Method for generating spectral data of a microparticle sample, method for analyzing microparticles, method for distinguishing microparticles, method for determining the presence or absence of exosomes derived from cancer cells, substrate for measuring microparticle spectra, device for measuring microparticle spectra, and apparatus for measuring microparticle spectra

Abstract

A method for generating spectral data according to the present invention involves generating spectral data pertaining to a microparticle sample that includes at least one microparticle, wherein: the method includes a step for acquiring a measurement spectrum from microparticles disposed within a through-hole in a substrate; the through-hole has an inclined structure in which the width continuously decreases from one surface of the substrate toward the other surface thereof; at least part of the inner surface of the through-hole is configured from a metal that exhibits plasmon resonance; and in the step for acquiring the measurement spectrum, the measurement spectrum is acquired while the interior of the through-hole is irradiated with light.

Description

[Technical Field]

[0001] The present invention relates to a method for generating (producing) spectral data of a particle sample. The present invention also relates to a method for analyzing particles, a method for distinguishing unidentified particles, a method for determining the presence or absence of exosomes derived from cancer cells in a sample derived from a body fluid containing exosomes, a substrate for measuring the spectrum of particles, a device for measuring the spectrum of particles, and an apparatus for measuring the spectrum of particles. [Background technology]

[0002] In various fields, analysis of microparticles or microparticle-based analyses is conducted based on the components contained in the microparticles. For example, for biological microparticles less than 5 μm in length, such as small microparticles like bacteria, exosomes, or viruses, the specific molecules present on their membrane surfaces can be elucidated to identify their type and determine the possibility of disease in the organism from which they originate. Furthermore, the component analysis of PM2.5, which is a type of microparticle less than 2.5 μm in diameter dispersed in the atmosphere, is used to analyze air pollution levels.

[0003] Among the methods for identifying the components contained in microparticles, spectroscopy is a technique that allows for non-destructive analysis. Raman spectroscopy, in particular, observes signals based on the vibrations of molecules that make up a substance, and can provide information based on each component that makes up the microparticle. Raman scattered light is very weak, but in recent years, techniques such as surface-enhanced Raman spectroscopy (SERS) have made it possible to perform more sensitive analysis. SERS is a method that utilizes the phenomenon in which the intensity of Raman scattered light from molecules is increased by localized surface plasmon resonance (SPR) on metals such as silver and gold that have rough surfaces. Patent Document 1 discloses an apparatus that uses SPR to analyze molecules such as polynucleotides.

[0004] Meanwhile, methods for identifying the type of object larger than a particle having a length of 5 μm or less are known. For example, Patent Document 2 discloses a method for identifying the type of cell contained in a sample using Raman spectra. In the method described in Patent Document 2, one Raman spectrum is obtained from one unclassified cell, and then a plurality of principal component spectra obtained by principal component analysis of a plurality of Raman spectra, each of which is obtained from a plurality of cells whose types are known, are analyzed to calculate a plurality of degrees of agreement indicating the degree to which the Raman spectrum of the unclassified cell matches with the spectra of the plurality of principal components, and the type of the unclassified cell is identified based on the classification of these degrees of agreement. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] International Publication No. 2009 / 030953 [Patent Document 2] International Publication No. 2019 / 117177 Summary of the Invention [Problem to be solved by the invention]

[0006] In the analysis method disclosed in Patent Document 2, a single Raman spectrum is acquired from a single unidentified microparticle for analysis. However, in actuality, paragraph

[0037] and Figure 4 of Patent Document 2 use eukaryotic cells such as rat basophilic leukemia cells (RBL) and Chinese hamster ovary cells (CHO), whose particle diameters are approximately 10 to 30 μm. Paragraph

[0036] of Patent Document 2 describes a Raman scattering light measurement device that changes the cells irradiated with laser light and acquires a single Raman spectrum from each of these large-diameter cells. However, the device described in Patent Document 2 is not applicable to microparticles with lengths of 5 μm or less, making it difficult to adapt to smaller microparticles.

[0007] Furthermore, in samples containing a large number of microparticles, analyses based on data from only one of the microparticles may not yield accurate results. Furthermore, the measurement results of one or a limited number of microparticles in an available sample may not be sufficient to perform the desired analysis. For example, even if specific components exist in exosomes derived from cancer cells, these are only a subset of the exosomes contained in the body fluid used for testing. Therefore, accurate determinations based on these specific components cannot be made based on the measurement results of one or a few exosomes. On the other hand, more common methods that measure a fixed amount of microparticles collectively do not readily reflect information on components contained in only a portion of the microparticles in a sample.

[0008] Furthermore, Patent Document 2 discloses a discrimination method that utilizes the results of principal component analysis of multiple spectra obtained from multiple cells whose types are known. However, to enable more reliable analysis, a method for easily obtaining a larger number of multiple spectra is desired.

[0009] An object of the present invention is to provide a method for generating spectral data for analyzing microparticles having a length of 5 μm or less or for analyzing using microparticles. Another object of the present invention is to provide a substrate, device, and apparatus for measuring the spectrum of microparticles that can be used in the above method. Another object of the present invention is to provide a method for analyzing microparticles, a method for distinguishing unidentified microparticles, and a method for determining the presence or absence of exosomes derived from cancer cells in a sample derived from a body fluid containing exosomes, all of which use the above method. [Means for solving the problem]

[0010] The present inventors have conducted extensive research to solve the above problems, and have completed a substrate capable of measuring single microparticles having a length of 5 μm or less with high sensitivity using a method completely different from the measurement method of Patent Document 2. They have also discovered a method for efficiently generating and acquiring measurement spectra for individual microparticles for a large number of microparticles using this substrate. Based on this method, further research has led to the completion of the present invention.

[0011] Specifically, the present invention is as follows.

[0012] [1] A method for generating (producing) spectral data for a particle sample containing at least one particle, comprising a step of acquiring a measured spectrum from a particle placed in a through-hole of a substrate, wherein the through-hole has a gradient structure in which the width continuously decreases from one side of the substrate to the other side, and at least a portion of the inner surface of the through-hole is made of a metal that exhibits plasmon resonance, and in the step of acquiring the measured spectrum, the measured spectrum is acquired while irradiating light into the through-hole. [2] The generation method described in [1], wherein in the step of acquiring the measured spectrum, a measured spectrum is acquired from each of a plurality of particles in the particle sample, and the spectral data is a bundle of a plurality of measured spectra. [3] The method according to [1] or [2], wherein the length of the fine particles is 10 nm to 5 μm. [4] The generation method described in [2], in the step of acquiring the measured spectrum, the plurality of particles are passed through the through-hole one by one, and the measured spectrum is acquired from each of the plurality of particles. [5] A generation method described in any one of [1] to [4], wherein in the step of acquiring the measurement spectrum, the microparticles dispersed in the liquid are moved into the through-hole by one or more methods selected from the group consisting of electrophoresis, dielectrophoresis, optical tweezers, Brownian motion, and Coulomb interaction. [6] The generation method according to any one of [1] to [5], wherein the measured spectrum is a Raman spectrum. [7] The generation method according to any one of [1] to [5], wherein the measured spectrum is a fluorescence spectrum. [8] A method for analyzing fine particles, comprising a step of performing statistical analysis of the spectral data obtained by the generation method described in [2] or [4]. [9] The step of performing statistical analysis of the spectral data includes forming a set of peaks having a high correlation coefficient among the plurality of measured spectra of the spectral data; The analytical method according to [8], further comprising a step of comparing the obtained set of peaks with the spectrum of a known substance to identify at least one component contained in the microparticles.

[10] The analytical method described in [8], wherein the step of performing statistical analysis of the spectral data includes a step of performing multivariate analysis on the multiple measured spectra of the spectral data, and a step of comparing the spectrum obtained by the multivariate analysis with the spectrum of a known substance to identify at least one component contained in the microparticles.

[11] A method for distinguishing unidentified particles, comprising the steps of: acquiring a measurement spectrum for each of a plurality of particles A whose type is known and a measurement spectrum for each of a plurality of particles B whose type is known, using the generation method described in claim 2; performing principal component analysis on spectral data including the measurement spectra of the plurality of particles A and the measurement spectra of the plurality of particles B, and determining an index for distinguishing between the measurement spectra of particle A and particle B from the scores of two or more principal components; acquiring a measurement spectrum from each of one or more unidentified particles in a particle sample using the generation method described in any one of [1] to [7]; calculating the scores of the two or more principal components for the measurement spectra of the unidentified particles; and comparing the scores of the unidentified particles with the index to distinguish them, wherein the length of the unidentified particles is 10 nm to 5 μm.

[12] The method of

[11] , wherein the unclassified microparticles are exosomes, the microparticles A are exosomes derived from cancer cells, and the microparticles B are exosomes derived from normal cells, and the step of performing the classification involves determining whether the exosomes as the unclassified microparticles are derived from cancer cells.

[13] A method for determining the presence or absence of cancer cell-derived exosomes in a sample derived from a body fluid containing exosomes, comprising the steps of generating spectral data consisting of a plurality of measurement spectra obtained from each of a plurality of exosomes in the sample by the generation method according to [2] or [4], and detecting a difference of 1087 cm between the plurality of measurement spectra. -1 ±5cm -1 A signal with a maximum at 1435 cm -1 ±5cm -1 and determining that cancer cell-derived exosomes are present in the sample when the correlation coefficient is equal to or greater than a certain value.

[14] A substrate for measuring the spectrum of microparticles, the substrate having through-holes penetrating from one surface to the other surface, the through-holes being of a size that allows the microparticles to pass through one by one, at least a portion of the inner surface of the through-holes being made of a metal that exhibits plasmon resonance, and the through-holes having a gradient structure in which the width continuously decreases from the one surface to the other surface of the substrate.

[15] The substrate according to

[14] , wherein the through-hole has a frustum shape.

