1. Introduction

Lung cancer is one of the malignant tumors with the highest morbidity and mortality in the world, which seriously endangers human health [1]. Non-small cell lung cancer (NSCLC) accounts for about 85% of all lung cancers and is the main type of lung cancer [2]. The early symptoms of NSCLC are hid­den and difficult to detect, most of the clinical symptoms such as fever, cough, hemoptysis and chest pain are presented in the late stage [3]. Most patients are diagnosed at an advanced stage and have a poor prognosis. Therefore, early detection and diagnosis of NSCLC will play an important role in improving the prognosis of patients. However, there are many problems in the commonly used clinical diagnosis methods [4]. For example, medical imaging is over-diagnosed and costly, while histopathological biopsy is invasive and has certain trauma to the body [4,5]. Excitingly, liquid biopsies use innovative minimally invasive techniques that are simpler and less expensive than traditional methods [6]. As a new detection technology, it is becoming a powerful weapon to achieve accurate decision-making in clinical oncology [7].

In recent years, with the development of molecular biology technology, several studies have shown that serum microRNA (miRNA) can be used as an ideal marker for the diagnosis of tumors [8]. MicroRNAs (miRNAs) are single-stranded, short-length endogenous non-coding RNAs. As a new biomarker, miRNA can play a positive role in the early diagnosis and prognostic monitoring of tumors, especially for patients with no obvious clinical symptoms and no specific indications [8,9]. A number of studies have shown that miR-125b and miR-21 are significantly correlated with NSCLC [10–13]. Accurate quantification of miRNA molecules is an important prerequisite for its application in clinical diagnosis such as early detection and screening of tumors. The commonly used miRNA detection methods include polymerase chain reaction (PCR), fluorescence, colorimetry, electrochemistry [14–17], which have the problems of complex operation, large sample consumption, high detection cos and low detection efficiency. Therefore, it is urgent to develop a new detection method for miRNA that is fast, simple, low cost and accurate with small sample size.

Surface-enhanced Raman spectroscopy (SERS) is a sensitive detection technique. By adsorbing the molecule of the measured substance on the surface of the crude metal (gold, silver, copper), the molecular signal of the measured substance is significantly enhanced, and the detection sensitivity is improved [18]. As a non-invasive detection method, SERS has the advantages of fingerprint identification, non-destructive detection, no interference from water peaks, real-time analysis [19]. With the development of optical, biochemical and nano-devices, SERS has been widely used in biological analysis, disease detection [20–22]. In recent years, Au or Ag-coated magnetic nanomaterials have shown great advantages in SERS analysis due to their good SERS enhancement effect and magnetic enrichment [23–25]. Fe₃O₄@AuNPs is more remarkable because the nanogap on the gold shell is very narrow, which can form lots of stable hot spots. Microfluidic chip can integrate the reaction, separation and detection functions of chemical, biological and medical related analysis processes on a single chip and automatically complete the entire analysis process [26–28]. Microfluidic chip has the advantages of small size, less sample consumption, fast reaction speed, portability and large-scale production. The microfluidic chip can be combined with SERS technology, their advantages can be comprehensively applied to achieve rapid, accurate, high sensitivity and specific detection of samples, which has broad application prospects in disease surveillance [29,30]. At present, many biosensors for miRNA detection have been developed by combining SERS technology with enzyme-assisted strategy or enzyme-free amplification strategy. For example, Chen et al. developed a ratiometric SERS biosensor based on mismatch catalyzed hairpin assembly (CHA) for sensitive and reproducible microRNA detection [31]. Hu et al. developed an ultrasensitive SERS sensor based on roll circle amplification (RCA) for nucleic acids detection [32]. Xu et al. quantified SERS detection of a variety of breast cancer miRNAs based on duplex-specific nuclease-mediated signal amplification [33]. However, these methods are complex and inefficient. Therefore, a simple and sensitive method is urgently needed to be combined with SERS microfluidic chips to achieve efficient detection of target miRNA.

