Enhancing the Sensitivity of Lateral Flow ... - ACS Publications

Subscriber access provided by Washington University | Libraries

Article

Enhancing the sensitivity of lateral flow immunoassay by centrifugation-assisted flow control Minjie Shen, Yiqi Chen, Yunzeng Zhu, Mangsuo Zhao, and Youchun Xu Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b00421 • Publication Date (Web): 05 Mar 2019 Downloaded from http://pubs.acs.org on March 6, 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 28 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

Enhancing the sensitivity of lateral flow immunoassay by

2

centrifugation-assisted flow control

3

Minjie Shen, a Yiqi Chen, a Yunzeng Zhu, a Mangsuo Zhao, b and Youchun Xu a, c*

4

a

5

China;

6

b

Tsinghua University Yuquan Hospital, Beijing 100049, China;

7

c

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

8

*Correspondence should be addressed to Y.X. ([email protected]).

9

Tel: (86)-10-62796071.

Department of Biomedical Engineering, Tsinghua University School of Medicine, Beijing 100084,

1 / 28

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

10

Abstract

11

Lateral flow immunoassay (LFIA) is widely used but is limited by its sensitivity. In this

12

study, a novel centrifugation-assisted lateral flow immunoassay (CLFIA) was proposed that

13

had enhanced sensitivity compared to traditional LFIA based on test strips. For CLFIA, a

14

vaulted piece of nitrocellulose membrane was prepared and inserted into a centrifugal disc.

15

Powered by the centrifugal force, the sample volume on the disc was not limited and the flow

16

rate of the reaction fluid was steady and adjustable at different rotation speeds. It was found

17

that lower rotation speeds and larger sample volumes resulted in greater signal intensity in the

18

nitrocellulose membrane as well as higher sensitivity, indicating that the actively controlled

19

flow on the disc allowed for sensitivity enhancement of CLFIA. To operate CLFIA on the

20

centrifugal disc, a portable and cost-effective operating device was constructed to rotate the

21

disc with a stepper motor and collect the results with a smartphone. The proposed method was

22

successfully applied to detect prostate specific antigen (PSA) in human serum. Standard curves

23

were established for CLFIA and LFIA, and both had correlation coefficients of up to 0.99.

24

Under optimal conditions (1500 rpm rotation speed, 120 μL sample volume), the detection limit

25

of CLFIA reached 0.067 ng/mL, showing a 6.2-fold improvement in sensitivity compared to

26

that of LFIA. With clinical serum samples, a good correlation was observed between PSA

27

concentrations measured by CLFIA and by a bulky commercial instrument in hospital. In

28

summary, this portable, cost effective, and easy-to-use system holds great promise for

29

biomarker detection with enhanced sensitivity compared to traditional LFIA.

2 / 28


Page 2 of 28

Page 3 of 28 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


30

3 / 28



31

Page 4 of 28

Introduction

32

Lateral flow immunoassay (LFIA) is one of the most popular methods for point-of-care

33

testing (POCT).1,2 The vast majority of LFIA is performed in the form of lateral flow strips

34

(LFS). For decades, LFS have been manufactured and widely applied to healthcare, food safety,

35

and environmental monitoring. Specific applications include pregnancy tests, infectious

36

diseases,3-6 inflammation,7,8 disease biomarkers,9-12 foodborne pathogens,13,14 toxins,15-17

37

pesticides,18 and drug residues.19,20 LFS have gained a great deal of interest and profit due to

38

the major advantages including their ease of use, rapid response, low cost, and long shelf life.

39

However, the analytical performance of traditional LFIA is still limited by its sensitivity.

40

First, to overcome this limitation, alternative materials have been used in place of traditional

41

gold nanoparticles (AuNPs) or to apply signal amplification. For instance, carbon

42

nanoparticles/nanotubes,21,22

43

materials,10 and magnetic nanoparticles26,27 have been reported as labels resulting in improved

44

LFIA sensitivity. In addition, silver and gold enhancement methods28 and enzyme/nanozyme-

45

based colorimetric amplification methods29,30 have also been exploited in LFIA. Although these

46

novel materials provide new approaches for highly sensitive detection, additional instruments

47

are more or less required for quantitative measurements due to specific optical, magnetic or

48

electrochemical properties of these materials. Besides, for the signal amplification methods,

49

additional assays will be needed which might otherwise make the corresponding instruments

50

for automated detection cumbersome. Second, in addition to advancements in signal

51

presentation, other characteristics of LFIA related to sensitivity have also been explored. Parolo

52

et al. altered the architecture of test strips, achieving a significant improvement in sensitivity.31

fluorescent

materials,23

quantum

4 / 28


dots,24,25

upconverting

Page 5 of 28 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


53

Rivas et al. constructed wax pillar modified strips, which caused microfluidic delay and

54

generated pseudoturbulence on the detection membrane, resulting in better performance.32

55

These improvements indicate another strategy to enhance LFIA’s sensitivity by changing the

56

sample flow on the strip. However, the change of the flow in these studies31,32 is not actively

57

adjustable. Third, the performance of LFIA is also constrained by intrinsic defects. A lateral

58

flow test strip basically consists of a sample pad, a conjugate pad, a nitrocellulose (NC)

59

membrane, and an absorbent pad. The immobilization of reaction reagents on a strip is

60

restrained by the capacity of these compartments as well as the sample loading volume.

