Magnetic SERS Strip Based on 4-mercaptophenylboronic Acid-Modified Fe3O4@Au for Active Capture and Simultaneous Detection of Respiratory Bacteria

The rapid diagnosis and detection of respiratory bacteria at the early stage can effectively control the epidemic spread and bacterial infection. Here, we designed a rapid, ultrasensitive, and quantitative lateral flow immunoassay (LFA) strip for simultaneous detection of respiratory bacteria S. aureus and S. pneumoniae. In this assay, the surface enhanced Raman scattering (SERS) tags were designed through combining magnetite Raman enhancement nanoparticle Fe3O4@Au/DTNB and recognition element 4-mercaptophenylboronic acid (4-MPBA). Further, 4-MPBA could capture multiple bacteria in a complex environmental solution. Based on the strategies, Fe3O4@Au/DTNB-mediated magnetic enrichment and 4-MPBA-mediated universal capture capabilities improved the detection sensitivity, the limits of detection for S. aureus and S. pneumoniae were as low as 8 and 13 CFU mL−1, respectively, which were more sensitive than those of colloidal gold method. The Fe3O4@Au/DTNB/Au/4-MPBA-LFA also exhibited good reproducibility, excellent specificity, and high recovery rates in sputum samples, indicating its potential application in the detection of respiratory bacteria samples.


Introduction
Common respiratory bacteria, mainly including Staphylococcus aureus, Escherichia coli, Streptococcus pneumoniae, and Pseudomonas aeruginosa [1], infections caused by these bacteria can induce serious respiratory diseases [2]. Accurate and sensitive respiratory bacteria detection is crucial to controlling bacterial infections and human health. Numerous analytical methods for pathogen detection have been developed, conventional detection method standard plate colony counting generally takes several days and the sensitivity is low [3]. Common molecular biological detection methods, such as polymerase chain reaction (PCR) [4], enzyme-linked immunosorbent assay (ELISA) [5], surface plasmon resonance [6], and nucleic acid hybridization [7], however, all these methods typically have the disadvantages of time-consuming, laborious sample pretreatment, and expensive equipment requirements, which limit their application of bacterial detection [8][9][10]. Therefore, a rapid and ultrasensitive detection technology for respiratory pathogens is highly required.
SERS has been used in various biomarker or pathogens detection field owing to its high sensitivity, specificity, and high speed [11]. In particularly, SERS can provide spectroscopic fingerprint and nondestructive data acquisition even in complex sample environment with fewer background interference [12]. For instance, Zhang has reported a potential biomarker exosomal Programmed cell death 1 ligand 1 (PD-L1) for immunotherapy and provided SERS detection platform for the trace detection of clinical cancers, the detection limit for PD-L1 and exosomes based on PD-L1 were 0.1 ng mL −1 and 4.8 × 10 6 particles mL −1 ,

Instruments
Transmission electron microscope (TEM) images were captured by a Hitachi H-7650 TEM (Tokyo, Japan) at 50 kV. High resolution transmission electron microscope (HRTEM) images were captured using the FEI Tecnai G2 F20 electron microscopy (Hillsboro, OR, USA) at 200 kV. Scanning electron microscope (SEM) images were obtained by a JEOL JSM-7001F instrument (Tokyo, Japan) at an operating voltage of 10 kV. Zeta potential data were tested by a Malvern Nano-ZS90 ZetaSizer (Malvern, UK) instrument. Raman spectra data were collected by a portable Raman system (B&W Tek, i-Raman Plus BWS465-785H spectrometer, Plainsboro, NJ, USA).

Fabrication of Fe 3 O 4 @Au/DTNB/Au/4-MPBA Tags
The design route of Fe 3 O 4 @Au/DTNB/Au/4-MPBA tags was demonstrated in Scheme 1. Fe 3 O 4 @Au/DTNB was fabricated by previously reported PEI−mediated seed growth method [28]. The specific synthesis details were provided in the Supporting Information S1. Fe 3 O 4 @Au/DTNB was dispersed in the PEI aqueous solution (2 mg mL −1 ) followed by sonication for 0.5 h and washed for three times, then mixed with colloidal 3-nm AuNPs and sonicated for another 0.5 h to obtain Fe 3 O 4 @Au/DTNB/Au. Finally, 200 µL of 4-MPBA (10 mM) was added to the prepared Fe 3 O 4 @Au/DTNB/Au and continuously sonicated for 2 h to fabricate Fe 3 O 4 @Au/DTNB/Au/4-MPBA tags. The Fe 3 O 4 @Au/DTNB/Au/4-MPBA tags were stored at room temperature for further use.

