Patent Grant
8211657
1. Field of the Invention
The present invention relates to a highly sensitive and convenient method using a capillary column as a bioseparator/bioreactor for detection of microbial contamination in food products.
2. Brief Description of the Related Art
Microbial contamination in food products is a major concern to the food industry, regulatory agencies and consumers [S. L. Wilkinson, C and CE News. 75 (1997) 24]. Foodborne illness caused by pathogenic microorganisms poses a serious threat to public health. The Center for Disease Control and Prevention (CDC) estimated that annually 76 million illnesses, 325,000 hospitalized and 5,000 deaths were caused by bacterial contamination of food products in the United States. It is imperative to develop more effective and rapid technology to detect specific pathogens such as Escherichia coli O157:H7, Salmonella Typhimurium and Listeria monocytogenes to ensure that safe food products reach the public [M. W. Griffiths, J. AOAC Int. 80 (1997) 1143] [H. Van der Zee, et al., J. AOAC Int. 80 (1997) 934] [P. C. Vasavada, Food Testing and Analysis. 47 (1997) 18]. Escherichia coli O157:H7 is one of the most dangerous pathogens [R. L. Buchanan, et al., Food Technol. 51 (1997) 69] [P. M. Griffin, et al., Epidemiol. Rev. 13 (1991) 60], and it has been associated with two important human diseases: hemorrhagic colitis and hemolytic uremic syndrome (HUS). It can be transmitted via contaminated foods such as raw ground beef and unpasteurized milk [Center of Disease Control, Morbid. Mortal. Weekly Rep. 42 (1993) 85] and from person to person [Center of Disease Control, Morbid. Mortal. Weekly Rep. 42 (1993) 85], e.g., in nursing houses and day care facilities. Salmonella Typhimurium, one of the pathogens most likely to be found in commonly slaughtered livestock (cattle, sheep, and swing) and poultry (chicken and turkey), causes acute diarrhea, vomiting, abdominal pain, and fever. Occasionally, it may cause blood stream infection and death. Symptoms occur 6-72 hours after eating contaminated foods [USDA Food Safety and Inspection Service (FSIS). 1999, 26 Aug.]. Salmonella contamination in food products also results in the large economical burden for the industry due to products recalls.
Conventional microbiological culture methods used for bacterial detection are cumbersome and time-consuming, requiring 3-4 days for confirmation. Immunosorbent assays also suffer from complexities in sample pretreatment and measuring procedures, because they use isotopes and are susceptible to interference caused by chromogens in food samples. Even though some methods have been developed in an effort to replace traditional techniques, the enzyme-linked immunosorbent assays (ELISA), which can achieve high specificity and sensitivity, still needs 24 hours or more to achieve the necessary detection limits [K. S. Cudjoe, et al., J. Food Microbiol. 27 (1995) 11].
Some methods have been developed for the rapid detection of E. coli O157:H7 in an effort to replace conventional techniques [E. D. Boer, et al., J. Appl. Microbiol. 88(S) (2000) 133S]. Several procedures, such as antibody direct epifluorescent filter technique [L. Restaino, et al., J. Food Prot. 59 (1996) 1072] [L. Restaino, et al., Lett. Appl. Microbiol. 24 (1997) 401], ATP bioluminescence [M. W. Griffiths, J. Dairy. Sci. 76 (1993) 3118], enzyme-linked immunosorbent assays [R. P. Johnson, et al., Appl. Environ. Microbiol. 61 (1995) 386] [P. M. Fratamico, et al., J. Food Prot. 61 (1998) 934] [J-M. Woody, et al., J. Food Prot. 61 (1998) 110] [N. V. Padhye, et al., Appl. Environ. Microbiol. 57 (1991) 2693], impedance [J. Dupont, et al., J. Appl. Bacteriol. 80 (1996) 81] [K. O. Colquhoun, et al., J. Appl. Bacteriol. 79 (1995) 635], multiplex polymerase chain reaction [J. P. Mckillip, et al., J. Appl. Microbiol. 89 (2000) 49] [P. M. Fratafico, et al., J. Clin. Microbiol. 33 (1995) 2188] [C. M. Gooding, et al., J. Dairy Research 64 (1997) 87] [J. L. Mckillip, et al., J. Food Prot. 63 (2000) 855], and flow cytometry [K. H. Seo, et al., J. Food Prot. 61 (1998) 812], have been reported. These procedures could reduce analysis time and give presumptive results within several hours to one day, when compared with cultural plating procedures that require two or more days. The detection limit of these methods varied from 103 to 105 cells/ml. Since E. coli O157:H7 cells in foods are usually present in small numbers, pre-enrichment is necessary to obtain a detectable signal for target bacteria in applications of these methods.