61

Moreover, liquid propulsion by capillary forces cannot be precisely controlled, which may

62

generate different liquid flow rates owing to variations in sample viscosity and surface

63

tension.33 Integrating the LFIA into centrifugal microfluidic platforms34 with active flow

64

control may address the aforementioned issues.

65

Over the past decade, we have been continuously working on the development of

66

microfluidic devices for biomedical-related applications,35,36 and recently dedicated in

67

innovation and development of centrifugal microfluidic platforms.37,38 Unlike using LFS as a

68

unit for final detection on microfluidic platforms,14,39 we herein present a controlled LFIA in

69

centrifugal discs, named centrifugation-assisted lateral flow immunoassay (CLFIA), for the

70

first time. In CLFIA, a vaulted piece of NC membrane, with a test and a control line, was

71

prepared and inserted into a centrifugal microfluidic disc. The AuNP-labeled antibody was

72

lyophilized and preloaded into the disc. Different volumes of sample can be loaded into the disc

73

in CLFIA without the limitation accompanied by the capacity of the strip in LFIA. The liquid

74

was driven through the NC membrane by the cooperation of the centrifugal and capillary forces. 5 / 28



75

Meanwhile, the liquid flow rate was much more stable than that of LFIA and could be adjusted

76

under different rotation speeds of the disc. Combining these merits, the improved sensitivity of

77

CLFIA was demonstrated compared to the traditional LFIA. To support the operation of this

78

centrifugal disc, a portable device was constructed to rotate the disc and capture the results

79

using a smartphone. To further assess our method, human prostate specific antigen (PSA) was

80

chosen as the target to validate the improved performance of our method.

6 / 28


Page 6 of 28

Page 7 of 28 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


81

Experimental section

82

Preparation of AuNP-labeled antibody

83

The AuNP-labeled antibody was prepared according to the protocol described previously.40

84

Briefly, incubation buffer was prepared by adding 3.09 g boric acid (Sinopharm Chemical

85

Reagent, Shanghai, China) into 500 mL ultra-pure water from the Milli-Q water purification

86

system (Millipore, Beijing, China), and the pH was adjusted to 8.0. Then, a volume of 1.5 mL

87

gold solution (20 nm, OD520 = 0.10, Beijing Kwinbon Biotechnology, Beijing, China) was

88

centrifuged at 10000 × g for 20 min and the buffer was exchanged with 0.8 mL incubation

89

buffer. An appropriate amount of conjugate antibody (10-3143, Fitzgerald, MA, USA) was

90

added into the gold solution and mixed using a roller mixer at 25°C. After incubation for 1 h,

91

incubation buffer (200 μL) containing 10% bovine serum albumin (BSA, Sigma-Aldrich,

92

Shanghai, China) was added into the gold solution and incubated for another 20 min to block

93

any unconjugated sites on the AuNPs. The solution was centrifuged at 10000 × g for 20 min at

94

4°C and the supernatant was discarded. The pellet was resuspended in 0.5 mL storage buffer

95

(incubation buffer containing 1% BSA) to eliminate any unconjugated antibody. Subsequently,

96

the solution was centrifuged again (10000 × g, 20 min, 4°C). The pellet was finally resuspended

97

in 200 μL storage buffer and stored at 4°C until use. After adding 10% trehalose (Sigma-Aldrich,

98

Shanghai, China), a portion of the final solution was lyophilized into small globules for use on

99

the disc.

100

Preparation of LFS

101

The NC membrane (HF13502S25, Millipore, Beijing, China), sample pad, and absorbent

102

pad (Beijing Kwinbon Biotechnology, Beijing, China) were incorporated onto a plastic backing 7 / 28



103

card successively. Capture antibody (2 mg/mL, 10-3142, Fitzgerald, MA, USA) and anti-mouse

104

IgG (1 mg/mL, ab7063, Abcam, Shanghai, China) were dispensed onto the NC membrane as

105

the test (T) and control (C) lines, respectively, using a dispenser (XYZ3050, BioDot, Shanghai,

106

China). The dispensing rate was 0.5 μL/cm. The AuNP-labeled antibody solution was dispensed

107

onto the conjugate pad using an IsoFlow dispenser (Imagene Technology, NH, USA) at the rate

108

of 16 μL/cm. The fabricated pad was then dried at 37°C for 10 h, cut into 3 mm wide strips,

109

and stored at room temperature in a desiccator.

