli (STEC) strains are major foodbome pathogens and have caused a large number of human illnesses. Among STEC strains associated with human infections, E. coli 0157:H7 is the most wellknown serotype and is responsible for 89 and 100% of STEC-related hospitalizations and deaths, respectively, in the United States (29). E. coli 0157:H7 causes diarrhea with or without blood, hemorrhagic colitis, and in more severe cases hemolytic uremic syndrome (15). The Centers for Disease Control and Prevention estimated that E. coli 0157:H7 causes approximately 63,000 illnesses, 2,100 hospitalizations, and 20 deaths annually in the United States (29). Most E. coli 0157:H7 infection outbreaks have been linked to the consumption of ground beef (9). However, human infections by other routes also have been reported, including the consumption of contaminated vegetables, drinking water, milk, and dairy products and direct contact with animals during visits to farms or petting * Author for correspondence. Tel: 352-392-1991; Fax: 352-392-9467; E-mail: [email protected].
zoos (10, 17, 19, 25, 26). Considering the impact of E. coli 0157:H7 on food safety and public health, rapid and sensitive detection methods for this pathogen in food and environmental samples are needed to monitor contamination of food products and to control outbreaks in a timely manner (14). Culture methods involving sorbitol-MacConkey agar or other selective agars such as CHROMagar and Rainbow agar are most commonly used for E. coli 0157:H7 detection (4,5, 23). These methods are also frequently combined with serological tests using antibodies to the 0157 and H7 antigens to confirm the presence of E. coli 0157:H7. However, these conventional methods are time-consuming and laborious. Most culture methods are based on screening of the non-sorbitol-fermenting 0157 serotype, and some traditional enteric media can even inhibit the growth of non0157 STEC (1, 8, 14, 16). Therefore, modification of these culture methods to simultaneously detect both E. coli 0157:H7 and other non-0157 STEC strains is difficult. Antibody and nucleic acid methods also have been developed and tested for detection of various STEC strains
J. Food Prot., Vol. 77, No. 1
101
DETECTION OF E. COU 0157:H7 IN FOODS USING IMMUNOASSAY
(2, 21, 33). These alternative methods could provide low
TABLE 1. Cross-reactivity of 1MB-TySA immunoassay with
detection limits, high specificity, and rapid processing; however, most of these methods have the disadvantages of cross-reactivity, limited multiplexing ability, and the presence of inhibitory substances in the sample. The sensitivity of these alternative methods also is highly affected by sample preparation, mostly because of interference from the sample matrix and low levels of the target pathogen mixed with high levels of background microflora. To develop a sensitive detection method, a sample preparation method is needed that can concentrate the target pathogens and isolate them from background microflora with minimal effect from the sample matrix. Immunomagnetic separation (IMS) has been used to effectively isolate and concentrate target organisms from complex matrices in food, clinical, and environmental samples (12, 31). IMS can reduce the total assay time and provide higher sensitivity for pathogen detection in complex food or environmental samples (3). Tyramide signal amplification (TySA), also known as catalyzed reporter deposition, is a signal amplification technique based on horseradish peroxidase (HRP)-catalyzed deposition of tyramide conjugates (6). This method has been used to increase detection sensitivity by amplifying chromogenic or fluorescent signals in many applications including immunohistochemistry, in situ hybridization, electron microscopy, and immunoassays (7, 18, 22, 32). Although TySA has been used in immunoassays for detection of protein markers in medical studies, its use in food safety applications for detecting foodbome pathogens has not been reported. In this study, we combined the IMS and TySA techniques to detect E. coli 0157:H7 in food and found that this approach was highly sensitive and minimally affected by other components in the food. The goal of this study was to develop a novel immunoassay based on IMS and TySA techniques for more rapid and sensitive detection of E. coli 0157:H7.