[16] The substrate according to

[15] , wherein the through-hole on the other surface of the substrate has an equivalent circle diameter of 10 nm to 5 μm.

[17] A device for measuring a spectrum of a fine particle, comprising: the substrate according to any one of

[14] to

[16] ; a first liquid tank having an inner wall including at least a portion of the one surface of the substrate that includes the through-hole; and a second liquid tank having an inner wall including at least a portion of the other surface of the substrate that includes the through-hole. A device for spectral measurement of particulate matter, comprising:

[18] An apparatus for measuring the spectrum of a microparticle, comprising: the device described in

[17] ; a guide unit for passing the microparticles one by one through the through-hole; a light source; and a detection unit for measuring the light generated when the light from the light source is irradiated onto the microparticle in the through-hole to obtain a measured spectrum.

[19] The apparatus according to

[18] , wherein the measured spectrum is a Raman spectrum.

[20] The apparatus according to

[18] , wherein the measured spectrum is a fluorescence spectrum. [Effects of the Invention]

[0013] The present invention provides a novel method for generating spectral data for analyzing microparticles having a length of 5 μm or less or for analyzing using microparticles. The present invention also provides a substrate, device, and apparatus for measuring the spectrum of microparticles that can be used in the above method. The present invention also provides a method for analyzing microparticles, a method for distinguishing unidentified microparticles, and a method for determining the presence or absence of cancer cell-derived exosomes in a sample derived from a body fluid containing exosomes, all of which use the above method. [Brief explanation of the drawings]

[0014] [Figure 1] 1A to 1C are diagrams showing the manufacturing procedure of the substrate used in the examples. [Figure 2] FIG. 2 is a diagram showing the structure of the substrate used in the examples. [Figure 3] FIG. 3 shows a schematic cross-sectional view of a device for spectrum measurement used in the examples. [Figure 4] FIG. 4 shows a schematic cross-sectional view of the Raman spectrum measurement system used in the examples. [Figure 5] FIG. 5 shows a circuit block diagram of the Raman spectrum measurement system used in the examples. [Figure 6] Figure 6 shows an overlapping image of measured Raman spectra of multiple (N=100) microparticles in an exosome sample derived from normal cells. [Figure 7] Figure 7 shows an overlapping image of measured Raman spectra of multiple (N=155) microparticles in an exosome sample derived from liver cancer cells. [Figure 8] FIG. 8 shows the results of mapping the correlation coefficients of the measured Raman spectra of multiple microparticles in an exosome sample derived from normal cells. [Figure 9]Figure 9 shows the results of mapping the correlation coefficients of the measured Raman spectra of multiple microparticles in an exosome sample derived from liver cancer cells. [Figure 10] FIG. 10 shows the spectrum of the first principal component obtained by performing principal component analysis on the measured Raman spectra of multiple microparticles in an exosome sample derived from normal cells. [Figure 11] Figure 11 is a plot of the first principal component score based on principal component analysis of the measured Raman spectra of multiple microparticles in exosome samples derived from liver cancer cells and normal cells, plotted as the abscissa and the second principal component score as the ordinate. [Figure 12] Figure 12 shows the plot of the first principal component score on the abscissa and the second principal component score on the ordinate based on principal component analysis of the measured Raman spectra of multiple microparticles in exosome samples derived from normal cells, X-ray-irradiated normal cells, liver cancer cells, blood from healthy individuals, and senescent cells. [Figure 13] FIG. 13 is a diagram showing an overlap of measured Raman spectra of a plurality of silica fine particles (N=100). [Figure 14] FIG. 14 is a diagram showing the measured fluorescence spectra of gold particles inside and outside the through-holes superimposed on each other. DETAILED DESCRIPTION OF THE INVENTION

[0015] The present invention will be described in detail below. The following description of the constituent elements may be based on representative embodiments or specific examples, but the present invention is not limited to such embodiments. In this specification, a numerical range expressed using "to" means a range that includes the numerical values ​​before and after "to" as the lower and upper limits.

[0016] <Method for generating spectral data of particulate samples> The method for generating spectral data of a particle sample of the present invention (hereinafter referred to as the generation method of the present invention) is a method for generating (producing) spectral data of a particle sample containing at least one particle, and includes obtaining a measurement spectrum from a particle placed in a through-hole of a substrate. As will be described later, the through-hole has a sloped structure in which the width continuously decreases from one surface of the substrate to the other, and at least a portion of the inner surface of the through-hole is made of a metal that exhibits plasmon resonance. Furthermore, in the step of acquiring the measurement spectrum, the measurement spectrum is acquired while irradiating light into the through-hole. If the particle sample contains multiple particles, the step of acquiring the measurement spectrum may acquire a measurement spectrum from each of the multiple particles in the particle sample. In this case, the spectral data is a bundle of multiple measurement spectra.

[0017] The present invention relates to a method for generating spectral data of a particle sample. More specifically, the present invention relates to a method for generating spectral data of a particle sample for analyzing particles based on components contained in the particles, or for performing analysis using the particles. In this specification, the term "components contained in the particles" is not particularly limited and may be one specific component or substantially all components, and is appropriately selected depending on the type of spectrum to be acquired and the purpose of the analysis. The components contained in the particles to be analyzed may be components contained in any part of the particles, and will vary depending on the type of spectrum to be acquired. However, when spectral data can be acquired without destroying the particles, it is possible to analyze components contained in the surfaces of the particles in particular.

[0018] When a particle sample contains multiple particles, the generation method of the present invention can provide spectral data of the particle sample as a bundle of measurement spectra based on each particle. Typically, spectral data of a particle sample is obtained by measuring a certain amount of particles collectively, so the data is averaged and it is difficult to reflect information about components contained in only some of the particles in the sample. Previously, analysis based on the measurement spectrum of a single particle was difficult. Meanwhile, even analysis based on the measurement spectrum of a single cell with a length exceeding 10 μm was limited to using the measurement spectrum of one or a limited number of cells, as described in Patent Document 2.

[0019] The generation method of the present invention provides spectral data of a particle sample containing multiple particles as a bundle of measured spectra based on each particle, making it possible to perform analysis using statistical processing. It also makes it possible to detect components contained in various particles that have not been detectable in the past. Furthermore, it also makes it possible to make judgments that were not possible with conventional methods, such as judgments based on the amount of particles containing a specific component among the particles in a sample. Hereinafter, the spectral data generated by the generation method of the present invention (a bundle of measured spectra based on each particle) is also referred to as the "new database obtained by the present invention."

[0020] (fine particles) In this specification, the term "microparticles" refers to particles having a length of 10 nm to 5 μm. The shape of the microparticles is not particularly limited, and may be, for example, spherical, ellipsoidal, cylindrical, cubic, pyramidal, other polyhedral, conical, or irregular.

[0021] In this specification, the length of a fine particle means the length of the longest line (major axis) among the two lines connecting the intersections of a line passing through the center of the cross section of the fine particle and the periphery of the cross section. In the present invention, the length of the fine particle is preferably 10 nm to 3 μm, more preferably 10 nm to 1 μm, particularly preferably 30 to 500 nm, and more particularly preferably 50 to 200 nm.

[0022] In this specification, the shortest line (minor axis) among the two lines connecting the intersections of a line passing through the center of the cross section of a fine particle and the periphery of that cross section is preferably 1 nm to 5 μm, more preferably 3 nm to 3 μm, particularly preferably 5 nm to 1 μm, more particularly preferably 10 to 500 nm, even more particularly preferably 30 to 300 nm, and most preferably 50 to 200 nm. When passing fine particles one by one through the through-holes of a substrate having through-holes, it is preferable that the range of the minor axis of the fine particles is within the above-mentioned preferred range in addition to the range of the length (major axis).

[0023] The ratio of the length (major axis) to the minor axis (aspect ratio) of the microparticles may be, for example, 99:1 to 50:50, preferably 90:10 to 50:50, more preferably 80:20 to 50:50, particularly preferably 70:30 to 50:50, and even particularly preferably 60:40 to 50:50. Microparticles with a relatively small aspect ratio are preferable to elongated microparticles with a large aspect ratio, such as polynucleotides and other polymer molecules, because it is easier to pass the microparticles one by one through the through-holes of a substrate having the microparticles.

[0024] Although microparticles may be composed of a single component or multiple components, the production method of the present invention is particularly useful when targeting microparticles composed of multiple components. Examples of microparticles include biological microparticles derived from living organisms and inorganic microparticles. Examples of biological microparticles include small microparticle-like bacteria, organelles such as mitochondria, viruses, and exosomes. Among biological microparticles, those with a lipid bilayer membrane structure are preferred, with viruses and exosomes being more preferred, from the perspective of passing the microparticles one by one through the through-holes of a substrate with through-holes and obtaining measurement spectra at the positions of the through-holes. Examples of inorganic microparticles include PM2.5.

[0025] Examples of components contained in biological microparticles include lipids, proteins, sugar chains, etc. Examples of components contained in inorganic microparticles include carbon components, nitrates, sulfates, ammonium salts, as well as silicon, sodium, aluminum, etc.

[0026] Among these, the microparticles are preferably viruses, exosomes, or particles dispersed in the air with a diameter of 2.5 μm or less, since they share the common property of being microparticles with a length of 10 nm to 3 μm and an aspect ratio within a preferred range.

[0027] Examples of small microparticle-like bacteria include fungi, bacteria, mycoplasma, etc. Typical examples include bacteria with a particle size of 100 nm to 5 μm, with bacteria with a particle size of 100 nm to 3 μm being preferred. An example of a component contained in bacteria that can be analyzed using spectral data obtained by the production method of the present invention is membrane protein. Analysis of membrane protein makes it possible to determine the type and state of bacteria, etc.