Kim and Park [I. S. Park, et al., Biosens. Bioelectron. 13 (1998) 1091] developed piezoelectric biosensors for detection of Salmonella in the range of 9.9×105 to 1.8×108 CFU/ml. Seo et al. [K. H. Seo, et al., J. Food Prot. 62 (1999) 431] developed an integrated optic interferometer system for detection of Salmonella in the range of 1.0×105 to 1.0×107 CFU/ml by observing the fringe shift generated by refractive index variation. By using membrane separation and electrochemical analysis, a biosensor possessed a linear response for S. Typhimurium from 5.0×103 to 5.0×106 CFU/ml within 2 hours [Z. P. Yang, et al., Electroanalysis 10 (1998) 913]. When immunomagnetic separation and electrochemical detection were applied, a biosensor system could be able to detect S. Typhimurium from 1.0×103 to 1.0×107 [Y. H. Che, et al., J. Rapid Meth. Auto. Microbiol. 7 (1999) 47] [Y. H. Che, et al., J. Food Prot. 63 (2000) 1043]. Brewster et al. [J. D. Brewster, et al., Anal. Chem. 68 (1996) 4153] prepared immunoelectrochemical sensors for the detection of Salmonella. Chang et al. [Y. H. Chang, et al., Biosci. Biotechnol. Biochem. 60 (1996) 1571] constructed a compact fiber optic-based biosensor for detection of Salmonella aureus by measuring laser light signal at 488 nm. These methods make it possible to miniaturize immunosensors in detection of bacteria. However, each of these methods has its particular disadvantages such as high detection limit and poor specificity, high cost of instruments and materials, and/or time-consuming.
Capillary columns offer the advantage of better surface-to-volume interaction and the reduced amount of reagent. More importantly, it would take shorter time for molecules to reach the surface in a capillary, resulting in a faster assay. Capillary columns have proven to be very successful in separation techniques [C. F. Poole, et al., Chromatography Today, Elsevier Science: New York, N.Y. 1991], but there are few reports on their applications in the separation and detection of pathogens. Recently, capillary columns were used as a bioseparator/bioreactor by chemically immobilizing anti-E. coli O157:H7 antibodies onto the inner wall of the column [Y. Liu, et al., Anal. Chem. 73 (2001) 5180]. After a sample and alkaline-phosphatase-labeled antibodies passed through the column and the “sandwich” immunocomplexes were formed, a substrate, p-phenol phosphate was added and then the absorbance of the enzymatic products was measured. A detection limit of 500 CFU/ml was obtained.
The invention described herein comprises a flow-injection analysis system with a bienzyme biosensor used to detect the product of enzymatic reaction instead of optical measurement. Phenol produced from the enzymatic reaction between alkaline phosphatase and its substrate, phenol phosphate was detected by a tyrosinase-horseradish peroxidase biosensor. Compared to UV spectroscopy, electrochemistry provided more sensitive detection [R. Q. Thompon, et al., Anal. Biochem. 192 (1991) 90]. With optimized conditions, a detection range from 8.8×101 to 8.8×106 CFU/ml was obtained for E. coli O157:H7, and the assay time was less than 2.5 hours without any enrichment. Capillary column has been used as a bioseparator/bioreactor for the detection of E. coli O157:H7 and a detection limit of 8.8×101 CFU/ml was obtained [Z. Y. Zhang, et al., Anal. Chim. Acta, 2002 (submitted)] [Y. Liu, et al., Anal. Chem. 73 (2001) 5180]. The capillary immunosensor was also used to detect S. Typhimurium with a total assay time less than 2 hours without any enrichment and a detection limit of 102 cfu/ml.
We demonstrate that this method is more sensitive and convenient compared to other methods developed for the detection of microbial contamination because capillary columns can offer the advantage of high surface-to-volume interaction.
References mentioned in this background section are not admitted to be prior art with respect to the present invention.
The present invention is directed to satisfying the need to detect microbial contamination of food products. The described bioseparator/bioreactor coupled with an optical/electrochemical biosensor was able to specifically detect E. coli O157:H7 from 8.8×101 to 8.8×106 CFU/ml in 2.5 hours without any enrichment. Using this invention, concentrations of S. Typhimurium ranging from 8.6×102 to 8.6×106 CFU/ml in pure culture were detected in 2 hours without any enrichment. The invention may also be used for the detection of S. Seftenberg, which has the same sensitivity as S. Typhimurium. Other pathogens such as L. monocytogenes and S. Heidleberg did not interfere with the detection. The optimum inner diameter of the 25 cm long column for the detection of E. coli O157:H7 and S. Typhimurium is 250 μm. The detection limit for other microbial pathogens may be controlled by changing the length of capillary columns, using higher concentration of the labeled antibodies, altering the flow rate and concentration of the substrate, and increasing the reaction temperature to 37° C.