110

Fabrication of CLFIA discs

111

The CLFIA disc was made of polymethyl methacrylate (PMMA) and engraved by

112

Hongyang Laser Co. Ltd. (Beijing, China). The lyophilized AuNP-labeled antibody was loaded

113

into the loading chamber on the disc. The lyophilizing process is only to facilitate the long-term

114

storage and easy use of the disc, and users can also directly load the buffer with AuNPs-labeled

115

antibody for disc using. The NC membrane for the disc was prepared using the same procedure

116

as described for LFS, except that the NC membrane was cut into the designed pattern (Fig. 1a)

117

using a cutting plotter (FC4500-50, Graphtec, Yokohama, Japan) with a width of 3 mm. The

118

NC membranes cut for LFS and CLFIA disc have similar total area (about 70 mm2). The back

119

of the patterned NC compartment was attached to an adhesive tape (Youbisheng Adhesive

120

Products, Hangzhou, China) and the tape was then stuck to the PMMA disc to seal the chambers

121

and channels. The NC compartment was aligned to the NC chamber when attaching the tape to

122

the disc.

123

Design of the operating device for CLFIA

124

The shell of the device was fabricated by 3D printing (WeNext Corp., Shenzhen, China). A 8 / 28


Page 8 of 28

Page 9 of 28 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


125

stepper motor (YZ-ACSD608, Yizhi Technology, Shenzhen, China) was attached to the bottom

126

of the device. A 3D printed pallet, on which the disc was set, was fixed to the spindle of the

127

motor. A photogate (Panasonic, Shanghai, China) and the positioning hole of the pallet were

128

used to stop the disc at the specific angle for final signal acquisition. For illumination, two flat

129

LED light sources (Baohui optoelectronics, Shenzhen, China) were attached to the

130

photographic chamber. A smartphone (Mi 4, Xiaomi, Beijing, China) was positioned over the

131

apparatus to photograph the results. The schematic of the device is shown in Fig. S1. The device

132

was powered by an external power supply and controlled by a computer.

133

Performance evaluation of liquid manipulation

134

The working buffer was 10 mM phosphate buffer (PBS, pH 7.4, Solarbio, Beijing, China)

135

containing 0.2% (V/V) Tween-20 (Sinopharm Chemical Reagent Co. Ltd., Shanghai, China).

136

First, the performance of liquid migration on LFS and on CLFIA discs was evaluated. For LFS

137

evaluation, the working buffer (120 μL) was transferred into a 1.5 mL centrifuge tube. Then,

138

the strip was added to the tube and the remaining volume of buffer was quantified at different

139

incubation times. For disc evaluation, the working buffer (120 μL) was transferred into the

140

loading chamber and centrifuged at a defined rotation speed (2000 rpm). Images were taken at

141

the different time points, and the remaining volume of buffer in the reaction chamber was

142

calculated by analyzing each picture (Fig. S2). Second, the flow rates of fluid through the NC

143

membrane on the disc were also measured at different rotation speeds. The same amount of

144

working buffer (60 μL) was loaded and centrifuged at 1250, 1500, 2000, 2500, or 2750 rpm.

145

The time required for all of the liquid to transfer from the reaction chamber to the waste

146

chamber at each rotation speed was recorded. 9 / 28



147

Immunoassay

148

PSA standard solutions at different concentrations were prepared by diluting a PSA stock

149

solution (30C-CP1017U, Fitzgerald, MA, USA) in the working buffer for performance

150

evaluation. First, with a fixed rotation speed, signals obtained with an increasing volume of

151

solution were measured on both LFS and CLFIA discs. Second, with a specified volume of

152

PSA standard solution, signals at different rotation speeds were measured on CLFIA discs.

153

Third, at a fixed reaction time, the signal intensity at different rotation speeds with a sufficient

154

volume of PSA standard solution were measured on CLFIA discs. For serum tests, PSA spiked

155

serum samples and clinical serum samples were analyzed on both platforms. Tests on LFS were

156

completed in 15 min. Tests on discs were performed at corresponding optimal conditions. All

157

experiments were performed at room temperature. Data were collected using a smartphone and

158

quantitative analysis was performed on a computer using the Image J software.

159

The study was approved by the Institutional Review Board (IRB) of Tsinghua University

160

of China (No. 20180022). All serum samples were provided by donors following IRB

161

guidelines.

10 / 28


Page 10 of 28

Page 11 of 28 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


162

Results and discussion

163

Principle of the CLFIA system

164

The details of the patterned NC membrane and the fabrication of a disc are shown in Fig.

165

1a-b and Fig. S3. For disc use, the sample is injected into the loading chamber (Fig. 1c). The

166

lyophilized AuNP-labeled antibody will be instantly dissolved. Then, the disc is rotated at a

167

defined speed. The NC membrane for CLFIA has a bridge-like shape and guides the liquid from

168

the upper chamber to the lower. After the reaction liquid transfers to the reaction chamber and

169

contacts the membrane, the liquid on the left of the NC bridge will be dragged by centrifugal

170

and capillary forces, which are in opposite directions. At rotation speeds below 3000 rpm, the

171

capillary force prevails, and the liquid will move forward along the NC bridge. The vaulted

172

structure of the NC bridge can lead the liquid to the right where the liquid will be continuously

173

impelled from the membrane to the waste chamber, forming a constant capillary force for the

174

left hand side of the NC bridge. Therefore, the sample is continuously transferred through the

175

NC bridge and the immunoassay steadily proceeds. For the immunoassay to detect PSA, the

176

antigen in the sample conjugates to the AuNP-labeled antibody and is then captured by another

177

antibody on the test line. The redundant AuNP-labeled antibody not bound to antigen will move

178

forward and be captured by the antibody on the control line. Thus, color signals can be

179

generated on the test and control lines. The gray intensity of the test line is obtained for

180

quantitative analysis. The schematic of LFIA was shown in Fig. S4. A brief video showing the

181

detection process is available in the Supporting Information.

