SERS array for rapid quantitative detection of MMP-2 enzyme activity, its preparation method and application

Abstract

This invention belongs to the field of molecular imaging technology, specifically relating to a rapid quantitative detection SERS array for MMP-2 enzyme activity, its preparation method, and its application. First, this invention uses computer-aided design to screen for fast-responding MMP-2 substrate peptides, improving the speed and selectivity of the enzymatic digestion reaction. Second, it constructs a Prussian blue-doped gold nanostar array, utilizing the stable Raman peak of Prussian blue in the bioquiescent region as an internal control, while retaining the SERS hotspots generated by the tips and internanospaces of the gold nanostars. Finally, it immobilizes preferred MMP-2 substrate peptides modified with Nile blue and lipoic acid on the array surface, causing MMP-2 enzymatic digestion to induce changes in Nile blue-related SERS signals, and performs ratio correction using the Prussian blue internal control peak. This invention overcomes the problems of traditional detection methods, such as long detection time, high fluorescence background, large fluctuations in absolute SERS signals, and lack of reliable quantitative internal controls, providing a rapid, stable, and array-compatible novel SERS technology solution for MMP-2 enzyme activity detection.

Description

Technical Field

[0001] This invention belongs to the field of molecular imaging technology, specifically relating to a SERS array for rapid quantitative detection of MMP-2 enzyme activity, its preparation method, and its application. Background Technology

[0002] Matrix metalloproteinase-2 (MMP-2, also known as gelatinase A) is an important zinc-dependent endopeptidase in the matrix metalloproteinase family that is closely related to extracellular matrix remodeling, basement membrane degradation, tumor invasion, and angiogenesis. [1] MMP-2 can degrade extracellular matrix components such as type IV collagen, gelatin, laminin, and fibronectin. Type IV collagen is an important structural component of the basement membrane. Therefore, after MMP-2 activation, it can directly promote tumor cells to break through the basement membrane barrier, invade surrounding tissues, enter the blood or lymphatic system, and further participate in the formation of distant metastases. [2] Previous studies have indicated that MMP-2, as a gelatinase, participates in pathological processes requiring basement membrane disruption, and its activation level is associated with tumor spread and poor prognosis. In MMP-2-deficient mice, tumor angiogenesis and tumor progression are both inhibited, suggesting that it is not only a tumor marker but may also participate in the process of tumor microenvironment remodeling. [3] .

[0003] In the process of tumorigenesis and development, the importance of MMP-2 is not only reflected in its traditional function of "degrading the extracellular matrix", but also in its regulation of the tumor microenvironment, pro-angiogenic factors, inflammatory signals and cell migration behavior. [4] Studies have shown that both MMP-2 and MMP-9 can degrade type IV collagen and are elevated in various human tumors; both can promote tumor angiogenesis, invasion, and growth through extracellular matrix degradation and activation of pro-angiogenic signals such as TGF-β and VEGF. [5] Measuring MMP-2 protein expression alone is often insufficient to reflect disease progression, because it is the catalytically active enzymes, rather than total protein or zymogens, that are truly involved in matrix degradation and microenvironment remodeling. [6, 7] Therefore, quantitative detection of MMP-2 "enzyme activity" is closer to its pathological functional state than simply detecting MMP-2 expression level, and is particularly suitable for evaluating tumor invasiveness, the degree of activation of the tumor microenvironment, the effect of drug inhibition, and rapid lesion localization in surgical settings.

[0004] Therefore, rapid, sensitive, and quantitative detection of MMP-2 enzyme activity is clearly needed in multiple scenarios. First, in tumor diagnosis and prognostic assessment, MMP-2 activity can serve as a functional indicator of tumor invasion, angiogenesis, and metastatic potential. Second, in drug screening, MMP-2 activity detection can be used to evaluate the effects of MMP inhibitors, anti-metastatic drugs, anti-angiogenic drugs, or drugs that modulate the tumor microenvironment. Third, in cell experiments and organoid models, MMP-2 activity can dynamically reflect extracellular matrix degradation, tumor cell migration, and matrix remodeling capabilities. Fourth, in clinical translational scenarios such as surgical margin determination, rapid tissue detection, and tumor specimen partitioning analysis, obtaining quantitative results of MMP-2 activity in a short time will help compensate for the inability of conventional pathological and immunological detection to reflect enzyme functional status in real time. Existing studies on peptide microarray fluorescence detection have also indicated that MMP activity detection is valuable for monitoring tumor progression, and MMP activity analysis in blood samples can serve as a potential supplement to liquid biopsy. [8] .

[0005] Currently, traditional techniques for detecting MMP-2 enzyme activity mainly include gelatin zymography, ELISA or immunocapture activity detection, fluorescence / FRET peptide substrate method, peptide microarray method, electrochemical sensing method, and Raman / surface-enhanced Raman scattering detection method developed in recent years.

[0006] Gelatin zymography is a classic method for detecting the activity of MMP-2 and MMP-9. [9] The basic principle is to perform non-reducing SDS-PAGE electrophoresis on a polyacrylamide gel containing gelatin substrate to separate gelatinases of different molecular weights in the sample; then, by removing SDS, enzyme refolding, incubation to degrade the gelatin in the gel, and finally Coomassie brilliant blue staining, the locations of gelatin degradation appear as transparent bands against a dark background.

[10] This method can distinguish between MMP-2 / MMP-9 and their proenzyme and activated forms, and has high sensitivity, thus it is widely used in basic research.

[11] However, the gelatin zymography method has significant shortcomings: its operation process includes gel preparation, sample loading, electrophoresis, renaturation, enzyme reaction, staining, destaining, and grayscale analysis, which takes a long time, usually several hours to more than a day; the detection results depend on the grayscale of the bands, and strictly speaking, it is semi-quantitative; this method is quite sensitive to the consistency of operation by the experimenters, the degree of sample concentration, the renaturation efficiency, and the color development time; at the same time, this method is usually used for conditioned culture media or tissue homogenates, and is not suitable for rapid, on-site, micro-sample or high-throughput array detection.

[0007] ELISA and immunological detection methods mainly rely on antibodies to recognize MMP-2 protein and can be used to detect MMP-2 content in serum, tissue homogenates, or cell culture supernatants.

[12] Traditional ELISA procedures are relatively standardized and suitable for batch testing of multiple samples, but its main problem is that it detects "protein content" rather than directly detecting "catalytic activity." For MMP-2, a sample may simultaneously contain precursor enzymes, active enzymes, inhibitor-bound forms, and degradation fragments; the total protein level is not equivalent to the functional enzyme activity. To address this issue, immunocapture activity assays have been developed that combine antibody capture with a FRET peptide substrate reaction. This involves first capturing MMPs in the sample with a pre-coated antibody, then adding a substrate peptide with fluorescent and quenching groups. The active MMPs cleave the substrate, generating a fluorescent signal. These methods are closer to functional detection than traditional ELISA, but they still depend on antibody quality, substrate selection, incubation conditions, and enzyme reaction time, and may be affected by non-specific proteases, matrix effects, and background fluorescence interference in complex samples.

[0008] Fluorescence / FRET peptide substrate method is one of the more commonly used techniques for detecting MMP activity.

[13] A typical design involves attaching a fluorescent donor and a quencher group to both ends of the MMP-cleavable peptide sequence, respectively. In the intact substrate state, fluorescence is quenched. When MMP-2 cleaves the peptide substrate, the fluorescent and quencher groups spatially separate, fluorescence is restored, and the fluorescence intensity or fluorescence growth rate is related to enzyme activity.

[14] Commercially available FRET MMP activity kits are available for use in 96-well plates and can detect the activities of various MMPs, including MMP-1, MMP-2, MMP-3, MMP-7, MMP-8, MMP-9, MMP-12, MMP-13, and MMP-14. This type of method offers advantages such as ease of operation, suitability for high-throughput processing, and intuitive readings, but it also has limitations: First, fluorescence detection is easily affected by sample autofluorescence, photobleaching, scattering, and absorption; second, the fluorescence spectrum is broad, leading to significant channel crosstalk during multiplex detection; third, MMP family members have similar catalytic structures, and if the substrate peptide specificity is insufficient, cross-cleavage by multiple proteases such as MMP-2, MMP-9, and MMP-13 may occur; fourth, the fluorescence signal generally reflects the average signal in the overall reaction solution, making it difficult to achieve stable, spatially resolved, and interference-resistant quantitative detection on a solid-phase array.

[0009] To improve throughput and substrate screening capabilities, existing research has developed fluorescence detection methods for MMP-2 peptide microarrays. For example, Jian et al. reported a fluorescence method based on peptide microarrays, where biotinylated peptide substrates were immobilized on aldehyde-based three-dimensional slides. Fluorescence was generated by the binding of FITC-labeled NeutrAvidin to biotin. When MMP-2 cleaved the substrate, the biotin terminus was released, leading to a decrease in fluorescence signal. The fluorescence change was correlated with MMP-2 activity. This study screened MMP-2 substrates from 11 candidate peptide substrates, achieving a detection limit of 14 pg / mL in buffer, and used it to analyze cellular secretory MMP-2 activity and serum MMP-2 activity. [8] This type of peptide microarray technology demonstrates advantages in substrate screening and high-throughput detection, but its signal is still essentially fluorescence, and it still faces problems such as fluorescence background, photobleaching, signal drift, and interference from complex samples. In addition, its detection relies on a fluorescent label affinity recognition step, which is relatively indirect and cannot meet the requirements for faster, more stable, and more resistant to background interference in enzyme activity quantification.