These and other features, objects and advantages of the present invention will become better understood from a consideration of the following detailed description and accompanying drawings.
With reference to
In this invention a critical step is the immobilization of antibodies onto the capillary column. In order to determine antibody binding, atomic force microscope and surface plasma resonance have been used to monitor the antibody immobilization on a gold electrode surface [J. Gau, et al., Biosens. Bioelectron. 16 (2001) 745]. I-labeled antibodies have also been used to determine the amount of antibodies immobilized on the silica surface [S. K. Bhatia, et al., Anal. Biochem. 178 (1989) 408].
However, direct analysis of the inside surface of a 250 μm inner diameter fused silica capillary is difficult. Hence, the absorbance at 280 nm of the antibody solution before and after the immobilization was measured with the procedure described by Sportsman and Wilson [J. R. Sportsman, et al., Anal. Chem. 52 (1980) 2013] to demonstrate that the antibody was coupled onto the inner surface of capillaries.
Optimization Parameters for Immobilizing the Primary Antibodies
Concentrations of 0.6 ml of 0.1 mg/ml antibodies were pumped into columns with different lengths at a flow rate of 0.5 ml/h. The first half of the 0.3 ml and the second half of the 0.3 ml solution after the immobilization were collected at the outlet separately. The absorbance was measured at 280 nm spectrophotometrically (HP 8453 UV-Vis-spectrophotometer, HP).
Table 1 shows the absorbance data collected before and after the antibody immobilization. When 0.3 ml of antibodies was pumped into the column, the reaction ratio of antibodies increased as the column length increased. This demonstrated that the longer the column, the more volume of antibodies needed. The result also showed that for the 3 m capillary column, less than 0.6 ml of antibodies are needed for the immobilization. Therefore, 0.4 ml of 0.1 mg/ml antibodies was selected for the antibody immobilization. Considering the instrumental noise, the difference between the initial absorbance of antibodies and that of the first half of the 0.3 ml antibodies after the immobilization was negligible. Therefore, 0.6 ml of antibodies are sufficient for the antibody immobilization when the capillary length is less than 3 m.
The procedure for the immobilization of anti-Salmonella antibodies is the same as that of anti-E. coli O157:H7 antibodies.
For S. Typhimurium an optimal temperature for the reaction is 37° C. Under the optimum temperature, a detection limit of 8.6×102 CFU/ml for S. Typhimurium was obtained, which is one order of magnitude lower than that at room temperature. At the same time, the linear range from 8.6×103 CFU/ml to 8.6×106 CFU/ml for S. Typhimurium was obtained, compared to 8.6×104 CFU/ml to 8.6×106 CFU/ml under room temperature.
The Length, Diameter, and Flow Rate of the Capillary Column
The length of the capillary column is an important parameter of the capillary bioseparator/bioreactor. Both a control devoid of E. coli O157:H7 and 4.3×105 CFU/ml E. coli O157:H7 pure culture, were analyzed at the same time to obtain sample and background signals for different lengths of the columns.
We demonstrate that changing the length of the column can alter the detection limit desired for particular concentration of bacteria. However, when the column is too long (>1 m), a larger volume of the sample is needed as well as a longer time period for antibody-antigen reaction to occur. Therefore, 25 and 40 cm long columns were selected for E. coli O157:H7 and S. Typhimurium, respectively.
Improvement of the detection limit may also be obtained through the optimization of capillary column diameter. The 25 cm long columns with 99, 152, 250, 325, and 525 μm inner diameter were tested to optimize the capillary diameter based on the ratio of the signal-to-background. Both a control devoid of E. coli O157:H7 and 8.8×105 CFU/ml E. coli O157:H7 pure culture were used to obtain signals from the column.
For each antigen-antibody pair, there is an optimum flow rate, which is related to the dissociation constant of antibodies [U. Narang, et al., Anal. Chem. 69 (1997) 2779]. Increasing the flow rate above this level may cause decreased antibody-antigen interaction time. Decreasing the flow rate too much may result in poor discrimination of the sample signal from the high background caused by nonspecific bounding in the capillary column. Thus, for columns with different diameters, linear flow velocity (=flow rate/cross-sectional area of the capillary) is a more accurate measurement for comparing the flow streams through the column.
