Rapid and automated detection of six contaminants in milk using a

Subscriber access provided by ALBRIGHT COLLEGE

Article

Rapid and automated detection of six contaminants in milk using a centrifugal microfluidic platform with two rotation axes Yiqi Chen, Yunzeng Zhu, Minjie Shen, Ying Lu, Jing Cheng, and Youchun Xu Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b01998 • Publication Date (Web): 24 May 2019 Downloaded from http://pubs.acs.org on May 25, 2019

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 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 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.

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 25 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

1

Rapid and automated detection of six contaminants in

2

milk using a centrifugal microfluidic platform with two

3

rotation axes

4 5

Yiqi Chen,†,‡ Yunzeng Zhu,†,‡ Minjie Shen, †,‡ Ying Lu†, Jing Cheng, †,‡,§ and Youchun Xu *,†,‡,§

6 7

†

8 9

‡

State Key Laboratory of Membrane Biology, Department of Biomedical Engineering, School of

Medicine, Tsinghua University, Beijing 100084, China. Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases, Hangzhou

310003, China.

10

§ National

11 12

*Corresponding

13

E-mail address: [email protected] (Y. Xu).

14

ABSTRACT: Antibiotic residues and illegal additives are among the most common contaminants in milk and

15

other dairy products, and they have become essential public health concerns. To ensure the safety of milk,

16

rapid and convenient screening methods are highly desired. Here, we integrated microarray technology into a

17

microfluidic device to achieve rapid, sensitive and fully automated detection of chloramphenicol, tetracyclines,

18

enrofloxacin, cephalexin, sulfonamides and melamine in milk on a centrifugal microfluidic platform with two

19

rotation axes. All the liquid reagent for immunoassay were pre-stored in the reagent chambers of the

20

microdevice and can be released on demand. The whole detection can be automatically accomplished within

21

17 min, and the limits of detection were defined as 0.92, 1.01, 1.83, 1.14, 1.96 and 7.80 μg/kg for

22

chloramphenicol, tetracycline (a typical drug of tetracyclines), enrofloxacin, cephalexin, sulfadiazine (a typical

23

drug of sulfonamides) and melamine, respectively, satisfying the national standards for maximum residue

24

limits in China. Raw milk samples were used to test the performance of the current immunoassay system, and

25

the recovery rates in the repeatability tests ranged from 80% to 111%, showing a good performance. In

26

summary, the immunoassay system established in this study can simultaneously detect six contaminants of

27

four samples in a fully automated, cost effective and easy-to-use manner, and thus has great promise as a

28

screening tool for food safety testing.

29

Corresponding Authors

30

*E-mail: [email protected]

Engineering Research Center for Beijing Biochip Technology, Beijing 102206, China

author at: Department of Biomedical Engineering, Tsinghua University School of Medicine, Beijing 100084, China. Tel.: +86 10 62796071.

1 / 25

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

31

Page 2 of 25

Introduction

32

Food safety has always been an issue of concern, and several recent food safety incidents

33

have stimulated the demand for the screening of the contaminants in milk 1. Among various

34

contaminants in dairy products, antibiotic residues are of particular interest due to their

35

frequent use in animal husbandry for infection treatment and growth promotion 2. Antibiotic

36

residues can cause severe damage to human health, including allergic reactions, drug

37

resistance, and enteric flora disturbance

38

typical illegal additive as well, because it artificially boosts the crude protein percentage.

39

Melamine and cyanuric acid mixtures can form insoluble crystals in the kidneys of babies 5.

40

Due to the severity of health risks caused by these contaminants in milk, many nations and

41

organizations have set strict regulations on residue limits for antibiotic residues and illegal

42

additives in milk. To screen these contaminants, the establishment of rapid, sensitive and

43

high-throughput detecting tools are necessary.

3,4.

In some developing countries, melamine is a

44

Several methods are currently available for the detection of antibiotic residues and

45

melamine, such as microbiological assays, high performance liquid chromatography (HPLC)

46

and immunoassays. Among these methods, microbiological assays are inexpensive but time-

47

consuming and limited in specificity and sensitivity 6. In particular, they cannot be used to

48

detect melamine. Chromatographic techniques are widely used in central laboratories 7-9 due

49

to their high sensitivity and specificity. However, they have some basic requirements,

50

including complex sample preparation, high-cost instrumentation, and skilled personnel.

