Salmonella typhimurium - ACS Publications - American

Subscriber access provided by READING UNIV

Article

Highly Sensitive and Rapid Detection of Salmonella typhimurium Using a Colorimetric Paper Based Analytical Device Coupled with Immunomagnetic Separation Monpichar Srisa-Art, Katherine E. Boehle, Brian J. Geiss, and Charles S. Henry Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b04628 • Publication Date (Web): 06 Dec 2017 Downloaded from http://pubs.acs.org on December 6, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Highly Sensitive and Rapid Detection of Salmonella typhimurium Using a Colorimetric Paper Based Analytical Device Coupled with Immunomagnetic Separation Monpichar Srisa-Arta, d, Katherine E. Boehlea, Brian J. Geissb,c, Charles S. Henrya,c* Department of Chemistry, bDepartment of Microbiology, Immunology, and Pathology, cSchool of Biomedical Engineering, Colorado State University, Fort Collins, Colorado 80523, United States

a

Department of Chemistry, Faculty of Science, Chulalongkorn University, Bangkok, 10330, THAILAND

d

ABSTRACT: Salmonella causes over a million foodborne illnesses per year in the United States resulting in more hospitalizations and deaths than any other foodborne bacteria pathogen. To help prevent outbreaks a rapid, portable, sensitive, and reliable method for onsite detection of bacteria that can be used in different sample matrices would be beneficial. Herein, we present a colorimetric paper-based analytical device (PAD) combined with immunomagnetic separation (IMS) for detecting Salmonella typhimurium. IMS anti-Salmonella coated magnetic beads was applied to capture and separate bacteria from the sample matrix and preconcentrated it into small volumes before testing on paper. To directly detect S. typhimurium after IMS, a sandwich immunoassay was implemented into the procedure with β-galactosidase (β-gal) as the detection enzyme. Using the antibody/enzyme complex, we performed a colorimetric assay with chlorophenol red-β-D-galactopyranoside (CPRG) for bacteria quantification. The method was confirmed to be highly specific to S. typhimurium without interference from other pathogenic bacteria like Escherichia coli. Using this system, the limit of detection of S. typhimurium was found to be 102 CFU mL-1 in culturing solution without any pre-enrichment. In addition, distance-based detection where the concentration is read as the length of colored band formed on reaction was also demonstrated. This assay had a detection limit of 102 CFU mL-1 for S. typhimurium, providing instrument-free quantitative analysis alternative to spot tests, which require image analysis. Finally, the proposed platform was applied for detection of S. typhimurium in inoculated Starling bird fecal samples and whole milk with detection limits of 105 CFU g-1 and 103 CFU mL-1 respectively, describing the first published paper-based detection method for S. typhimurium in bird feces and whole milk.

Salmonella is widely known as one of the most prevalent pathogens causing foodborne illness outbreaks.1 Per the United States Center for Disease Control and Prevention (CDC), Salmonella causes an estimated one million illnesses in the United States resulting in 19,000 hospitalizations and 380 deaths, more than any other pathogenic bacteria.2 Infected individuals encounter symptoms, such as fever, abdominal cramps, diarrhea, and even death within 12 to 72 h of exposure.3 Salmonella contaminates food products, like eggs, fruits, vegetables, meat, poultry, and milk4 typically through animal fecal contamination.5 This is because Salmonella lives and replicates in the intestinal tracts of humans and animals, and therefore is present in their feces.6 Studies have shown a strong correlation between skin and meat contamination of Salmonella, and prevalence in the animal’s feces, making it an important sample matrix to detect the pathogen’s presence.7 Traditional methods to detect Salmonella are based on bacterial culture, which is time-consuming and laborious, often taking at least a few days to complete detection.8,9 In addition, this conventional method

is not practical for on-site detection of bacteria because it requires a central laboratory. The need for expensive equipment and trained lab personnel increases testing costs, making large scale studies of Salmonella epidemiology difficult.10 Using polymerase chain reaction (PCR) as a method for detecting food borne pathogens has attracted considerable attention because of its high selectivity and accuracy. The overall analysis time of PCR methods was reported to take approximately 24 h which is much improved compared to the traditional culture method.11 However, PCR methods encounter a problem associated with sample matrix which could contain PCR inhibitors normally found in biological samples (i.e. urine, feces and blood), food, and humic substances.12 In addition, results from PCR could be unreliable when there is a low number of pathogens within a relatively large amount of sample, thus concentrating samples is unavoidably required.13 Accordingly, a rapid, userfriendly, cost-effective, reliable, and specific approach for Salmonella detection is required to overcome the drawbacks of conventional methods.

1 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Over the past decade, paper-based analytical devices (PADs) have become attractive analytical tools. The advantages of PADs include small sample and reagent consumption, rapid analysis, simple operation, disposability, and portability.14,15 PADs hold great promise for use as analytical tools in remote areas or areas where minimal instrumentation is available due to their natural fluid wicking properties and ability to store reagents. These properties make PADs attractive and simple platforms for analysis in fields such as environmental monitoring, medical diagnostics, point-of-care testing, and food safety control.14,16 However, there have been only a few reports on using PADs for rapid detection of bacteria, including Pseudomonas aeruginosa,17 Staphylococcus aureus,17,18 Escherichia coli,19-21 Salmonella typhimurium,20,22 Enterococcus spp.21 and Listeria monocytogenes.20 Our previous work demonstrated the application of PADs for detecting Salmonella spp., L. monocytogenes, E. coli, and antimicrobial resistant bacteria using both colorimetric detection 20,21,23 and electrochemical measurements.21 However, an enrichment step was required to detect lower bacteria concentrations and the detection motif relied solely on enzyme production by the bacteria. Immunomagnetic separation (IMS) is a powerful analytical method because of its ability to separate targets of interest from complex sample matrices. It can also be used as an alternative to culture enumeration as a preenrichment step for the target of interest. IMS is a procedure where antibodies specific to an analyte or cell are covalently attached to a magnetic particle.Antibodymodified magnetic particles are added to the sample matrix to adhere to the target and are separated from the matrix with a magnet and re-suspended in buffer. After separation from the sample matrix, many detection methods have been used including microscopy, broth enrichment, immunoassays, and PCR.24 Because IMS does not require bulky and expensive equipment to complete the procedure, this makes it ideal for in-field measurements. IMS is also a faster technique with total enrichment time of less than an hour compared to 4-16 h for culture enrichment depending on initial bacteria concentration. Unlike culture enrichment, IMS has been demonstrated for efficiently separating target analytes and cells from complex mixtures such as blood,25 milk,26 meat,27 cheese and yogurt,28 and even bovine feces.29 With IMS, the antibodies attached to the beads can be specific to any analyte or cell of interest. Because of this, IMS has been demonstrated for detecting many biomarkers,25,30 along with various bacteria27-29 and viruses.31-33 Combining IMS with paper-based devices has been previously described for the detection of E. coli in contaminated water.19 In this work, the authors describe

