Open Access
Add to BibSonomy
Abstract
A surface-enhanced Raman scattering (SERS) detection platform was constructed based on Au nano-dodecahedrons (AuNDs) functionalized with nucleic acid aptamer-specific binding and self-assembly techniques. SERS labels were prepared by modifying Raman signaling molecules and complementary aptamer chains and were bound on the aptamer-functionalized AuNDs array. Using this protocol, the limits of detection (LODs) of miR-21 and miR-18a in the serum were 6.8 pM and 7.6 pM, respectively, and the detection time was 5 min. Additionally, miR-21 and miR-18a were detected in the serum of a mouse model of colorectal cancer. The results of this protocol were consistent with quantitative real-time polymerase chain reaction (qRT-PCR). This method provides an efficient and rapid method for the simultaneous testing of miRNAs, which has great potential clinical value for the early detection of colorectal cancer (CRC).
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Colorectal cancer (CRC) is the third most prevalent malignancy worldwide and is the second most prevalent contributor to death after lung cancer [1], with a 5-year overall survival rate of only 63.5% when medical intervention is applied [2]. Mortality rates for colorectal cancer patients who received early diagnosis and treatment declined by about 35% from 1990 to 2007 [3]. In the last few years, the prevalence and mortality rates of colorectal cancer have been on the rise among Asians, especially Chinese, which is related to lifestyle changes and dietary changes [4,5]. Furthermore, the prevalence of early-onset colorectal cancer (individuals below the age of 50 who have been diagnosed with colorectal cancer) has escalated at an even greater rate, imposing a greater sense of urgency and demanding stricter protocols for the timely detection of CRC [6]. Colorectal cancer is currently diagnosed early through colonoscopy and pathologic diagnosis, which can be expensive, time-consuming, and reliant on the examiner's expertise [7,8]; Due to the invasive nature of colonoscopy, it is more difficult for younger people to undergo it than for older people who already use it as a routine medical test, which leads to an increase in the rate of missed diagnosis of early CRC [9]. Despite the availability of fecal occult blood tests, intestinal tumor biomarkers (such as CEA, CA-199, etc.), and other means of early screening, researchers have shown that the accuracy and sensitivity of these methods for detecting early colorectal cancer are still questionable [10], thus, there is an imperative to investigate novel biomarkers that can aid in the identification of early colorectal cancer.
Small non-coding ribonucleic acid (RNA) named microRNA (miRNA) can modulate the expression of target proteins by binding to their corresponding messenger RNA (mRNA) [11]. Several studies have found that the development of colorectal cancer is accompanied by alterations in the levels of miRNAs in body fluids [12–14]. Moreover, miRNA exhibits superior accuracy in identifying CRC compared to conventional blood biomarkers (such as CEA and CA-199) [15]. Hence, the use of miRNAs as detection markers will greatly enhance the possibility of early diagnosis of CRC. miR-21 is a miRNA that has been intensively investigated in colorectal cancer, and multiple research studies have unveiled that the expression of this sequence is altered in patients with colorectal cancer; Amal and colleagues made the discovery that elevated serum miR-21 expression levels correlated with poor prognoses such as increased rates of tumor node metastasis (TNM) classification, hematogenous metastasis, liver infiltrate, and recurrence rate [16]; The researchers in an academic study conducted by Mennatallah et al analyzed the levels of miR-21, miR-18a, and miR-92a in the blood sample to compare the expression between CRC patients and healthy individuals. The findings of the study revealed that colorectal cancer patients exhibited noticeably raised levels of serum miR-18a, miR-21, and miR-92a expression compared to healthy controls [17]. miR-21 and miR-18a may be genomic biomarkers for the early detection of colorectal cancer. Currently, various strategies have been developed to detect tumor-associated miRNAs, such as qRT-PCR, northern blotting, microarray, and electrochemical methods [18,19]; however, these methods still have some problems such as complicated operation, expensive equipment, and residual detection reagents when they are widely used in cancer diagnosis [20]. It is of concern that no single miRNA is highly specific for a particular tumor. To improve the accuracy of the assay and reduce the false-positive rate, simultaneous detection and mapping of multiple miRNA concentrations seems to be a promising approach [21,22]. When compared with single-indicator assays with low specificity towards cancer type, multiplexed assays for miRNAs help provide more comprehensive, robust, and accurate diagnostic information and have the advantage of saving time, reagents, and labor costs compared with single-assay testing of individual biomarkers.
Surface-enhanced Raman scattering (SERS), a method known for its ability to provide unique molecular fingerprinting information and its potential for sensitivity down to the single molecule level, has found extensive applications in liquid biopsy, bio-imaging, biochemical sensing, instant diagnostics, and related disciplines. Notably, SERS also offers the advantage of supporting multiplexed detection, resisting photo-bleaching and photodegradation, and being immune to water interference [23–26]. The key providers of Raman signal amplification are strongly enhanced regions of electromagnetic fields called “hot spots” generated by localized surface plasmon resonances (LSPR). LSPR forms when the frequency of electron vibrations conducted by metal nanoparticles is harmonized with the frequency of incident photons, resulting in these hot spots [27,28]. The core of a SERS sensor is the interaction between the substance to be detected and the substrate. The design and fabrication of high-performance SERS substrates are the key to driving the development of SERS technology. Researchers have extensively investigated various types of SERS substrates to obtain substrates capable of producing ample hotspots to enhance SERS [29–31]. Wu et al. modulated gallium nanoparticles of different diameters as a substrate, which produced a significant and stable enhancement effect, and the Raman signal intensity increased linearly with the increase of signal molecule concentration [32]. Tseng et al. applied the laser direct writing technique to silver oxide to prepare thin films with high SERS activity [33]. Substrates prepared from plasmon nanorods likewise yielded significant enhancements, demonstrating that the geometry of the nanomaterials is crucial for the strength of the SERS signal [34]. SERS substrates composed of Au polyhedral nanoparticles arranged in a highly organized manner have gained considerable interest in recent years [35–38]. Au nano-dodecahedrons (AuNDs) have been favored by researchers due to their stable spatial structure, large specific surface area, excellent spatial reproducibility, and high homogeneity. Additionally, AuNDs exhibit distinct plasmon resonance peaks and robust near-field coupling effects between particles, leading to their extensive utilization as highly efficient SERS-active substrates. Nucleic acid aptamers are a set of single-stranded oligonucleotide structures that can bind specifically to a target through non-covalent interactions and have been widely used to create molecular imaging labels due to their high affinity, affordability, and high stability [39,40]. Self-assembly, which is a facile and easy strategy to utilize the weak interactions between nanoparticle ligands to bring the particles together in an orderly manner [41,42], has been widely studied and applied in the preparation of monolayer membrane structures. The one-step hybridization paired reaction greatly reduces detection time, avoids cumbersome amplification, simplifies the reaction procedure, and ensures an easy and specific reaction [43]. Therefore, an effective combination of nucleic acid aptamers with self-assembly technology may be a reliable method to enhance the efficiency of miRNA detection and thus early diagnosis in CRC patients.
