Ultra-sensitive method based on time-resolved fluorescence immunoassay for detection of sulfamethazine in raw milk

ABSTRACT

A novel lateral flow assay (LFA) was developed by introducing Eu (III)-doped polystyrene nanoparticles (EuNPs) for rapid and ultra-sensitive detection of sulfamethazine (SM₂) in raw milk. The limit of detection and linear range of the proposed method were 0.0045 and 0.05–10 ng/mL, respectively. The recovery of LFA for the detection of SM₂ in raw milk was 96.1–108.2%. The proposed LFA provides a rapid and convenient strategy for fast and ultra-sensitive screening of SM₂ in raw milk. EuNP-LFA may be a remarkable method for the detection of other targets at low concentrations to ensure food safety.

GRAPHICAL ABSTRACT

KEYWORDS:

• Lateral flow assay
• Eu (III)-doped polystyrene nanoparticle
• sulfamethazine
• raw milk

1. Introduction

Sulfamethazine (SM₂) is widely used in veterinary medicine for treatment of infectious diseases. The abuse of using SM₂ in veterinary practice may lead to the presence of SM₂ residues in milk (Franek, Kolar, Deng, & Crooks, 1999; Galarini et al., 2014; Karageorgou, Manousi, Samanidou, Kabir, & Furton, 2016; Wang et al., 2012). Excessive SM₂ residue is harmful to consumers because of its carcinogenic potential and risk of antibiotic resistance (Barani & Fallah, 2015; Marshall & Levy, 2011). To protect consumers from the risk, China (Ministry of Agriculture), Canada (Food and Drug Regulation), and the European Union (EU Regulation) established the maximum residue limits of total sulfonamides at 100 µg/kg in milk. The Chinese Government have especially ruled out the maximum residue limits of SM₂ at 25 µg/kg in milk.

SM₂ was traditionally detected by high-performance liquid chromatography (Xu et al., 2010) and liquid chromatography–mass spectrometry (Shao et al., 2005). Although these methods are highly accurate, they are time consuming, costly, and generally unsuitable for use on site. Lateral flow assay (LFA), which is based on specific recognition of antigens and antibodies, is a rapid, simple, specific, and inexpensive technique (Hao et al., 2018; Peng, Pei, Zheng, Wang, & Wang, 2017; Sajid, Kawde, & Daud, 2015; Wang et al., 2018; Zhang et al., 2018). However, the traditional LFA based on colloidal gold as the label exhibits low sensitivity (Shim, Kim, Kim, & Chung, 2013). Hence, improving the sensitivity of LFA for SM₂ detection in actual samples is important and necessary.

Eu (III)-doped polystyrene nanoparticles exhibit a large stoke shift, narrow emission bands, and high fluorescence intensity (Hu, Luo, Xia, Xu, & Wu, 2017; NãReoja, Rosenholm, Lamminmã, & HãNninen, & E, 2017). Using Eu (III)-doped polystyrene nanoparticle (EuNP) as the label can considerably enhance the sensitivity of LFA (Luo, Hu, Guo, Wu, & Wu, 2017; Sun, Xie, Peng, Wang, & Xie, 2017).

Heterologous competitive strategy is a prevalent method for changing the recognition between antibody and antigen to improve the sensitivity of immunoassays. Scholars have focused on preparing heterologous coating antigen (HCA) to enhance the sensitivity of immunoassays (Peng, Liu, Kuang, Cui, & Xu, 2016a; Wang, Zhang, Ni, Zhang, & Shen, 2014).

In this study, a novel LFA was established based on EuNP as the label and HCA as the detective antigen (EuNP-HCA-LFA) for the sensitive detection of SM₂ in raw milk. EuNP-HCA-LFA may be a remarkable method for the sensitive detection of other targets at low concentrations to ensure food safety.

