PAM staining method and image data

Abstract

This invention provides a PAM staining method that enables staining suitable for magnified observation of various samples using a range of microscopes, from optical microscopes to electron microscopes. [Solution] A PAM staining method comprising the steps of: silvering a biological sample; coloring the biological sample; and fixing the color of the biological sample, wherein the biological sample is a biologically derived sample that has been oxidized; the silvering step involves immersing the biological sample in a methenamine silver aqueous solution at room temperature, raising the methenamine silver aqueous solution from room temperature to 62°C or higher and 68°C or lower, and carrying out the silvering reaction for 30 to 40 minutes from the start of the reaction at room temperature; the coloring step involves immersing the silvered biological sample in a 0.2% gold chloride aqueous solution at room temperature for 8 to 12 hours; and the fixing step involves immersing the sample in a hard film fixing solution containing 11% to 13% sodium thiosulfate and aluminum potassium sulfate dodecahydrate for 3 minutes or more at room temperature.

Description

Technical Field

[0001] The present invention relates to the PAM staining method and image data. Specifically, it relates to the PAM staining method that enables staining capable of observing the structures of living organisms and microorganisms with high resolution, and image data obtained by imaging a sample stained by the PAM staining method.

Background Art

[0002] Conventionally, in order to visualize the structures of living tissues and cells, and further microorganisms in vivo, techniques for staining living tissues and cells are known. The stained sample is observed under magnification using a microscope (such as an optical microscope, fluorescence microscope, electron microscope, etc.).

[0003] For example, as a staining method used for staining the basement membrane in kidney tissue, there is the periodic acid methenamine silver staining (hereinafter, PAM staining) method using the PAM staining method. In the PAM staining method, a section of kidney tissue oxidized with periodic acid is stained with methenamine silver to make it easier to observe the cross-sectional shape of the kidney basement membrane (for example, the glomerular basement membrane, Bowman's capsule basement membrane, renal tubule basement membrane, etc.) under an optical microscope. For example, in the glomerular tissue stained by the PAM staining method, it becomes easier to determine lesions of the kidney basement membrane such as membrane duplication and spike formation. Therefore, the PAM staining method plays an important role in the pathological diagnosis of kidney diseases. Also, as a staining method for fungi, the Grocott staining method is known, and as a staining method for acid-fast bacteria such as Mycobacterium tuberculosis, the Ziehl-Neelsen staining method is known.

[0004] Since the original method was invented in 1953, improvements aimed at speeding up and clarifying the staining have been devised for the PAM staining method. As an example of such an improved PAM staining method, for example, there is the PAM staining method (TSC-PAM staining method) described in Non-Patent Document 1. In the PAM staining method described in Non-Patent Document 1, it is said that the kidney basement membrane can be stained more quickly and clearly by reacting thiosemicarbazide after oxidizing a section of the kidney basement membrane with periodic acid. Also, the Grocott staining method of Non-Patent Document 2 and the Ziehl-Neelsen staining method based on Non-Patent Document 3 are useful for pathologically diagnosing microbial infections in tissues. [Prior art documents] [Non-patent literature]

[0005] [Non-Patent Document 1] Thiosemicarbazide used after periodic acid makes methenamine silver staining of renal glomerular basement membranes faster and cleaner. Hayashi I, Tome Y, Shimosato Y. Stain Technol. 1989 Jul;64(4):185-90. [Non-Patent Document 2] A stain for fungi in tissue sections and smears using Gomori's methanamine-silver nitrate technique. Grocott RG, Am. J. Clin. Pathol. 25: 975-979, 1955 [Non-Patent Document 3] F. Ziehl, "Zur Farbung des Tuberkelbacillus" (On the Staining of Tubercle Bacilli), Medizinische Microbiologie, 1882. [Overview of the project] [Problems that the invention aims to solve]

[0006] However, conventional PAM staining methods required different reagents and staining conditions depending on the sample being stained. Furthermore, while Grocott staining can selectively stain fungal cells, it requires the use of highly toxic chromic acid as an oxidizing agent, posing a problem in reagent management. Ziehl-Neelsen staining can selectively stain acid-fast bacteria, but its staining ability is poor for cells and surrounding tissues, making it particularly difficult to determine their localization within extracellular tissues.

[0007] For example, the PAM staining method described in Non-Patent Literature 1 is not suitable for staining thick (e.g., 3 μm or more) sections of the renal basement membrane, as it becomes difficult to observe the shape of the renal basement membrane when the thickness exceeds 3 μm. Therefore, in order to observe the renal basement membrane using the conventional PAM staining method, it was necessary to prepare sections of the renal basement membrane with a thickness of approximately 1 μm.

[0008] The preparation of 1 μm thick kidney basement membrane sections requires specialized tissue fixation methods and highly skilled technicians, posing challenges in terms of high costs and securing personnel.

[0009] Furthermore, the Grocott staining method (A stain for fungi in tissue sections and smears using Gomori's methanamine-silver nitrate technic. Grocott RG, Am. J. Clin. Pathol. 25: 975-979, 1955) is known as a method for staining fungi. Although this staining method can selectively stain fungal cells, it has the problem of requiring the use of highly toxic chromic acid as an oxidizing agent, thus necessitating careful reagent management.