Page 2 of 13

the use of IMS to pre-concentrate samples from contaminated water before lysing the bacteria and detecting bacterial enzymes β-galactosidase and β-glucuronidase. To the best of our knowledge, there has not been a paper-based device that is coupled with IMS for the detection of bacteria in more complicated sample matrices, such as animal feces and whole milk. Furthermore, despite the prevalence of Salmonella in bird feces, to the best of our knowledge, there has not been a proposed alternative detection method to traditional culture enrichment. Herein, we describe a technique that couples PADs with IMS for the colorimetric detection of S. typhimurium without performing cell culture for bacteria enrichment. IMS was applied to capture and separate target bacteria from the sample matrix and then preconcentrated into small volumes for further assays. A sandwich immunoassay was carried out to detect the presence of S. typhimurium in the sample. A second anti-Salmonella antibody was conjugated with biotin, which was bound to streptavidin linked to β−galactosidase (β-gal) to perform a colorimetric assay with chlorophenol red-β-Dgalactopyranoside (CPRG). This platform could detect 102 CFU mL-1 S. typhimurium in culture solution within 90 min and without any culture enrichment. Detection was demonstrated using both spot tests with image analysis and instrument-free distance-based detection. The PAD coupled with IMS was also demonstrated in detecting S. typhimurium in inoculated bird feces samples with a detection limit of 105 CFU g-1, making it a promising alternative detection method for bird feces compared to traditional culturing. To show the PAD’s promise for onsite detection of contaminated food products, this method was also demonstrated in the detection of S. typhimurium in inoculated whole milk with a detection limit of 103 CFU mL-1.

MATERIALS AND METHODS Materials and Reagents. Dynabeads M-280 Tosylactivated (Product no. 14203, Invitrogen) were purchased from Thermo Fisher Scientific Inc. (Waltham, MA, USA). Anti-Salmonella typhimurium 0-4 antibody (mouse monoclonal [1E6] against lipopolysaccharides, Product no. ab 8274) was obtained from Abcam (Cambridge, MA, USA). Salmonella antibody, biotin conjugate (4-5 mg mL1, Product no. PA1-73022, Invitrogen) and Streptavidin, β -galactosidase conjugate (Product no. S931, Life Technologies) were purchased from Thermo Fisher Scientific Inc. (Waltham, MA, USA). The streptavidin, β galactosidase conjugate was reconstituted in Milli Q water to a concentration of 2 mg mL-1 upon arrival. Chloro-


Page 3 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


phenol Red- β -D-galactopyranoside (CPRG, Product no 59767) was obtained from Sigma-Aldrich (St. Louis, MO, USA). Buffer Preparation. All buffers were prepared in MilliQ water and autoclaved at 121 °C for 15 min before use. All buffer recipes for immobilization were prepared in MilliQ water according to the protocol recommended by Invitrogen; 0.1 M sodium phosphate buffer pH 7.4 (sodium dihydrogen phosphate (NaH2PO4) and disodium hydrogen phosphate (Na2HPO4)); pre-washing buffer (0.1 M borate buffer pH 9.5) prepared from boric acid and adjusted to pH 9.5 using sodium hydroxide; coupling buffer pH 7.4 (0.1 M sodium phosphate buffer with 3 M ammonium sulphate); washing buffer (0.01 M sodium phosphate buffer pH 7.4 with 0.5% (w/v) bovine serum albumin (BSA)) and storage buffer (0.01 M sodium phosphate buffer pH 7.4 with 0.1% (w/v) BSA). Buffers used for IMS consisted of phosphate buffered saline (PBS) pH 7.4 obtained from Gibco by Life Technologies (Thermo Fisher Scientific Inc., Waltham, MA, USA) and washing buffer (PBS Tween) prepared from 0.01 M sodium phosphate buffer pH 7.4, with 0.05% (w/v) Tween 20 and 0.15 M NaCl. Paper-Based Device Fabrication. Two formats of paper devices were employed in this work; well-array 20 and distance-based (referred to as “chemometer”)34 patterns. Both configurations were fabricated on Whatman No. 4 filter paper (GE Healthcare Life Sciences, NY, USA) using wax printing as described below. The design of well-array paper-based devices followed the layout from the previous work.20 CorelDRAW software was used to design device features containing arrays of 7-mm diameter circular wells with 4-pt line thickness. After printing the design using a wax printer (Xerox Colorqube 8870), devices were heated on a hot plate at 175 °C for 50 s to melt the wax through the paper, creating a hydrophobic barrier. Finally, the backs of the paper devices were taped with Scotch packing tape to control fluid flow and prevent leaking during the assay. Application of reagents and samples were performed on the front (wax-printed) side. The design and fabrication of distance-based paper devices (chemometers) followed the process described previously.34 Device features containing a circular reservoir (6 mm diameter) and a straight channel (4 mm wide and 5.5 cm long) with 4-pt line thickness were generated using CorelDRAW software. A ruler was added parallel to the channel for easy reading of the distance of color development along the channel. The ruler design was first ink printed on the Whatman No. 4 filter paper to generate rulers for the chemometers, then wax printing