[0010] Raman spectroscopy is a spectroscopic analysis technique based on molecular vibrational information. It boasts advantages such as "fingerprint" characteristics, narrow peak width, strong multiplexing capability, low water background, and the ability to provide molecular structural information in complex biological systems. However, ordinary Raman scattering has a small cross-section and limited sensitivity. Surface-enhanced Raman scattering (SERS) generates a strong electromagnetic field enhancement through localized surface plasmon resonance on the surface of noble metal nanostructures, significantly amplifying the Raman signal of reporter molecules adsorbed on or near the metal surface. Literature indicates that SERS can amplify the Raman signal of molecules adsorbed on rough metal surfaces or nanostructures, making it valuable for high-sensitivity detection in complex biological and environmental matrices. Furthermore, "hotspot" structures such as nano-intervals and nucleo-satellite structures can further enhance the electromagnetic field, which is key to the design of SERS nanoprobes.

[15] Compared to fluorescence methods, SERS has a narrower signal peak, making it suitable for multi-channel detection; compared to ELISA, SERS can achieve faster spectral readings through changes in reporter molecules or ratio signals; compared to gelatin zymography, SERS does not require electrophoresis and staining, and is theoretically more suitable for rapid quantitative and array-based detection.

[0011] Previous studies have used SERS for MMP detection. Gong et al. reported a SERS-based platform for multiplex detection of MMP-2 and MMP-7. This method utilizes the specific reaction of enzyme-cleaved peptide substrates to bind a bimetallic film-over-nanosphere substrate with gold nanoparticles, achieving low detection limits and high specificity for the detection of MMP-2 and MMP-7.

[16] This approach is similar to the patented technology in that it utilizes MMP cleavage of peptide substrates to induce SERS signal changes. However, it primarily relies on specific nanoparticle-substrate assembly and signal changes after enzymatic cleavage, resulting in a complex detection system. Furthermore, it does not construct a stable array with a fixed internal control for rapid quantitative MMP-2 cleavage. Another study constructed a magnetic-fluorescence-plasma dual-mode nanosensor, using Rhodamine B as both a FRET donor and a Raman reporter molecule. After MMP-2 cleaves the peptide substrate, gold nanoparticles are released, and magnetic separation leads to a reduction in SERS hotspots and restoration of fluorescence, thereby achieving MMP-2 activity detection.

[17] This type of system has high sensitivity, but it requires steps such as nanocomposite assembly, enzymatic release, and magnetic separation, which is a long process and not conducive to the preparation of solid-phase arrays that can be directly read.

[0012] In recent years, research has also developed ratiometric SERS for MMP-2 detection. For example, Chen et al. reported a nucleus-satellite structured ratiometric SERS nanosensor for ultrasensitive detection of MMP-2 activity. The core idea is to improve the accuracy of MMP-2 detection by correcting SERS signal fluctuations with an internal reference signal.

[18] This indicates that ratiometric SERS is an important direction for solving the quantitative problem of MMP-2. However, existing related technologies are mostly dispersed nanoprobes or solution-phase self-assembled systems, which often rely on nanoparticle aggregation / deaggregation, core-satellite structure changes, or subsequent separation steps to achieve detection. They have not yet fully solved the quantitative error problems caused by uneven hotspot distribution, laser focusing differences, substrate batch differences, and complex sample backgrounds in solid-phase arrays. Additionally, SERS nanobarcode platforms can perform multiplex detection of MMP-1, MMP-2, MMP-3, MMP-7, and MMP-9, and verify the correlation with ELISA results. However, the reaction time between the enzyme and the nanotag in this type of detection can reach several hours. For example, in the literature, MMP enzyme solution and nanotag were incubated at 37°C for 3 hours, indicating that it is more suitable for high-sensitivity multiplex detection and not entirely suitable for the application requirement of "rapid quantification of enzyme activity".

[19] .

[0013] One of the key challenges of using SERS for quantitative detection is signal repeatability.

[20] Because SERS enhancement is highly dependent on the hotspot distribution of the metal nanostructure, the distance and orientation of the reporter molecule to the metal surface, the laser focusing position, substrate homogeneity, and sample matrix, the intensity of a single reporter peak is prone to significant fluctuations. Existing studies have clearly indicated that SERS quantitative analysis remains challenging due to the randomness and poor reproducibility of hotspot distribution; embedded internal standard molecules can correct for SERS signal fluctuations caused by different aggregation states and measurement conditions, thereby improving quantitative reliability.

[21] Therefore, in MMP-2 SERS detection, if only the absolute peak intensity of the reporter molecule after enzyme digestion is detected, it is easily affected by the number of local hot spots in the array, the difference in nanostructure morphology, the laser irradiation position and the sample drying state, making it difficult to stabilize and quantify the detection results.

[0014] Prussian blue (PB) provides a valuable material basis for SERS internal controls. Previous studies have reported that Prussian blue exhibits a strong and sharp single Raman peak in the cellular Raman silencing region, with a typical peak position of approximately 2156 cm⁻¹. This region is generally unaffected by endogenous Raman signals from biological samples, thus it can serve as a Raman reporter molecule with low background and high specificity.

[22] Further research shows that structures such as Au@PB can utilize the strong Raman peak of PB at approximately 2155 cm⁻¹ as an internal reference to correct for SERS signal fluctuations in target analytes, thereby enabling quantitative SERS analysis in complex systems.

[23] However, existing PB internal control SERS technology is mainly used for small molecule, immunoassay, or general quantitative detection, and has not established an integrated technical system of "stable internal control + enzyme-digestible reporter molecule + high hot spot gold nanostar array" for rapid detection of MMP-2 enzyme activity.

[0015] In summary, while existing MMP-2 enzyme activity detection technologies include various approaches such as gelatin zymography, ELISA / immunocapture activity detection, FRET fluorescent substrates, peptide microarrays, electrochemical sensing, and SERS nanoprobes, they still have the following shortcomings: First, gelatin zymography is time-consuming, complex, and has significant semi-quantitative properties, making it difficult to meet the needs of rapid detection; second, ELISA mainly reflects protein content and cannot fully represent the functional state of active enzymes; third, while FRET and fluorescent microarrays are relatively convenient, they are limited by autofluorescence, photobleaching, spectral crosstalk, and substrate specificity; fourth, existing MMP-2 SERS detections mostly rely on changes in the nanostructure of the solution phase, magnetic separation, or long-term incubation, resulting in complex systems and limited speed and arraying capabilities; fifth, SERS quantification itself is greatly affected by hotspot heterogeneity, and existing MMP-2 SERS arrays generally lack stable, fixed, low-background, and co-located reliable internal controls with the detection region. Based on the aforementioned shortcomings of existing technologies, it is necessary to develop a SERS array that can rapidly respond to MMP-2 enzyme digestion, while possessing stable internal reference calibration capabilities and high hotspot enhancement capabilities, for rapid, sensitive, reproducible, and quantitative detection of MMP-2 enzyme activity. Summary of the Invention

[0016] The purpose of this invention is to provide a rapid and stable SERS array for quantitative detection of MMP-2 enzyme activity, as well as its preparation method and application, to overcome the problems of traditional detection methods such as long detection time, high fluorescence background, large fluctuations in absolute SERS signal, and lack of reliable quantitative internal reference.

[0017] This invention first employs computer-aided design to screen for fast-responding MMP-2 substrate peptides, improving the speed and selectivity of the enzymatic digestion reaction. Second, it constructs a Prussian blue-doped gold nanostar array, utilizing the stable Raman peak of Prussian blue in the bioquiescent region as an internal control, while retaining SERS hotspots generated at the tips and internanospaces of the gold nanostars. Finally, it immobilizes preferred MMP-2 substrate peptides modified with Nile blue and lipoic acid onto the array surface, causing MMP-2 digestion to induce Nile blue-related SERS signal changes, which are then corrected using the Prussian blue internal control peak. This invention overcomes the problems of traditional detection methods, such as long processing time, high fluorescence background, large fluctuations in absolute SERS signals, and lack of reliable quantitative internal controls, providing a rapid, stable, and array- and quantifiable novel SERS technology solution for MMP-2 enzyme activity detection.

[0018] Specifically, the method for preparing a SERS array for the quantitative detection of MMP-2 enzyme activity provided by this invention integrates a screened fast-response MMP-2 substrate peptide, a Prussian blue biological silent region internal control, a gold nanostar hotspot array, a Nile blue Raman reporter molecule, and a lipoic acid immobilization strategy to construct a solid-phase SERS array that can be used for the rapid quantitative detection of MMP-2 enzyme activity. The specific steps are as follows:

[0019] (1) Screening and optimization of fast-response MMP-2 substrate peptides: Based on the MMP-2 protease substrate cleavage database, the enzymatic reaction rate constants of candidate substrate peptides were obtained; on this basis, combined with the AI / Rosetta structure-guided substrate extension strategy, bilateral sequence extension, three-dimensional complex modeling, main chain sequence optimization, molecular dynamics simulation and binding free energy analysis were performed on the core short peptides with MMP-2 selectivity. Furthermore, through high-performance liquid chromatography kinetic detection and solid-phase surface reaction efficiency detection, the preferred substrate peptide with a fast enzymatic cleavage response to MMP-2 was obtained: DITPAAMTSPP; In traditional MMP-2 enzyme activity detection methods, the commonly used substrates are mostly general MMP substrates reported in the literature or commercial FRET substrates. The cleavage rate, specificity and accessibility of such substrates after solid-phase fixation may not be suitable for rapid detection.

[0020] (2) Construction of Prussian blue-doped gold nanostar array: Specifically, Au@PB core-shell nanoparticles were prepared by using gold nanospheres as the core, followed by cyano pretreatment and Prussian blue epitaxial growth; then, using Au@PB core-shell nanoparticles as seeds, anisotropic growth of the outer gold shell was induced under acidic conditions by chloroauric acid, silver nitrate and ascorbic acid, so that Prussian blue co-migrated with the growth of gold spikes and doped into the interior of the gold nanostar spikes, thus obtaining Prussian blue-doped gold nanostar nanoparticles; then, the Prussian blue-doped gold nanostar nanoparticles were fixed on the surface of a silicon wafer solid substrate after aminosilanization treatment to form a Prussian blue-doped gold nanostar SERS array; finally, by utilizing the combined effects of local electric field enhancement induced by dielectric environment, cascade focusing of multi-level tip structures, plasmon coupling and chemical enhancement, a strong local electromagnetic field enhancement effect was formed inside the gold spikes, so that the array itself has a stable strong Raman signal located in the biological silent region; at the same time, the highly active sites on the surface were retained for the assembly of enzyme response elements.