[0010] Furthermore, the Ziehl-Neelsen staining method is known for staining acid-fast bacteria such as Mycobacterium tuberculosis. Although this staining method can selectively stain acid-fast bacteria, it has poor staining ability for cells and surrounding tissues, making it particularly difficult to determine their localization within extracellular tissues.

[0011] Based on the above circumstances, the present invention aims to provide a PAM staining method that enables staining suitable for magnified observation of various samples using a variety of microscopes, from optical microscopes to electron microscopes. [Means for solving the problem]

[0012] To solve the above problems, one aspect of the present invention includes the following:

[0013] Unless otherwise specified, the concentrations of solutions, etc., in this application are given as mass percentage concentrations.

[0014] [1] A PAM staining method comprising the steps of: silvering a biological sample; coloring the biological sample; and fixing the color of the biological sample, wherein the biological sample is a biologically derived sample that has been oxidized; the silvering step involves immersing the biological sample in a methenamine silver aqueous solution at room temperature, raising the methenamine silver aqueous solution from room temperature to 62°C or higher and 68°C or lower, and allowing the reaction to proceed for 30 to 40 minutes from the start of the reaction at room temperature; the coloring step involves immersing the silvered biological sample in a 0.2% gold chloride aqueous solution at room temperature for 8 to 12 hours; and the fixing step involves immersing the sample in a hard film fixing solution containing 11% to 13% sodium thiosulfate and aluminum potassium sulfate dodecahydrate for 3 minutes or more at room temperature.

[0015] [2] The PAM staining method according to [1], wherein the biological sample is a section sample cut from living tissue, and the section sample is obtained by oxidizing a thin section of the living tissue having a thickness of more than 1 μm.

[0016] [3] The PAM staining method according to [2], which includes a step of preparing the thin sections with a thickness of more than 1 μm and 30 μm or less, prior to the silvering step.

[0017] [4] A PAM staining method according to any one of [1] to [3], wherein the coloring step further involves immersing the biological sample, after immersing it in the gold chloride aqueous solution, in an oxalic acid aqueous solution.

[0018] [5] A PAM staining method according to any one of [1] to [4], comprising the step of oxidizing the biological sample with 0.5% periodic acid for 10 minutes prior to the silvering step.

[0019] [6] The PAM staining method according to any one of [1] to [5], wherein the methenamine silver aqueous solution consists of a mixture of 1.5% methenamine, 5% silver nitrate, and 5% borax, with a volume ratio of 100:5:5.

[0020] [7][1] to [6] Any one of the biological samples stained by the PAM staining method, which is image data obtained by imaging, and includes first image data corresponding to a first image imaged using an optical microscope, and a second image imaged using an electron microscope and corresponding to a second image with a higher magnification than the first image, and the second image is an enlarged image of a specific position in the first image.

[0021] [8] The first image data is the image data according to [7], which is the training data of deep learning performed in a deep learning processor.

Advantages of the Invention

[0022] According to the present invention, it is possible to provide a PAM staining method that enables staining suitable for magnified observation with various microscopes from an optical microscope to an electron microscope for various samples.

Brief Description of the Drawings

[0023] [Figure 1] An optical micrograph of a sample (stained section sample) stained by the conventional TSC-PAM staining method. [Figure 2] An optical micrograph of a sample (stained section sample) stained by the PAM staining method of the present embodiment. [Figure 3] An electron micrograph of a sample (stained section sample) stained by the conventional TSC-PAM staining method. [Figure 4] An electron micrograph of a sample (stained section sample) stained by the PAM staining method of the present embodiment. [Figure 5] An electron micrograph of a sample (stained section sample) stained by the conventional TSC-PAM staining method. [Figure 6] An electron micrograph of a sample (stained section sample) stained by the PAM staining method of the present embodiment. [Figure 7] A graph showing the ratio of GBM lesions judged using AI from the image data. [Figure 8] This graph shows the relationship between the COL4α5 preservation rate and the proportion of GBM lesions detected by AI. [Figure 9] This graph shows the relationship between urinary albumin / urinary creatinine concentration and the ratio of GBM lesions detected by AI. [Figure 10] This is an electron microscope image of a cell-derived organoid (stained biological sample) stained using the PAM staining method of this embodiment. [Figure 11] This is an optical microscope image of Mycobacterium tuberculosis (stained biological sample) stained using the PAM staining method of this embodiment. [Figure 12] This is an optical microscope image of a fungus (stained biological sample) stained using the PAM staining method of this embodiment. [Figure 13] This is an electron microscope image of Mycobacterium tuberculosis (stained biological sample) stained using the PAM staining method of this embodiment. [Figure 14] This is an electron microscope image of a fungus (stained biological sample) stained using the PAM staining method of this embodiment. [Modes for carrying out the invention]

[0024] One embodiment of the present invention will be described below. The embodiments shown below illustrate configurations for realizing the technical idea of ​​the present invention, and the present invention is not limited to these embodiments. Various modifications can be made to the technical idea of ​​the present invention within the technical scope defined by the claims described in the claims.

[0025] (PAM staining method) The PAM staining method in this embodiment includes the steps of oxidizing the biological sample, silver plating the biological sample, coloring the silver-plated biological sample, and fixing the color of the biological sample. This allows for suitable staining of the biological sample.

