Nanoparticle Enhanced Antibody and DNA Biosensors for Sensitive Detection of Salmonella [original]

materials

Article

Nanoparticle Enhanced Antibody and DNA

Biosensors for Sensitive Detection of Salmonella

Sumeyra Savas 1, Aylin Ersoy 1, Yakup Gulmez 1, Selcuk Kilic 2, Belkis Levent 2and

Zeynep Altintas 3,*

National Research Institute of Electronics and Cryptology, The Scientific and Technological Research Council

of Turkey (TUBITAK), Kocaeli 41470, Turkey; [email protected].tr (S.S.);

[email protected].tr (A.E.); [email protected].tr (Y.G.)

2Turkey Public Health General Headquarter, Ankara 06100, Turkey; [email protected] (S.K.);

[email protected] (B.L.)

3Institute of Chemistry, Technical University of Berlin, Straße des 17. Juni 124, Berlin 10623, Germany

*Correspondence: [email protected]; Tel.: +49-30-314-23727

Received: 12 August 2018; Accepted: 22 August 2018; Published: 27 August 2018





Abstract:

Bacteria-related pathogenic diseases are one of the major health problems throughout

the world. Salmonella is a genus of rod-shaped Gram-negative enterobacteria of which more than

2600 serotypes have been identified. Infection with Salmonella can cause salmonellosis, a serious

bacterial toxi-infection syndrome associated with gastroenteritis, and paralyphoid and typhoid

fevers. Its rapid and sensitive detection is a key to the prevention of problems related to health.

This paper describes the development of antibody and DNA sensors for Salmonella detection using

a microfluidic-based electrochemical system. Commercial Salmonella typhimurium and Salmonella

typhimurium from human stool samples were investigated using standard and nanomaterial-amplified

antibody sensors. S. typhimurium could be detected down to 1 cfu mL

−1

. The specificity of

immunoassay was tested by studying with non-specific bacteria including E. coli and S. aureus

that revealed only 2.01% and 2.66% binding when compared to the target bacterium. On the other

hand, the quantification of Salmonella DNA was investigated in a concentration range of 0.002–200

using the developed DNA biosensor that demonstrated very high specificity and sensitivity with

a detection limit of 0.94 nM. Our custom-designed microfluidic sensor offers rapid, highly sensitive,

and specific diagnostic assay approaches for pathogen detection.

Keywords:

Salmonella spp.; pathogen detection; antibody biosensor; DNA biosensor; microfluidic-

based electrochemical sensor; nanoparticle enhanced bio-detection; infectious diseases

1. Introduction

Foodborne diseases lead to diverse health problems worldwide [

]. The World Health

Organization reported that Salmonella typhimurium and Salmonella enteritidis are the most common

causes of foodborne illnesses all over the world. According to the reports of the European Food Safety

Authority, human salmonellosis has resulted in three billion euros loss per year [

]. In some regions,

more than 90% of Salmonella strains isolated from humans until 1970 was S. typhimurium, but now

incidence of S. enteritidis is also gradually increasing, which has been the most frequently isolated

serotype in the last 10 years [

]. The symptoms of Salmonella infection include abdominal pain, fever,

nausea, vomiting, diarrhea, dehydration, weakness, and loss of appetite, and the symptoms normally

appear 12–72 h after ingestion of contaminated foods or beverages [5].

There are several methods used for the detection of Salmonella serotypes, such as cultivation

techniques [

], enzyme-linked immunosorbent assays (ELISAs) [

], polymerase chain reaction (PCR)

Materials 2018,11, 1541; doi:10.3390/ma11091541 www.mdpi.com/journal/materials

Materials 2018,11, 1541 2 of 17

methods [

–

], and biosensors [

]. The gold standard for detection of Salmonella serotypes are

still conventional methods that are sensitive and inexpensive. However, these techniques require

more than five days to obtain a result and often lack in providing good specificity and sensitivity.

Additionally, the cultivation techniques are generally time consuming and the limit of detection for the

analysis is insufficient. The use of biosensor technology is a strong alternative to the other techniques

by offering highly sensitive, rapid, and easy-to-use bio-detection principles [

]. Today, biosensors

are widely used for pathogen detection and are able to measure bacteria down to 1 cfu mL

−1

. This is

particularly due to the significant impact of nanomaterials on the advancement of biosensors and

biosensing principles [

–

]. Furthermore, microbial biosensors often require sample volumes in the

microliter range and very short analysis time.

Various transducer systems based on surface plasmon resonance (SPR) [

], quartz crystal

microbalance (QCM) [

], and electrochemical sensing strategies [

] have been successfully employed

for bacteria detection. Offering very high sensitivity and good detection capacity electrochemical

sensors are among the widely used systems for Salmonella quantification [

]. Zhu et al.

developed a multichannel electrochemical immunosensor for Salmonella detection by combining the

rolling circle amplification with DNA-gold nanoparticles (AuNPs) probe [

]. Afonso et al. reported

a disposable immunosensor for electrochemical detection of Salmonella enterica subsp. Enterica serovar

Typhimurium LT2 using gold nanoparticles and magneto-immunoassay [

]. Sensitive amperometric

detection of Salmonella was also reported [

–

]. Even though great progress has been demonstrated

in the Salmonella detection, employing only one antibody in the biosensor design often results in

insufficient sensitivity, when the sensor could hardly discriminate between two LPS samples from two

different Gram-negative bacteria [25].

To overcome the selectivity problem, herein, we report antibody- and DNA-based biosensors

for Salmonella detection using a fully-automated custom-designed microfluidic sensing device

(MiSens) [

] that is composed of an electromechanical unit controlling the assay protocol via its

integrated software (MiCont

). Normal and gold-nanoparticle amplified sandwich assays were

developed and used for the detection of commercial Salmonella samples and real samples from human

stool. DNA biosensor was developed by capturing the surface DNA probe on the neutravidin (NA)

immobilized sensor surface and then measuring the target Salmonella DNA based on the hybridization

reaction that occurs between the target DNA and the surface probe. As the measurement system relies

on the enzymatic reaction between horseradish peroxidase (HRP) and 3,3

,5,5

-tetramethylbenzidine

(TMB), the detector antibody and the DNA detection probe were both labeled with HRP. We have

demonstrated that the developed antibody and DNA biosensors are capable of measuring trace

amounts of Salmonella and Salmonella DNA, respectively.

2. Materials and Methods

2.1. Materials and Reagents

A monoclonal anti-Salmonella antibody was bought from BIO-RAD (Puchheim, Germany).

Peroxidase-labeled goat anti-Salmonella secondary antibody (BacTrace

Anti-Salmonella CSA-1 Antibody)

and Salmonella typhimurium were purchased from SeraCare Life Sciences (Gaithersburg, MD, USA).

