Method for quantitative analysis of colloidal gold labeled aggregation degree in electron microscope

Abstract

The application discloses a kind of electron microscope colloidal gold label aggregation degree quantitative analysis method, including obtaining transmission electron microscope colloidal gold label picture, then the electron microscope colloidal gold label picture is pretreated, for the picture after pretreatment parameter-dependent semi-automatic division gold particle and manually selected region of interest, then the area density of real gold particle in region of interest is calculated, the aggregation degree of gold particle is calculated using average nearest neighbor analysis method to electron microscope colloidal gold in the range of 60-140 area density, finally identify ANNA numerical value between 0.6-1.4 electron microscope colloidal gold is labeled as qualified positive marker data.The application selects suitable gold particle positive marker data by the calculation of gold particle area density and aggregation degree, eliminates the interference of false positive to ANNA quantitative analysis, so that average nearest neighbor analysis method can be suitable for electron microscope colloidal gold label scene quantitative analysis.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of electron microscope colloidal gold labeling technology, and particularly relates to a method for quantitatively analyzing the aggregation degree of electron microscope colloidal gold labeling. BACKGROUND

[0002] The distribution of substances in a living organism mostly has spatial and temporal differences. Electron microscope technology is an important method for obtaining position information of a sample at a microscale. People use nanoscale particle-shaped colloidal gold as a marker to identify the position of a target object under an electron microscope, and the distribution mode thereof includes the following modes: random, aggregation, dispersion, etc. The method for obtaining content information under a microscope is mainly based on measurement of a microscopic image, i.e., counting colloidal gold particles in a specific region and calculating the occurrence frequency. In the field of antigen-antibody reaction, the quantity relationship between a target object and a colloidal gold label is basically linear. Based on this linear relationship, in the colloidal gold labeling method, not only can the position of an antigen be qualitatively described by using the position of colloidal gold, but also the content of the antigen can be quantitatively described by using the number (i.e., the density) of gold particles per unit area. Therefore, by using the electron microscope colloidal gold labeling technology, the position information and the content information of a specific component in a biological sample can be simultaneously obtained. However, the existing electron microscope colloidal gold labeling technology can only qualitatively describe the position, and there is no method for quantitatively describing and comparing the aggregation degree between different treatment groups. The image taken by an electron microscope contains rich information, and the information of the aggregation degree is contained in the above-mentioned picture. However, due to the insufficient types of image processing methods of the electron microscope colloidal gold labeling, the aggregation degree between different treatment groups cannot be quantitatively compared based on the image of the electron microscope colloidal gold labeling in the past.

[0003] The method commonly used for quantitatively describing and comparing the aggregation degree of target objects is average nearest neighboring analysis (ANNA). ANNA is a regression analysis, that is, the measured value and the expected value are compared, and the result presented is a numerical value. A numerical value equal to 1 indicates that the target objects are randomly distributed, a numerical value less than 1 indicates aggregation, and the smaller the absolute value, the greater the aggregation. A numerical value greater than 1 indicates dispersion, and the greater the absolute value, the greater the dispersion. However, ANNA cannot be directly applied to the field of electron microscope colloidal gold labeling. Compared with ANNA used in quantitative geography, the application of ANNA in electron microscope colloidal gold labeling has its own characteristics. In both plant ecology and quantitative geography, the target objects (such as trees, craters, and residential areas) used for quantitative analysis are basically real and have positive contribution to the scientific significance of the research field. However, in electron microscope colloidal gold labeling, gold particles are used as markers to represent the position of target protein, and the gold particles as secondary signals have false positives, which causes difficulties in calculating the ANNA value, especially for data sets with no positive and only false positive. The ANNA value has no biological significance and even has a negative contribution. Therefore, a method is needed to make the average nearest neighboring analysis (ANNA method) applicable to the quantitative analysis of electron microscope colloidal gold labeling. SUMMARY

[0004] The technical problem to be solved by the present application is to provide a method for quantitatively analyzing the aggregation degree of electron microscope colloidal gold labeling, so as to make the average nearest neighboring analysis (ANNA method) applicable to the quantitative analysis of electron microscope colloidal gold labeling.