182

The vaulted NC membrane in CLFIA system is a deliberate design inspired by the structure

183

of the siphon valve wildly used on centrifugal platform.41 With this bridge-like shape, the liquid 11 / 28



184

transferring from upstream to downstream chambers is steady and adjustable. Additionally, the

185

angle between the two hands of the NC membrane was definite. The two hands were located in

186

a radial pattern so that it is coincident with the direction of centrifugation force, facilitating the

187

uniformity of the lateral flow on the membrane.

188

The CLFIA system provides different approaches for the enhancement of LFIA, which are

189

mainly manifested in two aspects. First, sufficient time period is needed for the formation of

190

immune-complex at the test and control lines to ensure firm binding of the reagents at molecular

191

level. The flow rate is a determining factor for this time period. Second, with a certain flow

192

rate, sample volume determines the whole reaction time at the test and control lines. Longer

193

reaction time results in more immune-complexes at the reaction lines until saturate. Therefore,

194

the control of the flow rate and the adjustment of the sample volume may aid to improve the

195

sensitivity of LFIA.

196

Performance of liquid manipulation

197

The liquid on LFS is driven passively by the capillary force generated by the absorbent pad.

198

Once the process starts, the liquid movement is determined by the strip itself. Conversely, the

199

liquid on the NC bridge that inserted into the CLFIA disc is actively controlled by the

200

centrifugal and capillary forces acting on it. As shown in Fig. 2a, the liquid volume on LFS

201

increases rapidly at the beginning and then gradually a slow increase is observed until the liquid

202

volume reaches the maximum the strip can contain. The liquid on LFS at the beginning is

203

absorbed by the sample pad and conjugate pad. Therefore, the effective volume through the NC

204

membrane on LFS is less than 40 μL during the entire process. Whereas, the liquid volume

205

passing through the NC bridge in the CLFIA disc steadily increases over time without such 12 / 28


Page 12 of 28

Page 13 of 28 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


206

limitation. In addition, the slope at each point on the curves in Fig. 2a indicates that the liquid

207

flow rate on LFS decreased during strip tests while the flow rate through the NC bridge was

208

comparatively consistent under a constant rotation speed. In addition to consistent control, the

209

liquid flow rate can also be adjusted by altering the rotation speed of the CLFIA disc. As shown

210

in Fig. 2b, the flow rate on the CLFIA disc increased as the rotation speed increased from 1250

211

rpm to 2500 rpm. When the speed was higher than 2500 rpm, the flow rate decreases. Here, the

212

adjustable range of rotation speed is constrained between 1250 rpm to 2750 rpm to ensure the

213

lateral flow works on the disc. When the rotation speed is below than 1250 rpm, the liquid at

214

the right hand side of the NC bridge is retained by the membrane because the capillary force is

215

higher than the centrifugal force that tends to drive the liquid out of the NC bridge. When the

216

rotation speed is higher than 2750 rpm, the centrifugal force can impede the liquid at the left

217

hand side of the NC bridge to flow upward. These results show the overwhelming fluid control

218

capability of the CLFIA system over that of LFS. It demonstrates that the sample volume and

219

liquid flow rate through NC membrane are no longer non-adjustable on CLFIA discs.

220

Performance evaluation of CLFIA at adjustable conditions

221

There are plenty of factors that can affect the performance of CLFIA, which can be

222

generally divided into two groups determined by the biochemical characters of reaction system

223

and the spinning program, respectively. The first group includes the composition and viscosity

224

of the sample, the quality of AuNPs, and the specificity of the antibodies. The second group

225

includes the rotation speed and the reaction time period. Since the first group of factors is

226

determined by the characteristics of the testing object and the commercial reagents, there is

227

little possibility for us to make comprehensive optimization. Only the size, modification and 13 / 28



228

dosage of AuNPs were examined. For the size of AuNPs, it is reported that AuNPs with

229

comparatively large diameter benefit to produce stronger signals42,43 while they are more likely

230

to sediment under centrifugation. Therefore, AuNPs at diameter of 20 nm were chosen to

231

balance these two effects. For the modification of AuNPs, different dosages of antibody were

232

added for AuNPs labeling as shown in Fig. S5, and the result showed that 5 μg antibody was

233

sufficient for labeling. In addition, the dosage of AuNP-labeled antibody used for LFIA was

234

also optimized (Fig. S6), allowing the calculation of the amount of AuNP-labeled antibody

235

deposited on the LFS conjugate pad and CLFIA disc.