Instruments
Transmission electron microscope (TEM) images were captured by a Hitachi H-7650 TEM (Tokyo, Japan)at 50 kV. High resolution transmission electron microscope (HRTEM) images were captured using the FEI Tecnai G2 F20 electron microscopy (Oregon, USA) at 200 kV. Scanning electron microscope (SEM) images were obtained by a JEOL JSM-7001F instrument (Tokyo, Japan) at an operating voltage of 10 kV. Zeta potential data were tested by a Malvern Nano-ZS90 ZetaSizer (Malvern, UK) instrument. Raman spectra data were collected by a portable Raman system (B&W Tek, i-Raman Plus BWS465-785H spectrometer, Plainsboro, NJ, USA).

Fabrication of Fe3O4@Au/DTNB/Au/4−MPBA Tags
The design route of Fe3O4@Au/DTNB/Au/4−MPBA tags was demonstrated in Scheme 1. Fe3O4@Au/DTNB was fabricated by previously reported PEI−mediated seed growth method [28]. The specific synthesis details were provided in the Supporting Information S1. Fe3O4@Au/DTNB were dispersed in the PEI aqueous solution (2 mg mL −1 ) followed by sonication for 0.5 h and washed for three times, then mixed with colloidal 3-nm AuNPs and sonicated for another 0.5 h to obtain Fe3O4@Au/DTNB/Au. Finally, 200 μL of 4−MPBA (10 mM) was added to the prepared Fe3O4@Au/DTNB/Au and continuously sonicated for 2 h to fabricate Fe3O4@Au/DTNB/Au/4−MPBA tags. The Fe3O4@Au/DTNB/Au/4−MPBA tags were stored at room temperature for further use.

Fabrication of Lateral Flow Strip Bacteria Detection System
A LFA strip was prepared with absorbent pad, sample pad, and nitrocellulose membrane (NC membrane). S. aureus (0.6 mg mL −1 ) and S. pneumoniae (0.4 mg mL −1 ) antibodies were sprayed onto NC membrane as test line 1 (T1) and test line 2 (T2), respectively. NC membrane was placed at 37 • C for 2 h. Subsequently, absorption pad and sample pad were assembled onto the both ends of NC membrane to ensure liquid flow. The prepared card was then cut into the strips with a width of 3 mm and stored in a vacuum desiccator for further use.

Preparation of Bacterial Sample
Bacteria concentrations, including S. aureus and S. pneumoniae, were verified by plate counting method [29]. The bacteria were cultured using 5% sheep blood agar plates at 37 • C overnight. The initial bacterial solution was obtained by several bacterial colonies on the plates picked into 0.5 mL of sterile PBS solution. The initial solution was then diluted 10 4 to 10 6 times using PBS solution, 0.1 mL of diluted bacterial suspensions were withdrawn and dropped on blood agar plates followed by incubation at 37 • C overnight. The colony forming units (CFUs) on the plates were counted to calculate the concentration of initial bacterial solution, based on the calculated results, the testing concentration of 10 8 cells mL −1 was prepared for further use. The detection of S. aureus and S. pneumoniae was conducted in a 2 mL tube. First, 4 µL of liquid Fe 3 O 4 @Au/DTNB/Au/4-MPBA tags was added into 1 mL of each bacteria suspension and magnetically collected after incubation for 20 min. The bacteria suspension in the tube was removed and 80 µL of PBS running buffer was added followed by ultrasonic for 1 min, the prepared strip was inserted into the tube for 20 min, then the strip was dried and SERS intensities of test lines were tested with a portable Raman spectrometer under 785 nm laser excitation.