[0005] To solve the above technical problem, the present application provides a method for quantitatively analyzing the aggregation degree of electron microscope colloidal gold labeling, which comprises the following specific processes: obtaining a transmission electron microscope colloidal gold labeling picture, then pre-processing the electron microscope colloidal gold labeling picture in a graphics workstation, using parameter-dependent semi-automatic gold particle segmentation and manually selecting a region of interest for the pre-processed electron microscope colloidal gold labeling picture, then calculating the area density of the real gold particles in the region of interest, the electron microscope colloidal gold labeling with an area density in the range of 60-140 is recognized as positive labeling data suitable for subsequent average nearest neighboring analysis, and the aggregation degree of the gold particles is calculated by using the average nearest neighboring analysis method, and finally the electron microscope colloidal gold labeling with an ANNA value between 0.6 and 1.4 is recognized as qualified positive labeling data.

[0006] Improvements of the method for quantitatively analyzing the aggregation degree of electron microscope colloidal gold labeling according to the present application are as follows:

[0007] The pre-processing includes normalization processing and picture phase inversion processing of the brightness and contrast of the electron microscope colloidal gold labeling picture.

[0008] As a further improvement of the method for quantitative analysis of the aggregation degree of colloidal gold label in electron microscopy of the present application:

[0009] The specific process of the semi-automatic segmentation of gold particles is as follows:

[0010] The real gold particles and false signals to be segmented are screened according to the size factor parameters by three groups of filters to obtain all real colloidal gold particle models gold total in the electron microscope picture, and the three groups of filters are the integrated size and signal peak value, the definition of the boundary between the gold particles and the background, and the signal peak value.

[0011] As a further improvement of the method for quantitative analysis of the aggregation degree of colloidal gold label in electron microscopy of the present application:

[0012] The region of interest includes an inner region of interest ROI inner and an annular region extending outward from the inner region of interest ROI inner, and the outwardly extending annular region is a quality control region ROI peripheral; the sum of the inner region of interest ROI inner and the quality control region ROI peripheral is an outer region of interest ROI outer.

[0013] As a further improvement of the method for quantitative analysis of the aggregation degree of colloidal gold label in electron microscopy of the present application:

[0014] The calculation process of the surface density of the real gold particles is as follows:

[0015] The voxel number of the inner region of interest inner Number of voxel, the voxel number of the outer region of interest outer Number of voxel, the voxel size voxel size, the number of gold particles in the inner region of interest number of gold inner, and the number of gold particles in the outer region of interest number of gold outer of the electron microscope colloidal gold label picture are obtained, and the voxel size includes voxel X and voxel Y; and then the following calculations are sequentially performed:

[0016] (1) The area of the inner region of interest ROI inner is:

[0017] area of inner=inner Number of voxel*voxel X*voxel Y (1)

[0018] (2) The area of the outer region of interest ROI outer is:

[0019] area of outer = outer Number of voxel*voxel X*voxel Y (2)

[0020] (3) Area of the quality control zone ROI peripheral:

[0021] area of peripheral = area of outer - area of inner (3)

[0022] (4) Sorting out gold particle models distributed in the inner region of interest (ROI inner) from the total of all real gold particle models based on a distance threshold, including spots close to ROI inner models and spots far to ROI inner models;

[0023] Sorting out gold particle models distributed in the outer region of interest ROI outer from the total of all real gold particle models based on a distance threshold, including spots close to ROI outer and spots far to ROI outer models;

[0024] (5) Number of real gold particle models in the quality control zone ROI peripheral:

[0025] number of gold peripheral = number of gold outer - number of gold inner (4)

[0026] (6) Areal density of real gold particles in the inner region of interest ROI inner:

[0027] areal density inner = number of gold inner / area of inner (5)

[0028] (7) Areal density of real gold particles in the quality control zone ROI peripheral:

[0029] areal density peripheral = number of gold peripheral / area of peripheral (6)

[0030] (8) Calculation of background subtraction parameters:

[0031] ratio peripheral / inner = areal density peripheral / areal density inner (7)

[0032] (9) The areal density of the real gold particle model after background subtraction:

[0033] areal density = areal density inner - areal density peripheral (8)

[0034] (10) The sample with the areal density of the real gold particle greater than 30 / μm2 is identified as positive, and the sample with the areal density value of 60-140 / μm2 is selected for the calculation of the aggregation degree of the gold particle.

