A label-free colorimetric aptasensor based on controllable aggregation of AuNPs for the detection of multiplex antibiotics

Journal Pre-proofs A label-free colorimetric aptasensor based on controllable aggregation of AuNPs for the detection of multiplex antibiotics Yang-Yang Wu, Pengcheng Huang, Fang-Ying Wu PII: DOI: Reference:

S0308-8146(19)31491-8 https://doi.org/10.1016/j.foodchem.2019.125377 FOCH 125377

To appear in:

Food Chemistry

Received Date: Revised Date: Accepted Date:

20 December 2018 12 July 2019 17 August 2019

Please cite this article as: Wu, Y-Y., Huang, P., Wu, F-Y., A label-free colorimetric aptasensor based on controllable aggregation of AuNPs for the detection of multiplex antibiotics, Food Chemistry (2019), doi: https://doi.org/ 10.1016/j.foodchem.2019.125377

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier Ltd.

A label-free colorimetric aptasensor based on controllable aggregation of AuNPs for the detection of multiplex antibiotics Yang-Yang Wu, Pengcheng Huang, Fang-Ying Wu* College of Chemistry, Nanchang University, Nanchang, 330031, China

*Corresponding Author: Fang-Ying Wu, [email protected]. The first two authors contributed equally to this work. Tel: + 86 79183969882, Fax: + 86 79183969514. 1

Abstract: We devise a novel colorimetric aptasensor for multiplex antibiotics based on an ss-DNA fragment coordinately controlling gold nanoparticles (AuNPs) aggregation. The multifunctional aptamer (Apt) was elaborately designed to be adsorbed on AuNPs surfaces acting as a binding element for antibiotics and a molecular switch. Chloramphenicol (CAP) and tetracycline (TET) were selected as the model antibiotics. When one kind of antibiotics was added, the specifically recognized fragment of Apt can bind to it and dissociated, and the non-specific one coordinately controls AuNPs aggregation under high-salt conditions. Hence, different color changes of AuNPs solution can be used as the signal readout. The aptasensor exhibited remarkable selectivity and sensitivity for separate detection of TET and CAP, and the detection limits are estimated to be 32.9 and 7.0 nM, respectively. The analysis with the absorption spectroscopy and the smartphone are applied to detect antibiotics in real samples with consistent results and desirable recoveries. Keywords: Multiplex antibiotics, Colorimetric aptasensor, Controllable aggregation, Food safety

1. Introduction 2

Antibiotics such the broad-spectrum ones as tetracycline (TET) and chloramphenicol (CAP), are widely applied to treat bacterial infections and promote animal growth in livestock husbandry (Khoshbin, Verdian, Housaindokht, Izadyar, & Rouhbakhsh, 2018; Ma, Wang, Jia, & Xiang, 2018; Ramezani, Mohammad Danesh, Lavaee, Abnous, & Mohammad Taghdisi, 2015). In recent years, the overuse of antibiotics in human health, animal, ambient (including agriculture and aquaculture) is a great concern and leads to antimicrobial resistance through bioaccumulation (Caniça, Manageiro, Abriouel, Moran-Gilad, & Franz, 2018; Gai, Gu, Hou, & Li, 2017; Ge, Li, Du, Zhu, Chen, Shi, et al., 2018). Additionally, the long-term introduction of antibiotics into the water also poses a serious threat to environmental health (Ashbolt, Amezquita, Backhaus, Borriello, Brandt, Collignon, et al., 2013; J. Zhou, Nie, Chen, Yang, Gong, Zhang, et al., 2018). Therefore, official agencies around the world clearly declare the maximum limit of antibiotics residues in agriculture and animal products (S. Wang, Dong, & Liang, 2018; C. Yan, Zhang, Yao, Xue, Lu, Li, et al., 2018). Consequently, developing a simple, rapid and reliable method for visual determining antibiotic residues as a way of ensuring food safety attracts great attention worldwide. In recent years, some approaches to detecting antibiotic residues in food products have been published, such as fluorescence method (Tang, Gu, Wang, Song, Zhou, Lou, et al., 2018; Y. Wang, Gan, Zhou, Li, Cao, & Chen, 2017), enzyme-linked immunosorbent assay (Z. Wang, Mi, Beier, Zhang, Sheng, Shi, et al., 2015), electrochemical analysis (Gai, Gu, Hou, & Li, 2017; Liu, Wang, Xu, Leng, Wang, Guo, et al., 2017), high-performance liquid chromatography (HPLC) (Saxena, Rangasamy, Krishnan, Singh, Uke, Malekadi, et al., 2018), etc. Despite high sensibility, these methods suffer from such drawbacks as costly instruments, complicated operation procedures. Moreover, they are unsuited for on-site real-time detection which is highly desired for the prompt monitoring of multiplex harmful targets. Therefore, it is essential to design a novel antibiotic residues sensing platform, featuring simple construction, satisfied with highly sensible and effective detection in actual samples (Lin, Yan, Guo, Cao, Yu, Zhang, et al., 2018; Z. Yan, Gan, Wang, Cao, Chen, Li, et al., 2015). To meet this requirement, biosensors regarded as promising tools can detect various targets by converting a physically measurable signal. Aptamers (Apt), as candidate to antibodies, are specific short