[0021] Traditional gold nanostars are mainly used as reinforcing structures, while the gold nanostar array in this invention not only provides strong electromagnetic hotspots to enhance surface reporter molecular signals, but also provides stable Raman peaks to correct signal fluctuations caused by different detection points, different array batches, and different laser focusing conditions.

[0022] (3) Constructing a rapid quantitative detection SERS array for MMP-2 enzyme activity: Specifically, Nile blue, the MMP-2 substrate peptide obtained in step (1) and lipoic acid are connected in a 1:1:1 molecular relationship to construct a lipoic acid-substrate peptide-Nile blue probe; wherein lipoic acid is used as a gold surface immobilization group, the substrate peptide is used as an MMP-2 enzyme cleavage response element, and Nile blue is used as a Raman reporter molecule; then the Prussian blue doped gold nanostar SERS array obtained in step (2) is immersed in 10–100 μM (preferably 20–50 μM) of the lipoic acid-substrate peptide-Nile blue probe solution, so that the probe is immobilized on the surface of the gold nanostar through Au-S coordination, and a SERS array for quantitative detection of MMP-2 enzyme activity is obtained.

[0023] Traditional SERS methods for detecting MMP-2 employ solution-phase nanoprobes, such as core-satellite structures, magnetic nanoparticle-gold nanoparticle complexes, and enzyme-induced aggregation / deaggregation systems. These methods offer high sensitivity, but often require steps such as nanoparticle pre-assembly, enzyme digestion, magnetic separation, centrifugation, or redispersion, making the detection process relatively complex and difficult to standardize into a standardized array detection platform.

[0024] Furthermore:

[0025] In step (1), the screening and optimization of MMP-2 substrate peptides specifically includes the following process:

[0026] (1.1) The second-order reaction rate constants (kobs) of candidate peptides with respect to MMP-2 and MMP-9 were obtained from the protease substrate kinetics database, and the standardized cleavage score (Z-score) of the candidate peptides was obtained in combination with proteomics cleavage data; multiple independent measurements of the same sequence were normalized, and the median was taken as the representative value of the sequence, where the reaction rate difference was:

[0027] Δkobs = kobs(MMP-2) - kobs(MMP-9);

[0028] The differences in cutting scores are as follows:

[0029] ΔZ = Z(MMP-2) - Z(MMP-9);

[0030] Candidate sequences were sorted according to Δkobs and ΔZ to screen for candidate peptides that simultaneously exhibited high MMP-2 response rates and low MMP-9 responses.

[0031] (1.2) Based on the core short peptide with high selectivity for MMP-2, a complex model of the core short peptide and MMP-2 was constructed using AlphaFold2-Multimer, and the RFjoint algorithm was used to complete the sequence on both sides of the core short peptide to generate the extended substrate peptide backbone.

[0032] (1.3) The Rosetta software was used to optimize the sequence of the extended substrate peptide under the condition of fixed backbone, and the candidate amino acids of the extension site were restricted to the range of high-frequency amino acids obtained by MMP-2 substrate preference analysis.

[0033] (1.4) The optimized candidate substrate peptides were screened by MMP-2 enzyme digestion kinetics, solid-phase SERS interface response, molecular dynamics simulation, MM / PBSA binding free energy calculation and single residue energy decomposition analysis, and finally the preferred substrate peptide DITPAAMTSPP was obtained.

[0034] In step (2), the specific process for constructing the Prussian blue-doped gold nanostar array is as follows:

[0035] (2.1) Gold nanospheres were prepared by sodium citrate reduction method; HAuCl4 was added to boiling ultrapure water to make the final concentration of HAuCl4 0.1–1.0 mM, and then sodium tricitrate solution was added to carry out reduction reaction to obtain gold nanospheres; preferably, the final concentration of HAuCl4 was 0.25 mM, the concentration of sodium tricitrate was 38.8 mM, and the reaction time was 15 minutes;

[0036] (2.2) Using the gold nanospheres as seeds, gold nanospheres with an average particle size of 20–60 nm are gradually grown to obtain gold nanospheres; preferably, the particle size of the gold nanospheres is 40±3 nm.

[0037] (2.3) Take gold nanosphere colloids and add K3[Fe(CN)6] solution for cyano pretreatment to form an adsorbed CN- layer on the gold surface; preferably, add 1.0 mL of 5 mM K3[Fe(CN)6] solution to 10 mL of gold nanosphere colloids and stir at room temperature for 5–10 minutes.

[0038] (2.4) FeCl3 solution and K4[Fe(CN)6] solution are simultaneously added dropwise to the cyano-pretreated gold nanosphere colloid to allow Prussian blue to be epitaxially grown on the surface of the gold nanospheres to form Au@PB core-shell nanoparticles; preferably, the concentrations of FeCl3 and K4[Fe(CN)6] are both 0.5 mM, and they are added dropwise simultaneously in an equimolar ratio at a total flow rate of 30 μL / min. The stirring speed during the dropwise addition is 1000–2000 rpm, and stirring is continued for 1–4 hours after the dropwise addition is completed;

[0039] (2.5) Au@PB core-shell nanoparticles are dispersed in an acidic aqueous solution with pH 0.5-2.0, and HAuCl4, AgNO3 and ascorbic acid are added to induce anisotropic gold shell growth on the Au@PB surface, forming Prussian blue doped gold nanostars with spike structures; preferably, 4 mL of Au@PB colloid is added to 200 mL of an aqueous solution with pH 1.0, the aqueous solution containing about 0.25 mM HAuCl4, followed by the addition of 2.0 mL of 3 mM AgNO3 and 1.0 mL of 100 mM ascorbic acid, and the reaction is carried out at 1500 rpm for about 40 seconds;

[0040] (2.6) The obtained Prussian blue doped gold nanostars were purified by centrifugation and washed with water to obtain Prussian blue doped gold nanostar nanoparticles; preferably, the centrifugation conditions were 10,000 rpm for 15 minutes and the washing was performed 3 times.

[0041] (2.7) After the solid substrate is treated with aminosilanization, it is immersed in Prussian blue doped gold nanostar colloid for shaking incubation, so that the Prussian blue doped gold nanostar is fixed on the surface of the solid substrate to form a SERS array; preferably, the solid substrate is a silicon wafer.

[0042] In step (3), the specific process for building the SERS array is as follows:

[0043] (3.1) Using a silicon wafer as a solid substrate, the following processes were performed sequentially: oxidation cleaning, ethanol and water cleaning, aminosilanization treatment, acid activation, and drying treatment; wherein:

[0044] The aminosilanization treatment involves immersing the cleaned silicon wafer in a 2% anhydrous ethanol solution of 3-aminopropyltrimethoxysilane for 6–24 hours, preferably 12 hours.

[0045] The acid activation involves immersing the aminosilanized silicon wafer in a 0.1–0.5 M hydrochloric acid solution for 5–30 minutes, preferably in a 0.24 M hydrochloric acid solution for 10 minutes.

[0046] (3.2) Immerse the treated silicon wafer in Prussian blue doped gold nanostar colloid and incubate with shaking at 50–150 times / minute for 12–48 hours, preferably 90 times / minute for 36 hours, so that the Prussian blue doped gold nanostars are fixed on the silicon wafer surface through electrostatic adsorption and surface interaction, and a Prussian blue doped gold nanostar SERS array is obtained.

[0047] (3.3) Prepare a 10–100 μM probe solution of lipoic acid-substrate peptide-Nile blue probe, preferably 30 μM; wherein, lipoic acid, substrate peptide and Nile blue are covalently linked to form a single probe molecule, and the molecular linkage ratio of the three is 1:1:1.

[0048] (3.4) Immerse the SERS array obtained in step (3.2) in the probe solution and incubate gently with shaking at 0–10°C in the dark for 1–4 hours; preferably incubate at 4°C in the dark for 2 hours.

[0049] (3.5) After incubation, wash away unbound probes with ultrapure water and buffer, and dry with nitrogen to obtain a SERS array for quantitative detection of MMP-2 enzyme activity;

[0050] The probe is fixed by forming Au-S coordination bonds with gold atoms on the surface of gold nanostars through the disulfide ring structure in lipoic acid. The amount of probe solution used is such that the array is completely submerged. Preferably, 0.1–1.0 mL of probe solution is used per 1 cm² array.

[0051] The SERS array prepared by the method of this invention can be used for quantitative detection of MMP-2 enzyme activity.

[0052] The specific steps are as follows:

[0053] (1) The sample to be tested is dropped onto the surface of the SERS array, so that the MMP-2 in the sample to be tested reacts with the substrate peptide immobilized on the array surface;

[0054] (2) Incubate at room temperature for 1–16 minutes, preferably 8 minutes;

[0055] (3) The Raman spectra of the SERS array were acquired using a 785 nm excited Raman spectrometer;

[0056] (4) Select Nile blue at approximately 590 cm -1 The Raman peak at approximately 2100 cm⁻¹ was used as the MMP-2 response signal, with Prussian blue at approximately 2100 cm⁻¹ selected as the peak. -1 The Raman peak in the biological quiescent region was used as an internal reference signal;

[0057] (5) Calculate the Raman intensity ratio I 590 / I 2100 The activity of MMP-2 enzyme in the test samples was quantitatively analyzed according to the pre-established MMP-2 activity standard curve.