In this study, we present a novel technique for the detection of two specific miRNAs associated with colorectal cancer. The technique involves the use of Au nano-dodecahedrons (AuNDs) as a substrate that exhibits SERS activity for highly sensitive detection. The detection is made possible by utilizing the strong binding ability of nucleic acid aptamers to miRNAs (Fig. 1). Au nanocubes (AuNCs) modified with 4-mercaptobenzoic acid (4-MBA) and 5,5′-Dithiobis (2-nitrobenzoic acid) (DTNB) were used for the study, which was co-modified on the surface of AuNCs along with two complementary strands of nucleic acid aptamers (H1 and H2), which were attached to the self-assembled active substrate of AuNDs via two nucleic acid aptamer strands (cDNA1 and cDNA2) to generate a strong Raman signal. When the target miRNA was added to the reaction unit, the Raman signal was reduced due to more preferential hybridization, where the nucleic acid aptamer binds to the target nucleic acid strand, and the AuNCs carrying the Raman reporter gene competed away from the AuNDs active substrate. The sensitivity of the detection platform was evaluated by examining the relevance between the intensity of a specific peak and the concentration of the target. Determining the potential practical application value of our detection platform, we established a mouse model with colorectal cancer and measured the levels of miR-21 and miR-18a in the mouse serum at various stages. In comparison with traditional detection methods, the detection platform is characterized by rapid and convenient detection, accuracy, and sensitivity, and this work provides a new idea for multiple biomarker detection of colorectal cancer.
Fig. 1. (A) Synthesis of SERS label; (B) Establishment of SERS detection platform; (C) Principle of detection platform.
2. Experimental section
2.1 Materials
Chloroauric acid trihydrate (HAuCl4-3H2O), which is commonly used in Au nanoparticle synthesis, sodium borohydride (NaBH4), ascorbic acid (AA), Hexadecyl trimethyl ammonium chloride (CTAC), potassium iodide (KI), n-hexane (C6H14), phosphate-buffered saline (PBS), The carboxybenzenethiol, commonly referred to as 4-mercaptobenzoic acid (4-MBA), can adhere to diverse metal surfaces. 5,5′-Dithiobis (2-nitrobenzoic acid) (DTNB), a yellow water-soluble compound; N-hydroxysuccinimide (NHS) and 1-ethyl-(3-dimethylaminopropyl) carbodiimide (EDC) were used as an activating component of the reaction, tris(2-carboxyethyl) phosphine (TCEP). The above reagents were purchased from Sinopharm Chemical Reagent Suzhou Co. The purity of the reagents used in the experiment was analytically pure (AR). Shanghai Bioengineering Co provided unprocessed bovine serum albumin (BSA) was used to prevent the coagulation and precipitation of nanoparticles in solution while reducing the immunogenicity of the nanoparticles. GeteinBiotech (China) provided the quantitative real-time polymerase chain reaction (qRT-PCR) kit. The silicon wafers used for the experiments were purchased from Sinocrystal Electronic Technology (China), with a thickness of 475 ± 20 µm, doped with boron, and came out of the field cut into rounds of 1 cm diameter to be ready for use. Deionized water with a resistivity of 18.3 MΩ was prepared using a Millipore system. Shanghai Sangong Biotechnology (China) was the source of the nucleic acid aptamer chains (including miR-21 and miR-18a) and nucleic acid complementary aptamer chains mentioned in Table 1. The C57 mouse and MC38 cell lines used for the experiments were received from the School of Clinical Medicine of Yangzhou University.
Table 1. The experiment utilized nucleotide sequences listed in Table 1, read from 5′ to 3′
2.2 Equipment
The Tecnai 12 transmission electron microscope was used to take TEM images at an accelerating voltage of 120 kV. Investigations performed with the S-4800 field emission scanning electron microscope at 10 kV have been completed. The HRTEM and SAED used to characterize the structure of the material were taken with a Tecnai G2 F30 S-TWIN field-emission transmission electron microscope operating at 200 kV. Measurements of absorption spectra in the ultraviolet-visible-near infrared (UV-Vis-NIR) range were conducted using an Agilent UV absorption spectrometer. The DXRxi micro-Raman spectrometer was utilized for the measurement of Raman spectra.