was performed as above to create the remaining chemometer features. The ink-printed rulers are not affected by heating the devices on a hotplate. To prevent leaking and evaporation of reagents, the paper chemometers were thermally laminated at 110 °C by passing the chemometers enveloped in Scotchthermal laminating pouches through an Apache AL13P thermal laminator. A 4 mm (internal diameter) hole was punched through the reservoir using a 4 mm diameter disposable biopsy punch (Robbins Instruments, Inc., Chatham, NJ, USA). The back of the laminated chemometer was then taped with the Scotch shipping packing tape to allow reagent addition to the reservoir. Bacteria Strains and Culture Conditions. The bacterial strains used in this work were Salmonella enterica serovar Typhimurium (ATCC 14028, product no. 0363P) purchased from Microbiologics (St. Cloud, MN, USA) and DH5a Escherichia coli were purchased from Thermo Fisher Scientific. Both strains were grown in Difco Nutrient Broth (Product no. 234000, BD, Sparks, MD, USA) at 37 °C with gently shaking at 125 rpm for 24 h. Serial dilutions of bacteria suspension were prepared in the range from 102 to 108 CFU mL-1 in nutrient broth. The exact bacteria concentration was quantified by plating 100 µL of 102 and 103 CFU mL-1 dilutions onto Difco Nutrient agar plates (Product no. 213000, BD, Sparks, MD, USA). The plates were incubated at 37 °C for 24 h before counting the colonies. All experiments with bacteria in this work were performed in a biosafety level 2 (BSL-2) laboratory. Preparation of anti-Salmonella Dynabeads®. The anti-Salmonella typhimurium 0-4 antibody was covalently conjugated onto M-280 tosylactivated Dynabeads using the standard protocol from Invitrogen.35 Breifly, before antibody immobilization the beads were washed using pre-washing buffer. Beads were vortexed for approximately 30 s or until a homogeneous suspension was obtained. A volume of 165 µL (5 mg) of the beads was pipetted into a 1.5 mL Eppendorf tube followed by adding 165 µL of pre-washing buffer. Beads were thoroughly mixed, then separated from solution using a magnet (DynaMag-2 Magnet, product no. 12321D, Thermo Fisher Scientific Inc., Waltham, MA, USA) for 1 min and the supernatant was discarded. The washing process was repeated once with the pre-washing buffer before immobilization. To covalently immobilize antibody onto the beads, 100 µg of antibody (for 5 mg beads) was recommended by the manufacturer of the beads. 50 µL of anti-Salmonella typhimurium 0-4 antibody (2 mg mL-1, 250 µg) was added to re-suspend the washed beads. Pre-washing buffer was also added to give a total volume of 150 µL, and 100 µL



of coupling buffer was added to the beads mixture and thoroughly mixed on a rotator (RotoFlex, product no. R2000, Argos Technologies, Elgin, IL, USA) at 37 °C for 12-18 h. The bead-antibody mixture was placed on the magnet for 1 min and the supernatant was discarded. 1 mL of washing buffer was added into the beads and the tube was incubated on the rotator at 37 °C for 1 h, then the wash buffer was removed using the magnet as above. The beads were then washed using 1 mL of storage buffer, vortexed for 5-10 s, and the tube placed on the magnet to remove the supernatant. The washing step was performed twice. Finally, the beads immobilized with anti-Salmonella antibody were re-suspended in storage buffer to a final bead concentration of 20 mg mL1. The anti-Salmonella Dynabeads® were stored at 4 °C for further use. Immunomagnetic Separation and Incubation. Immunomagnetic separation (IMS) was employed to isolate S. typhimurium from culture media, bird feces, or whole milk. The IMS process was performed using antiSalmonella Dynabeads® that had been previously prepared. Subsequently, the bead-bacteria complex was conjugated to β-gal through a biotin-streptavidin linkage. Both processes are schematically illustrated in Fig. 2. For IMS of bacteria, the anti-Salmonella magnetic beads were vortexed until a homogeneous suspension was obtained. Subsequently, 5 µL of the magnetic beads were pipetted into a 1.5 mL Eppendorf tube and 1 mL of bacteria suspension was added into the tube and mixed thoroughly by pipetting. The mixture was incubated on a rotator at room temperature for 15 min. Next, the tube was placed in the magnet for 1 min before carefully removing the supernatant without disturbing the pellet of IMS beads attached on the side wall of the tube. Anti-Salmonella biotin conjugated antibody (Ab-biotin) was diluted to 0.02 mg mL-1 in phosphate buffered saline (PBS) pH 7.4. 100 µL of the diluted Ab-biotin (~2.0 µg) was added to the bead-bacteria complex from the IMS process and incubated on the rotator at room temperature for 20 min. The tube was then placed onto the magnet to remove the supernatant and washed twice using washing buffer (PBS Tween). 100 µL of streptavidin/ βgalactosidase conjugate (strep-β-gal), diluted at 1:1,000 v/v in PBS, was then pipetted into the bead-bacteria complex. This corresponded to a concentration of approximately 0.2 µg for strep-β-gal. The mixture was incubated on the rotator at room temperature for 10 min, placed onto the magnet to remove the supernatant, washed twice to remove unbound strep-β-gal, and resuspended in 100 µL PBS. The β-gal-labeled bacteria

Page 4 of 13

were detected with CPRG on paper devices as described below. Incubation times were optimized for (i) the antiSalmonella magnetic beads and bacteria suspension, (ii) the bead-bacteria complex and Ab-biotin and (iii) the bead-bacteria complex (conjugated with biotin) and strep-β-gal to achieve high sensitivity for the assay. The incubation time for the anti-Salmonella magnetic beads and bacteria suspension was tested at 5, 10, 15, 20 and 30 min, whereas the same incubation times of 1, 5, 10, 20 and 30 min were studied for both Ab-biotin and strep-βgal. Determining Capture and Retention Efficiency. The capture efficiency of the anti-Salmonella magnetic beads for Salmonella was considered using initial bacteria concentrations of 102, 103, 104 and 105 CFU mL-1. After the IMS process the bead-bacteria complex pellet was resuspended in 100 µL PBS. High-concentration samples were diluted to have final concentrations of approximately 101 and 102 CFU mL-1 before using 10 µL of the final suspension to spread onto DifcoTM nutrient broth agar plates. Colonies on the plates were counted after incubation for 24 h at 37 °C. The number of colonies before (nbefore IMS) and after (nafter IMS) IMS were compared to calculate the capture efficiency using the equation below; Capture efficiency (%) =

nafter IMS × 100 nbefore IMS

The number of bacteria before the IMS (nbefore IMS) was obtained from plating 100 µL of 101, 102, and 103 CFU mL-1 bacteria suspension on the agar plates before starting the IMS process. The number of colonies on each plate was averaged to determine an estimate of the bacteria concentration. The same bacteria concentrations used to investigate the capture efficiency were used to evaluate retention efficiency. After bacteria capture using the IMS procedure, the coupling process was performed as previously described. The suspension of the bead-bacteria-β-gal complex was diluted in PBS to final concentrations of 101 and 102 CFU mL-1 before using 10 µL for plating on the nutrient broth agar plates. After 24 h incubation at 37 °C, colonies on the plates were counted as the number of bacteria after the coupling process (nafter coupling) which was compared with before the coupling step (nbefore coupling which was equal to the number of bacteria after the IMS, nafter IMS). The retention efficiency was determined using the equation below;