236

As detailed in Fig. 2a-b, liquid volume and flow rates on CLFIA are adjustable, therefore,

237

we mainly focus on the second group of factors to improve the sensitively of immunoassay.

238

Fig. 3a shows the signal intensity as a function of sample volume in LFIA and CLFIA. As the

239

sample volume increased, signal of LFIA increased until it reached the maximum determined

240

by the capacity of the strip. Signal of CLFIA increased along with the sample volume. At the

241

same sample volume, the signal of CLFIA was much stronger than that of LFIA, because the

242

sample could be fully utilized in CLFIA but the sample on the strips might remain in the

243

sample pad or conjugate pad or evaporate. Another reason for poor signal of LFIA is that the

244

majority of the AuNP-labeled antibody is flushed out at the onset of testing, causing the

245

inefficient reaction of the remaining sample. As for performance evaluation at different flow

246

rates, signal intensity using equal volumes of sample at different rotation speeds was measured

247

(Fig. 3b). As the rotation speed increased from 1500 rpm to 2500 rpm, the flow rate increased

248

and the signal intensity decreased, indicating that increased flow rate reduced the reaction

249

efficiency between the antibodies and antigen. This is because the exposure time is reduced at 14 / 28


Page 14 of 28

Page 15 of 28 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


250

higher flow rates. However, the further reducing of the rotation speed (1250 rpm) may not be

251

helpful for signal enhancement because the molecules can get adequate time to contact and

252

bind to each other when the flow rate is low enough. As shown in Fig. 3b, the signals have no

253

significant difference when the rotation speed is below than 1500 rpm. At the rotation speed

254

of 2750 rpm, the AuNP-labeled antibody starts to precipitate, which affects the signal. Once

255

the reaction time of CLFIA is set, the signal will be a compromise between the sample volume

256

and flow rate at one rotation speed. Fig 3c shows that 2000 rpm is a turning point with

257

comparatively high signal intensity for the same reaction time as LFIA. To sum up, these

258

results indicate that CLFIA allows much greater flexibility in adjusting sample volume and

259

flow rate, which will be important parameters for CLFIA system to optimize the detection

260

signal.

261

Since the sensitivity and the operating period are two factors that should take into

262

consideration but have different importance in different applications, two optimal conditions

263

can be proposed for quick tests with acceptable sensitivity and sensitive tests with acceptable

264

duration, respectively. For quick tests (15 min), the optimal rotation speed is 2000 rpm, as

265

shown in Fig. 3c. The consumed sample in this condition (condition 1) is about 60 μL in

266

volume. For sensitive tests with acceptable duration (45 min), lower rotation speed (1500 rpm)

267

is facilitate to obtain stronger signal, as shown in Fig. 3b. The consumed sample in this

268

condition (condition 2) is about 120 μL in volume.

269

Performance validation of CLFIA using PSA spiked serum samples

270

Serum samples from five healthy females were used as control samples and to prepare PSA

271

spiked serum samples, as the PSA concentration in healthy female serum is far below the 15 / 28



272

detection capability of LFIA. It was further confirmed because the detected results of control

273

samples in CLFIA system remained low and stable without any nonspecific signal at increasing

274

sample volumes (Fig. S7). PSA spiked serum samples in the concentration range of 0-10 ng/mL

275

were analyzed both on LFS and CLFIA discs with optimized condition 1 and 2. Each sample

276

was diluted 4 times with the working buffer before each assay was performed. Despite the

277

optimization factors of the reaction system mostly determined by the characteristics of the

278

sample and commercial reagents, condition 2 is the most optimal among all the conditions

279

within adjustable range unless using a new disc with larger sample capacity.

280

Visualized with the naked eye, the color intensity of the T line increased as the PSA

281

concentration increased. The PSA concentrations that could be explicitly distinguished by LFIA,

282

CLFIA condition 1, and CLFIA condition 2 were 5 ng/mL, 1 ng/mL, and 1 ng/mL, respectively

283

(Fig. 4a-c). Using a smartphone to capture images of LFS and CLFIA discs, quantitative

284

analysis was performed using Image J. Standard curves are shown in Fig. 4d-f and the curve-

285

fitting equations and logistic correlation coefficients (R2) are also shown in the figures. The

286

limit of detections (LODs) of LFIA, CLFIA condition 1, and CLFIA condition 2 reached 0.41

287

ng/mL, 0.13 ng/mL, and 0.067 ng/mL, respectively. The R2 values were all above 0.99,

288

indicating a good correlation between the concentration of PSA and ΔGray intensity. The

289

reproducibility of the measurements for PSA in serum by different methods was also studied

290

and the relative standard deviations (RSD) were shown in Table 1. Lower RSD values at each

291

PSA concentration reflected better reproducibility of CLFIA than that of LFIA. We further

292

evaluated the specificity of our system by testing three proteins (alpha fetoprotein,

293

carcinoembryonic antigen, human chorionic gonadotropin) and a good specificity was obtained 16 / 28


Page 16 of 28

Page 17 of 28 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


294

as shown in Fig. S8. Besides, it is possible to insert more units on a disc for repeatable tests or

295

multiple detections, possessing the potential for better efficiency and reproducibility.

