Rapid pathogen detection by lateral-flow immunochromatographic assay with gold nanoparticle-assisted enzyme signal amplification Il-Hoon Cho, Arun Bhunia, Joseph Irudayaraj PII: DOI: Reference:
S0168-1605(15)00235-4 doi: 10.1016/j.ijfoodmicro.2015.04.032 FOOD 6897
To appear in:
International Journal of Food Microbiology
Received date: Revised date: Accepted date:
21 November 2014 13 April 2015 19 April 2015
Please cite this article as: Cho, Il-Hoon, Bhunia, Arun, Irudayaraj, Joseph, Rapid pathogen detection by lateral-flow immunochromatographic assay with gold nanoparticleassisted enzyme signal amplification, International Journal of Food Microbiology (2015), doi: 10.1016/j.ijfoodmicro.2015.04.032
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
PT
Rapid pathogen detection by lateral-flow immunochromatographic assay with gold nanoparticle-assisted enzyme signal amplification
RI
Il-Hoon Cho1,2, Arun Bhunia3, and Joseph Irudayaraj1,* 1
SC
Bindley Bioscience and Birck Nanotechnology Center, Department of Agricultural & Biological Engineering, Purdue University, West Lafayette, Indiana 47907 2
Department of Food Science, Purdue University, West Lafayette, Indiana 47907
MA
3
NU
Department of Biomedical Laboratory Science, College of Health Science, Eulji University, Seongnam 461-713, Republic of Korea
TE
D
Running head: Pathogen detection with enzyme signal amplification
*
AC CE P
Work was completed at Purdue University and the first author is now at Eulji University
Corresponding Author Joseph Irudayaraj, Professor 225 South University Street Department of Agricultural and Biological Engineering Purdue University West Lafayette, Indiana, 47907 Tel: 765-494-0388 Fax: 765-496-1115 E-mail: [email protected]
1
ACCEPTED MANUSCRIPT
PT
Abstract
To date most LF-ICA format for pathogen detection is based on generating color signals from
RI
gold nanoparticle (AuNP) tracers that are perceivable by naked eye but often these methods
SC
exhibit sensitivity lower than those associated with the conventional enzyme-based immunological methods or mandated by the regulatory guidelines. By developing AuNP avidin-
NU
biotin constructs in which a number of enzymes can be labeled we report on an enhanced LF-ICA
MA
system to detect pathogens at very low levels. With this approach we show that as low as 100 CFU/ml of E. coli O157:H7 can be detected, indicating that the limit of detection can be
D
increased by about 1000-fold due to our signal amplification approach. In addition, extensive
TE
cross-reactivity experiments were conducted (19 different organisms were used) to test and
AC CE P
successfully validate the specificity of the assay. Semi-quantitative analysis can be performed using signal intensities which was correlated with the target pathogen concentrations for calibration by image processing.
Key Words: Lateral-flow immunochromatography, Pathogen detection, Signal enhancement, Rapid detection, Enzyme immunoassay
2
ACCEPTED MANUSCRIPT
PT
1. Introduction
Foodborne illness or commonly known as food poisoning is a major public health concern often
RI
triggered by pathogens that enter into a host via ingestion of contaminated food samples.
SC
According to the Center for Disease Control and Prevention, each year 1 in 6 Americans get sick
NU
due to food poisoning associated with notorious pathogens such as E. coli O157:H7, S. typhimurium, S. enteritidis, L. monocytogenes, etc. These pathogens have also been responsible
MA
for the most number of cases of hospitalization and deaths arising from foodborne illness (DeWaal et al., 2012). With growing concern over public health, decision makers are imposing
D
stringent food safety regulatory codes on major food and processing industries to avoid outbreak.
TE
Thus simple and convenient tools that will help in rapid screening to provide quick assessment of
AC CE P
sample integrity can prevent the occurrence of disease or outbreak due to these pathogens and is of significant interest (Swaminathan and Feng, 1994) to both government and private agencies.
Reports on pathogen detection have usually focused on the following methodologies such as traditional culture collection (Reissbrodt, 2004; Velusamy et al., 2010), genetic approaches based on polymerase chain reaction (PCR) (Bennett et al., 1998; Gallegos‐Robles et al., 2009; Malorny et al., 2003), and immunological methods (Cho et al., 2014; Pandey et al., 2014), although other methods such as scattering spectroscopy (Banada et al., 2009; Bhunia, 2008; Rajwa et al., 2010) and phage-based detection (Brovko et al., 2012; Goodridge et al., 1999; Smartt et al., 2012) have also been used. Methods such as colony counting and biochemical methods are effective and reliable and are the gold standard, but takes several days and are accompanied with laborious steps. Such problems can be partly avoided if genetic and immunological methods are used 3
ACCEPTED MANUSCRIPT owing to their better traits (e.g., simplicity and sensitivity). However, these alternative methods still require multiple reaction steps, prolonged detection time, well-equipped facilities, well-
RI
PT
trained person, and even a pre-enrichment step resulting in a 20-24 hrs turnaround time.
Lateral Flow Immunochromatographic Assay (LF-ICA) is a simple immunological, practical, and
SC
a widely used technique for the detection of the presence (or absence) of a target analyte in
NU
sample matrices without the need for extensive steps (Cho et al., 2014; Cho and Irudayaraj, 2013; Pengsuk et al., 2013; Ravindranath et al., 2009; Subramanian et al., 2006; Wang and Irudayaraj,
MA
2010; Wang et al., 2010) A lateral flow through the immunostrip that consists of different types of membranes expedites the reaction by capillary flow phenomenon, allowing it to be completed
D
in a relatively short time (around 10 min) and further enabling the in situ separation of unreacted
TE
components via one-step analysis, which is a key advantage of this format. Other advantages of
AC CE P
this rapid immunoassay are: 1) colorimetric detection by naked-eye monitoring when gold nanoparticles are adopted as a label and 2) ease-of-use (no need for a trained person). However, the low analytical sensitivity compared with the conventional enzyme immunoassay (e.g., ELISA) is a drawback in this method (Cho et al., 2006; Cho et al., 2005), necessitating pre-enrichment steps to compensate for the sensitivity, requiring longer detection time.. To overcome this problem, enzyme tracers have been employed as an alternative to enhance the signals from the LF-ICA platform. An enzyme tracer can produce signals resulting from its relatively fast catalytic reaction along with a substrate, producing a visible color stain on the membrane. However the detection sensitivity is not sufficient to detect extremely low concentration of target pathogens in the sample.
4
ACCEPTED MANUSCRIPT To achieve higher detection sensitivity, we developed a simple and rapid LF-ICA system employing gold nanoparticles (AuNPs) as a carrier of enzyme. Since AuNPs have a large surface
PT
area, a number of enzymes technically can be coupled to the solid support, enabling a significant
RI
signal amplification to detect extremely low concentration of target pathogens. E. coli O157:H7
NU
SC
was used as a model pathogen in this study to assess the limit of detection and specificity.
TE
D
2.1. Materials
MA
2. Materials and Methods
Polyclonal antibodies (pAb) raised from goat which are reactive to E. coli O157:H7 (01-95-90)
AC CE P
was purchased from KPL (Gaithersburg, MD). 19 different bacteria (see Table 1 for strains and types) were obtained from the Center for Food Safety Engineering consortium at Purdue University. Sodium citrate, sodium carbonate, chloroauric acid (HAuCl4), casein (sodium salt type), Tween 20 and tetramethyl benzidine (TMB) were obtained from Sigma (St. Louis, MO). Horseradish peroxidase, streptavidin, sulfo-NHS-LC-Biotin, LC-SMCC, sulfo-LC-SPDP, dithiothreitol (DTT) and mouse anti-goat antibody labeled with HRP were purchased from Pierce (Rockford, IL). Glass fiber membrane (Ahlstrom 8980) and cellulose membrane (17 CHR) were supplied by Whatman (Kent, UK). Glass membrane (PT-R5) and nitrocellulose membrane (70 CNPH-N-SS40) were purchased from Advanced Microdevices (Ambala Cantt, India). Other reagents used were of analytical grade.