Page 5 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Retention efficiency (%) =

nafter coupling × 100 nbefore coupling

Enzymatic Assay. The detection procedure was carried out on both 7 mm well-array devices and chemometers using 10 µL of 2.5 mM CPRG in PBS and 10 µL of the β-gal-labeled bacteria. Images of red color development were taken using iPhone 5S after 30 min for quantifying the red color intensity. The color intensity was directly proportional to the number of bacteria. For ImageJ analysis, an image was split into red, green, and blue image components and only the green channel was selected to create an inversed image before measuring sample spots for intensity. For the assay on chemometers, the red-violet product (chlorophenol red) was wicked through the channel of the chemometer. Color development along the channel was proportional to the number of bacteria. For biosafety consideration, all assays on paper devices were carried out in covered petri dishes at BSL-2 containment. Assay Specificity. Salmonella suspensions of 103, 104 and 105 CFU mL-1 were inoculated with E. coli at ratios of 1:10, 1:100 and 1:1,000 (Salmonella:E. coli) for each Salmonella suspension. The mixed bacteria suspensions underwent the IMS procedure, Ab-biotin and strep-β-gal incubation, and enzymatic assay with CPRG as described above. Finally, the color intensities obtained from the mixed bacteria at different ratios were compared with those of 103, 104 and 105 CFU mL-1 Salmonella without E. coli (as a control) to determine the specificity of the proposed system. Analysis of Bird Fecal Samples. Bird fecal samples were obtained from captive Starling birds maintained by the National Wildlife Research Center, a division of the United States Department of Agriculture located in Fort Collins, Colorado, United States. Samples were collected and kept on ice or in a 4 °C refrigerator until testing. To perform IMS on bird feces samples, ~0.10 g fecal samples were weighed in Eppendorf tubes and 1 mL of Difco nutrient broth was added to the samples. The sample suspensions were vortexed for 30 s to break up the sample matrix and allow for separation of bacteria from the particulates. For inoculated fecal samples, at this point in the procedure, 10 µL of bacteria concentrations ranging from 102 CFU mL-1 to 108 CFU mL-1 were added to the suspensions and vortexed for an additional 15 s. For fecal samples without inoculation, this step was omitted. To observe the microorganism load in the fecal samples before immunomagnetic separation of Salmonella, 10 µL of each sample suspension was plated onto BBLxylose lysine deoxycholate (XLD) agar plates (BD, Sparks, MD,

USA) and labeled as “before IMS”. 5 µL of the antiSalmonella magnetic beads was added into the samples, vortexed for 5-10 s, and then immediately placed on the rotator to avoid fecal matter settling. After incubation for 15 min on the rotator, the samples were vortexed immediately before placing on the magnet to separate the bead-bacteria complex from the matrix. After placing the tubes on the magnet for 2 min, unbound particulate matter was removed using 1 mL pipette tips with the first 5 mm cut off to allow the pipette to remove fecal matter from the tubes without clogging the pipette tip. The beads were washed twice with 500 µL of 1x PBS pH 7.4 before re-suspending the beads in 100 µL of Ab-biotin (0.02 mg mL-1). The remainder of the IMS procedure was completed using the steps and incubation times optimized for detecting Salmonella in media. At the end of the IMS procedure, 10 µL was plated on XLD agar and labeled as “after IMS” to compare to the “before IMS” plates. Finally, 10 µL of re-suspended beads was used to perform the assay with CPRG as previously described. For quantitative determination of Salmonella in the sample, the color intensity of the sample spots was compared with that of standard solutions of known amounts of bacteria. Assessment of Beads Lost to Sample Matrix. To determine if beads were lost in the sample matrix, Abbiotin was directly conjugated to tosyl-activated Dynabeads®. The coupling process was the same as previously described. The antibody-conjugated beads were added directly to either media or bird fecal samples. Samples were incubated with the beads for 15 min on a rotator to replicate the first steps of the IMS procedure. After media or fecal matter was removed from the beads and washed, the beads were incubated with strep-β-gal for 10 min. After incubation, the beads were washed twice with washing buffer (PBS Tween), re-suspended in PBS, followed by the enzymatic assay with CPRG on paper devices as previously described. Analysis of Milk Samples. King Soopers City Market branded organic whole pasteurized milk was purchased from King Soopers in Fort Collins, Colorado, United States, on 05/31/2017. Milk was warmed to room temperature before diluting S. typhimurium in the sample matrix in 10-fold dilutions resulting in concentrations ranging from 101-107 CFU mL-1. The IMS procedure and sandwich immunoassay was carried out as previously described for media. After re-suspending the beads in 100 µL of 1x PBS, 10 µL of the beads was reacted with 10 µL of CPRG for 30 min before analyzing as previously described.



Page 6 of 13

RESULTS AND DISCUSSION A System for S. typhimurium Detection. Detection of S. typhimurium in this work was based on an enzymatic assay using β-galactosidase (β-gal) as the enzyme and CPRG as a substrate. It was reported previously that βgal has a relatively low limit of detection of 0.01 ± 0.01 µg mL-1 when reacting with CPRG,20 making it ideal for detection of S. typhimurium. Because S. typhimurium bacteria do not naturally produce β-gal, we developed a two-step assay to detect bacteria that consists of (i) incubating bacteria with a biotinylated anti-Salmonella antibody (Ab-biotin) and (ii) incubating bacteria with β-gal conjugated streptavidin (strep-β-gal) and colorimetric detection of β-gal retention with CPRG. A schematic of the S. typhimurium detection approach is presented in Fig. 1 (a) and the step-by-step IMS process is in Fig. 2. Before the incubation with Ab-biotin and strep-β-gal, IMS was performed to isolate S. typhimurium from the sample matrix. A S. typhimurium antibody, directed against lipopolysaccharides (LPS) on the exterior of the bacteria, was immobilized onto the magnetic bead surface to capture intact S. typhimurium bacteria from the samples. In addition to separation from the sample matrix, the IMS process was used for pre-concentrating the bacteria to enhance sensitivity of the assay. A second S. typhimurium antibody against “O” and “H” antigens of S. typhimurium conjugated with biotin was incubated with the bead-bacteria complex. For the final step of the assay, the bead-bacteria complex with biotinylated antibody was incubated with the strep-β-gal conjugate. Finally, β-gal labeled bacteria were detected with CPRG on paper devices. The β-gal on the bacteria complex catalyzed the hydrolysis of CPRG into chlorophenol red (Fig. 1 (b)), which is a red-violet product. Development of a red color in the assay indicated the presence of S. typhimurium.