Staphylococcus aureus and E. coli from human stool samples were obtained from the Public Health Agency

(Ankara, Turkey) for cross-reactivity studies. 11-Mercaptoundecanoic acid (MUDA), phosphate-buffered

saline tablets (PBS, 0.01 M phosphate buffer, 0.0027 M potassium chloride and 0.137 M sodium

chloride, pH 7.4), N-hydroxysuccinimide (NHS), analytical grade ethanol, horseradish peroxidase

(HRP), ethanolamine, and 3,3

,5,5

-tetramethylbenzidine (TMB) ready to use reagent with H

were purchased from Sigma Aldrich (Poole, UK). The gold nanoparticles in 15 nm (for DNA

assays) and 40 nm (for antibody assays) sizes were purchased from BBI International (Cardiff, UK).

Ultrapure water (18 M

Ω

−1

) produced by a Milli-Q water system was used for analyses (Millipore Corp.,

Tokyo, Japan). 1-Ethyl-3-(3-dimethylaminopropyl)-carbodiimide (EDC) and biotin were purchased

Materials 2018,11, 1541 3 of 17

from Thermo Fisher Scientific (Loughborough, UK). The oligonucleotide sequences (target sequence:

-ACCGACGGCGAGACCGACTTT-3

; surface probe: Biotin-5

-CTCACCAGGAGATTACAACATGG;

detection probe: Biotin-3

-AGTGGCTAAAAGTCGGTCTC; control surface probe: Biotin-5

CAATATTTGGCGTGAATGGGTCGGAAAACA) for DNA sensor development were obtained from

Sentromer DNA Technologies LLC (Istanbul, Turkey).

2.2. Isolation of Salmonella Typhimurium from Human Stool Samples

Human stool sample was inoculated onto Salmonella-Shigella and xylose lysine deoxycholate agar

mediums, and incubated at 37

◦

C overnight. Bacterial culture was identified by using standard

biochemical tests (the use of glucose, citrate, and indole by bacteria, the production of gas and H

upon lactose fermentation of bacteria, the determination of the mobility and the ability to split urea).

The cultured colonies used glucose, citrate and indole confirmed the presence of Salmonella bacteria.

The isolate was serogrouped and serotyped using polyvalent and monovalent Salmonella antisera (Statens

Serum Institut, Copenhagen, Denmark) according to the Kauffmann-White scheme [

]. The colonies that

produced H

S were considered as a pure colony based on their morphology. After the confirmation and

serotyping, the strain was immediately frozen in tryptic soy broth with 16% glycerol at −80 ◦C.

2.3. Fully-Automated Microfluidic-Based Electrochemical Sensor with a New Chip Design

A custom-designed fully-automated electrochemical sensor was used in this study to develop

antibody- and DNA-based biosensors for Salmonella detection. The sensor fabrication and its developmental

stages were fully described in our earlier studies [

]. Meanwhile, we have realized several drawbacks

dealing with device’s electronics and mechanics. One of the main problems was electromagnetic noise

in the signal. In order to decrease the noise level, two modifications were employed: (1) Increasing of

the distance between the microfluidic pumps and the potentiometer. The microfluidic pump spread

unwanted magnetic pulses by the pulse width modulation (PWM) mechanism; (2) Applying a coaxial

cable structure instead of unscreened cable structure between the biosensor and potentiostat circuit. Herein,

we designed and integrated a new chip (Figure 1) to improve the efficiency of the biosensing device and

the reproducibility of the assays. The electrodes were designed on the glass slide (10

20 mm) using a fine

metal mask, which was made of a laser-cut patterned stainless steel. Au was deposited on the wafer by

means of an electron beam evaporator (Ebeam system Nanovak NVEB-600, Nanovak, Ankara, Turkey).

Prior to the application of Au (200 nm), a 40 nm Ti layer was applied on to the wafer as an intermediary

adhesive layer to enhance the adhesion between the glass slide and the Au. Each array contains eight

working electrodes with a shared Au counter and quasi-reference electrodes. The experiments were

carried out using the MiCont

software (TUB˙

ITAK-B˙

ILGEM, Kocaeli, Turkey) running on a wireless PC.

The assay protocols were generated, saved, and used when required.

Materials 2018, 11, x FOR PEER REVIEW 3 of 17

control surface probe: Biotin-5′- CAATATTTGGCGTGAATGGGTCGGAAAACA) for DNA sensor

development were obtained from Sentromer DNA Technologies LLC (Istanbul, Turkey).

2.2. Isolation of Salmonella Typhimurium from Human Stool Samples

Human stool sample was inoculated onto Salmonella-Shigella and xylose lysine deoxycholate

agar mediums, and incubated at 37 °C overnight. Bacterial culture was identified by using standard

biochemical tests (the use of glucose, citrate, and indole by bacteria, the production of gas and H

upon lactose fermentation of bacteria, the determination of the mobility and the ability to split urea).

The cultured colonies used glucose, citrate and indole confirmed the presence of Salmonella bacteria.

The isolate was serogrouped and serotyped using polyvalent and monovalent Salmonella antisera

(Statens Serum Institut, Copenhagen, Denmark) according to the Kauffmann-White scheme [26]. The

colonies that produced H

S were considered as a pure colony based on their morphology. After the

confirmation and serotyping, the strain was immediately frozen in tryptic soy broth with 16%

glycerol at −80 °C.

2.3. Fully-Automated Microfluidic-Based Electrochemical Sensor with a New Chip Design

A custom-designed fully-automated electrochemical sensor was used in this study to develop

antibody- and DNA-based biosensors for Salmonella detection. The sensor fabrication and its

developmental stages were fully described in our earlier studies [18,27,28]. Meanwhile, we have

realized several drawbacks dealing with device’s electronics and mechanics. One of the main

problems was electromagnetic noise in the signal. In order to decrease the noise level, two

modifications were employed: (1) Increasing of the distance between the microfluidic pumps and the

potentiometer. The microfluidic pump spread unwanted magnetic pulses by the pulse width

modulation (PWM) mechanism. (2) Applying a coaxial cable structure instead of unscreened cable

structure between the biosensor and potentiostat circuit. Herein, we designed and integrated a new

chip (Figure 1) to improve the efficiency of the biosensing device and the reproducibility of the assays.

The electrodes were designed on the glass slide (10 × 20 mm) using a fine metal mask, which was

made of a laser-cut patterned stainless steel. Au was deposited on the wafer by means of an electron

beam evaporator (Ebeam system Nanovak NVEB-600, Nanovak, Ankara, Turkey). Prior to the

application of Au (200 nm), a 40 nm Ti layer was applied on to the wafer as an intermediary adhesive

layer to enhance the adhesion between the glass slide and the Au. Each array contains eight working

electrodes with a shared Au counter and quasi-reference electrodes. The experiments were carried

out using the MiCont

software (TUBİTAK-BİLGEM, Kocaeli, Turkey) running on a wireless PC.