5
ACCEPTED MANUSCRIPT 2.2. Preparation of conjugates Two conjugates were prepared. Conjugate I consists of the AuNP with biotin cross linkers. Here,
PT
1 mL of AuNP solution synthesized via citrate reduction (Makhsin et al., 2012) (40 nm in diameter) was sequentially mixed with 1 μL of 0.5 M of sodium carbonate, 100 μL of phosphate
RI
buffer (10 mM, pH 7.4) and 10 μg of E. coli O157:H7 antibody solution, which was stirred in a
SC
rotary unit for 2 hrs. 5% (w/v) casein dissolved in 10 mM phosphate buffer (122 μL) was added
NU
and then stirred for 1 hr to block residual surfaces of the AuNP. Unbound molecules were removed by centrifugation (12,000 rpm, 15 min) and the pellet was resuspended using 1 mL of
MA
PBS, which was repeated two additional times. This conjugate was biotinylated by adding 10 μg of sulfo-NHS-LC-biotin dissolved in PBS for 30 min followed by removal of unbound biotin-
TE
D
linkers by centrifugation at identical conditions as stated above.
AC CE P
Conjugate II is the streptavidin-HRP conjugate, where, streptavidin (300 μg) will be first labeled with LC-SMCC dissolved in DMSO at 50-fold molar excess at room temperature for 1 hr and the excess LC-SMCC reagents were removed by size exclusion chromatography (Sephadex G-15). HRP was also activated with sulfo-LC-SPDP dissolved in PBS at 25-fold molar excess, followed by DTT reduction (final 10 mM) and purification on Sephadex G-15 sequentially. The streptavidin was then mixed with the activated HRP which is 10 molar in excess and the reaction was carried out at room temperature for 4 hrs.
2.3. Analysis of conjugate components Enzyme immunoassay: 100 μL of E. coli O157:H7 (1 x 107 CFU/mL) resuspended in PBS was coated on the microwell at 37℃ for 1 hr. This incubation condition was identically applied to
6
ACCEPTED MANUSCRIPT perform the following reaction steps. The residual surface was blocked with 200 µL of 0.5% casein dissolved in PBS (Casein-PBS) for 1 hr and then 100 μL of conjugate I diluted in the
PT
Casein-PBS was reacted. For verification of antibody attachment and biotinylation of the gold
RI
surface, mouse anti-goat IgG coupled to HRP and streptavidin-HRP conjugate were reacted respectively. For signal generation 200 µL of substrate solution (50 mM sodium acetate: 1% (v/v)
SC
TMB: 3% (v/v) hydrogen peroxide = 1000: 10: 1) was added and maintained for 15 min. The
NU
reaction was terminated by adding 50 µL of 2 M sulfuric acid, the optical density was measured
MA
with an ELISA reader (VERSAmax; Molecular Device, Sunnyvale, CA) at 450 nm.
SDS-PAGE analysis: conjugate II, streptavidin and protein marker were treated with sample
D
buffer without mercaptoethanol (non-reducing condition) and loaded on to a 10% polyacrylamide
TE
gel. For electrophoretic separation, two voltages (80 V and 150 V) were sequentially applied for
AC CE P
20 min and 60 min respectively using a Mini-Protein Tetra Cell apparatus (Bio-Rad, Hercules, CA). The separating gel was sequentially immersed in a staining solution including 0.1% Coomassie blue R-250 for 15 min and the destaining solution for 1 hr for the development of protein bands.
2.4. Analytical procedure of LF-ICA assay E. coli O157:H7 antibody (0.5 μg) was used as a capture agent and coated on nitrocellulose membrane which was dried at 37℃. To initiate the immunoassay on the membrane strip, conjugate I and II were placed on the corresponding (absorption and conjugate pads respectively) followed by the introduction of the sample (in this demonstration, E. coli O157:H7 diluted in Casein-PBS (100 µL)) to the sample absorption pad (Fig 1). After 15 min of sample flow to allow 7
ACCEPTED MANUSCRIPT the immunoreaction to occur, the strip was washed with 40 µL of deionized water by connecting two absorption pads on the other side of the nitrocellulose membrane to induce cross-flow. For
PT
signal generation, a colorimetric substrate (i.e., TMB) was applied to the absorption pad and then
RI
washed with deionized water to stop the reaction. The color signals that appeared on the nitrocellulose membrane of the immunostrip was captured by scanning and the density values
NU
SC
were then obtained from the area where signal is produced.
2.5. Cross-reactivity test
MA
Specificity experiments were performed against 19 different bacteria (see Table 1 for the different strains and types of bacteria and Fig. 6), which were respectively cultured in 500 mL of
D
Tryptic Soy Broth (TSB) media at 37oC with shaking, for 18 hrs and then harvested by
TE
centrifugation (5000 rpm, 20 min) and stored in 50 mL of sterilized PBS. The concentration of
AC CE P
bacteria in these stock solutions was determined in triplicate by enumeration on plate count agar following incubation at 37°C for 24 hrs. The analytical procedure was the same as described above. The bacteria concentration was maintained at a high level (1 x 107 CFU/mL) in caseinPBS in the cross-reactivity experiments.
2.6. Food sample testing To test the feasibility of the method for pathogen detection in food, the assay was performed to test the presence of E. coli O157:H7 in intentionally inoculated ground beef samples. Uninoculated sample was used as control. The concentration of E. coli O157:H7 inoculated was determined by colony counting. The sampling procedure is as follows: a 10 g sample of ground beef was mixed with 10 mL of sterilized PBS and inoculated with E.coli O157:H7. After 1 hr of inoculation, the assay was performed in triplicate as described above. 8
ACCEPTED MANUSCRIPT
2.7. Biosafety concerns
PT
All experiments were conducted in BSL-2 approved facilities with proper safety training and
RI
approvals in place.
NU
SC
3. Results and Discussion
3.1. Analytical concept
MA
Our goal was to develop a high performance biosensor for onsite monitoring of select foodborne pathogens. The concept of LF-ICA was employed because the assay was simple and the signal
TE
D
can be amplified with enhancement tools. Since the signals obtained from the immunoreaction are directly proportional to the amount of enzyme (i.e., HRP) participating in the reaction, it is
AC CE P
crucial to tether as many HRP molecules to the AuNP surfaces used as a carrier of HRP to obtain high sensitivity (Fig. 1). Two conjugates are designed for this purpose: conjugate I (biotinylated gold nanoparticle) and a conjugate II (streptavidin-HRP). The conjugate II can be efficiently bound to biotin tethered to the AuNPs via the strong avidin-biotin interaction with minimal steric hindrance, resulting in a conjugate with multiple enzymes owing to the large surface area of AuNPs.
9
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
MA
Fig. 1. Schematic of gold nanoparticle (AuNP)-assisted signal amplification on LF-ICA for pathogen detection. The two conjugates used for signal enhancement consists of antibody, AuNP and biotin-terminated cross-linkers (conjugate I) and streptavidin and horseradish peroxidase
D
(conjugate II). Enzyme-catalyzed colorimetric signal will be produced by supplying a substrate
TE
(e.g., tetramethyl benzidine with hydrogen peroxide) in the cross-flow direction at the site of
AC CE P
antigen-antibody sandwich complex formation.
The analytical protocol is a two-step procedure consisting of (i) initiation of sample flow to induce immunocomplex formation (antibody-pathogen-conjugates binding) and (ii) supplying enzyme substrate for colorimetric signal generation in the cross-flow direction. First, the sample containing the target pathogen will be applied at the bottom of an immunostrip (sample absorption pad) to initiate a flow in the vertical (longitudinal) direction, inducing immunocomplex formation at a pre-determined site where the capture antibody is immobilized. During this flow, a number of SA-HRP conjugates at the conjugate pad can be coupled to the branched biotin on to conjugate I, to initiate an enzyme-network formation. Second, two horizontally arranged pads are placed on each lateral side of the immunostrip and the substrate containing tetramethyl benzidine and hydrogen peroxide is then added onto the substrate supply pad to initiate the enzymatic signal generation (horizontal flow). At the sites of complex 10
ACCEPTED MANUSCRIPT formation, colorimetric signals can be produced by supplying enzyme substrates at the time of
PT
signal generation.
RI
3.2. Characterization of the assay components
Since two key conjugates are synthesized, these need to be well characterized to ensure their
SC
physical and immunochemical properties. First, conjugate I was analyzed by three different
NU
analytical methods as follows: (a) UV/Vis spectrometry, (b) zeta potential measurement and (c) solid phase enzyme immunoassay (Fig. 2). From the spectra we note that the wavelength peak
MA
position of conjugate I was red-shifted by ~5 nm (527 nm to 532 nm) compared with that of bare AuNP. This is because of the two types of protein, i.e., antibody specific to E. coli O157:H7 and
D
casein used as a blocking agent were well adsorbed onto the gold surfaces. This is consistent with
TE
the previously reported results (Wang et al., 2011). This is also supported by the fact that no
was tested).
