Rapid Immunochromatographic Assay for Escherichia coli O157:H7 in Bovine Milk Using IgY Labeled by Fe₃O₄/Au Composite Nanoparticles

Abstract

A rapid pretreatment-free immunochromatographic assay (ICA) was developed for the screen of Escherichia coli O157:H7 (E. coli O157:H7) in bovine milk. Anti-E. coli O157:H7 immunoglobins of yolk (IgY) was conjugated with a core/shell-structured super-paramagnetic Fe₃O₄/Au composite nanoparticle (GoldMag) and showed a high coupling ratio of 8 µg IgY mL⁻¹ nanoparticles at pH 8.0. The conditions of ICA strips using this IgY conjugate were optimized and could be used to measure spiked E. coli O157:H7 samples with a detection limit of 10³ CFU mL⁻¹ obtained by visual detection. Additionally, the magnetic signal intensities of the assay can quantitatively measure E. coli O157:H7 concentration in the range of 10² – 10⁵ CFU mL⁻¹ both in phosphate buffer solution (pH 7.4) and pretreatment-free milk samples with a detection limit of 10² CFU mL⁻¹.

Introduction

Escherichia coli O157:H7 (E. coli O157:H7) is a widespread food-borne pathogen and capable of producing toxin and causing severe gastrointestinal disease in humans. It may lead to hemolytic uremic syndrome (HUS), and even to death, especially in children < 5 ages occasionally. E. coli O157:H7 can be transmitted through foods and water, which is estimated to cause 73,000 illnesses occurring each year (Mead et al. 1999). Dairy cattle are considered reservoirs of STEC (Enterotoxigenic E. coli) including E. coli O157:H7 and can impose a significant health risk to humans (Hussein and Sakuma 2005). The transmission to humans is associated mostly with consumption of contaminated food, unpasteurized milk and ready-to-eat products (Liptakova et al. 2004). The infective dose of E. coli O157:H7 is reported as low as 50 – 100 cells (Strachan et al. 2001). In addition, its extended survival in bovine feces emphasizes that environmental contamination may pose a serious risk to raw milk and ground beef. Milk-borne outbreaks of E. coli O157:H7 have been reported worldwide (Keene et al. 1997; Gaulin et al. 2012). In a review conducted by Baylis that numerous cases of food poisoning by E. coli O157:H7 due to unpasteurized milk products consumption were reported (Baylis, 2009). Therefore, prompt, careful monitoring and control of E. coli O157:H7 is extremely important to food safety evaluation.

Various methods have been developed to determine E. coli O157:H7 contamination in the food matrix. Conventional techniques used for detection of the E. coli O157:H7 in cases of food poisoning include bacteriological methods (Gehring et al. 2012), polymerase chain reaction (PCR) (Wang et al. 2014), surface plasmon resonance (Wang et al. 2013) (SPR) , enzyme-linked immunosorbent assays (ELISA) (Cavaiuolo et al. 2013), flow cytometry (Yankey et al. 2012) and so on. Culture-based methods are labor intensive, time consuming, and not sensitive. PCR yields fast results but it is prone to be affected by inhibitors in a food matrix. The other methods may need complicated preconcentration, time-consuming steps and high-cost instruments. Those rent themselves poorly to out-of-laboratory scenarios and on-spot detection.

Lateral-flow immunochromatographic assay (ICA) has been widely applied in the field of food safety supervision (Berlina et al. 2013; Byzova et al. 2014). This technique is utilizes antigen—antibody assay and provides rapid detection of analytes. The results judging by naked eye ensure the convenience of bioassays on-site. Colloidal gold was commonly used as a marker in ICA, mainly because of the nano-diameter and red color. A core/shell-structured super-paramagnetic Fe₃O₄/Au composite nanoparticle (GoldMag) has been introduced into immunochemistry as a marker for a visible color reaction to develop an ICA (Yang et al. 2013), which can take advantage of magnetic intensity except the red color over the colloidal gold. The sensitive and specific of ICA strip was determined by not only labeling marker used but also the antibodies. Immunoglobins of Yolk (IgY) is a kind of chicken antibodies extracted from egg yolks, which have been used as an alternative to mammalian IgG for serological tests (Cova 2005).