[0028] Known viruses include coronaviruses, noroviruses, influenza viruses, and Ebola viruses. Typical examples include viruses with particle sizes of 10 nm to 1 μm. Examples of components contained in viruses that can be analyzed using spectral data obtained by the generation method of the present invention include spike proteins on the surface. Analysis of spike proteins makes it possible to identify the type of virus and also to identify mutants.

[0029] As used herein, exosomes are synonymous with "extracellular vesicles." Extracellular vesicles are defined as "nucleusless (non-replicable) lipid bilayer-enclosed particles released from cells." Examples of components contained in exosomes that can be analyzed using spectral data obtained by the production method of the present invention include components within the lipid bilayer or components present on the surface of the lipid bilayer.

[0030] In recent years, "exosomes" have been considered a type of extracellular vesicle, and extracellular vesicles are classified into "exosomes," "microvesicles," and "apoptotic bodies" based on their production mechanism and size. Exosomes are vesicles derived from endosomes and have a diameter of approximately 50-150 nm. Microvesicles are vesicles secreted directly from cells and have a diameter of approximately 0.1-1 μm. Apoptotic bodies are cell fragments generated by cell death and have a diameter of approximately 1-4 μm. However, when referring to exosomes in this specification, it can refer to "exosomes," "microvesicles," "apoptotic bodies," or a mixture of these. For example, it can be a mixture of "exosomes" and "microvesicles."

[0031] Exosomes are microparticles secreted by cells and present in body fluids. Because they contain information about the cells that secreted them, they have attracted attention as biomarkers for diseases such as cancer. In particular, by examining the components present on the surface of the lipid bilayer membrane of exosomes, specifically molecules such as proteins, lipids, and glycans, it is possible to predict the type of cancer and the site of metastasis (Non-Patent Documents 1-3). However, conventional Raman spectroscopy measurements of exosomes yield spectra that display information about multiple components, such as proteins, lipids, and glycans, making it difficult to analyze specific components. The spectral data generation method of the present invention enables the analysis of specific components of exosomes using Raman spectroscopy, which may enable the identification of markers for diseases such as cancer. [Non-Patent Document 1] Haiying Zhang, et al., "Identification of distinct nanoparticles and subsets of extracellular vesicles by asymmetric flow field-flow fractionation", Nature Cell Biology, Vol. 20, pp. 332-343. [Non-Patent Document 2] Ayuko Hoshino, et al., “Tumor exosome integrins determine organotropic metastasis,” Nature, Vol. 527, pp. 329-335. [Non-Patent Document 3] Ayuko Hoshino, et al., “Extracellular Vesicle and Particle Biomarkers Define Multiple Human Cancers,” Cell, Vol. 182, pp. 1044-1061.

[0032] Cancer-derived exosomes are found in bodily fluids such as blood, urine, and saliva. Therefore, unlike conventional methods, testing and diagnosis using exosomes can be performed using non-invasive samples. Note that bodily fluids contain biological components other than exosomes. Because these components contribute to measurement noise, exosomes may be isolated from samples using known procedures, as described in the Examples below. Examples of cancers that secrete exosomes include gastric cancer, esophageal cancer, lung cancer, liver cancer, biliary tract cancer, pancreatic cancer, colorectal cancer, ovarian cancer, prostate cancer, bladder cancer, breast cancer, glioma, glioblastoma, melanoma, and medulloblastoma.

[0033] PM2.5 is a type of fine particle dispersed in the atmosphere with a diameter (particle size, preferably the longest axis) of 2.5 μm or less. PM2.5 is composed of salts such as sulfates and nitrates, silica components, metal components, black carbon, and various organic compounds, and is primarily suspended as an aerosol. Spectral data obtained using the generation method of the present invention can analyze the components on the surface of PM2.5 as well as each of the components that make up PM2.5. The composition of PM2.5 varies depending on the environment, and studying the composition of PM2.5 can elucidate the origin of PM2.5 and its effects on the body.

[0034] (particle sample) The method for generating spectral data for a particulate sample of the present invention includes obtaining a measured spectrum from each of n particles in a particulate sample containing multiple particles. Examples of particulate samples include cell culture media, cell suspensions, body fluids containing exosomes or suspensions containing exosomes isolated from body fluids, virus suspensions, body fluids containing viruses or suspensions containing viruses isolated from body fluids, filters that collect PM2.5 from the atmosphere, or solutions (aqueous solutions) in which PM2.5 extracted from such filters is dissolved.

[0035] The particle sample may be pre-treated according to the type of measurement spectrum to be obtained from the sample and the intended use of the spectral data. For example, the particle sample may be labeled with a fluorescent substance using molecules on the particle surface for detection by fluorescence. Furthermore, the molecules on the particle surface may be modified with reactive molecules to bind or adsorb the particles to the measurement substrate or inside the through-holes.

[0036] (Number of particles in the particle sample, n) A particle sample is typically a sample containing a large amount of particles, and may be a sample containing, for example, about 10 to the power of 20, or about 10 to the power of 15, particles. In the particle spectral data generation method of the present invention, a measured spectrum is obtained from each of n particles corresponding to all or a portion of the particle contained in the sample. n may be determined depending on the amount (number) of particles contained in the sample and the intended use of the spectral data, and may be 2 or more, 3 or more, 4 or more, 5 or more, 10 or more, 20 or more, 30 or more, 40 or more, 50 or more, 60 or more, 70 or more, 80 or more, 90 or more, 100 or more, etc. Statistical analysis is possible using the spectral data consisting of n measured spectra obtained by the generation method of the present invention. When measuring unknown particles, more reliable analysis is possible by using spectral data with a larger n (preferably 100 or more). On the other hand, when measuring particles with known spectral characteristics, a highly reliable analysis is possible even with an n number of 5 or less. Furthermore, by obtaining measured spectra from each of multiple particles, more accurate information about the particles can be obtained and characteristics exhibited by only a portion of the particles in the sample can be detected. There is no particular upper limit to n, but it may be 1000 or less, 500 or less, 300 or less, etc., taking into consideration the time and effort required to acquire a measurement spectrum and data processing.

[0037] (Measured spectrum) In the method for generating spectral data for a particle sample of the present invention, a measured spectrum is obtained from each of n particles. It is sufficient to measure one or more measured spectra per particle, but it is preferable that the number of spectra measured per particle is the same, and it is preferable to obtain one measured spectrum per particle. In other words, the spectral data obtained by the method for generating spectral data for a particle sample of the present invention is preferably a bundle of n measured spectra.

[0038] The bundle of n measurement spectra obtained by the method for generating spectral data of a particulate sample of the present invention is a bundle of n measurement spectra illustrated in two or more dimensions. The bundle of n measurement spectra means a collection of n measurement spectra, and does not mean a spectrum obtained by integrating n measurement spectra or averaging n measurement spectra. However, the bundle of n measurement spectra may be integrated or averaged as necessary. The bundle of n measurement spectra may also be data processed from n measurement spectra illustrated in two or more dimensions (for example, a matrix, heat map, or waterfall plot).

[0039] The measurement spectrum is not particularly limited, and may be any spectrum that can provide results based on one particle. Examples of the measurement spectrum include Raman spectrum, infrared spectrum, fluorescence spectrum, mass spectrum, etc. The measurement spectrum is preferably Raman spectrum or fluorescence spectrum, and particularly preferably Raman spectrum.

[0040] (Method of obtaining the measured spectrum) The method for generating spectral data of a particle sample of the present invention includes a step of obtaining a measurement spectrum from each of n particles in the particle sample. There are no particular limitations on the method for obtaining a measurement spectrum from the particles. For example, it is preferable that an external stimulus is applied to the particles according to the measurement spectrum to be obtained, and a signal derived from the particles is measured. For example, to obtain an infrared spectrum, infrared light is irradiated and the absorption of the irradiated light according to the wavenumber is observed. To obtain a fluorescence spectrum, monochromatic excitation light is irradiated and the resulting fluorescence is observed. To obtain a mass spectrum, electron beam irradiation, ion irradiation, laser irradiation, or the like required for ionizing molecules is performed.

[0041] The type of external stimulus depends on the measurement spectrum to be acquired as described above, but is preferably light. The light is preferably laser light, as laser light has high directionality and focusing ability and is suitable for measuring particles with lengths of 10 nm to 5 μm. The laser light can be selected from semiconductor lasers, gas lasers, solid-state lasers, etc., and can be used with the required wavelength.

[0042] To obtain a Raman spectrum, it is sufficient to irradiate a substance with monochromatic laser light. When a substance is irradiated with monochromatic light of a certain frequency, most of the scattered light produced has the same frequency as the incident light, but some scattered light (Raman scattered light) with a slightly different frequency is also produced. The difference in frequency between this scattered light and the incident light (Raman shift) corresponds to the vibrational state of the molecules that make up the substance. Therefore, a Raman spectrum, which plots the intensity of Raman scattered light against the Raman shift, is used as an analytical method to understand the structure and state of the molecules that make up a substance.

[0043] The wavelength of the monochromatic light used to obtain a Raman spectrum can be selected from wavelengths obtained using a known light source that irradiates external light (excitation light) with a wavelength capable of generating Raman scattered light. For example, a wavelength in the range of approximately 400 to 800 nm can be appropriately selected depending on the microparticles to be analyzed and the components contained in the microparticles. Since near-infrared light, which has little effect on biological samples, can be used as excitation light for Raman spectroscopy, it is also possible to observe living cells under culture conditions.