[0026] In this embodiment, the biological sample is a sample of biological origin. For example, it may be a microorganism such as fungi or acid-fast bacteria, or it may be a section cut from a living tissue or organoid derived from cells. Alternatively, it may be a section cut from living tissue containing microorganisms, or a suspension containing cells. In this specification, when the biological sample to be observed is a section cut from living tissue, the cut section is referred to as a "thin section," and the thin section that has undergone pre-treatment before staining is referred to as a "section sample."

[0027] (Sample cutting) In this embodiment, when the biological sample is a section, the thin section used is cut from living tissue. Living tissue includes, for example, kidney tissue, or organoids derived from cells, and living tissue containing microorganisms. Known methods can be used to cut the thin section. The thickness of the thin section to be cut is preferably greater than 1 μm and 30 μm or less, more preferably greater than 1 μm and 10 μm or less, and even more preferably between 3 μm and 5 μm.

[0028] In prior art, when the thickness of the thin section exceeded 1 μm, excessive binding of silver particles occurred during staining, resulting in poor contrast and making it difficult to observe the basement membrane. On the other hand, when attempting to prepare thin sections less than 1 μm thick, there was a risk that the sections would become unsuitable for diagnosis due to tissue crushing, and preparation required skilled techniques. With the PAM staining method of this embodiment, thin sections with a thickness greater than 1 μm but less than 30 μm can be used for staining, thus greatly simplifying the preparation of thin sections.

[0029] (Oxidation treatment of biological samples) Next, the biological sample is immersed in periodic acid for oxidation treatment. The concentration of periodic acid is preferably 0.2% to 0.8%, more preferably 0.4% to 0.6%. The treatment time is preferably 7 minutes to 13 minutes, more preferably 9 minutes to 11 minutes, and even more preferably 10 minutes. The oxidation treatment is preferably carried out at room temperature. By performing the oxidation treatment, the polysaccharides on the surface (basement membrane surface) of the biological sample contained in the thin section are oxidized to produce aldehydes.

[0030] The aldehyde produced by the oxidation treatment is necessary for the reaction with methenamine silver in the silver plating process described later; therefore, the oxidation treatment of the biological sample is carried out prior to the silver plating reaction.

[0031] (Sensitization reaction) Next, the oxidized biological sample is washed with running water and then with distilled water, and then sensitized by immersing it in a 0.5% thiosemicarbadite aqueous solution at room temperature. The immersion time is preferably 3 to 7 minutes, and more preferably 4 to 6 minutes. Performing the sensitization reaction shortens the time required for the subsequent silvering reaction.

[0032] (silver plated) Next, in the step of silvering the sample (biological sample) that has undergone the sensitization reaction, the biological sample is immersed in an aqueous solution of methenamine silver and the silvering reaction is carried out. The reaction temperature for the silvering reaction starts at room temperature, and is gradually increased until it reaches 62°C to 68°C, and then maintained at 62°C to 68°C for a predetermined time. The total time for the silvering reaction, from the start to the end of the reaction at room temperature, is preferably 30 minutes to 40 minutes. As an example, the above reaction can be carried out in a commercially available incubator or water bath with temperature control.

[0033] The methenamine silver aqueous solution in this embodiment consists of a mixture of 1.5% methenamine, 5% silver nitrate, and 5% borax, with a volume ratio of 100:5:5. The methenamine silver aqueous solution can be prepared, as an example, with the following composition.

[0034] Mix 20 ml of 3% methenamine solution (Merck), 2 ml of 5% silver nitrate solution (Fujifilm Wako Pure Chemical Industries), 2 ml of 5% borax solution, and 20 ml of distilled water in this order. When the 5% silver nitrate solution is added to the 3% methenamine solution, a white precipitate will form; mix thoroughly until this precipitate disappears. The 5% borax solution should be added just before using the methenamine silver solution for the silver plating reaction. The distilled water should be added after the methenamine solution, silver nitrate solution, and borax solution have been thoroughly mixed.

[0035] In this process, the aldehyde produced in the oxidation process reduces methenamine silver, and the resulting silver particles are deposited on the surface of the biological sample. For example, if the biological sample is kidney tissue, the silver particles will be deposited on the surface of the basement membrane.

[0036] In the prior art, the silvering reaction was carried out in a preheated incubator, which resulted in a fast reaction rate and difficulty in controlling the staining. Furthermore, in the prior art, it was necessary to precisely adjust the composition of the methenamine silver aqueous solution used in the silvering reaction in order to control the staining, making the process complicated.

[0037] In this embodiment, the silver plating reaction is started at room temperature, and the reaction temperature is gradually increased to between 62°C and 68°C. By maintaining this temperature throughout the reaction time, the staining process becomes easier to control. Furthermore, the composition ratio of the methenamine silver aqueous solution is simplified, making preparation easier.

[0038] Furthermore, in this embodiment, the control of staining has been made easier compared to the conventional technique, making it easier to suppress excessive binding of silver particles to the surface of biological samples when staining thicker biological samples compared to the conventional technique. As a result, it has become easier to apply staining suitable for microscopic observation to thicker biological samples compared to the conventional technique.