Detection of Respiratory Pathogens S. aureus and S. pneumoniae in Biological Samples
The recovery rates (%) of bacteria were calculated using sputum specimens. The bacterial concentrations of 10 6 CFU mL −1 and 10 3 CFU mL −1 were respectively spiked to the diluted sputum specimens, the specific operation was conducted according to the protocol in Section 2.6. The recovery rates were determined according to the previously established calibration curves and the tests were conducted in triplicate for each sample. In this detection system, Fe 3 O 4 @Au/DTNB/Au/4-MPBA was used as SERS tags to simultaneously capture two target bacteria and improve the sensitivity of LFA. Fe 3 O 4 @Au/ DTNB/Au/4-MPBA tags presented good dispersity, magnetic responsiveness, universal capture, and strong SERS activity in a complex solution. The operating procedure of the simultaneous detection of respiratory bacteria was displayed in Scheme 1. The detection principle was based on antibody-antigen interaction. First, Fe 3 O 4 @Au/DTNB/Au/4-MPBA tags were added into the solution containing S. aureus and S. pneumoniae followed by shaking for about 20 min. The target bacteria were captured by Fe 3 O 4 @Au/DTNB/Au/4-MPBA tags during incubation, the Fe 3 O 4 @Au/DTNB/Au/4-MPBA-bacteria complexes were separated from the bacteria solution using a magnet, then resuspended in the prepared running buffer solution and ultrasonic for several seconds. Finally, LFA strips were inserted into the running buffer solution, the complexes Fe 3 O 4 @Au/DTNB/Au/4-MPBA-bacteria slowly flowed along the strip from the sample loading pad to the absorbent pad under capillary forces. Due to the interaction between antibody and antigen, the complexes were captured by the S. aureus and S. pneumoniae antibodies pre-coated on the T1 and T2 lines, respectively, and then the sandwich−like immunocomplexes Fe 3 O 4 @Au/DTNB/Au/4-MPBA-bacteria-detection antibodies were formed at the test lines. There was no control line (C line) on NC membrane owing to lack 4-MPBA antibody. Although C line was used to evaluate whether the tags and NC membrane were effective, our study revealed that the target bacteria were captured by fabricated Fe 3 O 4 @Au/DTNB/Au/4-MPBA tags, then the Fe 3 O 4 @Au/DTNB/Au/4-MPBA-bacteria complexes were captured by the antibodies on NC membranes, our results demonstrated the tags possessed excellent capture performance and the immune binding with antibodies on the test lines was normal, which further indicated that the prepared NC membranes were available and our detection system was successfully fabricated, relevant studies of lack of C line on NC membrane have also been reported [30,31]. Moreover, the Raman signal intensities became higher with the increase of tags and bacteria, which provided the quantification detection method, after reaction for 20 min, two dark bands appeared on the strip in the presence of S. aureus and S. pneumoniae, only one dark band (T1 line or T2 line) was observed in the presence of S. aureus or S. pneumoniae, while no line was observed in absent of bacteria, which indicated that the detection system was successfully fabricated. Finally, Raman intensities were measured by a portable Raman spectrometer.

Characterization of Fe 3 O 4 @Au/DTNB/Au/4-MPBA Tags
As shown in Figure 1a The SERS activities of Fe 3 O 4 , Fe 3 O 4 @Au-seed/DTNB, and Fe 3 O 4 @Au/DTNB were tested, the feature peaks of DTNB at 736, 826, 1061, 1149, 1331, and 1557 cm −1 were detected, among them, 1331 cm −1 was the strongest peak ( Figure 1k) and the corresponding enhancement factor (EF) was 1.3 × 10 6 (S4 and Figure S2), therefore, 1331 cm −1 was recognized as the characteristic peak for further analysis. The ability of Fe 3 O 4 @Au (without modification of DTNB) to enhance Raman signal and act as SERS were also validated using conventional dyes CV (S3 and Figure S1). As shown in Figure 1l, the zeta potential of Fe 3 O 4 was −13.0 mV, the positively charged PEI and negatively charged 3-nm AuNPs were decorated onto the surface of Fe 3 O 4 under electrostatic interaction, zeta potentials of Fe 3 O 4 −PEI and Fe 3 O 4 @Au-seed were +27.2 mV and +6.7 mV, respectively, while the potential of Fe 3 O 4 @Au/DTNB was decreased to −21.2 mV owing to the reduction of HAuCl 4 . Similarly, the potential of Fe 3 O 4 @Au/DTNB/PEI was +39.4 mV and Fe 3 O 4 @Au/DTNB/Au was +6.7 mV after the second layer of PEI and 3-nm AuNPs were coated on Fe 3 O 4 @Au/DTNB, which confirmed that Fe 3 O 4 @Au/DTNB/Au was successfully fabricated [33,34].