Herein, a SERS microfluidic chip based on Fe₃O₄@AuNPs and double amplified signal Off-On strategy was constructed for the detection of miR-21 and miR-125b in NSCLC. Step one, prepare Fe₃O₄@AuNPs. As shown in Fig. 1(A), polyethyleneimine (PEI) was prepared on the surface of Fe₃O₄ nanoparticle. Then, AuNPs was cladded on the Fe₃O₄@PEI surface to obtain Fe₃O₄@AuNPs. Step two, preparation of SERS nanoprobes. As shown in Fig. 1 (B), the double strand consisting of Raman signaling molecules-labeled hpDNA and ssDNA was attached to the Fe₃O₄@AuNPs surface by sulfhydryl groups. Meanwhile, the ssDNA hybridized with hpDNA as a placeholder, destroying the hairpin structure of hpDNA, making the Raman signal molecules far away from the Fe₃O₄@AuNPs surface. The signal was turned off. Step three, SERS detection was performed. As shown in Fig. 1 (C), SERS nanoprobes and the sample were added to the two inlets of SERS microfluidic chip respectively. When the two parts were mixed, ssDNA bound to the target due to the presence of the target. At this time, Raman signaling molecules-labeled hpDNA returned to the hairpin conformation, the Raman signaling molecule was close to Fe₃O₄@AuNPs. The signal was turned on. The Raman signal was amplified once. Subsequently, the mixture flowed into the collection chamber and Fe₃O₄@AuNPs was enriched due to magnetic adsorption, forming the second amplification of SERS signal. By fitting the recorded Raman peak intensity and concentration, the quantitative detection of two markers can be realized. This work has the following innovations. The synthesized Fe₃O₄@AuNPs has abundant internal “hot spots” and narrow nanogaps of molecular magnetic properties, which showed good SERS property. In addition, significant SERS signal amplification can be achieved through magnetic aggregation. The double amplified signal Off-On strategy makes the detection process simple and efficient, without the need for target labeling and any subsequent washing steps. The SERS microfluidic chip applied with this strategy has an ultra-low detection limit, high sensitivity and specificity.

 

Fig. 1. Scheme. (A) the synthetic process of Fe₃O₄@AuNPs. (B) Functionalized Fe₃O₄@AuNPs preparation and signal Off-On strategy. (C) SERS microfluidic chip for detecting the target and the second amplification of SERS signal.

Download Full Size | PDF

2. Page layout and length

2.1 Materials and reagents

Chloroauric acid (HAuCl₄), trisodium citrate, ferroferric oxide (Fe₃O₄), (3-aminopropyl) triethoxysilane (APTES), phosphate-buffered saline (PBS), Tris (Hydroxymethyl) Aminomethane Hydrochloride (Tris-HCl), and magnesium chloride (MgCl₂) was purchased from Sangon Biotech (Shanghai) Co., Ltd. Cyanine 5 (Cy5) labeled hpDNA1, 5-carboxyfluorescein (5-FAM) labeled hpDNA2 and other nucleotide sequences used in the experiment (Table 1) were synthesized by Suzhou GENEWIZ Biotechnology Co. Polyethylene glycol (PEG) and polydimethylsiloxane (PDMS) was purchased from Sinopharm Chemical Reagent Co. Real-time quantitative polynucleotide chain reaction (qRT-PCR) kit was purchased from Guangzhou Ruibo Biotechnology Co. All glassware used in the experiments were cleaned using aqua regia immersion and all used in the experiments were deionized water (>18.3 Ω).

Table 1. Nucleotide sequences used in the experiment

View Table | View all tables in this article

2.2 Clinical sample

Human blood samples were obtained from the People's Hospital of Yangzhong City (Table S1). All experiments were conducted in accordance with relevant laws and consent forms were obtained from all subjects. Serum was collected by venipuncture in vacuum blood collection tubes and then centrifuged at 12000rpm (4 °C, 10 min). All specimens were stored at -80 °C until required. Table S1 lists the details of the volunteers

2.3 Preparation of AuNPs

The Au nanoparticles (AuNPs) were synthesized using a simple one-step synthesis. To obtain a burgundy-colored Au nanoparticle solution, 50 mL of HAuCl₄ solution (1 mM) was heated and boiled, then 7.5 mL of trisodium citrate solution (40 mM) was slowly added for 20 min, and the solution was concentrated by centrifugation (8500 rpm, 15 min) and then prepared for use.

2.4 Preparation of Fe₃O₄@AuNPs

To make Fe₃O₄ more dispersed, the Fe₃O₄ solution was ultrasonicated for 25 min. then, 5 mL of PEI solution (50%) was added to it and stirred stably at room temperature for 8 h. The above solution was centrifuged (10000 rpm, 30 min), washed with ethanol solution three times, and finally dispersed in 5 mL of ethanol. 1 mL of the prepared Fe₃O₄@PEI solution was taken and mixed with enough AuNPs solution (50 mL) and the solution was sonicated for 30 min to form the core-shell structure of Fe₃O₄@AuNPs. The product was magnetically separated and dispersed in 5 mL of ethanol solution.