[0035] As a further improvement of the method for quantitative analysis of the aggregation degree of colloidal gold labeled by an electron microscope of the present application:

[0036] The process for calculating the aggregation degree of the gold particle is:

[0037] (1) The inner layer ROI inner model of the inner layer region of interest ROI inner and the gold particle object gold ANNA calculated by the average nearest neighbor analysis method are displayed, and the point connecting line of the outermost gold particle model forms a closed polyline area as the calculation area ROI ANNA of the average nearest neighbor analysis method;

[0038] (2) The number of voxels of the calculation area ROI ANNA of the average nearest neighbor analysis method is obtained;

[0039] (3) The area calculation of the calculation area ROI ANNA of the average nearest neighbor analysis method:

[0040] area of ANNA = number of voxel * voxel size (9)

[0041] (4) The calculation of the expected value De of the regression analysis in the average nearest neighbor analysis method:

[0042]

[0043] wherein Nr gold represents the number of gold particles in the calculation area ROI ANNA of the average nearest neighbor analysis method;

[0044] (5) The actual distance value of the gold particle in the average nearest neighbor analysis method The measurement is performed as follows: the distance between each gold particle model in the calculation region ROI ANNA of the average nearest neighbor analysis method and each other gold particle model is measured in sequence to obtain data sets Distance1, Distance2, Distance n The minimum values in the data sets Distance1, Distance2, Distance n minimum1, Distance minimum2, Distance minimum n are recorded as the shortest distances Distance minimum1, Distance minimum2, Distance minimum n respectively, and the average value of the shortest distances Distance minimum1, Distance minimum2, Distance minimum

[0045] (6) The aggregation degree parameter of the gold particle is calculated:

[0046]

[0047] (7) The sample with the ANNA value between 0.6 and 1.4 is identified as an effective sample.

[0048] The beneficial effects of the present application mainly include:

[0049] 1. Compared with the ANNA method used in the quantitative geography scene, the present application screens the positive marker data of the gold particle by calculating the area density and aggregation degree of the gold particle, and eliminates the interference of false positives on the ANNA quantitative analysis;

[0050] 2. The density of the target object in plant ecology and quantitative geography is difficult to adjust conveniently, and therefore the ANNA calculation is interfered by the excessively high or low density of the target object. The present application establishes the applicable range of the gold particle concentration (represented by the parameter area density) as 60-140 per square micrometer, which can eliminate the interference of false positives on the ANNA calculation, and avoid the excessively high positive rate of the marker, the excessively dense gold particles, and the real aggregation being covered, which is reflected as the ANNA value representing the aggregation degree deviating to the dispersion direction;

[0051] 3. The target object marked by the electron microscope colloidal gold is a man-made marker, and its density can be adjusted. The present application adjusts the target object to a suitable density by the interval range of the area density of the electron microscope colloidal gold marker suitable for the ANNA effect, so as to facilitate the ANNA calculation to achieve the optimal condition. BRIEF DESCRIPTION OF DRAWINGS

[0052] The specific embodiments of the present application will be further described in detail below with reference to the accompanying drawings.

[0053] Figure 1 Flow chart of the method for electron microscope colloidal gold labeling quantitative analysis of the degree of aggregation of the present application;

[0054] Figure 2 Schematic diagram of the calculation region ROI ANNA of the average nearest neighbor analysis (ANNA) method;

[0055] Figure 3 Schematic diagram of the distance between a specific gold particle model and each of the other gold particle models in the gold ANNA object;

[0056] Figure 4 Schematic diagram of the experimental strategy for verifying the method of the present application by collecting data using proteins with known aggregation / dispersion patterns;

[0057] Figure 5 Schematic diagram of the electron microscope colloidal gold labeling results of the sterol methyltransferase protein that controls the bending of the membrane to form vesicles;

[0058] Figure 6 Schematic diagram of the electron microscope colloidal gold labeling results of the Nss endoplasmic protein of tomato spotted wilt virus (TSWV);

[0059] Figure 7 Schematic diagram of the results of the electron microscope colloidal gold labeling quantitative analysis of the degree of aggregation of proteins with known distribution patterns using the ANNA method;

[0060] Figure 8 Schematic diagram of the ANNA value distribution of the unreliable false positive group (Gn, Gc) in the application of the ANNA method to electron microscope colloidal gold labeling quantitative analysis of the degree of aggregation DETAILED DESCRIPTION

[0061] The present application will be further described below in conjunction with specific examples, but the scope of protection of the present application is not limited thereto:

[0062] Example 1, a method for electron microscope colloidal gold labeling quantitative analysis of the degree of aggregation, as shown in Figure 1 the specific process is as follows:

[0063] 1. Image preprocessing

[0064] Obtain the electron microscope colloidal gold labeling image by transmission electron microscopy, then perform preprocessing on the electron microscope colloidal gold labeling image in a graphics workstation, which includes normalization of brightness and contrast, and then perform phase inversion on the image so that the gold particles become signal peaks.