common foodbome pathogens
M A TER IA LS A N D M E TH O D S Microbial culture and growth conditions. E. coli 0157:H7 ATCC 43890 was used as the main test strain in this study. Other bacterial strains used are listed in Table 1. All non-0157:H7 STEC strains were obtained from the Citrus Research and Education Center (University of Florida, Lake Alfred), and all other strains (including E. coli 0157:H7 43890) were obtained from the University of Arkansas (Fayetteville). Culture of each bacterial strain was grown hi tryptic soy broth (BD, Franklin Lakes, NJ) at 37°C for 18 h. Cells were collected from 1 ml of culture by centrifugation (17,000 x g for 5 min) and resuspended in 1 ml of phosphate-buffered saline (PBS; 1 x , pH 7.6). The number of viable cells was determined by plate counting on MacConkey agar (Remel, Lenexa, KS). Food sample preparation. Ground beef and milk were purchased from local retailers 24 h before each experiment. Ground beef samples (25 g each) were mixed with 225 ml of 1 x PBS (pH 7.6) and homogenized with a stomacher for 1 min. The solutions were collected and then inoculated with the E. coli 0157:H7 culture. Milk samples were directly inoculated with E. coli 0157:H7 culture without dilution. The level of E. coli 0157:H7 in the food samples was adjusted to 1 to 106 CFU/ml.
Tested pathogen0
Detection results6
E. coli 0157:H7 (ATCC 43888) E. coli 0157:H7 (ATCC 43889) E. coli 0157:H7 (ATCC 43890) E. coli 0157:H7 (ATCC 43895) E. coli JM 109 (lab strain) E. coli O103:H2 E. coli O104:H4 E. coli 0145 E. coli 0111 Citrobacter freundii Enterobacter aerogenes Staphylococcus aureus S. epidermidis Listeria monocytogenes Salmonella Enteritidis Salmonella Typhimurium Salmonella Infantis Shigella dysenteriae S. flexneri
+ + + + — — —
“ All E. coli 0157:H7 strains were tested at 106 CFU/ml; the other pathogens were tested at 107 CFU/ml. b + , positive signal, i.e., greater than or equal to three times the standard deviation of the background signal; —, negative signal, i.e., less than three times the standard deviation of the background signal. The samples were used for assay immediately after inoculation without further incubation. Immunomagnetic bead-based immunofluorescence assay. The immunofluorescence assay with the immunomagnetic beads (IMBs) consisted of two reaction steps: (i) IMS and (ii) fluorescence immunoassay with TySA. For each assay, 1 x PBS (pH 7.6) and heat-killed E. coli 0157:H7 (KPL, Gaithersburg, MD) at 3 x 106 CFU/ml were added as a negative control and a positive control, respectively. For IMS, each well of a 96-well microtiter plate contained 3 pi of IMBs coated with anti-£. coli 0157 antibodies (Life Technologies, Grand Island, NY) and 100 pi of test sample. Plates were incubated at room temperature for 30 min with gentle agitation on a digital microplate shaker (IKA, Wilmington, NC). The I MB-/.’, coli 0157:H7 complexes were then separated from the solution with a magnet (DynaMag-96, Life Technologies) according to the manufacturer’s instructions, and the plates were washed three times with 150 pi of TNT buffer (0.1 M Tris-HCl, 0.15 M NaCl, and 0.05% Tween 20, pH 7.5). For the fluorescence immunoassay with TySA, 100 pi of 0.1 mg/ml HRP-labeled anti-£. coli 0157:H7 antibody (KPL) was added to the IMB-E. coli 0157:H7 complexes in each well. The plate was incubated at room temperature for 30 min with agitation and then washed three times with 150 pi of TNT buffer. TySA was performed using an ELAST ELISA (enzyme-linked immunosor bent assay) amplification system (Perkin Elmer, Waltham, MA) according to the manufacturer’s instructions. Biotinyl tyramide solution diluted to 50 pi with amplification diluent from the ELAST ELISA amplification system was added to each well. The plate was incubated for 10 min at room temperature with agitation and washed three times with 150 pi of TNT buffer, and 100 pi of 10 pg/ml streptavidin-Cy3 (Sigma-Aldrich, St. Louis, MO) was
102
AYDIN ET AL.