For enzymatic reactions, the flow rate of substrates had an effect on the peak current and on the detection limit. Therefore, for detection of E. coli O157:H7, a flow rate of 0.5 ml/h was used. For S. Typhimurium samples,
For E. coli O157:H7, the optimum contents of the substrate solutions were determined to be 1.0×10−3 M phenyl phosphate disodium and 1.0×10−2 M MgCl2 in pH 10, 0.1 M Tris buffer [C. M. Gooding, et al., J. Dairy Research 64 (1997) 87]. For S. Typhimurium, pH 9.0 Tris buffer is the best media for the enzymatic reaction and the maximum absorbance of 4-nitrophenol from the enzymatic reaction was observed at 1.0×10−2 M MgCl2.
In order to evaluate batch procedures, six 30 cm long capillary columns (250 μm i.d.) were coated with antibodies simultaneously. Three columns were pumped with 0.5 ml of 4.6×105 CFU/ml E. coli O157:H7 sample and three columns were pumped with 0.5 ml of 4.6×106 CFU/ml E. coli O157:H7 sample at a flow rate of 0.5 ml/h. The relative standard deviation (RSD) obtained by the six columns was 7.27% for the 4.6×106 CFU/ml E. coli O157:H7 sample, and 6.23% for the 4.6×105 CFU/ml E. coli O157:H7 sample, indicating reproducibility using a batch method.
Concentration of Antibodies
The concentration of labeled antibodies also has effect on the ratio of signal-to-background and that the ratio of signal-to-background increased as the concentration of labeled antibodies increased [J. Dupont, et al., J. Appl. Bacteriol. 80 (1996) 81]. In order to test the effect of the concentration of AP-labeled anti-E. coli 0157:H7 antibodies on the peak current measurement, 50 μl of AP-labeled anti-E. coli O157:H7 antibodies at different concentrations from 0.25 mg/l to 2.0 mg/l were used. A control devoid of E. coli O157:H7 and 8.8×105 CFU/ml E. coli O157:H7 pure culture were conducted at the same time.
Electrochemical Detection of Enzyme Labels
The sensitivity of an enzyme immunoassay is usually controlled by the sensitivity of the method used to detect the enzyme label. The enzyme label can be readily detected through the conversion of a substrate into a product, either using an optical method such as UV absorbance, fluorescence and luminescence, or by an electrochemical method. Electrochemical methods, especially electrochemical biosensors, have a very high sensitivity for the detection of alkaline phosphatase. In addition, electrochemical biosensors may be used to detect the binding of nanocrystals. Nanocrystals may be used to label antigen marker agents and the binding to the antigen marker agents can be determined by determining the level of binding of the nanocrystals using standard techniques known to those of skill in the art. Thus, nanocrystals may be used as detectable markers.
Recently, a bienzyme electrochemical biosensor based on tyrosinase and horseradish was developed and applied to the detection of zeptomolar concentrations of alkaline phosphatase in a flow injection system [C. Ruan, et al., Talanta 54 (2000) 1095]. The enhanced sensitivity of the bienzyme biosensor for detection of alkaline phosphatase was observed in comparison with its corresponding mono-enzyme biosensor.
The bienzyme biosensor was used to detect the enzyme label bound onto the capillary column through E. coli O157:H7. The enzyme labels catalyzed the hydrolysis of substrate, and produced phenol, which was detected by the bienzyme biosensor. The optimum contents of the substrate solutions were determined to be 1.0×10−3 M. phenyl phosphate disodium and 1.0×10−2 M MgCl2 in pH 9.0, 0.1 M Tris buffer at an optimum flow rate of 2 ml/h. The pH of the substrate had great influence on the enzymatic reaction. The peak current was much higher for the substrate in basic solutions than that in neutral or acidic solutions.
A linear relationship between the peak current of phenol and logarithmic value of Salmonella is shown in
The nonlinearity at high bacterial concentrations may be explained by concentrations of the alkaline phosphatase labeled antibodies reacting with surplus bacteria, thus forming a competing reaction between “sandwich” antibodies-bacteria-alkaline phosphatase labeled antibodies and the complexes of alkaline phosphatase labeled antibodies-bacteria. Thus, complexes of alkaline phosphatase labeled antibodies-bacteria were removed by separation, and therefore the response decreased at higher concentrations of S. Typhimurium. This is in turn prevented the formation of the “sandwich” and it is often termed the “hook effect” (prozone phenomenon), which occurs in most types of enzyme immunoassays [P. M. Fratafico, et al., J. Clin. Microbiol. 33 (1995) 2188].