[0044] Since the Raman scattered light from the components contained in a single particle is very weak, it is preferable to enhance it. Examples of methods for enhancement include resonance Raman scattering, tip-enhanced Raman scattering (TERS), and surface-enhanced Raman scattering (SERS), and these methods may be combined. Each method is well known. For example, by adsorbing a molecule to a metal that exhibits plasmon resonance, such as gold or silver, to obtain surface-enhanced Raman scattering, the scattered light intensity can be increased by 10 times compared to the Raman scattered light from an isolated molecule. 4It is known that the Raman scattering can be enhanced by 10 (to the fourth power) times or more. When a measurement spectrum is obtained using the substrate of the present invention having through-holes whose inner surfaces are made of a metal that exhibits plasmon resonance, an enhancement effect based on at least surface-enhanced Raman scattering can be obtained. Furthermore, by forming the through-holes into a predetermined shape, it is possible to enhance the Raman scattering by 10 8 An enhancement of 10 to the power of 8 can be obtained. The external stimulus used when using this substrate is not particularly limited as long as it induces plasmon resonance, but it is preferable to use a monochromatic laser, which is used for Raman spectrum measurement.

[0045] Examples of a method for obtaining a measured spectrum from each of n particles in a particle sample containing a plurality of particles (preferably, a method for generating a Raman spectrum from one particle) include the following.

[0046] First, one method involves scattering the particles in the sample on a substrate used for spectrum measurement, scanning the substrate with a probe such as a light irradiator and detector, and obtaining a measured spectrum of each particle under microscopic observation. Alternatively, the particles in the sample can be scattered on a substrate used for spectrum measurement, and the substrate can be scanned to obtain a measured spectrum of each particle. When the measured spectrum is a Raman spectrum, it is preferable to use a probe that induces plasmon resonance.

[0047] Another method involves moving the particles in a substantially horizontal direction on a substrate used for spectrum measurement using electrophoresis or optical tweezers under microscopic observation, while acquiring a measurement spectrum for each particle. For example, when the measurement spectrum is a Raman spectrum or a fluorescence spectrum, it is preferable to use a substrate provided with through-holes, at least the surface of which is made of a metal that exhibits plasmon resonance. This is because by holding the particles at the position of the through-holes by adsorption or the like and acquiring a measurement spectrum, it is possible to obtain surface-enhanced Raman scattering light or fluorescence enhancement due to plasmon resonance (surface plasmon-excited enhanced fluorescence). This procedure can be repeated to obtain a measurement spectrum from each of the n particles.

[0048] Another method involves moving each particle in a sample one by one in a direction approximately normal to the substrate, while obtaining a measurement spectrum for each particle. For example, electrophoresis or optical tweezers can be used to pass each particle through a through-hole in a substrate and measure the particle at the position of the through-hole. This procedure makes it possible to obtain many measurement spectra more quickly and easily. For example, when the measurement spectrum is a Raman spectrum or a fluorescence spectrum, it is preferable to use a substrate in which at least the inner surface of the through-hole is made of a metal that exhibits plasmon resonance. This is because obtaining a measurement spectrum inside the through-hole allows surface-enhanced Raman scattering or surface plasmon-excited enhanced fluorescence to be obtained. This procedure can be repeated to obtain a measurement spectrum from each of n particles.

[0049] In a method for obtaining a measurement spectrum from each of n particles in a particle sample containing multiple particles, obtaining the measurement spectrum while the substrate is fixed is preferable from the viewpoint of obtaining many measurement spectra more quickly and easily. Compared to the method of moving the sample holder and obtaining Raman spectra from each cell, as described in paragraphs

[0035] and

[0036] of Patent Document 2, obtaining the measurement spectrum while the substrate is fixed allows for faster measurement spectrum acquisition. Furthermore, it is preferable that the light irradiator and detector are also fixed.

[0050] Furthermore, from the viewpoint of further enhancing the Raman scattered light intensity or the fluorescent light intensity, it is more preferable to obtain the Raman spectrum or the fluorescent spectrum by a procedure that includes irradiating light from the side having the large opening of the through-hole.

[0051] <Method for analyzing fine particles> The method for analyzing microparticles of the present invention includes a step of performing statistical analysis of the spectral data obtained by the generation method of the present invention. By performing statistical analysis on the spectral data obtained by the method for generating spectral data of a microparticle sample of the present invention, i.e., a bundle of n measured spectra (a new database obtained by the present invention), analysis based on the components contained in the microparticles can be performed. The statistical analysis may be performed using machine learning or artificial intelligence (AI). When analyzing unknown microparticles by statistical analysis, in order to obtain sufficient accuracy and reliability when performing the generation method of the present invention (when constructing a new database obtained by the present invention), n is preferably 20 or more, more preferably 50 or more, even more preferably 80 or more, and particularly preferably 100 or more. On the other hand, when analyzing known microparticles or analysis using known microparticles, sufficient reliability may be obtained even with a small n, such as 2 to 5, when performing the generation method of the present invention (when constructing a new database obtained by the present invention).

[0052] (Multivariate analysis) One preferred embodiment of the analytical method of the present invention is an analytical method including the steps of performing multivariate analysis on n measured spectra of spectral data obtained by the generation method of the present invention, and comparing the spectrum obtained by the multivariate analysis with the spectrum of a known substance to identify at least one component contained in the microparticle.

[0053] Spectral data can be subjected to multivariate analyses such as principal component analysis, sparse principal component analysis, non-negative matrix factorization (multivariate waveform decomposition - alternating least squares), cluster analysis, independent component analysis, linear discriminant analysis, logistic regression analysis, or Gaussian mixture models to obtain information based on the components of the microparticles.

[0054] In the method of generating spectral data of fine particles according to the present invention, a bundle of measured spectra can be obtained for each fine particle, and therefore, information about the components contained therein can be obtained using the above-mentioned analytical method.

[0055] For example, if information about the constituent components of the measured microparticles is known, detailed analysis can be performed by analyzing the new database obtained by the present invention using spectral data of the known constituent components (known spectral database) as a reference spectrum, etc. Furthermore, whether or not a specific component (e.g., a cancer marker molecule) is present in the microparticles can also be determined by analysis based on the spectrum of the specific component.

[0056] Furthermore, by utilizing the above-described analytical method, it is possible to obtain information about the components contained in even unknown particles. For example, principal component analysis is performed on the spectral data generated by the generation method of the present invention, and the obtained spectra of the principal components are compared with known spectra (for example, a database of known spectra can be used) to identify the components contained in the particles.

[0057] (Analysis using correlation coefficients) Another example of a method for analyzing microparticles is analysis using a correlation coefficient. Specifically, a set of peaks with a high correlation coefficient between n measured spectra of the spectral data generated by the generation method of the present invention is formed, and the obtained set of peaks is compared with the spectrum of a known substance (for example, a known spectrum database can be used) to identify the components contained in the microparticles. The set of peaks may include two or more peaks, and may, for example, consist of only two peaks (for example, two peaks characteristic of the components), or may correspond to a spectrum.

[0058] In other words, if peak A increases and peak B also increases by the same amount, the correlation coefficient between A and B is 1; if peak B remains unchanged, the correlation coefficient is 0; and conversely, if peak A decreases by the same amount, the correlation coefficient between peaks A and B is -1. In the case of Raman spectra, peaks correspond to molecules, so the higher the correlation coefficient, the more likely it is that those peaks are signals from the same molecule. This correlation is examined for all the data obtained, that is, the correlation coefficient is calculated between each spectrum for the peak of interest in n spectra. By doing so, it is possible to identify peaks that are linked to each other. These linked peaks can be determined to be signals from the same molecule.

[0059] Specifically, the spectrum of each particle is organized into an N × M matrix as shown in Table 1. Then, the correlation coefficient of the intensity at each two wavenumber points (Yth and Y'th) is calculated based on a formula specifically shown in the Examples. If necessary, the maximum value of each spectrum may be normalized to 1, or a general standardization process may be performed, or both may be performed before the correlation coefficient is calculated.

[0060] [Table 1]

[0061] <Method for determining the presence or absence of cancer cell-derived exosomes in a sample derived from body fluid containing exosomes> Analysis using correlation coefficients also makes it possible to analyze the presence or absence of specific particles by utilizing signals based on the partial structure of components that exhibit the properties of the particles.

[0062] The determination method of the present invention is a method for determining the presence or absence of cancer cell-derived exosomes in a sample derived from a body fluid containing exosomes, and includes the steps of generating spectral data consisting of n measurement spectra obtained from n exosomes in the sample by the generation method of the present invention, and -1 ±5cm -1 A signal with a maximum at 1435 cm -1 ±5cm-1 and determining that cancer cell-derived exosomes are present in the sample when the correlation coefficient is equal to or greater than a certain value.

[0063] In the examples, 155 measured Raman spectra were acquired for exosomes in a sample derived from exosome-containing body fluids, and the presence of phosphorylated proteins was identified by confirming the correlation between two characteristic signals resulting from the structure of phosphorylated serine among the obtained spectra. Specifically, the presence of phosphorylated proteins was confirmed by confirming the correlation between two characteristic signals at 1087 cm -1 ±5cm -1 A signal with a maximum at 1435 cm -1 ±5cm -1 The presence of phosphorylated proteins is identified based on whether the signal with a maximum value is correlated with the signal with a maximum value, i.e., whether the correlation coefficient is above a certain value. Phosphorylated serine is a characteristic partial structure of phosphorylated proteins, and the presence of phosphorylated proteins is a characteristic of exosomes derived from cancer cells (Non-Patent Document 4). Therefore, the presence or absence of exosomes derived from cancer cells in a sample can be determined based on whether or not the above signals are correlated. This determination result can be used for cancer diagnosis. [Non-Patent Document 4] Shilian Dong et al., “Beehive-Inspired Macroporous SERS Probe for Cancer Detection through Capturing and Analyzing Exosomes in Plasma”, ACS Appl. Mater. Interfaces, Vol. 12, pp. 5136-5146.

[0064] The analysis using the correlation coefficient can also be performed using two-dimensional mapping of the spectrum as shown in the examples.

[0065] <Method for distinguishing unidentified particles> Furthermore, the spectral data of known particles obtained by the generation method of the present invention can be used to analyze unidentified (unknown) particles, for example, to distinguish between particles of the same type (e.g., viruses or exosomes) but with different properties (e.g., distinguishing between mutant viruses, distinguishing between particles derived from patients with a specific disease such as cancer, etc.).