Optimization of the Fe 3 O 4 @Au/DTNB/Au/4-MPBA-Based LFA Strip
The Fe 3 O 4 @Au/DTNB/Au/4-MPBA tags can capture bacteria because 4-MPBA binds lipopolysaccharide that locates on the bacterial cell wall [35]. The TEM images clearly revealed that Fe 3 O 4 @Au/DTNB/Au/4-MPBA could effectively capture S. aureus and S. pneumoniae (Figure 2a  Based on SERS strip, the specificity of antibodies was demonstrated by the solutions containing 10 5 CFU mL −1 of S. aureus, S. pneumoniae, and their mixture. As displayed in Figure 2f, two test lines of S. aureus and S. pneumoniae on NC membrane could be observed, indicating that there were two target bacteria in the detection sample, whereas only one test line was observed for S. aureus or S. pneumoniae that existed alone. Additionally, the SERS intensity values of S. aureus or S. pneumoniae alone were higher than that of S. aureus and S. pneumoniae simultaneous existence owing to broad-spectrum capture performance of 4-MPBA (Figure 2g Figure S3). Considering the detection performance of the system, several critical parameters were optimized and provided in the supporting information. The NC membranes of Prima40 (18 µm pore size), CN140 (8 µm pore size), and CN95 (15 µm pore size) were evaluated using 10 6 CFU mL −1 of S. aureus and S. pneumoniae. The signal-to-noise ratio (SNR) was calculated by the ratio of the SERS signal intensities of positive and negative samples, CN95 membrane displayed the highest SNR for target bacteria ( Figure S4). Owing to the immunobinding between Fe 3 O 4 @Au/DTNB/Au/4-MPBA-bacteria complexes and antibodies, the running buffer was further optimized, the results confirmed that PBS solution containing 2% (w/v) Tween and 15% (w/v) FBS could suppress the non-specific combination of Fe 3 O 4 @Au/DTNB/Au/4-MPBA on the test lines and obtained maximized SNR for target bacteria ( Figure S5). A serial usage of Fe 3 O 4 @Au/DTNB/Au/4-MPBA tags 1 µL, 2 µL, 4 µL, 6 µL, and 8 µL were respectively incubated with S. aureus and S. pneumoniae Based on SERS strip, the specificity of antibodies was demonstrated by the solutions containing 10 5 CFU mL −1 of S. aureus, S. pneumoniae, and their mixture. As displayed in Figure 2f, two test lines of S. aureus and S. pneumoniae on NC membrane could be observed, indicating that there were two target bacteria in the detection sample, whereas only one test line was observed for S. aureus or S. pneumoniae that existed alone. Additionally, the SERS intensity values of S. aureus or S. pneumoniae alone were higher than that of S. aureus and S. pneumoniae simultaneous existence owing to broad-spectrum capture performance of 4-MPBA (Figure 2g Figure S3).
Considering the detection performance of the system, several critical parameters were optimized and provided in the supporting information. The NC membranes of Prima40 (18 µm pore size), CN140 (8 µm pore size), and CN95 (15 µm pore size) were evaluated using 10 6 CFU mL −1 of S. aureus and S. pneumoniae. The signal-to-noise ratio (SNR) was calculated by the ratio of the SERS signal intensities of positive and negative samples, CN95 membrane displayed the highest SNR for target bacteria ( Figure S4). Owing to the immunobinding between Fe 3 O 4 @Au/DTNB/Au/4-MPBA-bacteria complexes and antibodies, the running buffer was further optimized, the results confirmed that PBS solution containing 2% (w/v) Tween and 15% (w/v) FBS could suppress the non-specific combination of Fe 3 O 4 @Au/DTNB/Au/4-MPBA on the test lines and obtained maximized SNR for target bacteria ( Figure S5). A serial usage of Fe 3 O 4 @Au/DTNB/Au/4-MPBA tags 1 µL, 2 µL, 4 µL, 6 µL, and 8 µL were respectively incubated with S. aureus and S. pneumoniae at room temperature, the results showed that 4 µL of tags provided the highest SNR for target bacteria detection ( Figure S6). Antibody concentrations of target bacteria were also optimized, 0.4 mg mL −1 of S. pneumoniae antibody and 0.6 mg mL −1 of S. aureus antibody could provide the highest SNR of SERS intensities ( Figure S7). Finally, the reaction time was optimized, SNR increased from 0 to 20 min and then slightly lower after 20 min ( Figure S8). Therefore, the reaction time of 20 min was enough for the whole process. Considering the detection performance of the system, several critical parameters were optimized and provided in the supporting information. The NC membrane of Prima40 (18 μm pore size), CN140 (8 μm pore size), and CN95 (15 μm pore size) were evaluated using 10 6 CFU mL −1 of S. aureus and S. pneumoniae. The signal−to−noise ratio (SNR) was calculated by the ratio of the SERS signal intensities of positive and negative samples, CN95 membrane displayed the highest SNR for target bacteria ( Figure S4). Owing to the immunobinding between Fe3O4@Au/DTNB/Au/4−MPBA−bacteria complexes and antibodies, the running buffer was further optimized, the results confirmed that PBS solution containing 2% (w/v) Tween and 15% (w/v) FBS could suppress the non-specific combination of Fe3O4@Au/DTNB/Au/4−MPBA on the test lines and obtained maximized SNR for target bacteria ( Figure S5). A serial usage of Fe3O4@Au/DTNB/Au/4−MPBA tags 1 μL, 2 μL, 4 μL, 6 μL, and 8 μL were respectively incubated with S. aureus and S. pneumoniae at room temperature, the results showed that 4 μL of tags provided the highest SNR for target bacteria detection ( Figure S6). Antibodies concentrations of target bacteria were also optimized, 0.4 mg mL −1 of S. pneumoniae antibody and 0.6 mg mL −1 of S. aureus antibody could provide the highest SNR of SERS intensities ( Figure S7). Finally, the reaction time was optimized, SNR increased from 0 to 20 min and then slightly lower after 20 min ( Figure S8). Therefore, the reaction time of 20 min was enough for the whole process.