[0058] As MMP-2 enzyme activity increases, MMP-2 cleaves substrate peptides on the array surface, causing Nile blue to move away from the hotspot region of the gold nanostar, resulting in a 590 cm⁻¹... -1 The Nile blue Raman signal decreases at 2100 cm⁻¹; while Prussian blue doping inside gold nanostars results in a lower signal at 2100 cm⁻¹. -1 The internal reference peak at point I remains stable, therefore I 590 / I 2100 It is negatively correlated with MMP-2 enzyme activity.

[0059] The MMP-2 enzyme activity detection range is 0–200 ng / mL, within which range I 590 / I 2100 It showed a linear negative correlation with MMP-2 enzyme activity, with a linear correlation coefficient R. 2 Not less than 0.98. The SERS array exhibits a selective response to MMP-2; and responds to MMP-9, bovine serum albumin, glutamate, arginine, cysteine, glutathione, ascorbic acid, and NO2. - Mg 2+ Ca 2+ ,ClO - O2 - In the presence of OH or H2O2, the I 590 / I 2100 No significant reduction occurred, while in the presence of MMP-2, the I 590 / I 2100 Significantly reduced.

[0060] The sample to be tested can be an in vitro enzyme solution, cell culture medium, tissue extract, brain tissue surface microdroplet extract, or intraoperative tissue microenvironment sample. When the sample to be tested is a brain tissue surface microdroplet extract, PBS buffered microdroplets are briefly brought into contact with the brain tissue surface to extract local free enzymes and metabolites. The microdroplets are then transferred to the surface of a SERS array for Raman detection, and the results are obtained through multiple spatial sampling points. 590 / I 2100 Plot the spatial distribution of MMP-2 enzyme activity using ratios.

[0061] This invention has the following technical features and advantages:

[0062] First, this invention does not directly use traditional general MMP substrates or commercial FRET substrates, but instead obtains the MMP-2 substrate peptide DITPAAMTSPP, which is more suitable for rapid detection in solid-phase SERS arrays, through data-driven screening, AI / Rosetta structure-guided extension design, solid-phase interface response screening, and molecular dynamics mechanism analysis.

[0063] Secondly, this invention incorporates Prussian blue into the spikes of gold nanostars, rather than simply adsorbing or coating it onto the surface of gold nanomaterials, thus enhancing the Prussian blue's 2100 cm⁻¹ density. -1 The Raman signal in the biosilent region is significantly enhanced in the hot spot region inside the gold nanostar, without occupying the probe assembly sites on the surface of the gold nanostar.

[0064] Third, this invention employs a ratio detection mode between the Nile blue signal and the Prussian blue internal reference signal, with I... 590 / I 2100 As a quantitative indicator, it can reduce the impact of substrate heterogeneity, sample matrix differences, laser power fluctuations, and local tissue heterogeneity on the detection results.

[0065] Fourth, the SERS array constructed in this invention can complete the detection of MMP-2 enzyme activity in a short time, preferably 8 minutes, which is suitable for rapid quantitative analysis and imaging of local enzyme activity distribution in tissues.

[0066] Fifth, the detection system of the present invention has high selectivity for MMP-2, and can effectively distinguish MMP-2 from MMP-9 and a variety of biologically active interfering substances, making it suitable for the detection of MMP-2 enzyme activity in complex biological samples. Attached Figure Description

[0067] Figure 1 A schematic diagram of the workflow for peptide design and screening.

[0068] Figure 2 The percentage of peak area of ​​the normalized intact peptides after incubation with MMP-2 enzyme for different times, including the control peptide (Ref.), high-scoring candidate peptides (Pep.1–Pep.3) screened from literature databases, and candidate peptides designed by SSE-AI (Pep.4).

[0069] Figure 3 The dose-response curves of normalized Raman characteristic peak intensity as a function of concentration after modifying the SERS substrate with candidate peptides (Pep.1, Pep.4) and control peptides (Ref.) and reacting with different concentrations (0~2000 ng / mL) of MMP-2 for 10 min are shown.

[0070] Figure 4 MM / PBSA binding free energy and residue degradation of the control peptide and Pep.4 bound to MMP-2. (a) Total binding free energy of the control peptide and Pep.4-MMP-2 complex calculated using MM / PBSA. (b) Binding free energy contributions of key catalytic residues, S1 / S2 / S3 pocket residues, and residues at the P / P' positions in Pep.4 (green) and the control peptide (gray). (c) Differences in contribution for each residue. ).

[0071] Figure 5 The docking model of the Pep.4-MMP-2 complex highlights the key interactions at the active sites.

[0072] Figure 6 The conformational kinetics of Pep.4 and the control peptide in the MMP-2 active site pocket are shown. Among them, (a) is the RMSD of Pep.4 and the control peptide. (b) is the main chain RMSF comparison by the relative P-position aligned residues P3-P1'.

[0073] Figure 7 This is a schematic diagram of the fabrication process for the APA array.

[0074] Figure 8 The images show the morphology and elemental distribution of Au@PB (top row) and APA (bottom row) nanoparticles. The left image is a transmission electron microscope (TEM) image, the middle image is a scanning transmission electron microscope (STEM) image, and the right image is the corresponding elemental surface scan (Au, Fe, N).

[0075] Figure 9 This is a scanning electron microscope image of an APA array.

[0076] Figure 10 The results of two-dimensional FDTD local electromagnetic field simulations of gold nanoparticles with different structures are shown.

[0077] Figure 11 Raman spectra of gold nanoparticles with different structures.

[0078] Figure 12 The stability of the APA array's Raman signal in PBS is shown. (a) is a thermogram of the Raman signal after 30 min of continuous 785 nm laser irradiation. (b) shows the Raman signal changes after adding buffer solutions of different pH values ​​(pH 2–10) to the APA array. (c) shows the stability curves of the APA array after 24 hours of storage in pure water, PBS, 10% FBS, and physiological saline. (d) shows the Raman signal stability of the APA array during long-term storage at room temperature (n = 5).

[0079] Figure 13 The construction process for MMP-APA arrays.

[0080] Figure 14 The quantitative Raman response of the MMP-APA array to the target enzyme is shown. (a) represents the response with increasing MMP-2 activity (0–200 ng / mL), and (b) represents the Raman intensity ratio (IL). 590 / I 2100 The linear fitting relationship between MMP-2 activity and the activity of MMP-2.

[0081] Figure 15 This represents the response of the MMP-APA array to various interfering factors.

[0082] Figure 16 Raman imaging of MMP-2 activity in exposed areas of mouse brain tissue. Detailed Implementation

[0083] The invention will be further described below with specific examples and accompanying drawings.

[0084] (I) Design of fast-response MMP-2 peptides based on a data-driven AI / Rosetta structure-guided substrate elongation strategy. The specific process is as follows (see...). Figure 1 ):

[0085] (1.1) Computer-aided peptide design and screening

[0086] By employing computational biology and computer-aided peptide design, MMP-2 substrate peptides with high response rates and high specificity were obtained. To ensure comprehensive and reliable screening, this invention employs two strategies: one is computational screening based on existing protease kinetic databases; the other is structure-guided substrate extension design screening based on AI and the Rosetta platform.

[0087] (1) Computational screening based on existing protease kinetics database

[0088] Based on existing peptide databases, the absolute reaction rates of peptides to MMP-2 and their selectivity relative to matrix metalloproteinase-9 (MMP-9) were comprehensively evaluated. First, approximately 1360 second-order reaction rate constants (k-values) of decapeptides to MMP-2 and MMP-9 were obtained from the research database established by Ratnikov et al. obs Combined with the large-scale proteomics analysis results of Kukreja et al., a standardized score (Z-score) for protease cleavage was obtained.

[24] Subsequently, the above data were standardized. When multiple independent measurements were obtained for the same sequence, the median was taken as the representative value for that sequence after unit normalization. In each independent dataset, the values ​​were standardized according to the protease type, and the response difference of each polypeptide sequence to MMP-2 and MMP-9 was calculated separately. The difference formula is defined as follows:

[0089]

[0090]

[0091] Filtering method: All sequences are sorted by Δk obs Sort in descending order, and for those with high k obs(MMP9) Weights are applied to the sequences; simultaneously, the sequences are sorted in descending order by ΔZ. The final candidate sequences must rank in the highest interval in both of these independent sorts. If there are sequences with the same joint score, the sequence with the higher k is preferred. obs(MMP2) sequence.

[0092] (2) Structure-guided substrate extension design screening based on AI and Rosetta platform

[0093] Other highly active and selective substrates for MMP-2 are mostly short peptides. In order to reduce steric hindrance and improve solid-phase cleavage kinetics while maintaining high activity and high Raman enhancement distance, this invention proposes a data-driven AI / Rosetta structure-guided substrate extension strategy (SSE-AI) to perform structure-guided extension design of known highly active core short peptides.

[0094] First, in the data-driven screening process, based on high catalytic efficiency, the substrate-catalytic efficiency ratio (K) is used. cat / K m The high selectivity for MMP-2 / MMP-9 led to the screening of three core hexapeptides: TPAAMT, STRPAEF, and IPLASL. Based on these, their blocs were extended to decapeptides, as shown in Table 1.

[0095] Table 1. Decapeptide extension sites and key tetrapeptide regions of the high MMP-2 selective core hexapeptide.

[0096] .

[0097] A three-dimensional spatial model of the complex of the three core hexapeptides and the MMP-2 protein was constructed using the deep learning model AlphaFold2-Multimer. Subsequently, the RFjoint algorithm was used for bidirectional sequence completion on both sides of the core sequence, generating a decapeptide backbone seeded by alanine. To ensure the extended sequences have reasonable chemical properties and spatial conformation, the generated backbone was optimized under fixed backbone conditions using Rosetta software. Furthermore, by introducing a candidate amino acid restriction set, the selectable amino acids at the extension sites (P5 to P5') were strictly limited to the range of high-frequency amino acids enriched in the MMP-2 substrate preference analysis, as shown in Table 2.

