Super-resolution imaging method based on chemical reaction luminescence and application

Abstract

The invention discloses a super-resolution imaging method based on chemical reaction luminescence and application, and the method comprises the following steps: marking an imaging target by adopting a specific marking system, and triggering a marking site to generate a chemical reaction through a reaction excitation system in situ to generate a photon signal; the method comprises the following steps: acquiring photon signals generated by chemical reaction by using an imaging system in a continuous multi-frame sampling manner to obtain a data set containing isolated photon spatio-temporal information; image reconstruction is carried out after the data set is preprocessed, information entropy and correlation cumulant are calculated pixel by pixel, and a preliminary super-resolution reconstruction image is obtained based on weighting of the information entropy and the correlation cumulant; and performing post-deconvolution on the preliminary super-resolution reconstructed image to obtain a final super-resolution reconstructed image. According to the invention, the problems of weak reaction luminescence signal and insufficient information utilization are solved, and super-resolution imaging of objects such as protein biomolecules, subcellular structures, bacteria, living organisms, single particles, catalysts and the like is realized.

Description

Technical Field

[0001] This invention belongs to the field of super-resolution imaging technology, specifically relating to a super-resolution imaging method and its application based on chemical reaction luminescence. Background Technology

[0002] Optical imaging, with its advantages of in-situ, real-time dynamic, and non-invasive nature, has been widely used to analyze complex material structures and life processes. However, limited by the optical diffraction limit, visualizing the fine internal structure of matter and cells with a spatial resolution below 200 nm has long been a scientific challenge in the field of optical imaging. The "super-resolution fluorescence microscopy" technique, which won the 2014 Nobel Prize in Chemistry, broke through this traditional limitation, revolutionizing the spatial resolution of traditional imaging systems by utilizing the linear or nonlinear optical response mechanism excited by fluorescent molecules. In recent years, with the evolution and iteration of fluorescence super-resolution microscopy, this technology has become the foundation for precise visualization and analysis of biological phenomena. For example, classic single-molecule localization microscopy, stimulated emission depletion microscopy, and super-resolution optical wave imaging can all track the spatial distribution of proteins within single cells with a spatial resolution below 100 nm. However, fluorescence imaging relies on external light source excitation; the introduction of lasers inevitably causes bleaching of fluorescent molecules and interference from background autofluorescence, increasing imaging uncertainty. Furthermore, high-power excitation light can cause phototoxicity to the sample, making long-term imaging observation of live samples difficult. Therefore, developing a novel optical imaging excitation mode that is fundamentally different from fluorescence excitation to achieve background-free, high-fidelity, and live-cell-friendly imaging technology has important application prospects.

[0003] Chemiluminescence is a form of emission that relies on intermolecular redox reactions to generate excited-state light. Since the reaction does not require an external light source, it fundamentally avoids the excitation light effect and holds promise as an alternative to fluorescence excitation. Chemiluminescence imaging can be categorized into three main areas based on its principles: chemiluminescence based on random collisions of free molecules in solution, electrochemiluminescence based on voltage-controlled triggering, and bioluminescence based on active enzyme catalysis. The selectable chemiluminescence principles endow chemiluminescence imaging modalities with advantages such as zero background, high signal-to-noise ratio, high sensitivity, and good biocompatibility. Therefore, this technology has been widely used in fields such as bioimaging. Currently, there are various bioimaging applications based on chemiluminescence reactions. For example, patent application CN121141625A provides a microfluidic chip for chemiluminescence imaging detection of single-cell secretions and its usage method; cell typing is also performed using electrochemiluminescence, such as patent application CN117629975A which provides a cell typing method based on electrochemiluminescence imaging technology and its application in identifying cellular heterogeneity; and in vivo apoptosis detection is also performed using bioluminescence, such as patent application CN113295679A which provides a BRET in vivo imaging probe for detecting apoptosis.

[0004] However, a single chemical molecule emits only one photon per reaction, and the low photon yield limits the spatiotemporal resolution of reactive luminescence imaging, hindering its wider application in bioimaging. To improve the spatial resolution of traditional reactive luminescence imaging, patent application CN117083518A provides a single-photon signal acquisition method, imaging system, and its application for electrochemiluminescence, achieving imaging of free reactant molecules in chemical reaction solutions and improving the spatial resolution of electrochemiluminescence imaging below the diffraction limit. However, this technology can only image the outline of objects attached to electrodes, lacking specificity and limiting its broader biological applications. Based on the above discussion, current imaging methods based on reactive luminescence have still failed to achieve specific super-resolution imaging, making it difficult to meet the growing demands of biological applications.

[0005] In summary, achieving super-resolution imaging based on chemiluminescence faces the following challenges: how to achieve specific labeling of analytes in chemiluminescence systems; how to improve the yield of chemiluminescence in biological systems, i.e., increase its luminescence intensity to meet the measurement baseline of the detector; how to collect the process signals of chemiluminescence to ensure maximum preservation of photon spatiotemporal information; and how to utilize the spatiotemporal information of chemiluminescence and achieve efficient, high-throughput super-resolution reconstruction. Summary of the Invention

[0006] In view of the above, the purpose of this invention is to provide a super-resolution imaging method and application based on chemical reaction luminescence. By constructing a universal and designable chemical reaction luminescence super-resolution imaging system, in-situ, zero-background super-resolution imaging of subcellular structures can be achieved, as well as dynamic long-term super-resolution imaging tracking of living cells can be realized.

