Introduction

Food safety issues caused by foodborne pathogens have become the focus of public health concerns all over the world [1]. As a highly pathogenic food-borne pathogenic bacterium, Staphylococcus aureus (S. aureus) can cause acute food poisoning [2]. Quickly and accurately identifying the pathogens throughout the food supply chain is a way to prevent and control foodborne disease outbreaks. Classical pathogen detection methods mainly include plate counting, enzyme-linked immune-sorbent assay (ELISA), polymerase chain reaction (PCR), and loop-mediated isothermal amplification (LAMP) [3,4,5]. However, they are restricted by time-consuming, labor-intensive, high detection limit, low automation, and complex sample pretreatment [6, 7].

In recent years, the emergence of nanotechnology opens the new horizon to overcome above obstacles. Some researchers have developed new bacterial isolation methods using magnetic beads (MB) to achieve the rapid isolation of target bacteria [8, 9]. It solves the problems of complex substrate and low bacterial concentration of food samples and realizes fast separation and efficient capture of target bacteria. Gold nanoparticles (Au NPs) have been favored by colorimetric researchers in recent years owing to their excellent chemical properties of size controlling, high biocompatibility, and long-term stability [10, 11]. However, the AuNPs-based colorimetric assay is usually susceptible to interference from ambient conditions (i.e., pH value, ion concentration, and temperature), thus affecting the sensitivity of bacterial detection [12, 13]. Copper (I) ions (Cu+) catalyzed click chemistry (alkyne-azide cycloaddition) can be selectively catalyzed by trace amount of Cu+, which provides a new way to improve the sensitivity of detection system. Therefore, due to the excellent properties of gold nanoparticles, the “click chemistry” reaction based on gold nanoparticles has attracted wide attention in recent years [14, 15]. In the presence of Cu+, a click reaction occurs between azide- and alkyne-functionalized AuNPs, resulting in the aggregation of nanoparticles and the color of the solution changing from red to blue [16]. In addition, Cu+ sources can be easily and inexpensively generated to reduce Cu2+ by ascorbic acid. Due to their high efficiency and selectivity, “click chemistry” has attracted more and more attention in colorimetric reactions [17, 18].

Inspired by this, we propose a colorimetric strategy based on the click chemistry indicator of AuNPs functionalized with alkyne and azide functional groups (alkyne and azide—AuNPs). In proposed method, the aptamer and ALP dual-labeled Fe3O4 NPs (IMB) serve as recognition and signal transduction element. The exogenous L-ascorbate-2-phosphate (AP) was first captured by ALP and then catalyzed to form to ascorbic acid (AA). Then, the Cu2+ in the solution was reduced to Cu+ by AA. The produced Cu+ triggered a click chemical reaction, causing color change recognized by the naked eyes. The binding of IMB to the bacterial surface prevents the substrate from entering the catalytically active site [19], inhibiting the colorimetric sensing of bacteria, which correlates with the concentration of S. aureus. Therefore, a colorimetric detection strategy with high sensitivity and specificity is proposed to quantify S. aureus.

Experimental section

Materials and reagents

Ferric chloride (FeCl3·6H2O), sodium citrate, and ethylene glycol were purchased from Tianjin Chemical Reagent Co., Ltd. (Tianjin, China) (http://5497614.1024sj.com). Chloroauric acid (HAuCl4·3H2O) was obtained from Aladdin Reagents Co., Ltd. (Shanghai, China) (https://www.aladdin-e.com/). CuSO4 and methanol were received from Beijing Chemical Reagent Co., Ltd. (Beijing, China) (http://www.crc-bj.com). N3-PEG-SH and CH-PEG-SH were purchased from Ponsure Company (Shanghai, China) (http://www.ponsure.com/). Poly(ethylene glycol) (PEG 6000), alkaline phosphatase (ALP), 2-phospho-L-ascorbic acid trisodium salt (> 95%, HPLC, AP), N-(3-(dimethylamino)-propyl)-N′-ethylcarbodiimide hydrochloride (EDC HCl), and N-hydroxysuccinimide (NHS) were bought from Sigma-Aldrich (http://www.sigmaaldrich.com). 5′-Amine-modified DNA aptamer (5′NH2—GCA ATG GTA CGG TAC TTC CTC GGC ACG TTC TCA GTA GCG CTC GCT GGT CAT CCC ACA GCT ACG TCA AAA GTG CAC GCT ACT TTG CTA A -3′) was selected according to literature [20] to identify S. aureus, which was synthesized by Sangon Biotech Co., Ltd. (Shanghai, China) (http://www.sangon.com). All other chemicals and reagents used were of analytical grade.