single-stranded

oligonucleotides that are selected via a process termed SELEX (systematic evolution of ligands by 3

exponential enrichment) (Famulok, M., Hartig, J. S., and Mayer, G, 2007). They are a novel molecular recognition element that can bind to targets with excellent affinity and specificity (Famulok, M.; Mayer, G, 2011). Moreover, they still have some unique features of being simple synthesis, easy modification, excellent chemical stability, and high flexibility (Phanchai, Srikulwong, Chompoosor, Sakonsinsiri, & Puangmali, 2018). Compared to traditional recognition molecules, such as antibodies, aptamers are more ideal molecular receptors and sensing elements for constructing sensing platforms (Hizir, Top, Balcioglu, Rana, Robertson, Shen, et al., 2016; Jiang, Shi, Liu, Wan, Cui, Zhang, et al., 2017). Base on the aforementioned developments, aptasensors of various performances are excavated, such as quartz-crystal microbalance, surface plasmon resonance, fluorescence, electrochemistry and colorimetry (Khoshbin, Verdian, Housaindokht, Izadyar, & Rouhbakhsh, 2018; W. Zhou, Saran, & Liu, 2017). Among them, gold nanoparticles (AuNPs) based colorimetric aptasensors have fascinated interests in the visual detection of relevant analytes in various fields, because of their advantages like unique optical properties and extremely high extinction coefficients (Deng, Liu, Zhang, Deng, Lei, Shen, et al., 2018; Guo, Zhang, Shao, Wang, Wang, & Jiang, 2014). However, unmodified AuNPs can be easily aggregated and flocculated induced by the passivating surface layer, ionic strength, pH, and temperature (Mirau, Smith, Chavez, Hagen, Kelley-Loughnane, & Naik, 2018; Park & Shumaker-Parry, 2014). The use of thiol groups of ssDNA covalently attached to AuNPs provides a high surface density to effectively prevent AuNPs aggregation in high ionic strength (Zhao, et al., 2008; Epanchintseva, Vorobjev, Pyshnyi, & Pyshnaya, 2018). Unfortunately, this classical strategy inevitably costs high and necessitates complex operation. Recent years have witnessed the critical role of aptamers with the random coil structure in stabilizing the AuNPs under salt conditions by means of the nucleobases via nonspecific electrostatic adsorption to the surface of AuNPs (Li, H; Rothberg, Lewis, 2004. ; Li, Xi, Kong, Liu, & Chen, 2018; W. Wang, Ding, He, Wang, & Lou, 2014). Simultaneously, a number of research groups have reported that the strong noncovalent adsorption of the ss-DNA to AuNPs surfaces can be affected by the sequence lengths (Mirau, Smith, Chavez, Hagen, Kelley-Loughnane, & Naik, 2018; Zhang, Servos, & Liu, 2012). They demonstrated that the shorter ss-DNA possessed a stronger binding affinity and better kinetics mainly because it had a higher surface density and was more conducive to the stability of AuNPs in high salt concentrations (Alsager, Kumar, Zhu, Travas-Sejdic, McNatty, & Hodgkiss, 2015; 4

Epanchintseva, Vorobjev, Pyshnyi, & Pyshnaya, 2018; Suzuki, K.; Hosokawa, K.; Maeda, M, 2009). Therefore, a novel colorimetric sensing platform may be provided based on unique interaction between AuNPs and ss-DNA with the fragments of different lengths. Herein, by bringing into full play the unique properties that both aptamer and AuNPs sensors have to offer, we designed a multifunctional colorimetric aptasensor platform that recognizes multiplex antibiotics based on the assembly of AuNPs and Apt. The Apt contains two recognition fragments with the unequal lengths where the length recognition fragment for TET is a fifth of that for CAP (Kwon, Ahmad Raston, & Gu, 2014; Mehta, Van Dorst, Rouah-Martin, Herrebout, Scippo, Blust, et al., 2011). Apt served as a molecular switch, which avoided surface modification and mediated controllable AuNPs aggregation in high ionic system. Simultaneously, the AuNPs solution also showed color changes as the signal readout. Compared with long fragment (CAP aptamer), the short one (TET aptamer) possesses a stronger capacity to increase the surface density of AuNPs so that AuNPs will be more stable in high-salt conditions. Accordingly, distinguishing color changes of AuNPs solution caused by TET and CAP can be realized. Therefore, we justify the ultrasensitive colorimetric label-free aptasensor that significantly improved the salt tolerance of AuNPs and achieved naked eye analysis of multiplex antibiotics. It is believed that the straightforward means and unique aptasensor design concept would show great potential in advance automated and convenient detection for multiplex targets.