We describe the development of a sandwich ICA for E. coli O157:H7 screening in bovine milk, using IgG as the capture antibodies and modified IgY as the detecting antibody. This method has some distinct advantages over traditional immunoassays-ease of procedure, rapid operation, low-cost and immediate results.

Material and Methods

Bacterial strains    The bacteria strains used in this study were listed in Table 1. E. coli O157:H7 ATCC 43895 was taken as the reference strain. It cannot ferment sorbitol. All bacteria were cultured, harvested and inactive in formaldehyde (0.3 – 0.4%) to obtain the suspensions (10⁹ CFU mL⁻¹) for use.

Table 1. Bacteria strains used in this study.

Bacteria strains		Source
Escherichia coli O157:H7	ATCC43895	CN-CDCa
	ATCC43889	CN-CDC
	ATCC43888	CN-CDC
	CMCC44828	CMCCb
	NCTC12900	NCTCc
	99X131	JS-CDCd
	99X132	JS-CDC
STEC O26:H11	ATCC BBA-2196	CN-CDC
STEC O111	ATCC BBA-2440	CN-CDC
STEC O103:H2	ATCC BBA-2210	CN-CDC
STEC O145	ATCC BBA-2206	CN-CDC
non-pathogen E. coli strains	BL21 DE3	Lab-preserved
	BL21RIPL	Lab-preserved
	JM109	Lab-preserved
	CMCC 44102	Lab-preserved
	ATCC 25922	Lab-preserved
	CMCC44103	Lab-preserved
vulgaris E. coli	CMCC 44102	Lab-preserved
Enteroinvasive E. coli	CMCC 44350	Lab-preserved
Staphylococcus aureus strains	CMCC26001	Lab-preserved
	CMCC26003	Lab-preserved
S. typhimurium	ATCC 13311	Lab-preserved
	S. urbana	Lab-preserved

^a  CN-CDC: Chinese Center for Disease Control and Prevention

^b  CMCC: China Medical Culture Collection, China

^c  NCTC: National Collection of Type Cultures, China

^d  JS-CDC: Jiang Su Province Center for Disease Control and Prevention, China

Preparation of GoldMag and IgY conjugate and immunochromatographic test strips    Polyclonal anti-E. coli O157:H7 IgY was produced as previously reported (Sunwoo et al. 2006). Purified IgY was used as a detection antibodies and conjugated to GoldMag (GM, Shaanxi Lifegen Co., Ltd, China). To examine a proper amount of antibodies and pH values for stabilizing GM nanoparticles, IgY-GM titration were performed following a modified colloidal gold conjugation protocol (Rudolf et al. 2009). The absorption difference (ΔAbs = Abs₅₅₀−Abs₆₀₀) used to estimate the optimum conjugation condition, where the difference reaches a constant level.

The ICA strip was assembled according to Byzova et al (Byzova et al. 2014). 25 µL conjugate solution was spayed on the conjugate pad, which was prepared by diluting with 0.02 M PBS (pH 8.0, 0.2% Tween-20, 1% PEG-2000, 1%BSA, 8% (w/v) sucrose) at ratios of 1:1. Pure cellulose fiber of 100% was used as the absorbent pad. The test line was coated by serially diluted IgG anti-E. coli O157:H7 (T line, Beijing Bioss Co., China), 0.3 mg mL⁻¹, 0.4 mg mL⁻¹, 0.5 mg mL⁻¹ and 0.6 mg mL⁻¹ with the volume of 10 µL. The control line was coated with l0 µL of rabbit anti-chicken IgG (C line, Beijing Bioss Co., China, 0.5 mg mL⁻¹). Positive solutions containing E. coli O157:H7 (1×105 CFU mL⁻¹) were tested to determine the optimal concentrations of coating antibody.