[0098] Table 2 Candidate amino acid restriction sets for each site in the Rosetta extended design

[0099] .

[0100] (1.2) Experimental screening of fast-response MMP-2 peptides

[0101] The four candidate peptides were synthesized using a solid-phase peptide synthesis method, and the widely used MMP-2 response sequence was introduced as a positive control (Ref.). To meet the requirements of subsequent SERS array assembly and Raman reporter molecule modification, specific amino acid residues were added to both ends of the peptides to increase their water solubility and simulate the modified peptides, as shown in Table 3.

[0102] Table 3 shows the five peptide sequences used for activity screening.

[0103] .

[0104] Five peptides were incubated with activated MMP-2 enzyme, and their cleavage kinetics were monitored by HPLC. The normalized percentage of intact peptide peak areas (normalized to the intact peptide peak area at 0 min) after incubation with MMP-2 enzyme for different times was calculated for the control peptide (Ref.), high-resolution candidate peptides screened from literature databases (Pep.1–Pep.3), and candidate peptide designed by SSE-AI (Pep.4). See [link to HPLC data]. Figure 2 The results showed that the peak area of ​​the control peptide (Ref.) decreased only slowly after 30 and 60 minutes of reaction, while the peak areas of candidate peptides Pep.1 and Pep.4 decreased sharply. At 60 minutes, the remaining proportions of Pep.1 and Pep.4 were significantly lower than those of Ref. and other candidate peptides, indicating that they have a faster reaction rate in the free liquid phase. Therefore, Pep.1 and Pep.4 were selected for the next round of solid-phase interface screening.

[0105] Since the liquid phase environment cannot accurately reflect the steric hindrance effect of peptides on the surface of a solid-phase array, this invention modifies the surface of a SERS array with Pep.1, Pep.4, and Ref., respectively, and detects their Raman signals after treatment with different active MMP-2s (0–2000 ng / mL) for 10 minutes. (See [link to relevant documentation]). Figure 3 The results showed a significant differentiation in the cleavage capabilities of the three peptides at the solid-phase interface. The Raman signal of the control peptide hardly decreased with increasing enzyme concentration, indicating its long reaction kinetics and steric hindrance, preventing it from effectively entering the enzyme's catalytic pocket. Pep.1 showed only a weak signal attenuation; at high enzyme concentrations, the Raman signal decreased by only about 25%, demonstrating that its sequence structure is ill-suited to the SERS solid-phase interface. In contrast, Pep.4, designed using the SSE-AI strategy, exhibited excellent solid-phase responsiveness: its Raman signal decreased rapidly and significantly with increasing enzyme concentration; even at lower concentrations of MMP-2, its signal intensity decreased markedly. These results demonstrate that Pep.4 has the highest response efficiency on the SERS array and is confirmed as a core element for subsequent detection.

[0106] (1.3) Calculation and analysis of preferred rapid peptide cleavage

[0107] This invention studies enzyme-substrate complexes based on molecular dynamics (MD) simulations and combined with free energy calculations, and analyzes the underlying mechanism by which the designed peptides have higher catalytic conversion rates compared to control peptides.

[0108] (1) Construction of the initial complex structure: First, the initial complex structures of the control peptide and the preferred peptide with MMP-2 were predicted using the Proteinix server. To accurately locate the catalytic water molecule, the predicted complex structure was spatially superimposed with the known MMP-2 crystal structure (PDB ID: 3AYU), and the catalytic water molecule in 3AYU was directly transferred into the theoretical model. Each constructed complex model retained a catalytic Zn 2+ A structure Zn 2+ And two Ca 2+

[25] The catalytic zinc ion is coordinated by a highly conserved His-His-His motif and a catalytic water molecule near the peptide bond to be cleaved, with nearby glutamate residues acting as a generalized base. This conformation is completely consistent with the reported MMP-2 catalytic mechanism. The modified complex structure was used as the initial conformation for subsequent MD simulations.

[0109] (2) Parameterization of the metal center: The MCPB.py file in the AmberTools package was used to parameterize the Zn catalyst. 2+ Center (Zn) 2 +-3His-H2O / substrate carbonyl) and the structure of Zn 2+ and Ca 2+ Parameterization was performed. The model contains nitrogen atoms in an imidazole ring with three histidine residues, and oxygen atoms that catalyze water or substrate carbonyl groups. The resulting bonding and nonbonding parameters were then combined with the standard force field parameters used for proteins and peptides.

[0110] (3) Molecular dynamics simulation process: Using Amber 22 software, the complex was first solvated in an explicit water box, and counterions were added to neutralize the charge. After energy minimization, the system was gradually heated to achieve density equilibrium. Subsequently, a 50 ns simulation was run under periodic boundary conditions, and the structural coordinates were saved periodically. Finally, the simulated trajectory was preprocessed using the cpptraj module: water molecules and ions were removed, the structure was centered, and a mirror image was generated to finally produce a simplified trajectory containing the protein-peptide-metal complex.

[0111] (4) Combining free energy calculation and residue energy decomposition: Using the MMPBSA.py module, the binding free energy (ΔG) of the substrate in the final state is calculated within the MM / PBSA framework. bind The topological files of the complex, receptor, and ligand were generated using ante-MMPBSA.py, and conformations were extracted from the equilibrium phase of each MD trajectory for calculation. Single-residue energy decomposition was further carried out to quantitatively assess the contributions of key catalytic residues (His115, His119, His125, Glu116 / 117), the S1 / S2 / S3 subunit pockets of the enzyme, and the P / P′ sites of the substrate to the total binding energy.

[0112] (5) Conformational kinetics: Using the average structure of the simulated product as a reference, the root mean square deviation (RMSD) and root mean square fluctuation (RMSF) of the heavy atoms in the main chain of the binding peptide (P2-P2′ region) were calculated. Finally, K-means clustering analysis was performed on the substrate main chain and surrounding active site residues using cpptraj. By evaluating the evolution of cluster occupancy over time, the flexibility and conformational heterogeneity of the substrate were quantified.

[0113] The MM / PBSA calculation results show that ( Figure 4 The control peptide binds much more tightly to MMP-2 than the preferred peptide. The binding free energy of the control peptide complex is approximately -45 kcal·mol⁻¹. -1 Pep.4 is approximately -12 kcal·mol⁻¹ -1 The difference is 33 kcal·mol -1This indicates that the control peptide exhibits an overly stable enzyme-substrate ground state, while Pep.4 displays a relatively loose binding complex.

[0114] Results of single-residue energy decomposition ( Figure 4 and Figure 5 This indicates that coordination catalysis of Zn 2+ The energy contributions of the three histidines are all within ±1 kcal·mol⁻¹ -1 The difference in total energy is not the primary cause of the overall energy variation. In contrast, the energy contribution of catalytic glutamate (Glu117) in the control system was as high as -56 kcal·mol⁻¹. -1 However, in the Pep.4 complex, it is only -1.4 kcal·mol⁻¹. -1 This extremely strong electrostatic and hydrogen bond interaction mainly stems from the tight binding between Glu117 and the arginine side chain and main chain of the control peptide at position P2′. Our sequence design breaks the excessive binding of the superanchor site, effectively redistributing the binding energy to the residues on both sides of the cleavage site.

[0115] In addition, the S1 / S2 / S3 pockets around the enzyme showed slightly stronger binding to the control peptide (total 3–4 kcal·mol⁻¹). -1 However, Pep.4 exhibits better interactions at P2, P1, and P1′, P3′, and P4′, which are adjacent to the cleavage sites. This localized energy enhancement results in a more balanced and favorable energy distribution for catalysis in the P2-P2′ region.

[0116] RMSD analysis shows that ( Figure 6 Pep.4 exhibits a compact and well-fitting conformation within the active pocket, with its RMSD values ​​remaining stable at 0.15–0.30 Å. In contrast, the control peptide shows significant RMSD fluctuations (0.25–0.50 Å) and exhibits pronounced "breathing motion" characteristics. RMSF curves around the cleavage site further reveal that the P1′ and P2′ residues of the control peptide fluctuate dramatically (0.41–0.60 Å), while the corresponding regions of Pep.4 are more stable. This indicates that Pep.4 maintains a more stable local geometry near the cleavage site, enabling more efficient catalytic reactions once it enters the near-attack conformation.

[0117] K-means clustering analysis revealed that over 70% of the trajectory frames of the control peptide were concentrated in a single dominant cluster, with very few transitions to other substates, indicating that it was easily confined to a highly stable but conformationally restricted enzyme-substrate configuration. In contrast, Pep.4 exhibited a wider cluster distribution (the three major clusters accounted for approximately 56%, 32%, and 10%, respectively), and experienced frequent inter-state transitions during simulations. This dynamic and less restrictive binding pattern allows it to more flexibly change to conformational substates favorable for catalysis within the active pocket.

[0118] The results showed that the control peptide formed an overly stable enzyme-substrate complex upon binding. This stable, strong binding limited the substrate's conformational flexibility and hindered its efficient rearrangement to the transition state. Our designed Pep.4 weakened the overstabilizing effect of Glu117, distributing the binding energy evenly on both sides of the cleavage site. This endowed the substrate with better dynamic adaptability in the pocket without disrupting the core catalytic geometry. This moderate binding affinity, combined with enhanced near-attack state sampling capability, significantly improved the catalytic conversion rate of the optimized peptide.

[0119] (II) Construction and characterization of the SERS basis of the silent region intrinsic reference

[0120] To achieve ratioistic quantitative analysis of target enzyme activity in heterogeneous tissue microenvironments, it is crucial to construct a SERS substrate with high Raman-enhanced activity and a stable internal reference signal. This invention employs a multi-step strategy of "seed growth-epithelial coating-spike in situ growth" to prepare a Prussian blue (PB)-doped gold nanostar (APA) array. The preparation process is as follows: Figure 7 PB exhibits a characteristic Raman peak in the biosilent region (BSR) (approximately 2100 cm⁻¹). Combining it with gold nanostars as an internal reference material can meet the requirements for high-sensitivity detection and interference-resistant quantitative analysis.