2. Experimental 2.1. Chemicals and Instruments Apt (5′-ACTTCAGTGAGTTGTCCCACGGTCGGCGAGTCGGTGGTAGCGGTGGTG-3′) was synthesized and HPLC-purified by Shanghai Sangon Biotechnology Co. Ltd. (Shanghai, China). Apt was diluted to 5 μM in 10 mM phosphate buffered solution (PBS, NaH2PO4-Na2HPO4) for use. CAP, TET, oxytetracycline (OTC), chlortetracycline (CTC), metronidazole, ciprofloxacin and amoxicillin were purchased from Aladdin (Shanghai, China). Chloroauric acid (HAuCl4·4H2O), citric acid, glucose and sucrose were obtained from Shanghai Chemical Reagent Co. Ltd. (Shanghai, China). Some amino acids, including L-serine, L-aspartic acid, histidine, L-threonine, tryptophan, were purchased from Shanghai Jingchun Technology Co. Ltd. (Shanghai, China). Ionic salts were purchased analytical grade from Sinopharm Chemical Reagent Company 5

(Beijing, China). Ultrapure water purified by a Milli-Q system was used throughout all experiments. Transmission electron microscopy (TEM) images were obtained by a JEOL JEM-2100. Dynamic light scattering (DLS) and potential analysis were conducted on a Zeta-sizer Nano-ZS90 zeta and size analyzer from MicrotracInc, USA. UV-vis absorption spectra were achieved using UV-2550 spectrophotometer (Shimadzu, Kyoto, Japan) with a 1.0 cm quartz cell at room temperature. 2.2. Preparation of AuNPs All the glassware were soaked by freshly prepared aqua regia (HNO3/HCl; 1:3) and rinsed in ultrapure water thoroughly and dried. We prepare the AuNPs via sodium citrate reduction according to our previous report (Huang, Gao, Li, & Wu, 2018). In brief, trisodium citrate (10.3 mL, 38.8 mM) was rapidly added to a boiling solution of HAuCl4 (10.3 mL, 4.0 g ·L−1). The color changed from pale yellow to purple in 5.0 min, until it became wine red in the end. After about 30 min, the heating stopped. The solution was stirred continuously about 15 min until it slowly cooled to room temperature. Then, the acquired solution was centrifuged at 12, 000 rpm for 20 min at 4 °C, and the resultant AuNPs were re-dispersed and filtered with 0.22 μm ultrafiltration membranes. Finally, the solution was diluted and stored in a refrigerator at 4 °C for further use. The concentration of AuNPs was calculated to be 11.5 nM based on Lambert Beer’s law where the extinction coefficient (ε) at λ = 520 nm for 13 nm AuNPs is 2.7 × 108 M−1 cm−1 (Huang, Gao, Li, & Wu, 2018). 2.3. Quantitative procedure The analysis was applied in 10.0 mM PBS and it was implemented based on the following procedures. Briefly, 88 μL of 11.5 nM AuNPs and 7.0 μL of 5.0 μM Apt solution were added into a centrifuge tube to form the Apt@AuNPs. A range of target concentrations were incubated for 6.0 min. Finally, the optimal NaCl concentration was rapidly added and the buffer solution was used to ensure ultimate volume of 500 μL. Photographic images and UV–vis Spectrophotometer were recorded 10.0 min later at room temperature. The absorbance ratio (A620/A520 for CAP and A650/A520 for TET) was employed to reflect the aggregation degree for AuNPs. 6

The limit of detection evaluation process was consistent with the above description. Ten blank samples in the absence of the target were measured in 10.0 mM PBS. It was calculated based on Eq. (1) as described by the International Union of Pure and Applied Chemistry (IUPAC) LOD = 3 × S0/K

(1)

where S0 is the standard deviation of blank measurements (n = 10) and K is the slope of calibration line.

2.4. Preparation of real samples Chicken and milk samples in a local supermarket were purchased to investigate the application of our strategy. The sample was treated before accurate analysis (Lin, Yu, Cao, Guo, Zhu, Dai, et al., 2018). Fresh chicken was triturated evenly by the crusher. Afterwards, 2.0 g of chicken or milk samples were added to 2.0 mL of 10 mM PBS (pH = 6.0). The mixture was ultrasonicated for 30 min and then centrifuged at 12 000 rpm (10 °C) for 10 min to eliminate the surplus solid. After that, proteins were precipitated by mixing with 100 μL of acetocaustin (50%) in the supernatant, centrifuged again and filtered with a 0.22 μm filter membrane. The filtrate was transferred to a flask and diluted to 5 mL for further analysis as the food sample. Three different concentrations (0.2, 0.4 and 1.0 μM) of target antibiotic were spiked into the pretreated real samples. Then, 100 μL of the above mixture was pipetted into 10 mM PBS (pH = 7.5). The specific detection procedure similar to that described in section 2.3 was conducted. And all experiments were performed in triplicate.