J. Food Prot., Vol. 77, No. 1
then washed three times with 150 j.il of TNT buffer. After washing, 100 (il o f HRP-labeled antibody (0.1 mg/ml) against E. coli 0157:H 7 (KPL) was added to each well. The plate was incubated at room temperature for 1 h with agitation and then washed with 150 pi o f TNT buffer. After washing, 100 p! of 3,3',5,5'tetramethylbenzidine (TMB) substrate solution (0.4 mg/ml; KPL) was added to each well, and the plate was incubated at room temperature for 10 min according to the manufacturer’s instruc tions. To stop the color development, 100 pi o f 1 M H2S 0 4 solution or TMB stop solution (KPL) was added to each well, and the color development was measured at 450 nm. PBS (1 x , pH 7.6) and heat-killed E. coli 0157:H 7 (KPL) were used as a negative and a positive controls, respectively, for each assay.
Y
antibody
Streptavidin-Cy3
FIGURE 1. IMB-based sandwich immunoassay with tyramide signal amplification (TySA). TySA can increase the sensitivity o f the assay by providing more binding sites fo r streptavidin-Cy3. then added to each well. The plate was incubated again for 30 min at room temperature with gentle agitation, and the Cy3-labeled IM B-E. coli 0157:H 7 complexes were then washed three times with 150 pi of TNT buffer and resuspended in 150 pi o f TNT buffer for detection of the fluorescent signal. Figure 1 illustrates the protocol used for the IMB-based sandwich assay with TySA.
Fluorescence signal measurement and data analysis. Fluorescence from Cy3 was measured with a multidetection microplate reader (POLARstar OPTIMA, BMG Labtech, Durham, NC) or a multimode microplate reader (Synergy HT, Biotek, Winooski, VT) with excitation and emission wavelengths of 544 and 590 nm, respectively. The data were recorded in the attached computer and exported to MS Excel (Microsoft, Redmond, WA) for further analysis. The average fluorescent signal from the negative control was used as a background signal, and a net signal for each sample was calculated by subtracting the background signal from its fluorescent signal. In this study, a positive signal was defined as any net fluorescent signal greater than three times the standard deviation of the background signal. Detection limits were estimated from the equation obtained from the dose-response curves drawn for sensitivity tests. The detection limit was defined as the cell level when the net fluorescent signal was equal to the positive signal threshold.
ELISA. In this study, the ELISA was conducted with and without modification. Assay conditions for both standard and modified ELISAs including incubation time and temperature, and concentrations of antibodies and substrates were optimized before running the assay. For the standard ELISA, each well of a microtiter plate was coated with 50 pi of 0.5 pg/ml anti-E. coli 0 1 5 7 antibodies (KPL) overnight at 4°C or for 2 h at room temperature. For the modified ELISA, which was performed in a manner similar to that of the IMB-based fluorescent assay, 3 pi of IMBs coated with anti—E. coli 0 1 5 7 antibodies (Life Technolo gies) was added to each well o f a microtiter plate for antibody immobilization. After antibody immobilization by either direct coating (standard ELISA) or employing IMBs (modified ELISA), 100 pi of test sample was added to each well, and the plate was incubated at room temperature for 45 min with gentle agitation. The IMB-bacteria complexes were then washed three times for 3 min each with 150 pi of TNT, and 100 pi o f 1% bovine serum albumin (10 mg/ml in 1 x PBS) was added to each well. The plate was incubated at room temperature for 30 min with agitation and