2.3 Preparation of Au nano-dodecahedrons (AuNDs)
Development of an improved procedure for the preparation of AuNDs according to Chung et al [44]. 0.32 g of CTAC was added to 9.9 mL of deionized water and dissolved at room temperature, 120 µL of 1% wt of HAuCl4 was injected into the CTAC solution with slow stirring. To obtain the gold seed solution, 0.6 mL of 10 mM chilled NaBH4 was added to the combined mixture and agitated for 2 min. The solution was left undisturbed for a duration of 2 h to facilitate the full decomposition of the excess NaBH4. The same beakers were configured in two identical growth solutions (marked as A and B), specifically, 0.32 g CTAC was dissolved in 9.485 mL of deionized water, and 250 µL of 10 mM HAuCl4 solution, 20 µL of 1 mM KI solution, and 220 µL of 40 mM AA solution were added sequentially to the two beakers, and the growth solution was obtained by shaking gently. In A beaker, 60 µL of completely decomposed gold seed solution was added, and the solution changed to a light pink color by gently shaking for 5 s. Immediately take 60 µL A beaker solution add it to B beaker, and let it stand for 10 min after gently shaking for 15 s. The growth of AuNDs was completed. The AuNDs were washed twice with deionized water for the next step of characterization and self-assembly.
2.4 Preparation of Au nano-cubes (AuNCs)
A mixture of 7.5 mL of 0.1 M CTAC solution and 0.25 mL of 0.01 M HAuCl4 solution was prepared and the reaction occurred at ambient temperature. Following the addition of 0.6 mL of 10 mM chilled NaBH4 and vigorous stirring for 2 min, the solution was placed unperturbed for 2 h to allow the total decomposition of the excess NaBH4. Subsequently, the solution was diluted by a factor of 10 to obtain the gold seed solution. To configure the growth solution, successively 3 mL of 0.1 M CTAC solution, 0.5 mL of 0.01 M HAuCl4 solution, and 2 mL of 0.1 M AA solution was poured into an unadulterated beaker containing 15 mL of deionized water and mixed homogeneously. Afterward, 10 µL of the seed solution was introduced into the growth solution, given a gentle shaking, and left undisturbed at an ambient temperature for 1 d to facilitate the growth process of AuNCs. Deionized water was washed 2 times for use.
2.5 Self-assembly of AuNDs active substrates and immobilization of nucleic acid aptamers
An oil-water interface was formed by mixing 3 ml of AuNDs solution with an equal volume of hexane. The same volume of anhydrous ethanol was quickly added to the system, which drove the nanoparticles to self-assemble at the oil-water interface to form a tightly arranged monolayer array. After about 2 min, clean silicon wafers hydrophilically treated with piranha solution were contacted with the arrays in parallel, and the arrays of AuNDs were transferred onto the silicon wafers and dried in a desiccator for 15 min. After that, the silicon wafer was submerged in NHS (45 mM) and EDC (180 mM) solutions for about 1.5 h. Subsequently, 200 µl of cDNA1 and cDNA2 (0.1 mM) were mixed with 20 µl of 1 M TCEP solution and reduced at room temperature for 0.5 h. The activated nucleic acid aptamers were coupled to the surface of the AuNDs active substrate, and the excess reagents were washed away with PBS after incubation for 2 h at 37 °C. 1% wt of BSA was used to repair the non-specific binding areas on the surface of the particles.
2.6 Preparation of SERS labels
The reporter genes 4-MBA and DTNB were employed as SERS labels (AuNCs@4-MBA@H1, AuNCs@DTNB@H2) in the experiment by Au-SH modification on the surface of the AuNCs, and the coupling agents were used as EDC and NHS. For the immobilization of Raman signaling molecules on the surface of the AuNCs, 200 µL of a 10 mM solution of 4-MBA dissolved in ethanol was introduced into a 10 mL solution containing AuNCs. The mixture was then vigorously stirred for a period of 2 h at room temperature. Follow this step to eliminate the 4-MBA excess. The centrifugation process was continued at 10,000 rpm for 10 min to dispose of the floating liquid, and the precipitated 4-MBA-labeled AuNCs were redistribution in 10 mL of PBS by a combination of sonication and oscillation, and the process was repeated twice to ensure complete removal of the signaling molecules. 100 µL of 150 mM EDC and 100 µL of 30 mM NHS were added one after the other to the aforementioned solution for a duration of 1.5 h. Subsequently, H1 (0.1 mM) activated by 20 µl of 1 M TCEP solution was added to it, and the mixture was incubated at a temperature of 37 °C for a duration of 2 h. Finally, after introducing 15 µL of a solution of 1% wt BSA into the mixture, it was allowed to undergo a 2 h reaction to hinder non-specific binding areas on the surface of the particles. Subsequently, the resulting solution was subjected to centrifugation at a speed of 8000 rpm for 15 min, enabling the removal of surplus reagents and nucleic acid strands. The precipitate was kept in a PBS solution at 4 °C until it was utilized. Additional SERS labels (AuNCs@DTNB@H2) were created using the identical method.
2.7 Establishment of a mouse model of colorectal cancer and serologic assays
All animal experiments were authorized by the Laboratory Animal Welfare Ethics Committee of Yangzhou University (No. 202307008). To establish a mouse model of colorectal cancer, 10 mice were randomly divided into 2 groups (8 mice in the transplanted tumor group and 2 mice in the blank control group). In the tumor group, the ventral side of the mice was injected with 2.43 × 106 MC38 cells dissolved in 0.2 mL PBS subcutaneously, while the control group was injected with an equal volume of PBS buffer at the same location. Tumor growth was observed daily for 3 consecutive days, and a hard mass of about 0.5 mm was palpable subcutaneously after 5-7 days, suggesting successful modeling. Mouse body weight and mouse tumor volume were continuously monitored during the experiment. All mice were blood sampled on days 1, 6, 11, 16, 21, 26, and 30 to obtain serum for SERS testing.
2.8 SERS measurement
To conduct the measurement, precise experimentation was carried out by gradually introducing 25 mL of the solution to be analyzed onto the prepared detection silicon wafer. Subsequently, the silicon wafer was incubated within a controlled-temperature incubator for a specific duration. Finally, the surface of the silicon wafer was thoroughly cleansed using ultrapure water. For 10 s, the Raman signal-detecting device was run at an excitation wavelength of 785 nm (using a 50× objective lens, the spot of 2 µm, and power of 5 mW), a wavelength that is a choice for assuring detection performance and reducing fluorescence backgrounds. Three repetitions were performed for each sample, with spectra acquired from three different locations (error bars indicate maximum and minimum intensities). The Raman shifts were adjusted to 600-1800cm-1. The obtained SERS spectra were processed with a flat baseline.