Analytical Performance of Fe 3 O 4 @Au/DTNB/Au/4-MPBA-Based LFA
To access the sensitivity and selectivity of our system, different concentrations of S. aureus and S. pneumoniae from 0 to 10 7 CFU mL −1 were mixed with Fe 3 O 4 @Au/DTNB/Au/ 4-MPBA tags and then detected by Fe 3 O 4 @Au/DTNB/Au/4-MPBA-LFA strips. As shown in Figure 3a, with the decreasing concentration of target bacteria, the color of the test lines gradually shallowed. It could be observed with the naked eyes that the visualization concentrations of target bacteria were 10 and 10 3 cells mL −1 , respectively. Moreover, the SERS signal intensities of test lines at 1331 cm −1 increased with increasing bacterial concentration, the calibration curves of two test lines displayed dynamic relationships between SERS intensities and the concentrations of target bacteria and the correlation coefficients (R 2 ) were 0.994 and 0.995 for S. aureus and S. pneumoniae, respectively (Figure 3c,d). Based on the calculation of blank SERS signal, LODs of Fe 3 O 4 @Au/DTNB/Au/4-MPBA were determined to be 8 and 13 cells mL −1 for S. aureus and S. pneumoniae, respectively, according to the International Union of Pure and Applied Chemistry (IUPAC) protocol (LOD = mean SERS signal intensities of blank groups + 3 × standard deviation of the control) [36]. We compared our method with standard colloidal gold (AuNP)-LFA method, the preparation of colloidal AuNPs was displayed in supporting information S2, the visualization concentrations of S. aureus and S. pneumoniae were 10 4 and 10 5 cells mL −1 for AuNP-LFA strip detection, respectively (Figure 3b). Therefore, our detection system Fe 3 O 4 @Au/DTNB/Au/4-MPBA-LFA presented excellent sensitivity. The reproducibility of Fe3O4@Au/DTNB/Au/4−MPBA−based LFA system is a cru factor for its practical application. The bacteria S. pneumoniae and S. aureus were dilu into a high-dose group of 10 6 CFU mL −1 , the results indicated good reproducibility w control) [36]. We compared our method with standard colloidal gold (AuNP)-LFA method the preparation of colloidal AuNPs was displayed in supporting information S2, the v sualization concentrations of S. aureus and S. pneumoniae were 10 4 and 10 5 cells mL − for AuNP-LFA strip detection, respectively (Figure 3b). Therefore, our detection system  The reproducibility of Fe 3 O 4 @Au/DTNB/Au/4-MPBA-based LFA system is a crucia factor for its practical application. The bacteria S. pneumoniae and S. aureus were dilute into a high-dose group of 10 6 CFU mL −1 , the results indicated good reproducibility wit relative standard deviation (RSD) of 6.32% and 5.12% for S. pneumoniae and S. aureus, respe tively (Figure 4a,b). The diluted medium-dose group of 10 4 CFU mL −1 was also conducte with RSD of 9.43% and 8.63% for S. pneumoniae and S. aureus, respectively (Figure 4c,d The specificity of Fe 3 O 4 @Au/DTNB/Au/4-MPBA-based LFA was further evaluated, sev eral respiratory viruses including FluB (10 5 copies mL −1 ), FluA H1N1 (10 5 copies mL −1 RSV (10 5 pfu mL −1 ), SARS-CoV-2 (100 ng mL −1 ), and respiratory bacteria S. aeruginos (10 6 CFU mL −1 ) were used as negative samples [37,38], as displayed in Figure 5a,b, th positive groups containing target bacteria generated higher SERS intensities on their corr sponding test lines, whereas all the negative samples displayed low SERS signal intensitie confirming Fe 3 O 4 @Au/DTNB/Au/4-MPBA-based LFA possessed good specificity. Finall the recovery rates of Fe 3 O 4 @Au/DTNB/Au/4-MPBA-LFA strips were tested in the sputum specimens, as shown in Table 1, the average recoveries for S. aureus and S. pneumonia ranged from 92.3% to 105.2%, and the RSD values of the SERS intensities on two test line The reproducibility of Fe 3 O 4 @Au/DTNB/Au/4-MPBA-based LFA system is a crucial factor for its practical application. The bacteria S. pneumoniae and S. aureus were diluted into a high-dose group of 10 6 CFU mL −1 , the results indicated good reproducibility with relative standard deviation (RSD) of 6.32% and 5.12% for S. pneumoniae and S. aureus, respectively (Figure 4a,b). The diluted medium-dose group of 10 4 CFU mL −1 was also conducted with RSD of 9.43% and 8.63% for S. pneumoniae and S. aureus, respectively (Figure 4c,d). The specificity of Fe 3 O 4 @Au/DTNB/Au/4-MPBA-based LFA was further evaluated, several respiratory viruses including FluB (10 5 copies mL −1 ), FluA H1N1 (10 5 copies mL −1 ), RSV (10 5 pfu mL −1 ), SARS-CoV-2 (100 ng mL −1 ), and respiratory bacteria P. aeruginosa (10 6 CFU mL −1 ) were used as negative samples [37,38], as displayed in Figure 5a,b, the positive groups containing target bacteria generated higher SERS intensities on their corresponding test lines, whereas all the negative samples displayed low SERS signal intensities, confirming Fe 3 O 4 @Au/DTNB/Au/4-MPBA-based LFA possessed good specificity. Finally, the recovery rates of Fe 3 O 4 @Au/DTNB/Au/4-MPBA-LFA strips were tested in the sputum specimens, as shown in Table 1, the average recoveries for S. aureus and S. pneumoniae ranged from 92.3% to 105.2%, and the RSD values of the SERS intensities on two test lines were less than 10%, therefore, Fe 3 O 4 @Au/DTNB/Au/4-MPBA-LFA strip had potential to detect S. aureus and S. pneumoniae in the real sputum specimens.
ing test lines, whereas all the negative samples displayed low SERS signal intensi firming Fe3O4@Au/DTNB/Au/4−MPBA−based LFA possessed good specificity the recovery rates of Fe3O4@Au/DTNB/Au/4−MPBA−LFA strips were tested in the specimens, as shown in Table 1, the average recoveries for S. aureus and S. pn ranged from 92.3% to 105.2%, and the RSD values of the SERS intensities on two were less than 10%, therefore, Fe3O4@Au/DTNB/Au/4−MPBA−LFA strip had po detect S. aureus and S. pneumoniae in the real sputum specimens.