296

The results clearly demonstrate that CLFIA has the capability to detect PSA at lower

297

concentrations than those detected by LFIA. In addition, CLFIA has the potential to detect even

298

lower concentrations of the target antigen using larger sample volumes. Compared to LFIA,

299

CLFIAs at the two tested conditions had better sensitivity, with an LOD improvement of up to

300

3.2-fold and 6.2-fold, respectively. Moreover, our method can be adjusted according to the

301

operation conditions, which can greatly increase its applicability.

302

Compared to the previous PSA detection methods (Table S1), our method utilizes the most

303

basic materials with an affordable cost and has a decent performance at the same time. In

304

addition, our method has inherent advantages for POCT applications due to its high automation

305

and quick response as well as a portable supporting device. More importantly, the

306

centrifugation-based strategy is compatible with many sensitivity enhancing methods reported

307

before to achieve more optimized results.

308

Determination of PSA in clinical serum samples

309

Analytical performance of CLFIA was tested with clinical serum samples. Serum samples

310

were diluted 4 times with the working buffer, and the PSA concentration in each sample was

311

determined by CLFIA under condition 2. PSA concentrations were calculated according to the

312

standard curve (Fig. 4f). Each sample result was plotted in Fig. 5. The regression coefficient of

313

the equation is close to 1 and the R2 value is above 0.97, indicating a good agreement between

314

the PSA concentrations measured by CLFIA and by electrochemiluminescence (ECL) with a

315

commercial bulky instrument in a hospital. Besides, compared with previous researches,7-10 the 17 / 28



316

relative deviations between CLFIA and ECL results were all within tolerance considering the

317

variability of serum samples (Table S2), indicating a good accuracy. The results demonstrate

318

that our method possesses the potential for clinical analysis. Moreover, our method can

319

additionally provide valuable monitoring data at low biomarker concentrations compared to

320

traditional LIFA (Fig. 4).

18 / 28


Page 18 of 28

Page 19 of 28 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

321


Conclusion

322

A novel centrifugation-assisted lateral flow immunoassay was developed for the first time

323

by inserting a piece of NC membrane into a centrifugal microfluidic disc. Powered by the

324

centrifugal force, the process of immunoassay was actively controlled. The sample volume and

325

liquid flow rate through the NC membrane, associated with the sensitivity, can be easily

326

adjusted. Under optimal conditions, our method showed 6.2-fold improvement in the LOD for

327

PSA measurements in serum compared to traditional LFIA by test strips. Additionally,

328

multistep operations including sample preparation and downstream amplification are readily

329

integrated, owing to the advantages of the centrifugal microfluidic platform. The centrifugal

330

immunoassay system proposed in this study provide a universal strategy that can be applied for

331

many analytes based on affinity interaction. Assisted by a portable and cost-effective

332

smartphone-based automatic operating device, this novel immunoassay method with increased

333

sensitivity shows great potential in healthcare, food safety, environmental monitoring, and

334

several other complex applications.

335

336

Acknowledgments

337

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

338

China (2016YFC0800703), the National Natural Science Foundation of China (31870853), and

339

the Beijing Lab Foundation.

340

341 342

Supporting Information Schematic of the operation device of CLFIA; volume measurement on captured images; 19 / 28



343

manufactural schematic of the vaulted NC membrane; schematic of LFIA; optimization of

344

AuNP labelling and the dosage of AuNP-labeled antibody; specificity test; summary of

345

different methods for PSA detection; relative deviations between CLFIA and ECL results.