3. Results and discussion 3.1. Design of the sensory principle The strategy for the colorimetric sensor detection of CAP and TET is illustrated in Scheme 1. The citrate stabilized AuNPs were highly dispersed in the aqueous solution because of electrostatic stabilization with the negative charge of citrates. Thus, the solution showed a strong absorbance at 520 nm wavelength (Fig. 1A, curve a) and a wine-red color (Figure 1A, inset a). It should also be noted that neither CAP nor TET could induce the aggregation of AuNPs (Fig. S1). 7

When Apt was added into AuNPs solution, it would be easily adsorbed on the surface of AuNPs to prohibit electrostatic screening effect of NaCl. This is because the ss-DNA unfolded sufficiently to expose

positively

charged

nucleobases,

which

facilitated

its

adsorption

onto

the

negatively-charged AuNPs surfaces through electrostatic attraction. Upon the addition of NaCl, AuNPs remained the dispersion state and the original color similar to bare AuNPs against salt-induced aggregation (Fig. 1A, curve b, inset b). In contrast, in the presence of the target (TET or CAP), the recognition fragment of Apt dissociated from the surfaces of AuNPs and strongly bound to the target to form the folded rigid structure. Simultaneously, the non-recognition fragment of Apt was unable to maintain the stability of the AuNPs leading to aggregation in the high salt conditions. The length of non-recognition fragment is critical to aggregation kinetics and sensor response. By comparison with the long fragment, the short one possesses a stronger capacity to the resistance of AuNPs to the salt-induced aggregation. Specifically, when adding CAP into the system, the short fragment was absorbed which caused slight aggregation of AuNPs in optimal high salt conditions (Fig. 1A, curve c) accompanied with a color change of the solution from wine red to purple (Fig. 1A, inset c). On the contrary, when TET was added into the system and bundled by Apt, the longer fragment did not effectively stabilize the AuNPs thus leading to serious aggregation in the same optimal high salt conditions (Fig. 1A, curve d). The solution color changed from wine red to blue (Fig. 1A, inset d). The state change of AuNPs enables the naked-eye readout and measurable signals by UV-vis spectroscopy.

8

Scheme 1. Schematic illustration of the detection TET/CAP based on AuNPs colorimetric aptasensors. The Apt acts as a molecular switch adjusting the AuNPs aggregation. When antibiotics remove the fragment of Apt from the AuNPs surface, unbalanced AuNPs was aggregated of different scales under high-salt conditions. It thereby causes colloidal color changes, which can be detected by UV-spectrum and Smartphone analysis, respectively.

To verify the feasibility of mechanism, we employ DLS and TEM to characterize the particle size changes of AuNPs. The TEM image showed that the prepared AuNPs were nearly monodisperse spheres with hydrodynamic diameter about 12.0 nm (Fig. S2A), and Apt@AuNPs has nearly the same dispersed state in high salt environments (Fig. S2B). However, the size of AuNPs changed obviously from original status to around 342.0 and 886.4 nm at the optimal salt concentration treated by CAP and TET, respectively, as displayed in Fig. 1B and C, which also justifies the possibility of using Apt for sensing multiplex antibiotics with sensitive analysis. Zeta-potentials were also employed to confirm the surface potential change of AuNPs (Fig. 1D). The zeta-potential of bare AuNPs was -88.4 mV. When Apt was incubated with AuNPs for 10 min, the zeta-potential were reduced to -27.6 mV, in which Apt was easily combined with AuNPs to neutralize the negative charges of citrates. While CAP was added into the system and incubated for another 6 min, the zeta-potential was increased to -42.31 mV, which is because the recognition fragment of Apt for CAP was desorbed. Compared with the long fragment (CAP aptamer), the short one (TET aptamer) possesses a stronger capacity to increase the surface density of AuNPs, therefore, when only TET was introduced to the probe system, the zeta-potential was again increased to -79.48 mV as the recognition fragment for TET was released. Overall, the feasibility of the mechanism was confirmed.

9

Fig. 1. (A) absorption spectra of AuNPs solutions, (a) bare AuNPs, (b) Apt@AuNPs with salt, (c) CAP and (d) TET present in the detection system. Inset: visual color changes corresponding to the spectra. (B) and (C) show TEM images of Apt@AuNPs treated with 1.5 μM CAP and TET. Inset: DLS corresponding to the aggregation of AuNPs TEM images. (D) Zeta potential changes. (a) AuNPs; (b) Apt@AuNPs; Apt@AuNPs incubated with (c) 1.5 μM CAP and (d) 3.0 μM TET. The error bar represents the standard deviation of three repetitive measurements.

