Introduction

Chloramphenicol (CAP) is widely used in veterinary medicine, but excessive levels of CAP in human blood may lead to diseases such as leukemia, aplastic anemia, and gray baby syndrome [1, 2]. Thus, the use of CAP is banned in food-producing animals in many countries, including China, the USA, and the European Union, and a minimum required performance limit value is set at 0.3 × 10−6 g·kg−1 [3]. As a group I carcinogen and “milk toxin,” aflatoxin B1 (AFB1)-derived aflatoxin M1 (AFM1) is hazardous to humans exposed to contaminated milk [4,5,6]. The limits on AFM1 levels in milk and dairy products were set, and the sources of AFM1 contamination need to be monitored [7].

A paper-based lateral flow immunochromatographic assay (ICA) was developed for sensitive and economical on-site determination of residues, including CAP and AFM1 [8,9,10]. For better ICA detection efficiency, several signal labels, including time-resolved fluorescence [11], up-converting phosphors [12], magnetic nanoparticles [13], quantum dots (QDs) [14], and QD submicrobead (QB), have been developed and discussed. QB has attracted great interest for its unique properties, such as narrow fluorescent emission spectra, high quantum yield, high photochemical stability, and high-throughput detection [15,16,17,18,19]. Single or several distinguishable signal-providing labels are applied in a single ICA detection for multiplexing and usually involves the modification of the architecture of the ICA strip [20,21,22,23,24,25,26,27,28].

However, one ubiquitous characteristic of the existing ICA system is that target-specific antigen or antibody immobilized on the testing zone hinders the development of another crucial property: universality. Transient and largely uncontrolled immune recognition on the testing zone has led to insufficient reaction and limited sensitivity.

In this work, a universal and sensitive ICA (USICA) for CAP and AFM1 was established using the anti-BSA antibody as coating element, and a preincubation was introduced. BSA-modified CAP (CAP–BSA) exposes the binding site thoroughly in mixture and competes with CAP for the binding site on the anti-CAP antibody-modified QB in the preincubation step for high sensitivity. Two kinds of immune complexes, namely, QB–mAb–CAP and QB–mAb–CAP–BSA, were formed. The QB–mAb–CAP–BSA is conjugated with the anti-BSA antibody on the test zone by the BSA terminal, and the QB–mAb–CAP is captured by the anti-IgG antibody on the control zone in the USICA strip. The immune complex QB–mAb–CAP–BSA is recognized by the BSA terminal, which is nontarget-specific and brings the foundation for universality. The AFM1 is detected for validation of universality by using the BSA-modified AFM1 and the USICA of the same structure. The properties of the two USICAs are systematically summarized and compared with those of the ICA based on traditional structure (TICA). The applicability of the USICA is demonstrated by analyzing spiked milk samples and its performance compared with the commercial ELISA kit.

Experimental

Materials and instruments

CAP and AFM1 standards, 1-ethyl-3-(3-(dimethylamino)propyl)-carbodiimide hydrochloride (EDC), and N-hydroxysuccinimide (NHS) were supplied by the J&K Scientific Ltd. (Shanghai, China, www.jkchemical.com/Company-inf.aspx?language=ch). The CAP−BSA and AFM1 − BSA conjugates, the commercial ELISA kits for CAP and AFM1, and the anti-CAP, anti-AFM1, anti-BSA, and goat antimouse IgG antibodies were supplied by the Shandong Landu Bio-Science & Technology Co., Ltd. (Shandong, China, www.11467.com/qiye/45274356.htm). Carboxylic group-modified CdSe/ZnS core–shell QBs (emission at 605 ± 5 nm) were purchased from the Beijing Najing Biological Technology Co., Ltd. (Beijing, China, www.najingbio.com/). The protein stabilizer solution (Cat: PR-SS-002, Huzhou Yingchuang Biological Technology Co., Ltd., Huzhou, China, www.innoreagents.com) was used for QB–mAb storage. Nitrocellulose (NC) membrane and the sample and the absorbent pads were obtained from the Kinbio Tech. Co., Ltd. (Shanghai, China, www.goldbio.cn/article-item-288.html). All other reagents (analytical grade) were supplied by the National Pharmaceutical Group Chemical Reagent Co., Ltd. (Shanghai, China, www.sinopharm.com/1156.html). A fluorescent strip reader was supplied by the Beijing Najing Biological Technology Co., Ltd. (Beijing, China). The USICA and TICA strips were prepared by using the BioDot XYZ platform combined with a BioJet Quanti3000k dispenser and motion controller from BioDot (Irvine, CA) and cut by an automatic programmable cutter from the Shanghai Jinbiao Biotechnology Co., Ltd. (Shanghai, China, www.goldbio.cn/article-item-288.html). Pure water was prepared using the Elix-3 and the Milli-QA system (Millipore Co., Bedford, MA, USA, http://www.well-honor.com/goods1-202.html).