The assay protocols were generated, saved, and used when required.

Figure 1. Illustration of MiSens electrochemical sensor and its new chip design.

Materials 2018,11, 1541 4 of 17

2.4. Sensor Chip Cleaning and SAM Deposition

Prior to forming a self-assembled monolayer (SAM) on the sensor chip, the electrode surfaces were

cleaned by employing nitrogen plasma [

]. A 2 mM concentration of MUDA was used to prepare

the thiol solution in absolute ethanol for SAM deposition [

]. The sensor chips were immersed in the

ethanolic solution for overnight followed by washing with ethanol and Milli-Q water, respectively.

Later on the sensor chips were dried thoroughly under a stream of nitrogen gas, vacuum-packed,

and stored at +4 ◦C till use.

2.5. Selection of HRP Concentration for Bioassays

Different concentrations of HRP were initially measured using MiSens device (TUB˙

ITAK-B˙

ILGEM,

Kocaeli, Turkey) to determine the optimum HRP amount for bioassays. For this, six different

concentrations of HRP were mixed with same amount of TMB and injected to the MUDA coated

surfaces. These optimization experiments were repeated three times and the optimum HRP

concentration was chosen based on the average sensor signals.

2.6. Characterization of SAM Coated Sensor Chips Using AFM

The bare gold, the SAM coated and the antibody immobilized sensor chips were visualized

by employing a Naio (Nanosurf AG, Liestal, Switzerland) atomic force microscope (AFM).

Commercially available AFM probes (NCLR) from NanoWorld (NanoWorld AG, Liestal, Switzerland)

were used for AFM measurements. The AFM analyses for the SAM-deposited surface and the antibody

immobilization were carried out after having completed three cycles of buffer wash flow. The sensor

surfaces during the DNA assays were also visualized using AFM. Hence, the NA immobilization,

the capturing of DNA surface probe on the NA layer, and the target DNA binding were confirmed.

The chips were undocked from the MiSens sensor after three cycles of buffer flow, dried using a gentle

nitrogen stream, and then immediately placed in a glass Petri dish. All AFM measurements were

performed at room temperature using the intermittent air mode.

2.7. Development of the Antibody Biosensor

Two different antibody sensors were developed in this work by using a monoclonal primary

antibody and a secondary polyclonal antibody for Salmonella. The primary antibody was used as

a surface ligand to specifically capture Salmonella bacterium, whereas the polyclonal antibody was

utilized as a detector antibody to increase the sensor signal. Two different bio-detection assays

were established: a standard sandwich assay (Scheme 1) and a nanoparticle enhanced sensor assay

(Scheme 2). In the latter case the polyclonal antibodies were initially conjugated with gold nanoparticles

prior to the Salmonella detection assay.

Materials 2018, 11, x FOR PEER REVIEW 4 of 17

2.4. Sensor Chip Cleaning and SAM Deposition

Prior to forming a self-assembled monolayer (SAM) on the sensor chip, the electrode surfaces

were cleaned by employing nitrogen plasma [29,30]. A 2 mM concentration of MUDA was used to

prepare the thiol solution in absolute ethanol for SAM deposition [18]. The sensor chips were

immersed in the ethanolic solution for overnight followed by washing with ethanol and Milli-Q

water, respectively. Later on the sensor chips were dried thoroughly under a stream of nitrogen gas,

vacuum-packed, and stored at +4 °C till use.

2.5. Selection of HRP Concentration for Bioassays

Different concentrations of HRP were initially measured using MiSens device (TUBİTAK-

BİLGEM, Kocaeli, Turkey) to determine the optimum HRP amount for bioassays. For this, six

different concentrations of HRP were mixed with same amount of TMB and injected to the MUDA

coated surfaces. These optimization experiments were repeated three times and the optimum HRP

concentration was chosen based on the average sensor signals.

2.6. Characterization of SAM Coated Sensor Chips Using AFM

The bare gold, the SAM coated and the antibody immobilized sensor chips were visualized by

employing a Naio (Nanosurf AG, Liestal, Switzerland) atomic force microscope (AFM).

Commercially available AFM probes (NCLR) from NanoWorld (NanoWorld AG, Liestal,

Switzerland) were used for AFM measurements. The AFM analyses for the SAM-deposited surface

and the antibody immobilization were carried out after having completed three cycles of buffer wash

flow. The sensor surfaces during the DNA assays were also visualized using AFM. Hence, the NA

immobilization, the capturing of DNA surface probe on the NA layer, and the target DNA binding

were confirmed. The chips were undocked from the MiSens sensor after three cycles of buffer flow,

dried using a gentle nitrogen stream, and then immediately placed in a glass Petri dish. All AFM

measurements were performed at room temperature using the intermittent air mode.

2.7. Development of the Antibody Biosensor

Two different antibody sensors were developed in this work by using a monoclonal primary

antibody and a secondary polyclonal antibody for Salmonella. The primary antibody was used as a

surface ligand to specifically capture Salmonella bacterium, whereas the polyclonal antibody was

utilized as a detector antibody to increase the sensor signal. Two different bio-detection assays were

established: a standard sandwich assay (Scheme 1) and a nanoparticle enhanced sensor assay

(Scheme 2). In the latter case the polyclonal antibodies were initially conjugated with gold

nanoparticles prior to the Salmonella detection assay.

Scheme 1. Principle of antibody sensor development for Salmonella detection using the standard assay.

Scheme 1.

Principle of antibody sensor development for Salmonella detection using the standard assay.

Materials 2018,11, 1541 5 of 17

Materials 2018, 11, x FOR PEER REVIEW 5 of 17

The running buffer used for immobilization was degassed buffered saline (PBS, pH 7.4) and this

buffer continuously flowed over the Au sensor surfaces between the injections. The flow rate of the

buffer/reagents for the assay was 50 µL min

−1

unless written otherwise. For the assays, the SAM

deposited sensor chip was initially inserted to the device and primed with the running buffer (PBS).