All the UV–Vis absorption spectra were collected using a TU-1810 DPC (PERSEE, Beijing, China). Transmission electron microscopy (TEM) images were measured on a JEOL JEM-2100F transmission electron microscope (Tokyo, Japan). Magnetic hysteresis loops were measured by a vibrating sample magnetometer (Lake Shore 7410 VSM). Fourier transform infrared spectra (FTIR) from 4000 to 500 cm−1 were recorded using a Nicolet 6700 FTIR spectrometer (Thermo Inc., USA). Zeta potential was carried out with a 90 Plus Zeta Nano Brook.

Culture of bacteria

Table S1 in ESM lists the strains used in the experiment. They were all preserved in 15% glycerol at − 80℃. Strains were resuscitated through scribing on Luria–Bertani (LB) agar plates. Each strain was cultivated by the conventional culture method and collected during the exponential growth period. The bacteria concentration was determined by conventional plate counting method. Various concentrations of S. aureus in series were obtained by gradient dilution and stored at 4℃ before use.

Synthesis of aptamer-functionalized ALP-labeled Fe3O4 NPs (IMB)

We prepared Fe3O4-COOH NPs through the previously reported method [21]. Aptamer-functionalized and ALP-labeled Fe3O4 NPs (IMB) were prepared with slight modification according to the method recorded in the literature [22]. Firstly, EDC and NHS (10 mg) were added to activate Fe3O4-COOH NPs (0.5 mg) for 0.5 h. Then, the mixture was collected by permanent magnet and washed three times with PBS buffer. 0.8 mL PBS buffer containing 1 mg ALP was added to the precipitate and gently rotated for 2 h. Subsequently, 0.1 μm S. aureus aptamer was put in and reacted overnight at gentle mixing. After washed with PBS, the IMB was finally suspended in PBS buffer and preserved at 4 ℃.

Synthesis of functionalized AuNPs

Referring to our previous work, the citric acid reduction method was used to synthesize AuNPs [23]. Alkyne- and azide-functionalized AuNPs solutions were prepared according to the literature with little modifications [24]. In brief, a solution of thiol-PEG-N3 (30 μL, 2 mM in 2:1 methanol/water) was added to the AuNPs solutions (0.5 mL). N3-PEG-AuNPs were obtained by rotating the mixture at room temperature for 12 h and centrifuging for 20 min. The obtained N3-PEG-AuNPs were finally re-dispersed in 2:1 methanol/water (0.25 mL). The CH-PEG-AuNPs were obtained by the same method, except that thiol-PEG-CH was replaced with thiol-PEG-N3. The N3-PEG-AuNPs and CH-PEG-AuNPs were mixed to obtain a uniform dispersed solution.

Optimization of experimental conditions

In order to maximize the capture efficiency of S. aureus, the concentration of IMB was optimized. S. aureus at 106 cfu mL−1 was taken as an example, and PBS was chosen as the control group. The S. aureus suspension were mixed with different concentration of IMB (0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6 mg·mL−1) for 0.5 h. After magnetic separation, the supernatant was spread on LB agar with appropriate dilution for colony count. Based on the plate count method [25], the capture efficiencies of S. aureus were obtained.

Our colorimetric method relies on the hydrolysis activity of alkaline phosphatase, so whether the ALP on the IMB surface has the ability to catalyze the hydrolysis of AP is the key point of the colorimetric reaction. According to the previous literature [24, 26], we determined that the detection mixture was composed of N3-PEG-AuNPs and CH-PEG-AuNPs and Cu2+ (1 mM). Based on the optimal amount of IMB, 0.5 mg IMB was incubated with 5 mM AP for 5 min. Then the introduction of detection mixture continued the color reaction. After 5 min, the absorption spectra of 400 to 900 nm were recorded by UV–Vis spectrophotometer.

S. aureus detection assay

Bacteria samples were prepared at different concentrations (10–106 cfu mL−1) using sterile PBS and freshly cultured inactivated bacteria. After optimization of experimental conditions, 0.1 mL of each concentration of S. aureus solution was incubated with a total of 0.9 mL PBS buffer containing 0.5 mg IMB. At the same time, sterile PBS without bacteria was used as a negative control. The suspension was then slowly shaken in a 1.5-mL plastic centrifuge tube at room temperature for 0.5 h. The solution containing AP (5 mM) was added to each tube after magnetic separation washing. Then, the produced AA was added into the detection mixture (comprising N3-PEG-AuNPs and CH-PEG-AuNPs and Cu2+). Finally, the absorption spectrum of the solution at 400 ~ 900 nm was recorded with UV–Vis spectrophotometer.