3. Results and discussion
3.1 Working principle of the SERS multi-detection platform
This new work developed a SERS and nucleic acid aptamer-based multiplexed detection platform with a high SERS activity substrate for specific capture of target sequences and SERS detection with the workflow shown in Fig. 1. 4-MBA and DTNB were selected as Raman reporter genes to prepare two SERS labels, which have well-characterized Raman peaks. Raman signaling molecules and complementary strands of nucleic acid aptamers (H1, H2) were co-modified on AuNDs (Fig. 1(A)). AuNDs substrates were successfully prepared and transferred onto clean silicon wafers, and attachment of nucleic acid aptamers to AuNDs substrates was used to synthesize active arrays that capture the target nucleic acid strands (Fig. 1(B)). The design principle of the detection process for colorectal cancer mouse serum samples is as follows (Fig. 1(C)): before adding the serum to be tested, SERS labels carrying Raman signaling molecules are attached to the nucleic acid aptamers of the active array through complementary chains, and this proximity of hot spots generated by the array leads to a significant enhancement of the SERS signals. After serum addition, the nucleic acid aptamers (cDNA1 and cDNA2) in the system were separated from the complementary strands and attached to the target nucleic acid strands (miR-21 and miR-18a) due to preferential hybridization, consequently reducing the SERS signals as the 4-MBA and DTNB molecules dissociated from the array surface. As the levels of miR-21 and miR-18a increase in the stream, the decrease in signal will be even more pronounced. It is thus feasible to measure the levels of miR-21 and miR-18a in the blood by assessing the extent of weakening of SERS intensity of 4-MBA and DTNB.
3.2 Characterization of AuNCs
Figure 2 characterizes the structure of AuNCs. Figure 2(A) and (B) display SEM and TEM images of AuNCs, illustrating the successful preparation of AuNCs with prominent prismatic edges. The measurements indicate that each edge of the AuNCs measures approximately 77 nm. Figure 2(C) is the HRTEM images of AuNCs, and Fig. 2(D) shows the local magnification of HRTEM images depicting the lattice stripe spacing of Au nanoparticles of 0.240 nm. Figure 2(E) shows that AuNCs have a strong absorption peak at 532 nm through the UV-Vis-NIR spectrum. Figure 2(F) was utilized to examine the enhancement effect of AuNCs on SERS by conducting a comparative analysis between the Raman signals of higher-concentration pure 4-MBA (10−2 M) and lower-concentration 4-MBA (10−8 M) labeled AuNCs. The Raman signal strength of pure 4-MBA is exceptionally low. In the spectra of 4-MBA (10−8 M) labeled AuNCs, the characteristic peaks of 4-MBA showed a remarkable amplification. This was caused by the creation of localized “hot spots” at the corners of individual AuNCs, along with a substantial enhancement of the LSPR effect between neighboring AuNCs. Therefore, AuNC was used as a SERS label.
Fig. 2. (A) SEM and (B) TEM image of AuNCs. (C) An HRTEM image of AuNC is also presented. (D) The HRTEM image is magnified locally. (E) The UV-Vis-NIR spectrum of AuNC is displayed. (F) The SERS spectra of 4-MBA in its pure form (1 × 10−2 M) and 4-MBA-labeled AuNCs (1 × 10−8 M) are compared.
3.3 Characterization of the AuNDs array
Figure 3 characterizes the structure of the AuNDs. Figure 3(A) and (B) show scanning and transmission electron microscope, respectively, images illustrating AuNDs, with the visualization providing evidence of the hexagonal morphologies of the AuNDs. The average particle size is about 46 nm, demonstrating the successful synthesis of uniformly sized Au nanoparticles with multiple vertices and prongs. The accompanying. Figure 3(C) displays a detailed HRTEM image of the AuNDs, and the individual AuNDs shown in Fig. 3(D) have a lattice-stripe pitch of 0.242 nm, and the electron diffraction pattern depicted in the inset image is a result of directing a high-energy electron beam perpendicular to the AuND. This pattern confirms the presence of single crystals within the incident region. Figure 3(E) shows the electromagnetic field distribution of a single AuND under 785 nm incident laser irradiation, the maximum electromagnetic field strength is found at the angular position of the particle, as indicated by the results. Figure 3(F). displays the AuNDs array, which was formed through self-assembly at the oil-water interface. The uniform distribution of Au nanoparticles is visible. The AuNDs array was further studied with Finite Difference Time Domain (FDTD) simulations to explore the mechanism behind its enhanced performance when widely distributed on a clean silicon wafer surface. The particle size of AuND was set to 46 nm and the inter-particle distance was set to 2 nm, which was consistent with the average actual size of the sample we obtained in the TEM test, with an excitation wavelength of 785 nm. Total-field scattered-field (TFSF) linearly polarized light waves, which are polarized in line with the X-axis with a wavelength range of 300 - 1200 nm, were injected into the unit cell along the -Z direction. Frequency-domain field profile monitors were localized at Z = 0 nm, 5.5, and 23 nm in the x-y plane, respectively. Based on Fig. 3 G, depict the distributions of electric fields in the X-Y plane (top view) at different heights along the Z-axis. In the top layer of the AuNDs array, the electric field has been greatly intensified at z = 23 nm and Z = 5.5 nm. The highest enhanced region of the electric field is found in the gap between adjacent Au dodecahedra, at Z = 0 nm. This is because the spacing between neighboring AuNDs is minimized at Z = 0 nm, and the collective oscillation of the surface electrons is substantially enhanced, resulting in a significant increase in the LSPR effect. The simulation results suggest that the AuNDs array experiences a strong local electromagnetic field, which generates a high density of “hot spots” and thus exhibits excellent SERS enhancement. Figure 3 H shows the UV-visible-near-infrared spectra of AuNDs, as a noble metal nanomaterial, with the maximum absorption peak located at 587 nm. To evaluate the signal enhancement effect of the prepared monolayer tightly arranged AuNDs array, SERS spectra were examined for substrates used with high concentrations of “pure” 4-MBA (10−2 M) and arrays of 4-MBA (10−8 M)-labeled AuNDs. As shown in Fig. 3(I), the enhanced intensity of the characteristic peak at 1593 cm-1 in the signaling molecule is evident on the AuNDs array compared to the pure 4-MBA, showing a good enhancement effect. To measure the SERS amplification effect produced by the array of AuNDs, which indicates a noticeable improvement. The formula $\textrm{EF}\; = \; ({\textrm{I}_{\textrm{SERS}}}/{\textrm{C}_{\textrm{SERS}}}\textrm{)}/\textrm{(}{\textrm{I}_{\textrm{Raman}}}/{\textrm{C}_{\textrm{Raman}}}\textrm{)}$ was used to calculate the enhancement ratio of the AuNDs array, where I and C represent the measured signal intensities and the corresponding concentrations. With the provided concentrations of 10−2 M and 10−8 M, the EF was calculated to be 1.65 ± 0.09 × 108.