[0007] To achieve the above-mentioned objectives, the present invention provides the following technical solution: This invention provides a super-resolution imaging method based on chemical reaction luminescence, comprising the following steps: A specific labeling system is used to label the imaging target, and a chemical reaction is triggered in situ at the labeling site by a reaction excitation system to generate a photon signal; The imaging system was used to collect photon signals generated by chemical reactions using a continuous multi-frame sampling method to obtain a dataset containing spatiotemporal information of isolated photons; After preprocessing the dataset, the spatiotemporal information of the photon signal is used for reconstruction, and then deconvolution processing is performed to obtain the final super-resolution reconstructed image.

[0008] More preferably, the specific labeling system comprises a pair of molecules that specifically bind to the imaging target, wherein one molecule is used to specifically identify the imaging target and the other molecule is coupled to an excitable luminescent probe, and the reaction excitation system comprises a chemical component that can react with the luminescent probe to generate a luminescent signal, and a sample carrier carrying the imaging target.

[0009] More preferably, the molecular pair that specifically binds to the imaging target is selected from antigen-antibody pairs, nucleic acid aptamers, recognition proteins and glycosyl groups, or luminescent enzymes and fluorescent protein chimeric compounds.

[0010] More preferably, when the chemical reaction is an electrochemical reaction luminescence, the excitable label is an electrochemiluminescent probe, and the chemical components include a conductive medium and a co-reactant; when the chemical reaction is a chemiluminescent reaction luminescence, the excitable label is a chemiluminescent probe, and the chemical components include a chemiluminescent reaction substrate; when the chemical reaction is a bioluminescent reaction luminescence, the excitable label is a bioluminescent enzyme probe, and the chemical components include a bioluminescent reaction substrate.

[0011] More preferably, the electrochemiluminescent probe is a goat anti-rabbit IgG antibody labeled with ruthenium tripyridine, the co-reactant is 0.01-0.2 M bis(2-hydroxyethyl)amino(tris(hydroxymethyl)methane) and the conductive medium is phosphate buffer; the chemiluminescent probe is a goat anti-rabbit IgG antibody labeled with horseradish peroxidase, the chemiluminescent reaction substrate is 0.05-1 mM luminol and 10 µM-30 mM hydrogen peroxide; the bioluminescent probe is a luciferase-green fluorescent protein chimeric compound, the bioluminescent reaction substrate is 2-10 μM furazine.

[0012] More preferably, the sample carrier carrying the imaging target includes fixed cells, live cells, fixed bacteria, live bacteria, live animals, catalysts or single particles, and other micro-nano scale research objects that can achieve super-resolution imaging by the method.

[0013] More preferably, the sampling exposure time in the imaging system is determined based on the photon emission characteristics of the corresponding luminescent system, so as to obtain the corresponding photon distribution sequence that is correlated in both time and space.

[0014] More preferably, the photon signal sampling exposure time in the electrochemical reaction luminescence system is 20 ms-1 s; and the photon signal sampling exposure time in the chemiluminescence and bioluminescence systems is 0.2 s-1 s.

[0015] More preferably, the imaging system relies entirely on the electron transfer of the chemical reaction itself to generate photon emission, without the need for any external excitation light source.

[0016] More preferably, pre-deconvolution is used to preprocess the dataset.

[0017] More preferably, the deconvolution preprocessing includes performing Richardson-Lucy deconvolution on the dataset.

[0018] More preferably, before preprocessing the dataset using pre-deconvolution, Gaussian filtering and / or interpolation upsampling are used to process the dataset to obtain a noise-suppressed densely sampled dataset.

[0019] More preferably, the Gaussian filtering process includes: generating a Gaussian function with a corresponding full width at half maximum (FWHM) based on the point spread function size, and then filtering the dataset.

[0020] More preferably, the interpolation upsampling preprocessing includes: transforming the dataset to the Fourier domain, performing upsampling with zero padding, and then transforming the upsampled image back to the spatial domain.

[0021] More preferably, the reconstruction using the spatiotemporal information of the photon signal includes: The information entropy and autocorrelation accumulation are calculated pixel by pixel based on the dataset and weighted to obtain the entropy-weighted correlation accumulation. The cross-entropy and cross-correlation accumulation between pixels are calculated based on the dataset and weighted to obtain the cross-entropy-weighted cross-correlation accumulation. The reconstructed image after resolution improvement is constructed based on the entropy-weighted correlation accumulation and the cross-entropy-weighted cross-correlation accumulation.

[0022] More preferably, post-deconvolution processing is performed based on joint constraints including continuity constraints and sparse constraints. Specifically, Richardson-Lucy deconvolution is used to maintain the coherence of the reconstruction results by utilizing continuity constraints and to improve spatial resolution by utilizing sparse constraints, so as to obtain the final super-resolution reconstructed image.

[0023] This invention also provides an application of the super-resolution imaging method based on chemiluminescence as described above, for super-resolution imaging of protein biomolecules, single cells, single bacteria, living organisms, single particles, or catalysts.

[0024] Compared with the prior art, the beneficial effects of the present invention include at least the following: (1) By constructing a specific labeling system, a reaction excitation system, an imaging system and a super-resolution reconstruction system, this invention breaks through the temporal and spatial resolution limitations of traditional chemiluminescence-based imaging. For the first time, it achieves continuous high-resolution imaging of subcellular structures such as mitochondria, microfilaments, microtubules and endoplasmic reticulum based on chemiluminescence. The spatial resolution of the imaging is better than 100 nm and the temporal resolution is 2 s. Compared with traditional cell imaging based on chemiluminescence, the spatiotemporal throughput is increased by nearly 100 times.

[0025] (2) By adopting a biocompatible reaction system and mild excitation conditions, this invention overcomes the limitations of photobleaching and phototoxicity in traditional fluorescence imaging, and can realize continuous observation of the dynamic process of living cells for up to 41 hours, breaking through the imaging time limitation of fluorescence super-resolution imaging.