51

Immunoassay-based methods, for instance, enzyme linked immunosorbent assays (ELISAs),

52

have been successfully applied to detect antibiotic residues and melamine

53

conventional ELISAs are labor-intensive and time-consuming, making them not suitable for 2 / 25


10-12.

But

Page 3 of 25 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


54

high-throughput screening.

55

Compared to conventional ELISAs, protein microarray is an excellent tool for the

56

simultaneous and high-throughput measurement of multiple analytes. In our previous studies

57

13-15, protein microarray had been utilized to screen and quantify the prohibited drugs in serum

58

and urine, as well as the antibiotic residues from animal tissue, and the reliability and accuracy

59

of the system was confirmed by thousands of real samples. In other studies, protein

60

microarrays were also proved as successfully screening tools

61

technology has its disadvantages as well. The operation of conventional microarray-based

62

immunoassay is sophisticated and time-consuming. To overcome these shortcomings, several

63

platforms have been developed 18-20 to simplify manual operation and shorten detection time.

64

Some of them are still semi-automatic 18, while the fully automated platforms that rely on

65

valves and pumps to control the external reagents 19-20 are still bulky and inconvenient to use.

66

Therefore, the development of a rapid, compact, automated and easy-to-use microarray-

67

based immunoassay system is still highly desired.

68

In our latest study

21,

16-17.

However, microarray

a centrifugal microfluidic platform with two rotation axes was

69

constructed, and the basic liquid operations including liquid transport, sequential release, and

70

mixing can be implemented in a simple manner, providing a new strategy for complex liquid

71

manipulation. In the present work, we retrofitted this centrifugal platform and designed a

72

matched microdevice with protein microarray for simultaneous screening and quantification

73

of antibiotic residues (chloramphenicol (CAP), tetracyclines (TCs), enrofloxacin (ENR),

74

cephalexin (CEX), sulfonamides (SAs)) and melamine (MEL) in milk samples. Liquid reagents

75

can be stored in the microdevices and released on demand, which significantly increases the 3 / 25



76

practicality of the platform and making the fully automated immunoassay possible. A compact

77

fluorescence detection system was designed and integrated into the platform for signal

78

recording, and four microdevices can be simultaneously processed, demonstrating its ability

79

for multiplexed target detection of multiple samples. Overall, the centrifugal immunoassay

80

system proposed in this study can successfully detect six contaminants in four milk samples

81

within 17 min, offering a great potential for rapid food safety testing.

82

Experiment

83

Reagents and materials

84

All the antibiotic standards (CAP, TCs, ENR, CEX, and SAs) were purchased from TargetMol

85

(Boston, MA). TCs and SAs are two groups of broad-spectrum antibiotics, while tetracycline

86

(TCN) and sulfadiazine (SDZ) are typical drugs of the TCs and SAs, respectively (Table S1). BSA-

87

SDZ and its antibody (Ab) were used to detect SAs, and BSA-TCN and its Ab were used to detect

88

TCs. The melamine standard was purchased from Dr. Ehrenstorfer GmbH (Augsburg,

89

Germany). Six types of target molecule conjugates (CAP-BSA, TCN-BSA, ENR-BSA, CEX-BSA,

90

SDZ-BSA, and MEL-BSA) and their corresponding murine monoclonal Abs were purchased

91

from the Huaan Magnech Bio-Tech Company Corp. (Beijing, China). BSA conjugates were used

92

as substrates to print the microarray. Stock solutions of the six contaminants were dissolved

93

in methanol or acetonitrile (analytical reagent grade) purchased from Beijing Chemical

94

Reagednts Company (Beijing, China) and stored at -20°C. Cy3-labeled goat anti-Mouse IgG was

95

purchased from Sigma-Aldrich (Shanghai, China), and Cy3-labeled BSA (Cy3-BSA) was

96

purchased from Bioss (Beijing, China). IgG produced in mice was purchased from Genia

97

(Beijing, China). The printing buffer was purchased from CapitalBio Corporation (Beijing, 4 / 25


Page 4 of 25

Page 5 of 25 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


98

China).

99

Glass slides were purchased from Gous Optics (Shanghai, China) and modified with epoxy

100

groups by CapitalBio. Ab solutions were prepared by serial dilution in 0.01 mol/L PBS (pH =

101

7.4), purchased from Solarbio Science & Technology Corp. (Beijing, China). PBST (PBS with 0.05%

102

Tween-20) was used as the washing buffer. Water in the experiments was purified with a Milli-

103

Q system (Millipore, Beijing, China).