AC CE P
aggregated forms were observed under high concentration of sodium chloride (up to 3 M NaCl
Fig. 2. Characterization of conjugate I used for signal enhancement using UV/Vis spectrometer (a), zeta potential measurement (b) and enzyme immunoassay (c).
11
ACCEPTED MANUSCRIPT The difference between bare AuNP and conjugate I can also be assessed from zeta potential values shown in Fig. 2b. Negatively charged bare AuNP was slightly shifted to the positive
PT
direction at conjugate I, which can be attributed to the masking of the surface via protein
RI
adsorption. Attachment of antibody and biotin to the surface was verified by enzyme immunoassay in which the analyte (i.e., E. coli O157:H7) was immobilized on a microwell
SC
surface and the two molecules were identified by HRP-labeled secondary antibody and
NU
streptavidin respectively. Obtaining absorbance signals from antibody and biotin shown in Fig. 2c indicates that antibodies and biotin-terminated cross-linkers were well coupled to the
AC CE P
TE
D
MA
nanoparticle surface.
Fig. 3. Characterization of conjugate II used for signal enhancement by (a) SDS-PAGE and (b) enzyme assay. The number of HRP molecules on the conjugate and minimal amount of HRP molecules can be determined by PAGE and enzyme analysis respectively. Combining the information with HABA analysis (c), the number of HRP molecules which can be bound to a single AuNP is estimated to be ~109 (see text for details).
12
ACCEPTED MANUSCRIPT Second the conjugate II was characterized by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) to measure the degree of conjugation between SA and HRP. From
PT
the result shown in Fig. 3a, the mole-to-mole ratio between SA and HRP at conjugate II can be
RI
estimated to be about 1 to 3. This was caused by the fact that the molecular weight of SA and HRP is 60 kDa and 44 kDa respectively. Thus the SA-HRP conjugate would be positioned
SC
around 190 kDa as shown in Fig. 3a. Another important parameter that needs to be assessed is
NU
the number of HRP molecules conjugated to one AuNP. The number of HRP molecules was determined by enzyme assay and HABA (4'-hydroxyazobenzene-2-carboxylic acid) analysis
MA
(Janolino et al., 1996). From the enzyme assay a minimal amount of HRP required for signal generation was calculated (see Fig. 3b) as ~0.5 attomole. Assuming all AuNPs are on an average
D
spherical in shape, with similar size characteristics and known concentration of AuNPs (e.g., 0.14
TE
pM at 40 nm-sized), the number of HRP molecules on the surface of a single AuNP can be
AC CE P
estimated from HABA analysis. Per our preliminary experiment, approximately 109 HRP molecules would be tethered to a single 40 nm size gold nanoparticle bearing biotin-containing chemical linkers (Fig. 3c).
3.3. Analytical performances
With this enzyme signal amplification strategy feasibility experiments were performed against different concentration of E. coli O157:H7. Gold nanoparticle that is considered as a conventional tracer in LF-ICA was used as a reference in comparison with the enzyme assay. From the results shown in Fig. 4a, it can be noted that the detection capability of the AuNP-based method is not sufficient to detect low number of bacteria. The limit of detection (LOD) using these constructs was 1 x 105 CFU/mL.
13
SC
RI
PT
ACCEPTED MANUSCRIPT
NU
Fig. 4. Comparison of conventional method based on AuNP with the developed method employing AuNP-assisted enzyme-based signal amplification on a LF-ICA platform against the
MA
target pathogen, E. coli O157:H7.
The low analytical sensitivity can be overcome by applying enzyme an amplification strategy
TE
D
proposed in this study. Compared with the result based on AuNP, extremely low numbers of E. coli O157:H7 could be detected (~100 CFU/mL) by applying AuNP-assisted enzyme-based
AC CE P
signal amplification (see Fig. 4b), providing an approximate 1000-fold signal enhancement over the existing LF methods. This indicates that numerous enzyme molecules bound to the AuNP could form an enzyme network via biotin-SA linkage and participate in this signal amplification event as expected without losing their enzymatic activity. Thus very low numbers of E. coli O157:H7 are detectable yielding a LOD of ~100 CFU/mL using this method. The LOD defined by multiplying the standard deviation at zero dose (0 CFU/mL) by three (Galikowska et al., 2011; Kim et al., 1999) were determined from dose response curve (Fig. 5b) as 100 CFU/mL.
14
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
Fig. 5. Analytical performance of AuNP-assisted enzyme signal amplification with LF-ICA to
MA
detect E. coli O157:H7. (a) signals from captured images produced on the immunostrip after the application of a chromogenic substrate solution. (b) dose response curve against E. coli O157:H7
TE
D
concentration.
Along with analytical sensitivity, specificity is also considered as one of the major factors that
AC CE P
affect the analytical performance of the biosensor. Since many antigenic sites of the bacteria surface are similar to others (e.g., lipopolysaccharides, K-, O-, and H- antigens), it is very crucial to differentiate other bacteria from the target pathogen to minimize false positives. Hence specificity of this method was examined against 19 different bacteria strains including E. coli O157:H7 (see Table 1 and Fig. 6) by performing enzyme immunoassay (Fig. 6a) and LF-ICA (Fig. 6b). Specificity experiments were performed at a higher concentration of bacteria (1 x 107 CFU/mL) to ensure a reliable performance of this method. For enzyme immunoassay, bacteria samples (S. typhimurium, S. enteritidis and L. monocytogenes including E. coli O157:H7) were coated onto a microtiter well plate surface and then evaluated by conjugate I and II sequentially. For LF-ICA experiments, each bacteria was detected by applying a sandwich immunoassay (i.e., capture antibody-bacteria-detection antibody) in which AuNP was employed as a tracer.
15
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
Table 1. List of bacteria strains for specificity test
16
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
Fig. 6. Specificity tests against various bacteria strains. All bacteria were cultured in TSB media and tested at a concentration of ~107 CFU/mL. (a) enzyme immunoassay and (b) LF-ICA.
From these experiments it was obvious that O157:H7 antibodies used in the proposed strategy for AuNP-assisted enzyme amplification had no cross-reactivity with the other competitive strains even at a high concentration of such bacteria in the sample and no signals were obtained, except for sample #5 and #11. The signal obtained from sample #11 (E. coli ATCC 35150) was due to the fact that the strain has the same antigenic properties as the serotype E. coli O157:H7. Specificity experiments against a wide range of bacteria strains provides a strong validation of the assay highlighting its specificity to the target pathogen (i.e., E. coli O157:H7) with high fidelity.
17
ACCEPTED MANUSCRIPT This AuNP-assisted signal amplified LF-ICA method was next used to test pathogens in food samples as a final validation step. Ground beef was adopted as a sample matrix and an initial
PT
concentration of E. coli O157:H7 was purposefully inoculated and enumerated by colony
RI
counting in triplicate. Three different inoculated samples (0, 437 and 984 CFU/mL) were maintained for 1 hr followed by testing of the supernatant for the presence of pathogens. As
SC
shown in Fig. 7, the signals obtained were consistent with those of the dose response in Fig. 5b,
NU
indicating that the developed method is robust and the antibodies are still reactive in a different environment. Although the samples were maintained 1 hr prior to analysis, our expectation is that
MA
most of the bacteria present in the lag phase did not affect bacteria proliferation. Therefore it can be concluded that a pre-enrichment step is not a requirement in this assay system unlike most of
AC CE P
TE
D
the pathogen immunoassays that need a prolonged culturing step.
Fig. 7. Performance of the LF-ICA platform in a food matrix (e.g., ground beef) innoculated with E. coli O157:H7. The initial number of E. coli was assessed by colony counting in LB agar plate. The assay was performed after 1 hr of incubation.
18
ACCEPTED MANUSCRIPT Results of our studies with the developed biosensor show high sensitivity and specificity, including detection of the pathogen in a relevant food matrix. We expect our approach can be
RI
PT
extended to other organisms and for multiplex detection of pathogens.