[0039] (Toning) In the process of coloring biological samples that have undergone a silver plating reaction, the biological samples are immersed in a 0.2% gold chloride aqueous solution. This causes gold particles to precipitate and deposit on the surface of the biological samples. The immersion temperature is preferably room temperature. The immersion time is preferably 8 to 12 hours. Compared to conventional techniques, by reacting with gold chloride for a sufficient amount of time, the amount of gold particles deposited on the biological samples can be increased, improving the contrast of the biological samples under both optical and electron microscopes.

[0040] If there is an excess of gold particle binding, the biological sample may be further immersed in an oxalic acid solution after being immersed in a gold chloride solution. The oxalic acid solution used can be concentrated to, for example, 1% to 3%. The immersion time can be 1 to 5 minutes. By immersing the biological sample in an oxalic acid solution, the excess silver and gold particles can be removed from the biological sample.

[0041] (Establishment) The color-tuned biological sample is immersed in a hardening solution of the following composition to remove unreacted silver and gold ions and perform a fixing reaction to enhance the clarity of the stain. A commercially available photographic acidic hardening solution can be used as the hardening solution. The concentration of the photographic acidic hardening solution is preferably a 2:1 dilution of the stock solution (2.5 times the manufacturer's recommended concentration). The immersion time is preferably 3 minutes or more, and more preferably 5 minutes. The reaction temperature is preferably room temperature. By increasing the concentration of the photographic acidic hardening solution compared to conventional techniques, unreacted silver and gold ions can be removed more reliably. This is expected to improve the clarity of the stain.

[0042] (Composition of fixing solution) The dural fixative contains at least five compounds: sodium thiosulfate pentahydrate, sodium sulfite, acetic acid, boric acid, and aluminum potassium sulfate dodecahydrate. The concentration of sodium thiosulfate is preferably 11% to 13%, and more preferably 11.5% to 12.5%. The concentrations of the five compounds are, for example, 12% sodium thiosulfate pentahydrate, 0.75% sodium sulfite, 0.65% acetic acid, 0.4% boric acid, and 0.75% aluminum sulfate dodecahydrate.

[0043] The PAM staining method described above allows for the appropriate staining of thin sections that are thicker than those considered appropriate for conventional staining methods. Furthermore, it enables safer staining of biological samples that previously required highly toxic chemicals. In addition, it improves the visibility during observation of biological samples that previously had poor staining properties for cells and surrounding tissues, making it particularly difficult to determine their localization within extracellular tissues (method for producing stained biological samples).

[0044] Using the resulting stained biological sample, the biological sample can be observed under magnification as follows.

[0045] (Creating image data) The stained biological sample obtained by the method described above can be used in both optical and electron microscopes. The image data of this embodiment is image data obtained by imaging a biological sample stained by the PAM staining method described above, and comprises a first image data corresponding to a first image taken using an optical microscope, and a second image data corresponding to a second image taken using an electron microscope with a higher magnification than the first image. The second image is image data that is a magnified image of a specific position in the first image.

[0046] Conventional staining techniques are not suitable for staining thick samples (e.g., 3 μm or more). Therefore, samples used in conventional techniques are typically prepared to a thickness of approximately 1 μm, making them unsuitable for observing the three-dimensional structure of the basement membrane. Consequently, to perform magnified observation using an electron microscope, it was necessary to prepare a separate sample specifically for electron microscopy from the one used for optical microscopy.

[0047] Consequently, when observing lesions using stained biological samples obtained with conventional staining methods, even if a suspected lesion was found through magnified observation with an optical microscope, it was necessary to use a different sample for magnified observation with an electron microscope than the one observed with the optical microscope. Therefore, it was difficult to compare optical microscope findings with electron microscope findings.

[0048] On the other hand, this staining method allows for the staining of thick samples, thus enabling optical microscope samples to have the necessary thickness for observing their three-dimensional structure. Therefore, with this method, optical microscope samples can be used directly as electron microscope samples.

[0049] As a result, when kidney tissue is stained and observed using the staining method of this embodiment, if tissue suspected to be a lesion is found during observation with an optical microscope, the same stained sample can be used for magnified observation with an electron microscope. Therefore, by using a sample stained with the staining method of this embodiment, it is possible to compare the findings of the optical microscope and the electron microscope, allowing for a more accurate diagnosis of the lesion.

[0050] Furthermore, by accumulating comparative results between optical microscope findings and electron microscope findings, it becomes possible to more accurately diagnose lesions based on observations using an optical microscope.

[0051] The electron microscope used for imaging may be a scanning electron microscope or a transmission electron microscope. Compared to a transmission electron microscope, which requires specialized resin embedding and ultrathin sectioning techniques, it is preferable to use a scanning electron microscope that allows direct observation of the stained sample under low vacuum conditions.

[0052] Images captured using an optical microscope correspond to the "first image" in this invention, and the image data of the first image corresponds to the "first image data" in this invention. Furthermore, images captured using an electron microscope correspond to the "second image" in this invention, and the image data of the second image corresponds to the "second image data" in this invention. The second image has a higher magnification than the first image and is a magnified image of a specific location in the first image.

[0053] This makes it possible to observe the same area using both an optical microscope and an electron microscope, which was difficult with conventional methods, and also simplifies the sample preparation process.

[0054] Image data of a sample can be generated by imaging a sample stained using this method. When imaging a sample stained using this method, either an optical microscope or an electron microscope may be used. Known imaging techniques may be used to image a sample stained using this method.