104

Instrumentation

105

The immunoassay system was developed based on a centrifugal microfluidic platform

106

with two rotation axes, and the structure of the system is shown in Figure 1. The main stage

107

was driven by a servo motor with its driver (YZ-ACSD608, Yizhi Technology, Shenzhen, China).

108

Four micro-servo motors (RB-15PG, ALSRobot Technology, Harbin, China) were installed on

109

the main stage to precisely control the angular position of each microdevice on demand. Both

110

the power supply and the signal transmission were implemented through an electric slip ring

111

(MOFLON, Shenzhen, China). The pulse signals to control the micro-servo motors were

112

generated by a microcontroller (STC89C52) according to the instructions from the computer.

113

To fix the microdevices, four plastic carriers fabricated by 3D printing (WeNext Corp.,

114

Shenzhen, China) were mounted onto the micro-servo motors. The platform can be controlled

115

by a laptop with a custom-made LabVIEW program (National Instruments, Austin, TX).

116

For fluorescence detection, a green LED (530 nm, LUXEON, Alberta, Canada) with

117

corresponding lenses was used to excite the fluorescent dye. An industrial CCD (MV-BS30U,

118

CREVIS, Yongin-si, Korea), a microscope objective lens (MRH00100, Nikon, Shanghai, China),

119

and an optical filter (BD570, BODIAN Optical, Beijing, China) were used to capture the images.

5 / 25



120

Design and fabrication of the microdevice

121

As shown in Figure 2, the microdevice is composed of four layers, including a pressure

122

sensitive adhesive tape layer (Youbisheng Adhesive Product Corp., Hangzhou, China), a

123

polymethyl methacrylate (PMMA) layer, a double-sided adhesive layer (QL-9970-025, Wuxi

124

Bright Technology, Wuxi, China), and a glass layer. Three metal pins were inserted into the

125

microdevice as part of the three pillar valves.

126

The structure of the PMMA layer is shown in Figure 2A, which was machined by a

127

computerized numerical control (CNC) milling machine (Hongyang Laser Co., Ltd., Beijing,

128

China). A sample chamber and three reagent storage chambers (two chambers for washing

129

buffer and one chamber for secondary antibody solution) are connected to the reaction

130

chamber by microchannels. There is a pillar valve (Figure S1) located between each reagent

131

storage chamber and the reaction chamber. The pillar valve consists of a metal pillar and a

132

through-hole of the PMMA layer. The metal pillar is embedded in the through-hole of PMMA

133

layer, and the top of the connecting chamber is sealed by the PSA layer. The connecting

134

chamber (4 mm in length, 4 mm in width, 0.5 mm in depth) was machined with a spherical

135

cutter to get a concave bottom with rounded corner so that the PSA layer can be deformed to

136

seal it perfectly. The through-hole and the pillar is under interference fit to ensure no liquid

137

leakage in chip using. Once the microdevice is placed on the carrier, the pillar valves will be

138

opened because the metal pins are lifted up by the pillars on the carrier (Figure S1). Thus, the

139

liquid reagent can be released under the centrifugation as needed.

140

Among all of the chambers in the PMMA layer, the reaction chamber is perforative, and

141

a small glass slide is fixed to the reaction chamber from the backside using the double-sided

142

adhesive layer, with all the probes pre-printed and immobilized on the surface. 6 / 25


Page 6 of 25

Page 7 of 25 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

143


Printing and immobilization of the probes

144

A printing robot (Personal ArrayerTM 16, CapitalBio) was used to spot the probes onto the

145

specific area of the slides interfacing with the reaction chamber (Fig, 2A). The volume of each

146

probe was around 0.8 nL. After spotting, the epoxy group-modified glass slides were dried

147

under vacuum at 37℃ overnight. The epoxy groups of the slides would react with the free

148

amino groups of the protein, covalently linking the protein to the glass substrate.

149

Two types of microdevices were contact printed. First, different concentrations of each

150

BSA-antigen (Ag) solution were contact printed on the glass slides with six replicates to

151

optimize the spotting concentrations (Figure S2A). Second, once the spotting concentrations

152

of the Ags were determined, the six BSA-Ags of the optimal concentration and the four control

153

probes (printing buffer and BSA solution as negative controls, Cy3-BSA as a printing control,

154

and mouse-IgG as a positive and hybridization control) were printed on the microdevices for

155

detection (Figure S2B).