2. Materials and methods

2.1. Materials and reagents

SM₂, sulfameter, sulfasoxazole, sulfamethoxypyridazine, sulfachloropyridazine, sulfamethoxazole, sulfamonomethoxine, sulfadimethoxine, sulfathiazole, and sulfamerazine were obtained from J&K Scientific Ltd (Shanghai, China), and the stock solution of SM₂ was prepared in methanol (2 mg/mL) for further use. Bovine serum albumin (BSA), 1-(3-dimethylaminopropyl)-3-ethyl-carbodiimide (EDC), borate for preparation of borate buffer (BB; 0.05 M, pH 6.0), and phosphate for preparation of phosphate-buffered saline (PBS; 0.01 M, pH 7.4) were purchased from Sigma-Aldrich (St. Louis, MO, USA). EuNPs (excitation wavelength = 365 nm, emission wavelength = 615 nm) was obtained from Microdetection Bio-tech Co., Ltd. (Nanjing, China). Raw milk was provided by Jiangxi Sunshine Dairy Co., Ltd. (Jiangxi, China). Anti-SM₂ monoclonal antibody (mAb), prepared using SM₂-DZ-BSA as immunogen, was provided by Wuxi Zodoboer Biotech. Co., Ltd. (Wuxi, China). In addition, two detective antigens, namely, SM2-DZ-BSA and SM₂-GA-BSA, were prepared in our laboratory (the UV–vis spectra of the two antigens are shown in Supporting Information Figure S1). SM₂-DZ-BSA and SM₂-GA-BSA were prepared by coupling SM₂ to BSA through diazotization (Li et al., 2009) and glutaraldehyde method (Peng, Liu, Kuang, Cui, & Xu, 2016b), respectively. Goat anti-mouse IgG was purchased from Meridian Life Science Inc. (Memphis, TN, USA). All other chemicals and reagents were of analytical grade and acquired from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China).

A portable strip reader was obtained from Fenghang Scientific Instrument Co., Ltd. (Zhejiang, China) and used for the LFA test strip (excitation wavelength= 365 nm, emission wavelength = 610 nm). Nitrocellulose (NC) membrane, sample pad, conjugate pad, and absorbent pad were purchased from Millipore (Bedford, MA, USA). Bio-layer interferometry (BLI; ForteBio, Menlo Park, CA, USA) and anti-mouse Fc biosensor (AMC sensor) were provided by Pall ForteBio LLC (CAE, USA).

2.2. Preparation of labelled mAb

EuNP–mAb was prepared using a previously reported method with some modifications (Luo et al., 2017). EuNP (50 µg) was added to 0.5 mL of 0.05 M BB (pH 7.5) containing 1.75 µg of EDC to activate the carboxyl of the surface of materials for 15 min. Anti-SM₂ mAb (0.625, 1.25, 2.50, 5.0, and 7.5 µg) was added to the mixture to form EuNP–mAb complexes. The remaining active sites of the complex solution were blocked with 50  µL of 0.05 M BB containing 10% (w/v) BSA. Finally, the mixtures were centrifuged at 12,000 rpm at 4 °C for 20 min. The precipitates were dissolved in 0.1 mL of 0.05 M BB (pH 7.5) containing 0.2% (w/v) BSA and 0.5% (v/v) Tween-20. The concentrations of mAb coupled with EuNP in the final solution were 6.25, 12.5, 25.0, 50.0, and 75 µg/mL.

2.3. Characterization of the EuNP and EuNP–mAb complexes

The morphology and sizes of the EuNP were characterized using a transmission electron microscope (TEM, JEOLJEM 2100, Tokyo, Japan). The zeta potential, size intensity, and polydispersity index (PDI) of the EuNP and EuNP–mAb complexes were analysed (Xing et al., 2018) by a particle size analyzer (Nano-ZS, Malvern Instruments Ltd., Worcestershire, UK).

2.4. Preparation of LFA test strip

The detective antigens, namely, SM₂-DZ-BSA and SM₂-GA-BSA (1.0 mg/mL), and goat anti-mouse IgG (0.5 mg/mL) were sprayed on the NC membrane as the test and control lines, respectively. These membranes were dried at 37 °C overnight. The sample pad, conjugate pad, NC membrane, and absorbent pad were assembled as LFA test strip (Scheme 1).

Scheme 1. Schematic of LFA based on competitive format immunoassay principle to quantitatively analyse SM₂. When the sample was free of SM₂ (control), the EuNP–mAb complexes migrated along the NC membrane under the action of capillary force and were available to bind to SM₂-DZ-BSA or SM₂-GA-BSA immobilized on the test line. The EuNP–mAb complexes would be captured by the antigen when the sample was presented SM₂ (positive). The fluorescence intensity of the test line changed from strong to weak as the concentration of SM₂ increased. The EuNP–mAb complexes should be captured by goat anti-mouse IgG immobilized on the control line regardless of the presence of the target SM₂.

2.5. Assays with LFA test strip

SM₂ stock solution was diluted in 0.01 M PBS (pH 7.4). As shown in Scheme 1, the control or positive sample solution (100 µL) was mixed with the EuNP–mAb complexes (4 µL) in an ELISA well for 5 min and added to the LFA test strip. After 30 min of reaction, the fluorescence intensity of the test line (FI_T) was recorded. All experiments were performed in triplicate.