RESULTS AND DISCUSSION Specificity of IMB-based fluorescence assay. The specificity of the developed assay (IMB-TySA immunoas say) was validated by testing the assay with four strains of E. coli 0157:H7 and other foodbome pathogens, including non0157:H7 STEC strains. All samples were prepared as described above, and the number of viable cells was determined by plate counting. All E. coli 0157:H7 strains were tested at 106 CFU/ml, and all other pathogens were tested at a higher level of 107 CFU/ml. PBS (1 x , pH 7.6) and heat-killed E. coli 0157:H7 were also used in the assay as negative and positive controls, respectively. The results are shown in Table 1. The IMB-TySA immunoassay also was tested with 74 E. coli 0157:H7 strains isolated from cattle farms, and all strains were successfully identified (data not shown). All E. coli 0157:H7 strains tested in the present study were detected with a positive signal; in contrast, none of the 15 nontarget strains cross-reacted with the IMB-TySA immunoassay. The positive threshold was defined as a net fluorescent signal equal to three times the standard deviation of the background. The background signal was obtained from the fluorescence signal of the negative control. The results indicate that the IMB-TySA immunoassay can identify the target organism E. coli 0157:H7 without cross-reactivity with other common foodbome pathogens, including non0157:H7 STEC or common enteric bacteria. In this study, commercial IMBs with anti-E. coli 0157 antibodies were used for capturing the targets, and polyclonal antibodies against E. coli 0157:H7 were used as detection antibodies. The results of the specificity tests suggest that the pairing of anti-E. coli 0157 and anti-E. coli 0157:H7 antibodies allows differentiation of the 0157:H7 serotype from other STEC serotypes. The specificity of the IMB-TySA immuno assay might be enhanced even further by employing different types of capture antibodies such as anti-H7 or anti-E. coli 0157:H7 antibodies. Sensitivity of IMB-based fluorescence assay. The IMB-TySA immunoassay was validated with bacterial cultures to evaluate the feasibility of the assay in food safety testing. The sensitivity of the IMB-TySA immunoassay also was compared with that of an ELISA, which is the most commonly used immunoassay for pathogen detection. In this study, the ELISA was mn with and without modification. For the modified ELISA (IMS-ELISA), IMS was employed instead of directly immobilizing antibodies on the microplate
DETECTION OF E. COL1 0157:H7 IN FOODS USING IMMUNOASSAY
J. Food Prot., Vol. 77, No. 1
TABLE 2. Sensitivity of IMB-TySA immunoassay compared with ELlSAa Cell level (CFU/ml)
10° 101 102 103 104 105 106
Net fluorescence signal for IMB-TySA immunoassay
[87 [145 430 1,800 4,300 7,900 10,050
± 60] ± 100] ± 150 ± 340 + 580 ± 1,031 + 1,200
Net absorbance signal ELISA
[0.01 [0.03 [0.05 [0.11 0.24 0.52 0.81
IMS-ELISA
+ 0.01] [0.01 ± 0.01] ± 0.03] [0.04 + 0.03] + 0.04] ± + + ±
0.05] 0.07 0.10 0.12
[0.09 0.21 0.47 0.79 0.98
0.02] 0.05 0.09 0.10 + 0.13
+ ± + ±
° Values in brackets are below the cutoff for a positive signal, which was defined as greater than or equal to three times the standard deviation of the background signal. The standard deviation of the background signal was 78 for the IMB-TySA immunoassay and 0.04 for both ELISAs, and the threshold limits for a positive signal were calculated as 234 (3 x 78) and 0.12 (3 x 0.04), respectively.
as occurs in the standard ELISA. This way, we were able to separately determine the effects on sensitivity of two important assay components: (i) IMS of the target organism and (ii) TySA. Detection limits of E. coli 0157:H 7 in these assays were determined based on analysis of serial dilutions (1 to 106 CFU/ml) of overnight cultures of E. coli 0157:H7. Results from the IMB-TySA immunoassay and both the standard and modified ELISAs are presented in Table 2. Dose-response curves and dynamic ranges for detection of E. coli 0157:H 7 from the IMB-TySA immunoassay with and without a 3-h enrichment step are shown in Figure 2. The positive threshold was defined as the net fluorescent signal equal to three times the standard deviation of the background as described above. Detection limits of E. coli 0157:H 7 in the IMB-TySA immunoassay with and without enrichment were
Cell Numbers (logN CFU/ml)
103