NU
SC
4. Conclusions
MA
In conclusion, we report a highly sensitive AuNP-assisted enzyme signal amplification method based on a LF-ICA format to detect very low number of pathogens. The proposed format is rapid
D
(the whole procedure can be completed within 15 min), simple (only two steps are required;
TE
vertical and cross flow for immunoreaction and signal generation respectively), low cost (no need to use fluorometer or luminometer because signal is amplified for visual inspection) and practical.
AC CE P
Very high specificity was demonstrated by testing against numerous bacteria strains and food pathogens as well as in a food sample. We expect our signal enhancing approach can be efficiently utilized for timely detection of foodborne pathogens to prevent an outbreak in the field.
Acknowledgements Research was supported by funds from USDA-ARS project number 1935-42000-049-00D in conjunction with the Center for Food Safety Engineering at Purdue University.
19
ACCEPTED MANUSCRIPT References
AC CE P
TE
D
MA
NU
SC
RI
PT
Banada, P.P., Huff, K., Bae, E., Rajwa, B., Aroonnual, A., Bayraktar, B., Adil, A., Robinson, J.P., Hirleman, E.D., Bhunia, A.K., 2009. Label-free detection of multiple bacterial pathogens using light-scattering sensor. Biosensors and Bioelectronics 24, 1685-1692. Bennett, A., Greenwood, D., Tennant, C., Banks, J., Betts, R., 1998. Rapid and definitive detection of Salmonella in foods by PCR. Letters in applied microbiology 26, 437-441. Bhunia, A.K., 2008. Biosensors and bio‐based methods for the separation and detection of foodborne pathogens. Advances in food and nutrition research 54, 1-44. Brovko, L.Y., Anany, H., Griffiths, M.W., 2012. 6 Bacteriophages for Detection and Control of Bacterial Pathogens in Food and Food-Processing Environment. Advances in food and nutrition research 67, 241. Cho, I.-H., Bhandari, P., Patel, P., Irudayaraj, J., 2015. Membrane filter-assisted surface enhanced Raman spectroscopy for the rapid detection of E. coli O157: H7 in ground beef. Biosensors and Bioelectronics 64, 171-176. Cho, I.-H., Irudayaraj, J., 2013. Lateral-flow enzyme immunoconcentration for rapid detection of Listeria monocytogenes. Analytical and bioanalytical chemistry 405, 3313-3319. Cho, I.-H., Mauer, L., Irudayaraj, J., 2014. In-situ fluorescent immunomagnetic multiplex detection of foodborne pathogens in very low numbers. Biosensors and Bioelectronics 57, 143148. Cho, J.-H., Han, S.-M., Paek, E.-H., Cho, I.-H., Paek, S.-H., 2006. Plastic ELISA-on-a-chip based on sequential cross-flow chromatography. Analytical Chemistry 78, 793-800. Cho, J.-H., Paek, E.-H., Cho, I.-H., Paek, S.-H., 2005. An enzyme immunoanalytical system based on sequential cross-flow chromatography. Analytical Chemistry 77, 4091-4097. DeWaal, C.S., Roberts, C., Catella, C., 2012. Outbreak Alert! 1999-2008. Center for Science in the Public Interest. http://www. cspinet. org/foodsafety/PDFs/Outbreak_Alert_1999-2008. pdf. Galikowska, E., Kunikowska, D., Tokarska-Pietrzak, E., Dziadziuszko, H., Łoś, J.M., Golec, P., Węgrzyn, G., Łoś, M., 2011. Specific detection of Salmonella enterica and Escherichia coli strains by using ELISA with bacteriophages as recognition agents. European journal of clinical microbiology & infectious diseases 30, 1067-1073. Gallegos‐Robles, M., Morales‐Loredo, A., Álvarez‐Ojeda, G., Osuna‐García, J., Martinez, I., Morales‐Ramos, L., Fratamico, P., 2009. PCR detection and microbiological isolation of Salmonella spp. from fresh beef and cantaloupes. Journal of food science 74, M37-M40. Goodridge, L., Chen, J., Griffiths, M., 1999. Development and characterization of a fluorescentbacteriophage assay for detection of Escherichia coli O157: H7. Applied and environmental microbiology 65, 1397-1404. Janolino, V.G., Fontecha, J., Swaisgood, H.E., 1996. A spectrophotometric assay for biotinbinding sites of immobilized avidin. Applied Biochemistry and Biotechnology 56, 1-7. Kim, J.W., Jin, L.Z., Cho, S.H., Marquardt, R.R., Frohlich, A.A., Baidoo, S.K., 1999. Use of chicken egg‐yolk antibodies against K88+ fimbrial antigen for quantitative analysis of enterotoxigenic Escherichia coli (ETEC) K88+ by a sandwich ELISA. Journal of the Science of Food and Agriculture 79, 1513-1518. Makhsin, S.R., Razak, K.A., Noordin, R., Zakaria, N.D., Chun, T.S., 2012. The effects of size and synthesis methods of gold nanoparticle-conjugated MαHIgG4 for use in an immunochromatographic strip test to detect brugian filariasis. Nanotechnology 23, 495719. 20
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
Malorny, B., Tassios, P.T., Rådström, P., Cook, N., Wagner, M., Hoorfar, J., 2003. Standardization of diagnostic PCR for the detection of foodborne pathogens. International journal of food microbiology 83, 39-48. Pandey, S.K., Vinayaka, A.C., Rishi, D.B., Rishi, P., Suri, C., 2014. Immuno-fluorescence based Vi capsular polysaccharide detection for specific recognition of Salmonella enterica serovar Typhi in clinical samples. Analytica Chimica Acta. Pengsuk, C., Chaivisuthangkura, P., Longyant, S., Sithigorngul, P., 2013. Development and evaluation of a highly sensitive immunochromatographic strip test using gold nanoparticle for direct detection of Vibrio cholerae O139 in seafoodsamples. Biosensors and Bioelectronics 42, 229-235. Rajwa, B., Dundar, M.M., Akova, F., Bettasso, A., Patsekin, V., Dan Hirleman, E., Bhunia, A.K., Robinson, J.P., 2010. Discovering the unknown: Detection of emerging pathogens using a label‐free light‐scattering system. Cytometry Part A 77, 1103-1112. Ravindranath, S.P., Mauer, L.J., Deb-Roy, C., Irudayaraj, J., 2009. Biofunctionalized magnetic nanoparticle integrated mid-infrared pathogen sensor for food matrixes. Analytical Chemistry 81, 2840-2846. Reissbrodt, R., 2004. New chromogenic plating media for detection and enumeration of pathogenic Listeria spp.—an overview. International journal of food microbiology 95, 1-9. Smartt, A.E., Xu, T., Jegier, P., Carswell, J.J., Blount, S.A., Sayler, G.S., Ripp, S., 2012. Pathogen detection using engineered bacteriophages. Analytical and bioanalytical chemistry 402, 31273146. Subramanian, A., Irudayaraj, J., Ryan, T., 2006. A mixed self-assembled monolayer-based surface plasmon immunosensor for detection of E. coli O157: H7. Biosensors and Bioelectronics 21, 998-1006. Swaminathan, B., Feng, P., 1994. Rapid detection of food-borne pathogenic bacteria. Annual Reviews in Microbiology 48, 401-426. Velusamy, V., Arshak, K., Korostynska, O., Oliwa, K., Adley, C., 2010. An overview of foodborne pathogen detection: in the perspective of biosensors. Biotechnology advances 28, 232254. Wang, C., Irudayaraj, J., 2010. Multifunctional magnetic–optical nanoparticle probes for simultaneous detection, separation, and thermal ablation of multiple pathogens. Small 6, 283-289. Wang, Y., Lee, K., Irudayaraj, J., 2010. Silver nanosphere SERS probes for sensitive identification of pathogens. The Journal of Physical Chemistry C 114, 16122-16128. Wang, Y., Ravindranath, S., Irudayaraj, J., 2011. Separation and detection of multiple pathogens in a food matrix by magnetic SERS nanoprobes. Analytical and bioanalytical chemistry 399, 1271-1278.