Real sample detection

Spiked water and food samples were used to test the suitability of this colorimetric method. The real samples were prepared according to the literature with little modification [27]. The preparation of pork homogenate was as follows: The pork used in the experiment was bought from the local market. Five grams of samples and 10 mL sterilized PBS were ground until homogenized. The collected lake water samples and pork homogenate were filtered through a 0.22-μm filter. We used the filtrate as a solution instead of PBS buffer to simulate the real sample detection. The filtrate was inoculated with S. aureus at different concentrations to prepare the real sample standard. The procedure was identical to the steps described in the part of S. aureus analysis, similarly, with food filtrate used as a negative control.

Results and discussion

Strategy for S. aureus detection

As shown in Fig. 1a, this new visualized biosensor is based on ALP to catalyze the hydrolysis of AP to form ascorbic acid and the reduction product Cu+ to trigger “click chemistry.” The aggregation state of AuNPs changes during signal generation and transduction, and the color of the solution changes from red to blue correspondingly. In the presence of S. aureus, due to an antigen–antibody immune response between IMB and S. aureus, immune complexes were formed. Immune complexes were collected by magnetic separation and delivered to the reporter probe system. The binding of IMB to the S. aureus on surface prevented the substrate from entering the catalytically active site, resulting in a decrease in Cu+ yield; thus, the color of the detection system gradually changes from blue to red (or purple) with the increase of S. aureus concentration, facilitating plasmonic readout with bare eyes (Fig. 1b).

Fig. 1
figure 1

a Cu+-mediated click chemistry response mechanism. b Schematic diagram of the proposed colorimetric assay for detection of S. aureus

Full size image

Characterization of IMB and functionalized AuNPs

We performed a series of characterization of the IMB, including TEM, FTIR, zeta potential, and magnetic hysteresis loop. According to the characterization results (please see Supplementary Information), we successfully synthesized spherical IMB with superparamagnetic properties and fast magnetic response, and the prepared IMB has excellent enrichment effect on S. aureus.

Alkyne- and azide-functionalized AuNPs were successfully prepared by ligand-exchange reaction. TEM, UV–Vis absorption spectrum, and zeta potential were used to measure and characterize alkyne- and azide-functionalized gold nanoparticles (Fig. S2). These characterization results implied the successful preparation of functionalized AuNPs.

Experimental condition optimization

In order to get the best capture performance of S. aureus, various concentrations of IMB were investigated by using S. aureus of 106 cfu·mL−1. The typical photographs of S. aureus colony formation captured by IMB are shown in Fig. S3. Compared with the positive control image in Fig. S3 h, the results showed that with the increase of IMB concentration. The number of colonies in enriched supernatant decreased gradually (Fig. S3 a–g). The average colony numbers of S. aureus are listed in Table S2. When the concentration was 0.5 mg·mL−1, the capture efficiency was 94.96%, indicating that almost all bacteria were captured by IMB. Further increasing the concentration of IMB, it had weak additional beneficial effects. Hence, follow-up experiments were carried out with 0.5 mg·mL−1 IMB.

To verify the ALP activity on the IMB surface, excessive AP was incubated together with IMB. ALP on the surface of IMB hydrolyzes AP to generate AA. The supernatant containing AA reacted with the detection mixture. Cu2+ in the detection mixture was reduced by AA to form Cu+, which triggered the click chemistry reaction of functionalized AuNPs, so that the color of the solution changes from red to blue (Fig. 2a), confirming that the ALP activity still existed. The changes in aggregation states of the detection mixture were also observed by the UV–Vis absorption spectra and the TEM images (Fig. 2b, c, d).