[0066] As an example, if a principal component analysis is performed using a bundle of measurement spectra obtained by adding together bundles of measurement spectra obtained using the generation method of the present invention for different particle samples, and an index for distinguishing between different particles is found, the index can be used to distinguish between the particles.

[0067] Among the particle discrimination methods using the spectral data of known particles obtained by the generation method of the present invention, it is preferable to apply it to the discrimination method of the present invention described below.

[0068] The discrimination method of the present invention is a discrimination method for discriminating undistinguished particles, comprising the steps of: N of particles A whose type is known A The measured spectra of the particles and other known particles of B B acquiring each of the measured spectra by the generation method of the present invention; N A +N B a step of performing principal component analysis on the spectral data consisting of the measured spectra, and determining an index for distinguishing the measured spectra of particle A and particle B from the scores of two or more principal components; obtaining a measured spectrum from each of one or more unidentified particles in the particle sample; calculating scores of two or more principal components for the measured spectrum of the unclassified particle; The method includes a step of comparing the score of the unclassified particles with an index to perform classification, and the length of the unclassified particles is 10 nm to 5 μm.

[0069] The following describes possible indicators for distinguishing target specimen particles (unidentified particles) as either particle A (known specimen particle 1) or particle B (known specimen particle 2).

[0070] A bundle of N1 measured spectra of analyte particle 1 and a bundle of N2 measured spectra of analyte particle 2 are added together to obtain spectral data consisting of a bundle of N1 + N2 measured spectra. Principal component analysis is performed on the spectral data consisting of the N1 + N2 measured spectra, and an index for distinguishing the measured spectra of analyte particle 1 and analyte particle 2 is obtained from the scores of two or more principal components. N1 and N2 are each preferably 20 or more, more preferably 50 or more, and even more preferably 100 or more. As a specific example, assuming that there are MM spectral data points, (N1 + N2) × MM matrices are created. For example, if there are 100 spectral bundles for each of analyte particle 1 and analyte particle 2, and the total number of spectral data points is 1000, the sum of the 200 spectral bundles results in a 200 × 1000 matrix. A variance-covariance matrix for this matrix is ​​calculated, and the eigenvalues ​​and eigenvectors for this variance-covariance matrix are then calculated. In this process, the spectral bundles may be normalized and standardized as necessary. The principal components are ranked in descending order of eigenvalue, and the eigenvectors at each eigenvalue are used to create a score plot. If a score plot of two or more principal component scores, for example, the first and second principal components, can be created and a boundary line can be formed between each particle, this can be used as the index.

[0071] That is, by calculating the scores of the principal components for the measurement target spectrum obtained from an unidentified particle and comparing the results with the score plot, it is possible to determine which particle it is. The measurement target spectrum may be a measured spectrum of a single particle, or may be spectral data generated by the generation method of the present invention.

[0072] The preferred embodiment for obtaining the spectral data of unidentified particles as a measurement spectrum of a single particle is the same as the preferred embodiment for the generation method of the present invention. Furthermore, when obtaining the unidentified particles as a measurement spectrum of a single particle, it is preferred to obtain the measurement spectrum of a single particle using the particle spectrum measurement substrate of the present invention.

[0073] (Method for determining whether or not a target exosome is derived from a cancer cell) Among the methods for distinguishing unidentified microparticles, the method of the present invention is a method for determining whether or not a target exosome is derived from a cancer cell, A step of obtaining N1 measurement spectra of exosomes derived from cancer cells and N2 measurement spectra of exosomes derived from normal cells using the production method of the present invention; performing principal component analysis on spectral data consisting of N1+N2 measured spectra, and determining an index for discriminating between measured spectra derived from cancer cells and normal cells from the scores of two or more principal components; calculating scores of two or more principal components for the spectrum of the target exosome; It is preferable to include a step of comparing the score of the target exosome with an index to make a distinction.

[0074] Other preferred embodiments of the discrimination method include a discrimination method, learning method, and discrimination device that use microparticles having a length of 5 μm or less instead of cells, as described in paragraphs

[0006] to

[0085] of Patent Document 2, and the contents of the paragraphs in Patent Document 2 are incorporated herein by reference.

[0075] Specifically, it is preferable to obtain one measurement spectrum (e.g., a Raman spectrum or a fluorescence spectrum) from one unidentified particle, calculate multiple degrees of agreement indicating the degree to which the measurement spectrum of the unidentified particle matches multiple principal component spectra obtained by principal component analysis of multiple measurement spectra consisting of measurement spectra obtained one by one from multiple particles of known types, and classify the multiple degrees of agreement based on the results of classifying the multiple principal component scores corresponding to each of the multiple particles of known types obtained by principal component analysis by type using a learning model using supervised learning, thereby identifying the type of the unidentified particle.

[0076] As a learning model using supervised learning, a support vector machine is preferably used, but other learning models may also be used. It is preferable to perform machine learning of a learning model such as a support vector machine using a plurality of principal component scores corresponding to a plurality of particles whose types are known and the types of each particle as training data. Furthermore, it is preferable that the discrimination device that discriminates the types of particles acquires training data from an external source.

[0077] On the other hand, classification may be performed using a learning model that uses unsupervised learning, such as clustering.

[0078] It is preferable that the discrimination device externally acquires multiple principal component spectra obtained by principal component analysis of multiple measurement spectra obtained from multiple particles of known types, and the results of classifying multiple principal component scores corresponding to each of the multiple particles of known types by type using a learning model.

[0079] Furthermore, when classifying the multiple principal component scores corresponding to each of the multiple particles whose types are known, it is preferable to divide a coordinate space containing coordinate points whose components are the multiple principal component scores into multiple regions using a learning model such as a support vector machine. For example, the coordinate points are two-dimensional coordinate points whose components are the first principal component score and the second principal component score. Depending on the division of the coordinate space, the multiple principal component scores are classified by particle type.

[0080] (Discrimination method by comparing two-dimensional mapping of correlation coefficients) In addition, the unknown particles to be measured can be identified by comparing the two-dimensional mapping of correlation coefficients obtained as described above for known particles with the two-dimensional mapping of correlation coefficients similarly obtained for the unknown particles to be measured to determine whether they are different or similar.

[0081] <Substrate for measuring the spectrum of fine particles> The substrate for measuring the spectrum of microparticles of the present invention has through-holes that penetrate from one side to the other side of the substrate, the through-holes being large enough to allow microparticles to pass through one at a time, at least a portion of the inner surface of the through-holes being made of a metal that exhibits plasmon resonance, and the through-holes having a gradient structure in which the width continuously decreases from one side of the substrate to the other side. As described above, by using a substrate with through-holes to obtain measurement spectra from each of n microparticles in a sample containing multiple microparticles, it is possible to obtain many measurement spectra more quickly and easily. Below, a substrate for measuring the spectrum of microparticles of the present invention having such through-holes will be described.

[0082] The substrate for particle spectrum measurement of the present invention can be applied to particles having a length of 10 nm to 5 μm, as well as to particles having a length greater than or smaller than that. However, the substrate for particle spectrum measurement of the present invention is preferably applied to particles having a length of 10 nm to 5 μm. In particular, it is more preferable to apply the substrate for particle spectrum measurement of the present invention to particles having a length of 10 nm to 5 μm when performing the method for generating spectral data of a particle sample of the present invention, the method for analyzing particles of the present invention, and the method for distinguishing particles of the present invention.

[0083] (substrate) The substrate is not particularly limited as long as it can be used for spectrum measurement. A sheet-like substrate may be used as the substrate. Examples of substrate materials include inorganic materials and organic materials such as polymeric materials. The substrate material is preferably electrically insulating. Examples include insulating materials used in the field of semiconductor manufacturing technology, such as silicon (Si), as well as glass, quartz, gold, silver, copper, aluminum, polydimethylsiloxane (PDMS), polytetrafluoroethylene (PTFE), polystyrene, and polypropylene. The thickness of the substrate is not particularly limited as long as it can have through-holes, and may be, for example, about 50 nm to 1 mm, preferably 50 nm to 500 μm, and more preferably 1 μm to 500 μm. The size of the substrate is not particularly limited as long as it can have through-holes.

[0084] The substrate may be produced by a known method, or a commercially available product may be used. The through-holes may be formed by known means such as etching or photolithography. The entire surface or a part of the surface of the substrate (for example, the inside or vicinity of the through-holes) may be surface-treated.

[0085] (Through hole) The substrate has through-holes that penetrate from one surface to the other. The size of the through-holes need only allow the target particles to pass through one at a time. Therefore, the size of the through-holes can be adjusted appropriately so that it is larger than the maximum cross-sectional area of ​​the target particles when they pass through, but not too large. For example, the equivalent circle diameter of the through-holes can be in the range of 10 nm to 5 μm. When the target particles are exosomes, the equivalent circle diameter of the through-holes can be 50 nm to 4 μm at their smallest point. In this specification, the equivalent circle diameter of a through-hole refers to the diameter of a circle with an area equal to the cross-sectional area of ​​the through-hole (a cross-section parallel to one surface or the other surface).