The sensor surface was activated with a mixture of EDC (0.4 M) and NHS (0.1 M) in a 1:1 volume

ratio by 4 min injection prior to the covalent immobilization of the primary antibody (50 µg mL

−1

prepared in NaAc buffer, pH: 4.5) during the 4 min injection (50 µL min

−1

, 200 µL). The antibody-free

areas of the sensing surface were then blocked using a 100 µg mL

−1

BSA solution (50 µL min

−1

, 200

µL) and 1 M of ethanolamine (pH: 8.5, 50 µL min

−1

, 200 µL), respectively. Bio-detection capacity of

the sensor was tested in the investigation range of 1‒5.41 × 10

cfu mL

−1

using two different Salmonella

sample sources: commercially available S. typhimurium and real S. typhimurium samples from human

stool. The samples were prepared in a particular PBS buffer containing 200 µg mL

−1

BSA, 0.5 M NaCl,

500 µg mL

−1

dextran, and 0.5% Tween 20. Each sample was injected to the sensor surface (50 µL min

−1

200 µL) followed by the injection of the HRP-labelled secondary antibody (50 µL min

−1

, 200 µL).

Amperometric measurements were then carried out at −0.1 V using TMB reagent (4 min, 50 µL min

−1

)

followed by buffer injection (4 min, 100 µL min

−1

). The sensor surface regeneration was achieved with

the injection of 0.1 M HCl (1 min, 100 µL min

−1

) twice for the sequential detection of different

Salmonella concentrations.

The gold nanoparticle-enhanced bioassays (Scheme 2) were performed using the same

procedure. In this case, the HRP-labelled secondary antibody was conjugated with AuNPs prior to

the experiments to amplify the sensor signal. The AuNP conjugation protocol was described in our

earlier studies [18] and used as is in this work. The secondary antibody-labelled AuNPs were stored

at +4 °C and warmed to room temperature before use. The concentration of nanoparticles was

calculated at 525 nm wavelength and the AuNP solution was diluted based on the dilution factor

calculated by considering the OD value. Commercially available S. typhimurium and real S.

typhimurium samples from human stool were detected in a concentration range from 1 to 5.41 × 10

cfu mL

−1

using the amperometric sensor.

Scheme 2. Principle of the gold nanoparticles enhanced immunosensor for Salmonella detection.

The running buffer used for immobilization was degassed buffered saline (PBS, pH 7.4) and this

buffer continuously flowed over the Au sensor surfaces between the injections. The flow rate of the

buffer/reagents for the assay was 50

L min

−1

unless written otherwise. For the assays, the SAM

deposited sensor chip was initially inserted to the device and primed with the running buffer (PBS).

The sensor surface was activated with a mixture of EDC (0.4 M) and NHS (0.1 M) in a 1:1 volume

ratio by 4 min injection prior to the covalent immobilization of the primary antibody (50

g mL

−1

prepared in NaAc buffer, pH: 4.5) during the 4 min injection (50

L min

−1

, 200

L). The antibody-free

areas of the sensing surface were then blocked using a 100

g mL

−1

BSA solution (50

L min

−1

, 200

and 1 M of ethanolamine (pH: 8.5, 50

L min

−1

, 200

L), respectively. Bio-detection capacity of the

sensor was tested in the investigation range of 1–5.41

cfu mL

−1

using two different Salmonella

sample sources: commercially available S. typhimurium and real S. typhimurium samples from human

stool. The samples were prepared in a particular PBS buffer containing 200

g mL

−1

BSA, 0.5 M

NaCl, 500

g mL

−1

dextran, and 0.5% Tween 20. Each sample was injected to the sensor surface

(50

L min

−1

, 200

L) followed by the injection of the HRP-labelled secondary antibody (50

L min

−1

200

L). Amperometric measurements were then carried out at

−

0.1 V using TMB reagent (4 min,

L min

−1

) followed by buffer injection (4 min, 100

L min

−1

). The sensor surface regeneration was

achieved with the injection of 0.1 M HCl (1 min, 100

L min

−1

) twice for the sequential detection of

different Salmonella concentrations.

The gold nanoparticle-enhanced bioassays (Scheme 2) were performed using the same procedure.

In this case, the HRP-labelled secondary antibody was conjugated with AuNPs prior to the experiments

to amplify the sensor signal. The AuNP conjugation protocol was described in our earlier studies [

]

and used as is in this work. The secondary antibody-labelled AuNPs were stored at +4

◦

C and

warmed to room temperature before use. The concentration of nanoparticles was calculated at

525 nm wavelength and the AuNP solution was diluted based on the dilution factor calculated by

considering the OD value. Commercially available S. typhimurium and real S. typhimurium samples

from human stool were detected in a concentration range from 1 to 5.41

cfu mL

−1

using the

amperometric sensor.

Materials 2018,11, 1541 6 of 17

2.8. Development of the DNA Biosensor

2.8.1. Immobilization of Neutravidin and DNA Capture Probe on the Sensor Chip Surface

The SAM formation was initially performed on the sensor chip using 2 mM MUDA as described

in Section 2.4. The chip and the PMMA cassette were then assembled together using a double-sided

sticky tape and docked to the sensor device that was primed with degassed Dulbecco’s modified

phosphate-buffered saline (PBS, pH: 7.4). The same reagent was used as the running buffer during NA

immobilization and continuously flowed over the sensor surfaces between the injections. The flow rate

of the sensor was 50

L min

−1

during the DNA bioassays unless written otherwise. The SAM-coated

electrode surfaces of the sensor chip were activated using amine coupling chemistry [

]. For this,

a mixture of EDC (0.4 M)-NHS (0.1 M) at 1:1 volume ratio was injected to the surface during 4 min.

The NA prepared in PBS (50

g mL

−1

, 250

L) was then immobilized to the sensor surface followed

by the injection of ethanolamine (1 M, 250

−1

pH: 8.5) for blocking of the NA-free areas on the

surface [

]. After immobilization of NA, the running buffer was changed to Tris buffer (20 mM

Tris-HCI, 150 mM NaCl, and 1 mM EDTA, pH: 7.0) and it was used during the capturing of the

biotinylated complementary surface probe and the detection of Salmonella DNA. The surface probe

was prepared in Tris buffer in the concentration of 10

M and injected to the NA-immobilized surface

for 5 min, followed by a 1 min injection of 10 mM biotin to block the remaining active sites of the NA

surface. To obtain a control surface, a biotinylated non-specific surface probe was captured on the

NA layer and used for the target detection assays. The sensor signals obtained upon the binding of

Salmonella DNA on the non-specific surfaces were subtracted from the results obtained on the target

specific surfaces.

2.8.2. Preparation of HRP-NA-Labelled AuNPs

The HRP-NA labelled AuNPs were used for the sensitive detection of the target DNA. In this

assay, HRP is required to convert TMB

red

to TMB

, and produce a measurement signal using the

amperometric sensor (Scheme 3). It was well defined that proteins can physically bind to the gold

surface via electrostatic interactions [

]. Herein, the same approach was used for modification of

AuNPs with HRP and NA. For this, a 1 mL solution of the gold nanoparticles (15 nm) was derivatised

with NA (1.5

L, 1 mg mL

−1

) and HRP (2.5

L, 1 mg mL

−1

) in an Eppendorf tube that was covered by

aluminum foil to prevent exposure to light. This solution was incubated for 45 min on a shaker at room

temperature prior to centrifugation for discarding excess reagents. The supernatant was removed,

and 30

L of BSA (10 mg mL

−1

) and 100

L Tris buffer (20 mM) were added to the tube, respectively.