2.5 Synthesis of SERS nanoprobes

Briefly, 10 µL of hpDNA (hpDNA1, hpDNA2) solution (10 µM) and 10 µL of single-stranded DNA (ssDNA1, ssDNA2) solution were mixed in PBS buffer and incubated at 37 °C for 12 h to generate complementary double strands, and then centrifuged (8000 rpm, 10 min) to remove excess unlinked single-stranded DNA. Next, 40 µL (resuspension with PBS buffer) of the above mixed solution and 0.5 mL of MgCl₂ solution (3.5 mM) were added to 5 mL of Fe₃O₄@AuNPs solution prepared in Section 2.4, and the reaction was stirred homogeneously for 12 h. Next, mPEG-SH (1 µM) was added to the mixed solution and the reaction was performed for 1 h for more stability. Centrifugation was processed (8000 rpm, 10 min) and re-dissolved using 10 mM Tris-HCl buffer (pH 8.0) to synthesize the SERS nanoprobes.

2.6 Fabrication of the microfluidic chip

In this paper, a pump-free microfluidic chip with a self-contained magnet device was proposed, as shown in Scheme 1. The chip consisted of three parts: the PDMS lid, the glass sheet, and the built-in magnet. The PDMS cover included two inlets, a serpentine channel reaction zone, a rectangular detection zone, and an outlet, among other refinements. The draft design was completed by AutoCAD software. Subsequently, a master of the PDMS cover was created using UV lithography. The PDMS prepolymer and curing agent were mixed in a 10: 1 weight ratio, and then the mixture was evacuated in a vacuum chamber for 30 min to remove air bubbles. Next, the mixture was poured onto a master mold and placed on a hot plate at 72 °C for 2 h to cure. After cooling to room temperature, the PDMS layer was separated from the template and the inlets and outlets were drilled into the PDMS layer using a hole punch. At this point, the PDMS coverslips and slides were placed in a plasma cleaner and subjected to plasma treatment for 30 s. After removing the coverslips, the coverslips and slides were laminated, and most of the microfluidic chip preparation had been completed. Finally, PEG immersion was employed to hydrophilize the chip for autonomous liquid flow.

2.7 Measurement and characterization

SERS nanoprobes and samples were added to each of the two inlets of the SERS microfluidic chip. The two parts of the flow entered the reaction zone and mixed. 30 min later the SERS assay was performed. Scanning electron microscopy (SEM) image was obtained by a field-emission scanning electron microscope (FE-SEM, Hitachi S-4800) operated at 5.0 kV. The transmission electron microscopy (TEM) image was obtained by a transmission electron microscope (TEM, Philips Tecnai 12) operated at 120 kV. The crystal structure of Fe₃O₄@AuNPs was characterized by a field-emission transmission electron microscope (FE-TEM, FEI Tecnai G2 F30 S-TWIN) operated at 300 kV. The ultraviolet-visible (UV-vis) absorption spectra were recorded via a Cary 5000 spectrophotometer (Varian). Raman spectra were measured via a Renishaw inVia Raman microscope with a laser power of 5 mW. SERS spectra were recorded in the reaction region at 785 nm with a 50-objective. The exposure time was 10 s in all measurements.