2.6. Measurements of the affinities of mAb for the two detective antigens

The procedure was performed in accordance with a previous work (Li, Duan, Xu, Huang, & Xiong, 2016) with some modifications. The AMC sensor was first pre-wetted and equilibrated with PBS buffer (0.01m M, pH 7.4). The antibody was non-covalently loaded on the surface of the AMC sensor and incubated for an addition time of 600 s off line. After a 300 s equilibration step, the association assay of antibody to coating antigens was conducted for 180 s. Dissociation assays were carried out with PBS for 120 s. Finally, the association rate (k_a), dissociation rate (k_d), and affinity constant (KD) of mAb for SM₂-DZ-BSA and SM₂-GA-BSA were determined by loading the experimental data into Octet Data Analysis software.

2.7. Establishment of standard calibration curve in PBS

SM₂ (0, 0.01, 0.05, 0.1, 0.5, 1.0, 5.0, 10.0, 20 ng/mL) was added to PBS. A standard curve was established using the logarithm of the concentration of SM₂ and T/T₀, where T indicates FI_T when SM₂ is present in samples, and T₀ represents FI_T when the solution only contains PBS. The limit of detection (LOD) (Adrian et al., 2009) is defined as IC₉₀.

2.8. Specificity analysis

The specificity of the proposed LFA was assessed by testing 10 sulfonamide drugs. All experiments were repeated in triplicate (Chen et al., 2017).

2.9. Detection of SM₂ in raw milk

2.9.1. Establishment of standard calibration curve in raw milk

SM₂ (0, 0.01, 0.05, 0.1, 0.5, 1.0, 5.0, 10.0, and 20 ng/mL) was added to raw milk. A standard curve was established using the logarithm of the concentration of SM₂ and T/T₀, where T refers to FI_T when spiked-SM₂ is present in raw milk, whereas T₀ represents FI_T when raw milk possesses no spiked SM₂.

2.9.2. Recovery of SM₂ in raw milk on the EuNP-HCA-LFA system

SM₂ (0.05, 0.5, and 5 ng/mL) was added to raw milk. The recovery and coefficient of variation (CV) of SM₂ in raw milk were quantitatively analysed on the LFA system.

3. Results and discussion

3.1. Characterization of EuNP and EuNP–mAb complexes

The morphology and size of EuNP were characterized by TEM (Figure 1(A)). EuNP exhibited homogenous size and an average diameter of 193 nm. The sizes (Figure 1(B)) of EuNP and EuNP–mAb complex were 220 and 255 nm, respectively, indicating that mAb was successfully coupled on the EuNP surface. In Figure 1(C), the PDI of EuNP was 0.002. After the combination of EuNP with mAb, the PDI slightly changed to 0.025 (6.25 μg/mL), 0.039 (12.5 μg/mL), 0.033 (25 μg/mL), 0.01 (50 μg/mL), and 0.012 (75 μg/mL). This finding revealed the good dispersibility of EuNP–mAb and the stability for the application of EuNP–mAb complexes in EuNP-LFAs. The zeta potential (Figure 1(D)) of EuNP increased from −50 mV to >–36 mV after coupling with mAb. Hence, mAb combined with EuNP.

Figure 1. Characterization of EuNP and EuNP–mAb complexes. (A) TEM image of EuNP. (B) Size characterization of EuNP and EuNP–mAb complexes by dynamic light scattering. (C) PDI of EuNP and EuNP–mAb complexes. (D) Zeta potential of EuNP before and after coupling with mAb.

3.2. Optimization of the concentration of mAb coupled with EuNP

When SM₂-GA-BSA was used as the detective antigen, the concentration of mAb (6.25, 12.5, 25.0, 50.0, and 75.0 µg/mL) coupled with EuNP was optimized. PBS solution without SM₂ (control sample) and positive sample spiked with SM₂ in PBS solution (0.1 ng/mL) were prepared. The signal intensity and inhibition rate were determined by analysing FI_T0 and (1−T/T₀), respectively. High signal value and inhibition rate are the reference for selecting the optimized concentration of mAb coupled with EuNP (Kong, Liu, Song, Suryoprabowo, & Li, 2016). As shown in Figure 2, in solutions with mAb coupled with EuNP at 6.25, 12.5, 25.0, 50.0, and 75.0 µg/mL, the corresponding signal intensities were 87.5, 33.0, 64.0, 708.0, and 1222.5, respectively, and corresponding (1–T/T₀) were 0.97, 0.98, 0.98, 0.74, and 0.43, respectively. The FI was too low and had no increasing trend when the concentration of EuNP–mAb were 6.25, 12.5, and 25.0 µg/mL. When the concentration of EuNP–mAb were 50.0 and 75.0 µg/mL, FI increased rapidly. However, the higher amount of EuNP–mAb had a lower sensitivity of the LFA because of competitive format. Hence, 50 µg/mL was selected as the optimized concentration of mAb coupled with EuNP, When the SM₂-DZ-BSA was used as detective antigen, the optimized concentration of mAb coupled with EuNP was 25 µg/mL (Supporting Information Figure S2).