The modified AuNPs were stored at +4

◦

C until use. The AuNP concentration was determined using

a spectrophotometer at a wavelength of 525 nm.

Materials 2018, 11, x FOR PEER REVIEW 6 of 17

2.8. Development of the DNA Biosensor

2.8.1. Immobilization of Neutravidin and DNA Capture Probe on the Sensor Chip Surface

The SAM formation was initially performed on the sensor chip using 2 mM MUDA as described

in Section 2.4. The chip and the PMMA cassette were then assembled together using a double-sided

sticky tape and docked to the sensor device that was primed with degassed Dulbecco’s modified

phosphate-buffered saline (PBS, pH: 7.4). The same reagent was used as the running buffer during

NA immobilization and continuously flowed over the sensor surfaces between the injections. The

flow rate of the sensor was 50 µL min

−1

during the DNA bioassays unless written otherwise. The

SAM-coated electrode surfaces of the sensor chip were activated using amine coupling chemistry

[29,31]. For this, a mixture of EDC (0.4 M)-NHS (0.1 M) at 1:1 volume ratio was injected to the surface

during 4 min. The NA prepared in PBS (50 µg mL

−1

, 250 µL) was then immobilized to the sensor

surface followed by the injection of ethanolamine (1 M, 250 µL

−1

pH: 8.5) for blocking of the NA-free

areas on the surface [31]. After immobilization of NA, the running buffer was changed to Tris buffer

(20 mM Tris-HCI, 150 mM NaCl, and 1 mM EDTA, pH: 7.0) and it was used during the capturing of

the biotinylated complementary surface probe and the detection of Salmonella DNA. The surface

probe was prepared in Tris buffer in the concentration of 10 µM and injected to the NA-immobilized

surface for 5 min, followed by a 1 min injection of 10 mM biotin to block the remaining active sites of

the NA surface. To obtain a control surface, a biotinylated non-specific surface probe was captured

on the NA layer and used for the target detection assays. The sensor signals obtained upon the

binding of Salmonella DNA on the non-specific surfaces were subtracted from the results obtained on

the target specific surfaces.

2.8.2. Preparation of HRP-NA-Labelled AuNPs

The HRP-NA labelled AuNPs were used for the sensitive detection of the target DNA. In this

assay, HRP is required to convert TMB

red

to TMB

, and produce a measurement signal using the

amperometric sensor (Scheme 3). It was well defined that proteins can physically bind to the gold

surface via electrostatic interactions [32]. Herein, the same approach was used for modification of

AuNPs with HRP and NA. For this, a 1 mL solution of the gold nanoparticles (15 nm) was derivatised

with NA (1.5 µL, 1 mg mL

−1

) and HRP (2.5 µL, 1 mg mL

−1

) in an Eppendorf tube that was covered by

aluminum foil to prevent exposure to light. This solution was incubated for 45 min on a shaker at

room temperature prior to centrifugation for discarding excess reagents. The supernatant was

removed, and 30 µL of BSA (10 mg mL

−1

) and 100 µL Tris buffer (20 mM) were added to the tube,

respectively. The modified AuNPs were stored at +4 °C until use. The AuNP concentration was

determined using a spectrophotometer at a wavelength of 525 nm.

Scheme 3. Principle of the DNA assay using the microfluidic-based electrochemical biosensor.

Materials 2018,11, 1541 7 of 17

2.8.3. DNA Detection Assay

Specific gene for Salmonella spp. was selected based on the literature [

]. For DNA assays,

the biotinylated surface probe was initially captured by the NA immobilized surface. The concentrations

of capture (10

M) and detection (10

M) probes were kept the same for all experiments. The biotin-free

Salmonella DNA was incubated with the biotinylated detection probe for 10 min at 55

◦

C using a thermal

cycler. The hybridized DNA molecules were then injected to the sensor surface during 5 min (250

followed by the injection of the HRP-NA modified AuNPs for 4 min (200

L). The modified AuNPs

bound to the sensor surface via interaction between the biotin label of the detection probe and the

NA label of the AuNPs. The electrochemical signal was obtained using a real-time electrochemical

profiling

(REP

, TUB˙

ITAK-B˙

ILGEM, Kocaeli, Turkey) assay at a fixed potential of

−

0.1 V and the

current was measured continuously. The current values were recorded during Tris buffer injection and

this signal was considered as a baseline. The subsequent injection of 200

L TMB revealed a current

change and the subtraction between the current obtained during TMB and buffer was used as the sensor

signal (Figure 2).

Materials 2018, 11, x FOR PEER REVIEW 7 of 17

2.8.3. DNA Detection Assay

Specific gene for Salmonella spp. was selected based on the literature [33]. For DNA assays, the

biotinylated surface probe was initially captured by the NA immobilized surface. The concentrations

of capture (10 µM) and detection (10 µM) probes were kept the same for all experiments. The biotin-

free Salmonella DNA was incubated with the biotinylated detection probe for 10 min at 55 °C using a

thermal cycler. The hybridized DNA molecules were then injected to the sensor surface during 5 min

(250 µL) followed by the injection of the HRP-NA modified AuNPs for 4 min (200 µL). The modified

AuNPs bound to the sensor surface via interaction between the biotin label of the detection probe

and the NA label of the AuNPs. The electrochemical signal was obtained using a real-time

electrochemical profilingTM (REPTM, TUBİTAK-BİLGEM, Kocaeli, Turkey) assay at a fixed potential of

−0.1 V and the current was measured continuously. The current values were recorded during Tris

buffer injection and this signal was considered as a baseline. The subsequent injection of 200 µL TMB

revealed a current change and the subtraction between the current obtained during TMB and buffer

was used as the sensor signal (Figure 2).

Figure 2. The current response of an individual electrode at −0.1 V potential is recorded

uninterruptedly during the consequent injections of buffer and TMB substrate to the HRP-bound

electrode to obtain a sensorgram. The flow rate of the device is 50 µL min−1 during the injections.

3. Results and Discussion

3.1. Antibody Sensor for Salmonella Detection

Prior to developing the bioassays, we initially investigated the optimum HRP concentration for

use in electrochemical sensor. The sensor signal gradually increased from 1.5 ng mL−1 to 12 ng mL−1

of HRP, whereas it showed a clear decrease afterwards (Figure 3). Hence, 12 ng mL−1 was selected as

the optimum concentration for HRP and used for the entire study.