[0065] The method of normalization is through the normalize command in the Image process menu of the Imaris software; the method of phase inversion is through the invert command in the Image process menu of the Imaris software. Common image processing software can perform normalization and phase inversion operations on images. The Imaris used in this case is an interactive microscope image analysis software of the Oxford Instrument Group.

[0066] 2. Calculate the surface density of gold particles for the pre-processed electron microscope colloidal gold label picture

[0067] The surface density of gold particles includes the surface density of the inner region of interest (ROI inner) and the surface density in the 100-nanometer annular region outside the inner region of interest, which is used as a quality control area (ROI peripheral) for background subtraction to improve the signal-to-noise ratio of the electron microscope colloidal gold label. The specific process is as follows:

[0068] 2.1. Semi-automatic segmentation rendering dependent on gold particle parameters

[0069] The gold particles are segmented by the spots algorithm of the Imaris software, with a size factor parameter of 10 nanometers. Three filters are used to screen real gold particles and false signals to be segmented: the quality function screens by combining size and signal peak value parameters; the intensity STDEV function screens by detecting the clarity of the boundary between gold particles and background; and the intensity median function screens by detecting signal peak value. Finally, a gold total model of all real colloidal gold particles in an electron microscope picture is obtained.

[0070] 2.2. Manually segment each region of interest (ROI) by the surface algorithm of the Imaris software, including the inner region of interest (ROI inner) and the annular region extending 100 nanometers outward from the inner region of interest, which is used as a quality control area (ROI peripheral) for background subtraction (the quality control area does not include the inner region of interest). The specific method is as follows:

[0071] Open the contour function under the edit tab of the surface module in Imaris software, click draw in select mode with the mouse, draw a polyline along the boundary of the region of interest until it is closed, and obtain the surface model of the inner region of interest (ROIinner). Then create a new surface algorithm, push the boundary of the inner region of interest (ROIinner) outward by about 100 nanometers, draw a closed polyline, and obtain the surface model of the outer region of interest (ROIouter), wherein the annular region expanded outward by 100 nanometers is the quality control area (ROIperipheral), and the outer region of interest (ROIouter) is the sum of the inner region of interest (ROIinner) and the quality control area (ROIperipheral).

[0072] 2.3, obtain the parameters of the inner region of interest of the picture, such as the number of voxels inner Number of voxel, the number of voxels outer Number of voxel, the voxel size, the number of gold particles in the inner region of interest number of gold inner, and the number of gold particles in the outer region of interest number of gold outer, by Imaris software.

[0073] Wherein, the voxel size includes voxel X and voxel Y, which are obtained automatically by reading the picture parameters after opening the picture properties geometry of Imaris software.

[0074] The inner Number of voxel, outer Number of voxel, number of gold inner, and number of gold outer are the actual measured values of the corresponding number of gold particle aggregation regions.

[0075] 2.4, calculate the area of the inner region of interest (ROIinner):

[0076] area of inner=inner Number of voxel*voxel X*voxel Y (1)

[0077] 2.5, calculate the area of the outer region of interest (ROIouter):

[0078] area of outer=outer Number of voxel*voxel X*voxel Y (2)

[0079] 2.6, Calculate the area of the quality control zone (ROI peripheral):

[0080] area of peripheral = area of outer - area of inner (3)

[0081] 2.7, Sort all the real gold particle models gold total using the algorithm (find spots close to surface plug-in in XTension module of Imaris software) based on distance threshold. Both inner region of interest (ROI inner) and outer region of interest (ROI outer) need to be sorted, and get the groups of gold particle models distributed inside and outside the inner region of interest (ROI inner), and the groups of gold particle models distributed inside and outside the outer region of interest (ROI outer) respectively.

[0082] The operation object is total spots and corresponding surface, and the input distance factor (magnification 40,000 times, voxelsize 9 nanometers 0.1, slightly changed under different magnifications). The sorting operation automatically generates four new spots models, namely spotsclose to ROI inner, spots far to ROI inner, spots close to ROI outer and spots farto ROI outer four models.

[0083] 2.8, Calculate the number of real gold particle models in the quality control zone (ROI peripheral) for background subtraction:

[0084] number of gold peripheral = number of gold outer - number of gold inner (4)

[0085] 2.9, Calculate the areal density of real gold particles in the inner region of interest (ROI inner):

[0086] areal density inner = number of gold inner / area of inner (5)

[0087] 2.10, Calculate the areal density of real gold particles in the quality control zone (ROI peripheral) for background subtraction:

[0088] areal density peripheral = number of gold peripheral / area of peripheral (6)

[0089] 2.11, Calculation of background subtraction parameters:

[0090] ratio peripheral / inner = areal density peripheral / areal density inner (7)

[0091] 2.12, Final areal density of real gold particles after background subtraction:

[0092] areal density = areal density inner - areal density peripheral (8)

[0093] 2.13, Samples with areal density of real gold particles after background subtraction greater than 30 particles per square micron are considered positive. Samples with areal density values between 60 and 140 particles per square micron are selected for subsequent quantitative analysis of aggregation degree.