156

Sample preparation and recovery experiments

157

Both spiked milk samples and raw milk samples were used in the experiments. Spiked milk

158

samples were used as standards to acquire the standard curves, and raw milk samples were

159

used to test the performance of the system. Commercial milk was purchased from local

160

supermarkets (Beijing, China) to prepare the standard samples, and it was determined to be

161

free of studied contaminants by HPLC-MS.

162

The milk samples were prepared using the following protocol 20: the milk sample (2 mL)

163

was properly mixed with 15 mL of 1% trichloroacetic acid and 3 mL of acetonitrile in a 50-mL

164

polypropylene test tube by vortexing for 1 min to precipitate the proteins in milk. Then, the

165

sample was centrifuged at 7000 × g for 2 min. After the layers were thoroughly separated, the 7 / 25



166

middle layer was taken for subsequent analysis.

167

The microdevice-based immunoassay

168

Before use, the channel and reaction chamber of microdevices were incubated with PBST

169

(30 μL) containing 2% BSA for 1 h to minimize non-specific binding of the Ags to the

170

microdevices. Then the washing buffers and the secondary antibody solution (12 μL each)

171

were added and sealed into the microdevice, which was then stored at 4°C before use.

172

For chip using, the middle layer of the processed milk sample (1 ml) was mixed with Ab

173

solutions (9 ml, six Ab solutions with the optimal concentrations) and loaded to the sample

174

chamber. Then, the microdevice was fixed onto the carrier at the original position (α = 0°).

175

After that, the automated operating procedure for immunoassay was performed as follows

176

(Figure 3, Figure S3): the platform started to rotate, and the mixture in the sample chamber

177

was centrifuged into the reaction chamber to perform the first affinity binding reaction. At

178

this time, the microdevice was driven to tilt around the original position (α1 = -37°, α2 = +15°)

179

continuously at a frequency of 1 Hz. This step was allowed to last for 200 s to accelerate the

180

reaction. Subsequently, the microdevice tilted to the pre-determined angular position (α =

181

+58°), and the waste was driven into the waste chamber. After the first affinity binding step,

182

three more steps followed: the first washing step, the second affinity binding step, and the

183

second washing step. During each step, the corresponding reagents (washing buffer Ⅰ, the

184

secondary antibody solution and washing buffer Ⅱ , respectively) in the chambers were

185

released to the reaction chamber sequentially at specified angular positions (α = -47°, -52°,

186

and -57°), while the microdevice tilted at the same frequency (1 Hz) for 200 s, 200 s and 300

187

s, respectively.

8 / 25


Page 8 of 25

Page 9 of 25 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


188

It was found that the signal intensity can be enhanced by increasing the reaction time.

189

However, too long time for affinity binding will prolong the whole detection time and thus

190

restrict the practicability of the system. Therefore, the reaction time was chosen as 200 s by

191

taking the above factors into consideration.

192

Data acquisition and analysis

193

Once the on-chip immunoassay was completed, the microdevice were illuminated by LED

194

and the images captured by CCD. The differences in the median fluorescence intensity (MFI)

195

between the sample and the blank (MFISample-Blank) were extracted by the LuxScan 3.0 program

196

(CapitalBio).

197 198

Data analysis was performed by using Eq. (1): 𝑀𝐹𝐼𝐼𝐶𝑎 =

100 ― 𝑎 100

× 𝑀𝐹𝐼0 (1)

199

where a (%) is the inhibition ratio of the immunoassay, and MFI0 is the MFI at zero

200

contaminant concentration. With the data extracted from the images, standard curves for the

201

six contaminants were plotted, and the concentration of each contaminant in real samples

202

was determined by referring to these standard curves.

9 / 25



203

Results and discussion

204

The efficiency of the microdevice-based immunoassay

205

No extra pumps or valves were used in the current immunoassay system, and all steps

206

were performed automatically after sample loading. Because the microdevices were tilting

207

continuously during the entire process of the immunoassay, both the reactions and the

208

washing steps were performed in a dynamic manner. As shown in Table S2, with the similar

209

reaction results generated, the current immunoassay system consumes less time for affinity

210

binding and washing compared to the conventional protein chip. With the accelerated

211

reaction efficiency induced by continuously tilting of the microdevice, the entire duration of

212

detection for the immunoassay can be shorten to as less as 17 min.