[0121] Synthesis of gold nanospheres. Gold nanospheres of approximately 20 nm were prepared using the classic sodium citrate reduction method. 100 mL of ultrapure water was placed in a round-bottom flask equipped with a reflux condenser and heated to a vigorous boil while maintaining strong stirring (800–1000 rpm). 1.0 mL of 25 mM HAuCl4 (final concentration 0.25 mM) was rapidly added, followed by a single addition of 2.5 mL of 38.8 mM sodium tricitrate solution. The solution color gradually changed from pale yellow to wine red. The reaction was continued at boiling for 15 minutes, after which heating was stopped, and the mixture was allowed to cool naturally to room temperature and stored away from light. To prepare monodisperse gold nanospheres with larger particle sizes, the above-mentioned 20 nm colloid was used as a seed, and the reaction was carried out at 90°C using a stepwise seed growth method: equal volumes of HAuCl4 (25 mM, 1.0 mL) and sodium tricitrate (38.8 mM, 1.0 mL) were added sequentially every 10 minutes. After 3 rounds, gold nanospheres with an average particle size of 40 ± 3 nm were obtained, and after 6 rounds, gold nanospheres (Au) with a particle size of 60 ± 4 nm were obtained.

[0122] Synthesis of Au@PB core-shell nanoparticles. Using gold spheres as the core, through "CN..." –Au@PB was obtained using a two-step method of "pretreatment-PB epitaxial growth". Specifically, 10 mL of gold nanospheres were taken and 1.0 mL of 5 mM K3[Fe(CN)6] was added while stirring at room temperature (about 650 rpm). The reaction was carried out for 5–10 minutes to form adsorbed CN on the gold surface. – Layer (Au@CN). Subsequently, under the control of two syringe pumps, equimolar concentrations of FeCl3 and K4[Fe(CN)6] precursor solutions (0.5 mM each) were injected at a total flow rate of 30 µL / min. –1 The product was simultaneously added dropwise to the Au@CN colloid, and the stirring was increased to approximately 1500 rpm. After the addition was complete, stirring was continued for 3 hours. The product was purified by centrifugation at 10000 rpm for 18 minutes. The supernatant was discarded, and the product was resuspended in ultrapure water and washed three times. The final resuspended volume was adjusted to 4 mL. The shell thickness could be continuously controlled within the range of several nanometers to tens of nanometers by adjusting the concentration of the PB precursor and the total amount of feed. As the shell thickness increased, the solution color gradually turned dark blue.

[0123] Growth of gold shells on PB-doped gold nanostars (APA). 200 mL of ultrapure water was adjusted to pH 1.0 with dilute hydrochloric acid and stirred at 650 rpm at room temperature. 206 µL of 242 mM HAuCl4 was added to form a gold precursor solution of approximately 0.25 mM; subsequently, 4 mL of Au@PB colloid was added. Almost simultaneously, 2.0 mL of 3 mM AgNO3 and 1.0 mL of 100 mM Vc were rapidly added while stirring at 1500 rpm. After approximately 40 s, the solution color deepened, indicating the formation of the outer gold shell. The product was purified by centrifugation at 10,000 rpm for 15 minutes and washed three times with water to obtain APA nanoparticles. Following a similar method, gold nanospheres were replaced with Au@PB colloid to prepare PB-free gold nanostars (Austar). According to "CN..." – Austar@PB and Au@PB with PB surface coating were prepared by a two-step method of "pretreatment-PB epitaxial growth".

[0124] Construction of the APA array: A silicon wafer was placed in 100 mL of boiling pure water, 20 mL of ammonia and 30% H2O2 were added, and the mixture was stirred until homogeneous. The mixture was heated to boiling and maintained for 30 minutes. It was then rinsed three times with pure water and anhydrous ethanol. Next, it was immersed in a 2% anhydrous ethanol solution of 3-aminopropyltrimethoxysilane for 12 hours. It was then rinsed three times with anhydrous ethanol and pure water, and dried in an oven at 95°C. Finally, it was immersed in 0.24 M dilute hydrochloric acid for 10 minutes and rinsed three times with pure water. The wafer was then immersed in an APA gold nanoparticle sol and shaken at low speed for 36 hours to obtain the complete APA array.

[0125] The morphology and elemental composition of nanomaterials at each stage of synthesis were characterized using transmission electron microscopy and energy-dispersive X-ray spectroscopy. Figure 8 As shown, Au@PB core-shell nanoparticles synthesized via a two-step method of "CN⁻ pretreatment—PB epitaxial growth" exhibit a uniformly dispersed spherical morphology. High-resolution imaging and elemental surface scanning analysis revealed a uniformly distributed shell material with a thickness of 5–10 nm on the surface of the gold core, which has a diameter of approximately 40 nm. This shell is rich in iron (Fe) and nitrogen (N), confirming that the PB shell was successfully coated on the surface of the gold spheres.

[0126] Using Au@PB as a seed, radially arranged gold star spikes were grown on its surface through an induced reduction reaction. During the anisotropic growth of the spikes, due to the strong interfacial interaction and lattice stress pull between PB and reduced gold atoms, the PB originally covering the outer layer co-migrated as the gold spikes extended, ultimately forming PB-doped gold nanostar particles (APA) within the spikes. Transmission electron microscopy images showed that the APA nanoparticles had abundant sharp spikes, a uniform overall morphology, and a diameter distribution of approximately 100 nm. Elemental analysis revealed that Fe and N elements, representing PB, were no longer confined to the gold sphere shell but were radially doped into the interior of the gold shell along the growth direction of the gold spikes.

[0127] The assembled APA array was characterized by scanning electron microscopy. The APA gold nanostars exhibited a high-density, large-area, uniform arrangement on the silicon substrate. Figure 9 .

[0128] To investigate the effect of PB doping structure on Raman enhancement, this invention employs a two-dimensional FDTD method to simulate the spatial electric field distribution (|E / E0|) of PB-coated gold nanospheres (Au@PB), PB-coated gold nanostars (Austar@PB), and PB-doped gold nanostars (APA) under 785 nm excitation. (See [link to FDTD method]). Figure 10 Calculations show that, compared to the smooth-surfaced Au@PB (with a maximum field strength ratio of only 3.6), the sharply curved Autar@PB exhibits a significant electromagnetic field enhancement at the spike peaks, reaching a maximum of 29. However, when PB is distributed as a dopant within the gold spikes, its local electromagnetic field is significantly enhanced, with the maximum field strength ratio increasing to 118. Spatially, the strongest superhotspots do not appear only at the outermost tip of the gold spikes, but penetrate deep into the inner layer of the spikes near the tip of the PB nanocrystal doped layer. The enhancement effect in non-tip regions shows a clear gradient weakening, and the enhancement effect on the surface of the core gold sphere is significantly shielded.

[0129] Raman signals from three different substrates were acquired using a 785 nm handheld Raman spectrometer. (See [reference needed]) Figure 11 Spectral results show that it is located at BSR 2100 cm⁻¹ −1The PB characteristic peak signal is strongest in the nearby APA structure, significantly decreases in Autar@PB, and is very weak in Au@PB. The APA with internally doped PB exhibits a high internal reference signal, which not only closely matches the internal super-hot spot predicted by FDTD, but also indicates that this structure achieves significant amplification of the BSR internal reference signal without occupying surface binding sites.

[0130] The stability of the PB signal, used as a quantitative internal reference, under complex physiological conditions was then examined; see [reference needed]. Figure 12 During 30 minutes of continuous irradiation with a 785nm laser, the intensity of the PB characteristic peak signal showed no significant attenuation, demonstrating excellent resistance to photobleaching. After contact with buffer solutions ranging from pH 2 to 10, the Raman intensity of the PB from the APA array maintained stability of over 96% over a wide acid-base range, exhibiting acid and alkali resistance. Furthermore, regardless of incubation for up to 24 hours in pure water, PBS buffer, culture medium containing 10% FBS, or physiological saline, or long-term storage at room temperature for up to 28 days, the average fluctuation of the PB signal from the APA array did not exceed 5%. These results fully demonstrate that the PB tightly doped within the gold lattice receives substantial physicochemical protection, with its 2100 cm⁻¹ peak intensity remaining stable. −1 The signal can serve as a stable internal reference for quantitative Raman detection.

[0131] (III) Construction and characterization of MMP-2 enzyme activity response array

[0132] The N-terminus of the peptide is covalently modified with lipoic acid, utilizing the Au-S coordination bond in its disulfide ring structure to anchor the peptide to the AuS array surface; the C-terminus of the peptide is coupled with Nile Blue (NB), a reporter molecule with strong Raman activity, which exhibits a Raman shift of 590 cm⁻¹ under 785 nm laser excitation. -1 A very strong Raman resonance characteristic peak was generated. The peptide modification was customized by Hefei Sener Biotechnology Co., Ltd., with a purity greater than 90%. A solution containing 30 μM lipoic acid-DITPAAMTSPP-NB was prepared, and the APA substrate was immersed in the solution and incubated gently with shaking at 4°C in the dark for 2 hours. The substrate was gently rinsed with ultrapure water and buffer, and gently dried with nitrogen gas to complete the construction of the MMP-2 enzyme activity-responsive SERS array (MMP-APA). See [link to documentation]. Figure 13 .