[0094] 3. For the pre-processed electron microscope colloidal gold labeled image, average nearest neighbor analysis is performed to calculate the aggregation degree of gold particles, the specific process is

[0095] 3.1, Use Imaris software to determine the calculation area of average nearest neighbor analysis (ANNA method): display the spots close to ROI inner model in the inner layer region of interest (ROI inner), and take it as the gold particle object (gold ANNA) for ANNA calculation, and form a closed polyline area with the point connecting line of the outermost gold particle model (so that the area enclosed is as large as possible), as shown by the area enclosed by the dashed line in Figure 2 , to obtain a surface model as the calculation area (ROI ANNA) of average nearest neighbor analysis (ANNA) method;

[0096] 3.2, Obtain the voxel number of the calculation area (ROI ANNA) of average nearest neighbor analysis (ANNA method) by Imaris software;

[0097] 3.3, Area calculation of the calculation area (ROI ANNA) of average nearest neighbor analysis (ANNA) method:

[0098] area of ​​ANNA=number of voxel*voxel size (9)

[0099] 3.4 Calculation of the expected value De in regression analysis using the Average Nearest Neighbor Analysis (ANNA) method:

[0100]

[0101] Where Nr gold represents the number of gold particles in the computational region (ROI ANNA) of the Average Nearest Neighbor Analysis (ANNA) method, which was obtained by Imaris through actual measurement of each electron micrograph of colloidal gold labeling. area of ​​ANNA represents the area of ​​the computational region (ROI ANNA) of the Average Nearest Neighbor Analysis (ANNA) method, which was obtained by Imaris through actual measurement of each electron micrograph of colloidal gold labeling.

[0102] 3.5. Using Imaris software to analyze the measured values ​​of regression analysis in ANNA analysis. The measurement is performed as follows: Select a specific gold particle model and measure its distance (geometric center to geometric center) from every other gold particle model in the gold ANNA object (e.g., ...). Figure 3 As shown, the spacing of the first gold particle is represented by a thin solid line, and the spacing of the second gold particle is represented by a thin line with thickened endpoints, resulting in the data set Distance1. Then, the shortest spacing value in the data set Distance1 is selected and denoted as Distance minimum1 (e.g., ...). Figure 3 As shown in the diagram (represented by thick solid lines); and so on, the distances between each gold particle model and every other gold particle model in the computational region ROI ANNA of the Average Nearest Neighbor Analysis (ANNA) are measured sequentially, resulting in data sets Distance2, Distance3...Distance. n Where n is the computational region (ROI) of the Average Nearest Neighbor Analysis (ANNA) method, and n is the number of gold particle models in the ANNA region. Then, each data group Distance2, Distance3...Distance is divided into groups. n The minimum values ​​in these values ​​are correspondingly denoted as Distanceminimum2, Distanceminimum3, ... Distanceminimum... n Finally, the minimum spacing Distanceminimum1, Distance minimum2, ... Distance minimum are calculated as follows: n The average value is used as the actual spacing value.

[0103] 3.6, Final calculation of ANNA value, i.e. calculation of the aggregation degree of gold particles:

[0104] ANNA = Do / De (11)

[0105] 3.7, Calculate the aggregation degree of gold particles ANNA of the sample with areal density value of 60-140 / μm2 obtained in step 2 according to steps 3.1-3.6.

[0106] 3.8, Samples with aggregation degree ANNA value between 0.6-1.4 are identified as valid samples, and finally 50 valid samples are selected for subsequent scientific analysis.

[0107] 4, Use of ANNA method of colloidal gold labeling by electron microscopy

[0108] Based on the areal density of gold particles obtained in step 2 and the aggregation degree of gold particles obtained in step 3, whether the areal density is within the range of 60-140 / μm2 is used to determine whether the data of colloidal gold labeling by electron microscopy is suitable for step 3 average nearest neighbor analysis, then calculate and obtain ANNA value, using ANNA value whether in 0.6-1.4 between the colloidal gold labeling by electron microscopy to determine whether the colloidal gold labeling by electron microscopy experiment is a qualified positive marker, and whether the target protein identified by the gold particles is distributed in the form of aggregation in the vesicle, or is distributed in the form of dispersion in the organelle matrix, or both distribution patterns.