213

The optimization of the concentrations of Ags and Abs

214

The concentrations of BSA-Ags and monoclonal Abs are closely related to the sensitivity

215

of an immunoassay. On one hand, the hybridization signals should be strong enough so that

216

wide linear ranges of the standard curves can be obtained. On the other hand, the Ag

217

concentrations should be kept as low as possible to lower the cost. As shown in Figure S4,

218

when the concentrations of the Ab remained constant, the fluorescence intensities increased

219

as a function of the printing concentration of BSA-Ags. Similarly, for a specific concentration

220

of BSA-Ag, the fluorescence intensities increased with the concentrations of Ab.

221

Considering both principles collectively, the spotting concentrations of CAP-BSA, TCN-BSA,

222

ENR-BSA, CEX-BSA, SDZ-BSA and MEL-BSA were determined as 0.6, 0.6, 0.2, 0.4, 0.6 and 0.4

223

mg/mL, respectively, and the optimal concentrations of the corresponding Abs were 2.0, 2.0,

224

5.0, 2.0, 1.0 and 2.0 μg/mL, respectively. For the concentration of the secondary antibody

225

solution, the increasing of it facilitates the signal intensity but the background of chip is also 10 / 25


Page 10 of 25

Page 11 of 25 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


226

enhanced. Therefore, an appropriate dilution ratio for the fluorescent labeled secondary

227

antibody was chosen as 1:400.

228

The determination of the specificity

229

The specificity among the BSA-Ags and their Abs was evaluated by Ag-Ab cross-reactivities

230

(CRs). Apart from the corresponding Ag-Ab at 100%, all of the other CRs were calculated using

231

the ratio of MFIs between the two hybridization signals. For example, the CR between MEL-

232

BSA and CAP-Ab was calculated by using Eq. (2): 𝑀𝐹𝐼(𝑀𝐸𝐿 ― 𝐵𝑆𝐴) 𝑎𝑛𝑑 (𝐶𝐴𝑃 ― 𝐴𝑏)

233

CR(%) = 𝑀𝐹𝐼(𝑀𝐸𝐿 ― 𝐵𝑆𝐴) 𝑎𝑛𝑑 (𝑀𝐸𝐿 ― 𝐴𝑏) × 100% (2)

234

As shown in Figure 4, all of the CRs were lower than 5%, suggesting high specificity of the

235

interaction between the six Ags and their corresponding Abs 22.

236

The reactivity of Abs to free antibiotics may be slightly different from the reactivity of Abs

237

to BSA-Abs, and the former method is a common way to evaluate the specificity between the

238

Abs and Ags 22. In our experiments, no serious cross-reactivity was found between different

239

Abs and Ags (both free antibiotics and BSA coupled antibiotics).

240 241

The CRs of the structurally related antibiotics were calculated according to Eq. (3): CR(%) =

𝐼𝐶50 (𝑆𝐷𝑍/𝑇𝐶𝑁) 𝐼𝐶50 (𝑎𝑛𝑎𝑙𝑦𝑡𝑒)

× 100% (3)

242

Cross-reactivity may be a valuable asset for protein microarray because it enables

243

detection of more analytes with limited number of antibodies. As shown in Table S1, the anti-

244

SDZ mAb exhibited high cross-reactivity against the other eight SAs, and the anti-TCN mAb

245

exhibited high cross-reactivity with the other three TCs due to their similar molecular

246

structures. Therefore, this immunoassay system can be used as a screening tool to determine

247

the total amount of SAs and TCs in milk rapidly and the other tools such as HPLC are needed

11 / 25



248

to further distinct the molecules with similar structures.

249

The calibration curves and sensitivity

250

Competitive immunoassays were used for the detection of six contaminants in milk. In

251

principle, the fluorescent signal intensity of corresponding probes decreases when the target

252

substance is present because the target molecules in the samples compete with the same

253

molecules immobilized on the array.

254

Spiked standard samples with a series of different concentrations for each contaminant

255

were prepared and analyzed following the operation protocol to obtain the standard curves.

256

The ratios of MFIs for each standard and the MFI for blank milk were used to generate

257

standard curves (Figure 5A). The limits of detection (LODs) were defined as the concentrations,

258

which were equivalent to 10% inhibition (IC10). The limits of the working range at the high and

259

low concentrations were defined as an inhibition of 80% (IC80) and 20% (IC20), respectively.

260

The performance of the detection, including the IC50, the LODs and the working ranges is listed

261

in Table 1.