estimated from dose-response curves (Fig. 2) to be 5 and 50 CFU/ml, respectively. This 10-fold increase in sensitivity with the addition of a 3-h enrichment step suggests that the IMB-TySA immunoassay can detect a lower number of cells after a longer enrichment period. When enrichment was employed, the intensity of the signal reached saturation at high levels of E. coli 0157:H 7 (> 1 0 5 CFU/ml). This hook effect commonly occurs in sandwich assays in the presence of high levels of analyte (13). The detection limit of the IMBTySA immunoassay for E. coli 0157:H7 (50 CFU/ml without enrichment) also was compared with the limits of the standard ELISA and IMS-ELISA. These limits were also estimated from dose-response curves (not shown) plotted using data from Table 2 and were 1,260 and 200 CFU/ml, respectively. These results indicate that the IMB-TySA immunoassay is 25 times and 4 times more sensitive than the standard ELISA and the IMS-ELISA, respectively. The detection limit of the ELISA decreased more than six times when IMS was incorporated into the standard ELISA protocol. These data indicate that both TySA and IMS contributed to the higher sensitivity of the IMB-TySA immunoassay. In the dose-response curves obtained from the assay, we observed a dynamic range of three orders of magnitude, between 102 and 105 CFU/ml and between 103 and 106 CFU/ml with and without enrichment, respectively. This finding is in agreement with the finding reported previously that well-performing sandwich assays with a multiwell plate format or other array format should exhibit a dynamic range of two to three orders of magnitude (24). The linear regression equations obtained were Y = 2955.2x — 5121.2 (R2 = 0.996) for enriched samples and Y = 2835x — 6745 (R2 = 0.991) for nonenriched samples. R 2 values of dynamic ranges, which are close to 1.0, indicate that the IMBTySA immunoassay is highly reliable and could be used for quantification of a target pathogen in a limited range of levels. Various antibody-based methods for detection of E. coli 0157:H 7 have been reported in the past few years. In one
Cell Numbers (logN CFU/ml)
FIGURE 2. Dose-response curves (A) and dynamic ranges (B) of the developed E. coli 0157:H7 detection assay with and without enrichment. (A) Solid and dotted lines represent the polynomial curve fits for dose-responses of the assay with and without enrichment, respectively. Dashed horizontal line is the positive threshold of the assay, defined as a net signal equal to three times the standard deviation of the background signal. (B) Solid and dotted lines represent the linear fit for the signal intensity with and without enrichment with R2 values of 0.996 and 0.991, respectively. Each data point represents the average of triplicate measurements.
104
AYDIN ET AL.
Cell Numbers (logN CFU/ml) FIGURE 3. Dose-response cuiyes from the IMB-TySA immuno assay fo r E. coli 0157:H7 in pure culture (9 , dashed line), ground beef (M, solid line), and milk ( 4 , dotted line). Solid horizontal line is the positive threshold o f the assay, defined as a net signal equal to three times the standard deviation o f the background signal. Each data point represents the average o f triplicate measurements.
recent study of a fluorescence immunoassay combined with IMS, the detection limit was 10 CFU/ml for E. coli 0157:H7 (33). In another study employing an amperometric immunosensor, the detection limit for E. coli 0157:H7 was approximately 250 CFU/ml (20). Two recently developed bead-based immunoassays had E. coli 0157:H7 detection limits of 10 to 50 CFU/ml (11, 31). Qi et al. (27) reported that an immunochromatography assay combined with nanoparticles could detect E. coli 0157:H7 at as low as 1,000 CFU/ml. Based on these previous reports, we suggested that our IMB-TySA immunoassay has higher sensitivity than most immunoassays for detection of E. coli 0157:H7 and the detection limit is comparable to that of the most sensitive immunoassay reported. In addition to high sensitivity, the IMB-TySA immunoassay has additional advantages over other immunoassays: it is simple to run without expensive equipment or reagents, and it can facilitate sample preparation and target concentration by employing IMBs. Detection of E. coli 0157:H7 in food samples. The performance of the IMB-TySA immunoassay was evaluated with ground beef and milk samples spiked with E. coli 0157:H7 at 1 to 106 CFU/ml. All food samples were tested without enrichment. Dose-response curves obtained from ground beef and milk samples spiked with E. coli 0157:H7 are shown in Figure 3 and compared with the dose-response curve from pure E. coli 0157:H7 culture. Detection limits of E. coli 0157:H7 in ground beef and milk samples were estimated from dose-response curves to be 250 and 100 CFU/ml, respectively, i.e., five- and twofold increases from the 50 CFU/ml achieved with pure culture of E. coli 0157:H7. Signal intensities decreased by as much as 52 and