3.2. Optimization of the sensing conditions A series of systematic optimization experiments were executed to obtain the optimal assay performances. We used A and A0 to express the presence and the absence of target antibiotics respectively. We recorded the states of AuNPs at various concentrations of NaCl. As shown in Fig. S3A, the highest A/A0 ratio was found in the presence of 90 mM NaCl. Then, we explored the influence of Apt concentration to ensure high sensitivity for antibiotics detection (Fig. S3B). The A/A0 ratio rose up more and more steeply and presented the highest value when 70 nM Apt, so such a concentration of Apt was chosen in the future experiment. The AuNPs concentration is also 10

an important factor effecting color changes. As displayed in Fig. S4, the best response to the targets was obtained when AuNPs concentration was 2.0 nM. We also take into account the effect of reaction time and the pH value. As shown in Fig. S5A and B, 6 min and pH 7.5 were considered as best experimental conditions to achieve accurate detection. And the optimum detection time of AuNPs was obtained at 10 min (Fig. S6). 3.3. Sensitivity of the aptasensor for antibiotics detection Under optimal sensing conditions, we investigated the sensitivity of our assay toward two antibiotics. The characteristic absorption bands of the AuNPs were measured with UV–Vis absorption spectra, in which different concentrations of CAP and TET were quantitatively evaluated (Fig. 2). It can be seen that as the concentration of antibiotics increased, the absorbance ratio (A620/A520 for CAP and A650/A520 for TET) gradually increased. For CAP detection，the linear regression equations at low and high concentrations are A620/A520 = 0.1747 − 0.3364 C1 and A620/A520 = 0.8456 C2 − 0.2185 (Fig. 2B, a, b), respectively, with the corresponding correlation coefficients of 0.9945 and 0.9977. The limit of detection (LOD), according to Eq. (1), was calculated to be 7.0 nM in the range of 0.05 − 1.8 μM. Despite the LOD of CAP could not be lower than the maximum residue limits of CAP about 0.928 nM defined by European Union, the developed aptasensor displayed unique advantages, such as label-free, simple principle and great tolerance to high salt concentrations. Concurrently, for TET detection，it showed a good linear correlation in the range of 0.05 − 3.0 μM (Fig. 2B, d, e). The linear regression equation was: A650/A520 = 0.2761 C + 0.1268 with a correlation coefficient of 0.9961. The LOD was calculated to be 32.9 nM according to Eq. (1), which was lower than the maximum residue limits (about 225 nM) defined by European Commission. As shown in Table S1, the most reported approaches are time-consuming and complicated, and have higher LODs. The unique advantages of our approach were displayed, such as simple operation, better sensitivity and the detection capability for multiplex antibiotics. Correspondingly, heat maps were employed to amplify color recognition (Fig. 2, c, f). These results visually displayed the trend of the color change with the gradual increment of antibiotics, which provides the extreme significance for development of the novel molecular profiling.

11

Fig. 2. (A ) and (B) show the sensitivity of the aptasensor and a heat map for antibiotics detection in solution. (a) and (d) present the changes of UV-vis absorption spectra in the presence of CAP and TET at various concentrations, respectively; The relationship between the absorbance ratio (A620/A520 for CAP and A650/A520 for TET) and the concentration of CAP (b) and TET (e); Heat maps of CAP (c) and TET (f) highlighting the profiling and comparison of target antibiotics for color recognition; Each measurement was replicated for three times; the error bars represent standard deviations.

3.4 Selectivity of the proposed aptasensor

12

The selectivity is also other vital features to evaluate our proposed aptasensor’s performance in practical applications. The selectivity experiment was carried out in the simulated environment of real sample. Thus, another five common antibiotics, OTC, CTC, metronidazole, amoxicillin, and ciprofloxacin, were employed as the negative controls to validate the specificity of this aptasensor platform. Fig. 3A and B show only the presence of the CAP and TET caused obvious differences in the absorbance ratio value. It was apparent that the non-ignorable interference from OTC and CTC for the detection of TET as they have similar structures to TET. However, they were hardly added simultaneously in drugs. We also explored potential interference from the common substances usually contained in food samples including ions, amino acids, carbohydrates, and food additives. As shown in Fig. 3C and D, even when the concentration of K+, CH3COO-, L-serine (Ser), L-threonine (Thr), L-aspartic acid (Asp), tryptophan (Trp), histidine (His), sucrose (Suc) and glucose (Glc) was 1000 times of antibiotics concentration, the concentration of Ba2+ and SO42- was 100 times, and the concentration of Mg2+ and Ca2+ was 20.0 times, the absorbance ratio value had only slight changes. Meanwhile, when 3.0 μM of TET or 1.5 μM of CAP and the interference coexisted, the absorbance ratio value increased obviously and reached the similar level to that upon sole addition of TET or CAP. These observations substantially demonstrated that the proposed colorimetric aptasensor exhibits an excellent discriminating ability for targets CAP or TET even in very complicated food samples.

13

Fig. 3. The selectivity of the aptasensor for CAP (A) and TET (B) analysis in aqueous solution. The concentrations of OTC, CTC, metronidazole, amoxicillin, and ciprofloxacin were all 3.0 μM. (C) and (D) show the interference of the coexisting ions and compounds. The concentration of antibiotic was 3 μM, the concentrations of K+, CH3COO-, amino acids and sugars were 3.0 mM, the concentration of Ba2+ and SO42- were 300 μM, and the concentration of Mg2+ and Ca2+ were 60 μM. Each measurement was replicated for three times; the error bars represent standard deviations.