[0026] (3) This invention is based on a two-step deconvolution signal processing framework and introduces entropy-weighted correlation cumulative analysis to effectively extract spatiotemporal correlation information in sparse scintillation signals. It combines the continuity prior and the relative sparsity prior for biological structures to form a joint constraint, and constructs a super-resolution reconstruction algorithm suitable for chemical reaction luminescence imaging to achieve zero background and high throughput super-resolution reconstruction. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 This is a flowchart illustrating the super-resolution imaging method based on chemiluminescence provided in an embodiment of the present invention. Figure 2 A schematic diagram of the structure of the system of this invention; Figure 3 This is a flowchart of the super-resolution reconstruction system; Figure 4 This is a schematic diagram of the electrochemiluminescence imaging principle and device; Figure 5 This is a diagram showing the optimized conditions for the electrochemiluminescence reaction system. Figure 6 This is a super-resolution reconstruction comparison image based on electrochemiluminescence; Figure 7 This is a quantitative analysis diagram based on super-resolution results of electrochemiluminescence; Figure 8 This is a comparison of the results of electrochemiluminescence super-resolution and fluorescence super-resolution. Figure 9 This is a large-field super-resolution imaging result based on electrochemiluminescence; Figure 10 This is a three-dimensional super-resolution imaging result of cells based on electrochemiluminescence; Figure 11 This is a schematic diagram of the bioluminescence imaging principle and device; Figure 12 This is a dynamic super-resolution imaging result of mitochondria based on bioluminescence; Figure 13 This is a quantitative analysis diagram of dynamic super-resolution imaging results based on bioluminescence; Figure 14 This is a continuous 41-hour mitochondrial tracking and quantitative analysis diagram based on bioluminescence super-resolution; Figure 15 This is a diagram of intercellular mitochondrial transfer results based on bioluminescence super-resolution. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the scope of protection of this invention.

[0031] This invention discloses a super-resolution imaging method based on chemiluminescence, encompassing three general and designable reactive luminescence super-resolution imaging techniques: electrochemiluminescence, chemiluminescence, and bioluminescence. Specifically, it includes a specific labeling system for the imaging target, a reaction excitation system, an imaging system, and a super-resolution reconstruction system based on the luminescence signal. The specific labeling system and reaction excitation system are used to in-situ trigger the luminescence reaction at the analyte, generating a detectable photon signal. The imaging system collects the photon emission from the chemical reaction, which possesses spatiotemporal resolution information. The super-resolution reconstruction system uses algorithms to improve the spatiotemporal resolution of the imaging based on the spatiotemporal information of the chemical reaction luminescence. The correspondence between each system is as follows: Figure 2 As shown. Unlike traditional chemiluminescence systems, the specific labeling and reaction excitation systems designed in this invention can achieve in-situ triggering of luminescence reactions in subcellular structures and generate positively labeled photon signals. Unlike traditional long-exposure, single-frame acquisition strategies, the photon signal acquisition strategy designed in this invention employs short-exposure, continuous multi-frame sampling to fully record the information flow of chemiluminescence fluctuations. Targeting the unique photon fluctuation characteristics of chemiluminescence, the designed super-resolution reconstruction algorithm maximizes the utilization of relevant information contained in the acquired data, achieving the highest resolution subcellular structure positively labeled super-resolution image reconstruction based on chemiluminescence to date.

[0032] like Figure 1 As shown, the embodiment provides a super-resolution imaging method based on chemical reaction luminescence, including the following steps: S1 uses a specific labeling system to label the imaging target, and then triggers a chemical reaction at the labeling site in situ through a reaction excitation system to generate a photon signal.

[0033] Specifically, depending on the different imaging scenario requirements, the present invention designs and adopts different chemical reaction luminescence imaging modes.

[0034] The specific labeling and reaction excitation system based on electrochemiluminescence described in this invention is realized through a targeted labeling part, an electrochemical reaction part, an electrochemical triggering part, and a host part.

[0035] In one embodiment, the specific labeling method for imaging targets includes an immunolabeling method based on antigen-antibody specific recognition, a direct labeling method based on click chemistry, a labeling method based on nucleic acid aptamers, a labeling method based on biotin and streptavidin specific recognition, and a labeling method based on target protein and glycosyl protein specific recognition.

[0036] In one embodiment, the luminescent probe in the electrochemiluminescence system comprises ruthenium terpyridine, iridium terpyridine, or other transition metal complexes, or nanoparticles or quantum dots coated with the aforementioned complexes. Highly specific targeted labeling of the analyte can be achieved by attaching any of the aforementioned electrochemiluminescence probes to the recognition molecule in the molecular pair of the labeling method described above.

[0037] In one embodiment, the electrochemiluminescence reaction excitation system includes a sample cell containing co-reactant molecules and the specifically labeled object to be imaged. The co-reactant molecules include reducing co-reactants such as tripropylamine and bis(2-hydroxyethyl)amino(tris(hydroxymethyl)methane), or oxidizing co-reactants such as hydrogen peroxide. The sample cell also contains the three electrodes commonly used in electrochemiluminescence reactions: a working electrode, a counter electrode, and a reference electrode. The labeled object to be imaged is adhered to the working electrode, and the working and counter electrodes are immersed in the solution. The three electrodes do not interfere with each other. The sample cell is connected to an electrochemical workstation via external wires, or to a signal triggering device such as a data acquisition card, signal generator, or electrochemical workstation with similar signal control functions, enabling flexible control of the applied signal.