In addition, as the bacteria concentration increased, the number of enzymes present during the incubation increased. This may result in a substrate-limited enzyme reaction, thus reducing the peak current value. Adding more substrate might broaden the detection range, but may also increase the background noise. The cell numbers of Salmonella associated with processed food products usually is less than 1,000 CFU/ml, therefore, Salmonella titers ranging from 102 to 106 is a reasonable range in most investigations.
A significant difference (p<0.05) in the response between background control and samples containing 8.8×101 CFU/ml of E. coli O157:H7 was observed. No significant difference (p>0.05) was found for E. coli O157:H7 concentrations between 8.8×101 and 8.8×103 CFU/ml, indicating no linear relationship between the number of E. coli O157:H7 and the current response when the concentration of E. coli O157:H7 is less than 8.8×103 CFU/ml.
Compared to UV spectroscopy, electrochemistry provides a more sensitive detection method that is crucial to trace analysis [J. L. Mckillip, et al., J. Food Prot. 63 (2000) 855]. Using electrochemical detection we observed greater limits than those obtained using optical measurement [Y. Liu, et al., Anal. Chem. 73 (2001) 5180]. For E. coli O157:H7 the detection limit obtained may be as low as 8.8×101 CFU/ml, and the working range from 8.8×101 CFU/ml to 8.8×106 CFU/ml. For S. Typhimurium the detection limit obtained by electrochemical measurement was 8.6×102 CFU/ml, which is two orders of magnitude better than by absorption spectrophotometry. Even though absorption spectrophotometry is more convenient to use and much faster, electrochemical measurement remains the choice when attempting to detect lower concentrations of S. Typhimurium.
Antibody Specificity
The detection specificity of this method was investigated by comparing the detection results of samples containing individual isolates of E. coli O157:H7, L. monocytogenes, S. Seftenberg, S. Heidleberg, S. Typhimurium, samples containing the combinations of E. coli O157:H7, L. monocytogenes, and S. Typhimurium, and samples containing combinations of S. Typhimurium, S. Seftenberg, and S. Heidleberg.
Unlike the specificity for detection of E. coli O157:H7 that is detected with a monoclonal antibody, the anti-Salmonella antibody is a polyclonal antibody. As shown in Table 2, no significant difference (p>0.05) was observed between the background and signal of the samples containing only 5.0×106 CFU/ml of L. monocytogenes, 5.0×106 CFU/ml of E. coli O157:H7, or 5.0×106 CFU/ml of S. Heidleberg. This response may be due to the nonspecific binding by L. monocytogenes, E. coli O157:H7 or S. Heidleberg to the anti-Salmonella antibody. Similarly, the signals from the samples containing 5.0×106 S. Typhimurium and 5.0×106 S. Seftenberg were not significantly different (p>0.05).
S. Typhimurium
E. coli O157:H7
L. monocytogenes
S. Heidleberg
S. Typhimurium
S. Seftenberg
[a]Mix 1 is the sample containing the same concentration (5.0 × 106 CFU/ml) of S. Typhimurium, L. monocytogenes, and E. coli O157:H7.
[b]Mix 2 is the sample containing the same concentration (5.0 × 106 CFU/ml) of S. Typhimurium, S. Seftenberg, and S. Heidleberg.
Using the anti-E. coli O157:H7 antibody, selectivity was greatly increased among different bacteria. As shown in
Detection of S. Typhimurium in Food Products
As shown in
Anti-E. coli O157:H7 antibodies (1 mg) and alkaline phosphatase-labeled affinity purified antibodies to E. coli O157:H7 (0.1 mg), purchased from Kirkegaard & Perry Laboratories (Gaithersburg, Md.), were rehydrated with 1 ml of 50% glycerin water solution. Dilution of 1:10 of the antibodies and 1:100 of the alkaline phosphatase-labeled antibodies were prepared before use.
Primary and secondary antibodies, anti-Salmonella antibodies and alkaline phosphatase-labeled affinity purified anti-Salmonella antibodies, were obtained from Kirkegaard & Perry Laboratories (Gaithersburg, Md.). One mg of anti-Salmonella antibodies and 0.1 mg of alkaline phosphatase-labeled affinity purified anti-Salmonella antibodies were rehydrated with 1 ml of 50% glycerin water solution to obtain 1:10 and 1:100 dilutions, respectively.
