Introduction

Staphylococcus aureus (S. aureus) is one of the most common pathogens associated with dangerous diseases including septicemia, osteomyelitis, pneumonias and endocarditis [1, 2]. Thus, sensitive, rapid, cost-efficient and specific detection of this pathogen is of great importance.

Various methods have already been established to detect S. aureus. The conventional method based on bacteria culture and colony counting possesses some merits, such as high sensitivity and good reliability. Nevertheless it is still time-consuming and laborious [3, 4]. Nucleic acid-based approaches such as polymerase chain reaction are culture-free and highly efficient. However, complicated sample pretreatments such as cell disruption and nucleic acid extraction limit their application in field detection [5, 6]. Some molecular recognition modes for whole cells detection have been established to speed up and simplify the detection process. The utilized molecular recognition agents include antibodies [7], bacteriophages [8], aptamers [9], antibiotics [10], lectins [11], and glycan [12]. Specific agents such as antibodies and aptamers always suffer from poor stability, high cost and varying activity depending on batch and source [13]. Other agents such as glycan lack specificity to identify certain species of bacteria.

Peptides are much cheaper to be produced and modified compared with other molecular recognition agents. Antimicrobial peptides (AMPs) such as magainin I are broad-spectrum antimicrobial agents against various pathogens [14, 15], however they usually suffer from poor specificity. Phage-displayed peptides (PDPs) show ideal specificity to the target bacteria as bacteriophages possess highly specific mechanism for recognizing pathogens [16, 17]. For example, via biopanning, Rao et al. [18] have obtained a PDP that can specifically bind with the cell membrane protein of S. aureus. Both AMPs and PDPs can be modified with maintenance of their binding activity to bacteria, when the modification is occurred at the C-terminal [15, 19].

Sandwich fluorimetric detection using molecular recognition agents exerts high sensitivity and ideal specificity, thus has been adopted to detect S. aureus [20,21,22] and other pathogens [15, 20, 23]. In these previous works, antibodies, thioglycolic acid, aptamers and oligonucleotides were utilized to recognize the pathogens. In this study, two peptides were utilized to establish a fluorimetric method for specific detection of S. aureus on magnetic particles (MPs) platform. Briefly, a PDP exerting high affinity and ideal specificity to S. aureus was utilized to functionalized MPs. Then the functionalized MPs were applied to capture S. aureus from the sample matrix. Fluorescein isothiocyanate (FITC)-tagged magainin I was used as a second recognition peptide and fluorescent tracer to form a sandwich biological complex of S. aureus for fluorimetry.

Experimental

Materials and instrumentations

Strains of S. aureus, Staphylococcus epidermidis (S. epidermidis), Escherichia coli (E. coli), Pseudomonas aeruginosa (P. aeruginosa), Enterococcus faecium (E. faecium), Streptococcus and Micrococcus luteus (M. luteus) were obtained from Chongqing Center for Disease Control and Prevention (Chongqing, China, www.cqcdc.org). The amino acid sequences of PDP and Magainin I were LHVNPGLQTLQG and GIGKFLHSAGKFGKAFVGEIMKS, respectively. FITC-tagged magainin I, biotin-tagged PDP and FITC-tagged PDP were synthesized by GL Biochem Co., Ltd. (Shanghai, China, www.glbiochem.com) according to the amino acid sequences provided by us. Streptavidin-modified MPs with a diameter of 1 μm were provided by Seebio Co., Ltd. (Shanghai, China, www.seebio.com). Gelatin was purchased from Sigma-Aldrich (St. Louis, USA, www.sigma-aldrich.com). Bovine serum albumin (BSA) and SuperBlock® T20 was provided by Thermo Fisher Scientific Inc. (Waltham, USA, www.thermofisher.com). All other reagents were of analytical grade and used as received.

All aqueous solutions were prepared in ultrapure water (18.2MΩ) produced by an ELGA PURELAB Classic system (France, www.elgalabwater.com). Confocal microscopy micrographs were obtained from a LSM710 confocal laser scanning microscope (Carl Zeiss AG, Germany, www.zeiss.com). Fluorescence signals were detected using an Infinite 200 PRO multifunctional microplate reader (Tecan, Switzerland, www.tecan.com). Scanning electron micrograph was obtained using a S3000 N scanning electron microscope (Hitachi, Japan, www.hitachi.com).