20 / 28


Page 20 of 28

Page 21 of 28 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


346

References

347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387

(1) Posthuma-Trumpie, G. A.; Korf, J.; van Amerongen, A. Anal. Bioanal. Chem. 2009, 393, 569582. (2) Quesada-Gonzalez, D.; Merkoci, A. Biosens. Bioelectron. 2015, 73, 47-63. (3) Ang, G. Y.; Yu, C. Y.; Yean, C. Y. Biosens. Bioelectron. 2012, 38, 151-156. (4) Li, J.; Zou, M.; Chen, Y.; Xue, Q.; Zhang, F.; Li, B.; Wang, Y.; Qi, X.; Yang, Y. Anal. Chim. Acta 2013, 782, 54-58. (5) Lee, J. H.; Seo, H. S.; Kwon, J. H.; Kim, H. T.; Kwon, K. C.; Sim, S. J.; Cha, Y. J.; Lee, J. Biosens. Bioelectron. 2015, 69, 213-225. (6) Watanabe, Y.; Ito, T.; Ibrahim, M. S.; Arai, Y.; Hotta, K.; Phuong, H. V.; Hang Nle, K.; Mai le, Q.; Soda, K.; Yamaoka, M.; Poetranto, E. D.; Wulandari, L.; Hiramatsu, H.; Daidoji, T.; KubotaKoketsu, R.; Sriwilaijaroen, N.; Nakaya, T.; Okuno, Y.; Takahashi, T.; Suzuki, T., et al. Biosens. Bioelectron. 2015, 65, 211-219. (7) Oh, Y. K.; Joung, H. A.; Han, H. S.; Suk, H. J.; Kim, M. G. Biosens. Bioelectron. 2014, 61, 285289. (8) Wang, H.; Wang, H.; Chen, S.; Dzakah, E. E.; Kang, K.; Wang, J.; Wang, J. Clin. Chim. Acta 2015, 444, 37-42. (9) Liang, R.; Xu, X.; Liu, T.; Zhou, J.; Wang, X.; Ren, Z.; Hao, F.; Wu, Y. Anal. Chim. Acta 2015, 891, 277-283. (10) Yang, X.; Liu, L.; Hao, Q.; Zou, D.; Zhang, X.; Zhang, L.; Li, H.; Qiao, Y.; Zhao, H.; Zhou, L. PLoS One 2017, 12, e0171376. (11) Lou, D.; Fan, L.; Cui, Y.; Zhu, Y.; Gu, N.; Zhang, Y. Anal. Chem. 2018, 90, 6502-6508. (12) Fang, C.; Chou, C.; Yang, Y.; Tsai, W.; Wang, Y.; Chan, Y. Anal. Chem. 2018, 90, 2134-2140. (13) Bruno, J. G. Pathogens 2014, 3, 341-355. (14) Park, B. H.; Oh, S. J.; Jung, J. H.; Choi, G.; Seo, J. H.; Kim, D. H.; Lee, E. Y.; Seo, T. S. Biosens. Bioelectron. 2017, 91, 334-340. (15) Krska, R.; Molinelli, A. Anal. Bioanal. Chem. 2009, 393, 67-71. (16) Anfossi, L.; Calderara, M.; Baggiani, C.; Giovannoli, C.; Arletti, E.; Giraudi, G. Anal. Chim. Acta 2010, 682, 104-109. (17) Anfossi, L.; D'Arco, G.; Calderara, M.; Baggiani, C.; Giovannoli, C.; Giraudi, G. Food Addit. Contam. Part A Chem. Anal. Control Expo. Risk Assess. 2011, 28, 226-234. (18) Wang, S.; Zhang, C.; Zhang, Y. Methods Mol. Biol. 2009, 504, 237-252. (19) Zhao, Y.; Zhang, G.; Liu, Q.; Teng, M.; Yang, J.; Wang, J. J. Agric. Food Chem. 2008, 56, 12138-12142. (20) Douglas, D.; Banaszewski, K.; Juskelis, R.; Al-Taher, F.; Chen, Y.; Cappozzo, J.; McRobbie, L.; Salter, R. S. J. Food Prot. 2012, 75, 1270-1277. (21) Suarez-Pantaleon, C.; Wichers, J.; Abad-Somovilla, A.; van Amerongen, A.; Abad-Fuentes, A. Biosens. Bioelectron. 2013, 42, 170-176. (22) Sun, W.; Hu, X.; Liu, J.; Zhang, Y.; Lu, J.; Zeng, L. Biosci. Biotechnol. Biochem. 2017, 81, 1874-1882. (23) Song, C.; Zhi, A.; Liu, Q.; Yang, J.; Jia, G.; Shervin, J.; Tang, L.; Hu, X.; Deng, R.; Xu, C.; Zhang, G. Biosens. Bioelectron. 2013, 50, 62-65.