QB modification with anti-target mAb

Carboxy-modified QB was activated and modified using the antitarget mAb through the active ester method [29]. Details are described in the Electronic Supporting Material (ESM).

Fabrication of the USICA and the TICA systems

The two ICA systems were similar in structure. These systems contain three parts: (1) the NC membrane in the middle, (2) the glass fiber, and (3) absorption pads on both ends overlapping nearly 2 mm with NC membrane on the backing card. The USICA system had an anti-BSA antibody at 0.25 mg·mL−1 immobilized at a density of 3 μL·cm−1 in the NC membrane as test line, whereas the TICA system had a BSA-modified target (CAP–BSA) at a density of 3 μL·cm−1. The goat antimouse IgG antibodies (0.5 mg mL−1) were immobilized in both ICA systems at a density of 3 μL·cm−1 in the NC membrane as the control line. The formed ICA system was dried at 37 °C in a blasting drying trunk for 2 h, cut into strips (3.5 mm in width) by using the automatic cutter, and stored at 4 °C until use.

Quantitative procedure of the USICA and the TICA systems

Parameters, such as the amount of mAb modified on QB, pH value, ionic strength, the amount of antigen coated on T line, the amount of surfactant, dilution ratio of QB–mAb, and ICA reaction time, was optimized in the TICA system for best detection results. In addition to the parameters optimized in the TICA system, the amount of antigen used for preincubation, the amount of anti-BSA mAb immobilized on the T line, and the incubation time were optimized in the USICA system. The optimization was carried out in competitive inhibition mode, in which the target was spiked at the final concentrations of 0 and 1.25 ng·mL−1, and the parameter inducing the most obvious inhibition was selected. The analyte was dissolved in methanol, and the final methanol concentration was 5% (v/v) in the spiked samples. All experiments were performed in triplicate. The fluorescence intensity (FI) in the T line was inversely proportional to the analyte content, and a calibration plot was established on this basis.

AFM1- and CAP-spiked samples were pretreated. Trichloroacetic acid was added to the spiked milk samples to a final concentration of 10% (w/v). After thorough mixing and centrifugation, 2.5 μL NaOH (5 mol·L−1) was used per 100 μL supernatant for neutralizing. The insoluble materials that appeared during neutralizing were removed by centrifugation, and the neutral supernatant was used for ICA. The CAP- and AFM1-free milk samples, which were confirmed by LC–MS/MS, were collected from the local market. Accuracy and precision analyses were carried out. The universality was validated by detecting the CAP and the AFM1 using the USICA strip.

Comparative study with commercial ELISA kit

Spiked milk samples were determined using the USICA and the TICA strips and the commercial ELISA kit. Sample pretreatment and the detection procedure for the commercial ELISA kit was performed according to the manufacturer’s instructions. A correlation coefficient (R2) of the two methods was calculated through a fit plot.