J. Food Prot., Vol. 77, No. 1
48% in ground beef and milk samples, respectively, relative to those obtained from the pure E. coli 0157:H7 culture. The results from the IMB-TySA immunoassay also were compared with those from the IMS-ELISA. The IMSELISA detection limits were 1.4 x 103 CFU/g in ground beef and 1.1 x 103 CFU/ml in milk, respectively (data not shown). The decrease in sensitivity observed in both assays can be explained by the effects of the complex matrices of ground beef and milk. A similar decrease in sensitivity or signal intensity in food samples has been observed by others. Reinders et al. (28) detected E. coli 0157:H7 in milk samples using IMS methods with a sensitivity of 105 CFU/ml. They found that the sensitivity did not change regardless of the presence of milk in the samples, but the signal was lower in the presence of milk. In another study using immunochromatography with immunomagnetic nanoparticles to detect E. coli 0157:H7, Shen et al. (30) noted that the sensitivity of the assay was reduced by 100fold in food samples, including milk and beef, to a detection limit of 105 CFU/ml, whereas the limit was 103 CFU/ml in culture. When Shen et al. (30) used quartz crystal microbalance immunosensors the detection limits were 23 and 53 CFU/ml for E. coli 0157:H7 in PBS and milk, respectively, which indicates the sensitivity of the assay decreased by more than twofold in food samples. From the comparison with other studies, our results indicate that even in complex food samples the IMBTySA immunoassay is able to detect E. coli 0157:H7 with only moderate loss of sensitivity, and the sensitivity achieved is still higher or comparable to that of the most sensitive immunoassays previously reported. In this study, an IMB-based immunoassay combined with TySA was developed for rapid and sensitive detection of E. coli 0157:H7. E. coli 0157:H7 is a primary concern for the food industry, and rapid and sensitive detection of this pathogen is very important to ensure food safety. The IMB-TySA immunoassay was able to detect E. coli 0157:H7 with a detection limit of 50 CFU/ml within 3 h without enrichment with a dynamic range of at least three orders of magnitude. The sensitivity of the assay was further enhanced 10-fold after a 3-h enrichment step. This immunoassay could also be applied to food samples with detection limits of 250 and 100 CFU/ml in ground beef and milk, respectively, suggesting only a moderate loss of sensitivity. The IMB-TySA immunoassay was very specific for E. coli 0157:H7, and no cross-reactivity was observed for other non-0157:H7 STEC or common foodborne pathogens. This assay still has limitation of not being able to discriminate live cells from dead cells; however, it will be a valuable tool for highly sensitive and specific screening of a large number of samples before the additional techniques are employed to verify the viability of the target pathogen. The results from this study indicate this IMB-based assay combined with TySA can be a simple, sensitive, and rapid system for detection of foodborne pathogens. The IMBTySA immunoassay also could be used to simultaneously detect multiple pathogens when used with beads internally coded with addressing dyes. This assay is a rapid and sensitive alternative to current pathogen screening systems.
J. Food Prot., Vol. 77, No. 1
DETECTION OF E. COU 0157:H7 IN FOODS USING IMMUNOASSAY
ACKNOWLEDGMENTS The authors are grateful to Dr. Michael Johnson (Department of Food Science, University of Arkansas, Fayetteville) and Dr. Michelle Danyluk (Department of Food Science and Human Nutrition, University of Florida, Gainesville) for providing all STEC and other pathogen strains used in this study.
17.
18.
REFERENCES 1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15. 16.