3.5 Antibiotics detection in real samples To justify the potential of our method in point-of-care testing (POCT) assay in adopting general devices, we performed the automated signal readout through RGB analysis for the photos of the final AuNPs solutions with the color scanner APP (Color Grab) installed in a smartphone. As depicted in Fig. 4A, as the concentration of targets increases, the color of the solution gradually changes from red to purple or blue, indicating that the blue/red (B/R) value of each image was closely associated with the concentration of antibiotics. As shown in Fig. 4B and C, when the concentration of target antibiotics ranged from 0 to 3.0 μM, the linear regression equations for CAP and TET were B/R = 0.569 + 0.445 C and B/R = 0.585 + 0.243 C respectively, 14

with the correlation coefficients of 0.995 and 0.993. Therefore, it can be concluded that this strategy with smartphone analysis is very simple in operation and has a wide linear range. Consequently, the concentrations of antibiotics in the blank and spiked samples were determined by using RGB analysis based on the smartphone device, and the analysis results were compared with those by UV–Vis absorption spectra analysis method to check the precision and accuracy. Table S2 and S3 clearly depicted that the determination results of various spiked samples by RGB analysis were in good agreement with those by the UV–Vis absorption spectra analysis. The recoveries of 96.11 - 109.88% and the RSD (n = 3) of 1.06 - 5.57% were achieved, which further proved that the practical use of the RGB analysis were good.

Fig. 4. (A) The RGB analysis for the color changes of AuNPs solution corresponding to TET and CAP using the smartphone. The liner relationship between the B/R values in RGB analysis and the antibiotics concentrations ((B) CAP, (C) TET). N = 10 indicates the number of data collections per group.

4. Conclusion In conclusion, we have successfully developed a colorimetric platform for the detection of multiplex antibiotics by using Apt mediated controllable label-free AuNPs aggregation. When one target is separately added into the detection system, the specifically recognized fragment of Apt binds to it and then dissociated, while the non-specific one coordinately controls AuNPs 15

aggregation under high-salt conditions. Since the shorter fragment of Apt has a stronger capability to prevent the AuNPs aggregation than the longer one in high salt conditions, distinct color changes could be observed for multiplex antibiotics (e.g., CAP and TET in this study). Different from traditional colorimetric aptasensors for antibiotics, this strategy bears some significant advantages. For example, it does not require any chemical modification and displays great tolerance to high salt concentrations, which allows for multiplex antibiotics detection by the naked eye. Moreover, the novel method is simple, efficient, reliable and free of sophisticated instruments, which would promote the practical application in the immediate detection via RGB analysis by using the smartphone. It thus has great potential to be applied in on-site detection in the field of food safety.

Acknowledgments This work was supported by the National Natural Science Foundation of China (Nos. 21765014 and 21864018); and the Opening Project of Guangzhou Key Laboratory of Analytical Chemistry for Biomedicine (No. 2018001). Appendix A. Supplementary data Supplementary data to this article can be found online at http:// Notes The authors declare no competing financial interest.

References Alsager, O. A., Kumar, S., Zhu, B., Travas-Sejdic, J., McNatty, K. P., & Hodgkiss, J. M. (2015). Ultrasensitive colorimetric detection of 17beta-estradiol: the effect of shortening DNA aptamer sequences. Analytical Chemistry, 87(8), 4201-4209. Ashbolt, N. J., Amezquita, A., Backhaus, T., Borriello, P., Brandt, K. K., Collignon, P., Coors, A., Finley, R., Gaze, W. H., Heberer, T., Lawrence, J. R., Larsson, D. G., McEwen, S. A., Ryan, J. J., Schonfeld, J., Silley, P., Snape, J. R., Van den Eede, C., & Topp, E. (2013). Human Health Risk Assessment (HHRA) for environmental development and transfer of antibiotic resistance. Environmental Health Perspectives, 121(9), 993-1001. Caniça, M., Manageiro, V., Abriouel, H., Moran-Gilad, J., & Franz, C. M. A. P. (2018). Antibiotic resistance in foodborne bacteria. Trends in Food Science & Technology. DOI.org/10.1016/j.tifs.2018.08.001 Deng, C., Liu, H., Zhang, M., Deng, H., Lei, C., Shen, L., Jiao, B., Tu, Q., Jin, Y., Xiang, L., Deng, W., Xie, Y., & Xiang, J. (2018). Light-Up Nonthiolated Aptasensor for Low-Mass, Soluble Amyloid-beta40 Oligomers at High Salt Concentrations. Analytical Chemistry, 90(3), 1710-1717. Epanchintseva, A., Vorobjev, P., Pyshnyi, D., & Pyshnaya, I. (2018). Fast and Strong Adsorption of Native 16