Tyrosinase (EC 1.14. 18. 1, from mushroom, 3000 units/mg), bovine serum albumin (BSA), horseradish peroxidase (EC 1. 11.1. 7, type II, 240 units/mg), tris (hydroxymethyl)aminomethane (Tris, 99.9%), pH 7.4 phosphate buffer saline (PBS), and pH 7.4, 1% BSA were purchased from Sigma (St. Louis, Mo.). Glutaric dialdehyde (50 wt. % solution in water), 3-aminopropyltrimethoxysilane (97%), hydrogen peroxide (30 wt. %), phenol (99%), phenyl phosphate disodium (98%) and acetonitrile were obtained from Aldrich (Milwaukee, Wis.). Na2HPO4, NaH2PO4 H2O and MgCl2 were purchased from Fisher (Pittsburgh, Pa.). Fused-silica capillary columns (100, 150, and 250 μm i.d.) were from Polymicro Technologies (Phoenix, Ariz.). Other chemicals were of analytical grade and were used without further purification.
A stock solution of 0.1 M phenol was prepared by dissolving an appropriate amount of phenol in acetonitrile and stored at 4° C. Standard phenol was diluted in PBS (pH 6.9) on the day of use from the stock solution. A stock solution of 0.1 M hydrogen peroxide was prepared in water and stored at 4° C.
Escherichia coli O157:H7 (ATCC 43888) as a target pathogen and Salmonella Typhimurium (ATCC 14028) and Listeria monocytogenes (FDA 10143) as competing bacteria were obtained from American Type Culture Collection (Rockville, Md.). The pure culture of E. coli O157:H7, S. Typhimurium, and L. monocytogenes were grown in brain heart infusion (BHI) broth (Remel, Lenexa, Kans.) at 37° C. for 20 hours before use. The culture was serially diluted to 10−8 with physiological saline solution (PSS) and viable cell number was determined by plate count. Microbial tests for E. coli O157:H7 was performed by surface plating 0.1 ml of dilutions on MacConkey sorbitol agar (remel, Lenexa, Kans.). After incubation at 37° C. for 24 hours, E. coli O157:H7 colonies on the plate were counted to determine the number of colony forming units per ml (CFU/ml). Cell numbers of S. Typhimurium and L. monocytogenes in pure culture were determined by the same method with the exception of the use of xylose lysine tergitol agar (XLT4) (remel, Lenexa, Kans.) and Oxoid agar (Oxoid LTD, Basingstoke Hampshire, U.K.), respectively.
Alternatively, pure culture of S. Typhimurium (ATCC 14028) as a target pathogen, Escherichia coli O157:H7 (ATCC 43888), Salmonella Seftenberg (ATCC 46845) and Salmonella Heidleberg (8326) as competing bacteria were obtained from American Type Culture Collection (Rockville, Md.). Listeria monocytogenes (FDA 10143), as a competing bacteria, was obtained from the Food and Drug Administration. All cultures were grown in brain heart infusion (BHI) broth (remel, Lenexa, Kans.) at 37° C. for 20 hours before use. The culture was serially diluted to 10−8 with physiological saline solution (PSS) and viable cell number was determined by plate count. S. Typhimurium, S. Seftenberg and S. Heidleberg culture counting was performed by surface plating 0.1 ml of dilutions on xylose lysine tergitol agar (XLT4) (remel, Lenexa, Kans.). After incubation at 37° C. for 24 hours, the colonies on the plate were counted to determine the number of colony forming units per ml (CFU/ml). The cell numbers of E. coli O157:H7 and L. monocytogenes were determined by the same method with the exception of the use of MacConkey sorbitol agar (remel, Lenexa, Kans.) and Oxoid agar (Oxoid LTD, Basingstoke Hampshire U.K.), respectively.
The desired antibody was immobilized onto the inner wall of columns basically according the method described by Liu and Li [Y. Liu, Y. Li, Anal. Chem. 73 (2001) 5180]. Using a series of syringes on a Harvard PHD 2000 advanced syringe pump (Harvard Apparatus, Holliston, Mass.), a series of fused-silica capillary columns (3 m) were treated with 1 M NaOH and 1M HCl, and dried out overnight. Dried capillary columns were treated with 3-aminopropyltrimethoxysilane (1% methanol solution), incubated for 70° C. for 4 hours to allow the formation of the aminopropyl derivative of glass, rinsed with methanol, dried overnight, reacted with glutaraldehyde, and washed with PBS (pH=7.4) solution. Our study showed that the concentration of antibodies had great effect on the detection limit. Therefore, 0.1 mg/ml concentration of the antibody was used in order to decrease the detection limit. For E. coli detection, 0.4 ml of anti-E. coli O157:H7 antibodies (0.1 mg/ml) was pumped into each of the columns for coupling, and followed by washing with PBS (pH=7.4) solution. These anti-E. coli antibody modified columns were cut into 25 cm long pieces and put into a refrigerator at 4° C. for later use. For Salmonella detection, anti-Salmonella antibodies (0.1 mg/ml) were pumped into prepared columns for coupling, and followed by washing columns with PBS (pH=7.4) solution. These anti-Salmonella antibody modified columns were cut to 40 cm long pieces and put into a refrigerator at 4° C. for later use.