21
ACCEPTED MANUSCRIPT
Highlights
AC CE P
TE
D
MA
NU
SC
RI
PT
1) Detection of O157:H7 in < 20 min in food matrices 2) Highly sensitive with a limit of detection of ~100 cfu/ml for routine onsite measurements (a 1000-fold increment compared to the conventional lateral flow systems). 3) Excellent potential to be generalized and expanded to other pathogens for detection of pathogens in food, water, and human samples. 4) Highly specific as demonstrated by testing a robust set of strains for detection specificity
22
S0168-1605(15)00235-4 doi: 10.1016/j.ijfoodmicro.2015.04.032 FOOD 6897
To appear in:
International Journal of Food Microbiology
Received date: Revised date: Accepted date:
21 November 2014 13 April 2015 19 April 2015
Please cite this article as: Cho, Il-Hoon, Bhunia, Arun, Irudayaraj, Joseph, Rapid pathogen detection by lateral-flow immunochromatographic assay with gold nanoparticleassisted enzyme signal amplification, International Journal of Food Microbiology (2015), doi: 10.1016/j.ijfoodmicro.2015.04.032
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
PT
Rapid pathogen detection by lateral-flow immunochromatographic assay with gold nanoparticle-assisted enzyme signal amplification
RI
Il-Hoon Cho1,2, Arun Bhunia3, and Joseph Irudayaraj1,* 1
SC
Bindley Bioscience and Birck Nanotechnology Center, Department of Agricultural & Biological Engineering, Purdue University, West Lafayette, Indiana 47907 2
Department of Food Science, Purdue University, West Lafayette, Indiana 47907
MA
3
NU
Department of Biomedical Laboratory Science, College of Health Science, Eulji University, Seongnam 461-713, Republic of Korea
TE
D
Running head: Pathogen detection with enzyme signal amplification
*
AC CE P
Work was completed at Purdue University and the first author is now at Eulji University
Corresponding Author Joseph Irudayaraj, Professor 225 South University Street Department of Agricultural and Biological Engineering Purdue University West Lafayette, Indiana, 47907 Tel: 765-494-0388 Fax: 765-496-1115 E-mail: [email protected]
1
ACCEPTED MANUSCRIPT
PT
Abstract
To date most LF-ICA format for pathogen detection is based on generating color signals from
RI
gold nanoparticle (AuNP) tracers that are perceivable by naked eye but often these methods
SC
exhibit sensitivity lower than those associated with the conventional enzyme-based immunological methods or mandated by the regulatory guidelines. By developing AuNP avidin-
NU
biotin constructs in which a number of enzymes can be labeled we report on an enhanced LF-ICA
MA
system to detect pathogens at very low levels. With this approach we show that as low as 100 CFU/ml of E. coli O157:H7 can be detected, indicating that the limit of detection can be
D
increased by about 1000-fold due to our signal amplification approach. In addition, extensive
TE
cross-reactivity experiments were conducted (19 different organisms were used) to test and
AC CE P
successfully validate the specificity of the assay. Semi-quantitative analysis can be performed using signal intensities which was correlated with the target pathogen concentrations for calibration by image processing.
Key Words: Lateral-flow immunochromatography, Pathogen detection, Signal enhancement, Rapid detection, Enzyme immunoassay
2
ACCEPTED MANUSCRIPT
PT
1. Introduction
Foodborne illness or commonly known as food poisoning is a major public health concern often
RI
triggered by pathogens that enter into a host via ingestion of contaminated food samples.
SC
According to the Center for Disease Control and Prevention, each year 1 in 6 Americans get sick
NU
due to food poisoning associated with notorious pathogens such as E. coli O157:H7, S. typhimurium, S. enteritidis, L. monocytogenes, etc. These pathogens have also been responsible
MA
for the most number of cases of hospitalization and deaths arising from foodborne illness (DeWaal et al., 2012). With growing concern over public health, decision makers are imposing
D
stringent food safety regulatory codes on major food and processing industries to avoid outbreak.
TE
Thus simple and convenient tools that will help in rapid screening to provide quick assessment of
AC CE P
sample integrity can prevent the occurrence of disease or outbreak due to these pathogens and is of significant interest (Swaminathan and Feng, 1994) to both government and private agencies.
Reports on pathogen detection have usually focused on the following methodologies such as traditional culture collection (Reissbrodt, 2004; Velusamy et al., 2010), genetic approaches based on polymerase chain reaction (PCR) (Bennett et al., 1998; Gallegos‐Robles et al., 2009; Malorny et al., 2003), and immunological methods (Cho et al., 2014; Pandey et al., 2014), although other methods such as scattering spectroscopy (Banada et al., 2009; Bhunia, 2008; Rajwa et al., 2010) and phage-based detection (Brovko et al., 2012; Goodridge et al., 1999; Smartt et al., 2012) have also been used. Methods such as colony counting and biochemical methods are effective and reliable and are the gold standard, but takes several days and are accompanied with laborious steps. Such problems can be partly avoided if genetic and immunological methods are used 3
ACCEPTED MANUSCRIPT owing to their better traits (e.g., simplicity and sensitivity). However, these alternative methods still require multiple reaction steps, prolonged detection time, well-equipped facilities, well-
RI
PT
trained person, and even a pre-enrichment step resulting in a 20-24 hrs turnaround time.
Lateral Flow Immunochromatographic Assay (LF-ICA) is a simple immunological, practical, and
SC
a widely used technique for the detection of the presence (or absence) of a target analyte in
NU
sample matrices without the need for extensive steps (Cho et al., 2014; Cho and Irudayaraj, 2013; Pengsuk et al., 2013; Ravindranath et al., 2009; Subramanian et al., 2006; Wang and Irudayaraj,
MA
2010; Wang et al., 2010) A lateral flow through the immunostrip that consists of different types of membranes expedites the reaction by capillary flow phenomenon, allowing it to be completed
D
in a relatively short time (around 10 min) and further enabling the in situ separation of unreacted
TE
components via one-step analysis, which is a key advantage of this format. Other advantages of
AC CE P
this rapid immunoassay are: 1) colorimetric detection by naked-eye monitoring when gold nanoparticles are adopted as a label and 2) ease-of-use (no need for a trained person). However, the low analytical sensitivity compared with the conventional enzyme immunoassay (e.g., ELISA) is a drawback in this method (Cho et al., 2006; Cho et al., 2005), necessitating pre-enrichment steps to compensate for the sensitivity, requiring longer detection time.. To overcome this problem, enzyme tracers have been employed as an alternative to enhance the signals from the LF-ICA platform. An enzyme tracer can produce signals resulting from its relatively fast catalytic reaction along with a substrate, producing a visible color stain on the membrane. However the detection sensitivity is not sufficient to detect extremely low concentration of target pathogens in the sample.
4
ACCEPTED MANUSCRIPT To achieve higher detection sensitivity, we developed a simple and rapid LF-ICA system employing gold nanoparticles (AuNPs) as a carrier of enzyme. Since AuNPs have a large surface
PT
area, a number of enzymes technically can be coupled to the solid support, enabling a significant
RI
signal amplification to detect extremely low concentration of target pathogens. E. coli O157:H7
NU
SC
was used as a model pathogen in this study to assess the limit of detection and specificity.
TE
D
2.1. Materials
MA
2. Materials and Methods
Polyclonal antibodies (pAb) raised from goat which are reactive to E. coli O157:H7 (01-95-90)
AC CE P
was purchased from KPL (Gaithersburg, MD). 19 different bacteria (see Table 1 for strains and types) were obtained from the Center for Food Safety Engineering consortium at Purdue University. Sodium citrate, sodium carbonate, chloroauric acid (HAuCl4), casein (sodium salt type), Tween 20 and tetramethyl benzidine (TMB) were obtained from Sigma (St. Louis, MO). Horseradish peroxidase, streptavidin, sulfo-NHS-LC-Biotin, LC-SMCC, sulfo-LC-SPDP, dithiothreitol (DTT) and mouse anti-goat antibody labeled with HRP were purchased from Pierce (Rockford, IL). Glass fiber membrane (Ahlstrom 8980) and cellulose membrane (17 CHR) were supplied by Whatman (Kent, UK). Glass membrane (PT-R5) and nitrocellulose membrane (70 CNPH-N-SS40) were purchased from Advanced Microdevices (Ambala Cantt, India). Other reagents used were of analytical grade.