[0109] Compared with other scenarios using ANNA method (such as ANNA method used in quantitative geography), the application of ANNA method in colloidal gold labeling by electron microscopy has its uniqueness, i.e. it is necessary to use the method for colloidal gold labeling by electron microscopy to exclude the interference of false positive marker data. The interference of false positive data on the reliability of ANNA value in quantitative geography is generally ignored. This method uses the high and low of areal density parameter and the interval of ANNA value to determine whether the colloidal gold labeling by electron microscopy is a positive marker. Areal density (unit: pieces / μm2) value higher than 30 is recognized as positive, lower than 30 is false positive, and higher than 60 is typical positive. The results of pre-experiment show that the ANNA value of typical positive data group is between 0.6-1.4, while the ANNA value of false positive data group is higher than 1.7. The ANNA method applied to colloidal gold labeling by electron microscopy is only used in data group with areal density higher than 60 and ANNA value between 0.6-1.4; data group with areal density lower than 60 and ANNA value higher than 1.7 is not trusted and discarded.

[0110] In the electron microscope colloidal gold labeling, the ANNA method can be used to quantitatively analyze the aggregation degree of the target object, and the labeling efficiency (represented by the surface density) of the gold particles of the label can be artificially adjusted to a range suitable for reflecting the ANNA effect. The surface density (unit: pieces per square micrometer) of 60-140 is a range suitable for reflecting the ANNA effect. The surface density lower than 30 is false positive, and the data is not reliable; the surface density higher than 140 is too dense, and the ANNA has a tendency to deviate from aggregation to randomness, causing false appearance. Therefore, in the operation of the electron microscope colloidal gold labeling, the concentration of the antibody and the labeling conditions are controlled to make the surface density of the labeling in the range of 60-140, so as to better reflect the effect of the ANNA analysis.

[0111] In summary, in the electron microscope colloidal gold labeling process, the ANNA method can ensure true positive, and the applicable range of the gold particle concentration (represented by the parameter surface density) is 60-140 pieces per square micrometer. This parameter can not only exclude the interference of false positive on ANNA calculation, but also avoid the situation that the true aggregation is covered due to too high labeling positive rate and too dense gold particles, which is reflected as the ANNA value representing the aggregation degree deviating to the dispersion direction.

[0112] Experiment:

[0113] This experiment is to collect data by using proteins with known aggregation / dispersion modes to verify the method of applying the ANNA method to the quantitative analysis of the aggregation degree of the electron microscope colloidal gold labeling of the present application. The target objects related to vesicles (having the property of aggregation) or not related to vesicles (not having the property of aggregation) are selected for gold labeling, and the quantitative data of the aggregation degree are obtained by calculating the aggregation degree of the target objects by using the method of the electron microscope colloidal gold labeling of the present application. The experimental strategy is shown in Figure 4 The target objects related to vesicles include: endoplasmic reticulum localization signal tetraamino acid short peptide (named HDEL), protein sterol methyltransferase (named SMT1) for controlling the formation of vesicles by membrane bending (the electron microscope colloidal gold labeling result is shown in Figure 5 , and the gold particles are in an aggregated distribution state), and biomembrane-related cytoskeleton microfilament protein (named actin). The target objects not related to vesicles include: tomato spotted wilt virus TSWV inclusion protein (named Nss) (the electron microscope colloidal gold labeling result is shown in Figure 6, gold particles are in discrete distribution state). The target double-stranded RNA (named J2) related to both vesicles and non-vesicles, which exists in both vesicle localization (TSWV is an RNA virus, its nucleic acid replication needs to be completed in vesicles, so the virus produces double-stranded RNA intermediates in vesicles) and non-vesicle localization (host-derived double-stranded RNA). In addition, two completely non-specific negative controls are set: TSWV-encoded proteins named Gc and Gn (which are not localized in vesicles and also not localized in fibrous inclusions, and there is no positive labeling on the target area of fibrous material, only non-specific false positive labeling) are used to verify whether the ANNA method has no discrimination with positive labeling when the number of non-specific labeling gold particles is small. Double-stranded RNA (J2) is labeled at two concentrations to achieve a significantly different surface density. One label is called J2-dilute30, and the surface density of electron microscopy colloidal gold labeling is 220±120; one label is called J2-dilute50, and the surface density of electron microscopy colloidal gold labeling is 100±40. A total of 7 groups of data, 50 pictures were taken repeatedly for each group, and the gold particle aggregation degree analysis was performed by using the electron microscopy colloidal gold labeling aggregation degree quantification analysis method of the present application, and the aggregation degree data of HDEL, SMT1, Nss, J2-dilute30, J2-dilute50, Gc and Gn were obtained respectively, and a violin plot was drawn as shown in Figure 7 , wherein the electron microscopy colloidal gold labeling of SMT1 is the data of typical gold particle aggregation distribution; and the electron microscopy colloidal gold labeling of Nss is the data of typical gold particle discrete distribution as shown in Figure 7 and Figure 8 . The HDEL protein and SMT1 protein distributed in vesicles have ANNA values significantly less than 1, indicating that they have a strong aggregation tendency. The aggregation degree of SMT1 protein is more than that of HDEL protein. The gold particle ANNA value representing the position of SMT1 protein shows a multi-cluster phenomenon, indicating the diversity of its aggregation degree, which can be divided into three obvious types. Nss distributed on fibers has an ANNA value significantly greater than 1, indicating that it has a strong discrete tendency. Double-stranded RNA (J2-dilute50) has two distribution modes in theory, virus-derived double-stranded RNA may be distributed in vesicles, and host-derived double-stranded RNA is not distributed in vesicles. The ANNA result shows that its value crosses 1, showing two obvious clusters, both aggregation and dispersion, suggesting that it is composed of two types with different distribution modes Figure 7 . The proteins Gn and Gc labeled with electron microscopy colloidal gold on the target fibrous material show false positive, and the median of their ANNA average value is higher than 1.8, and the peak shape in the violin plot is abnormal Figure 8The results from the Gn and Gc groups are highly distinguishable from true positive results and can be easily screened out and discarded. This is a unique treatment in electron microscopy colloidal gold labeling where the ANNA method is applied compared to other fields. HDEL and SMT1, distributed in vesicles, have ANNA values ​​significantly less than 1, indicating a strong aggregation tendency; Nss, distributed on fibers, has ANNA values ​​significantly greater than 1, indicating a strong dispersion tendency; double-stranded RNA (J2-dilu50) from different sources exhibits two distribution patterns, with ANNA values ​​spanning across 1, forming two distinct clusters in the aggregation and dispersion regions. Figure 7 (The area of ​​expansion in group J2-dilu50 in the violin diagram).

[0114] The effect of different areal densities of colloidal gold-labeled particles on the ANNA analysis of this invention was studied using different dilution concentrations of J2 double-stranded RNA antibody. J2-dilu50, diluted 50-fold with the antibody, achieved a gold particle areal density of 100±40, with a median ANNA value slightly less than 1.0. However, apart from antibody concentration, the other identically labeled J2-dilu30 (antibody diluted 30-fold) had a gold particle areal density of 220±120, with a median ANNA value slightly greater than 1.0. This indicates that as the antibody concentration increases, the gold label positivity rate increases, and the areal density of the gold particles increases, causing the ANNA values ​​to shift towards a more discrete direction. Figure 7 (J2-dilu50 group and J2-dilu30 group). The positive rate achieved by colloidal gold labeling under electron microscopy is more suitable for ANNA analysis in this invention when the gold particle areal density is 100±40.