Fig. 2
figure 2

Photographs (a) and UV–Vis spectra (b) of experiments to verify ALP activity on IMB surfaces (1, detection mixture; 2, the supernatant of the reaction between IMB and AP + detection mixture). c TEM image of the detection mixture. d TEM image of the supernatant of the reaction between IMB and AP + detection mixture

Full size image

Sensitivity detection of S. aureus

To test the sensitivity of the detection system, we conducted experiments on various concentrations of S. aureus (10–106 cfu·mL−1) under optimum measurement conditions. The photographs in Fig. 3a show that after reacting for 5 min, the detection solution gradually changed from blue to blue-purple, purplish-red, and red with the concentration of S. aureus increasing. The naked-eye detection limit of S. aureus was determined as 50 cfu·mL−1, which had significant difference from the color of the control sample. The corresponding changes were observed in UV–Vis spectra at the same time (Fig. S4). With the increase of S. aureus, the absorption peak gradually increased at 530 nm and decreased at 760 nm, meaning that the aggregation of functionalized AuNPs gradually decreases. We employ the net extinction ratio (Δ(A530/A760) = (A530/A760)Sample-(A530/A760)Blank) to evaluate the aggregation of the mixture AuNPs solution. Δ(A530/A760) was linearly correlated with the concentration of S. aureus at 10–106 cfu mL−1. The calibration equation was Δ(A530/A760) = 0.137lgC-0.062, R2 = 0.998 (Fig. 3b). The detection limit (LOD) for S. aureus was calculated as low as 2.4 cfu mL−1 (calculated by dividing the standard deviation of the threefold negative control sample by the slope of the fitting standard curve). In order to narrow the influence of large deviations at large concentrations on the regression line, we plotted the linear relationship equation for low concentrations of S. aureus. As shown in Fig. 3c, the linear regression equation is Δ(A530/A760) = 0.132lgC-0.050, R2 = 0.992. Compared with the related reported detection strategies for S. aureus (list in Supplementary Material, Table S3), the colorimetric assay we introduced shows excellent performance in terms of detection time, detection steps, and detection limits. This is due to our preparation of a novel dual-labeled magnetic probe modified with ALP and aptamer of S. aureus. This dual-labeled nanoprobe has multiple functions of rapid identification, enrichment of target bacteria, and signal transduction. This dual-labeled nanoprobe converted the quantitative signal of S. aureus. Through the catalysis of ALP and the specific recognition of aptamer, it is converted into the change of Cu+ concentration in solution. As the carrier of ALP and aptamer, magnetic nanoprobe also takes advantage of its magnetic function to simplify the pretreatment steps. Secondly, we choose alkyne- and azide-functionalized AuNPs as the optical sensors. The click chemistry reaction triggered by Cu+ led to aggregation of functionalized AuNPs. The state changes of functionalized AuNPs were accompanied by obvious color change, and the semi-quantitative analysis of S. aureus can be realized by naked eyes. And our proposed method broadens the application of click chemistry in bacterial detection.

Fig. 3
figure 3

a Photographs of detection system after incubation with different concentrations of S. aureus. (1 → 10: 0, 10, 5 × 10, 102, 5 × 102, 103, 5 × 103, 104, 105, and 106 cfu·mL−1). b The calibration plot for S. aureus (Δ(A530/A760) vs. the logarithm of S. aureus concentration). c The calibration plot for S. aureus at 0, 10, 5 × 10, 102, 5 × 102, and 103 cfu·mL−1 (Δ(A530/A760) vs. the logarithm of S. aureus concentration). Error bars represent the standard deviation of three replicates

Full size image

Specificity detection of S. aureus

To evaluate the selectivity of our assay for the target bacteria, the detection system was applied to other interference bacteria, including E. coli O157:H7; V. parahaemolyticus; L. monocytogenes; and S. typhimurium, and the absorption spectrum was recorded. Selectivity experiments were performed by identifying S. aureus with 102 cfu mL−1 and interfering bacteria with 103 cfu mL−1. The PBS solution was also used as reagent blank. As predicted, all interfering samples and blank samples emerged with a blue color change along with gathered functionalized AuNPs. By contrast, a great color change which transformed from blue to light red was observed, while there is S. aureus only and S. aureus that coexisted with other interfering bacteria (Fig. 4a). As shown in Fig. 4b, just the existence of S. aureus could present a strong absorbance ratio, while the group of interfering substances appeared a bare absorbance ratio. The results showed that 10 folds of other interfering bacteria will not disturb the identification of the target bacteria, demonstrating that the proposed colorimetric assay possessed a good selectivity for S. aureus, which can attribute to the synthesized aptamer with high specificity of target bacteria.