-800

-700

-600

-500

-400

-300

-200

-100

0 50 100 150 200 250 300 350 400 450 500

Sensor signal (nA)

Tris buffer

Sensor Signal (nA)

Time (s)

TMB Tris buffer

Figure 2.

The current response of an individual electrode at

−

0.1 V potential is recorded

uninterruptedly during the consequent injections of buffer and TMB substrate to the HRP-bound

electrode to obtain a sensorgram. The flow rate of the device is 50 µL min−1during the injections.

3. Results and Discussion

3.1. Antibody Sensor for Salmonella Detection

Prior to developing the bioassays, we initially investigated the optimum HRP concentration for

use in electrochemical sensor. The sensor signal gradually increased from 1.5 ng mL

−1

to 12 ng mL

−1

of HRP, whereas it showed a clear decrease afterwards (Figure 3). Hence, 12 ng mL

−1

was selected as

the optimum concentration for HRP and used for the entire study.

Materials 2018,11, 1541 8 of 17

Materials 2018, 11, x FOR PEER REVIEW 8 of 17

1.53 6 121824

100

150

200

250

HRP-Signal Relationship

Sensor Signal (nA)

HRP Concentration (ng mL

-1

)

(A)

(B)

Figure 3. Overall results of HRP-TMB optimization assays in a concentration range of 1.5–24 ng mL−1

(A) and real-time sensorgrams obtained with six different HRP concentrations (B).

3.1.1. Standard Sandwich Assay for the Analysis of Commercial and Real Samples

The initial investigations were carried out with commercially available Salmonella typhimurium

samples using a standard sandwich assay. The bare, the SAM-coated, and the antibody-immobilized

sensor surfaces were characterized by AFM (Figure 4). The 3D surface topology images in 1 × 1 µm

scanning area resulted in the heights of 11.5 nm, 12.3 nm, 22 nm for bare (Figure 4A), MUDA-coated

(Figure 4B), and antibody-immobilized (Figure 4C) surfaces, respectively. The gradual increase of the

surface height confirmed the successful preparation of the sensor chip for bacteria detection assays.

-350

-300

-250

-200

-150

-100

-50

0 50 100 150 200 250 300 350 400

Sensor Signal (nA)

Time (s)

a: 1.5 ng mL

-1

HRP

b: 3 ng mL

-1

HRP

c: 6 ng mL

-1

HRP

d: 24 ng mL

-1

HRP

e: 18 ng mL

-1

HRP

f: 12 ng mL

-1

HRP

Figure 3.

Overall results of HRP-TMB optimization assays in a concentration range of 1.5–24 ng mL

−1

(A) and real-time sensorgrams obtained with six different HRP concentrations (B).

3.1.1. Standard Sandwich Assay for the Analysis of Commercial and Real Samples

The initial investigations were carried out with commercially available Salmonella typhimurium

samples using a standard sandwich assay. The bare, the SAM-coated, and the antibody-immobilized

sensor surfaces were characterized by AFM (Figure 4). The 3D surface topology images in 1

scanning area resulted in the heights of 11.5 nm, 12.3 nm, 22 nm for bare (Figure 4A), MUDA-coated

(Figure 4B), and antibody-immobilized (Figure 4C) surfaces, respectively. The gradual increase of the

surface height confirmed the successful preparation of the sensor chip for bacteria detection assays.

Materials 2018,11, 1541 9 of 17

Materials 2018, 11, x FOR PEER REVIEW 9 of 17

A) B)

C) D)

Figure 4. AFM analysis of bare (A), MUDA-coated (B) and antibody-immobilized (C,D) sensor

surfaces.

Salmonella detection was investigated in the concentration range of 1–5.41 × 107 cfu mL−1, which

provided a limit of detection (LOD) of 2.7 × 101 cfu mL−1 for both commercial and real samples. The

average signal ratio in the entire concentration range between commercial and real sample analyses

was determined as 1.1. The overall results of Salmonella detection were subjected to the logarithmic

regression analysis that resulted in the correlation coefficients of 0.9671 and 0.9505 for commercial

and real samples, respectively (Figure 5).

Figure 5. Overall results of Salmonella detection in a concentration range of 27—5.41 × 107 using a

standard sandwich assay (n = 6).

y = 37.238ln(x) - 83.938

R² = 0.9505

y = 39.769ln(x) - 67.166

R² = 0.9671

100

200

300

400

500

600

1 100 10000 1000000 100000000

Sensor Signal (nA)

Bacteria concentration (cfu mL-1)

Real samples Commercial samples

Figure 4.

AFM analysis of bare (

), MUDA-coated (

) and antibody-immobilized (

) sensor surfaces.

Salmonella detection was investigated in the concentration range of 1–5.41

cfu mL

−1

which provided a limit of detection (LOD) of 2.7

cfu mL

−1

for both commercial and real samples.

The average signal ratio in the entire concentration range between commercial and real sample analyses

was determined as 1.1. The overall results of Salmonella detection were subjected to the logarithmic

regression analysis that resulted in the correlation coefficients of 0.9671 and 0.9505 for commercial and

real samples, respectively (Figure 5).

Materials 2018, 11, x FOR PEER REVIEW 9 of 17

A) B)

C) D)

Figure 4. AFM analysis of bare (A), MUDA-coated (B) and antibody-immobilized (C,D) sensor

surfaces.

Salmonella detection was investigated in the concentration range of 1–5.41 × 107 cfu mL−1, which

provided a limit of detection (LOD) of 2.7 × 101 cfu mL−1 for both commercial and real samples. The

average signal ratio in the entire concentration range between commercial and real sample analyses

was determined as 1.1. The overall results of Salmonella detection were subjected to the logarithmic

regression analysis that resulted in the correlation coefficients of 0.9671 and 0.9505 for commercial

and real samples, respectively (Figure 5).

Figure 5. Overall results of Salmonella detection in a concentration range of 27—5.41 × 107 using a

standard sandwich assay (n = 6).

y = 37.238ln(x) - 83.938

R² = 0.9505

y = 39.769ln(x) - 67.166

R² = 0.9671

100

200

300

400

500

600

1 100 10000 1000000 100000000

Sensor Signal (nA)

Bacteria concentration (cfu mL-1)

Real samples Commercial samples

Figure 5.

Overall results of Salmonella detection in a concentration range of 27–5.41

using

a standard sandwich assay (n = 6).