[0038] Preferably, the imaging labeling method is an immunolabeling method based on the specific recognition of antigen and antibody. Preferably, the immunolabeling component includes a rabbit monoclonal antibody linked to an imaging target and a goat anti-rabbit IgG antibody linked to a rabbit monoclonal antibody, wherein the goat anti-rabbit IgG antibody needs to be coupled to a corresponding electrochemiluminescent probe.

[0039] Preferably, to suit biological applications, unlike traditional tripropylamine (TPrA) co-reactant molecules, the co-reactant molecule of this invention uses neutral bis(2-hydroxyethyl)amino(tris-hydroxymethyl)methane (Bis-tris), which exhibits superior electrochemiluminescence performance compared to traditional acidic solutions. It is also dissolved in 0.01 M phosphate buffer. This strategy not only solves the problem of weak electrochemiluminescence signals in traditional labeling systems but also ensures a good imaging environment for biological samples, achieving a 1000-fold increase in electrochemiluminescence intensity compared to traditional methods, thus resolving the issue of weak electrochemiluminescence signals in specific traditional labeling systems.

[0040] The electrochemical triggering section includes an electrochemical workstation or other voltage-applying triggering device. The electrochemical workstation is connected to the reference electrode, counter electrode, working electrode, and host. The host is used to regulate the electrochemical workstation to regulate the magnitude of the voltage applied to the electrochemical reaction system, thereby controlling the intensity of the luminescence emitted by the reaction.

[0041] The specific labeling and reaction excitation system based on chemiluminescence or bioluminescence described in this invention is achieved through a targeted labeling part and a reaction triggering part.

[0042] In one embodiment, the targeting marker portion may be a specific recognition marker based on antigen-antibody molecular pairs, an intermolecular specific recognition based on click chemistry, a recognition marker based on transient protein expression self-assembly, or a stable cell line construction based on gene editing.

[0043] In one embodiment, the chemiluminescent probe comprises a goat anti-rabbit IgG antibody labeled with horseradish peroxidase, and the bioluminescent probe comprises a luciferase-fluorescent protein chimera.

[0044] In one embodiment, the chemiluminescent or bioluminescent reaction excitation system includes a sample cell containing a reaction substrate, the specifically labeled object to be imaged, and a live-cell workstation, which ensures cell viability during the imaging process. The chemiluminescent reaction substrate is a classic chemiluminescent substrate such as luminol or hydrogen peroxide, while the bioluminescent reaction substrate is a classic bioluminescent substrate such as D-luciferin, coelenterate, or furazolidone.

[0045] Preferably, the targeting marker is a recognition marker for transient protein expression self-assembly. A TOMM20 plasmid and a bioluminescent system plasmid specifically targeting the outer mitochondrial membrane are designed, and the two plasmid segments are fused for expression. The plasmid delivery part includes methods such as phospholipid transfection or gene knock-in.

[0046] Preferably, the probe of the bioluminescent system is a bioluminescent energy resonance transfer system based on the intercalation of NanoLuc luciferase and green fluorescent protein.

[0047] As a preferred option, the reaction substrate is a live-cell optimized and compatible furazolidone.

[0048] S2 uses an imaging system to collect photon signals generated by chemical reactions in a continuous multi-frame sampling manner to obtain a dataset containing spatiotemporal information of isolated photons.

[0049] Specifically, the imaging system includes a microscopic imaging system, a photon detector, and a host computer. The microscopic imaging system is an inverted imaging system placed below the sample cell to collect emission signals. It includes an oil immersion objective with a high numerical aperture for collecting weak photon information from the chemical reaction emission system. The end of the microscopic imaging system is connected to the photon detector, which receives photon signals and converts the collected photon information into electrical information. The photon detector is either a high-quantum-efficiency and low-readout-noise electron multiplier camera or a high-sensitivity photodetector and its array, used for collecting weak photon signals. Its end is connected to the host computer, and it sequentially sends the collected signals to the host computer. The collected spatiotemporally isolated and correlated photon signals include their pixel position at the current moment and their grayscale values ​​relative to multiple neighboring pixel photons. The host computer displays the converted grayscale image.

[0050] In one embodiment, unlike the long-exposure, single-frame acquisition strategy of traditional chemiluminescence and bioluminescence imaging, this invention employs a short-exposure, continuous multi-frame acquisition strategy in its imaging system. This strategy fully records the entire process of molecular reaction and emission, utilizing the rich temporal-spatial correlation information contained within this process to achieve super-resolution reconstruction of the target structure. The isolated, scintillation photon signals used for super-resolution reconstruction described in this invention are obtained by controlling the reactant concentration and the exposure time. Excessively high reactant concentrations increase the frequency of molecular scintillation, making it difficult to distinguish adjacent emission signals; excessively low reactant concentrations result in incomplete excitation of the labeling system, making it impossible to reconstruct the overall distribution of the analyte. Conversely, excessively high exposure times ignore the temporal correlation information of photons on the same pixel at different times, while excessively low exposure times lead to sparsity of molecular signals, which is detrimental to subsequent image reconstruction.

[0051] In one embodiment, the concentration of the co-reactant in the electrochemiluminescence reaction ranges from 0.01 to 0.2 M, and the exposure time is 20 ms to 1 s.

[0052] In one embodiment, the concentration range of the chemiluminescent or bioluminescent reaction substrate is 2-10 μM, and the exposure time is 0.2 s-1 s.

[0053] S3 preprocesses the dataset, then uses the spatiotemporal information of the photon signal to reconstruct the image, and finally performs deconvolution processing to obtain the final super-resolution reconstructed image.