3. Results and discussions

3.1 Characterization of nanomaterials

The Fe₃O₄@AuNPs prepared in this paper were characterized using SEM and TEM. As illustrated in Fig. 2(A), the resulting Fe₃O₄ microspheres had an excellent spherical morphology with a diameter of about 280 nm, which was the basis for the subsequent preparation of morphologically stable and uniformly sized Fe₃O₄@AuNPs. Observation of the magnetic nanomaterial Fe₃O₄@AuNPs revealed that the AuNPs were uniformly coated on the surface of the magnetic core (Fe₃O₄) and exhibited excellent morphological homogeneity and good dispersion (Fig. 2(B)). The Fe₃O₄@AuNPs under TEM were clearly hierarchical structures with the surface gold shells possessing nanoscale cracks (Fig. 2(C)), and this rough nanogap could generate massive hotspots to enhance the SERS performance. Figure 2(D) was an HRTEM image of the gold shells on the surface of Fe₃O₄@AuNPs, showing clear lattice streaks with a layer spacing of 0.24 nm, corresponding to the {111} face of Au. In order to gain insight into the composite structure of Fe₃O₄@AuNPs, high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and energy dispersive X-ray (EDX) spectroscopy were applied to observe them. The HAADF-STEM image in Fig. 2(E) revealed that Fe₃O₄@AuNPs consisted of dense AuNP shells and magnetic cores. Figures 2(F-H) were the elemental mapping images of Au (blue), Fe (orange), and O (red) composing the composite nanomaterials, further demonstrating the elemental composition of Fe₃O₄@AuNPs. The EDX spectrum of Fig. 2(I) indicated that Fe₃O₄@AuNPs was mainly composed of Au, Fe, and O, while the peaks of Cu were caused by the Cu lattice that carries the sample, and no other elements were found. The optical properties of Fe₃O₄@AuNPs were particularly important. As shown in Fig. 2(J), the plasmon resonance band appears near 512 for AuNPs with a diameter of 16 nm. After the AuNPs were coated with Fe₃O₄ magnetic nanoparticles, the absorption band broadened and red-shifted. This shift was caused by the coupling of surface plasma excitations induced by the highly rough surface of Fe₃O₄@AuNPs.

 

Fig. 2. (A) SEM image of Fe₃O₄ microspheres. (B) SEM and (C) TEM images of Fe₃O₄@AuNPs. (D) HRTEM and (E) HAADF-STEM images of Fe₃O₄@AuNPs. (F-H) elemental mappings of Fe₃O₄@AuNPs. (I) EDX and (J) UV-vis absorption spectrums of Fe₃O₄@AuNPs. (K) SERS spectra of pure DTNB and DTNB labled Fe₃O₄@AuNPs.

Download Full Size | PDF

To evaluate the SERS enhancement effect of Fe₃O₄@AuNPs, Fig. 2(K) recorded SERS spectra of pure DTNB (1 M) and DTNB-labeled Fe₃O₄@AuNPs (10⁻⁶ M), respectively. The intensity of the characteristic peak of DTNB at 1335 cm^-1 was selected to calculate the enhancement factor (EF) to quantify the SERS activity of Fe₃O₄@AuNPs. The formula was as follows: EF = (I_SERS/C_SERS)/(I_RS/C_RS). Here, EF = 3.09 × 10⁷ when C_SERS = 10⁻⁶ M and C_RS = 1 M, indicating that the material has prominent SERS enhancement.

3.2 Experimental optimization

The concentration of hpDNA in the experiment and the detection time were optimized for the best performance. The concentration of hpDNA was gradually increased while the SERS signal intensity was stabilized at 10 µM by ensuring the same spiked amount (Fig. 3(A)). This was because the hpDNA (labeled with Raman signaling molecules) modified on Fe₃O₄@AuNPs has reached saturation. Therefore, 10 µM hpDNA was used to prepare SERS probes in subsequent experiments. In addition, to evaluate the optimal detection time of the SERS microfluidic chip, the SERS signals in the collection area were obtained at different times after sample addition and incubation (Fig. 3(B)), and it was seen that the signals had basically reached the threshold at 10 min, and then the SERS signals did not appear to be significantly elevated with the increase of time. Therefore, the detection time was chosen as 10 min.

 

Fig. 3. (A) The hpDNA concentration optimization and (B) assay time optimization.

Download Full Size | PDF

3.3 Performance evaluation of SERS microfluidic chips

Figure 4(A) demonstrated the pump-free microfluidic chip fabricated in this work, which was 8.5 cm in length and 2.5 cm in width, featuring the advantages of small size and convenience. To assess the effect of the hydrophilicity of the chip and the closure of the system, red and blue inks were added dropwise to each of the two spiking ports and allowed to flow spontaneously under the action of a capillary pump. As indicated in Fig. 4(B), the ink flowed programmatically in the hydrophilic microchannels and could fill the entire channel in 60 s without any leakage. Therefore, this device allowed for subsequent automated integrated testing without the requirement for a heavy-duty syringe pump. To evaluate the ability of this microfluidic device to collect magnetic nanomaterials, the rectangular detection chamber was microscopically photographed during chip use and documented in Fig. 4(C). It could be easily seen that the aggregation of magnetic nanomaterials in the detection chamber was obvious with time, and a clear aggregation could already be observed at 2 min. PDMS as the raw material of the chip itself had a Raman signal, so it was necessary to verify whether it affected the results. The chip’s two regions (I and II) were selected for SERS detection, as marked in Fig. 4(D). Region I represented the serum samples and PDMS that were not reacted with the Raman probe. In contrast, region II represented the products of the Raman probe after binding to the target. It was obvious that both at 1176 cm^-1 and 1376 cm^-1, the signal intensity measured in Region I was much smaller than that in Region II, so the chip of PDMS material would not affect the subsequent Raman detection.