Fluorescent characterization of affinity between S. aureus and peptides

FITC was introduced to the peptides (PDP and magainin I) by linking a C6 linker (triglycine) and a FITC-tagged lysine at the C-terminal of the amino acid sequence according to the previously reported protocol [24, 25]. The aim of using C6 linker is to minimize the loss of peptides binding activity that was resulted from chemical modification. Then the binding activity of FITC-tagged peptides towards some common bacteria including S. aureus, S. epidermidis, E. coli, P. aeruginosa, E. faecium, Streptococcus and M. luteus were investigated using fluorescence imaging. In detail, 1.0 mL of bacterial culture at 1.2 × 109 cfu·mL−1 was centrifuged at 6544 rcf for 3 min to discard the supernatant. The obtained bacteria pellet was incubated with 1.0 mL of 10 μg mL−1 FITC-tagged peptides at room temperature (RT) for 45 min. After that, the stained bacteria were centrifugally washed twice and re-suspended in 200 μL of phosphate buffered saline (PBS) at pH 8.0. Finally, 10 μL of the resultant stained bacterial solution was mounted onto a microscopic slide, and observed under a confocal laser scanning microscope with a magnification of 1000. The excitation and emission wavelengths for FITC were set to be 488 nm and 525 nm, respectively.

Detection procedure for S. aureus

Biotin-tagged PDP was prepared by using the same protocol for the preparation of FITC-tagged peptides. Then 200 μg of biotin-tagged PDP was incubated with 1.0 mL of streptavidin-modified MPs suspension at 2.0 mg mL−1 and pH 7.4 at RT for 45 min. The obtained PDP-functionalized MPs were washed twice with PBS at pH 7.4, then blocked by 1% gelatin for 60 min at RT. For sandwich fluorimetric detection of S. aureus, 5.0 μL of PDP-functionalized MPs was washed twice and then incubated with 1.0 mL of S. aureus sample at RT for 45 min. After magnetic separation and twice washing, the resulted S. aureus-loading MPs were incubated with 1.0 mL of FITC-tagged magainin I at 5.0 μg mL−1 at RT for 45 min. The formed sandwich complex of PDP-functionalized MPs/S. aureus/FITC-tagged magainin I was wash twice and re-suspended in 200 μL of PBS at pH 8.0. Lastly the suspension was transferred into a microplate to conduct fluorescence detection with excitation wavelength of 488 nm and emission wavelength of 525 nm.

Results and discussion

The principle of fluorimetric detection of S. aureus

PDP was used as a highly specific recognition agent for S. aureus to functionalize MPs. The specific binding between PDP and S. aureus was tested by laser confocal fluorescence microscopy. As shown in Fig. 1a and b, strong fluorescent emission was observed on the cells of S. aureus. However, for S. epidermidis, E. faecium, E. coli, P. aeruginosa, Streptococcus and M. luteus, fluorescence staining was not observed (Fig. S1 Electronic Supporting Material). The above results demonstrated the ideal binding specificity between S. aureus and the artificially synthesized PDP. The binding behavior between magainin I and bacteria was also observed by the same procedure. Typical fluorescence micrographs of S. aureus stained with FITC-tagged magainin I are showed in Fig. 1c and d. The similar results were also obtained for other investigated bacteria (Fig. S2 in Electronic Supporting Material). They demonstrated that magainin I was a broad-spectrum AMP capable of binding with all the investigated bacteria.

Fig. 1
figure 1

a Fluorescence micrograph of S. aureus stained with FITC-tagged PDP; b bright field micrograph of S. aureus stained with FITC-tagged PDP; c fluorescence micrograph of S. aureus stained with FITC-tagged magainin I; d bright field micrograph of S. aureus stained with FITC-tagged magainin I. The excitation and emission wavelengths for FITC were set to be 488 nm and 525 nm, respectively

Full size image

Based on the above results, a sandwich method for fluorimetric detection of S. aureus was established and schematically illustrated in Fig. 2. MPs were functionalized with PDP through biotin-streptavidin reaction, and then the PDP-functionalized MPs were used to capture and enrich S. aureus from sample matrix. FITC-tagged magainin I acted as the signal tracer to form fluorescence-stained sandwich complex of S. aureus, and the fluorescent signal was used to quantitate S. aureus.

Fig. 2
figure 2

Schematic illustration of sandwich fluorimetric detection of S. aureus based on dual-peptide recognition strategy

Full size image

Capture of S. aureus by PDP-functionalized MPs

The loading amount of PDP was measured by UV spectrum to be 75 μg of biotin-tagged PDP per gram of streptavidin-modified MPs. The capture of S. aureus by PDP-functionalized MPs was observed by the scanning electron microscope, and the micrograph is shown in Fig. 3. As seen in this micrograph, PDP-functionalized MPs and S. aureus were attached with each other, demonstrating the binding behavior between them. To investigate the capture efficiency (CE) of S. aureus by PDP-functionalized MPs, 5.0 μL of PDP-functionalized MPs at 2.0 mg mL−1 was incubated with 1.0 mL of S. aureus at 1.0 × 103 cfu·mL−1 (C 0 ) under constant stirring at RT for 45 min. After magnetic separation, the bacteria concentration in the supernatant (C 1 ) was measured using agar plate counting. The CE value was calculated by the following equation:

$$ \mathrm{CE}=\left(1- C1/ C0\right)\times 100\% $$
(1)
Fig. 3
figure 3

Scanning electron micrograph of S. aureus binding with PDP-functionalized MPs. The concentrations of S. aureus was 1.0 × 103 cfu·mL−1