[0115] Finally, it should be noted that the above examples are merely some specific embodiments of the present invention. Obviously, the present invention is not limited to the above embodiments and many variations are possible. All variations that can be directly derived or conceived by those skilled in the art from the disclosure of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A method for quantitative analysis of the aggregation degree of colloidal gold labels using electron microscopy, characterized in that... The specific process includes acquiring transmission electron microscopy (TEM) images of colloidal gold labels, preprocessing these images in a graphics workstation, using a parameter-dependent semi-automatic segmentation method for the preprocessed images to segment gold particles, manually selecting regions of interest (ROIs), and calculating the areal density of the actual gold particles within the ROIs. Colloidal gold labels with an ROI between 60 and 140 are considered suitable for subsequent mean nearest neighbor (ANN) analysis. The aggregation degree of the gold particles is then calculated using ANN analysis, and finally, colloidal gold labels with ANN values ​​between 0.6 and 1.4 are considered qualified positive labels. The process for calculating the degree of aggregation of gold particles is as follows: (1) Based on the distance threshold, sort out the gold particle models distributed in the inner region of interest (ROI) from all real colloidal gold particle models (gold total), including the spot close to ROI inner model and the spot far to ROI inner model; The innermost region of interest (ROI) is displayed as the spot close to the innermost model and is used as the gold particle object calculated by the mean nearest neighbor analysis (NNN) method. The outermost gold particle model is connected to form a closed polyline region, which is used as the calculation region of the mean nearest neighbor analysis (ROI) ANNA. (2) Obtain the number of voxel parameters of the ROI ANNA for the mean nearest neighbor analysis method; (3) Calculation of the area of ​​the ROI ANNA using the mean nearest neighbor analysis method: area of ​​ANNA=number of voxel×voxel size(9) (4) Calculation of the expected value De in regression analysis using the mean nearest neighbor analysis method: ； （10） Where Nr gold represents the number of gold particles in the ROI ANNA calculated by the average nearest neighbor analysis method; (5) The actual spacing value of gold particles in the mean nearest neighbor analysis method The measurement was performed as follows: the distance between each gold particle model and every other gold particle model in the ROI ANNA of the mean nearest neighbor analysis was measured sequentially, resulting in data sets Distance1, Distance2...Distance. n The data groups Distance1, Distance2...Distance are respectively... n The minimum value in the interval is denoted as the minimum distance. 1、 Distance minimum2……Distance minimum n Finally, the minimum spacing (Distance) is calculated. 1、 Distance minimum2……Distance minimum n The average value is used as the actual spacing value. ; (6) Calculate the aggregation degree parameters of gold particles: （11） (7) Samples with ANNA values ​​between 0.6 and 1.4 are considered valid samples.

2. The method for quantitative analysis of the aggregation degree of colloidal gold labels using electron microscopy according to claim 1, characterized in that: The preprocessing includes normalizing the brightness and contrast of the electron microscope colloidal gold-labeled image and performing image phase inversion processing.

3. The method for quantitative analysis of the aggregation degree of colloidal gold labels using electron microscopy according to claim 2, characterized in that: The specific process of the semi-automatic gold particle segmentation is as follows: The real gold particles and spurious signals to be segmented were filtered according to the size factor parameter to obtain the total gold particle model of all real colloidal gold particles in the electron microscope image. The three sets of filters are the total size and signal peak value, the sharpness of the gold particle boundary with the background, and the signal peak value.

4. The method for quantitative analysis of the aggregation degree of colloidal gold labels using electron microscopy according to claim 3, characterized in that: The region of interest includes an inner region of interest (ROI inner) and an outwardly extending ring region from the inner region of interest (ROI inner). The outwardly extending ring region is the quality control region (ROI peripheral). The sum of the inner region of interest (ROI inner) and the quality control region (ROI peripheral) is the outer region of interest (ROI outer).

5. The method for quantitative analysis of the aggregation degree of colloidal gold labels using electron microscopy according to claim 4, characterized in that: The calculation process for the areal density of the real gold particles is as follows: Obtain the inner number of voxels, the outer number of voxels, the voxel size, the number of gold particles in the inner region of interest, and the number of gold particles in the outer region of interest from the electron microscope colloidal gold labeled image. The voxel size includes voxel X and voxel Y. Then, perform the following calculations sequentially: (1) Area of ​​the inner region of interest (ROI): area of ​​inner=inner Number of voxel×voxel X×voxel Y(1) (2) Area of ​​the outer region of interest (ROI): area of ​​outer=outer Number of voxel×voxel X×voxel Y(2) (3) Area of ​​the quality control area (ROI) peripheral: area of ​​peripheral=area of ​​outer-area of ​​inner (3) (4) Based on the distance threshold, sort out the gold particle models distributed in the outer region of interest (ROIouter) from all the real colloidal gold particle models, including spot close to ROI outer and spot far to ROI outer models; (5) The number of real gold particle models in the quality control area (ROI) peripheral: number of gold peripheral=number of gold outer - number of gold inner (4) (6) Areal density of real gold particles in the inner region of interest (ROI): areal density inner=number of gold inner / area of ​​inner(5) (7) Areal density of real gold particles in the quality control area (ROI peripheral): areal density peripheral=number of gold peripheral / area of ​​peripheral (6) (8) Calculation of background subtraction parameters: ratio peripheral / inner=areal density peripheral / areal density inner (7) (9) The areal density of the real gold particle model after removing the background: areal density= areal density inner- areal density peripheral (8) (10) Samples with an area density greater than 30 gold particles per square micrometer were identified as positive. Samples with an area density of 60-140 gold particles per square micrometer were selected for the calculation of the aggregation degree of the gold particles.

Images