Results and discussion

Figure 1 presents a schematic for USICA and TICA using CAP as an example. In USICA, preincubation was introduced for high sensitivity, in which the target (CAP) and the BSA-modified target (CAP–BSA) competed for the binding site on the QB–mAb under optimal conditions. The anti-BSA antibody was introduced in the test line (T line) for universal detection, and the nontarget-specific and BSA-containing complexes were conjugated. The intensity of QB in the T line was inversely proportional to the amount of the target, which was the basis of establishment of the novel USICA system. Meanwhile, in the TICA system, the immune recognition that occurred on the testing zone was transient and largely uncontrolled, and universal detection was not discussed.

Fig. 1
figure 1

Schematic for USICA and TICA. Preincubation was introduced for competitive recognition of high sensitivity, and the T line was coated with anti-BSA antibody for universal detection (a); the competitive recognition occurred at the T line in TICA (b); Excess modified QBs (e.g., QB–CAP–mAb) were captured by element in the C line

Full size image

QB modification with antitarget mAb

The morphology of the anti-AFM1 mAb-modified QB under a scanning electron microscope (SEM) is shown in Fig. 2a. A layer of material appeared on the surface (Fig. A insert) after modification. Fourier transform infrared (FTIR) spectroscopy was applied to characterize the result of modification. As shown in Fig. 2b, the characteristic absorption peaks corresponding to protein amide bands I (1641 cm−1) and II (1530 cm−1) existing in the QB–mAb proved successful modification. The anti-CAP mAb-modified QB was prepared using the same method and applied in the following experiment.

Fig. 2
figure 2

SEM image of QB after conjugation (a) and a layer of mAb coated on the surface (A insert). FTIR spectra of QB (curve a), anti-AFM1 antibody (curve b), and QB–mAb (curve c) (b). The characteristic absorption peaks of protein that appeared in the final products proved the success of the modification

Full size image

Property and analytical performance of the USICA and the TICA systems

In the TICA for small molecule detection, competitive recognition occurred on the test line. Target-specific antigen immobilized on the T line was adverse for full exposure of the binding site, and the transient and largely uncontrolled recognition on the T line negatively affected the degree of thorough recognition. The USICA system was established in this work to address these problems. Preincubation was introduced for high sensitivity. The target-specific antigen competed with the target for the binding site on the QB–mAb in the preincubation solution. The binding site was fully exposed in the solution, and competitive recognition was carried out in optimal conditions, including ample time. The anti-BSA antibody was introduced for universality. Two kinds of complexes were formed in preincubation, namely, QB–mAb–CAP and QB–mAb–CAP–BSA. Only the latter complexes were captured by the anti-BSA antibody through the BSA terminal in the complexes. The BSA in the target-specific antigen was nontarget-specific and brought the foundation for universality. Any analyte that can be modified with BSA and antibody-available may be detected by this method. CAP was detected using the USICA system under optimal conditions. For sensitivity comparison, CAP was also detected using the TICA system. The AFM1 was detected to validate the universality of USICA system, and the TICA system was used after systematic optimization. A calibration plot was established by plotting the B/B0 × 100% against the target amount in logarithm, where B and B0 represent the FI on the test line (FIT) with and without the presence of spiked analyte, respectively.

The LOD was defined as the concentration of the spiked target, at which 10% inhibition (IC10) of the total inhibition was reached. Optimal parameters for CAP and AFM1 using USICA and TICA are summarized in Table E1 (shown in ESM) After optimization, the LOD of USICA for CAP was nearly three times that of TICA (Table 1). The LOD for AFM1 improved about 13 times by using USICA compared with that using TICA. These results verify the high sensitivity and the universality of the USICA system. Compared with the recently reported nanomaterial-based optical methods for the determination of CAP and AFM1 in Table 1, the unique merit of USICA is its universality.