200 µL E. coli O157:H7 simulated sample was tested by the strips and the different intensity of the red color can be observed by naked eye. The magnetic signal (MS) of the cartridge mentioned above was detected by Magnetic Assay Reader (MAR, Assay Development System, Magna-biosciences, LLC, USA). Data were reported as relative magnetic units (RMUs) which are in direct proportion to the amount of magnetic complexes formed in the T line (IgG-antigen-IgY-GM complexes) or C line (2^ndIgG-IgY-GM complexes). All experiments were performed in Triplicate, and the average of the Triplicates was used in the analysis.

Sensitivity and specificity test    E. coli O157:H7 was diluted from 1×10¹ to 1×10⁶ CFU mL⁻¹ to evaluate the sensitivity of the strips. And Cow's milk with a fat content of 1.5% and 3.5% was purchased in retail stores. Formaldehyde inactive E. coli O157:H7 was added to the milk to generate the simulated milk samples in the concentration range of 1×10¹ to 1×10⁶ CFU mL⁻¹. 200 µL of each particular dilution was used to assess the sensitivity and PBS (0.02 M, pH 8.0) was used as the blank control.

The formaldehyde (0.3 – 0.4%) inactive bacteria solutions, including other kinds of E. coli O157:H7 strains, Non-pathogenic laboratory E. coli, Salmonella and Staphylococcus aureus with the same concentration of 10⁵ CFU mL⁻¹ in milk samples were prepared to evaluate the specificity of the ICA strip.

All ICA strips were stored at 4°C for 12 weeks to evaluate the stability during storage. 1×10⁵ CFU mL⁻¹ of E. coli O157:H7 and E.coli in milk was used as samples, with PBS (0.02 M, pH 8.0) as the blank control.

Results and Discussion

Optimal conditions for conjugation    Different concentrations from 0 to 20 µg antibody mL⁻¹ GM solutions were prepared to estimate the optimal conditions for IgY-GM conjugate coupling. The optimal coupling ratio was determined at pH 8.0. As shown in Fig. 1A, the optimal IgY concentration for labeling was 8 µg IgY mL⁻¹ GM particles, where the saturation curve reached a plateau suggesting the colloidal particles stabilized. The amount of IgY used for GM conjugation was usually in excess by 30%, that is, 10.5 µg IgY mL⁻¹ GM particles.

Fig. 1.

Titration curve of IgY-GM conjugate, Absorbance difference between 550 nm and 600 nm. (A) Titration curve for the optimal concentration of IgY coupled with GM. (B) Titration curve of IgY coupled with GM at different pH.

The pH-value has a high impact on the charge state of IgY and GM particles. Considering that the pI of IgY is not a clearly defined point but rather an area of about two pH units, the GM solution was adjusted to 4 different pH values between 6.8 and 8.5 with 0.1 M K₂CO₃ to find the ideal conditions for antibody coupling. Fig. 1B showed that the saturation curve reached a constant level at pH 6.8, 7.5, 8.0, yet this saturation cannot be reached at pH 8.5. It might be that the pI of IgY are lower than that of IgG and then the repellants between antibody and GM particles are too high to cover particles with enough antibodies to reach a stable state at pH 8.5. However, more alkaline pH values are benefit for antibody binding to GM particles via their Fc regions (Dávalos-Pantoja et al., 2001), the optimal pH was determined as pH 8.0.

Optimization of the ICA conditions    In terms of the coating of IgG anti-E. coli O157:H7 at T line, the color intensity deepened gradually with the increase of concentration, and maintained when the IgG concentration reached 0.5 mg mL⁻¹. The appropriate amount applied to NC membrane test line was 10 µL at concentration of 0.5 mg mL⁻¹.

For sandwich lateral flow test, the sample plays an important role of providing adequate quantities of antigen to link the two antibodies. More sample volume provides more E. coli O157:H7 at the same concentration. Hence the sample volume was tested. No significant growth of red color intensity was observed when the volume reaching 200 µL. The reason might be that it is hard to complete the reaction between antigen and antibody in a solid-phase sandwich form in short time. The excess E. coli O157:H7-IgY-GM complex may just pass through the test line without capturing by IgG anti-E. coli O157:H7. As a result, 200 µL was determined as the optimal sample volume for use.