Materials 2018,11, 1541 10 of 17

3.1.2. Nanoparticle Enhanced Sandwich Assay for the Analysis of Commercial and Real Samples

The use of nanomaterials significantly enhanced the sensor signal and allows detecting trace

amounts of pathogens that has critical role in diagnostic [

]. To obtain a more sensitive diagnostic assay

for Salmonella detection we, therefore, used the AuNP-modified sandwich assay. The sensor signal

obtained with AuNP conjugated detector antibody is proportional to the amount of bacteria captured

on the surface by the primary antibody. The commercial and real Salmonella samples were investigated

in a concentration range of 1–5.41

cfu mL

−1

(Figure 6A), which resulted in a 1 cfu mL

−1

LOD

(Figure 6A,B). The logarithmic regression analysis of all results revealed a good reproducibility for

the assays with correlation coefficients of 0.9931 and 0.9902 for commercial and real sample analysis,

respectively (Figure 6C). The signal ratio between the two assays was calculated based on the detection

limit responses of two assays and it was found to be 1.098. In the entire investigation range, the average

signal ratio of commercial sample analysis to real sample analysis was found to be 1.03. Therefore,

for the first time in this study, we demonstrated that the sensor can reliably measure Salmonella samples

from different sources with a very high accuracy. Additionally, 50 times higher sensitivity was obtained

in this work when compared to our earlier work [

] and this is most likely due to the new chip design.

Materials 2018, 11, x FOR PEER REVIEW 10 of 17

3.1.2. Nanoparticle Enhanced Sandwich Assay for the Analysis of Commercial and Real Samples

The use of nanomaterials significantly enhanced the sensor signal and allows detecting trace

amounts of pathogens that has critical role in diagnostic [11]. To obtain a more sensitive diagnostic

assay for Salmonella detection we, therefore, used the AuNP-modified sandwich assay. The sensor

signal obtained with AuNP conjugated detector antibody is proportional to the amount of bacteria

captured on the surface by the primary antibody. The commercial and real Salmonella samples were

investigated in a concentration range of 1–5.41 × 107 cfu mL−1 (Figure 6A), which resulted in a 1 cfu

mL−1 LOD (Figure 6A,B). The logarithmic regression analysis of all results revealed a good

reproducibility for the assays with correlation coefficients of 0.9931 and 0.9902 for commercial and

real sample analysis, respectively (Figure 6C). The signal ratio between the two assays was calculated

based on the detection limit responses of two assays and it was found to be 1.098. In the entire

investigation range, the average signal ratio of commercial sample analysis to real sample analysis

was found to be 1.03. Therefore, for the first time in this study, we demonstrated that the sensor can

reliably measure Salmonella samples from different sources with a very high accuracy. Additionally,

50 times higher sensitivity was obtained in this work when compared to our earlier work [18] and

this is most likely due to the new chip design.

(A)

(B)

-4000

-3500

-3000

-2500

-2000

-1500

-1000

-500

500

1000

0 100 200 300 400 500 600

Sensor Signal (nA)

Time (s)

1 cfu mL

-1

5.41×10

cfu mL

-1

Increased S. typhimurium

concentration- CS

-4000

-3500

-3000

-2500

-2000

-1500

-1000

-500

500

1000

0 100 200 300 400 500 600

Sensor Signal (nA)

Time (s)

1 cfu mL

-1

5.41×10

cfu mL

-1

Increased S. typhimurium

concentration- RS

Figure 6. Cont.

Materials 2018,11, 1541 11 of 17

Materials 2018, 11, x FOR PEER REVIEW 11 of 17

(C)

Figure 6. Real-time sensorgrams of commercial (A) and real S. typhimurium (B) detection in a

concentration range of 1–5.41 × 107 cfu mL−1. Overall results of the AuNP-enhanced sandwich assays

for the determination of commercial and real samples (n = 6) (C). CS: commercial sample, RS: real

sample.

Liebana et al. investigated Salmonella detection in milk using an electrochemical magneto-

immunosensor where the bacteria were captured and preconcentrated from milk samples by utilizing

magnetic beads via an immunological interaction [24]. A limit of detection of 7.5 × 103 cfu mL−1 in

milk diluted 1/10 in lysogeny broth was achieved in 50 min without any pretreatment. Another

magneto-immunosensor for Salmonella quantification was reported using nano- and micro-sized

magnetic particles that allowed measuring S. enterica down to 5 × 104 and 1 × 104 cfu mL−1, respectively

[15]. A disposable immunosensor, combining a permanent magnet under the screen-printed carbon

electrode and AuNPs as signal amplification agents, offers an alternative platform for sensitive

detection of Salmonella enterica subsp. enterica serovar Typhimurium LT2 (S) [14]. This sensor was

able to measure the target bacteria in a linear concentration range from 103 to 106 cfu mL−1 with a

detection limit of 143 cfu mL−1. The AuNP enhanced biosensor technique combined with magnetic

field application provided more sensitive detection platform than the conventional commercial

method carried out for comparison purposes. The total sensor assay time using the disposable

immunosensor was 1:30 h per sample [14]. Electrochemical sensing of Salmonella typhimurium in milk

samples was investigated using iron oxide/gold core/shell nanomagnetic probes and CdS biolabels

[21]. The screen-printed carbon electrodes were used as the sensor substrate and the measurements

were carried out by square-wave anodic stripping voltammetry through the use of CdS nanocrystals.

The bacterium was measured between 1 × 101 and 1 × 106 cfu mL−1, and a LOD of 13 cfu mL−1 was

recorded. The determination of Salmonella in milk samples using the biosensor required less than 1 h

[21]. On the other hand, the AuNP modified sandwich assay in our work could detect S. typhimurium

in a concentration range of 1‒5.41 × 107 with LOD of 1 cfu mL−1, which demonstrates much higher

sensitivity than all of the reported works. Additionally, the preparation of our antibody sensor

required 30 min and the detection of a sample was achieved in only 12 min. Furthermore, the same

sensor surface could be used multiple times in this work, avoiding the preparation of the antibody-

immobilized surface all the time. The reliability and accuracy of the sensor are also very high, which

could be demonstrated by studying with different sample sources.

y = 192.02ln(x) + 339.86

R² = 0.9902

y = 196.4ln(x) + 364.84

R² = 0.9931

500

1000

1500

2000

2500

3000

3500

4000

4500

1 100 10000 1000000 100000000

Sensor Signal (nA)

Bacteria Concentration (cfu mL-1)

S. typhimurium samples from human stool

Commercial S. typhimurium samples

Figure 6.

Real-time sensorgrams of commercial (

) and real S. typhimurium (

) detection in

a concentration range of 1–5.41

cfu mL

−1

. Overall results of the AuNP-enhanced sandwich

assays for the determination of commercial and real samples (n = 6) (

). CS: commercial sample, RS:

real sample.