5
ACCEPTED MANUSCRIPT 2.2. Preparation of conjugates Two conjugates were prepared. Conjugate I consists of the AuNP with biotin cross linkers. Here,
PT
1 mL of AuNP solution synthesized via citrate reduction (Makhsin et al., 2012) (40 nm in diameter) was sequentially mixed with 1 μL of 0.5 M of sodium carbonate, 100 μL of phosphate
RI
buffer (10 mM, pH 7.4) and 10 μg of E. coli O157:H7 antibody solution, which was stirred in a
SC
rotary unit for 2 hrs. 5% (w/v) casein dissolved in 10 mM phosphate buffer (122 μL) was added
NU
and then stirred for 1 hr to block residual surfaces of the AuNP. Unbound molecules were removed by centrifugation (12,000 rpm, 15 min) and the pellet was resuspended using 1 mL of
MA
PBS, which was repeated two additional times. This conjugate was biotinylated by adding 10 μg of sulfo-NHS-LC-biotin dissolved in PBS for 30 min followed by removal of unbound biotin-
TE
D
linkers by centrifugation at identical conditions as stated above.
AC CE P
Conjugate II is the streptavidin-HRP conjugate, where, streptavidin (300 μg) will be first labeled with LC-SMCC dissolved in DMSO at 50-fold molar excess at room temperature for 1 hr and the excess LC-SMCC reagents were removed by size exclusion chromatography (Sephadex G-15). HRP was also activated with sulfo-LC-SPDP dissolved in PBS at 25-fold molar excess, followed by DTT reduction (final 10 mM) and purification on Sephadex G-15 sequentially. The streptavidin was then mixed with the activated HRP which is 10 molar in excess and the reaction was carried out at room temperature for 4 hrs.
2.3. Analysis of conjugate components Enzyme immunoassay: 100 μL of E. coli O157:H7 (1 x 107 CFU/mL) resuspended in PBS was coated on the microwell at 37℃ for 1 hr. This incubation condition was identically applied to
6
ACCEPTED MANUSCRIPT perform the following reaction steps. The residual surface was blocked with 200 µL of 0.5% casein dissolved in PBS (Casein-PBS) for 1 hr and then 100 μL of conjugate I diluted in the
PT
Casein-PBS was reacted. For verification of antibody attachment and biotinylation of the gold
RI
surface, mouse anti-goat IgG coupled to HRP and streptavidin-HRP conjugate were reacted respectively. For signal generation 200 µL of substrate solution (50 mM sodium acetate: 1% (v/v)
SC
TMB: 3% (v/v) hydrogen peroxide = 1000: 10: 1) was added and maintained for 15 min. The
NU
reaction was terminated by adding 50 µL of 2 M sulfuric acid, the optical density was measured
MA
with an ELISA reader (VERSAmax; Molecular Device, Sunnyvale, CA) at 450 nm.
SDS-PAGE analysis: conjugate II, streptavidin and protein marker were treated with sample
D
buffer without mercaptoethanol (non-reducing condition) and loaded on to a 10% polyacrylamide
TE
gel. For electrophoretic separation, two voltages (80 V and 150 V) were sequentially applied for
AC CE P
20 min and 60 min respectively using a Mini-Protein Tetra Cell apparatus (Bio-Rad, Hercules, CA). The separating gel was sequentially immersed in a staining solution including 0.1% Coomassie blue R-250 for 15 min and the destaining solution for 1 hr for the development of protein bands.
2.4. Analytical procedure of LF-ICA assay E. coli O157:H7 antibody (0.5 μg) was used as a capture agent and coated on nitrocellulose membrane which was dried at 37℃. To initiate the immunoassay on the membrane strip, conjugate I and II were placed on the corresponding (absorption and conjugate pads respectively) followed by the introduction of the sample (in this demonstration, E. coli O157:H7 diluted in Casein-PBS (100 µL)) to the sample absorption pad (Fig 1). After 15 min of sample flow to allow 7
ACCEPTED MANUSCRIPT the immunoreaction to occur, the strip was washed with 40 µL of deionized water by connecting two absorption pads on the other side of the nitrocellulose membrane to induce cross-flow. For
PT
signal generation, a colorimetric substrate (i.e., TMB) was applied to the absorption pad and then
RI
washed with deionized water to stop the reaction. The color signals that appeared on the nitrocellulose membrane of the immunostrip was captured by scanning and the density values
NU
SC
were then obtained from the area where signal is produced.
2.5. Cross-reactivity test
MA
Specificity experiments were performed against 19 different bacteria (see Table 1 for the different strains and types of bacteria and Fig. 6), which were respectively cultured in 500 mL of
D
Tryptic Soy Broth (TSB) media at 37oC with shaking, for 18 hrs and then harvested by
TE
centrifugation (5000 rpm, 20 min) and stored in 50 mL of sterilized PBS. The concentration of
AC CE P
bacteria in these stock solutions was determined in triplicate by enumeration on plate count agar following incubation at 37°C for 24 hrs. The analytical procedure was the same as described above. The bacteria concentration was maintained at a high level (1 x 107 CFU/mL) in caseinPBS in the cross-reactivity experiments.
2.6. Food sample testing To test the feasibility of the method for pathogen detection in food, the assay was performed to test the presence of E. coli O157:H7 in intentionally inoculated ground beef samples. Uninoculated sample was used as control. The concentration of E. coli O157:H7 inoculated was determined by colony counting. The sampling procedure is as follows: a 10 g sample of ground beef was mixed with 10 mL of sterilized PBS and inoculated with E.coli O157:H7. After 1 hr of inoculation, the assay was performed in triplicate as described above. 8
ACCEPTED MANUSCRIPT
2.7. Biosafety concerns
PT
All experiments were conducted in BSL-2 approved facilities with proper safety training and
RI
approvals in place.
NU
SC
3. Results and Discussion
3.1. Analytical concept
MA
Our goal was to develop a high performance biosensor for onsite monitoring of select foodborne pathogens. The concept of LF-ICA was employed because the assay was simple and the signal
TE
D
can be amplified with enhancement tools. Since the signals obtained from the immunoreaction are directly proportional to the amount of enzyme (i.e., HRP) participating in the reaction, it is
AC CE P
crucial to tether as many HRP molecules to the AuNP surfaces used as a carrier of HRP to obtain high sensitivity (Fig. 1). Two conjugates are designed for this purpose: conjugate I (biotinylated gold nanoparticle) and a conjugate II (streptavidin-HRP). The conjugate II can be efficiently bound to biotin tethered to the AuNPs via the strong avidin-biotin interaction with minimal steric hindrance, resulting in a conjugate with multiple enzymes owing to the large surface area of AuNPs.
9
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
MA
Fig. 1. Schematic of gold nanoparticle (AuNP)-assisted signal amplification on LF-ICA for pathogen detection. The two conjugates used for signal enhancement consists of antibody, AuNP and biotin-terminated cross-linkers (conjugate I) and streptavidin and horseradish peroxidase
D
(conjugate II). Enzyme-catalyzed colorimetric signal will be produced by supplying a substrate
TE
(e.g., tetramethyl benzidine with hydrogen peroxide) in the cross-flow direction at the site of
AC CE P
antigen-antibody sandwich complex formation.
The analytical protocol is a two-step procedure consisting of (i) initiation of sample flow to induce immunocomplex formation (antibody-pathogen-conjugates binding) and (ii) supplying enzyme substrate for colorimetric signal generation in the cross-flow direction. First, the sample containing the target pathogen will be applied at the bottom of an immunostrip (sample absorption pad) to initiate a flow in the vertical (longitudinal) direction, inducing immunocomplex formation at a pre-determined site where the capture antibody is immobilized. During this flow, a number of SA-HRP conjugates at the conjugate pad can be coupled to the branched biotin on to conjugate I, to initiate an enzyme-network formation. Second, two horizontally arranged pads are placed on each lateral side of the immunostrip and the substrate containing tetramethyl benzidine and hydrogen peroxide is then added onto the substrate supply pad to initiate the enzymatic signal generation (horizontal flow). At the sites of complex 10
ACCEPTED MANUSCRIPT formation, colorimetric signals can be produced by supplying enzyme substrates at the time of
PT
signal generation.
RI
3.2. Characterization of the assay components
Since two key conjugates are synthesized, these need to be well characterized to ensure their
SC
physical and immunochemical properties. First, conjugate I was analyzed by three different
NU
analytical methods as follows: (a) UV/Vis spectrometry, (b) zeta potential measurement and (c) solid phase enzyme immunoassay (Fig. 2). From the spectra we note that the wavelength peak
MA
position of conjugate I was red-shifted by ~5 nm (527 nm to 532 nm) compared with that of bare AuNP. This is because of the two types of protein, i.e., antibody specific to E. coli O157:H7 and
D
casein used as a blocking agent were well adsorbed onto the gold surfaces. This is consistent with
TE
the previously reported results (Wang et al., 2011). This is also supported by the fact that no
was tested).