Performance and sensitivity of ICA test strip    The experiments of the ICA using the optimized test strip and E. coli O157:H7 standard solutions range from 1×10¹ – 10⁶ CFU mL⁻¹ were preformed in triplicate. The representative results are showed in Fig. 2. It can be seen that a negative control without E. coli O157:H7 gave only a clear red color at the control line and the same intensity of red color appeared on control lines in different strips, indicating validity of the assay. E. coli O157:H7 concentrations at 10¹ caused no color and 10² CFU mL⁻¹ caused weak color development at the test line. While, with the concentration of E. coli O157:H7 increased, the intensity of red color on test lines was appeared and increased. Furthermore, the intensity of red color on test line at 10³ CFU mL⁻¹ of E. coli O157:H7 caused a distinguishable difference compared to the negative control. Therefore, the visual detection limit for E. coli O157:H7 was considered to be 10³ CFU mL⁻¹. In addition, we also evaluated the effect of a food sample matrix on the ICA strips. A series of test strips was applied to characterize the efficiency of the E. coli O157:H7 detection in milk matrices. Compared to E. coli O157:H7 detection in a buffer (PBS), the milk matrices showed the same threshold judging from the appearance of visual staining in the test line of the test strip (10³ CFU mL⁻¹).

Fig. 2.

Representive results of the lateral flow ICA for standard solutions of E. coli O157:H7.

The magnetic signal (MS) intensities of T line of the experiments mentioned above were measured by MAR (Magnetic Assay Reader), and the results were expressed as relative magnetic units (RMUs). As shown in Fig. 3, the signals grew with the increase of E. coli O157:H7 concentrations. A significant linear relationship presented a log-log plot between E. coli O157:H7 concentration and MS of T line in the range from 1×10¹ to 1×10⁶, with a linear coefficient of 0.9954. This provides a quantitive method for E. coli O157:H7 detection. The calculated detection limit was 32 CFU mL⁻¹ (three times the negative sample signal, 3×65 RMUs), which was lower than the detection limit judging by naked eye (about 10³ CFU mL⁻¹) and also closer to the infective dose of E. coli O157:H7 (50 – 100 cells).

Fig. 3.

Standard curves of the ICA results determined by magnetic intensity shown in a log-log plot. X and Y axis are all expressed on logarithm. Note: MS, magnetic signal; RMUs, relative magnetic units.

The results of the milk matrix assays were also analyzed by MS intensity. As can be seen from Fig. 3, MS intensities still presented linear relationship with E. coli O157:H7 concentration in log-log plot. The linear range for quantitive measure has been narrowed down to 10² – 10⁵ CFU mL⁻¹, although the MS intensities indicated the detection limit of 100 CFU mL⁻¹, the same level with that in PBS. The transport velocity of the reagents across the membrane also somewhat slowed down in the assays for milk samples. The results showed that presence of the milk matrix exhibited interference with ICA. Nevertheless, the development of the color in the test zone and control zone can be completed the in 20 min. Taken together, this ICA strip can be applied to screen E. coli O157:H7 in water or polluted milk. The results can be determined by both visual detection and MS intensity. The latter one can quantitively measure the E. coli O157:H7 concentrations at the same time with a lower detection limit.

The capacity of the sandwich reaction was determined by both capture and detection antibodies with high antigen specificity and the sensitive labeling methods. The detections of the bacteria are often poor due to single targeting of the bacteria antigen using mAb both as capture and detection antibodies. For instance, a sandwich ELISA using a mAb to the LPS of E. coli O157 has a detection limit of 10⁵ CFU mL⁻¹ (Kerr et al. 2001). Sunwoo et al. use IgY as a detection antibody showed the detection limit of 10³ CFU mL⁻¹ by ELISA (Sunwoo et al. 2006) . The ICA developed using IgY as the detection antibodies here can achieve a detection limit of 10³ CFU mL⁻¹ without pretreatment of the milk samples, which can be also analyzed by MS intensity with a detection limit of 10² CFU mL⁻¹. This procedure is more rapid and did not use any devices or dangerous compounds and provides an alternative semiquantitive measurement with lower limit of detection at the same time. Therefore, this method is suitable for on-spot quality test of raw milk purchases.