[0086] The shape of the through-hole is not particularly limited, but may be, for example, a truncated cone, a truncated pyramid, or the like. The shapes of the two bases (upper and lower bases) of the truncated cone are usually similar, but do not necessarily have to be. For example, one base of the truncated cone may be square and the other rectangular. It is also preferable to select the shape and width of the through-hole depending on the wavelength of the laser used and the refractive index near the through-hole. For example, when a 785 nm laser is used in an aqueous solution, through-holes of any of the above shapes can be formed in a substrate such that the equivalent circle diameter of the opening formed by the through-hole on one side of the substrate is 10 nm to 5 μm (preferably 50 nm to 3 μm, more preferably 50 nm to 1 μm, particularly preferably 50 to 500 nm, and even more particularly preferably 100 to 200 nm) and the equivalent circle diameter of the opening formed by the through-hole on the other side is 10 nm to 500 μm (preferably 50 nm to 1 μm). It is preferable that the through-hole has a shape having a gradient structure that continuously decreases from one side of the substrate to the other side. In this case, the equivalent circle diameter of the through-hole at the surface where the cross-sectional area of ​​the through-hole is largest is preferably 500 nm or more, and the equivalent circle diameter of the through-hole at the surface where the cross-sectional area of ​​the through-hole is smallest is preferably 50 nm or more, and more preferably 100 nm or more and 1 μm or less.

[0087] As described above, the substrate for measuring the spectrum of fine particles has through-holes penetrating from one side to the other side of the substrate, and the through-holes are of a size that allows the fine particles to pass through one by one, and at least the surface inside the through-holes is made of a metal that exhibits plasmon resonance, and the through-holes preferably have a shape that becomes continuously smaller from one side of the substrate to the other (a gradient structure). Hereinafter, this preferred form of substrate for measuring the spectrum of fine particles may be referred to as the substrate of the present invention. The inventors have found that by using the substrate of the present invention, particularly for measuring Raman spectra, the measurement results are 10 times better than when this substrate is not used (when a substrate with cylindrical through-holes is used). 8 We found that a sensitivity enhancement of 10 to the power of 8 was observed. One of the reasons for this enhancement is that the through-holes have a shape that continuously decreases in size from one side of the substrate to the other (a gradient structure, preferably a truncated square pyramid shape). This allows incident light to be focused along the gradient structure into the through-holes on the side where the through-hole width is smallest, enabling efficient use of the incident light. This enhancement is equivalent to the value generally defined as the enhancement factor (EF) in surface-enhanced Raman spectroscopy. This value was obtained by measuring 4-aminothiophenol molecules (Sigma-Aldrich Co. LLC), which are commonly used as a reference sample in Raman spectroscopy, and determining how many times the Raman scattering light is enhanced with or without through-holes.

[0088] By passing the particles one by one through the through-holes of the substrate of the present invention and measuring the particles at the positions of the through-holes, it is possible to continuously obtain the measured spectra for each particle. Furthermore, in the method for generating spectral data by measuring Raman spectra using the substrate of the present invention, it is possible to efficiently obtain spectral data obtained by highly sensitive measurement without labeling with fluorescent molecules or immobilization treatment on a metal surface.

[0089] The through-hole of the substrate of the present invention preferably has a truncated quadrangular pyramid shape, and the angle formed by the inner surface of the through-hole (the side surface of the trapezoid in the case of a truncated quadrangular pyramid shape) and the surface of the substrate is preferably 30 to 90 degrees, more preferably 30 to 60 degrees, and particularly preferably more than 45 degrees but not more than 60 degrees.

[0090] In the substrate of the present invention, at least the surface inside the through-hole is made of a metal that exhibits plasmon resonance. The substrate of the present invention is preferably made of a metal that exhibits plasmon resonance. Examples of metals that exhibit plasmon resonance include gold, silver, copper, aluminum, and the like, or a combination of two or more of these metals, with gold or silver being preferred, and gold being more preferred.

[0091] The substrate of the present invention may be made of an insulating material commonly used in the field of semiconductor manufacturing technology, from the viewpoint of ease of through-hole formation or cost, and may have a metal layer exhibiting plasmon resonance on at least the surface inside the through-hole of the substrate. Examples of insulating materials that can be used as the substrate include Si, Ge, Se, Te, GaAs, GaP, GaN, InSb, InP, Si3N4 (silicon nitride), SiO2 (silicon dioxide), or a combination of two or more of these. Of these, Si is preferred as the substrate, and a substrate body made of Si is more preferred. As with the substrate used in the examples, it is also preferred to use a material having a Si3N4 layer or SiO2 layer on the surface of a sheet-like substrate body made of Si.

[0092] The through-holes in the substrate of the present invention can be formed, for example, by etching. A preferred method for forming the through-holes is a manufacturing method for forming the through-holes in the form of a truncated quadrangular pyramid in the substrate for spectrum measurement of fine particles of the present invention by etching. Specifically, the manufacturing method for the substrate of the present invention shown below is preferred.

[0093] The substrate manufacturing method of the present invention includes performing anisotropic wet etching on one side of a substrate including a substrate body made of Si to form square pyramidal holes, and then performing wet etching on the other side of the substrate at an opposing position to penetrate the holes and form through-holes. An example of this procedure is shown in the examples described below. In the examples, anisotropic wet etching is performed on one side of a silicon wafer (a sheet-like substrate made of Si) to form square pyramidal holes, and then wet etching is performed on the other side at an opposing position. By forming through-holes using this manufacturing method, the holes can be penetrated and the size of the openings can be adjusted to a preferred range. The size of the openings can be adjusted to the desired size after forming a metal layer that exhibits plasmon resonance.

[0094] In the substrate of the present invention, the metal layer exhibiting plasmon resonance provided on at least the surface inside the through-hole of the base material may be formed only on the surface inside the through-hole, or may be formed on either one surface of the base material or on both surfaces of the base material. The metal layer exhibiting plasmon resonance may be formed on the surface of the base material by a method such as vapor deposition.

[0095] In the substrate of the present invention, the thickness of the metal layer exhibiting plasmon resonance provided at least on the surface inside the through-holes of the base material may be 50 nm or more, and preferably 50 μm or more. Since the thicker the better, there is no particular upper limit, but from the viewpoint of ease of production and cost, it is preferably 100 μm or less.

[0096] In the substrate of the present invention, the cross-sectional shape of the opening formed by the through-hole on either side of the substrate is preferably a rectangle with each side measuring 10 nm to 5 μm. In the substrate of the present invention, the width of the through-hole at either opening of the substrate where the through-hole width is smallest is preferably 50 to 500 nm, more preferably 100 to 500 nm. In particular, this region is preferably a rectangle with each side measuring 50 nm to 5 μm × 50 nm to 5 μm, more preferably 100 nm to 5 μm × 100 nm to 5 μm, and even more preferably 100 to 500 nm × 50 to 500 nm. In this case, the thickness of the substrate is preferably 50 μm to 500 μm. A substrate of this size enables efficient and highly sensitive measurement of exosomes, particularly exosomes with a particle size of approximately 50 to 150 nm.

[0097] The cross-sectional shape of the opening of the through-hole may have a minor axis and a major axis. The ratio (aspect ratio) of the length (major axis) of the cross-sectional shape of the opening of the through-hole to the width (minor axis) of the opening may be, for example, 99:1 to 50:50, preferably 90:10 to 50:50, and more preferably 80:20 to 50:50. Among quadrangles, the cross-sectional shape of the opening of the through-hole is preferably rectangular. One preferred example of the cross-sectional shape of the opening of the through-hole is, for example, a rectangle of 100 nm × 300 nm (aspect ratio 76:24).

[0098] (Method for obtaining a measurement spectrum using a substrate with through-holes) -Process for adjusting the microparticles to a dispersed state in a liquid- When measuring a spectrum using a substrate with through-holes, it is preferable that the particles are dispersed in a liquid. This is because the particles are easily guided to and passed through the through-holes. An electrolyte is preferred as the liquid in which the particles are dispersed. There are no particular restrictions on the electrolyte, but a solution in which an ionic substance is dissolved in a polar solvent is preferred. An example of an electrolyte is TE buffer (Tris-EDTA buffer: manufactured by Nippon Gene Co., Ltd.).

[0099] -The process in which fine particles pass through the through-holes- The step of passing the microparticles through the through-holes preferably includes a step of guiding the microparticles into or near the through-holes, a step of retaining the guided microparticles inside or near the through-holes, and a step of removing the retained microparticles from inside or near the through-holes. It is sufficient that the spectrum is measured in the retaining step. The retaining position may be inside the through-holes, but may also be near the through-holes, i.e., a portion of the substrate surface contacting the through-holes, depending on the properties of the microparticles to be measured and the spectrum to be measured. However, when measuring a spectrum that does not require the microparticles to be retained inside or near the through-holes during spectrum measurement, the retaining step may be omitted. It is preferable that the microparticles enter the through-holes through an opening on one surface side of the substrate in the through-holes provided in the substrate and exit the through-holes through an opening on the other surface side of the substrate.

[0100] The particles can be guided into or near the through-holes by using electrophoresis, dielectrophoresis, optical tweezers, Brownian motion, Coulomb interaction, or the like. In the present invention, it is preferable to pass the particles dissolved in a liquid (electrolyte) through the through-holes by one or more methods selected from the group consisting of electrophoresis, dielectrophoresis, optical tweezers, Brownian motion, and Coulomb interaction. A combination of multiple actions and methods may also be used.

[0101] The microparticles may be retained within or near the through-holes by, for example, adsorption within or near the through-holes. Examples of adsorption methods include, for example, a method in which, when the biological microparticles have a negative charge, the through-holes are positively charged to adsorb them using Coulomb force, a method using optical tweezers, and a chemical adsorption method. The chemical adsorption method utilizes reactive molecules on the microparticles and the nanostructure surface, and for this purpose, either one or both may be modified or coated in advance. For example, adsorption utilizing an antigen-antibody reaction may also be used. For example, the inside of the through-holes or the vicinity of the through-holes may be modified with a group that binds to an antibody that recognizes molecules present on the microparticle surface as an antigen, and further adsorption may be performed using the antibody (see, for example, Patent Document 3). [Patent Document 3] International Publication No. 2018 / 221271

[0102] The retention time of the microparticles may be determined depending on the microparticles to be measured and the properties of the spectrum to be measured, and can be selected, for example, from the range of 0.00005 seconds to 100 seconds. The upper limit of the preferred range of the retention time of the microparticles is more preferably 10 seconds or less, particularly preferably 1 second or less, and even more particularly preferably 0.1 seconds or less. The retention step as described above may be omitted, and for example, measurement may be performed while the microparticles are slowing down inside or near the through-holes compared to the movement speed of the microparticles up to the entrance of the through-holes.