262

The results showed that the LODs of this immunoassay system were as low as 0.92, 1.01,

263

1.83, 1.14, 1.96 and 7.80 μg/kg for CAP, TCN, ENR, CEX, SDZ and MEL, respectively, and all of

264

the working ranges were three orders of magnitude. According to the legislation of the China

265

Food and Drug Administration (CFDA), the maximum residue limits (MRLs) of TCs, CEX, SAs

266

and ENR are all 100 μg/kg, falling within the working ranges of our system. Compared to other

267

common methods, our system is rapid, automated and stable, with sufficient detection

268

performance (Table S3).

269

In the current study, the sample preparation is off-chip because some reagents and

12 / 25


Page 12 of 25

Page 13 of 25 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


270

operations is hard to be integrated on the chip. In the future, it is possible to realize “sample-in

271

and answer-out” detection by developing the sample pretreatment that is compatible with

272

on-chip immunoassay, which is the goal of our future work.

273

Analysis of raw milk samples

274

Raw milk samples (n = 35) provided by CapitalBio were used to evaluate the analysis

275

capabilities of the established system. The detection results were compared with data

276

acquired by HPLC-MS method, as listed in Table 2. Even though the raw samples were more

277

complex due to the existence of potential interfering substances, no false positives or false

278

negatives occurred in the experiments. The deviations of detection results between our

279

method and HPLC-MS method were all below 20%, demonstrating that our system can

280

simultaneously detect the six contaminants in raw milk samples and the results are reliable 23-

281

24.

282

microdevices to ensure the consistency of detections. In this immunoassay system, more

283

microdevices can be implanted on demand so that the calibration will be more convenient.

In practical application, the standard curves and calibration are needed for each batch of

13 / 25



284

Conclusion

285

In this study, an automated immunoassay system was constructed based on a centrifugal

286

microfluidic platform with two rotation axes. Simultaneous detection of six contaminants

287

(CAP, TCs, ENR, CEX, SAs and MEL) in milk on a single microdevice was achieved within 17 min,

288

and the system was able to process four samples at a time. The calibration experiments were

289

successfully performed, and the LODs and working ranges of all six contaminants were

290

determined. The applicability was tested with 35 raw milk samples, and reliable results were

291

obtained. In summary, our novel immunoassay system offers a rapid and automated method

292

for contaminants screening and quantification in milk, and this system has the potential to be

293

adapted to detect other proteins or small molecules to further aid food safety testing.

294

ASSOCIATED CONTENT

295

Supporting Information

296

Additional information as noted in text. This material is available free of charge via the Internet at

297

http:// pubs.acs.org.

298

Schematic of the pillar valve; The printing design of the microdevices; Work flow of the

299

microdevice; Optimization of the concentration of each Ag and primary Ab; Cross-Reactivity of

300

tetracyclines and sulfonamides using tetracycline (TCN) and sulfadiazine (SDZ) as standard curves,

301

respectively; Comparison of the reaction efficiency between the current immunoassay system and

302

the conventional protein chip; Comparison of the proposed method with other methods for the

303

detection of milk contaminant (TCN) (PDF).

304

Working video of the immunoassay (mp4).

14 / 25


Page 14 of 25

Page 15 of 25 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


305

AUTHOR INFORMATION

306

Corresponding Authors

307

*E-mail: [email protected]

308

Acknowledgments

309

This study was supported by the National Key Research and Development Program of

310

China (2016YFC0800703), the National Natural Science Foundation of China (31500691), and

311

the Beijing Lab Foundation.

15 / 25



312

Figures

313

Figure 1. Schematic and photograph of the prototype of the immunoassay system (A-B). The

314

size of the prototype is 16 cm × 16 cm × 23 cm.

315 316

16 / 25


Page 16 of 25

Page 17 of 25 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


317

Figure 2. Structure of the microdevice. The size of the microdevice is 30 mm × 22 mm × 4.8

318

mm. (A) Top view of the PMMA layer. (B) Cross-section of the microdevice. It consists of a top

319

layer (pressure sensitive adhesive tape), a PMMA layer, a double-sided adhesive tape, a glass

320

slide with a pre-spotted probe array, and three metal pins (parts of the pillar valves). (C)

321

Photograph of the microdevice.

322 323

17 / 25



324

Figure 3. Work flow of the microdevice. The entire process can be divided into four parts (A-

325

D), each consisting of the following steps (1-4): 1) the microdevice is tilted to the original

326

position (α = 0°); 2) the microdevice is tilted to the specified angular positions (α = 0°, -47°, -

327

52°, -55°), and the reagent or washing buffer is driven into the reaction chamber; 3) the

328

microdevice is repeatedly tilted around the original position (α1 = -37°, α2 = +15°); and 4) the

329

microdevice is tilted (α = 65°), and the buffer is driven into the waste chamber. (A) The process

330

of the first affinity binding step. (B) The process of the first washing step. (C) The process of

331

the second affinity binding step. (D) The process of the second washing step.