Andreoli, S. P., H. Trachtman, D. W. Acheson, R. L. Siegler, and T. G. Obrig. 2002. Hemolytic uremic syndrome: epidemiology, pathophysiology, and therapy. Pediatr. Nephrol. 17:293-298. Anklam, K. S., K. S. Kanankege, T. K. Gonzales, C. W. Kaspar, and D. Dopfer. 2012. Rapid and reliable detection of Shiga toxinproducing Escherichia coli by real-time multiplex PCR. / . Food Prot. 12:643-650. Benoit, P. W., and D. W. Donahue. 2003. Methods for rapid separation and concentration of bacteria in food that bypass timeconsuming cultural enrichment. J. Food Prot. 66:1935—1948. Bettleheim, K. A. 1998. Reliability of CHROMagar 0157 for the detection of enterohaemorrhagic Escherichia coli (EHEC) but not EHEC belonging to other serogroups. J. Appl. Microbiol. 85:425428. Bettleheim, K. A. 1998. Studies of Escherichia coli cultured on rainbow agar 0157 with particular reference to enterohaemorrhagic Escherichia coli (EHEC). Microbiol. Immunol. 42:265-269. Bobrow, M. N„ K. J. Shaughnessy, and G. J. Lift. 1991. Catalyzed reported deposition, a novel method of signal amplification. II. Application to membrane immunoassays. J. Immunol. Methods 137: 103-112. Brouns, I., L. Van Nassauw, J. Van Genechten, M. Majewski, D. W. Scheuermann, J. Timmermans, and D. Adriaensen. 2002. Triple immunofluorescence staining with antibodies raised in the sample species to study the complex innervation pattern of intrapulmonary chemoreceptors. J. Histochem. Cytochem. 50:575-582. Centers for Disease Control and Prevention. 2007. Laboratoryconfirmed non-0157 Shiga toxin-producing Escherichia coli— Connecticut, 2000-2005. Morb. Mortal. Wkly. Rep. 56:29-31. Centers for Disease Control and Prevention. 2009. OutbreakNet foodbome outbreak online database. Available at: http://wwwn.cdc. gov/foodbomeoutbreaks/. Accessed 25 May 2013. Centers for Disease Control and Prevention. 2009. Outbreak of Shiga toxin-producing Escherichia coli 0157 infection associated with a day camp petting zoo—Pinellas Country, Florida, May-June 2007. Morb. Mortal. Wkly. Rep. 58:426-428. Clotilde, L. M., C. Bernard IV, G. L. Hartman, D. K. Lau, and J. M. Carter. 2011. Microbead-based immunoassay for simultaneous detection of Shiga toxins and isolation of Escherichia coli 0157 in foods. J. Food Prot. 74:373-379. Favrin, S. J., S. A. Jassim, and M. W. Griffiths. 2003. Application of a novel immunomagnetic separation-bacterophage assay for the detection of Salmonella Enteritidis and Escherichia coli 0157:H7 in food. Int. J. Food Microbiol. 85:63—71. Fernando, S. A., and G. S. Wilson. 1992. Studies of the “ hook” effect in the one-step sandwich immunoassay. J. Immunol. Methods 151:47-66. Gould, L. H., C. Bopp, N. Strockbine, R. Atkinson, V. Baselki, B. Body, R. Carey, C. Crandall, S. Hurd, R. Kaplan, M. Neill, S. Shea, P. Somsei, M. Tobin-D’Angelo, P. M. Griffin, P. Gemer-Smidt, and Centers for Disease Control and Prevention. 2009. Recommendations for diagnosis of Shiga toxin-producing Escherichia coli infections by clinical laboratories. Morb. Mortal. Wkly. Rep. Recomm. Rep. 58:114. Hunt, J. M. 2010. Shiga toxin-producing Escherichia coli (STEC). Clin. Lab. Med. 30:21-45. Hussein, H. S., and L. M. Bollinger. 2008. Influence of selective media on successful detection of Shiga toxin-producing Escherichia coli in food, fecal, and environmental samples. Foodbome Pathog. Dis. 5:227-244.
19.
20.
21.
22.
23.
24. 25.
26.
27.
28.
29.
30.
31.
32.
33.