[0055] (Deep learning) Of the image data of a sample stained using this method, the first image data can be used as training data for deep learning performed in an AI-based deep learning processor. Deep learning is supervised learning using a multi-layered neural network. By using the first image data of a sample stained using this method as training data for AI-based deep learning, and using the physician's diagnosis results using the second image for specific locations in the first image as training data, the AI ​​can be trained to automatically distinguish lesions from the image data of the sample.

[0056] Automatic differentiation of lesions by AI means, for example, that the AI ​​automatically distinguishes between abnormal GBM regions, normal GBM regions, and other regions within the glomerular region of kidney tissue shown in an image. By using image data of samples stained with this method for AI deep learning, it is believed that the AI ​​will be able to automatically diagnose diseases from image data of samples stained with this method. [Examples]

[0057] The present invention will be described in detail with reference to the following examples. The present invention is not limited to the following examples.

[0058] (Example 1) (Sample cutting) The kidneys of Alport mice (female, 22 weeks old) were used as the sample material. Formalin-fixed paraffin-embedded blocks were prepared by fixing the kidneys with 10% neutral buffered formalin and embedding them in paraffin. Samples (thin sections) were prepared by cutting 3 μm thick sections from the formalin-fixed paraffin-embedded blocks.

[0059] (oxidation process) The cut sample was deparaffinized using xylene, then dexylened again using ethanol, and immersed in distilled water. The ethanol used may contain water if it is possible to remove the xylene impregnated in the sample by the deparaffinization process. A high concentration of ethanol is preferable, and pure ethanol (100% ethanol) is more preferable, as it facilitates dexylened. The sample thus prepared was immersed in a 0.5% periodic acid aqueous solution at room temperature for 10 minutes to carry out the oxidation reaction and obtain sectioned samples.

[0060] (Sensitization process) After washing the oxidized section samples with distilled water, they were immersed in a 0.5% thiosemicarbadite aqueous solution at room temperature for 5 minutes to perform the sensitization reaction.

[0061] (Silver plating process) The silver plating process was carried out by the following method. The sample (section sample) that underwent the sensitization reaction was thoroughly washed with running water, then washed again with distilled water, and then immersed in a methenamine silver aqueous solution at room temperature (20°C to 28°C) and placed in an incubator. The temperature of the incubator was gradually increased from room temperature (20°C to 28°C) to between 62°C and 68°C. The reaction time was set to be between 30 and 40 minutes from the start of the gradual temperature increase from room temperature. The methenamine silver aqueous solution was prepared by the following method.

[0062] (Preparation of methenamine silver aqueous solution) 20 ml of 3% methenamine solution (Merck), 2 ml of 5% silver nitrate solution (Fujifilm Wako Pure Chemical Industries), 2 ml of 5% borax aqueous solution, and 20 ml of distilled water were mixed in this order. When the 5% silver nitrate solution was added to the 3% methenamine solution, a white precipitate formed, but the mixture was thoroughly mixed until this precipitate disappeared. The 5% borax aqueous solution was added immediately before using the methenamine silver solution for the silver plating reaction.

[0063] (Toning process) After observing under a light microscope to confirm that the basement membrane had been silver-plated and turned dark brown, the silver-plated sample was washed with distilled water. Next, the sample was color-tuned by immersion in a 0.2% chloroauric acid solution at room temperature overnight.

[0064] (Fixing process) The color-tuned samples were immersed in a photographic acidic hardening fixative (manufactured by Muto Chemical Co., Ltd.) at room temperature for 5 minutes to fix the color. The fixative was diluted to 2 times the original concentration. By following these steps, staining was performed in this example, and stained section samples were obtained.

[0065] (Observation under an optical microscope) A 3 μm thick sample stained using the conventional TSC-PAM staining method and a sample of the same thickness stained using the present method were observed and compared under an optical microscope. The specific procedure for the TSC-PAM staining method will be described below. Figure 1 is an optical microscope image of a sample stained using the TSC-PAM staining method (stained section sample), and Figure 2 is an optical microscope image of a sample stained using the present method (stained section sample). The scale bar is 10 μm. Figure 2 corresponds to the first image in the present invention, and the image data in Figure 2 corresponds to the first image data in the present invention.

[0066] (TSC-PAM staining method) The conventional TSC-PAM staining method was performed in the same manner as the present method, except for the following: In the silvering step, the section samples were immersed in a methenamine silver aqueous solution preheated to 58°C, and the silvering reaction was carried out at 58°C for more than 2 minutes. This reaction was observed every 30 seconds thereafter, and stopped when the glomerular basement membrane turned brown by thoroughly washing the section samples with distilled water.

[0067] Next, in the coloring step, a coloring reaction was carried out in a 0.2% gold chloride aqueous solution at room temperature for 15 minutes. In the fixing step, a fixing reaction was carried out in a 2% sodium thiosulfate aqueous solution or a 5-fold diluted photographic acidic hardening fixer at room temperature for 1 minute. After washing the fixed section samples with running water, hematoxylin-eosin (HE) staining was performed as a post-staining step, followed by dehydration, clearing, and mounting with Marinol. Samples stained using the conventional TSC-PAM staining method were obtained in this manner.