Oligonucleotides on Citrate-Coated Gold Nanoparticles. Langmuir, 34(1), 164-172. Famulok, M., Hartig, J. S., and Mayer, G. (2007), Functional aptamers and aptazymes in biotechnology, diagnostics, and therapy. Chemical Reviews, 107, 3715– 3743. Famulok, M.; Mayer, G. (2011). Aptamer Modules as Sensors and Detectors. Accounts of Chemical Research, 44, 1349– 1358. Gai, P., Gu, C., Hou, T., & Li, F. (2017). Ultrasensitive Self-Powered Aptasensor Based on Enzyme Biofuel Cell and DNA Bioconjugate: A Facile and Powerful Tool for Antibiotic Residue Detection. Analytical Chemistry, 89(3), 2163-2169. Ge, L., Li, H., Du, X., Zhu, M., Chen, W., Shi, T., Hao, N., Liu, Q., & Wang, K. (2018). Facile one-pot synthesis of visible

light-responsive

BiPO4/nitrogen

doped

graphene

hydrogel

for

fabricating

label-free

photoelectrochemical tetracycline aptasensor. Biosensors and Bioelectronics, 111, 131-137. Guo, Y., Zhang, Y., Shao, H., Wang, Z., Wang, X., & Jiang, X. (2014). Label-free colorimetric detection of cadmium ions in rice samples using gold nanoparticles. Analytical Chemistry, 86(17), 8530-8534. Hizir, M. S., Top, M., Balcioglu, M., Rana, M., Robertson, N. M., Shen, F., Sheng, J., & Yigit, M. V. (2016). Multiplexed Activity of perAuxidase: DNA-Capped AuNPs Act as Adjustable Peroxidase. Analytical Chemistry, 88(1), 600-605. Huang, P.-C., Gao, N., Li, J.-F., & Wu, F.-Y. (2018). Colorimetric detection of methionine based on anti-aggregation of gold nanoparticles in the presence of melamine. Sensors and Actuators B: Chemical, 255, 2779-2784. Jiang, Y., Shi, M., Liu, Y., Wan, S., Cui, C., Zhang, L., & Tan, W. (2017). Aptamer/AuNP Biosensor for Colorimetric Profiling of Exosomal Proteins. Angewandte Chemmie International Edition, 56(39), 11916-11920. Khoshbin, Z., Verdian, A., Housaindokht, M. R., Izadyar, M., & Rouhbakhsh, Z. (2018). Aptasensors as the future of antibiotics test kits-a case study of the aptamer application in the chloramphenicol detection. Biosensors and Bioelectronics, 122, 263-283. Kwon, Y. S., Ahmad Raston, N. H., & Gu, M. B. (2014). An ultra-sensitive colorimetric detection of tetracyclines using the shortest aptamer with highly enhanced affinity. Chemical Communications, 50(1), 40-42. Li, H; Rothberg, Lewis, (2004). Colorimetric detection of DNA sequences based on electrostatic interactions with unmodified gold nanoparticles. Proceedings of the National Academy of Sciences of the United States of America, 101(39), 14036-14039. Li, J., Xi, H., Kong, C., Liu, Q., & Chen, Z. (2018). "Aggregation-to-Deaggregation" Colorimetric Signal Amplification Strategy for Ag(+) Detection at the Femtomolar Level with Dark-Field Microscope Observation. Analytical Chemistry, 90(19), 11723-11727. Lin, B., Yan, Y., Guo, M., Cao, Y., Yu, Y., Zhang, T., Huang, Y., & Wu, D. (2018). Modification-free carbon dots as turn-on fluorescence probe for detection of organophosphorus pesticides. Food Chemistry, 245, 1176-1182. Lin, B., Yu, Y., Cao, Y., Guo, M., Zhu, D., Dai, J., & Zheng, M. (2018). Point-of-care testing for streptomycin based on aptamer recognizing and digital image colorimetry by smartphone. Biosensors and Bioelectronics, 100, 482-489. Liu, S., Wang, Y., Xu, W., Leng, X., Wang, H., Guo, Y., & Huang, J. (2017). A novel sandwich-type electrochemical aptasensor based on GR-3D Au and aptamer-AuNPs-HRP for sensitive detection of oxytetracycline. Biosensors and Bioelectronics, 88, 181-187. Ma, Q., Wang, Y., Jia, J., & Xiang, Y. (2018). Colorimetric aptasensors for determination of tobramycin in milk and chicken eggs based on DNA and gold nanoparticles. Food Chemistry, 249, 98-103. Mehta, J., Van Dorst, B., Rouah-Martin, E., Herrebout, W., Scippo, M.-L., Blust, R., & Robbens, J. (2011). In vitro 17