A

B

Figure 1. System for Detecting S. typhimurium. (A) Schematic of selected approach for S. typhimurium detection. (B) S. typhimurium detection based on an enzymatic assay between β-galactosidase and CPRG, resulting in chlorophenol red as a red-violet product.

Incubation Optimization. Optimization of incubation time for each step of the IMS procedure was performed to enhance assay sensitivity and minimize the total analysis time. PADs with an array of wells were used for all optimizations. The incubation of antiSalmonella magnetic beads with bacteria suspension was carried out for 5, 10, 15, 20, and 30 min on a rotator, while the incubation times for Ab-biotin and strep-β-gal were kept constant at 30 min for each step. Gentle continuous agitation using a rotator was performed to prevent the beads from settling while increasing the bacterial capture probability. Longer incubation times resulted in higher signals for 102 and 103 CFU mL-1 (Fig. S1(a) in Supplementary Information), indicating that more bacteria were captured by the beads when longer incubation times were used. However, high concentrations of bacteria (104 CFU mL-1 or more) resulted in relatively constant signals for all incubation times (Fig S1(a)), implying that when there was a high number of bacteria, the number of bacteria was sufficiently high to saturate the assay with chlorophenol red. The number of anti-Salmonella beads used in this work was 100 µg (5 µL of 20 mg mL-1), which was sufficient to differentiate signal intensity of 103 CFU mL-1 and lower when using different incubation times. Because analyses are usually focused on detecting and distinguishing low concentrations of bacteria in samples, 100 µg of beads were used for all remaining


Page 7 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


experiments. As seen in Fig S1(a), the signal intensity was stabilized when the beads were incubated with 103 CFU mL-1 bacteria suspension at least 15 min, whereas it needed at least 20 min incubation for 102 CFU mL-1 to obtain stable signal. However, because there was not a significant difference between 15 and 20 min incubation times for 102 CFU mL-1, 15 min incubation was selected to minimize total analysis time.

Figure 2. Schematic diagram illustrates the process for detection of S. typhimurium, which is based on immunomagnetic separation (15 min), enzyme conjugation (20 min for Ab-biotin and 10 min for strep-β-gal), and the enzymatic assay between β-gal and CPRG (30 min). The assay is carried out in a petri dish for biosafety consideration.

Incubation time for Ab-biotin was assessed for 1, 5, 10, 20 and 30 min by maintaining 15 min incubation for the beads and 30 min for strep-β-gal. The concentration of Ab-biotin was added in excess to the bead-bacteria complex to ensure that all bacteria were captured by the biotinylated anti-Salmonella. Unbound Ab-biotin was removed through two washing steps. As expected, the signal intensity increased with incubation time for 102 and 103 CFU mL-1 (Fig. S1(b)). The signal intensity curves plateaued after 10 min incubation. Therefore, 10 min incubation for Ab-biotin was selected for further optimization of incubation time for strep-β-gal. Dilution of 2 mg mL-1 strep-β-gal was performed at 1:1,000 v/v in PBS and 100 µL of the diluted solution was added to all experiments, which is equal to 0.20 µg strep-β-gal in the assay. It was found that higher concentrations of strep-βgal increased background signal when reacting with CPRG and the red color intensity could not be differentiated between low and high concentrations of S. typhimurium. In addition, false-positive results were observed from E. coli when using high concentrations of strep-βgal, which is likely due to nonspecific binding of strep-βgal to unconjugated spots on the magnetic beads that

were not washed away during washing steps. Conversely, using more diluted strep-β-gal resulted in inadequate β-gal on the bacteria, which was insufficient to detect low concentrations of Salmonella. Like beads and Abbiotin, longer incubation times for strep-β-gal resulted in higher signal intensities (Fig. S1(c)). The signal intensity was stable after strep-β-gal incubation for 10 min. However, it was observed that when using 20 min incubation for Ab-biotin and the bead-bacteria complex, a better signal intensity was obtained, especially with lower concentrations of bacteria. This may be due to longer incubation for Ab-biotin increased the number of biotin molecules on the bacteria complex. This would enhance binding between biotin and streptavidin (conjugated with β-gal), leading to increased signals. Therefore, 20 min incubation for Ab-biotin and 10 min incubation for strep-β-gal were selected as optimized incubation times. Consequently, total time required to complete the IMS and incubation processes was 45 min. Longer incubation time for each step could enhance the sensitivity of the proposed approach, especially at low concentrations of bacteria; however, assay optimization is a compromise between the total analysis time and sensitivity. In addition, it was determined that the reaction between β-gal and CPRG required 30 min to obtain stable color intensity (as shown in Fig. S2). Accordingly, the total analysis time of the proposed approach for detection of S. typhimurium was 75 min. Final Assay and Sensor LOD. Using the optimized incubation times described above, the limit of detection for this approach was determined, as shown in Fig. 3 (a). The proposed system detected S. typhimurium at concentrations as low as 100 CFU mL-1. E. coli was also applied to investigate the selectivity of the proposed approach. As seen in Fig. 3 (a), E. coli only showed a positive result at a relatively high concentration of 107 CFU mL-1 with signal intensity equal to the intensity of only 100 CFU mL-1 of S. typhimurium. Normally, E. coli can produce intracellular β-gal; however, the result showing a relatively low signal of 107 CFU mL-1 E. coli implied that the enzyme is not secreted and is not detectable with CPRG in intact organisms. The slightly positive result from 107 CFU mL-1 of E. coli could be due to non-specific binding of the anti-Salmonella antibody. However, non-specific signals only occurred at an extremely high concentration of E. coli, therefore should not be a concern for realworld samples.

A



B

Figure 3. Assay limit of detection and specificity using the optimized incubation times and demonstrated on two different PADs; (A) well-array devices and (B) chemometers.