Figure 2. Optimization of the concentration of mAb coupled with EuNP with SM₂-GA-BSA as detective antigen.

3.3. Affinities of mAb to two detective antigens

BLI is an optical technique that allows for the real-time monitoring of the interactions between mAb and antigens without the requirements for enzymatic, fluorescent, or radioactive labels. In the measuring process, the interaction of mAb with an immobilized ligand on a biosensor surface forms a molecular layer, and the data obtaining is the basis of the changes in the thickness of the molecular layer which is related to the concentrations and the characters of some different antigens. In this study, the affinity properties between mAb and SM₂-DZ-BSA and SM₂-GA-BSA were determined using a BLItz instrument under the same conditions. As shown in Figure 3, the KD values of mAb for SM₂-DZ-BSA and SM₂-GA-BSA were 3.875×10⁻⁹ and 3.214×10⁻⁵ M, respectively. The results revealed that the affinity of mAb to HCA (SM₂-GA-BSA) was lower than that to the homologous antigen (SM₂-DZ-BSA).

Figure 3. Affinity properties of mAb to SM₂-DZ-BSA and SM₂-GA-BSA as measured by a BLItz instrument.

3.4. Standard curve of the two LFAs in SM₂-spiked PBS

The standard curves of the two LFAs were constructed by plotting the T/T₀ against the logarithm of the concentrations of SM₂. Figure 4(A, B) indicated that the regression equations of the two LFAs with SM₂-GA-BSA and SM₂-DZ-BSA as detective antigens were y= –0.2939 log (x) + 0.2084 (R² =0.9808) and y= –0.3757 log (x) + 0.2730 (R² =0.9827), where y is the T/T₀ and x is the concentration of SM₂, respectively. A comparison of the IC₅₀, LOD (Adrian et al., 2009), and linear ranges of the two LFAs in PBS is presented in Table 1. SM₂-GA-BSA as HCA could clearly enhance the sensitivity of LFA. Our result was consistent with that of Goodrow (Goodrow, Harrison, & Hammock, 1990), who found that the sensitivity of immunoassays is enhanced when HCA is used as the detective antigen because the affinity of mAb to HCA is less than that to a homologous antigen.

Figure 4. Standard calibration curve of the two LFAs in PBS were obtained by determining the T/T₀ against the logarithm of the concentrations of SM₂ under SM₂-GA-BSA (A) and SM₂-DZ-BSA (B) as detective antigens.

Table 1. Sensitivity and linear range of the two EuNP-LFAs in PBS.

Detective antigens	IC50 (ng/mL)	Limit of detection (ng/mL)	Linear range (ng/mL)
SM2-GA-BSA	0.10	0.0044	0.01–5
SM2-DZ-BSA	0.25	0.021	0.01–5

3.5. Specificity of the EuNP-HCA-LFA

The specificity of the EuNP-HCA-LFA was assessed by testing the 10 sulfonamide drugs (10 ng/mL) with SM₂-GA-BSA as the detective antigen. As shown in Figure 5, the T/T₀ value of SM₂-spiked PBS sample was 0.01 and much lower than that with the nine other sulfonamide drugs. The result indicated that the EuNP-HCA-LFA could specifically determine SM₂.

Figure 5. Specificity of the EuNP-HCA-LFA over 10 sulfonamide drugs with SM₂-GA-BSA as detective antigen.

3.6. Detection of SM₂ in raw milk

3.6.1. Standard curve of the two LFAs in SM₂-spiked raw milk

The standard curves of the two LFAs in raw milk were constructed by plotting the T/T₀ against the logarithm of the concentrations of SM₂. Figure 6(A, B) indicated that the regression equations of the two LFAs with SM₂-GA-BSA and SM₂-DZ-BSA as the detective antigens were y= –0.2037 log (x) + 0.3781 (R² =0.9782) and y= –0.2605 log (x) + 0.5079 (R² =0.9836), respectively, where y is the T/T₀, and x is the concentration of SM₂, respectively. The IC₅₀, LOD, and linear ranges of the two LFAs in raw milk are compared in Table 2. Similarly, the EuNP-HCA-LFA also offered advantages to detect SM₂ in raw milk. The IC₅₀, LOD, and linear range were 0.25, 0.0045, and 0.05–10 ng/mL, respectively.