Full size image

The CE value was measured to be 86.3%, showing very strong capture ability of the PDP-functionalized MPs to S. aureus. In the further investigation, MPs without PDP was also used to detect S. aureus with the same protocol. The obtained fluorescence signal was found to be 6.2% of that using PDP-functionalized MPs, which demonstrated that the capture of S. aureus was produced by PDP.

Optimization of detection condition

The following parameters were optimized: (a) the blocking agent; (b) the incubation time for S. aureus and FITC-tagged magainin I; (c) the concentration of FITC-tagged magainin I; (d) the amount of PDP-functionalized MPs. Respective data and figures (Fig. S3–6) are given in Electronic Supporting Material. The following experimental conditions were found to give the best results: (a) 1% gelatin as the blocking agent; (b) 45 min as the incubation time; (c) 5.0 μg mL−1 as the concentration of FITC-tagged magainin I; (d) 5.0 μL as the amount of PDP-functionalized MPs.

Analytical performance

Under the chosen experimental conditions, the fluorescence signal was proportional to the logarithm value of S. aureus concentration within a linear range of 1.0 × 101–1.0 × 105 cfu·mL−1 (Fig. S7). The detection limit for S. aureus was 9.0 cfu·mL−1 at a signal to noise ratio of 3. For a five-point calibration curve, the regression equation was I (a . u.) = 13414 + 8156 lg C (cfu · mL − 1), with a correlation coefficient (R 2) of 0.999. Here I and C are the signal intensity and the bacterial concentration, respectively. The relative standard deviation (RSD) values of the five detection concentrations ranged from 7.4% to 2.1% (n = 3), demonstrating its acceptable repeatability. As seen in Table S1, this work showed comparable sensitivity but much lower cost compared with the previously reported methods for S. aureus detection [9, 22, 26,27,28,29].

Specificity for S. aureus detection

The specificity of the established method was evaluated by investigating the potential interference from some common gram-positive bacteria (E. faecium and Streptococcus) and gram-negative bacteria (E. coli, P. aeruginosa and M. luteus). In the interference investigation, the concentration of all bacteria was 1.0 × 103 cfu·mL−1. The fluorescence signals of all the investigated bacteria are shown as Fig. 4. The interference factor (IF) values of these interfering bacteria were calculated according to the following formula:

$$ \mathrm{IF}=\left( IS- BS\right)/\left( SS- BS\right)\times 100\% $$
(2)
Fig. 4
figure 4

Fluorescence signals of S. aureus and the interfering bacteria. The concentrations of all bacteria were 1.0 × 103 cfu·mL−1. The excitation and emission wavelengths for FITC were set to be 488 nm and 525 nm, respectively. All other detection conditions were the chosen optimal conditions (n = 3)

Full size image

Here, IS, SS, and BS mean the signals of interfering bacteria, S. aureus and blank sample (plain phosphate buffered saline), respectively. The IF values of M. luteus, E. coli, P. aeruginosa, Streptococcus and E. faecium, were 5.1%, 6.7%, 7.3%, 8.3% and 7.6%, respectively, all showing negligible interference.

Application in real sample assay

To evaluate the application potential of this method, lake water, human urine and apple juice were sterilized and spiked with S. aureus standard samples at known concentrations. All these samples were adjusted to pH 7.4 and filtered through 0.22-μm filter membrane before they were spiked with S. aureus. These spiked real samples were adopted for recovery tests. The results presented in Table 1 shows acceptable recovery values ranging from 81% to 110%, with RSD values all below 5.1%. They demonstrated the reliability of this method for application in environmental, biological and food sample matrixes.

Table 1 Recovery tests of S. aureus spiked in real samples (n = 3)
Full size table

Conclusion

In summary, a novel sandwich fluorimetric method for detecting whole cells of S. aureus has been established by utilizing two artificially synthesized peptides as the molecular recognition agents. This dual-peptide recognition-based method did not demand complex pretreatments such as bacteria culture or nucleic acid extracting, thus achieved a rapid detection of S. aureus. PDP with inherent specificity originated from bacteriophage ensured the exclusion of interference from other bacteria. Compared with conventional molecular recognition agents such as antibodies and bacteriophages, artificially synthesized peptides were much more stable and inexpensive. Utilization of MPs platform greatly improved its sensitivity and achieved a very low detection limit. It showed greatly promising application potential in medical diagnosis, food safety and bioterrorism defense.