[0054] Specifically, in super-resolution reconstruction systems, due to the low photon budget of reactive luminescence, Poisson and Gaussian noise during camera acquisition can interfere with the molecular luminescence scintillation signal. Therefore, as... Figure 3As shown, this invention constructs a signal processing framework based on two-step deconvolution, introduces entropy-weighted correlation cumulant analysis to effectively extract spatiotemporal correlation information in sparse flicker, and introduces joint constraints on the continuity prior and the relative sparsity prior for biological structures to achieve image super-resolution reconstruction.

[0055] In one embodiment, the dataset is preprocessed using pre-deconvolution (specifically, the Richardson-Lucy deconvolution algorithm). Pre-deconvolution increases the on / off contrast of the reactive light flicker signal while suppressing sampling noise.

[0056] Optionally, before deconvolution preprocessing, Gaussian filtering is first applied to the dataset to suppress high-frequency random noise in the data.

[0057] Alternatively, the Gaussian-filtered data can be spatially upsampled in two dimensions to increase pixel density and provide pixel support for subsequent resolution improvements.

[0058] After signal preprocessing, a super-resolution reconstruction algorithm targeting the scintillation characteristics of reactive light emission is designed. Reactive light-emitting molecules react independently and randomly with reactants and release photons, exhibiting independent and random scintillation information. Based on this characteristic, this invention proposes an entropy-weighted correlation cumulant to extract information from the reactive light emission scintillation signal to achieve super-resolution reconstruction.

[0059] In one embodiment, information entropy is first calculated pixel-by-pixel based on the dataset to obtain a two-dimensional entropy map. In the distribution region of reactant molecules, random luminescence scintillation signals have high information content, resulting in high entropy values; conversely, only the background region with sampling noise contains less information, leading to lower entropy values. The entropy map reflects the spatial distribution of reactant molecules and filters out sampling noise. Subsequently, autocorrelation cumulative values ​​are calculated pixel-by-pixel based on the dataset to highlight the scintillation signals of isolated molecules while suppressing mixed scintillation signals of multiple molecules, achieving molecular localization resolution within the point spread function distance. Finally, the entropy map is weighted by the autocorrelation cumulative values ​​to obtain a super-resolution reconstructed image with uniform molecular distribution and a high signal-to-noise ratio.

[0060] In another embodiment, the present invention further calculates the cross-entropy and cross-correlation accumulation between pixels based on the dataset, and obtains the cross-entropy weighted cross-correlation accumulation by weighting, and constructs the reconstructed image with improved resolution based on the above-mentioned entropy weighted correlation accumulation and cross-entropy weighted cross-correlation accumulation.

[0061] Furthermore, this invention introduces post-deconvolution based on joint constraints. Considering the continuity of biological structures and the Nyquist sampling theorem, biological structure imaging possesses prior characteristics of continuity and relative sparsity. Therefore, performing post-deconvolution on the super-resolution reconstruction results from the previous step based on continuous and relatively sparse prior constraints (specifically using the Richardson-Lucy deconvolution algorithm) can further improve spatial resolution and suppress reconstruction artifacts that may be introduced by deconvolution, thereby maximizing the resolution of the super-resolution reconstruction.

[0062] To explain the purpose and technical approach of this invention in detail, the super-resolution imaging method based on chemical reaction luminescence proposed in this invention will be further described below through specific embodiments and in conjunction with the accompanying drawings.

[0063] Example 1: Super-resolution imaging of immobilized intracellular microtubules based on electrochemiluminescence

[0064] An electrochemiluminescence super-resolution method for imaging immobilized intracellular microtubules comprises the following steps: specific recognition of tubulin and electrochemiluminescence probe labeling, triggering an electrochemiluminescence reaction at the microtubule, acquiring the photon signal emitted by the reaction, and super-resolution processing based on the signal.

[0065] Step 1: Specific recognition and labeling of tubulin In the fixed-cell labeling imaging process, specific recognition of tubulin was achieved by adding 6.48 μg / mL of anti-tubulin rabbit monoclonal antibody to fixed, permeabilized cells and incubating overnight at 4°C. Then, rabbit anti-tubulin on the cells was specifically recognized by 4 μg / mL of ruthenium tripyridine-labeled goat anti-rabbit IgG antibody, and incubated at 37°C for 1 h. The ruthenium tripyridine-modified goat anti-rabbit IgG antibody was achieved through a covalent chemical reaction between the ruthenium tripyridine modified with the succinimide group and the naked amino group of the goat anti-rabbit antibody.

[0066] Step 2: Triggering the electrochemiluminescence reaction The electrochemiluminescence reaction system includes: cells labeled with a ruthenium terpyridine luminescent probe, obtained in step one and coated onto the working electrode; a co-reactant solution; a counter electrode; a reference electrode; and an electrochemical workstation. The imaging cells are African green monkey kidney cells. The working electrode is a transparent indium tin oxide electrode. The co-reactant solution is 5 mL of 0.1 M bis(2-hydroxyethyl)amino(tris(hydroxymethyl)methane). The counter electrode is a platinum electrode, and the reference electrode is a silver / silver chloride electrode. The voltage required to trigger the electrochemiluminescence reaction is applied by the electrochemical workstation and is 1.0–1.4 V. The specific electrochemiluminescence system is as follows: Figure 4 As shown.

[0067] Since the signal intensity of the electrochemiluminescence reaction is affected by factors such as voltage, type of co-reactant, and concentration of co-reactant, the above imaging parameters are optimized, such as... Figure 5 As shown, the specific steps are as follows: (1) For example Figure 5 As shown in (a), 0.1 M tripropylamine (TPrA) and bis(2-hydroxyethyl)amino(trihydroxymethyl)methane (Bis-tris) were added to the above electrochemiluminescence reaction system, and a DC voltage of 1.2 V was applied to the system. The intensity of the electrochemiluminescence reaction was tested as a function of time.