[0103] Fine particles can be removed from inside or near the through-holes by applying discharge or reverse bias (in the case of Coulomb force), turning off light (in the case of optical tweezers), or by desorption reaction (in the case of chemical adsorption).

[0104] The movement of the microparticles and their retention within or near the through-holes can be adjusted under microscopic observation. Measurements can be performed by observing the movement of each microparticle individually under microscopic observation. Alternatively, in a method that combines electrophoresis and optical tweezers, the above process involving the movement of microparticles can be automated by adjusting the concentration of the microparticles in the sample, the electrolyte used, the on / off cycle of the light, and other factors. For example, the ionic current flowing through the through-hole decreases when a microparticle passes through the through-hole. Therefore, by observing this ionic current, the timing of the passage of the microparticle through the through-hole can be determined. Therefore, by using this ionic current as a trigger signal, the on / off control of the measurement can be automated. Electrodes for observing the ionic current may be placed on the sidewall of the through-hole or immediately adjacent to the through-hole.

[0105] -Repeated spectrum measurements- After spectrum measurement, the measured particle is removed from inside or near the through-hole, and then the next particle is introduced into or near the through-hole and measured. This procedure is repeated, thereby efficiently obtaining a large number of measurement spectra based on one particle (e.g., 20 or more, 30 or more, 40 or more, 50 or more, 60 or more, 70 or more, 80 or more, 90 or more, 100 or more). The time required to complete obtaining measurement spectra from n particles is preferably 0.00005 to 100 seconds, converted into the average time per particle. The upper limit of the preferred range of the time required to complete obtaining measurement spectra from n particles is more preferably 10 seconds or less, particularly preferably 1 second or less, and even particularly preferably 0.1 second or less.

[0106] - Acquisition of measurement spectrum - The measurement spectrum is obtained by detecting a signal generated by applying an external stimulus to a particle inside or near the through-hole. When the through-hole has a shape with a gradient structure in which the size continuously decreases from one surface (surface A) of the substrate to the other surface (surface B), it is preferable to obtain the measurement spectrum by a procedure that includes irradiating the through-hole with light from the side with the larger opening, from the viewpoint of further enhancing the intensity of the Raman scattered light. In other words, it is preferable that the external stimulus is applied from the side with the larger opening (surface A). For example, it is preferable to irradiate the laser from the side with the larger opening of the through-hole. Furthermore, it is preferable that the size of the through-hole opening on the laser irradiation side is larger than the laser spot diameter.

[0107] It is also preferable that detection of signals for spectrum measurement is carried out from the surface having the larger opening (surface A).

[0108] During spectral measurement, it is preferable to use a confocal lens and an objective lens in combination with a light source to irradiate and focus external light from the light source onto the microparticles inside or near the through-holes. It is preferable that the microparticles are measured in a liquid, and the objective lens is preferably an immersion type. Signal detection can also be performed via the objective lens. The distance from the objective lens to the through-hole is preferably the same as the working distance of the objective lens. In the device used for spectral measurement, optical filters, half mirrors, etc. can be used with reference to known techniques to suppress background signals and achieve a higher S / N ratio.

[0109] <Devices and equipment for measuring the spectrum of fine particles> The substrate of the present invention or other substrates for particle measurement having through-holes can be used as components of a device for measuring particle spectra. For example, a spectrum measurement device can be configured to include a substrate, a first liquid tank having an inner wall including at least a portion of the through-hole on one side of the substrate, and a second liquid tank having an inner wall including at least a portion of the through-hole on the other side of the substrate. With this configuration, particles in a liquid sample can be guided into or near the through-hole using electrophoresis or optical tweezers, as described above, and passed through the through-hole. In particular, among such configurations, a device for measuring particle spectra using a "substrate for measuring particle spectra of the present invention" as a substrate is referred to as a device for measuring particle spectra of the present invention (hereinafter also referred to as a device of the present invention). Furthermore, an apparatus for measuring particle spectra of the present invention can be configured using the device of the present invention. The apparatus for measuring particle spectra of the present invention (hereinafter also referred to as an apparatus of the present invention) includes the device of the present invention, a guide unit for passing particles one by one through the through-holes, a light source, and a detection unit for measuring the light (signal) generated when light from the light source is irradiated onto the particles in the through-holes to obtain a measurement spectrum. Preferred embodiments of the device of the present invention and the apparatus of the present invention are described below.

[0110] (First liquid tank, second liquid tank) When measuring particles using this device of the present invention, the particles are guided from a first liquid tank to a through-hole, and after measurement, they move from the through-hole to a second liquid tank. If the substrate in the device has through-holes that become continuously smaller from one side of the substrate to the other, it is preferable that the liquid tank on the side with the larger opening be the first liquid tank, and the liquid tank on the side with the smaller opening be the second liquid tank.

[0111] The material for forming the first liquid tank and the material for forming the second liquid tank are not particularly limited as long as they are capable of forming a liquid tank filled with a liquid (such as an electrolyte), but they are preferably made of an electrically and chemically inert material, such as glass, sapphire, ceramic, resin, rubber, elastomer, Si3N4 (silicon nitride), SiO2 (silicon dioxide), and Al2O3.

[0112] The first and second liquid tanks are formed to sandwich the substrate and are configured so that microparticles introduced into the first liquid tank can pass through the through-holes and move to the second liquid tank. The first and second liquid tanks may be provided with holes for filling or discharging the sample solution or electrolyte, or holes for inserting electrodes and / or leads for electrophoresis.

[0113] The electrodes can be made of known conductive metals such as silver / silver chloride, aluminum, copper, platinum, gold, silver, and titanium. The two electrodes are placed in the first and second liquid tanks, respectively, and direct current is applied to move the microparticles. The applied voltage can be determined depending on the conditions, such as the type of microparticles to be moved, and can be set, for example, between 0.01 V and 1.5 V.

[0114] The electrode placed in the first liquid tank may be connected to a power supply and earth via a lead. The electrode placed in the second liquid tank may be connected to an ammeter and earth via a lead. The connection positions of the power supply and ammeter may be interchanged between the first liquid tank side and the second liquid tank side, or the power supply and ammeter may be provided on the same electrode side.

[0115] There are no particular limitations on the power supply as long as it can pass a direct current between the electrodes. There are no particular limitations on the ammeter as long as it can measure the ion current generated over time when a current is passed through it. If necessary, a noise removal circuit, a voltage stabilization circuit, etc. may be provided.

[0116] The spectral measurement of the particles may be performed using an apparatus including a means for applying an external stimulus to the particles (light source) and a means for detecting a signal (detection unit). When the spectral measurement of the particles is performed using a substrate having through-holes, the apparatus preferably further includes a means for passing the particles through the through-holes one by one (guiding unit).

[0117] (Means for applying external stimuli to particles (light source)) The means for applying the external stimulus to the microparticles may be a light source. A preferred light source is a laser. When the measured spectrum is a Raman spectrum or a fluorescence spectrum, the means for applying the external stimulus to the microparticles may be a laser light source.

[0118] The device of the present invention preferably includes an inlet for irradiating the through-hole with a laser light source.

[0119] (Means for detecting a signal (detection unit)) The detection unit that detects the signal, for example, measures light generated by an external stimulus to obtain a measurement spectrum. Examples of light generated by an external stimulus include Raman scattered light and fluorescence generated when light from a laser light source is irradiated onto a particle. The detection unit includes, for example, a spectrometer for spectrally dividing the light and a detector (light-receiving device) for detecting the light. It is more preferable that the detection unit includes a spectrometer for spectrally dividing the Raman scattered light and a detector for detecting the spectrally divided Raman scattered light. Furthermore, in the device of the present invention, it is more preferable that the measurement spectrum is a Raman spectrum, the light source is a laser light source, and the detection unit includes a spectrometer for spectrally dividing the Raman scattered light and a detector for detecting the spectrally divided Raman scattered light.

[0120] (Means for passing fine particles one by one through the through-holes (guiding section)) The means for passing the microparticles one by one through the through-holes is not particularly limited, and examples thereof include the means described in the preferred embodiment of the process for passing the microparticles through the through-holes of the substrate of the present invention. Examples include electrophoresis means and optical tweezers means. The electrophoresis means may include a first liquid tank filled with an electrolyte solution on at least the surface including the through-holes on one side of the device as described above, a second liquid tank filled with an electrolyte solution on at least the surface including the through-holes on the other side of the substrate, a first electrode formed in the first liquid tank (which can be placed in the liquid to be filled), a second electrode formed in the second liquid tank (which can be placed in the liquid to be filled), and a power source for applying a voltage to the first and second electrodes.

[0121] (analysis methods, etc.) The particle spectrum measurement device may include analysis means for carrying out the particle analysis method described above (such as a program for creating principal component analysis or correlation coefficient mapping); control means such as a general-purpose CPU or processor for executing the analysis means; storage means for storing the analysis means; recording means such as memory or RAM for recording the obtained measurement spectrum and other data; input means such as a mouse or keyboard; output means; display means such as a display; and other software or hardware. Some or all of these may be stored in a PC or a cloud system. [Example]

[0122] The present invention will be explained in more detail below with reference to examples. The materials, reagents, amounts and proportions of substances, procedures, etc. shown in the following examples can be changed as appropriate without departing from the spirit of the present invention. Therefore, the scope of the present invention is not limited to the following examples.