[0133] MMP-2 solutions of 0, 10, 50, 100, and 200 ng / mL were dropped onto the surface of the MMP-APA array. After incubation at room temperature for 8 minutes, spectra were acquired using a handheld Raman spectrometer. A 590 cm⁻¹ wavelength was selected. 1 The NB characteristic peak at 2100 cm⁻¹ is used as the response signal in the biological silent region. -1The characteristic peak of PB was used as a quantitative internal reference. The absolute intensities of the two target peaks were statistically analyzed, and the Raman intensity ratio I was plotted. 590 / I 2100 Linear relationship fitting curve between I and MMP-2 activity. As MMP-2 activity increased from 0 to 200 ng / mL, I 590 The absolute intensity showed a significant concentration-dependent decrease, while the internal reference peak (I) at 2100 cm⁻¹ was attributed to the latter. 2100 () Remained stable. Further quantitative analysis showed that () Figure 14 Raman signal ratio I after internal reference calibration 590 / I 2100 It showed a linear negative correlation with MMP-2 activity in the range of 0–200 ng / mL (R0). 2 = 0.9865).

[0134] The following interfering solutions were individually added to the MMP-APA array: blank PBS buffer (pH 7.4), 200 ng / mL MMP-9, 200 μg / mL BSA, Glu, Arginine (Arg), Cys, GSH, Vitamin C, and 100 μM NO2. - Mg 2+ Ca 2+ ,ClO - O2 - ·OH, H2O2. After incubation for 8 minutes, the Raman spectra were measured using a handheld Raman spectrometer. The interaction between each interfering substance and MMP-2 was analyzed at I... 590 / I 2100 Differences in the intensity ratio. For example... Figure 15 As shown, under the condition of various high concentrations of interfering substances, the key Raman intensity ratio I of the array 590 / I 2100 The levels were almost identical to the blank PBS control group, with no significant change. Only in the presence of the specific target MMP-2 did I... 590 / I 2100 Only then did a significant reduction occur. This result confirms that the MMP-APA array is resistant to interference from non-specific impurity molecules and has high MMP-2 selectivity.

[0135] (iv) Rapid visualization of enzyme activity in animal models using MMP-APA array

[0136] To evaluate the ability of the MMP-APA array to quantitatively detect MMP-2 activity in vivo, this invention constructed an LPS-induced mouse neuroinflammation model and a KA-induced mouse orthotopic epilepsy model, respectively. After model establishment, Raman signals were acquired using a "microdroplet in-situ extraction-in vitro SERS array detection" strategy: PBS buffer microdroplets were briefly contacted with mouse brain tissue to extract local metabolites and free enzymes, and then rapidly added to the surface of the MMP-APA array. Raman spectra were acquired using a handheld Raman spectrometer. A sample of approximately 1.2 × 0.7 cm² was collected within a very short time (< 10 minutes). 2 Point-by-point Raman signal acquisition was performed at 80–90 gridded sampling points within the target brain region. Raman imaging was conducted in live mice using a handheld Raman instrument to generate a brain tissue-level MMP-2 activity distribution map. Figure 16 Imaging results showed almost no Raman signal fluctuations observed in the brain microenvironment of normal mice. In the LPS neuroinflammation model, MMP-2 distribution exhibited a large area of ​​high-activity signal (green dashed line area), successfully locating extensive inflammatory boundaries. In the KA epilepsy model, MMP-2 activity distribution also showed high-level signal and a wide range of inflammation.

[0137] References

[0138] [1]WOLOSOWICZ M, PROKOPIUK S, KAMINSKI T W. The Complex Role ofMatrix Metalloproteinase-2 (MMP-2) in Health and Disease [J]. Int J Mol Sci, 2024, 25(24).

[0139] [2]EBLE JA, NILAND S. The extracellular matrix in tumor progression and metastasis [J]. Clin Exp Metastas, 2019, 36(3): 171-98.

[0140] [3]KESSENBROCK K, PLAKS V, WERB Z. Matrix Metalloproteinases:Regulators of the Tumor Microenvironment [J]. Cell, 2010, 141(1): 52-67.

[0141] [4]QUINTERO-FABIAN S, ARREOLA R, BECERRIL-VILLANUEVA E, et al. Roleof Matrix Metalloproteinases in Angiogenesis and Cancer [J]. Front Oncol,2019, 9: 1370.

[0142] [5]SHOARI A, COBAN M A, RADISKY E S. Therapeutic potential ofinhibitors of gelatinases MMP-2 and MMP-9 for the treatment of breast cancer[J]. Febs J, 2025.

[0143] [6]RAK-PASIKOWSKA A, KAMINSKA M, NIECHCIALA M, et al. siRNA-mediatedinhibition of NTT-MMP-2 reduces oxidative stress and apoptotic signaling inan ex vivo model of ischemia / reperfusion injury [J]. Scientific Reports,2025, 16(1).

[0144] [7]SCHMITT R, TSCHEUSCHLER A, LASCHINSKI P, et al. A potential keymechanism in ascending aortic aneurysm development: Detection of a linearrelationship between MMP-14 / TIMP-2 ratio and active MMP-2 [J]. Plos One,2019, 14(2).

[0145] [8]JIAN M H, SU M, GAO J X, et al. Peptide microarray-basedfluorescence assay for quantitatively monitoring the tumor-associated matrixmetalloproteinase-2 activity [J]. Sensor Actuat B-Chem, 2020, 304.

[0146] [9]SNOEK-VAN BEURDEN P A M, VON DEN HOFF J W. Zymographic techniquesfor the analysis of matrix metalloproteinases and their inhibitors [J].Biotechniques, 2005, 38(1): 73-83.

[0147]

[10] DESHMUKH A A, WEIST J L, LEIGHT J L. Detection of proteolyticactivity by covalent tethering of fluorogenic substrates in zymogram gels[J]. Biotechniques, 2018, 64(5): 203-10.

[0148]

[11] REN Z, CHEN J, KHALIL R A. Zymography as a Research Tool in theStudy of Matrix Metalloproteinase Inhibitors [J]. Methods Mol Biol, 2017,1626(1940-6029 (Electronic)): 79-102.

[0149]

[12] KIRCHHAIN A, POMA N, SALVO P, et al. Biosensors for measuringmatrix metalloproteinases: An emerging research field [J]. Trac-Trend AnalChem, 2019, 110: 35-50.

[0150]

[13] AKERS W J, XU B G, LEE H, et al. Detection of MMP-2 and MMP-9Activity

[0151] with a Triple-Helical Peptide Optical Probe [J]. Bioconjugate Chem,2012, 23(3): 656-63.

[0152]

[14] FIELDS G B. Using fluorogenic peptide substrates to assay matrixmetalloproteinases [J]. Methods Mol Biol, 2010, 622: 393-433.

[0153]

[15] MARTÍNEZ V R, LEE E, OH J W. Exploring Plasmonic StandaloneSurface-Enhanced Raman Scattering Nanoprobes for Multifaceted Applications inBiomedical, Food, and Environmental Fields [J]. Nanomaterials-Basel, 2024, 14(22).

[0154]

[16] GONG T X, KONG K V, GOH D, et al. Sensitive surface enhancedRaman scattering multiplexed detection of matrix metalloproteinase 2 and 7cancer markers [J]. Biomed Opt Express, 2015, 6(6): 2076-87.

[0155]

[17] LIU L, CHU H Y, YANG J K, et al. Construction of a magnetic-fluorescent-plasmonic nanosensor for the determination of MMP-2 activitybased on SERS-fluorescence dual-mode signals [J]. Biosens Bioelectron, 2022,212.

[0156]

[18] CHEN Y Y, ZHU D, ZHONG H Y, et al. Ultrasensitive Detection ofMatrix Metalloproteinase 2 Activity Using a Ratiometric Surface-EnhancedRaman Scattering Nanosensor with a Core-Satellite Structure [J]. Acs AppliedMaterials & Interfaces, 2024, 16(3): 4160-8.

[0157]

[19] LIN D, HSIEH C L, HSU K C, et al. Geometrically encoded SERSnanobarcodes for the logical detection of nasopharyngeal carcinoma-relatedprogression biomarkers [J]. Nature Communications, 2021, 12(1).

[0158]

[20] SLOAN-DENNISON S, WALLACE G Q, HASSANAIN W A, et al. AdvancingSERS as a quantitative technique: challenges, considerations, and correlativeapproaches to aid validation [J]. Nano Converg, 2024, 11(1).

[0159]

[21] TIAN Y F, ZHAO L, SHEN X F, et al. Self-assembled core-shellnanoparticles with embedded internal standards for SERS quantitativedetection and identification of nicotine released from snus products [J].Front Chem, 2024, 12.

[0160]

[22] YIN Y M, LI Q, MA S S, et al. Prussian Blue as a Highly Sensitiveand Background-Free Resonant Raman Reporter [J]. Anal Chem, 2017, 89(3):1551-7.

[0161]

[23] LI M, WANG J Y, CHEN Q Q, et al. Background-Free QuantitativeSurface Enhanced Raman Spectroscopy Analysis Using Core Shell Nanoparticleswith an Inherent Internal Standard [J]. Anal Chem, 2019, 91(23): 15025-31.

[0162]

[24] KUKREJA M, SHIRYAEV S A, CIEPLAK P, et al. High-ThroughputMultiplexed Peptide-Centric Profiling Illustrates Both Substrate CleavageRedundancy and Specificity in the MMP Family [J]. Chemistry & Biology, 2015,22(8): 1122-33.