[0068] (2) For example Figure 5 As shown in (b), bis(2-hydroxyethyl)amino(trihydroxymethyl)methane with concentrations of 0.01 M, 0.1 M, and 0.2 M were added to the above electrochemiluminescence reaction system, and a DC voltage of 1.2 V was applied to the system. The intensity of the electrochemiluminescence reaction was tested as a function of time.

[0069] (3) such as Figure 5 As shown in (c), 0.1 M of bis(2-hydroxyethyl)amino(trihydroxymethyl)methane was added to the above electrochemiluminescence reaction system, and DC voltages of 1.0, 1.2, and 1.4 V were applied to the system respectively to test the reaction intensity of electrochemiluminescence.

[0070] (4) Electrochemiluminescence intensity analysis is performed by reading and analyzing the average gray value of the luminescent cell region using Matlab, and comparing its average value and standard deviation. In addition to the direct imaging intensity, the contrast of cell imaging is also analyzed, which is defined as the intensity ratio of the cell region to the background region. Figure 5 This is a condition optimization diagram of the electrochemiluminescence cell imaging system. Through analysis, the optimal conditions were determined to be the use of bis(2-hydroxyethyl)amino(trihydroxymethyl)methane with a co-reactant concentration of 0.1 M and the application of a DC voltage of 1.2 V.

[0071] Step 3: Acquire photon signals The steps for acquiring photon signals on single-cell microtubes using the above photon signal acquisition system under the optimized electrochemiluminescence reaction conditions in step two (4) are as follows: (1) Place the fixed cells marked in step one above into the sample cell of the electrochemiluminescence reaction system, and apply voltage to the electrochemiluminescence reaction system according to the optimized conditions in step two (4).

[0072] (2) The microscopic imaging system below the sample cell is equipped with a 100X oil immersion objective lens with a high numerical aperture to collect low-photon signals. After the photons are absorbed by the objective lens, they are detected by the electron multiplication camera at the back end and transmitted to the host. The acquisition area is 512*512 pixels (81.92*81.92 µm), the single frame exposure time is 20 ms, and a total of 100 frames are acquired for each cell.

[0073] Step 4: Super-resolution reconstruction The processing steps for super-resolution reconstruction of the acquired time-series photon signal dataset are as follows: (1) Preprocessing to enhance signal detection. The acquired time-series photon signal dataset was subjected to two-dimensional Gaussian pre-filtering with a Gaussian kernel standard deviation of 1 to suppress high-frequency random sampling noise. Subsequently, the dataset underwent 2×2 lateral Fourier upsampling, increasing the pixel base by 4 times while preserving the signal, providing pixel support that satisfies the Nyquist sampling theorem for subsequent resolution improvements. Then, the dataset was deconvolved using the Richardson-Lucy deconvolution algorithm, with the convolution kernel fitted using a Gaussian function based on the imaging parameters. The deconvolution iterations were performed 5 times. This enhanced the flickering contrast of the reactive photon signal and suppressed pixel-level random fluctuation noise.

[0074] (2) Cumulative Value Calculation for Super-Resolution Reconstruction. An entropy map is obtained by calculating the information entropy pixel-by-pixel on the preprocessed and enhanced reactive luminescence signal dataset. The reactive luminescence time-series signal is divided into 100 bins within its minimum to maximum value range for information entropy calculation. Second-order correlation cumulative values ​​are also calculated pixel-by-pixel. Simultaneously, cross-entropy and cross-correlation cumulative values ​​between pixels are calculated based on the dataset and weighted to obtain cross-entropy weighted cross-correlation cumulative values. Uniform super-resolution reconstruction is achieved by combining the entropy-weighted correlation cumulative values ​​and the cross-entropy weighted cross-correlation cumulative values.

[0075] (3) Post-deconvolution maximizes resolution reconstruction. A joint constraint deconvolution step is applied to the super-resolution reconstruction result of the previous step, introducing continuous prior constraints and sparse prior constraints. The continuous prior constraints are implemented using the Hessian matrix; the joint constraint optimization problem is solved by split Braggman algorithm, with 100 iterations. The deconvolution iterations are 10. Super-resolution reconstruction of the fixed cell microtubule structure is achieved.

[0076] After reconstruction using the aforementioned super-resolution algorithm, super-resolution imaging of microtubules in African green monkey kidney cells based on electrochemiluminescence can be achieved. Figure 6 The images show a before-and-after comparison of super-resolution processing, clearly demonstrating a significant improvement in resolution compared to before processing. Figure 7 As shown, bimodal analysis and Fourier ring correlation analysis reveal that the imaging spatial resolution of this system can reach 97 nm, and as... Figure 8Analysis shows that it can achieve the same resolution level as existing fluorescence super-resolution imaging based on wave analysis.

[0077] The super-resolution reconstruction algorithm proposed in this invention also brings excellent temporal resolution (2 s), such as Figure 9 As shown, this method can achieve rapid scanning super-resolution imaging with a large field of view in a short time (18 s, 0.53 × 0.53 mm). 2 ).

[0078] Leveraging the excellent spatiotemporal controllability of electrochemiluminescence reactions, three-dimensional super-resolution imaging of intracellular structures can be achieved through variable voltage excitation-assisted layer scanning technology, such as... Figure 10 As shown, electrochemiluminescence can be used to achieve microtube imaging within a single cell with an excitation depth of 1.8 µm, with an axial resolution of 235 nm and a lateral resolution of 116 nm.