[0068] For observation under an optical microscope, a SLIDEVIEW VS200 (EVIDENT) was used. The TSC-PAM staining method is optimized for staining samples with a section thickness of approximately 1 μm and observing the cross-sectional shape of the basement membrane from a vertical direction, so the tangential region of the basement membrane appears black. As a result, in samples with a section thickness of 3 μm, overlapping basement membranes were excessively stained, making it difficult to observe the cross-sectional and surface shape of the basement membrane due to poor contrast.

[0069] In contrast, this method results in a brownish tint in the tangential region of the basement membrane. Therefore, even with thicker section thicknesses compared to conventional techniques, overlapping basement membranes are not excessively stained, resulting in improved contrast and easier observation of the cross-sectional and surface shape of the basement membrane even with 3 μm section thicknesses. The ease of observing thicker section thicknesses compared to conventional techniques eliminates the need for 1 μm thick sample preparation, which requires advanced skills, and is expected to significantly simplify the preparation of kidney tissue samples.

[0070] (Observation under a low-vacuum scanning electron microscope) The same samples observed under an optical microscope were also observed and compared under a low-vacuum scanning electron microscope (LVSEM) (Miniscope TM4000PlusII (Hitachi High-Tech Corporation)). Figures 3 and 5 are electron microscope images of samples stained with the TSC-PAM staining method, and Figures 4 and 6 are electron microscope images of samples stained with this method. The sample in Figure 1 is the same as the samples in Figures 3 and 5, and the sample in Figure 2 is the same as the samples in Figures 4 and 6. The scale bars in Figures 3-6 are 10 μm.

[0071] As shown in Figure 5, it was difficult to observe lesions in samples stained with the TSC-PAM staining method. In contrast, as shown in Figure 6, the resolution of samples stained with this method improved under LVSEM observation, making it easier to observe lesions with LVSEM. Since the sample in Figure 6 is the same sample as those in Figures 2 and 4, which were stained with this method, it was confirmed that this method does not require the preparation of a separate sample, and that the same area of ​​the same sample observed under an optical microscope can be magnified and observed with LVSEM.

[0072] While attempts have been reported to observe samples stained with the conventional TSC-PAM staining method under electron microscopy, including LVSEM, the information obtainable with TSC-PAM staining was limited unless a transmission electron microscope was used. This new method improves resolution under LVSEM observation, enabling observation even with a smaller, less expensive LVSEM compared to a transmission electron microscope, thus simplifying and reducing the cost of electron microscopy.

[0073] (Image analysis using AI) To verify whether image data of samples stained using this method can be used for diagnosis using AI-based image analysis, the following procedure was performed.

[0074] (Preparing the sample) Using CRISPR / Cas9, we established an XLAS model mouse with a nonsense mutation (R471X) in COL4A5, the same mutation seen in human X-linked Alport syndrome (Hashikami K, Asahina M, Nozu K, Iijima K, Nagata M, Takeyama M: Establishment of X-linked Alport syndrome model mice with a Col4a5 R471X mutation. Biochem Biophys Rep 2019, 17:81-86.). While the above study used male mice that developed severe lesions similar to human males, this study uses female mice that developed mild lesions similar to human females. The animals were aged 22, 30, and 42 weeks for the XLAS model mice, and 42 weeks for the control wild-type mice. Female mice were euthanized upon reaching the aforementioned ages, and kidney tissue was collected and prepared as specimens.

[0075] (Preparation of the first image data) Sectional tissue samples were collected from the kidneys of model mice with X-linked Alport syndrome (XLAS) and wild-type mice, and stained using this method. The stained sectional samples were scanned at 40x magnification using Z-stack extended focus imaging (EFI) with a SLIDESHOW VS200 (EVIDENT) to prepare image data (first image data) of the samples.

[0076] The glomerular regions of the sample image data were classified using the HALO AI network (DenseNet V2 classifier). The HALO AI network is AI software that can be trained to identify user-specified features from loaded image data.

[0077] In this example, image data of the sample was used as training data. In addition, LVSEM observation image data was also referred to, and image data in which a human judged the glomerular basement membrane (GBM) shape to be one of three types: an abnormal GBM region, a normal GBM region, or other regions was used as training data. The AI ​​was trained using this training data and training data.

[0078] Using the trained HALO AI network described above, the glomerular region was identified as abnormal, normal, or other regions from the image data of the sample. The ratio of GBM lesions detected by the AI ​​was calculated as (area of ​​abnormal GBM region / area of ​​glomerular region) × 100. Figure 7 is a graph showing the ratio of GBM lesions determined by the AI ​​from the image data. Compared to wild-type mice, the ratio of GBM lesions was significantly higher in XLAS model mice at all ages.

[0079] (AI-based image analysis evaluation: urinary protein) To evaluate AI-based image analysis of samples, urine samples were collected from the same mice from which PAM-stained samples (stained section samples) were taken using this method, using the following procedure: Before sample collection, each mouse was placed in a separate metabolic cage, and samples were collected over 16 hours under conditions of 20-26°C and 40-70% humidity, with free access to water and food.

[0080] The urinary albumin and creatinine concentrations in the collected urine samples were measured using the LBIS Mouse Urinary Albumin Assay Kit and L-type Wako CRE-M, respectively, and the urinary albumin / creatinine concentration was calculated.

[0081] (Evaluation of image analysis by AI: COL4α5 preservation rate) To evaluate the AI-based image analysis of samples, the COL4α5 preservation rate in the same mice from which the PAM-stained samples were collected was calculated using the following method.