Chemometers were implemented for detection of S. typhimurium, as shown in Fig. 3 (b), to provide instrument-free CPRG signal quantification. At a high concentration of bacteria (104-106 CFU mL-1), color development was not significantly different, likely due to assay saturation. However, for 102-104 S. typhimurium, the distances that color developed along the channels were proportional to the bacteria concentration. Because of this, a calibration curve was generated as seen in Fig. S3 by plotting color distance vs. S. typhimurium concentration. The chemometer also showed a detection limit of 100 CFU mL-1 for S. typhimurium detection. Compared with spot tests on the well-array paper devices, performing the assay on chemometers is more convenient for users because the signal can be read directly on the device without image capture and analysis. However, wellarray paper devices are slightly easier, faster, and more convenient for fabrication and preliminary testing. Results obtained from both formats of paper devices confirmed that detection of S. typhimurium using the proposed system was achieved at as low as 100 CFU mL1, with analysis time of 75 min and an additional 10-15 min for washing steps. The analysis time for the entire procedure should be within 90 min without any requirement for complex instruments and culture enrichment steps for bacteria incubation. Therefore, the proposed approach has shown its ability as an alternative to culturing to be a sensitive, easy, rapid, instrument-free, reliable, and portable method for detection of S. typhi-

Page 8 of 13

murium and could be an ideal platform for on-site analysis, especially in chemometer format. Capture and Retention Efficiencies. To determine how efficient anti-Salmonella magnetic beads were for capturing bacteria in solution, the capture efficiencies were calculated. This was done by comparing the S. typhimurium concentration pulled down with antiSalmonella magnetic beads with the original concentrations of S. typhimurium (ranging from 102-105 CFU mL-1). The calculated capture efficiencies ranged from 8.8421.3%, decreasing as bacteria concentration increased (Table 1). This is likely due to bead concentration remaining constant as bacteria concentration increases, therefore fewer bacteria are captured by beads. Because the magnetic beads are smaller than the bacteria, more bacteria in the solution could also lead to steric hindrance, making it harder for bacteria to bind to the antibodies. These capture efficiencies are lower than expected as literature citing the use of polyclonal antibodies conjugated to tosyl-activated Dynabeads® for immunomagnetic separation reports ~40% capture efficiency for 103 CFU mL-1 of Mycobacterium avium subsp. Paratuberculosis.36 However, different antibodies will have different binding affinities, which can impact capture efficiencies. In addition, mycobacteria are much smaller than Salmonella bacteria, so more could potentially bind to the beads. The large standard deviations are likely due to the error associated with plating lower concentrations of bacteria. Moreover, pipetting variability could also be an issue because bacteria are not necessarily evenly distributed throughout the media. This caused variation in the number of bacteria pipetted and rendered quantification of small concentrations of bacteria inaccurate. This error could also explain why retention efficiencies were on average >100% with one of the values at 13% and the other three concentrations resulting in well over 100% efficiency. This mostly consistent >100% efficiency could also be due to possible bacteria enumeration during the IMS procedure. Although the capture efficiencies can likely be improved, retention efficiencies are high and so few bacteria are lost throughout the IMS procedure after initial capture. Increasing the capture efficiency using demonstrated techniques will be the subject of future studies. One possible factor could be further studied to improve the capture efficiency is the volume of beads used for bacteria capture. However, this should be compromised between the capture efficiency and the background signal possibly arising from non-specific binding of the reagents used for the sandwich assay and due to Dynabeads® color being similar to chlorophenyl red, our detectable product. Grant and coworkers also inves-


Page 9 of 13

tigated how different antibodies and magnetic beads increase capture efficiency of their final IMS assay36, something that can be done as future work to ultimately achieve a more sensitive assay. Bacteria concentration (CFU mL-1)

A

Efficiency (%) RetenCapture tion

2.3 × 102

21.3 ± 10.3

13.0 ± 3.1

2.3 × 103

20.2 ±10.8

115.0 ±70.5

2.3 × 104

11.1 ± 4.9

196.0 ± 30.0

2.3 × 105

8.8 ± 5.1

114.0 ± 57.4

Table 1. Capture and retention efficiencies. Determining the anti-Salmonella conjugated Dynabeads® efficiency to capture S. typhimurium in solution and retain bacteria throughout IMS process.

B

150

2

10 cells 3 10 cells 4 10 cells

120

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


90 60 30 0

Assay Specificity. In real-world samples S. typhimurium bacteria are found mixed with many other bacterial species, so a key question is if the assay can specifically detect S. typhimurium in the presence of other bacteria. The specificity of the proposed system was evaluated by mixing S. typhimurium and E. coli cells at different concentration ratios. Three different concentrations of S. typhimurium (102, 103 and 104 CFU mL-1) were used as representatives of low, medium and high concentrations, respectively. Each S. typhimurium suspension was inoculated with different ratios of E. coli to observe the effect of E. coli on the signal intensity from the assay. Results from Fig. 4(a) show that no difference was observed from the color developed in the assays obtained from the bacterial mixtures and the S. typhimurium control, even when the number of E. coli was 1,000 times higher than S. typhimurium. The measured color intensities (Fig. 4(b)) confirmed that there was no significant difference among different E. coli ratios and the S. typhimurium control. In addition, the signal intensities of pure E. coli cultures were relatively low and constant for the concentration range of 102-104 CFU mL-1 and the intensity was found to be approximately 6 times lower than pure S. typhimurium culture at 102 cells. These results show high specificity of the proposed approach for detection of S. typhimurium without interference from E. coli.

E. coli

control

1:10

1:100

1:1,000

Figure 4. Specificity study of the proposed approach for determination of Salmonella in the presence of E. coli at different concentration ratios. (A) Images of the assays and (B) intensities measured from image in (A) and plotted as bar charts for comparison.