[0079] Example 2: Continuous dynamic super-resolution imaging of mitochondria in living cells based on bioluminescence

[0080] A bioluminescent super-resolution method for imaging mitochondria in live cells is provided, comprising the following steps: targeted labeling of mitochondria, triggering a bioluminescent reaction, acquiring the photon signal emitted by the reaction, and super-resolution processing based on the signal. The targeted labeling and reaction triggering parts differ from those in Example 1.

[0081] Step 1: Targeted labeling of mitochondria Mitochondrial targeting is achieved by labeling the mitochondrial targeting protein Tomm20 with a bioluminescent probe. The bioluminescent probe in this invention is a bioluminescent energy resonance transfer system based on a NanoLuc luciferase and green fluorescent protein (GFP) chimeric compound. It relies on NanoLuc luciferase to catalyze the emission of bioluminescence from a furazine substrate at a wavelength of 473 nm, followed by energy resonance transfer to the adjacent GFP, emitting light at a wavelength of 517 nm. The coupling of the targeted Tomm20 protein with the bioluminescent system is achieved through in vitro plasmid construction and liposome transfection, specifically through the following steps: (1) A mitochondrial targeting plasmid was constructed in vitro, and its protein peptide sequence is as follows: MVGRNSAIAAGVCGALFIGYCIYFDRKRRSDPNFKDRLRERRKKQKLAKERAGLSKLPDLKDAEAVQKFFLEEIQLGEELLAQGDYEKGVDHLTNAIAVCGQPQQLLQVLQQTLPPPVFQMLLTKLPTISQRIVSAQSFGEDDVE.

[0082] (2) The fusion plasmid constructed in step (1) was delivered to African green monkey kidney cells with a growth density of 70% using the transfection reagent Lipo 3000 and cultured continuously for 24 h.

[0083] Step 2: Triggering the bioluminescence reaction The bioluminescent reaction system includes the specific mitochondrial-targeting cells constructed in step one, the bioluminescent reaction substrate, and a live-cell workstation. The specific luminescence principle is as follows: Figure 11 As shown. The bioluminescent substrate has a certain degree of cell membrane permeability, allowing it to enter the cell and directly participate in the bioluminescent reaction. Therefore, triggering the bioluminescent reaction only requires adding a 4 µM concentration of substrate directly to the cell culture medium. The live cell workstation is used to maintain the physiological conditions required for normal cell life activities, with culture parameters set at 37°C, 5% carbon dioxide concentration, and 80% humidity.

[0084] Step 3: Acquire photon signals The photon signal acquisition hardware device for the bioluminescence system is consistent with the photon signal acquisition system described in step three of Embodiment 1 above. The photon information collected by the microscopic imaging system is transmitted to the electron multiplication camera. However, due to differences in luminescence intensity and fluctuation characteristics between the two, the exposure time for acquiring the corresponding images differs. The exposure time for bioluminescence imaging is set to 0.5 s, and the acquisition area is 512*512 pixels (81.96*81.96 µm). Data for each cell is continuously acquired until cell apoptosis.

[0085] Step 4: Super-resolution reconstruction The processing steps for super-resolution reconstruction of the acquired time-series photon signal dataset are as follows: (1) Preprocessing to enhance signal detection. The acquired time-series photon signal dataset was subjected to a two-dimensional Gaussian pre-filter with a Gaussian kernel standard deviation of 1 to suppress high-frequency random sampling noise. Subsequently, the dataset underwent 2×2 lateral Fourier upsampling, increasing the pixel base by 4 times while preserving the signal, providing pixel support that satisfies the Nyquist sampling theorem for subsequent resolution improvements. Then, the dataset was subjected to Richardson-Lucy deconvolution, with the convolution kernel fitted using a Bessel function based on the imaging parameters. The deconvolution iterations were performed 5 times. This enhances the flickering contrast of the reactive photoluminescence signal and suppresses pixel-level random fluctuation noise.

[0086] (2) Cumulative Value Calculation for Super-Resolution Reconstruction. The information entropy of the preprocessed and enhanced reactive luminescence signal dataset is calculated pixel-by-pixel to obtain an entropy map. The reactive luminescence time-series signal is divided into 100 bins within its minimum to maximum value range for information entropy calculation. Second-order correlation cumulative values ​​are also calculated pixel-by-pixel. Simultaneously, cross-entropy and cross-correlation cumulative values ​​between pixels are calculated based on the dataset and weighted to obtain cross-entropy weighted cross-correlation cumulative values. Uniform super-resolution reconstruction is achieved by combining entropy weighted correlation cumulative values ​​and cross-entropy weighted cross-correlation cumulative values. Further, iterative wavelet analysis is used to estimate and remove defocus noise. The signal is decomposed into 7 levels of different frequency bands using db6 wavelet basis wavelet analysis. The lowest frequency band is used as the initial estimate of the continuously slowly changing defocus signal, and it is compared with half the maximum value of the original signal to prevent filtering of useful information. This step is iterated three times to obtain the super-resolution reconstruction result with defocus noise removed.

[0087] (3) Post-deconvolution maximizes resolution reconstruction. A joint constraint deconvolution step is applied to the super-resolution reconstruction result of the previous step, introducing continuous prior constraints and sparse prior constraints. The continuous prior constraints are implemented using the Hessian matrix; the joint constraint optimization problem is solved by split Braggman algorithm, with 100 iterations. The deconvolution iterations are 5. Super-resolution reconstruction of the mitochondrial structure of living cells is achieved.

[0088] (4) The super-resolution reconstruction of live cell data is based on the principle of rolling reconstruction. The number of original image frames required for reconstruction of a single super-resolution image is 30 frames, and the rolling step is 10 frames.