selection and characterization of DNA aptamers recognizing chloramphenicol. Journal of Biotechnology, 155(4), 361-369. Mirau, P. A., Smith, J. E., Chavez, J. L., Hagen, J. A., Kelley-Loughnane, N., & Naik, R. (2018). Structured DNA Aptamer Interactions with Gold Nanoparticles. Langmuir, 34(5), 2139-2146. Park, J. W., & Shumaker-Parry, J. S. (2014). Structural study of citrate layers on gold nanoparticles: role of intermolecular interactions in stabilizing nanoparticles. Journal of the American Chemical Society, 136(5), 1907-1921. Phanchai, W., Srikulwong, U., Chompoosor, A., Sakonsinsiri, C., & Puangmali, T. (2018). Insight into the Molecular Mechanisms of AuNP-Based Aptasensor for Colorimetric Detection: A Molecular Dynamics Approach. Langmuir, 34(21), 6161-6169. Ramezani, M., Mohammad Danesh, N., Lavaee, P., Abnous, K., & Mohammad Taghdisi, S. (2015). A novel colorimetric triple-helix molecular switch aptasensor for ultrasensitive detection of tetracycline. Biosensors and Bioelectronics, 70, 181-187. Saxena, S. K., Rangasamy, R., Krishnan, A. A., Singh, D. P., Uke, S. P., Malekadi, P. K., Sengar, A. S., Mohamed, D. P., & Gupta, A. (2018). Simultaneous determination of multi-residue and multi-class antibiotics in aquaculture shrimps by UPLC-MS/MS. Food Chemistry, 260, 336-343. Suzuki, K.; Hosokawa, K.; Maeda, M. (2009). Controlling the number and positions of oligonucleotides on gold nanoparticle surfaces. Journal of the American Chemical Society, 131, 7518– 7519. Tang, Y., Gu, C., Wang, C., Song, B., Zhou, X., Lou, X., & He, M. (2018). Evanescent wave aptasensor for continuous and online aminoglycoside antibiotics detection based on target binding facilitated fluorescence quenching. Biosensors and Bioelectronics, 102, 646-651. Wang, S., Dong, Y., & Liang, X. (2018). Development of a SPR aptasensor containing oriented aptamer for direct capture and detection of tetracycline in multiple honey samples. Biosensors and Bioelectronics, 109, 1-7. Wang, W., Ding, X., He, M., Wang, J., & Lou, X. (2014). Kinetic adsorption profile and conformation evolution at the DNA-gold nanoparticle interface probed by dynamic light scattering. Analytical Chemistry, 86(20), 10186-10192. Wang, Y., Gan, N., Zhou, Y., Li, T., Cao, Y., & Chen, Y. (2017). Novel single-stranded DNA binding protein-assisted fluorescence aptamer switch based on FRET for homogeneous detection of antibiotics. Biosensors and Bioelectronics, 87, 508-513. Wang, Z., Mi, T., Beier, R. C., Zhang, H., Sheng, Y., Shi, W., Zhang, S., & Shen, J. (2015). Hapten synthesis, monoclonal antibody production and development of a competitive indirect enzyme-linked immunosorbent assay for erythromycin in milk. Food Chemistry, 171, 98-107. Yan, C., Zhang, J., Yao, L., Xue, F., Lu, J., Li, B., & Chen, W. (2018). Aptamer-mediated colorimetric method for rapid and sensitive detection of chloramphenicol in food. Food Chemistry, 260, 208-212. Yan, Z., Gan, N., Wang, D., Cao, Y., Chen, M., Li, T., & Chen, Y. (2015). A "signal-on'' aptasensor for simultaneous detection of chloramphenicol and polychlorinated biphenyls using multi-metal ions encoded nanospherical brushes as tracers. Biosensors and Bioelectronics, 74, 718-724. Zhang, X., Servos, M. R., & Liu, J. (2012). Surface science of DNA adsorption onto citrate-capped gold nanoparticles. Langmuir, 28(8), 3896-3902. Zhao, W.; Chiuman, W.; Lam, J. C. F.; McManus, S. A.; Chen, W.; Cui, Y.; Pelton, R.; Brook, M. A.; Li, Y. (2008). DNA Aptamer Folding on Gold Nanoparticles: From Colloid Chemistry to Biosensors. Journal of the American Chemical Society, 130, 3610−3618. Zhou, J., Nie, W., Chen, Y., Yang, C., Gong, L., Zhang, C., Chen, Q., He, L., & Feng, X. (2018). Quadruplex gold immunochromatogaraphic assay for four families of antibiotic residues in milk. Food Chemistry, 256, 304-310. 18

Zhou, W., Saran, R., & Liu, J. (2017). Metal Sensing by DNA. Chemical Reviews, 117(12), 8272-8325.

19

Highlights

1.

A novel colorimetric aptasensor for the detection of multiple antibiotics based on an ss-DNA fragment coordinately controlling gold nanoparticles aggregation is presented.

2.

This colorimetric aptasensor allows for multiplex antibiotics detection by the naked eye.

3.

It does not require any chemical modification and displays great tolerance to high salt concentrations.

4.

The method is efficient, reliable and free of sophisticated instruments, which would promote the practical application in the immediate detection via RGB analysis of the digital images.

20