Specificity of the ICA system    The cross-reactivity of the ICA strips with four kinds of E. coli O157:H7 strains and other related bacteria were tested. Table 2 showed the observed results. Six E. coli O157:H7 strains could be detectable by the ICA. Four other STEC strains showed weak cross-reactivity. However, among seven Non-pathogenic laboratory E. coli strains tested, none of them showed cross-reactivity. The interferences of the other bacteria examined to the assays were negligible, especially for the Staphylococcus aureus (S. aureus) which contains protein A in the cell wall capable of covalently binding to the Fc region of IgG of most mammalian species with high affinity (Xiong et al. 2014) but not IgY. In addition, it have been reported that some specific IgY antibodies can be raised in egg yolk (Ferreira Junior et al. 2012). As a result, using as a detecting antibody, IgY can avoid the false positive results in application.

Table 2. Specificity of the ICA strips.

Strain	Number of strains	Results of ICA (Cross-reactivity)
E. coli O157:H7 (ATCC 43895)	1	++
E. coli O157:H7a	6	++
Other STEC E.coli b	4	+
Non-pathogenic laboratory E. coli c	7	−
Salmonella d	2	−
Staphylococcus aureuse	2	−

++: positive result; +: positive result; −: negative result

^a  E. coli O157:H7 strains : ATCC 43888, ATCC 43889, CMCC44828, NCTC 12900, 99X131 and 99X132 JS-CDC

^b  Other STEC E.coli : O26:H11, O111, O103:H2, O145

^c  Lab strains : BL21 DE3, BL21RIPL, JM109, CMCC44102, ATCC25922, CMCC44103, CMCC44350

^d  Salmonella : S. typhimurium ATCC 13311, S. urbana.

^e  Staphylococcus aureus : CMCC26001, CMCC26003

Repetition and Stability of ICA    Spiking samples were analyzed by ICA and the results were observed both by naked eyes and MAR (Table 3). The consistent results were obtained in the same levels with twenty replications, showing good reproducibility and strip-to-strip performance either analyzed by color or by MAR.

Table 3. Results of ICA for spiked samples.

Concentration of E. coli O157:H7 (CFU mL−1)	Visual result	FN rate (%) a	Mean MS of T line (RMUs)	SD of MS b
102	−	95	285.35	±21.65
103	+	10	860.05	±34.25
104	+	0	2550.85	±54.60
105	+	0	8669.05	±80.50

+: positive; − :negative.

^a  FN rate, negative rate was calculated as: %FN=(fn/tp)×100, where fn are negative test samples; tp are truly positive samples.

^b  SD standard deviation.

The stability of MS was evaluated by measuring the ICA strip by MAR on the first and 12th weeks since sample addition. The MS intensity presented good stability with a slightly decline after 12-week exposure at room temperature. The results were in coincide with what has been reported by Xu et al. (Xu et al. 2009). The results supported the idea that MS of this ICA test can remain relatively stable in a long term, compared with signal based on optical measurement.

Conclusions

In this study, it has been demonstrated that the super-paramagnetic nanoparticles with red color (GM) can be employed for the development of ICA strip. An IgY specific to E. coli O157:H7 was prepared and labeled by GM to develop a sandwich ICA for E. coli O157:H7. The ICA was optimized and testified with a limit of 10³ CFU mL⁻¹ for visual detection. Using MS intensity detection, the ICA can quantitively measure E. coli O157:H7 concentration with a detection limit of 32 CFU mL⁻¹ in PBS and 10² CFU mL⁻¹ in pretreatment-free milk samples. The assay could be completed in less than 20 min. The visual detection is fit for supervising the water environment or rapid screening of milk on spot of purchasesand the result are also can be preserved and quantitively reassay by MS intensity.

Acknowledgement    This work was supported by the National Natural Science Foundation of China (Grant No. 31301405), Science and Technology Research and Development Program of Shaanxi Province (Grant No. 2013KTZB02-02-05, 2013KTCQ02-09), China and Science and Technology Project of Xi'an City (No. CXY1434(4)), China.