[0089] After reconstruction using the aforementioned super-resolution algorithm, continuous super-resolution imaging of mitochondria in African green monkey kidney cells based on bioluminescence can be achieved. Figure 12 The results show the continuous dynamic changes of mitochondria over 60 minutes after super-resolution processing. The spatial resolution of the continuous dynamic imaging over 41 hours was 104 nm. Figure 13 ).

[0090] The good biocompatibility of bioluminescence allows for continuous dynamic observation of single cells for up to 41 hours. Figure 14 The results represent the statistical analysis of mitochondrial morphology in a single cell over a period of 41 hours.

[0091] Continuous dynamic observation also allows for real-time imaging of mitochondrial transfer processes between cells, with specific trends in these processes as follows: Figure 15 Show.

[0092] The specific embodiments described above illustrate the technical solution and beneficial effects of the present invention in detail. It should be understood that the above description is only the most preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, additions, and equivalent substitutions made within the scope of the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A super-resolution imaging method based on chemical reaction luminescence, characterized in that, Includes the following steps: A specific labeling system is used to label the imaging target, and a chemical reaction is triggered in situ at the labeling site by a reaction excitation system to generate a photon signal; The imaging system was used to collect photon signals generated by chemical reactions using a continuous multi-frame sampling method to obtain a dataset containing spatiotemporal information of isolated photons; After preprocessing the dataset, the spatiotemporal information of the photon signal is used for reconstruction, and then deconvolution processing is performed to obtain the final super-resolution image.

2. The super-resolution imaging method based on chemiluminescence according to claim 1, characterized in that, The specific labeling system comprises a pair of molecules that specifically bind to the imaging target, wherein one molecule is used to specifically identify the imaging target and the other molecule is coupled to an excitable luminescent probe; the reaction excitation system comprises a chemical component that can react with the luminescent probe to generate a luminescent signal and a sample carrier that carries the imaging target.

3. The super-resolution imaging method based on chemiluminescence according to claim 2, characterized in that, The molecular pairs that specifically bind to the imaging target are selected from antigen-antibody pairs, nucleic acid aptamers, recognition proteins and glycosyl groups, or luminescent enzymes and fluorescent protein chimeric compounds.

4. The super-resolution imaging method based on chemiluminescence according to claim 2, characterized in that, When the chemical reaction is an electrochemical reaction luminescence, the excitable label is an electrochemiluminescent probe, and the chemical components include a conductive medium and a co-reactant; when the chemical reaction is a chemiluminescent reaction luminescence, the excitable label is a chemiluminescent probe, and the chemical components include a chemiluminescent reaction substrate; when the chemical reaction is a bioluminescent reaction luminescence, the excitable label is a bioluminescent enzyme probe, and the chemical components include a bioluminescent reaction substrate.

5. The super-resolution imaging method based on chemiluminescence according to claim 4, characterized in that, The electrochemiluminescent probe is an IgG antibody labeled with ruthenium tripyridine, the co-reactant is bis(2-hydroxyethyl)amino(trihydroxymethyl)methane, and the conductive medium is phosphate buffer; the chemiluminescent probe is an IgG antibody labeled with horseradish peroxidase, and the chemiluminescent reaction substrate is luminol and hydrogen peroxide; the bioluminescent probe is a luciferase-fluorescent protein chimeric compound, and the bioluminescent reaction substrate is furazine.

6. The super-resolution imaging method based on chemiluminescence according to claim 2, characterized in that, The sample carriers carrying the imaging targets include fixed cells, live cells, fixed bacteria, live bacteria, live animals, catalysts or single particles, and other micro- and nano-scale research objects that can achieve super-resolution imaging by the method.

7. The super-resolution imaging method based on chemiluminescence according to claim 1, characterized in that, The exposure time for continuous multi-frame sampling in the imaging system is determined based on the photon emission characteristics of the corresponding luminescent system, so as to obtain the corresponding photon distribution sequence that is correlated in both time and space.

8. The super-resolution imaging method based on chemiluminescence according to claim 1, characterized in that, This imaging system relies entirely on the electron transfer of the chemical reaction itself to generate photon emission, without the need for any external excitation light source.

9. The super-resolution imaging method based on chemiluminescence according to claim 1, characterized in that, Preprocessing of the dataset is performed using pre-deconvolution.

10. The super-resolution imaging method based on chemiluminescence according to claim 9, characterized in that, Before preprocessing the dataset using pre-deconvolution, Gaussian filtering and / or interpolation upsampling are applied to obtain a densely sampled dataset with noise suppression.

11. The super-resolution imaging method based on chemiluminescence according to claim 1, characterized in that, The reconstruction using the spatiotemporal information of photon signals includes: The information entropy and autocorrelation accumulation are calculated pixel by pixel based on the dataset and weighted to obtain the entropy-weighted correlation accumulation. The cross-entropy and cross-correlation accumulation between pixels are calculated based on the dataset and weighted to obtain the cross-entropy-weighted cross-correlation accumulation. The reconstructed image after resolution improvement is constructed based on the entropy-weighted correlation accumulation and the cross-entropy-weighted cross-correlation accumulation.

12. The super-resolution imaging method based on chemiluminescence according to claim 1, characterized in that, Post-deconvolution processing is performed based on joint constraints including continuity constraints and sparse constraints. The continuity constraints are used to maintain the coherence of the reconstruction results, while the sparse constraints are used to improve the spatial resolution, so as to obtain the final super-resolution reconstructed image.

13. An application of the super-resolution imaging method based on chemiluminescence as described in any one of claims 1 to 12, characterized in that, Used for super-resolution imaging of protein biomolecules, single cells, single bacteria, living organisms, single particles, or catalysts.

Images