Fig. 3. (A) SEM and (B) TEM image of AuNDs. (C) HRTEM image of AuND. (D) Local magnification of HRTEM image. (E) Electromagnetic field distribution of a single AuND. (F) SEM image of an array of AuNDs. (G) Electromagnetic field strength simulation of an array of AuNDs. (H) UV-VIS-NIR spectra of AuNDs. (I) SERS profiles of 1 × 10−2 M pure solution of 4-MBA and 1 × 10−8 M concentration of 4-MBA-labeled AuNDs solution were analyzed.
3.4 Homogeneity and stability testing of 4-MBA-labeled AuNDs
The characteristics of the AuNDs array will have a prominent impact on the detection. The homogeneity of the AuNDs array was analyzed by generating SERS spectra at six randomly selected points on the surface. The Raman intensity at 1593 cm-1 was illustrated in these plots, and it was subsequently visualized as a histogram with a relative standard deviation of 4.46%, and both Fig. 4(A) and (B) indicate that the AuNDs array prepared by us have significant homogeneity. As shown in Fig. 4(C) and (D), to assess the reproducibility of the AuNDs array, SERS mapping was conducted on six distinct batches of 4-MBA-labeled AuNDs array, revealing consistent spectral waveforms among them. Additionally, the signals at 1593 cm-1 exhibited an RSD value of 4.13%, which showed good reproducibility. Finally, SERS analysis was performed on arrays of AuNDs preserved for 1 d, 7 d, and 14 d under ambient conditions. The outcomes are presented in Fig. 4(E) and (F). Following 14 days, the 5.26% decline in the SERS signal's intensity at a frequency of 1593 cm-1 was noted, which confirms that the AuNDs array has excellent temporal stability. The good characteristics of the AuND array demonstrate the ability of the prepared assay platform to perform SERS detection on the SERS detection of miRNAs.
Fig. 4. (A) The SERS spectra of six randomly chosen locations from a collection of Au nanoparticle-decorated surfaces coated with 4-MBA were analyzed. (B) Bar graph of signal intensity at 1593 cm-1. (C) The SERS spectra of six separate sets of AuNDs array labeled with 4-MBA were analyzed. (D) A graph displaying the intensity of signals at 1593 cm-1 in the form of bars. (E) The AuNDs array labeled with 4-MBA was stored for 1 d, 7 d, and 14 d, and the SERS spectra were recorded for each time point. (F) A line graph depicting the changes in signal intensity at 1593 cm-1 over time was generated to correspond with the aforementioned storage periods of the AuNDs array labeled with 4-MBA at room temperature.
3.5 Utility validation and condition optimization
To evaluate whether the detection platform could successfully distinguish miR-21 and miR-18a, the SERS assay was performed in the following cases: both miR-21 and miR-18a were present, only miR-21 was present, only miR-18a was present, and both target chains were missing. As depicted in Fig. 5(A), the coexistence of both the desired target strands produced weaker Raman signals at 1337 cm-1 and 1593 cm-1 than when both target strands were missing. Simultaneously, when a sole miRNA is present, the magnitudes of the corresponding distinguishing peaks’ signals notably decrease without any adverse overlap. Therefore, the practicality and specificity of the detection method were confirmed for simultaneously detecting miR-21 and miR-18a. The reaction incubation time is a key parameter of the assay, and the performance of the detection platform can be further improved by optimizing the reaction time. Figure 5(B) shows that as the duration of incubation increased, a progressive decline in the SERS intensity at 1337 cm-1 was observed until it eventually stabilized at around 5 min, and the intensity change at 1593 cm-1 showed the same result. Therefore, the optimal incubation time was 5 min.
Fig. 5. (A) SERS spectra in the presence of both miR-21 and miR-18a, in the presence of only miR-21, in the presence of only miR-18a, and in the absence of both target strands. (B) Signal intensity changes with prolonged incubation time.