 

Fig. 4. (A) Physical photo of a pump-free microfluidic chip (B) digital images of the automatic flow of red and blue ink over time in a hydrophilic microchannel. (C) photomicrographs of rectangular inspection chambers according to time. (D) SERS intensity obtained at 1176 cm^-1 and 1367 cm^-1 by Raman detection according to time.

Download Full Size | PDF

3.4 Reproducibility and stability analysis

After verifying the mobility of the SERS microfluidic chip and the magnetic attraction effect in the collection area, we evaluated the SERS microfluidic chip's reproducibility and stability. The reproducibility of a microchip affects whether it was reproducible or not, so we prepared four different batches of SERS microfluidic chips and used them for the same samples. The assay results indicated that the SERS spectra (Fig. 5(A)) obtained from different batches of prepared SERS microfluidic chips had consistent waveforms (with some differences in intensities); the intensities of the characteristic peaks corresponding to the positions at 1176 cm^-1 and 1367 cm^-1 were shown in Fig. 5(B). Therefore, this SERS microfluidic chip had good reproducibility. In addition, the SERS microfluidic chip was stored at room temperature for different times (0d, 7d, 14d, 21d, 28d), and then the same sample was tested, and the obtained SERS spectra and corresponding characteristic peak intensities were shown in Fig. 5(C) and 5(D). Compared with no storage, the intensity of the SERS spectra obtained from the SERS microfluidic chip after 28 d of storage only decreased by about 11.02%. Overall, the stability of this SERS microfluidic chip can meet the practical applications.

 

Fig. 5. (A) SERS spectra obtained from different batches of SERS microfluidic chips and (B) the folded plots of the SERS intensities corresponding to 1176 cm^-1 and 1367 cm^-1. (C) SERS spectra obtained from SERS microfluidic chip after storage for different times (0d, 7d, 14d, 21d, 28d) and (D) the SERS intensity histograms corresponding to 1176 cm^-1 and 1367 cm^-1.

Download Full Size | PDF

3.5 Specificity and sensitivity measurements of SERS microfluidic chip

The ability to effectively distinguish the target was of paramount importance for the practical application of this device. In this investigation, we assessed various target (including Blank, Random, MT1-1 & MT1-2, MT3-1 & MT3-2, miR-21 & miR-125b) using the SERS microfluidic chip. The resulting spectra were depicted in Fig. 6(A), highlighting two distinctive peaks. Signal intensities are presented in Fig. 6(B) and 6(C). Notably, when the target was present, the SERS signals obtained were markedly higher than those originating from other interfering substances. Furthermore, the microfluidic chip demonstrates the capability to differentiate single-base and three-base mismatched sequences (MT1-1 & MT1-2, MT3-1 & MT3-2). Consequently, this SERS microfluidic chip exhibits excellent specificity in target discrimination amidst interfering elements.

 

Fig. 6. (A) SERS spectra obtained using a SERS microfluidic chip to detect different targets (Blank, Random, MT1-1 & MT1-2, MT3-1 & MT3-2, miR-21 & miR-125b), (B) the signal intensity of the characteristic peak at 1176 cm^-1 and (C) the signal intensity of the characteristic peak at 1367 cm^-1. (D) SERS spectra obtained after addition of different concentrations (10 aM, 100 aM, 1 fM, 10 fM, 100 fM, 1 pM, 10 pM) of the target in PBS buffer. (E) Standard curve of the intensity of the characteristic peak at 1176 cm^-1 versus the logarithm of miR-21 concentration. (F) Standard curve of the intensity of the characteristic peak at 1367 cm^-1 versus the logarithm of miR-125b concentration.