Table 1 Overview of the recently reported nanomaterial-based optical methods for the determination of CAP and AFM1
Full size table

The specificity of the USICA and the TICA systems for CAP was investigated using the cross-reaction (CR) value. Similar results among structural analogs, such as thiamphenicol (TAP) and florfenicol (FF), and other contaminants, such as kanamycin (KNM), streptomycin (SPM), and ceftiofur (CTF), were observed. The CR value was calculated using the eq. CR% = [(IC50 analyte) / (IC50 analog)] × 100 [36]. Figure 3 shows that the structural analogs TAP (CR = 14.3% in USICA) and FF (CR = 2.3% in USICA) had obvious CR against CAP, and the other potential pollutants (KNM, SPM, and CTF) had CR values lower than 0.01. The specificity of the USICA and the TICA systems for AFM1 was also evaluated by CR. Results show that the structural analogs aflatoxin M2 (AFB1, CR = 58% in USICA), aflatoxin G2 (AFG2, CR = 55% in USICA), and aflatoxin B2 (AFG2, CR = 20% in USICA) had obvious CR against AFM1 and that the CR values of aflatoxin G1 (AFG1), aflatoxin B1 (AFB1), deoxynivalenol (DON), fumonisin B1 (FMB1), and zearalenone (ZEN) were negligible (< 0.01%). These results suggest that the USICA and the TICA system for CAP and AFM1 are specific. Although specific, analogs might fake the presence of the analyte, especially during the analysis of complex samples. Some specificity data were usually demonstrated for reference in actual use (Table 1).

Fig. 3
figure 3

Cross-reactivity of TICA and USICA to analogs. A negative control (5% methanol in standard solution) and analogs, including TAP, FF, KNM, SPM, and CTF spiked at 1.25 ng·mL−1 for CAP (a) and AFG2, AFM2, AFB2, AFB1, AFG1, DON, ZEN, and FMB1 spiked at 5 ng·mL−1 for AFM1 (b)

Full size image

The accuracy and precision of the USICA and the TICA systems were evaluated through the recovery of the intra- and inter-assay, and the results are summarized in Table 2. The analysis was carried out in triplicate at each spiked concentration. The intra-assay was completed within 1 day, and the inter-assay was performed continuously for 3 days. The average recoveries of the two ICA systems ranged from 85% to 116%, which is acceptable for ICA quantitative analysis [37].

Table 2 Recovery of the USICA and TICA systems for CAP- and AFM1-spiked samples
Full size table

Comparative study with a commercial ELISA kit

The performances of the USICA and the TICA systems for CAP or AFM1 were compared with that of the commercial ELISA kit to demonstrate reliability and practicability. For the USICA and the TICA systems, the spiked milk samples were pretreated using trichloroacetic acid, which was mentioned in the “quantitative procedure of the USICA and the TICA systems” section. For the ELISA, the sample pretreatment was performed according to the manufacturer’s instructions.

The results in Fig. 4 show that the USICA (R2 = 0.9854 for CAP and R2 = 0.9828 for AFM1) and the TICA (R2 = 0.9562 for CAP and R2 = 0.9664 for AFM1) systems were in good agreement with the ELISA method. In addition to their universality and high sensitivity, the USICA and the TICA systems were easier to operate and took no more than 40 min to complete one sample analysis. The traditional ELISA involved multiple incubation, washing, and coloring, which took nearly 90 min. The main limit of the established method is the increased susceptibility induced by the delicate preincubation. Each step should be performed exactly to address the side effect.

Fig. 4
figure 4

Correlation between the ICA systems (x-axis) and the ELISA (y-axis). The plot fits for CAP between USICA and ELISA (a) and between TICA and ELISA (b) and AFM1 between USICA and ELISA (c) and between TICA and ELISA (d). Eighteen blank samples were spiked with analyte standard solutions under different concentrations in each group

Full size image

Conclusion

The USICA system combines the advantages of sensitivity and universality. The USICA system is beneficial for quantitative analysis due to its potential to provide a uniform ICA strip for different analytes. Any analyte can be detected using the USICA system if analyte-specific mAb and BSA-modified analytes are available. The core merit of the USICA is its universality, and the susceptibility induced by the delicate preincubation may be reduced by standard operations.

The availability of the ICA was improved by the USICA system, paving the way for its wider application. On the basis of this work, sensitive and universal detection systems for other analytes may be further investigated.