[0082] Fresh frozen kidneys were collected from the same mice from which PAM-stained samples were taken, and 6 μm thick sections were obtained from them using a cryostat CM3050S (Leica Biosystems).

[0083] The obtained sections were stained according to the manufacturer's protocol using a mixed monoclonal antibody (CFT45325; manufactured by Shigei Medical Research Institute) with fluorescein isothiocyanate (FITC) against COL4α5 (clones H53 and B51) and Texas Red labeled antibody against COL4α2 (clone H25). Since the anti-FITC sensitization process depends on the development of the fluorescent dye, staining was performed in a light-shielded wet box.

[0084] After washing the tissue sections with distilled water, they were treated with washing buffer and incubated with hydrogen peroxide for 5 minutes to remove endogenous peroxidase. Following the removal of endogenous peroxidase, the tissue sections were washed again with washing buffer and incubated with 200-fold diluted peroxidase-labeled anti-FITC rabbit polyronal antibody (Agilent Technologies) for 30 minutes.

[0085] After washing the sections with washing buffer, staining was performed using 3,3'-diaminobenzidine (DAB). The sections were reacted with methenamine silver solution prepared in the same manner as the TSC-PAM staining method described above, thoroughly washed with distilled water, and then immersed overnight in a 0.2% silver chloride aqueous solution. Subsequently, the samples were thoroughly washed with distilled water, treated with photographic fixer (manufactured by Muto Chemical Co., Ltd.) for 5 minutes, and rinsed with running water. A 2-fold dilution of the photographic fixer (2.5 times the manufacturer's recommended concentration) was used. Samples stained with the mixed monoclonal antibody were obtained as described above.

[0086] When samples stained with mixed monoclonal antibodies were analyzed using LVSEM with a Miniscope TM4000PlusII (Hitachi High-Tech Corporation), it was confirmed that the basement membrane in areas where COL4α5 remained showed a dense structure and was healthy, while areas where COL4α5 was deficient showed a lesion with a disrupted basement membrane structure (basket weave).

[0087] Samples stained with mixed monoclonal antibodies were converted to whole slide images using NanoZomer S60 (Hamamatsu Photonics). The obtained image data was used as training data, and data specifying glomeruli, non-glomeruli, and non-tissue regions was used as target data. A supervised machine learning algorithm (random forest) was trained using HALO software v3.6 (Indica Labs) to recognize glomeruli, non-glomeruli, and non-tissue regions.

[0088] Using Indica Labs Area Quantification FL v2.2.2, image data of samples stained with mixed monoclonal antibodies were evaluated by dividing them into FITC-positive regions (F), Texas Red-positive regions (T), or FITC / Texas Red bipositive regions (D) within the glomerular region.

[0089] The COL4α5 preservation rate was calculated as follows. COL4α5 preservation rate (%)=F / (F+TD)×100 (In the formula, F represents the area of ​​the FITC-positive region within the glomerular region, T represents the area of ​​the Texas Red-positive region within the glomerular region, and D represents the area of ​​the FITC / Texas Red bipositive region within the glomerular region.)

[0090] (Evaluation results of AI-based image analysis) Figure 8 is a graph showing the relationship between the COL4α5 preservation rate and the proportion of GBM lesions detected by AI. Figure 9 is a graph showing the relationship between urinary albumin / urinary creatinine concentration and the proportion of GBM lesions detected by AI. It showed a positive correlation with urinary albumin / urinary creatinine concentration and a negative correlation with the COL4α5 preservation rate. This suggests that samples stained using this method can be used for AI-based image analysis, and that image analysis can be used for the diagnosis of lesions.

[0091] (Example 2) (Organoid staining and imaging) To investigate whether this method can be used for staining biological samples containing cell-derived organoids, organoids derived from human breast cancer tissue, prepared as described below, were stained with this method and observed and imaged using LVSEM in the same manner as in Example 1. Figure 10 shows electron microscope images of human breast cancer tissue-derived organoids stained with the PAM staining method of this embodiment. It was confirmed that clear images could be obtained with LVSEM for cell-derived organoids.

[0092] (Organoid fabrication) The samples were obtained under the "Applied Research for Establishing Organoids and Novel Therapeutic Targets from Patient Cancer Tissue (R06-007)". Breast cancer tissue collected from patients was washed with 10 ml of AdDF+++, and the tissue was sculpted into 1-3 mm sculpts. 3 The meat was minced. After digesting with 10 ml of AdDF+++ containing 1-2 mg / ml of collagenase in an orbital shaker at 37°C for 1-2 hours, the mixture was pipetted with a glass Pasteur pipette and passed through a cell strainer with approximately 10 ml of AdDF+++. This step was repeated twice.

[0093] After resuspending in 2% FCS, the pellet was spun down at 400 rcf, then resuspended again in 10 ml of AdDF+++, and spun down again at 400 rcf. The resulting pellet was resuspended in 10 mg / ml cold Caltrex growth factor-reduced BME type 2. 40 μl was dropped into a 24-well suspension culture plate and allowed to solidify at 37°C for 20 minutes. 400 μl of BC organoid medium (composition described below) was added, and the culture was incubated at 37°C under 5% CO2 conditions. The medium was changed every 3-4 days.