3.6 Simultaneous quantification of the expression levels of both miR-21 and miR-18a in a given sample
The good characteristics of the detection system have been verified, and the lower limit of detection is an important parameter that affects its practical application. Mixed solutions of nucleic acids containing different concentrations of mi-21 and miR-18a, ranging from 10 µM to 10 pM, were prepared by dispersing them in a PBS buffer. The SERS spectra were detected after adding 25 µL of mixed solutions with varying concentrations to the SERS detection silica and incubating them in a constant temperature incubator for 5 min. Figure 6(A) displayed the SERS spectra of miRNAs at different concentrations can be observed. As the concentration of the target nucleic acid increases, the corresponding positions’ SERS signal intensity decreases, and the SERS signal intensities demonstrate a linear correlation with the logarithm of various nucleic acid concentrations, as observed in Fig. 6(B) and (C), and the two are linearly related. The linear regression equations for the relationship between SERS signal intensity and the focus of miR-21 and miR-18a in a PBS buffer containing target nucleic acids ranging from 10 µM to 10 pM can be represented as y = -578.130x-1007.965 (R2 = 0.981) and y = -516.480x-2039.520 (R2 = 0.990), corresponding to which the limit of detection (LOD) of miR-21 and miR-18a in PBS buffer can be calculated as 4.9 pM and 7.4 pM, respectively. The LOD was calculated based on the characteristic peaks of the SERS spectra using the equation: $\textrm{LOD}\; = \; 3{S_b}/b$, where ${S_b}$ is the standard deviation of the SERS intensity of the blank samples at 1337 cm-1 and 1593 cm-1, and b represents the slope of the plotted calibration curve. Meanwhile, different levels of miR-21 and miR-18a were assessed in serum, as depicted in Fig. 6(D), which showed similar results to those in the PBS buffer. As the levels of miR-21 and miR-18a rose, the intensity of the SERS signal at the corresponding position decreased. As shown in Fig. 6(E) and (F), in serum with target nucleic acid concentrations from 10 µM to 10 pM, a linear regression equation was derived to correlate the SERS signal intensity with the concentration's logarithm of miR-21, the outcome is the formation of the equation: y = -595.110x-1303.231 (R2 = 0.983), with a LOD of 6.8 pM, and the miR-18a's corresponding linear regression equation was determined: y = -508.480x-2036.761 (R2 = 0.989), and the LOD is 7.6 pM. A comparison of our test results with other assays in Table 2 reveals that the detection limit of this method is close to that of most of the other methods, and most prominently, it greatly shortens the time required for detection. It has been proved that this SERS assay platform can quantitatively detect two CRC-associated miRNAs at the same time, and its unique advantage makes it highly suitable for clinical diagnosis and offers promising potential for clinical applications.
Fig. 6. (A) The SERS spectra of different concentrations of miR-21 and miR-18a suspended in PBS buffer were examined through investigation. (B) A relationship was established between the peak intensity at 1337 cm-1 and (C) 1593 cm-1 and the logarithm of miR-21 and miR-18a concentrations, respectively, by creating calibration curves. (D) The SERS spectra were obtained from serum samples containing various concentrations of miR-21 and miR-18a. Calibration curves were constructed using serum samples to relate peak intensity at (E) 1337 cm-1 and (F) 1593 cm-1 with the logarithm of concentrations for miR-21 and miR-18a.
Table 2. Comparison of the SERS detection platform reported in this paper with methods documented in prior research
3.7 Characterization of colorectal cancer in homozygous mice
Photographs were taken to compare the growth of the subcutaneous tumors on day 1 and day 30 after loading. As shown in Fig. 7(A) and (B), it was obvious that the tumors were growing and shaped at day 30, and the tumor tissues were of uniform size. The weight of the mice was monitored synchronously before each blood collection after tumor loading. Compared with the normal control group, the difference in body weight of the transplanted tumor group was not statistically significant (Fig. 7(C)). The curves of the average volume of tumor tissues of the eight tumor-bearing mice over time are shown in Fig. 7(D), which are all suggestive that the modeling was successful. Serum could be collected for testing of the real samples.
Fig. 7. (A) Pictures of mice on day 1 after loading the tumor. (B) Picture of mice on day 30. (C) Average mouse weight curve. (D) Average tumor volume curve.
3.8 Simultaneous detection of two miRNAs in mouse serum
To further evaluate the practical application value of this method, we performed a SERS assay of miR-21 and miR-18a levels in the serum of mice at different stages (1, 6, 11, 16, 21, 26, and 30 d). The accuracy of this assay platform's determination was made by comparing it with qRT-PCR, using the average quantities of miR-21 and miR-18a in serum as the control. As shown in Fig. 8(A) and (C), the average SERS profiles of normal control mice and hormonal mice at different stages were collected, and the SERS intensity in hormonal mice decreased with the growth of tumor tissue. In Fig. 8(B) and (D), there was no detectable change in the intensity of the average SERS spectra at 1337 cm-1 and 1593 cm-1 for the control mice, whereas the transplanted tumor group exhibited a steady reduction in intensity at those frequencies over different stages, specifically with a significant change on day 6 of early tumor formation. After analyzing the detection curves, the calculated average containment level of miR-21 and miR-18a showed an increase. The obtained results were further in contrast to qRT-PCR results, as presented in Table 3. The findings of this technique in identifying the existence of miR-21 and miR-18a corresponded with qRT-PCR, suggesting a high level of accuracy in quantitatively detecting miR-21 and miR-18a. Additionally, this assay platform is well-suited for analyzing actual clinical samples.
Fig. 8. (A) Blank mouse sera from day 1, 6, 11, 16, 21, 26, and 30 were studied to analyze the SERS spectra of miR-21 and miR-18a. (B) Corresponding intensity histograms at 1337 cm-1 and 1593 cm-1 in blank mice. (C) SERS spectra of miR-21 and miR-18a in mice's tumor growth serum were analyzed. (D) Intensity histograms that correspond to each other at 1337 cm-1 and 1593 cm-1 in tumor-bearing mice.
Table 3. Average results of serum SERS and qRT-PCR in hormonal mice
4. Conclusion
To summarize, our team has created a platform effectively for SERS detection built upon a nucleic acid aptamer and a self-assembled highly ordered AuNDs array for the simultaneous testing of miR-21 and miR-18a levels in serum to detect colorectal cancer at an early stage. The tightly packed, ordered AuNDs array was also shown to have a large number of “hotspots” in their interstitial spaces, which could greatly amplify the SERS signal intensity. Multiple tests have shown that the method is characterized by high specificity, reproducibility, and stability, as well as rapidity and efficiency, and the process of detection can be finished in just 5 min, and the LOD can reach the level of pM. A comparison of the results of two nucleic acids in the serum of the CRC mouse model with qRT-PCR verified the accuracy and practical applicability of this method. Given these benefits, this approach holds promise for the effective and simultaneous identification of diverse biomarkers associated with CRC, making it highly valuable in real-world clinical settings.