AC CE P
aggregated forms were observed under high concentration of sodium chloride (up to 3 M NaCl
Fig. 2. Characterization of conjugate I used for signal enhancement using UV/Vis spectrometer (a), zeta potential measurement (b) and enzyme immunoassay (c).
11
ACCEPTED MANUSCRIPT The difference between bare AuNP and conjugate I can also be assessed from zeta potential values shown in Fig. 2b. Negatively charged bare AuNP was slightly shifted to the positive
PT
direction at conjugate I, which can be attributed to the masking of the surface via protein
RI
adsorption. Attachment of antibody and biotin to the surface was verified by enzyme immunoassay in which the analyte (i.e., E. coli O157:H7) was immobilized on a microwell
SC
surface and the two molecules were identified by HRP-labeled secondary antibody and
NU
streptavidin respectively. Obtaining absorbance signals from antibody and biotin shown in Fig. 2c indicates that antibodies and biotin-terminated cross-linkers were well coupled to the
AC CE P
TE
D
MA
nanoparticle surface.
Fig. 3. Characterization of conjugate II used for signal enhancement by (a) SDS-PAGE and (b) enzyme assay. The number of HRP molecules on the conjugate and minimal amount of HRP molecules can be determined by PAGE and enzyme analysis respectively. Combining the information with HABA analysis (c), the number of HRP molecules which can be bound to a single AuNP is estimated to be ~109 (see text for details).
12
ACCEPTED MANUSCRIPT Second the conjugate II was characterized by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) to measure the degree of conjugation between SA and HRP. From
PT
the result shown in Fig. 3a, the mole-to-mole ratio between SA and HRP at conjugate II can be
RI
estimated to be about 1 to 3. This was caused by the fact that the molecular weight of SA and HRP is 60 kDa and 44 kDa respectively. Thus the SA-HRP conjugate would be positioned
SC
around 190 kDa as shown in Fig. 3a. Another important parameter that needs to be assessed is
NU
the number of HRP molecules conjugated to one AuNP. The number of HRP molecules was determined by enzyme assay and HABA (4'-hydroxyazobenzene-2-carboxylic acid) analysis
MA
(Janolino et al., 1996). From the enzyme assay a minimal amount of HRP required for signal generation was calculated (see Fig. 3b) as ~0.5 attomole. Assuming all AuNPs are on an average
D
spherical in shape, with similar size characteristics and known concentration of AuNPs (e.g., 0.14
TE
pM at 40 nm-sized), the number of HRP molecules on the surface of a single AuNP can be
AC CE P
estimated from HABA analysis. Per our preliminary experiment, approximately 109 HRP molecules would be tethered to a single 40 nm size gold nanoparticle bearing biotin-containing chemical linkers (Fig. 3c).
3.3. Analytical performances
With this enzyme signal amplification strategy feasibility experiments were performed against different concentration of E. coli O157:H7. Gold nanoparticle that is considered as a conventional tracer in LF-ICA was used as a reference in comparison with the enzyme assay. From the results shown in Fig. 4a, it can be noted that the detection capability of the AuNP-based method is not sufficient to detect low number of bacteria. The limit of detection (LOD) using these constructs was 1 x 105 CFU/mL.
13
SC
RI
PT
ACCEPTED MANUSCRIPT
NU
Fig. 4. Comparison of conventional method based on AuNP with the developed method employing AuNP-assisted enzyme-based signal amplification on a LF-ICA platform against the
MA
target pathogen, E. coli O157:H7.
The low analytical sensitivity can be overcome by applying enzyme an amplification strategy
TE
D
proposed in this study. Compared with the result based on AuNP, extremely low numbers of E. coli O157:H7 could be detected (~100 CFU/mL) by applying AuNP-assisted enzyme-based
AC CE P
signal amplification (see Fig. 4b), providing an approximate 1000-fold signal enhancement over the existing LF methods. This indicates that numerous enzyme molecules bound to the AuNP could form an enzyme network via biotin-SA linkage and participate in this signal amplification event as expected without losing their enzymatic activity. Thus very low numbers of E. coli O157:H7 are detectable yielding a LOD of ~100 CFU/mL using this method. The LOD defined by multiplying the standard deviation at zero dose (0 CFU/mL) by three (Galikowska et al., 2011; Kim et al., 1999) were determined from dose response curve (Fig. 5b) as 100 CFU/mL.
14
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
Fig. 5. Analytical performance of AuNP-assisted enzyme signal amplification with LF-ICA to
MA
detect E. coli O157:H7. (a) signals from captured images produced on the immunostrip after the application of a chromogenic substrate solution. (b) dose response curve against E. coli O157:H7
TE
D
concentration.
Along with analytical sensitivity, specificity is also considered as one of the major factors that
AC CE P
affect the analytical performance of the biosensor. Since many antigenic sites of the bacteria surface are similar to others (e.g., lipopolysaccharides, K-, O-, and H- antigens), it is very crucial to differentiate other bacteria from the target pathogen to minimize false positives. Hence specificity of this method was examined against 19 different bacteria strains including E. coli O157:H7 (see Table 1 and Fig. 6) by performing enzyme immunoassay (Fig. 6a) and LF-ICA (Fig. 6b). Specificity experiments were performed at a higher concentration of bacteria (1 x 107 CFU/mL) to ensure a reliable performance of this method. For enzyme immunoassay, bacteria samples (S. typhimurium, S. enteritidis and L. monocytogenes including E. coli O157:H7) were coated onto a microtiter well plate surface and then evaluated by conjugate I and II sequentially. For LF-ICA experiments, each bacteria was detected by applying a sandwich immunoassay (i.e., capture antibody-bacteria-detection antibody) in which AuNP was employed as a tracer.
15
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
Table 1. List of bacteria strains for specificity test
16
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
Fig. 6. Specificity tests against various bacteria strains. All bacteria were cultured in TSB media and tested at a concentration of ~107 CFU/mL. (a) enzyme immunoassay and (b) LF-ICA.
From these experiments it was obvious that O157:H7 antibodies used in the proposed strategy for AuNP-assisted enzyme amplification had no cross-reactivity with the other competitive strains even at a high concentration of such bacteria in the sample and no signals were obtained, except for sample #5 and #11. The signal obtained from sample #11 (E. coli ATCC 35150) was due to the fact that the strain has the same antigenic properties as the serotype E. coli O157:H7. Specificity experiments against a wide range of bacteria strains provides a strong validation of the assay highlighting its specificity to the target pathogen (i.e., E. coli O157:H7) with high fidelity.
17
ACCEPTED MANUSCRIPT This AuNP-assisted signal amplified LF-ICA method was next used to test pathogens in food samples as a final validation step. Ground beef was adopted as a sample matrix and an initial
PT
concentration of E. coli O157:H7 was purposefully inoculated and enumerated by colony
RI
counting in triplicate. Three different inoculated samples (0, 437 and 984 CFU/mL) were maintained for 1 hr followed by testing of the supernatant for the presence of pathogens. As
SC
shown in Fig. 7, the signals obtained were consistent with those of the dose response in Fig. 5b,
NU
indicating that the developed method is robust and the antibodies are still reactive in a different environment. Although the samples were maintained 1 hr prior to analysis, our expectation is that
MA
most of the bacteria present in the lag phase did not affect bacteria proliferation. Therefore it can be concluded that a pre-enrichment step is not a requirement in this assay system unlike most of
AC CE P
TE
D
the pathogen immunoassays that need a prolonged culturing step.
Fig. 7. Performance of the LF-ICA platform in a food matrix (e.g., ground beef) innoculated with E. coli O157:H7. The initial number of E. coli was assessed by colony counting in LB agar plate. The assay was performed after 1 hr of incubation.
18
ACCEPTED MANUSCRIPT Results of our studies with the developed biosensor show high sensitivity and specificity, including detection of the pathogen in a relevant food matrix. We expect our approach can be
RI
PT
extended to other organisms and for multiplex detection of pathogens.
NU
SC
4. Conclusions
MA
In conclusion, we report a highly sensitive AuNP-assisted enzyme signal amplification method based on a LF-ICA format to detect very low number of pathogens. The proposed format is rapid
D
(the whole procedure can be completed within 15 min), simple (only two steps are required;
TE
vertical and cross flow for immunoreaction and signal generation respectively), low cost (no need to use fluorometer or luminometer because signal is amplified for visual inspection) and practical.