Liebana et al. investigated Salmonella detection in milk using an electrochemical magneto-

immunosensor where the bacteria were captured and preconcentrated from milk samples by utilizing

magnetic beads via an immunological interaction [

]. A limit of detection of 7.5

cfu mL

−1

in milk

diluted 1/10 in lysogeny broth was achieved in 50 min without any pretreatment. Another magneto-

immunosensor for Salmonella quantification was reported using nano- and micro-sized magnetic particles

that allowed measuring S. enterica down to 5

and 1

cfu mL

−1

, respectively [

]. A disposable

immunosensor, combining a permanent magnet under the screen-printed carbon electrode and AuNPs as

signal amplification agents, offers an alternative platform for sensitive detection of Salmonella enterica subsp.

enterica serovar Typhimurium LT2 (S) [

]. This sensor was able to measure the target bacteria in a linear

concentration range from 10

to 10

cfu mL

−1

with a detection limit of 143 cfu mL

−1

. The AuNP enhanced

biosensor technique combined with magnetic field application provided more sensitive detection platform

than the conventional commercial method carried out for comparison purposes. The total sensor assay

time using the disposable immunosensor was 1:30 h per sample [

]. Electrochemical sensing of Salmonella

typhimurium in milk samples was investigated using iron oxide/gold core/shell nanomagnetic probes

and CdS biolabels [

]. The screen-printed carbon electrodes were used as the sensor substrate and

the measurements were carried out by square-wave anodic stripping voltammetry through the use of

CdS nanocrystals. The bacterium was measured between 1

and 1

cfu mL

−1

, and a LOD of

13 cfu mL

−1

was recorded. The determination of Salmonella in milk samples using the biosensor required

less than 1 h [

]. On the other hand, the AuNP modified sandwich assay in our work could detect

S. typhimurium in a concentration range of 1–5.41

with LOD of 1 cfu mL

−1

, which demonstrates

much higher sensitivity than all of the reported works. Additionally, the preparation of our antibody

sensor required 30 min and the detection of a sample was achieved in only 12 min. Furthermore,

the same sensor surface could be used multiple times in this work, avoiding the preparation of the

antibody-immobilized surface all the time. The reliability and accuracy of the sensor are also very high,

which could be demonstrated by studying with different sample sources.

3.1.3. Cross-Reactivity Studies for Salmonella

To determine the specificity of the developed assays for the detection of Salmonella, the interaction

of non-specific bacteria with the Salmonella-specific antibody-immobilized surfaces was studied.

Related document tools

Review originality and document trust

Plag helps identify passages that may need closer source checking. Identific adds another layer when trust and verification matter. They are useful when quality and trust matter.

plag.ai

Materials 2018,11, 1541 12 of 17

S. typhimurium was used as the target bacterium, whereas S. aureus and E. coli were investigated

as the non-specific bacteria. Being a Salmonella serotype the binding of S. enteritidis on the antibody

immobilized surface was also tested. A fixed concentration (10

cfu mL

−1

) of each bacterium was used

for cross-reactivity tests. The non-specific binding responses of S. aureus and E. coli were calculated as

2.66% and 2.01%, respectively, suggesting the high specificity of the bioassay for Salmonella (Figure 7).

The binding of S. enteritidis on the Salmonella antibody immobilized surface was determined as 14.35%.

Nevertheless, the specificity of antibody sensors for bacteria is often a problem, which motivated us to

develop a DNA biosensor in this work as an alternative method.

Materials 2018, 11, x FOR PEER REVIEW 12 of 17

3.1.3. Cross-Reactivity Studies for Salmonella

To determine the specificity of the developed assays for the detection of Salmonella, the

interaction of non-specific bacteria with the Salmonella-specific antibody-immobilized surfaces was

studied. S. typhimurium was used as the target bacterium, whereas S. aureus and E. coli were

investigated as the non-specific bacteria. Being a Salmonella serotype the binding of S. enteritidis on

the antibody immobilized surface was also tested. A fixed concentration (107 cfu mL−1) of each

bacterium was used for cross-reactivity tests. The non-specific binding responses of S. aureus and E.

coli were calculated as 2.66% and 2.01%, respectively, suggesting the high specificity of the bioassay

for Salmonella (Figure 7). The binding of S. enteritidis on the Salmonella antibody immobilized surface

was determined as 14.35%. Nevertheless, the specificity of antibody sensors for bacteria is often a

problem, which motivated us to develop a DNA biosensor in this work as an alternative method.

Figure 7. Cross-reactivity tests for S. typhimurium specific antibody sensor assay (n = 3).

3.2. DNA Sensor for Salmonella Detection

For the first time, we investigated the potential of our custom-designed MiSens device for the

development of a DNA biosensor for pathogenic bacteria detection in this study. The detection of

target Salmonella DNA was studied in a concentration range of 0.002–200 µM using the DNA surface

probe that was initially captured by a neutravidin-immobilized layer on the sensor chip (Figure 8A).

The sensor could differentiate the lowest concentration of DNA (0.002) and negative control (Figure

8B) and showed a good reproducibility in a wide investigation range (Figure 8C). The LOD was

verified based on the linear portion of the saturation graph by calculating three times the standard

deviation of the blank response (0.4 × 3 = 1.2 nA) and extrapolating the sensor signal in the linear

calibration curve to convert the value to concentration [34]. The limit of detection was determined to

be 0.94 nM with a linear range from 2 nM to 20 nM, as shown in Figure 8D.

(A)

400

800

1200

1600

2000

2400

2800

3200

S. typhimurium S. aureus E. coli S. enteritidis

Sensor Signal (nA)

Binding of different bacteria on Salmonella

antibody immobilized sensor surfaces

-250

-200

-150

-100

-50

200 250 300 350 400 450 500 550 600

Sensor Signal (nA)

Time (s)

0.002 µM target DNA

200 µM target DNA

Figure 7. Cross-reactivity tests for S. typhimurium specific antibody sensor assay (n = 3).

3.2. DNA Sensor for Salmonella Detection

For the first time, we investigated the potential of our custom-designed MiSens device for the

development of a DNA biosensor for pathogenic bacteria detection in this study. The detection of

target Salmonella DNA was studied in a concentration range of 0.002–200

M using the DNA surface

probe that was initially captured by a neutravidin-immobilized layer on the sensor chip (Figure 8A).

The sensor could differentiate the lowest concentration of DNA (0.002) and negative control (Figure 8B)

and showed a good reproducibility in a wide investigation range (Figure 8C). The LOD was verified

based on the linear portion of the saturation graph by calculating three times the standard deviation of

the blank response (0.4

3 = 1.2 nA) and extrapolating the sensor signal in the linear calibration curve

to convert the value to concentration [

]. The limit of detection was determined to be 0.94 nM with

a linear range from 2 to 20 nM, as shown in Figure 8D.