Funding
Jiangsu Commission of Health (ZD2021038); YangZhou Municipal Science and Technology Bureau (YZ2023147).
Disclosures
The authors declare that there are no conflicts of interest related to this article.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. E. Dekker, P. J. Tanis, J. L. A. Vleugels, et al., “Colorectal cancer,” Lancet 394(10207), 1467–1480 (2019). [CrossRef]
2. S. Aguiar Junior, M. M. de OLIVEIRA, D. R. M. e Silva, et al., “Survival of patients with colorectal cancer in a cancer centeR,” Arq. Gastroenterol. 57(2), 172–177 (2020). [CrossRef]
3. K. Thanikachalam and G. Khan, “Colorectal Cancer and Nutrition,” Nutrients 11(1), 164 (2019). [CrossRef]
4. S. G. Patel, J. J. Karlitz, T. Yen, et al., “The rising tide of early-onset colorectal cancer: a comprehensive review of epidemiology, clinical features, biology, risk factors, prevention, and early detection,” Lancet Gastroenterol Hepatol 7(3), 262–274 (2022). [CrossRef]
5. A. N. Burnett-Hartman, J. K. Lee, J. Demb, et al., “An update on the epidemiology, molecular characterization, diagnosis, and screening strategies for early-onset colorectal cancer,” Gastroenterology 160(4), 1041–1049 (2021). [CrossRef]
6. N. Li, B. Lu, C. Luo, et al., “Incidence, mortality, survival, risk factor and screening of colorectal cancer: a comparison among China, Europe, and northern America,” Cancer Lett. 522, 255–268 (2021). [CrossRef]
7. N. Vatandoost, J. Ghanbari, M. Mojaver, et al., “Early detection of colorectal cancer: from conventional methods to novel biomarkers,” J Cancer Res Clin Oncol 142(2), 341–351 (2016). [CrossRef]
8. M. A. Scaffidi, S. C. Grover, H. Carnahan, et al., “Impact of experience on self-assessment accuracy of clinical colonoscopy competence,” Gastrointest Endosc 87(3), 827–836.e2 (2018). [CrossRef]
9. W. Wu, J. Huang, Y. Yang, et al., “Adherence to colonoscopy in cascade screening of colorectal cancer: a systematic review and meta-analysis,” J Gastroenterol Hepatol 37(4), 620–631 (2022). [CrossRef]
10. A. K. Tiwari, H. S. Laird-Fick, R. K. Wali, et al., “Surveillance for gastrointestinal malignancies,” World J Gastroenterol 18(33), 4507–4516 (2012). [CrossRef]
11. N. Zhang, X. Hu, Y. Du, et al., “The role of miRNAs in colorectal cancer progression and chemoradiotherapy,” Biomed. Pharmacother. 134, 111099 (2021). [CrossRef]
12. J. Xu, Q. Meng, X. Li, et al., “Long noncoding RNA MIR17HG promotes colorectal cancer progression via miR-17-5p,” Cancer Res 79(19), 4882–4895 (2019). [CrossRef]
13. A. Siskova, K. Cervena, J. Kral, et al., “Colorectal adenomas-genetics and searching for new molecular screening biomarkers,” Int. J. Mol. Sci. 21(9), 3260 (2020). [CrossRef]
14. L. Gmerek, K. Martyniak, K. Horbacka, et al., “MicroRNA regulation in colorectal cancer tissue and serum,” PLoS One 14(8), e0222013 (2019). [CrossRef]
15. B. Chen, Z. Xia, Y.-N. Deng, et al., “Emerging microRNA biomarkers for colorectal cancer diagnosis and prognosis,” Open Biol. 9(1), 180212 (2019). [CrossRef]
16. A. F. Ghareib, R. H. Mohamed, A. R. Abd el-Fatah, et al., “Assessment of serum microRNA-21 gene expression for diagnosis and prognosis of colorectal cancer,” J Gastrointest Cancer 51(3), 818–823 (2020). [CrossRef]
17. M. Elaguizy, M. Sheta, N. Ibrahim, et al., “Serum microRNA-18a, microRNA-21 and microRNA-92a as diagnostic markers in colorectal cancer patients,” J buon 25, 1443–1448 (2020).