S. typhimurium Detection in Bird Feces. After optimizing the IMS procedure in media, the method was used with bird feces to demonstrate this assay could be used to accurately detect S. typhimurium in animal samples. Fecal samples were collected from Starling birds maintained by the National Wildlife Research Center, a division of the United States Department of Agriculture located in Fort Collins, Colorado, United States. First, it was determined whether the same sensitive LOD would transfer from media to fecal samples despite a complex matrix. 0.1 g of fecal sample was inoculated with different concentrations of S. typhimurium before completing the IMS procedure as optimized in media. Final bacteria concentrations in the fecal matrix ranged from 101-107 CFU g-1. To compare bacteria composition of the sample from before the IMS procedure to after, 10 µL of sample



was plated on an XLD agar plate before and after the IMS procedure for comparison. The LOD of S. typhimurium in bird fecal samples was significantly higher than in media at 105 CFU g-1 (Fig. 5(a)). This higher LOD could be due to several factors. Because bird feces are a complex matrix, there could be numerous components that affect specificity of antibodies, reducing bacteria binding. In addition, the acidity of bird feces could decrease binding efficiency of the antibody to the antigen.37 Solid matter in the solution could also affect the beads’ ability to access bacteria in the sample. Lastly, it was observed that the number of beads retrieved from fecal samples varied qualitatively, indicating that beads were likely getting caught in matrix and not adequately recovered by the magnet. When using IMS on fecal samples that had not been inoculated, one slightly positive result was obtained (Fig. 5(b)). However, this result was not confirmed on XLD agar plates. All other fecal samples showed a negative result, and these negatives were confirmed on XLD agar plates. Therefore, of the 10 noninoculated fecal samples, there was one possible false positive, and no false negatives. To confirm whether beads were lost in the sample matrix, Ab-biotin modified beads were added to fecal samples and allowed to incubate on the rotator like traditional IMS. After removing all fecal matter from the solution, Strep-β-gal was conjugated to the biotin and then reconstituted in PBS before reacting with CPRG. Therefore, unlike traditional IMS, regardless of bacteria present in the sample, the beads should react with CPRG if present in the final solution. Fecal samples were placed on the magnet for 1, 3, and 5 min to determine whether extended time would result in improved bead recovery (Fig. S4). Extended time on the magnet does not appear to assist with bead recovery, and as observed, significant variation was observed in beads lost in the sample matrix when compared to beads in media. There is still the possibility of sample matrix conditions affecting the antibodies conjugated to Dynabeads®, and this will be the subject of future research. While the LOD is not as low in fecal samples, IMS has demonstrated in inoculated bird fecal samples the ability to selectively purify S. typhimurium from other bacteria species present in the sample (Fig. S5). S. typhimurium presence is indicated on XLD plates by black colonies while other bacteria species are yellow. Although no S. typhimurium was detected in fecal samples without inoculation, no other bacterial species grew on the “after IMS” agar plates. This indicates that the beads were specific to S. typhimurium and would not adhere to other

Page 10 of 13

bacterial species, resulting in no bacteria after the IMS procedure if S. typhimurium is not present.

Figure 5. Detecting S. typhimurium in bird fecal samples. (A) The LOD of S. typhimurium in inoculated bird fecal samples is 105 CFU g-1, compared to 100 CFU mL-1 in solution. (B) Detecting S. typhimurium in non-inoculated bird feces samples yielded one slightly positive that was not confirmed via traditional plating methods.

S. typhimurium Detection in Milk. To demonstrate the onsite food safety potential of this method, S. typhimurium detection was verified in pasteurized whole milk. Utilizing the optimized incubation times and concentrations demonstrated in media, the detection limit of S. typhimurium in whole milk was 103 CFU mL-1 (Fig. 6). While lower than bird feces, this detection limit is still an order of magnitude higher than media. Although not as complex as feces, milk is still a very complex matrix with solids comprising over 10% of the matrix composition, including numerous proteins, fat globules, and lactose.38 Any of these solids could affect the specificity of the antibodies during capture, thus raising the detection limit. A capture efficiency study was not conducted in any other sample matrix other than media, but will be the subject of future research to determine how different sample conditions affect antibody capture. This detection limit should still be efficient in determining whether whole milk has been contaminated by bacteria, and shows promise as a quick and accurate detection method for infield food safety measurements.


Page 11 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry We thank members of the Henry and Geiss laboratories for useful discussions, and Dr. Mary Jackson for this use of her BSL-2 laboratory for bacteriological work. This work was supported by USDA grant #16-7400-0589 through the National Wildlife Research Center.

REFERENCES

Figure 6. IMS detection of S. typhimurium was demonstrated in whole milk for food applications where the detection limit was 103 CFU mL-1.

CONCLUSIONS Colorimetric PADs coupled with IMS have been developed for simple, rapid and sensitive detection of S. typhimurium in complex sample matrices without preenrichment. A sandwich immunoassay was applied to directly detect S. typhimurium in the sample, with the retention of β-gal on the beads and colorimetric conversion of CPRG to chlorophenyl red as a sensitive detection modality. The detectable level of S. typhimurium was found to be 102 CFU mL-1 in pure culturing solution within 90 min and without pre-enrichment. In addition, the proposed approach was confirmed to be highly selective with S. typhimurium without any interference from E. coli. Chemometers were also used for detection of S. typhimurium and showed the same detection limit as the well-array paper devices, providing a more userfriendly device where image processing was not required. IMS was also demonstrated for the first time in the positive detection of inoculated S. typhimurium in bird fecal samples and whole milk samples with detection limits of 105 CFU g-1 and 103 CFU mL-1, respectively. Low sensitivity of the developed system for detection of S. typhimurium in fecal samples could be due to the matrix effect in feces. However, this method is the first demonstration of an alternative to traditional culture methods, and has shown its potential for onsite detection of S. typhimurium in complex sample matrices.

AUTHOR INFORMATION Corresponding Authors *E-mail: [email protected]. Phone: +1-970-4912852.