The 0.5 ml of E. coli O157:H7 sample and 50 μl of alkaline phosphatase-labeled affinity purified antibodies (1 mg/l) were pumped into the column (length=25 cm) at a flow rate of 0.5 ml/h to form the sandwich immunocomplexes on the inner wall of the column. The column was then rinsed with 2×1 ml of PBS (0.01 M, pH 7.4, 1% BSA) and 1 ml of Tris buffer (0.1 M, pH 8.0). The substrate, composed of 1.0×10−3 M phenyl phosphate disodium and 1.0×10−2 M MgCl2 in pH 9.0, 0.1 M Tris buffer, was pumped through the column at a flow rate of 2 ml/h. The product of the enzymatic reaction from the bioreactor was collected at the outlet. After the enzymatic reaction, 10 μl of 0.1 M H2O2 was added to the solution to amplify the electrochemical signal of the phenol generated, which was then determined by the bienzyme biosensor in a flow injection system.
Procedures for fabrication of tyrosinase-horseradish peroxidase electrodes were described in detail in the report by Ruan and Li [C. Ruan, Y. Li, Talanta 54 (2000) 1095]. Dual glassy carbon electrodes (BAS, West Lafayette, Ind.) were polished and sonicated, washed with deionized water and dried in the air. Two mg tyrosinase, 2 mg horseradish peroxidase and 100 μg bovine serum albumin were separately dissolved in 100 μl, 0.05 M phosphate buffer solution (PBS, pH 7.0, without KCl). Forty μl of the above dissolved tyrosinase solution, 10 μl of the above dissolved horseradish peroxidase solution, 10 μl of BSA and 10 μl of 5% glutaric dialdehyde and 20 μl of PBS (pH 7.0, without KCl) were mixed thoroughly. A 2.5 μl of the resulting mixture was spread on the surface of a glassy carbon disk electrode (3 mm diameter). All modified enzyme electrodes were dried overnight at 4° C. and washed with deionized water before use.
For the electrochemical measurement using a bienzyme electrode, a flow injection analysis (FIA) system connected with an electrochemical cell was used for the detection of phenol production. The bienzyme electrode was inserted into a cross-flow thin-layer amperometric cell comprising an Ag/AgCl (3.0 M NaCl) reference electrode and a stainless steel auxiliary electrode (BAS, West Lafayette, Ind.). The amperometric cell was connected to a single channel flow injection system consisting of a six ports injection valve (Rheodyne, Berkley, Calif.) with a 100 μl injection loop, to an electrochemical detector (model 800, CH Instruments, Dallas, Tex.). An HP1100 series isocratic pump (Hewlett-Packard, Waldbronn, Germany) was used to drive the carrier solution at a rate of 1.0 ml/min. All measurements were performed at an applied potential of −0.2 V (vs. Ag/AgCl). A 0.05 M, pH 7.0 PBS was used throughout the tests as a carrier solution in the FIA system.
Before all electrochemical measurements, each electrode was potentiostated at the working potential, allowing the background current to decay to a steady state value. The electrode was activated for 5 min using a solution of 10−5 M standard phenol (containing 10−3 M H2O2).
The 0.5 ml of S. Typhimurium sample and 100 μl of alkaline phosphatase-labeled affinity purified antibodies (2 mg/l) were pumped into the columns (length=40 cm) to form the sandwich immunocomplex on the inner wall of the columns under the temperature of 37° C. The columns were then rinsed with 2×1 ml of PBS (0.01 M, pH 7.4, 1% BSA) and 1 ml of Tris buffer (0.1 M, pH 8.0). The electrochemical substrate, composed of 1.0×10−3 M phenyl phosphate disodium and 1.0×10−2 M MgCl2 in 1.0 M, pH 10.0 Tris buffer, or optical substrate, composed of 2.0×10−4 M 4-nitrophenyl phosphate and 1.0×10−2 M MgCl2 in 1.0 M pH 9.0 Tris buffer, was pumped through the columns. The product of the enzymatic reaction from the capillary bioreactor was collected at the outlet. After the enzymatic reaction, for electrochemical measurement, a flow injection analysis (FIA) system connected with an electrochemical cell was used for the detection of phenol production. The bienzyme electrode was inserted into a cross-flow thin-layer amperometric cell comprising an Ag/AgCl (3.0 M NaCl) reference electrode and a stainless steel auxiliary electrode (BAS, West Lafayette, Ind.). The amperometric cell was connected to a single channel flow injection system consisting of a six ports injection valve (Rheodyne, Berkley, Calif.) with a 100 μl injection loop, to an electrochemical detector (Model 800, CH Instruments, Dallas, Tex.). An HP1100 series isocratic pump (Hewlett-Packard, Waldbronn, Germany) was used to drive the carrier solution at a rate of 0.5 ml/min. All measurements were performed at an applied potential of −0.2 V (vs. Ag/AgCl). A 0.05 M, pH 7.0 PBS was used throughout the tests as a carrier solution in the FIA system. Before all electrochemical measurements, each electrode was potentiostated at the working potential, allowing the background current to decay to a steady state value. The electrode was activated for 5 min using a solution of 10−5 M standard phenol (containing 10−3 M H2O2). 10 μl of 0.1 M H2O2 was added to the solution to amplify the electrochemical signal of the phenol generated, which was then determined by the bienzyme biosensor in a flow injection system.