List of abbreviations

E. coli O157:H7

Escherichia coli O157:H7

IgY

immunoglobins of Yolk

ICA

immunochromatographic assay

GM

Goldmag, A core/shell-structured super-paramagnetic Fe₃O₄/Au composite nanoparticles

MAR

Magnetic Assay Reader

RMUs

Relative magnetic units

MS

Magnetic signal

EI

ELISA Index

NC

Nitrocellulose

References

•  Baylis,  C. L. (2009). Raw milk and raw milk cheeses as vehicles for infection by Verocytotoxin-producing Escherichia coli. Int. J. Dairy Tech., 62, 293-307.
•  Berlina,  A.N.,  Taranova,  N.A.,  Zherdev,  A.V.,  Vengerov,  Y.Y., and  Dzantiev,  B.B. (2013). Quantum dot-based lateral flow immunoassay for detection of chloramphenicol in milk. Anal. Bioanal.Chem., 405, 4997-5000.
•  Byzova,  N.A.,  Smirnova,  N.I.,  Zherdev,  A.V.,  Eremin,  S.A.,  Shanin,  I.A.,  Lei,  H.-T.,  Sun,  Y., and  Dzantiev,  B.B. (2014). Rapid immunochromatographic assay for ofloxacin in animal original foodstuffs using native antisera labeled by colloidal gold. Talanta, 119, 125-132.
•  Cavaiuolo,  M.,  Paramithiotis,  S.,  Drosinos,  E.H., and  Ferrante,  A. (2013). Development and optimization of an ELISA based method to detect Listeria monocytogenes and Escherichia coli O157 in fresh vegetables, Anal. Methods, 5, 4622-4627.
•  Cova,  L. (2005). DNA-designed avian IgY antibodies: novel tools for research, diagnostics and therapy. J. Clin. Virol. 34, S70-S74.
•  Ferreira Junior,  A.,  Santiago,  F.M.,  Silva,  M.V.,  Ferreira,  F.B.,  Macedo Junior,  A.G.,  Mota,  C.M.,  Faria,  M.S.,  Silva,  H.H.,  Silva,  D.A.O.,  Cunha-Junior,  J. P.,  Mineo,  J.R., and  Mineo,  T.W.P. (2012). Production, Characterization and Applications for Toxoplasma gondii-Specific Polyclonal Chicken Egg Yolk Immunoglobulins. Plos One, 7, e40391.
•  Gaulin,  C.,  Levac,  E.,  Ramsay,  D.,  Dion,  R.,  Ismail,  J.,  Gingras,  S., and  Lacroix,  C. (2012). Escherichia coli O157:H7 Outbreak Linked to Raw Milk Cheese in Quebec, Canada: Use of Exact Probability Calculation and Case-Case Study Approaches to Foodborne Outbreak Investigation. J. Food Protect., 75, 812-818.
•  Gehring,  A.G.,  Albin,  D.M.,  Bhunia,  A.K.,  Kim,  H.,  Reed,  S.A., and  Tu,  S.-I. (2012). Mixed culture enrichment of Escherichia coli O157:H7, Listeria monocytogenes, Salmonella enterica, and Yersinia enterocolitica. Food Control, 26, 269-273.
•  Hussein,  H.S. and  Sakuma,  T. (2005). Invited Review: Prevalence of Shiga Toxin-Producing Escherichia coli in Dairy Cattle and Their Products. J. Dairy Sci., 88, 450-465.
•  Keene,  W.E.,  Hedberg,  K.,  Herriott,  D.E.,  Hancock,  D.D.,  McKay,  R.W.,  Barrett,  T.J., and  Fleming,  D.W. (1997). A prolonged outbreak of Escherichia coli O157:H7 infections caused by commercially distributed raw milk. J. Infect. Dis., 176, 815-818.
•  Kerr,  P.,  Chart,  H.,  Finlay,  D.,  Pollock,  D.,  MacKie,  D., and  Ball,  H. (2001). Development of a monoclonal sandwich ELISA for the detection of animal and human Escherichia coli O157 strains. J. Appl. Microbiol., 90, 543-549.