Download Full Size | PDF

In addition, assay sensitivity was crucial in practical applications, especially for low-abundance biomarkers in the early stages of cancer. Sample solutions (miR-21: miR-125b = 1:1) were made by adding different concentrations of miR-21 & miR-125b in PBS buffer, which were used to analyze the detection sensitivity (Fig. 6(D)). It can be seen that the obtained SERS signals gradually increased with increasing concentration. Here, there was a good linear relationship between the logarithm of miR-21 concentration and the intensity of the characteristic peak at 1176 cm^-1 with the linear regression equation: y = 2586.97x + 47778.44 (R²= 0.982) (Fig. 6(E)). Similarly, the logarithm of miR-125b concentration and the intensity of the characteristic peak at 1367 cm^-1 had a linear regression equation: y = 2521.23x + 46682.74 (R²= 0.983) (Fig. 6(F)). The corresponding LODs of 6.38 aM and 7.94 aM were calculated by the limit of detection formula ($LOD = 10\frac{{\left( {{C_{blank}} + 3SD} \right) - a}}{b}$), where C_blank was the SERS intensity of the blank sample, SD was the standard deviation, a and b were the variables obtained with a linear regression of the signal-concentration curve. The sensitivity of the proposed SERS microfluidic chip also belongs to the forefront after comparing it with the currently reported methods (Table 2).

Table 2. Comparison of the proposed method with currently reported methods

View Table | View all tables in this article

3.6 Clinical diagnosis of lung cancer

A case of a female patient treated in the People’s Hospital of Yangzhong. X-ray examination results (Fig. 7(A)) showed nodules in the right third and fourth anterior interrib. After CT examination, nodules were found in the middle lobe of the right lung, with a clear boundary and superficial leaves (Figs. 7(B-D)). HE staining of surgically removed samples confirmed lung cancer. The histological type was mucinous adenocarcinoma (Fig. 7(E)). Immunohistochemistry found that ALK (Fig. 7(F)) and CK5-6 (Fig. 7(G)) proteins were negatively expressed in tumor cells. CK7, EGFR, Ki67, Napsin A, P53, and the TTF-1 protein, were positively expressed in the tumor cells (Figs. 7(H-M)). Special staining found that the elastic fiber staining was positive (Fig. 7(N)).

 

Fig. 7. Clinical diagnosis of lung cancer. (A) X-ray examination results. (B-D) CT findings for transverse, coronal and sagittal sites. (E) HE staining of tumor samples after surgery (200×). (F-N) Immunohistochemical staining (ALK, CK5-6, CK7, EGFR, Ki67, Napsin A, P53, TTF-1, and elastic fibers, respectively, (200×).

Download Full Size | PDF

3.7 Clinical sample analysis

Serum spectra of 25 healthy individuals and 25 NSCLC patients were acquired using this SERS microfluidic chip and averaged to obtain an average spectrum (Fig. 8(A)), corresponding to the characteristic peak intensities at 1176 cm^-1 and 1367 cm^-1 were shown in Fig. 8(B), which shows that the intensity of SERS was significantly higher in NSCLC patients than in healthy individuals. Meanwhile, the expression levels of the markers calculated from the characteristic peak intensities at 1176 cm^-1 and 1367 cm^-1 using standard curves were compared with the detection results of the qRT-PCR method, which was used to validate the detection accuracy and the clinical detection effect of this microarray (Fig. 8(C and D)). In addition, the comparison of the obtained SERS with the detection results of qRT-PCR using Student's t-test showed no statistical difference (Table S2). Therefore, the detection efficacy of this microarray was comparable to that of the qRT-PCR method and has good prospects for application.

 

Fig. 8. (A) Spectra of miR⁃21 and miR⁃125b in serum of healthy people and NSCLC patients. (B) The histogram corresponding to the characteristic peak intensity of 1176 cm^-1 and 1367 cm^-1. SERS and qRT-PCR were used to detect (C) miR-21 and (D) miR-125b in the serum of healthy people and NSCLC patients (p > 0.05).

Download Full Size | PDF

4. Conclusion

The proposed SERS microfluidic chip integrates a signal switching strategy and magnetic attraction effect to realize dual signal amplification for sensitive and rapid detection of NSCLC-associated miRNAs. In this SERS microfluidic chip, miniature magnets aggregate the reaction end-products into the collection area. By detecting the SERS signal in the collection zone and substituting it into the corresponding standard curve equation, the level of markers present in the sample can be quantitatively analyzed. Meanwhile, the LOD of this protocol for miR-21 and miR-125b in the detection range of 10 aM-10 pM was 6.38 aM and 7.94 aM, respectively, with good specificity and sensitivity. In addition, the detection accuracy of this protocol was comparable to that of qRT-PCR, which is a promising new method for early diagnosis of NSCLC.