332 333 18 / 25


Page 18 of 25

Page 19 of 25 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

334


Figure 4 Cross-reactivities among the six Ags and Abs.

335 336

19 / 25



337

Figure 5 The establishment of calibration curves. (A) Calibration curves of the six contaminants

338

in milk. (B) Representative fluorescence images of microarrays during the calibration

339

experiments with simulated standard samples with CAP. The concentrations of CAP are 1, 2,

340

5, 10, 20, 50, 100 and 200 μg/kg (from left to right and top to bottom).

341 342

20 / 25


Page 20 of 25

Page 21 of 25 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

343


Table 1. Detection performance of the system for the six contaminants (𝑥 ± s, n = 3). Analyte

LOD (μg/kg)

IC50 (μg/kg)

Working range (μg/kg)

MRL (μg/kg)

CAP

0.92

12.78

1.78-127.58

-

TCN

1.01

14.46

1.97-148.05

100

ENR

1.83

20.34

3.34-167.28

100

CEX

1.14

12.55

2.08-102.27

100

SDZ

1.96

23.80

3.66-211.39

100

MEL

7.80

266.96

18.86-4876.70

-

344

MRL stands for the maximum residue limit according to the legislations of CFDA;

345

“-” represents not allowed to exist.

21 / 25



346

Table 2. The analytical results of 35 raw milk samples. Sample No. S2 S5 S7 S11 S13 S14 S16 S20 S21 S28 S32 S35 Others

347

Page 22 of 25

Analytes CAP CEX MEL CAP TCN SDZ ENR CAP TCN ENR SDZ CEX

HPLC-MS (μg/kg)

Our system (μg/kg)

113.1 83.1 308.1 48.6 18.2 159.7 246.4 109.7 143.6 15.8 40.7 21.6

99.9 78.7 329.7 45.2 20.1 138.1 197.6 102.3 129.8 14.2 45.1 19.4 ND

Each was analyzed with 3 repeats; ND means not detectable.

22 / 25


Recovery rate (%) 88.3 94.7 107.0 93.0 110.4 86.5 80.2 93.2 90.4 89.9 110.8 89.8

Page 23 of 25 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


348

References:

349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390

[1] Song, Y.; Yu, H.; Lv, W. Risk analysis of dairy safety incidents in China. Food Control 2018, 92, 63-71. [2] Boeckel, T. P. V.; Brower, C.; Gilbert, M.; Grentfell, B.T.; Levin, S. A.; Robinson, T. P.; Teillant, A.; Laxminarayan, R. Global trends in antimicrobial use in food animals. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 5649-5654. [3] Fonseca, G. P.; Cruz, A. G.; Faria, J. A. F.; Silva, R.; Moura, M. R. L.; Carvalho, L. M. J. Antibiotic residues in Brazilian UHT milk: a screening study. Ciênc. Tecnol. Aliment. 2009, 29, 451-453. [4] Marshall, B. M.; Levy, S. B. Food Animals and antimicrobials: impacts on human health. Clin. Microbiol. Rev. 2011, 24, 718-733. [5] Hau, K. C.; Kwan, T. H.; Li, K. T. J. Melamine toxicity and the kidney. Am. Soc. Nephrol. 2009, 20, 245-250. [6] Nouws, J. F.; Van, E. H.; Loeffen, G. Schouten, J.; Keukens, H.; Smulders, I.; Stegeman, H. Suitability of the charm HVS and a microbiological multiplate system for detection of residues in raw milk at EU maximum residue levels. Vet. Q. 1999, 21, 21-27. [7] Karageorgou, E.; Armeni, M.; Moschou, I.; Samanidou, V. Ultrasound-assisted dispersive extraction for the high pressure liquid chromatographic determination of tetracyclines residues in milk with diode array detection. Food Chem. 2014, 150, 328-334. [8] Filigenzi, M. S.; Tor, E. R.; Poppenga, R. H.; Aston, L. A.; Puschner, B. The determination of melamine in muscle tissue by liquid chromatography/tandem mass spectrometry. Rapid Commun. Mass Spectrom. 2007, 21, 4027-4032. [9] Mahony, J. O.; Clarke, L.; Kennedy, R. O.; Lehotary, S. J.; Danaher, M. The use of ultra-high pressure liquid chromatography with tandem mass spectrometric detection in the analysis of agrochemical residues and mycotoxins in food - Challenges and applications. J. Chromatogr. A. 2013, 1292, 83-95. [10] Song, E.; Yu, M.; Wang, Y. HU, W.; Cheng, D.; Swihart, M. T.; Song, Y. Multi-color quantum dotbased fluorescence immunoassay array for simultaneous visual detection of multiple antibiotic residues in milk. Biosens. Bioelectron. 2015, 72, 320-325. [11] Ho, T. Y.; Chan, C. C.; Chan, K.; Wang, Y. C.; Lin, J. T.; Chang, C. M.; Chen, C. S. Development of a novel bead-based 96-well filtration plate competitive immunoassay for the detection of Gentamycin. Biosens. Bioelectron. 2013, 49, 126-132. [12] Sun, F.; Liu, L.; Xu, H. K. C. Development of ELISA for melamine detection in milk powder. Food Agric. Immunol. 2013, 24, 79-86. [13] Du, H.; Yang, W.; Xing, W. Su, Y.; Cheng, J. Parallel detection and quantification using nine immunoassays in a protein microarray for drug from serum samples. Biomed. Microdevices 2005, 7, 143-146. [14] Du, H.; Wu, M.; Yuan, G. Sun, Y.; Lu, Y.; Zhao, S.; Du, Q.; Wang, J.; Yang, S.; Pan, M.; Lu, Y.; Wang, S.; Cheng, J. Development of miniaturized competitive immunoassays on a protein chip as a screening tool for drugs. Clin. Chem. 2005, 51, 368-375. [15] Chen, A.; Wang, G.; Cao, Q.; Wang, Y.; Zhang, Z.; Sun, Y.; Wang, H.; Xu, C.; Zhou, Q.; Mitchelson, K. R.; Cheng, J. Development of an antibody hapten-chip system for detecting the residues of multiple antibiotic drugs. J. Forensic Sci. 2009, 54, 953-960. [16] Zuo, P.; Ye, B. C. Small molecule microarrays for drug residue detection in foodstuffs. J. Agric. 23 / 25



391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412

Food Chem. 2006, 54, 6978-6983. [17] Zuo, P.; Zhang, Y.; Liu, J. Ye, B. C. Determination of β-adrenergic agonists by hapten microarray. Talanta 2010, 82, 61-66. [18] Noroozi, Z.; Kido, H.; Peytavi, R.; Nakajima-Sasaki, R.; Jasinskas, A.; Micic, M.; Felgner, P. L.; Madou, M. J. A multiplexed immunoassay system based upon reciprocating centrifugal microfluidics. Rev. Sci. Instrum. 2011, 82, 64303. [19] Knecht, B. G.; Strasser, A.; Dietrich, R.; Märtlbauer, E.; Niessner, R.; Weller, M. G. Automated microarray system for the simultaneous detection of antibiotics in milk. Anal. Chem. 2004, 76, 646654. [20] Guo, H.; Zhou, X.; Zhang, Y.; Song, B.; Zhang, J.; Shi, H. Highly sensitive and simultaneous detection of melamine and aflatoxin M1 in milk products by multiplexed planar waveguide fluorescence immunosensor (MPWFI). Food Chem. 2016, 197(Pt A), 359-366. [21] Zhu, Y.; Chen, Y.; Meng, X. Wang, J.; Lu, Y.; Cheng, J. Comprehensive study of the flow control strategy in a wirelessly charged centrifugal microfluidic platform with two rotation axes. Anal. Chem. 2017, 89, 9315-9321. [22] Wang, Y.; Liu, N.; Ning, B. Simultaneous and rapid detection of six different mycotoxins using an immunochip. Biosens. Bioelectron. 2012, 34, 44-45. [23] Karageorgou, E.; Armeni, M.; Moschou, I. Ultrasound-assisted dispersive extraction for the high pressure liquid chromatographic determination of tetracyclines residues in milk with diode array detection. Food Chem. 2014, 150, 328-334. [24] Wang, G.; Zhang, H.; Liu, J. A receptor-based chemiluminescence enzyme linked immunosorbent assay for determination of tetracyclines in milk. Anal. Biochem. 2019, 564, 40-46.

24 / 25


Page 24 of 25

Page 25 of 25 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

413


For Table of Contents Only

414

25 / 25


Rapid and automated detection of six contaminants in milk using a

Recommend Documents