[0163]

[25] DÍAZ N, SUÁREZ D, SUÁREZ E. Kinetic and binding effects inpeptide substrate selectivity of matrix metalloproteinase-2: Moleculardynamics and QM / MM calculations [J]. Proteins-Structure Function andBioinformatics, 2010, 78(1): 1-11。

Claims

1. A method for preparing a SERS array for quantitative detection of MMP-2 enzyme activity, characterized in that, A solid-phase SERS array for rapid quantitative detection of MMP-2 enzyme activity was constructed by integrating screened fast-response MMP-2 substrate peptides, Prussian blue biosilent region internal controls, gold nanostar hotspot arrays, Nile blue Raman reporter molecules, and lipoic acid immobilization strategies. The specific steps are as follows: (1) Screening and optimization of fast-response MMP-2 substrate peptides: Based on the MMP-2 protease substrate cleavage database, the enzymatic reaction rate constants of candidate substrate peptides were obtained; on this basis, combined with the AI / Rosetta structure-guided substrate extension strategy, bilateral sequence extension, three-dimensional complex modeling, fixed backbone sequence optimization, molecular dynamics simulation and binding free energy analysis were performed on the core short peptides with MMP-2 selectivity. Furthermore, through high-performance liquid phase dynamics detection and solid-phase surface reaction efficiency detection, the preferred substrate peptide with fast enzymatic cleavage response to MMP-2 was obtained: DITPAAMTSPP; (2) Construction of Prussian blue-doped gold nanostar array: Specifically, Au@PB core-shell nanoparticles were prepared by cyano pretreatment and epitaxial growth of gold nanospheres as the core; Au@PB core-shell nanoparticles were then used as seeds to induce anisotropic growth of the outer gold shell under acidic conditions by chloroauric acid, silver nitrate and ascorbic acid, so that Prussian blue co-migrated and doped into the interior of the gold nanostar spikes during the growth process, thus obtaining Prussian blue-doped gold nanostar nanoparticles; Subsequently, the Prussian blue-doped gold nanostar nanoparticles were fixed on the surface of a silicon wafer solid substrate after aminosilanization treatment to form a Prussian blue-doped gold nanostar SERS array; Finally, the local electric field enhancement induced by dielectric environment, the cascade focusing of multi-level tip structures, plasmon coupling and chemical enhancement were combined to form a strong local electromagnetic field enhancement effect inside the gold spikes, so that the array itself has a stable strong Raman signal located in the biological silent region; at the same time, the high-activity sites on the surface are retained for the assembly of enzyme response elements; (3) Constructing a rapid quantitative detection SERS array for MMP-2 enzyme activity: Specifically, Nile blue, the MMP-2 substrate peptide obtained in step (1) and lipoic acid are connected in a 1:1:1 molecular relationship to construct a lipoic acid-substrate peptide-Nile blue probe; wherein lipoic acid is used as a gold surface immobilization group, the substrate peptide is used as an MMP-2 enzyme cleavage response element, and Nile blue is used as a Raman reporter molecule; then the Prussian blue doped gold nanostar SERS array obtained in step (2) is immersed in 10–100 μM of the lipoic acid-substrate peptide-Nile blue probe solution, so that the probe is immobilized on the surface of the gold nanostar through Au-S coordination, and a SERS array for quantitative detection of MMP-2 enzyme activity is obtained; Traditional SERS methods for detecting MMP-2 employ solution-phase nanoprobes, such as core-satellite structures, magnetic nanoparticle-gold nanoparticle complexes, and enzyme-induced aggregation / deaggregation systems. These methods offer high sensitivity, but often require steps such as nanoparticle pre-assembly, enzyme digestion, magnetic separation, centrifugation, or redispersion, making the detection process relatively complex and difficult to standardize into a standardized array detection platform.

2. The preparation method according to claim 1, characterized in that, The screening and optimization of MMP-2 substrate peptides in step (1) specifically includes the following steps: (1.1) The second-order reaction rate constants (kobs) of candidate peptides with respect to MMP-2 and MMP-9 were obtained from the protease substrate kinetics database, and the standardized cleavage score (Z-score) of the candidate peptides was obtained in combination with proteomics cleavage data; multiple independent measurements of the same sequence were normalized, and the median was taken as the representative value of the sequence, where the reaction rate difference was: Δkobs = kobs(MMP-2) - kobs(MMP-9); The differences in cutting scores are as follows: ΔZ = Z(MMP-2) - Z(MMP-9); Candidate sequences were sorted according to Δkobs and ΔZ to screen for candidate peptides that simultaneously exhibited high MMP-2 response rates and low MMP-9 responses. (1.2) Based on the core short peptide with high selectivity for MMP-2, a complex model of the core short peptide and MMP-2 was constructed using AlphaFold2-Multimer, and the RFjoint algorithm was used to complete the sequence on both sides of the core short peptide to generate the extended substrate peptide backbone. (1.3) The Rosetta software was used to optimize the sequence of the extended substrate peptide under the condition of fixed backbone, and the candidate amino acids of the extension site were restricted to the range of high-frequency amino acids obtained by MMP-2 substrate preference analysis. (1.4) The optimized candidate substrate peptides were screened by MMP-2 enzyme digestion kinetics, solid-phase SERS interface response, molecular dynamics simulation, MM / PBSA binding free energy calculation and single residue energy decomposition analysis, and finally the preferred substrate peptide DITPAAMTSPP was obtained.

3. The preparation method according to claim 2, characterized in that, The specific process for constructing the Prussian blue-doped gold nanostar array in step (2) is as follows: (2.1) Gold nanospheres were prepared by sodium citrate reduction method; HAuCl4 was added to boiling ultrapure water to make the final concentration of HAuCl4 0.1–1.0 mM, and then sodium tricitrate solution was added to carry out reduction reaction to obtain gold nanospheres. (2.2) Using the gold nanospheres as seeds, gold nanospheres with an average particle size of 20–60 nm were obtained by stepwise growth; (2.3) Take gold nanosphere colloids and add K3[Fe(CN)6] solution for cyano pretreatment to form an adsorbed CN⁻ layer on the gold surface; (2.4) FeCl3 solution and K4[Fe(CN)6] solution were simultaneously added to the cyano-pretreated gold nanosphere colloid to allow Prussian blue to grow epitaxially on the surface of the gold nanospheres, forming Au@PB core-shell nanoparticles; (2.5) Au@PB core-shell nanoparticles were dispersed in an acidic aqueous solution with pH 0.5–2.0, and HAuCl4, AgNO3 and ascorbic acid were added to induce anisotropic gold shell growth on the surface of Au@PB, forming Prussian blue doped gold nanostars with spike structures. (2.6) The obtained Prussian blue doped gold nanostars were purified by centrifugation and washed with water to obtain Prussian blue doped gold nanostar nanoparticles; (2.7) After the solid substrate is treated with aminosilanization, it is immersed in Prussian blue doped gold nanostar colloid for shaking incubation, so that the Prussian blue doped gold nanostar is fixed on the surface of the solid substrate to form a SERS array.

4. The preparation method according to claim 1, characterized in that, The specific process for building the SERS array in step (3) is as follows: (3.1) Using a silicon wafer as a solid substrate, the following processes were performed sequentially: oxidation cleaning, ethanol and water cleaning, aminosilanization treatment, acid activation, and drying treatment; wherein: The aminosilanization treatment involves immersing the cleaned silicon wafer in a 2% anhydrous ethanol solution of 3-aminopropyltrimethoxysilane for 6–24 hours. The acid activation involves immersing the aminosilanized silicon wafer in a 0.1–0.5 M hydrochloric acid solution for 5–30 minutes. (3.2) The treated silicon wafer is immersed in Prussian blue doped gold nanostar colloid and shaken and incubated for 12–48 hours, so that the Prussian blue doped gold nanostars are fixed on the surface of the silicon wafer through electrostatic adsorption and surface interaction, thus obtaining a Prussian blue doped gold nanostar SERS array. (3.3) Prepare a 10–100 μM probe solution by preparing the lipoic acid-substrate peptide-Nile blue probe; wherein, lipoic acid, substrate peptide and Nile blue are covalently linked to form a single probe molecule, and the molecular linkage ratio of the three is 1:1:1; (3.4) Immerse the SERS array obtained in step (3.2) in the probe solution and incubate gently with shaking for 1–4 hours at 0–10°C in the dark. (3.5) After incubation, the unbound probes were washed away with ultrapure water and buffer, and dried with nitrogen to obtain a SERS array for quantitative detection of MMP-2 enzyme activity.

5. A SERS array obtained by the preparation method according to any one of claims 1-4.

6. The application of the SERS array as described in claim 5 in the quantitative detection of MMP-2 enzyme activity.

7. The application according to claim 6, characterized in that, The specific steps are as follows: (1) The sample to be tested is dropped onto the surface of the SERS array, so that the MMP-2 in the sample to be tested reacts with the substrate peptide immobilized on the array surface; (2) Incubate at room temperature for 1–16 minutes; (3) The Raman spectra of the SERS array were acquired using a 785 nm excited Raman spectrometer; (4) Select Nile blue at approximately 590 cm -1 The Raman peak at approximately 2100 cm⁻¹ was used as the MMP-2 response signal, with Prussian blue at approximately 2100 cm⁻¹ selected as the peak. -1 The Raman peak in the biological quiescent region was used as an internal reference signal; (5) Calculate the Raman intensity ratio I 590 / I 2100 The activity of MMP-2 enzyme in the test samples was quantitatively analyzed according to the pre-established MMP-2 activity standard curve.

8. The application according to claim 7, characterized in that, The MMP-2 enzyme activity detection range is 0–200 ng / mL, within which range I 590 / I 2100 The SERS array exhibits a linear negative correlation with MMP-2 enzyme activity, with a linear correlation coefficient R² of not less than 0.98; the SERS array also shows a selective response to MMP-2.

9. The application according to claim 7, characterized in that, The sample to be tested is an in vitro enzyme solution, cell culture medium, tissue extract, brain tissue surface microdroplet extract, or intraoperative tissue microenvironment sample. When the sample to be tested is a microdroplet extract from the surface of brain tissue, PBS buffer microdroplets are briefly brought into contact with the brain tissue surface to extract local free enzymes and metabolites. The microdroplets are then transferred to the surface of a SERS array for Raman detection, and the results are obtained through multiple spatial sampling points. 590 / I 2100 Plot the spatial distribution of MMP-2 enzyme activity using ratios.

Images