[0094] (Composition of organoid medium) The composition of the organoid medium was based on previous studies (Sachs N, de Ligt J, Kopper O, Gogola E, Bounova G, Weeber F, Balgobind AV, Wind K, Gracanin A, Begthel H, Korving J, van Boxtel R, Duarte AA, Lelieveld D, van Hoeck A, Ernst RF, Blokzijl F, Nijman IJ, Hoogstraat M, van de Ven 11;172(1-2):373-386 (2018).).

[0095] The specific composition of the organoid culture medium is as follows: Culture medium name: Advanced DMEM / F12 Additive: N-Acetylcysteine ​​(1.25mM), GlutaMax(100x:1X), Nicotinamide (5mM) HEPES (10mM), Penicillin / Streptomycin (1x), Primocin (50g / ml), B27 supplement (1x), EGF (5 ng / ml), R-Spondin 1 (10%) conditioned medium or R-Spondin 3 (250 ng / ml), Noggin (100 ng / ml), A83-01 (500nM), Neuregulin (1.5 nM), Y-27632(5M), FGF7 (5 ng / ml), SB202190 (500nM), FGF10 (20 ng / ml)

[0096] (Example 3) (Staining and imaging of tissues other than kidney tissue and organoids) To investigate whether this method can be used to stain cells other than kidney tissue, biological samples containing Mycobacterium tuberculosis and fungi (Aspergillus) derived from autopsy tissue were stained using this method, and observed and imaged using an optical microscope and LVSEM in the same manner as in Example 1.

[0097] Figure 11 is an optical microscope image of Mycobacterium tuberculosis stained with the PAM staining method of this embodiment, and Figure 12 is an optical microscope image of a fungus stained with the PAM staining method of this embodiment. Figure 13 is an electron microscope image of Mycobacterium tuberculosis stained with the PAM staining method of this embodiment, and Figure 14 is an electron microscope image of a fungus stained with the PAM staining method of this embodiment. The scale bars in Figures 11-14 are all 10 μm.

[0098] In both Mycobacterium tuberculosis and fungi, this method was confirmed to produce clear images of the same sample using both an optical microscope and LVSEM. While the conventional Ziehl-Neelsen staining method for acid-fast bacilli, including Mycobacterium tuberculosis, selectively stains acid-fast bacilli, it has poor staining properties for cells and surrounding tissues, making it particularly difficult to determine their localization within extracellular tissues. Samples stained with this method are suitable for observation using both an optical microscope and LVSEM, and it was confirmed that the staining properties for cells and surrounding tissues are improved, making it easier to observe the localization of acid-fast bacilli within extracellular tissues compared to conventional methods.

[0099] The conventional Grocott staining method for fungi selectively stains fungi, allowing the stained samples to be used for observation with LVSEM. However, it has the problem that chromic acid, the oxidizing agent used for staining, is a highly toxic substance, requiring careful handling of the reagents. This method uses periodic acid as the oxidizing agent, which is expected to simplify reagent handling and improve safety. Furthermore, it was confirmed that the relationship between fungi and surrounding extracellular tissue is easier to observe compared to the conventional method.

Claims

1. The process of silvering biological samples, The process of coloring the aforementioned biological sample, The process includes fixing the color of the biological sample, The aforementioned biological sample is a biological sample that has been oxidized. The silvering step involves immersing the biological sample in a methenamine silver aqueous solution at room temperature, raising the methenamine silver aqueous solution from room temperature to 62°C or higher and 68°C or lower, and carrying out the silvering reaction for 30 to 40 minutes from the start of the reaction at room temperature. The aforementioned color-matching step involves immersing the silver-plated biological sample in a 0.2% gold chloride aqueous solution at room temperature for 8 to 12 hours. The aforementioned fixing step involves immersing the sample in a hard film fixing solution containing 11% to 13% sodium thiosulfate and dodecahydrate of aluminum potassium sulfate for 3 minutes or more at room temperature, in a PAM staining method.

2. The aforementioned biological sample is a section sample cut from living tissue. The PAM staining method according to claim 1, wherein the section sample is obtained by oxidizing a thin section of the biological tissue having a thickness of more than 1 μm.

3. The PAM staining method according to claim 2, further comprising the step of preparing the thin sections with a thickness of more than 1 μm and 30 μm or less, prior to the silvering step.

4. The PAM staining method according to claim 1, further comprising the step of immersing the biological sample, which has been immersed in the gold chloride aqueous solution, in an oxalic acid aqueous solution.

5. The PAM staining method according to claim 1, further comprising the step of oxidizing the biological sample with 0.5% periodic acid for 10 minutes prior to the silvering step.

6. The PAM staining method according to claim 1, wherein the methenamine silver aqueous solution consists of a mixture of 1.5% methenamine, 5% silver nitrate, and 5% borax, with a volume ratio of 100:5:

5.

7. Image data obtained by imaging the biological sample stained by the PAM staining method described in any one of claims 1 to 6, First image data corresponding to the first image captured using an optical microscope, It comprises a second image data corresponding to a second image with a higher magnification than the first image, which is imaged using an electron microscope, The second image is image data that is an enlarged image of a specific location in the first image.

8. The image data according to claim 7, wherein the first image data is training data for deep learning performed in a deep learning processor.

Images