ACKNOWLEDGMENT

(1) Newell, D. G.; Koopmans, M.; Verhoef, L.; Duizer, E.; AidaraKane, A.; Sprong, H.; Opsteegh, M.; Langelaar, M.; Threfall, J.; Scheutz, F.; van der Giessen, J.; Kruse, H. International Journal of Food Microbiology 2010, 139, S3-S15. (2) Scallan, E.; Hoekstra, R. M.; Angulo, F. J.; Tauxe, R. V.; Widdowson, M. A.; Roy, S. L.; Jones, J. L.; Griffin, P. M. Emerg Infect Dis 2011, 17, 7-15. (3) Scharff, R. L. J. Food Prot. 2012, 75, 123-131. (4) Glisson, J. R. Poult. Sci. 1998, 77, 1139-1142. (5) Narvaez-Bravo, C.; Rodas-Gonzalez, A.; Fuenmayor, Y.; FloresRondon, C.; Carruyo, G.; Moreno, M.; Perozo-Mena, A.; Hoet, A. E. International Journal of Food Microbiology 2013, 166, 226-230. (6) Renter, D. G.; Gnad, D. P.; Sargeant, J. M.; Hygnstrom, S. E. J Wildl Dis 2006, 42, 699-703. (7) Dione, M. M.; Ieven, M.; Garin, B.; Marcotty, T.; Geerts, S. Journal of Food Protection 2009, 72, 2423-2427. (8) Costello, M.; Rhee, M. S.; Kang, D. H. '. Rapid Methods Autom. Microbiol. 2002, 10, 19-25. (9) Kwon, D.; Joo, J.; Lee, J.; Park, K.-H.; Jeon, S. Analytical Chemistry 2013, 85, 7594-7598. (10) Singer, R. S.; Cooke, C. L.; Maddox, C. W.; Isaacson, R. E.; Wallace, R. L. J Vet Diagn Invest 2006, 18, 319-325. (11) Malorny, B.; Paccassoni, E.; Fach, P.; Bunge, C.; Martin, A.; Helmuth, R. Applied and Environmental Microbiology 2004, 70, 70467052. (12) Scheu, P. M.; Berghof, K.; Stahl, U. Food Microbiology 1998, 15, 13-31. (13) Maurer, J. J. In Annual Review of Food Science and Technology, Vol 2, Doyle, M. P.; Klaenhammer, T. R., Eds., 2011, pp 259-279. (14) Mettakoonpitak, J.; Boehle, K.; Nantaphol, S.; Teengam, P.; Adkins, J. A.; Srisa-Art, M.; Henry, C. S. Electroanalysis 2016, 28, 1420-1436. (15) Yang, Y. Y.; Noviana, E.; Nguyen, M. P.; Geiss, B. J.; Dandy, D. S.; Henry, C. S. Analytical Chemistry 2017, 89, 71-91. (16) Cate, D. M.; Adkins, J. A.; Mettakoonpitak, J.; Henry, C. S. Analytical Chemistry 2015, 87, 19-41. (17) Li, C.-z.; Vandenberg, K.; Prabhulkar, S.; Zhu, X.; Schneper, L.; Methee, K.; Rosser, C. J.; Almeide, E. Biosensors and Bioelectronics 2011, 26, 4342-4348. (18) Wang, Y.; Ping, J.; Ye, Z.; Wu, J.; Ying, Y. Biosensors and Bioelectronics 2013, 49, 492-498. (19) Hossain, S. M. Z.; Ozimok, C.; Sicard, C.; Aguirre, S. D.; Ali, M. M.; Li, Y. F.; Brennan, J. D. Analytical and Bioanalytical Chemistry 2012, 403, 1567-1576. (20) Jokerst, J. C.; Adkins, J. A.; Bisha, B.; Mentele, M. M.; Goodridge, L. D.; Henry, C. S. Analytical Chemistry 2012, 84, 29002907. (21) Adkins, J. A.; Boehle, K.; Friend, C.; Chamberlain, B.; Bisha, B.; Henry, C. S. Analytical Chemistry 2017, 89, 3613-3621. (22) Park, T. S.; Li, W.; McCracken, K. E.; Yoon, J.-Y. Lab on a Chip 2013, 13, 4832-4840. (23) Boehle, K. E.; Gilliand, J.; Wheeldon, C. R.; Holder, A.; Adkins, J. A.; Geiss, B. J.; Ryan, E. P.; Henry, C. S. Angew Chem Int Ed Engl 2017, 56, 6886-6890.



(24) Olsvik, O.; Popovic, T.; Skjerve, E.; Cudjoe, K. S.; Hornes, E.; Ugelstad, J.; Uhlen, M. Clin Microbiol Rev 1994, 7, 43-54. (25) Cohen, S. J.; Punt, C. J.; Iannotti, N.; Saidman, B. H.; Sabbath, K. D.; Gabrail, N. Y.; Picus, J.; Morse, M.; Mitchell, E.; Miller, M. C.; Doyle, G. V.; Tissing, H.; Terstappen, L. W.; Meropol, N. J. J Clin Oncol 2008, 26, 3213-3221. (26) Brandao, D.; Liebana, S.; Campoy, S.; Alegret, S.; Isabel Pividori, M. Talanta 2015, 143, 198-204. (27) Skjerve, E.; Rorvik, L. M.; Olsvik, O. Appl Environ Microbiol 1990, 56, 3478-3481. (28) Skjerve, E.; Olsvik, O. International Journal of Food Microbiology 1991, 14, 11-18. (29) Chapman, P. A.; Wright, D. J.; Siddons, C. A. J Med Microbiol 1994, 40, 424-427. (30) Hyun, K.-A.; Kim, J.; Gwak, H.; Jung, H.-I. Analyst 2016, 141, 382-392. (31) Bakkali Kassimi, L.; Gonzague, M.; Boutrouille, A.; Cruciere, C. Journal of Virological Methods 2002, 101, 197-206. (32) Monceyron, C.; Grinde, B. Journal of Virological Methods 1994, 46, 157-166.

Page 12 of 13

(33) Park, Y.; Cho, Y. H.; Jee, Y.; Ko, G. Appl Environ Microbiol 2008, 74, 4226-4230. (34) Cate, D. M.; Dungchai, W.; Cunningham, J. C.; Volckens, J.; Henry, C. S. Lab on a Chip 2013, 13, 2397-2404. (35) Arun, R. K.; Halder, S.; Chanda, N.; Chakraborty, S. Lab Chip 2014, 14, 1661-1664. (36) Foddai, A.; Elliott, C. T.; Grant, I. R. Applied and Environmental Microbiology 2010, 76, 7550-7558. (37) Devanaboyina, S. C.; Lynch, S. M.; Ober, R. J.; Ram, S.; Kim, D.; Puig-Canto, A.; Breen, S.; Kasturirangan, S.; Fowler, S.; Peng, L.; Zhong, H. H.; Jermutus, L.; Wu, H.; Webster, C.; Ward, E. S.; Gao, C. S. Mabs 2013, 5, 851-859. (38) Ceballos, L. S.; Morales, E. R.; Adarve, G. T.; Castro, J. D.; Martinez, L. P.; Sampelayo, M. R. S. Journal of Food Composition and Analysis 2009, 22, 322-329.


Page 13 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Graphical Abstract Here

ACS Paragon Plus Environment

13

Salmonella typhimurium - ACS Publications - American

Recommend Documents