For optical measurement, the product of 4-nitrophenol was detected by measuring the absorbance at 400 nm using a spectrophotometer (HP 8453 UV-Visible Spectrophotometer, Agilent Technologies, Palo Alto, Calif.).
For E. coli measurements, the difference between the background and sample responses was analyzed using t test. For the detection of S. Typhimurium the means and standard deviations of electrochemical and optical signals were calculated, and the difference between the background response and the response of samples was analyzed using t test. Statistical analysis for both setups was performed using the JMP program (SAS Institute Inc., Cary, N.C.).
Milk was purchased from a local grocery store. 0.2 ml of diluted S. Typhimurium pure culture (ranging from 8.6×102 to 8.8×107 CFU/ml) was added to 1.8 ml milk to get the final inoculation concentration ranging from 8.8×101 to 8.8×106 CFU/ml. 0.5 ml of milk without bacteria was used as the background control.
The present application is a Continuation of U.S. patent application Ser. No. 10/425,510, filed Apr. 28, 2003 now abandoned, which claims the benefit of U.S. Provisional Application No. 60/376,608, filed Apr. 29, 2002, which is incorporated herein by reference.
This invention was made with Government support under the terms of Grant Nos. 2001-35201-10056 and 2001-34211-10288 awarded by the USDA/NRI and USDA/FSC. The Government has certain rights in this invention.
| Number | Name | Date | Kind |
|---|---|---|---|
| 4072572 | Lanham et al. | Feb 1978 | A |
| 4340671 | Gibson | Jul 1982 | A |
| 4525452 | Jones et al. | Jun 1985 | A |
| 4590157 | Chandler et al. | May 1986 | A |
| 4615979 | Perloe | Oct 1986 | A |
| 4844869 | Glass | Jul 1989 | A |
| 4918025 | Grenner | Apr 1990 | A |
| 5183740 | Ligler et al. | Feb 1993 | A |
| 5250439 | Musho et al. | Oct 1993 | A |
| 5354661 | Doyle et al. | Oct 1994 | A |
| 5447617 | Shieh | Sep 1995 | A |
| 5447836 | Wolber et al. | Sep 1995 | A |
| 5462646 | Shieh | Oct 1995 | A |
| 5496700 | Ligler et al. | Mar 1996 | A |
| 5516409 | Kambara | May 1996 | A |
| 5601988 | Gordon | Feb 1997 | A |
| 5624850 | Kumar et al. | Apr 1997 | A |
| 5637467 | Meltzer | Jun 1997 | A |
| 5897993 | Sato | Apr 1999 | A |
| 6001564 | Bergeron et al. | Dec 1999 | A |
| 6020209 | Narang et al. | Feb 2000 | A |
| 6177266 | Krishnamurthy et al. | Jan 2001 | B1 |
| 6245296 | Ligler et al. | Jun 2001 | B1 |
| 6274323 | Bruchez et al. | Aug 2001 | B1 |
| 6323042 | Narang et al. | Nov 2001 | B1 |
| 6375817 | Taylor et al. | Apr 2002 | B1 |
| 20020029968 | Tan et al. | Mar 2002 | A1 |
| Number | Date | Country |
|---|---|---|
| 0010584 | Mar 2000 | WO |
| 0016803 | Mar 2000 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 20110165555 A1 | Jul 2011 | US |
| Number | Date | Country | |
|---|---|---|---|
| 60376608 | Apr 2002 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 10425510 | Apr 2003 | US |
| Child | 12966340 | US |