Conclusions
In this work, we fabricated a novel strip based on Fe3O4@Au/DTNB/Au/4 SERS nanotags for simultaneous detection of two respiratory bacteria. The SERS t sessed magnetic enrichment and broad−spectrum bacteria capture capabilities. sults showed that Fe3O4@Au/DTNB/Au/4−MPBA−LFA biosensor was ultrasensi the LODs of S. aureus and S. pneumoniae were as low as 8 and 13 CFU mL −1 , resp Moreover, the biosensor displayed good reproducibility, excellent specificity, a

Conclusions
In this work, we fabricated a novel strip based on Fe 3 O 4 @Au/DTNB/Au/4-MPBA SERS nanotags for simultaneous detection of two respiratory bacteria. The SERS tags possessed magnetic enrichment and broad-spectrum bacteria capture capabilities. Our results showed that Fe 3 O 4 @Au/DTNB/Au/4-MPBA-LFA biosensor was ultrasensitive and the LODs of S. aureus and S. pneumoniae were as low as 8 and 13 CFU mL −1 , respectively. Moreover, the biosensor displayed good reproducibility, excellent specificity, and high recovery rates in the sputum specimens. If the antibodies on NC membrane are altered, the strip can be used to detect other specific bacteria. Thus, the proposed multifunctional tool provided a universal platform for the detection of respiratory pathogens.