105
Islam, M., M. P. Doyle, S. C. Phatak, P. Millner, and X. Jiang. 2004. Persistence of enterohemorrhagic Escherichia coli 0157:H7 in soil and on leaf lettuce and parsley grown in fields treated with contaminated manure composts or irrigation water. J. Food Prot. 67:1365-1370. Karsten, S. L., V. M. D. Van Deerlin, C. Sabatti, L. H. Gill, and D. H. Geschwind. 2002. An evaluation of tyramide signal amplification and archived fixed and frozen tissue in microarray gene expression analysis. Nucleic Acids Res. 30:E4. Lahti, E. M., M. Eklund, P. Ruutu, A. Siitonen, L. Rantala, P. Nuorti, and T. Honkanen-Buzalski. 2002. Use of phenotyprng and genotyping to verify transmission of Escherichia coli 0157:H7 from dairy farms. Eur. J. Clin. Microbiol. Infect. Dis. 21:189-195. Li, Y., P. Cheng, J. Gong, L. Fang, J. Deng, W. Liang, and J. Zheng. 2012. Amperometric immunosensor for the detection of Escherichia coli 0157:H7 in food specimens. Anal. Biochem. 421:227-233. Lorusso, V., A. Dambrosio, N. C. Quaglia, A. Parisi, G. Lasalandra, G. Mula, S. Virgilio, G. Lucifora, M. Dario, and G. Normanno. 2011. Development of a multiplex PCR for rapid detection of verocytotoxin-producing Escherichia coli 026 in raw milk and ground beef. J. Food Prot. 74:13-17. Luk, C., G. Giovannoni, D. R. Williams, A. J. Lees, and R. de Silva. 2009. Development of a sensitive ELISA for quantification of threeand four-repeat tau isoforms in tauopathies. J. Neurosci. Methods 180:34-42. March, S. B., and S. Ratnam. 1986. Sorbitol-MacConkey medium for detection of Escherichia coli 0 157 :H7 associated with hemorrhagic colitis. J. Clin. Microbiol. 23:869-872. Nielsen, U. B., and B. H. Geierstanger. 2004. Multiplexed sandwich assay in microarray format. J. Immunol. Methods 290:107-120. O ’Brien, S. J., G. K. Adak, and C. Gilham. 2001. Contact with farming environment as a major risk factor for Shiga toxin (verocytotoxin)-producing Escherichia coli 0157 infection in humans. Emerg. Infect. Dis. 7:1049-1051. Olsen, S. J., G. Miller, M. Kennedy, C. Higgins, J. Walford, G. McKee, K. Fox, W. Bibb, and P. Mead. 2002. A waterborne outbreak of Escherichia coli 0157:H7 infections and hemolytic uremic syndrome: implications for rural water systems. Emerg. Infect. Dis. 8:370-375. Qi, H., Z. Zhong, H. X. Zhou, C. Y. Deng, H. Zhu, J. F. Li, X, L. Wang, and F. R. Li. 2011. A rapid and highly sensitive protocol for the detection of Escherichia coli 0157:H7 based on immunochromatography assay combined with the enrichment technique of immunomagnetic nanoparticles. Int. J. Nanomed. 6:3033-3039. Reinders, R. D., A. Bama, L. J. A. Lipman, and P. G. H. Bijker. 2002. Comparison of the sensitivity of manual and automated immuno magnetic separation methods for detection of Shiga toxin-producing Escherichia coli 0157:H7 in milk. J. Appl. Microbiol. 92:1015—1020. Scallan, E., R. M. Hoekstra, F. J. Angulo, R. V. Tauxe, M. A. Widdowson, S. L. Roy, J. L. Jones, and P. M. Griffin. 2011. Foodbome illness acquired in the United States— major pathogens. Emerg. Infect. Dis. 17:7-15. Shen, Z. Q„ J. F. Wang, Z. G. Qiu, M. Jin, X. W. Wang, Z. L. Chen, J. W. Li, and F. H. Cao. 2011. QCM immunosensor detection of Escherichia coli 0157:H7 based on beacon immunomagnetic nanoparticles and catalytic growth of colloidal gold. Biosens. Bioelectron. 26:3376-3381. Wang, H., Y. Li, A. Wang, and M. Slavik. 2011. Rapid, sensitive, and simultaneous detection of three foodbome pathogens using magnetic nanobead-based immunoseparation and quantum dot—based multiplex immunoassay. J. Food Prot. 74:2039-2047. Zaidi, A. U., H. Enomoto, J. Milbrandt, and K. A. Roth. 2000. Dual florescent in situ hybridization and immunohistochemical detection with tyramide signal amplification. J. Histochem. Cytochem. 48: 1369-1375. Zhu, P„ D. R. Shelton, S. Li, D. L. Adams, J. S. Kams, P. Amstutz, and C. M. Tang. 2011. Detection of E. coli 0157:H7 by immunomagnetic separation coupled with fluorescence immunoas say. Biosens. Bioelectron. 30:337-341.
Copyright of Journal of Food Protection is the property of Allen Press Publishing Services Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