21 / 28



388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419

(24) Cheng, X.; Pu, X.; Jun, P.; Zhu, X.; Zhu, D.; Chen, M. Int. J. Nanomedicine 2014, 9, 56195626. (25) Yan, X.; Wang, K.; Lu, W.; Qin, W.; Cui, D.; He, J. Nanoscale Res. Lett. 2016, 11, 138. (26) Yang, D.; Ma, J.; Zhang, Q.; Li, N.; Yang, J.; Raju, P. A.; Peng, M.; Luo, Y.; Hui, W.; Chen, C.; Cui, Y. Anal. Chem. 2013, 85, 6688-6695. (27) Lu, X.; Liang, X.; Dong, J.; Fang, Z.; Zeng, L. Anal. Bioanal. Chem. 2016, 408, 6703-6709. (28) Rodriguez, M. O.; Covian, L. B.; Garcia, A. C.; Blanco-Lopez, M. C. Talanta 2016, 148, 272278. (29) Parolo, C.; de la Escosura-Muniz, A.; Merkoci, A. Biosens. Bioelectron. 2013, 40, 412-416. (30) Duan, D.; Fan, K.; Zhang, D.; Tan, S.; Liang, M.; Liu, Y.; Zhang, J.; Zhang, P.; Liu, W.; Qiu, X.; Kobinger, G. P.; Gao, G. F.; Yan, X. Biosens. Bioelectron. 2015, 74, 134-141. (31) Parolo, C.; Medina-Sanchez, M.; de la Escosura-Muniz, A.; Merkoci, A. Lab Chip 2013, 13, 386-390. (32) Rivas, L.; Medina-Sanchez, M.; de la Escosura-Muniz, A.; Merkoci, A. Lab Chip 2014, 14, 4406-4414. (33) Haeberle, S.; Zengerle, R. Lab Chip 2007, 7, 1094-1110. (34) Madou, M.; Zoval, J.; Jia, G.; Kido, H.; Kim, J.; Kim, N. Annu. Rev. Biomed. Eng. 2006, 8, 601-628. (35) Cheng, J.; Sheldon, E. L.; Wu, L.; Uribe, A.; Gerrue, L. O.; Carrino, J.; Heller, M. J.; O'Connell, J. P. Nat. Biotechnol. 1998, 16, 541-546. (36) Xu, Y.; Lv, Y.; Wang, L.; Xing, W.; Cheng, J. Biosens. Bioelectron. 2012, 32, 300-304. (37) Meng, X.; Zhu, Y.; Chen, Y.; Lu, Y.; Xu, Y.; Cheng, J. Sens. Actuators, B 2017, 242, 710-717. (38) Zhu, Y.; Chen, Y.; Meng, X.; Wang, J.; Lu, Y.; Xu, Y.; Cheng, J. Anal. Chem. 2017, 89, 93159321. (39) Kim, T. H.; Park, J.; Kim, C. J.; Cho, Y. K. Anal. Chem. 2014, 86, 3841-3848. (40) Thobhani, S.; Attree, S.; Boyd, R.; Kumarswami, N.; Noble, J.; Szymanski, M.; Porter, R. A. J. Immunol. Methods 2010, 356, 60-69. (41) Ducree, J.; Haeberle, S.; Lutz, S.; Pausch, S.; von Stetten, F.; Zengerle, R. J. Micromech. Microeng. 2007, 17, S103-S115. (42) Haiss, W.; Thanh, N. T.; Aveyard, J.; Fernig, D. G. Anal. Chem. 2007, 79, 4215-4221. (43) Zhan, L.; Guo, S. Z.; Song, F.; Gong, Y.; Xu, F.; Boulware, D. R.; McAlpine, M. C.; Chan, W. C. W.; Bischof, J. C. Nano Lett. 2017, 17, 7207-7212.

22 / 28


Page 22 of 28

Page 23 of 28 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


420 421

Fig. 1 Schematic of the CLFIA system. (a) The patterned NC membrane. T and C show the

422

positions of the test and control lines, respectively. (b) The configuration of the CLFIA disc.

423

(c) Illustration of the patterned PMMA layer. (d) An assembled CLFIA disc. (e) The

424

smartphone-based operating device.

23 / 28



425 426

Fig. 2 Performance of liquid manipulation. (a) Comparison of the liquid volume on LFS and

427

through the NC membrane on CLFIA discs at different reaction times. (b) Liquid flow rates

428

through the NC membrane of CLFIA at different rotation speeds. The curves in this figure are

429

fitted as guides to the eye. Error bars show standard deviations of triplicate measurements.

24 / 28


Page 24 of 28

Page 25 of 28 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


430 431

Fig. 3 Signal intensities at different conditions. (a) Signal intensity using different volumes of

432

PSA standard solution at a concentration of 3 ng/mL and rotation speed of 2000 rpm.

433

Photographic images of (i) LFIA results and (ii) CLFIA results with different sample volumes.

434

(iii) Quantitative comparison of LFIA and CLFIA results. “Blank” means the background using

435

120 μL working buffer without PSA. (b) Signal intensity of CLFIA at different rotation speeds

436

using PSA standard solution (60 μL) at a concentration of 1 ng/mL. (i) Photographic images of

437

CLFIA results and (ii) quantitative signal analysis at different rotation speeds. (c) Signal

438

intensity of CLFIA at different rotation speeds for 15 min with sufficient PSA standard solution

439

(1 ng/mL). (i) Photographic images of CLFIA results and (ii) quantitative signal analysis at

440

each rotation speed. ΔGray intensity = signal intensity - background intensity. Error bars show

441

standard deviations of triplicate measurements.

25 / 28



442 443

Fig. 4 Detection results with PSA spiked serum samples. (a-c) Photographic images of LFIA,

444

CLFIA condition 1, and CLFIA condition 2, respectively, with PSA spiked samples at different

445

concentrations. (d-f) Calibration curves for LFIA, CLFIA condition 1, and CLFIA condition 2,

446

respectively. Error bars show standard deviations of triplicate measurements.

26 / 28


Page 26 of 28

Page 27 of 28 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


447

Table 1 Relative standard deviations of results at different PSA concentrations measured by

448

LFS, CLFIA condition 1, and CLFIA condition 2. (n = 3)

449

27 / 28



450 451

Fig. 5 Correlation between PSA concentrations of clinical serum samples measured by CLFIA

452

and by ECL.

28 / 28


Page 28 of 28

Enhancing the Sensitivity of Lateral Flow ... - ACS Publications

Recommend Documents