18. K. Saliminejad, H. R. Khorram Khorshid, S. Soleymani Fard, et al., “An overview of microRNAs: biology, functions, therapeutics, and analysis methods,” J Cell Physiol 234(5), 5451–5465 (2019). [CrossRef]
19. A. Bawazeer and D. C. Prince, “Detection of miRNAs,” Methods Mol Biol 2630, 1–11 (2023). [CrossRef]
20. E. Ban and E. J. Song, “Considerations and suggestions for the reliable analysis of miRNA in plasma using qRT-PCR,” Genes 13(2), 328 (2022). [CrossRef]
21. Q. Zhu, T. Yan, Y. Yang, et al., “Programmable microparticle array for in situ modification and multiple miRNA detection,” ACS Sens. 7(12), 3654–3659 (2022). [CrossRef]
22. T. Jet, G. Gines, Y. Rondelez, et al., “Advances in multiplexed techniques for the detection and quantification of microRNAs,” Chem. Soc. Rev. 50(6), 4141–4161 (2021). [CrossRef]
23. Z. D. Schultz, D. N. Shoup, A. E. Smith, et al., “Super-resolution SERS spectral bioimaging,” Proc SPIE Int Soc Opt Eng 12203, 13 (2022). [CrossRef]
24. V. Moisoiu, S. D. Iancu, A. Stefancu, et al., “SERS liquid biopsy: an emerging tool for medical diagnosis,” Colloids Surf., B 208, 112064 (2021). [CrossRef]
25. K. Xu, R. Zhou, K. Takei, et al., “Toward flexible surface-enhanced Raman scattering (SERS) sensors for point-of-care diagnostics,” Adv. Sci. 6(16), 1900925 (2019). [CrossRef]
26. H. Liu, X. Gao, C. Xu, et al., “SERS tags for biomedical detection and bioimaging,” Theranostics 12(4), 1870–1903 (2022). [CrossRef]
27. A. Issatayeva, E. Farnesi, D. Cialla-May, et al., “SERS-based methods for the detection of genomic biomarkers of cancer,” Talanta 267, 125198 (2024). [CrossRef]
28. G. M. Das, S. Managò, M. Mangini, et al., “Biosensing Using SERS active gold nanostructures,” Nanomaterials 11(10), 2679 (2021). [CrossRef]
29. T. Shi, Z. Tang, P. Liang, et al., “ZrO2@Ag@SiO2 sandwich structure with high SERS enhancement effect and stability,” J. Phys. Chem. C 124(47), 25967–25974 (2020). [CrossRef]
30. C. Li and M. Chen, “Active site-dominated electromagnetic enhancement of surface-enhanced Raman spectroscopy (SERS) on a Cu triangle plate,” RSC Adv. 10(69), 42030–42037 (2020). [CrossRef]
31. S. Kaja and A. Nag, “Bimetallic Ag-Cu Alloy microflowers as SERS substrates with single-molecule detection limit,” Langmuir 37(44), 13027–13037 (2021). [CrossRef]
32. P. C. Wu, C. G. Khoury, T.-H. Kim, et al., “Demonstration of surface-enhanced Raman scattering by tunable, plasmonic gallium nanoparticles,” J. Am. Chem. Soc. 131(34), 12032–12033 (2009). [CrossRef]
33. M. L. Tseng, C. M. Chang, B. H. Cheng, et al., “Multi-level surface enhanced Raman scattering using AgOx thin film,” Opt. Express 21(21), 24460–24467 (2013). [CrossRef]
34. P. Wang, M. E. Nasir, A. V. Krasavin, et al., “Plasmonic metamaterials for nanochemistry and sensing,” Acc Chem Res 52(11), 3018–3028 (2019). [CrossRef]
35. S. Gómez-Graña, C. Fernández-López, L. Polavarapu, et al., “Gold nanooctahedra with tunable size and microfluidic-induced 3d assembly for highly uniform SERS-active supercrystals,” Chem. Mater. 27(24), 8310–8317 (2015). [CrossRef]
36. X. Li, T. Zhang, Z. Chen, et al., “Au polyhedron array with tunable crystal facets by PVP-assisted thermodynamic control and its sharp shape as well as high-energy exposed planes co-boosted SERS activity,” Small 18, e2105045 (2022). [CrossRef]
37. B.-B. Fu, X.-D. Tian, J.-J. Song, et al., “Self-calibration 3D hybrid SERS substrate and its application in quantitative analysis,” Anal. Chem. 94(27), 9578–9585 (2022). [CrossRef]
38. D. Dong, L. W. Yap, D. M. Smilgies, et al., “Two-dimensional gold trisoctahedron nanoparticle superlattice sheets: self-assembly, characterization and immunosensing applications,” Nanoscale 10(11), 5065–5071 (2018). [CrossRef]
39. B. Guan and X. Zhang, “Aptamers as versatile ligands for biomedical and pharmaceutical applications,” Int J Nanomedicine 15, 1059–1071 (2020). [CrossRef]
40. L. Bohrmann, T. Burghardt, C. Haynes, et al., “Aptamers used for molecular imaging and theranostics - recent developments,” Theranostics 12(9), 4010–4050 (2022). [CrossRef]
41. S. Shi and T. P. Russell, “Nanoparticle assembly at liquid-liquid interfaces: from the nanoscale to mesoscale,” Adv Mater 30, e1800714 (2018). [CrossRef]
42. L. Costa, G. Li-Destri, N. H. Thomson, et al., “Real space imaging of nanoparticle assembly at liquid-liquid interfaces with nanoscale resolution,” Nano Lett 16(9), 5463–5468 (2016). [CrossRef]
43. C. Y. Yu, B.-C. Yin, Bang-Ce Ye, et al., “A universal real-time PCR assay for rapid quantification of microRNAs via the enhancement of base-stacking hybridization,” Chem Commun 49(74), 8247–8249 (2013). [CrossRef]
44. P. J. Chung, Lian-Ming Lyu, M. H. Huang, et al., “Seed-mediated and iodide-assisted synthesis of gold nanocrystals with systematic shape evolution from rhombic dodecahedral to octahedral structures,” Chemistry 17(35), 9746–9752 (2011). [CrossRef]
45. X. Lu, D. Li, Z. Luo, et al., “A dual-functional fluorescent biosensor based on enzyme-involved catalytic hairpin assembly for the detection of APE1 and miRNA-21,” Analyst 147(12), 2834–2842 (2022). [CrossRef]
46. S. Lv, F. Chen, C. Chen, et al., “A novel CdTe quantum dots probe amplified resonance light scattering signals to detect microRNA-122,” Talanta 165, 659–663 (2017). [CrossRef]
47. J. Y. Lim and S. S. Lee, “Sensitive detection of microRNA using QCM biosensors: sandwich hybridization and signal amplification by TiO2 nanoparticles,” Anal Methods 12(42), 5103–5109 (2020). [CrossRef]
48. A. Miglione, A. Raucci, J. Amato, et al., “Printed electrochemical strip for the detection of miRNA-29a: a possible biomarker related to Alzheimer's disease,” Anal Chem 94(45), 15558–15563 (2022). [CrossRef]
49. K. Treerattrakoon, P. Roeksrungruang, T. Dharakul, et al., “Detection of a miRNA biomarker for cancer diagnosis using SERS tags and magnetic separation,” Anal Methods 14(20), 1938–1945 (2022). [CrossRef]