AC CE P
Very high specificity was demonstrated by testing against numerous bacteria strains and food pathogens as well as in a food sample. We expect our signal enhancing approach can be efficiently utilized for timely detection of foodborne pathogens to prevent an outbreak in the field.
Acknowledgements Research was supported by funds from USDA-ARS project number 1935-42000-049-00D in conjunction with the Center for Food Safety Engineering at Purdue University.
19
ACCEPTED MANUSCRIPT References
AC CE P
TE
D
MA
NU
SC
RI
PT
Banada, P.P., Huff, K., Bae, E., Rajwa, B., Aroonnual, A., Bayraktar, B., Adil, A., Robinson, J.P., Hirleman, E.D., Bhunia, A.K., 2009. Label-free detection of multiple bacterial pathogens using light-scattering sensor. Biosensors and Bioelectronics 24, 1685-1692. Bennett, A., Greenwood, D., Tennant, C., Banks, J., Betts, R., 1998. Rapid and definitive detection of Salmonella in foods by PCR. Letters in applied microbiology 26, 437-441. Bhunia, A.K., 2008. Biosensors and bio‐based methods for the separation and detection of foodborne pathogens. Advances in food and nutrition research 54, 1-44. Brovko, L.Y., Anany, H., Griffiths, M.W., 2012. 6 Bacteriophages for Detection and Control of Bacterial Pathogens in Food and Food-Processing Environment. Advances in food and nutrition research 67, 241. Cho, I.-H., Bhandari, P., Patel, P., Irudayaraj, J., 2015. Membrane filter-assisted surface enhanced Raman spectroscopy for the rapid detection of E. coli O157: H7 in ground beef. Biosensors and Bioelectronics 64, 171-176. Cho, I.-H., Irudayaraj, J., 2013. Lateral-flow enzyme immunoconcentration for rapid detection of Listeria monocytogenes. Analytical and bioanalytical chemistry 405, 3313-3319. Cho, I.-H., Mauer, L., Irudayaraj, J., 2014. In-situ fluorescent immunomagnetic multiplex detection of foodborne pathogens in very low numbers. Biosensors and Bioelectronics 57, 143148. Cho, J.-H., Han, S.-M., Paek, E.-H., Cho, I.-H., Paek, S.-H., 2006. Plastic ELISA-on-a-chip based on sequential cross-flow chromatography. Analytical Chemistry 78, 793-800. Cho, J.-H., Paek, E.-H., Cho, I.-H., Paek, S.-H., 2005. An enzyme immunoanalytical system based on sequential cross-flow chromatography. Analytical Chemistry 77, 4091-4097. DeWaal, C.S., Roberts, C., Catella, C., 2012. Outbreak Alert! 1999-2008. Center for Science in the Public Interest. http://www. cspinet. org/foodsafety/PDFs/Outbreak_Alert_1999-2008. pdf. Galikowska, E., Kunikowska, D., Tokarska-Pietrzak, E., Dziadziuszko, H., Łoś, J.M., Golec, P., Węgrzyn, G., Łoś, M., 2011. Specific detection of Salmonella enterica and Escherichia coli strains by using ELISA with bacteriophages as recognition agents. European journal of clinical microbiology & infectious diseases 30, 1067-1073. Gallegos‐Robles, M., Morales‐Loredo, A., Álvarez‐Ojeda, G., Osuna‐García, J., Martinez, I., Morales‐Ramos, L., Fratamico, P., 2009. PCR detection and microbiological isolation of Salmonella spp. from fresh beef and cantaloupes. Journal of food science 74, M37-M40. Goodridge, L., Chen, J., Griffiths, M., 1999. Development and characterization of a fluorescentbacteriophage assay for detection of Escherichia coli O157: H7. Applied and environmental microbiology 65, 1397-1404. Janolino, V.G., Fontecha, J., Swaisgood, H.E., 1996. A spectrophotometric assay for biotinbinding sites of immobilized avidin. Applied Biochemistry and Biotechnology 56, 1-7. Kim, J.W., Jin, L.Z., Cho, S.H., Marquardt, R.R., Frohlich, A.A., Baidoo, S.K., 1999. Use of chicken egg‐yolk antibodies against K88+ fimbrial antigen for quantitative analysis of enterotoxigenic Escherichia coli (ETEC) K88+ by a sandwich ELISA. Journal of the Science of Food and Agriculture 79, 1513-1518. Makhsin, S.R., Razak, K.A., Noordin, R., Zakaria, N.D., Chun, T.S., 2012. The effects of size and synthesis methods of gold nanoparticle-conjugated MαHIgG4 for use in an immunochromatographic strip test to detect brugian filariasis. Nanotechnology 23, 495719. 20
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
Malorny, B., Tassios, P.T., Rådström, P., Cook, N., Wagner, M., Hoorfar, J., 2003. Standardization of diagnostic PCR for the detection of foodborne pathogens. International journal of food microbiology 83, 39-48. Pandey, S.K., Vinayaka, A.C., Rishi, D.B., Rishi, P., Suri, C., 2014. Immuno-fluorescence based Vi capsular polysaccharide detection for specific recognition of Salmonella enterica serovar Typhi in clinical samples. Analytica Chimica Acta. Pengsuk, C., Chaivisuthangkura, P., Longyant, S., Sithigorngul, P., 2013. Development and evaluation of a highly sensitive immunochromatographic strip test using gold nanoparticle for direct detection of Vibrio cholerae O139 in seafoodsamples. Biosensors and Bioelectronics 42, 229-235. Rajwa, B., Dundar, M.M., Akova, F., Bettasso, A., Patsekin, V., Dan Hirleman, E., Bhunia, A.K., Robinson, J.P., 2010. Discovering the unknown: Detection of emerging pathogens using a label‐free light‐scattering system. Cytometry Part A 77, 1103-1112. Ravindranath, S.P., Mauer, L.J., Deb-Roy, C., Irudayaraj, J., 2009. Biofunctionalized magnetic nanoparticle integrated mid-infrared pathogen sensor for food matrixes. Analytical Chemistry 81, 2840-2846. Reissbrodt, R., 2004. New chromogenic plating media for detection and enumeration of pathogenic Listeria spp.—an overview. International journal of food microbiology 95, 1-9. Smartt, A.E., Xu, T., Jegier, P., Carswell, J.J., Blount, S.A., Sayler, G.S., Ripp, S., 2012. Pathogen detection using engineered bacteriophages. Analytical and bioanalytical chemistry 402, 31273146. Subramanian, A., Irudayaraj, J., Ryan, T., 2006. A mixed self-assembled monolayer-based surface plasmon immunosensor for detection of E. coli O157: H7. Biosensors and Bioelectronics 21, 998-1006. Swaminathan, B., Feng, P., 1994. Rapid detection of food-borne pathogenic bacteria. Annual Reviews in Microbiology 48, 401-426. Velusamy, V., Arshak, K., Korostynska, O., Oliwa, K., Adley, C., 2010. An overview of foodborne pathogen detection: in the perspective of biosensors. Biotechnology advances 28, 232254. Wang, C., Irudayaraj, J., 2010. Multifunctional magnetic–optical nanoparticle probes for simultaneous detection, separation, and thermal ablation of multiple pathogens. Small 6, 283-289. Wang, Y., Lee, K., Irudayaraj, J., 2010. Silver nanosphere SERS probes for sensitive identification of pathogens. The Journal of Physical Chemistry C 114, 16122-16128. Wang, Y., Ravindranath, S., Irudayaraj, J., 2011. Separation and detection of multiple pathogens in a food matrix by magnetic SERS nanoprobes. Analytical and bioanalytical chemistry 399, 1271-1278.
21
ACCEPTED MANUSCRIPT
Highlights
AC CE P
TE
D
MA
NU
SC
RI
PT
1) Detection of O157:H7 in < 20 min in food matrices 2) Highly sensitive with a limit of detection of ~100 cfu/ml for routine onsite measurements (a 1000-fold increment compared to the conventional lateral flow systems). 3) Excellent potential to be generalized and expanded to other pathogens for detection of pathogens in food, water, and human samples. 4) Highly specific as demonstrated by testing a robust set of strains for detection specificity
22