•  Liptakova,  A.,  Siegfried,  L.,  Rosocha,  J.,  Podracka,  L.,  Bogyiova,  E., and  Kotulova,  D. (2004). A family outbreak of haemolytic uraemic syndrome and haemorrhagic colitis caused by verocytotoxigenic Escherichia coli O157 from unpasteurised cow's milk in Slovakia. Clin. Microbiol. Infect., 10, 576-578.
•  Mead,  P.S.,  Slutsker,  L.,  Dietz,  V.,  McCaig,  L.F.,  Bresee,  J.S.,  Shapiro,  C.,  Griffin,  P.M., and  Tauxe,  R.V. (1999). Food-related illness and death in the United States. Emerging Infect. Dis., 5, 607-625.
•  Reddy,  P.K.,  Shekar,  A.,  Kingston,  J.J.,  Sripathy,  M.H., and  Batra,  H. (2013). Evaluation of IgY capture ELISA for sensitive detection of Alpha hemolysin of Staphylococcus aureus without staphylococcal protein A interference. J. Immunol. Methods, 391, 31-38.
•  Rudolf,  J.,  Führer,  M.,  Galler,  B.,  Ansari,  P.,  Hasenhindl,  C., and  Baumgartner,  S. (2009). Differences in usability of rabbit IgG and chicken IgY after clean-up and impact on gold labelling properties. J. Immunol. Methods, 350, 79-88.
•  Strachan,  N.J.,  Fenlon,  D.R., and  Ogden,  I.D. (2001). Modelling the vector pathway and infection of humans in an environmental outbreak of Escherichia coli O157. FEMS Microbiol. Lett., 203, 69-73.
•  Sunwoo,  H.H.,  Wang,  W.W., and  Sim,  J.S. (2006). Detection of Escherichia coli O157:H7 using chicken immunoglobulin Y. Immunol. Lett., 106, 191-193.
•  Wang,  L.,  Li,  P.,  Zhang,  Z.,  Chen,  Q.,  Aguilar,  Z.P.,  Xu,  H.,  Yang,  L.,  Xu,  F.,  Lai,  W.,  Xiong,  Y., and  Wei,  H. (2014). Rapid and accurate detection of viable Escherichia coli O157:H7 in milk using a combined IMS, sodium deoxycholate, PMA and real-time quantitative PCR process. Food Control, 36, 119-125.
•  Wang,  Y.,  Ye,  Z.,  Si,  C., and  Ying,  Y. (2013). Monitoring of Escherichia coli O157:H7 in food samples using lectin based surface plasmon resonance biosensor. Food Chem., 136, 1303-1308.
•  Xiong,  Q.,  Cui,  X.,  Saini,  J.K.,  Liu,  D.,  Shan,  S.,  Jin,  Y., and  Lai,  W. (2014). Development of an immunomagnetic separation method for efficient enrichment of Escherichia coli O157:H7. Food Control, 37, 41-45.
•  Xu,  Q.,  Xu,  H.,  Gu,  H.,  Li,  J.,  Wang,  Y., and  Wei,  M. (2009). Development of lateral flow immunoassay system based on superparamagnetic nanobeads as labels for rapid quantitative detection of cardiac troponin I. Mater. Sci. Eng.: C, 29, 702-707.
•  Yang,  D.,  Ma,  J.,  Zhang,  Q.,  Li,  N.,  Yang,  J.,  Raju,  P.A.,  Peng,  M.,  Luo,  Y.,  Hui,  W.,  Chen,  C., and  Cui,  Y. (2013). Polyelectrolyte-Coated Gold Magnetic Nanoparticles for Immunoassay Development: Toward Point of Care Diagnostics for Syphilis Screening. Anal. Chem., 85, 6688-6695.
•  Yankey,  W.B.,  Chen,  Y.-C., and  Chen,  H. (2012). Evaluation of USEPA Time-Temperature Requirement for Escherichia coli Destruction by Standard Culture Method and Flow Cytometry. J. Residuals Sci. Technol., 9, 9-19.