Figure 6. Standard calibration curve of the two LFAs in raw milk were obtained by determining the T/T₀ against the logarithm of the concentrations of SM₂ under SM₂-GA-BSA (A) and SM₂-DZ-BSA (B) as detective antigens.

Table 2. Sensitivity and linear range of the two EuNP-LFAs in raw milk.

Coating antigens	IC50 (ng/mL)	Limit of detection (ng/mL)	Linear range (ng/mL)
SM2-GA-BSA	0.25	0.0045	0.05–10
SM2-DZ-BSA	1.07	0.076	0.05–10

3.6.2. Recovery of EuNP-HCA-LFA for the detection of SM₂ in raw milk

The recovery of EuNP-HCA-LFA for the determination of 0.05, 0.5, and 5 ng/mL SM₂ in raw milk was examined using the test strips with SM₂-GA-BSA as the detective antigen. Table 3 revealed the average EuNP-HCA-LFA recoveries of 96.1–108.2% with CVs of 6.5–10.8%. These data indicated that the EuNP-HCA-LFA could reliably and precisely determine SM₂ in raw milk samples.

Table 3. Recovery of SM₂ on the EuNP-HCA-LFA in SM₂-spiked milk sample.

Spiked SM2 (ng/mL)	Recovery (%)^a	SD	CV (%)
0.05	108.2	9.2	8.5
0.5	96.1	6.3	6.5
5	105.3	11.4	10.8

^aRecovery= (detection concentration/spiked concentration) × 100%.

3.7. Comparison of EuNP-HCA-LFA with other methods for SM₂ detection

A comparison of the linear range and sensitivity of the EuNP-HCA-LFA in our study with those in other works is listed in Table 4. The results showed that EuNP-HCA-LFA exhibited the broadest linear range of 0.05–10 ng/mL and the lowest LOD of 0.0045 ng/mL in milk samples, which occurred because of the coupling of EuNP with HCA on the LFA system. As previously mentioned, EuNP, which has a large stokes shift, narrow emission bands, and high fluorescence intensity, increased the sensitivity. The use of HCA could further enhance the sensitivity of the method by fivefolds, thereby creating an ultra-sensitive LFA for SM₂ detection in raw milk samples.

Table 4. Comparison of EuNP-HCA-LFA and other methods for SM₂ detection.

Methods	Target	Detection samples	Linear range	LOD	References
MI-ELISA	Sulfamethazine	Swine muscle	100−3200 μg/L	20.4 μg/kg	Peng et al. (2017)
Electroanalytical assay	Sulfamethazine	Honey	278.33−139165 μg/L	44.53 μg/L	Su et al. (2017)
Microfluidic Chips	Sulfamethazine	Milk	2−200 μg/L	0.6 μg/L	Wang et al. (2012)
IC-ELISA	Sulfamethazine	Chicken	2–32 μg/L	17.4 ug/kg	Zhou et al. (2014)
LFA	Sulfamethazine	Egg	0.5−100 μg/L	2.0 μg/L	Shim et al. (2013)
LFA	Sulfamethazine	Swine urine	0.625–80.0 μg/L	1.0 μg/L	Li et al. (2009)
NIR-LFA	Sulfadimidine	Milk	0.1–3.98 μg/L	0.1 μg/L	Chen et al. (2016)
EuNP-LFA	Sulfamethazine	Milk	0.05−10 μg/L	0.0045 μg/L	This work

4. Conclusions

EuNP combined with HCA was applied to the LFA for the first time for the rapid and ultra-sensitive detection of SM₂ in raw milk samples. The novel method could dramatically enhance the sensitivity of LFA compared with other methods, demonstrating a linear range of 0.05–10 ng/mL and LOD of 0.0045 ng/mL in milk samples. This type of LFA offers promising application in the detection of other targets at low concentration.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Song Hu http://orcid.org/0000-0003-2862-4248

Additional information

Funding

This work was supported by National Undergraduate Training Program for Innovation and Entrepreneurship [grant number 201701055]; Jiangxi Special Fund for Agro-scientific Research in the Collaborative Innovation [grant number JXXTCX201703-1]; earmarked fund for Jiangxi Agriculture Research System [grant number JXARS-03]; National Science Foundation for Young Scientists of China [grant number 31800776].