Fig. 4
figure 4

Selectivity test between the target bacteria S. aureus and potential interferences. The concentration of S. aureus was 102 cfu mL−1, and that of other bacteria was 103 cfu mL−1. a Photographs of the proposed colorimetric assay in the presence of different analytes (1 → 8: blank, V. parahaemolyticus, E. coli O157:H7, L. monocytogenes, S. typhimurium, mixture, S. aureus + mixture, and S. aureus). b The ratio of Δ(A530/A760) of the AuNPs in response to different analytes. Error bars represent the standard deviation of three replicates

Full size image

Analysis of real samples

To study the application performance of the proposed method in real samples, we selected the spiked lake water and food as the sample model for determination. The proposed method was employed to evaluate different concentrations of S. aureus in lake water and pork homogenate. Figure 5 shows the photographs and standard calibration curves of two real samples. Figure 5a and b show the photographs of different concentrations of S. aureus in two real sample matrices. The color of the sample tube changes from dark blue, light blue, blue-violet, to red with the increase of S. aureus. For the pork matrix, when the concentration is at 103 cfu mL−1, there was a clear color difference between the sample and the blank, so the naked-eye readout was 103 cfu mL−1, and the naked eye detection limit is 102 cfu mL−1 in lake water matrix. Because of some impurities in the food matrix, the standard calibration plots of the two real matrix samples were slightly different from those obtained in PBS buffer (Fig. 5c, d). The calibration equation were Δ (A530/A760) = 0.1559lgC-0.0491, R2 = 0.974 (pork matrix) and Δ (A530/A760) = 0.1853lgC-0.1567, R2 = 0.990 (lake water matrix). The LOD for S. aureus in real samples was calculated to be 9.7 cfu mL−1 (pork) and 8.5 cfu mL−1 (lake water) (calculated by dividing the standard deviation of the threefold negative control sample by the slope of the fitting standard curve). According to Chinese National Microbiological Criteria, the Australia/New Zealand Standard Methods for Food Microbiology and Microbiological Criteria for Food Stuffs of the European Communities, the coagulase-positive Staphylococci limits for meat and tap water were 100 CFU/g and 100 CFU/mL, respectively [28]. This means that the proposed method could be sufficient to detect S. aureus in real world. As described in Table 1, the spiked samples in low concentration (5.0 × 10 cfu mL−1), medium concentration (5.0 × 102 cfu mL−1), and high concentration (5.0 × 103 cfu mL−1) were measured. The results showed that the average recovery rate was 91.15–106.36%, and the relative standard deviations (RSD) ranged from 7.30 to 13.87%, which proves that the method has good stability and consistency. Meanwhile, it should be noted that the color of all sample’s solution can be distinguished by the naked eye (Fig. S5). Thus, the colorimetric assay we proposed can easily realize the one-step detection and examination of S. aureus in real samples.

Fig. 5
figure 5

Detect S. aureus in real samples. Photographs of detection system after incubation with different concentrations of S. aureus in pork (a) and lake water matrix (b) (1 → 7: 0, 10, 102, 103, 104, 105, and 106 cfu·mL−1). The standard calibration plots of the pork (c) and lake water matrix (d)

Full size image
Table 1 The recoveries and RSD values of detecting bacteria in spiked pork samples (n = 3)
Full size table

The analytical performance and detection effect of the sensor for S. aureus were investigated comprehensively, and the results are basically satisfactory. However, the proposed method still has some limitations. One of the biggest defects was that our materials cannot be reused. Although the cost of material synthesis was very low, there were still cumbersome shortcomings in the synthesis process. Secondly, our proposed method largely depends on the catalysis of ALP. Although ALP has specificity and high catalytic activity, as a natural enzyme, it will also be affected by external factors and reduce its stability. Therefore, if this sensor is to be applied to a wider range of fields, it must be committed to developing reusable probes to further improve the detection activity.

Conclusion

A reliable colorimetric method with high sensitivity and good selectivity for the detection of S. aureus in food was established. Our proposed approach has the following highlights: (1) the aptamer and ALP dual-labeled magnetic nanoparticles could not only capture S. aureus but also hydrolyze AP to induce “click chemistry.” Benefit from its excellent enrichment efficiency and high enzyme catalytic activity, the analysis can be completed within 1 h. (2) The click chemistry reaction induced the aggregation of functionalized AuNPs accompanied by a visible color change from blue to red. The concentration of bacteria can be judged semi-quantitatively by color change. At the same time, the proposed method broadens the application of click chemistry reaction in bacterial detection. (3) The application of multifunctional nanoprobes makes the method less demanding and easy to operate. These advantages make us believe that the colorimetric method developed by us has great application potential in monitoring S. aureus and on-site detection.