[0123] <Preparation of substrate> A 300 μm thick silicon wafer (E&M CO., LTD) with a (100) surface orientation and 50 nm silicon nitride (SiN) film on both surfaces (Figure 1A) was cut into 25 mm squares. One side of the substrate was covered with an etching-preventive metal mask with 400 μm × 400 μm square through-holes. Using a reactive ion etching system (RIE-10NR, SAMCO CO., LTD), only the 400 μm × 400 μm square area of ​​the SiN film exposed within the through-holes was removed, exposing the silicon surface. Similarly, a 1000 μm × 1000 μm square area of ​​the SiN film on the other side of the substrate was removed in the same manner, exposing the silicon surface. Next, the exposed silicon portion of the 400 μm × 400 μm silicon surface (Surface A) was selectively wet-etched using a potassium hydroxide solution (Wako Pure Chemical Industries, Ltd.) on a hot plate (Hot plate NINOS ND-1, As One Co., Ltd.) at 125 °C for approximately 3 hours. Then, the exposed silicon portion of the other 1000 μm square silicon surface (Surface B) was selectively wet-etched using a potassium hydroxide solution (Wako Pure Chemical Industries, Ltd.) on a hot plate at 125 °C until through-holes were formed in the silicon substrate. This process resulted in the formation of silicon through-holes with a shape that continuously decreases in size from one side of the substrate to the other (a gradient structure) (Figure 1B). Then, 100 nm of Au, a metal that exhibits plasmon resonance, was deposited on the top and bottom surfaces of the substrate by sputtering (SVC-700LRF, Sanyu Electronics Co., Ltd.) (Figure 1C).

[0124] Figure 2 shows a photograph of the surface A of the fabricated substrate and a scanning electron microscope photograph of the through-hole area of ​​the substrate. The opening of the through-hole was a rectangle measuring 100 nm x 300 nm. The angle between the inclined surface of the through-hole (the side of the truncated square pyramid) and the surface of the substrate was 54.7 degrees (Figure 1B).

[0125] In the configuration shown in Figure 3, liquid tanks (the first liquid tank and the second liquid tank, respectively) were formed on one side (surface A) and the opposite side (surface B) of the substrate fabricated above to form a device for measuring one particle. Both liquid tanks were formed from a hydrophilic resin so that the portion of the substrate surface containing the through-holes became part of the inner wall.

[0126] <Exosome sample preparation> Exosomes, biological particles with a lipid bilayer membrane, were used as the microparticle sample. (1) Culture supernatants from normal cells (TIG-3 (human lung-derived adherent cells); American Type Culture Collection Co., Ltd., USA) were collected and centrifuged at 3,000 g for 15 minutes at 4°C to remove dead cells and cell debris. This was used as the culture supernatant sample. (2) 20 mL of the culture supernatant sample was placed in an ultracentrifuge tube and centrifuged at 110,000 g for 80 minutes at 4°C. (3) Carefully removing the supernatant, taking care not to touch the inner wall of the tube with the pipette, the particles adhering to the inner wall of the tube were dispersed by multiple pipetting with 1 mL of PBS (Nippon Gene Co., Ltd.) filtered through a 0.22 μm filter. (4) 19 mL of PBS was added, and the tube was again centrifuged at 110,000 g for 80 minutes at 4°C. (5) During centrifugation, care was taken to avoid touching the inner wall of the tube with the pipette. The supernatant was removed, and the adherent material was thoroughly dispersed in 1 mL of PBS to obtain an exosome sample derived from normal cells.

[0127] (6) Exosome samples derived from liver cancer were prepared using the same procedure as for the preparation of exosome samples derived from normal cells, except that cells derived from liver cancer cells (HepG2 (human liver cancer cells); American Type Culture Collection Co., Ltd., USA) were used instead of normal cells.

[0128] <Raman spectroscopy of exosomes> The first and second liquid tanks of the device prepared above were filled with TE buffer (Tris-EDTA buffer, manufactured by Nippon Gene Co., Ltd.) as the electrolyte. Furthermore, silver / silver chloride electrodes were placed in both liquid tanks. The exosome sample prepared above was added to the electrolyte in the first liquid tank. This device was placed on the sample stage of a HORIBA Raman spectrometer (LabRAM ARAMIS), and the measurement system shown in Figures 4 and 5 was constructed. An immersion-type objective lens was placed above the first liquid tank so that the distance from the through-hole to the objective lens surface was 2 mm. Measurements were performed using a 785 nm laser.

[0129] By applying a voltage of 0.1 V to the electrode placed in the liquid tank, the exosomes in the first liquid tank were electrophoretically guided into the through-holes, and then passed through the through-holes one particle at a time. Optical tweezers (using a 785 nm laser for Raman measurements) were then used to adjust the speed at which the microparticles passed through the through-holes, adsorbing them to the through-holes and holding them there so that they would decelerate within the through-holes. A laser was irradiated from the side with the larger opening of the through-hole for 1 to 10 seconds per microparticle, and the laser was applied at a wavelength of 400 to 1600 cm. -1 Spectra up to 1cm -1 The Raman spectrum was obtained at the position of the through-hole, with a resolution of 100 kHz. After the measurement, the light was turned off and the particle was peeled off. After the held microparticle was removed from inside the through-hole, the next microparticle was measured. By repeating this process, multiple Raman spectra were obtained for each microparticle.

[0130] 100 microparticles were measured for the normal cell-derived exosome sample, and 155 microparticles for the liver cancer cell-derived exosome sample, and spectral data consisting of bundles of 100 and 155 measured spectra were obtained for each sample. In other words, according to the present invention, it was found that spectral data for analysis using microparticles with lengths of 5 μm or less can be generated from each microparticle. Figure 6 shows the result of overlaying multiple Raman spectra measured for the normal cell-derived exosome sample, and Figure 7 shows the result of overlaying multiple Raman spectra measured for the liver cancer cell-derived exosome sample. For the liver cancer cell-derived exosome sample, the peak at 1100 cm -1 It can be seen that signals are observed in the spectra of many fine particles in the vicinity.

[0131] <Data analysis using correlation coefficients> For the spectral data obtained for each sample as described above, the observed Raman scattering light intensity was normalized (the maximum value in each spectrum was set to 1), and then an N × M matrix was formed. N is the number of spectra (100 and 155), and M is the range from 400 to 1600 cm. -1 Measurement resolution: 1cm -1 The Raman spectrum points (all 1200) were divided by . For the Raman spectrum group (N) of each sample, the correlation coefficient of the intensity between two Raman shift points (Y and Y') was calculated using the following formula:

[0132]

number

[0133] The correlation coefficients were calculated for all combinations of YY' and the correlation coefficients were mapped. The results are shown in Figure 8. Figure 8 shows the results for an exosome sample derived from normal cells. In the correlation coefficient mapping shown in the figure, higher correlation coefficients are indicated by darker colors. In the results for the exosome sample derived from normal cells in Figure 8, -1Peaks showing a high correlation coefficient (0.5 or higher) are circled. These peaks move in tandem and are therefore thought to be signals from the same molecule. This group of peaks coincides with the peak of an integrin molecule (α5β1) (Non-Patent Document 5), and therefore this group of peaks is thought to be derived from an integrin molecule. [Non-Patent Document 5] Mustafa H. Chowdhury, et al., "Use of surface-enhanced Raman spectroscopy for the detection of human integrins," Journal of Biomedical Optics, Vol. 11, pp. 024004.

[0134] Furthermore, the results of mapping the correlation coefficient of the exosome sample derived from liver cancer cells (data (155 spectra) in Figure 7) are shown in Figure 9. -1 Near and 1435cm -1 The color at the intersection of the 1087cm and 1087cm peaks is darker, indicating that these peaks are linked. -1 The signal at 1435 cm is a peak obtained from non-phosphorylated proteins. -1 This signal was found to be derived from phosphorylated proteins (phosphorylated serine) because it was linked to nearby proteins.

[0135] <Principal component analysis> Principal component analysis was performed using the N × M matrix created in the same manner as above. The spectrum (eigenvector) of the first principal component obtained is shown in Figure 10 (exosome sample derived from normal cells). All peaks in the obtained spectrum were consistent with the data for integrin (α5β1).

[0136] Spectral data (N1: 100 spectra) obtained from the exosome sample derived from normal cells obtained as described above and spectral data (N2: 155 spectra) obtained from the exosome sample derived from liver cancer cells were used to prepare spectral data, which is a bundle of 255 measured spectra (N1 + N2). The number of data points of the spectra was set to 1200 (M), and principal component analysis was performed using a 255×1200 matrix based on this spectral data. The eigenvectors of the first to 100th principal components were obtained using a program created in Python. The results of the score plot of the first principal component (PC1) and the second principal component (PC2) are shown in Fig. 11. As shown by the boundary line in the figure, it was found that the measured spectra of exosomes derived from normal cells and the measured spectra of exosomes derived from liver cancer cells could be discriminated from the results of this plot.

[0137] <Principal Component Analysis of Raman Spectra of Exosomes Derived from X-Ray-Irradiated Cells> In the same procedure as above, the Raman spectra of exosomes derived from normal cells (TIG-3), exosomes derived from normal cells (TIG-3) irradiated with X-rays, exosomes derived from liver cancer cells (HepG2), exosomes derived from the blood of healthy individuals, and exosomes derived from senescent cells were measured and principal component analysis was performed. The results of the score plot of the first principal component (PC1) and the second principal component (PC2) are shown in Fig. 12. In Fig. 12, the black circular marks indicate exosomes derived from X-ray-irradiated normal cells, the light gray diamond marks indicate exosomes derived from normal cells, the dark gray square marks indicate exosomes derived from the blood of healthy individuals, the light gray triangular marks indicate exosomes derived from senescent cells, and the dark gray cross marks indicate exosomes derived from liver cancer cells.

[0138] As shown by the dashed line in Fig. 12, it was found that the measured spectra of exosomes derived from X-ray-irradiated normal cells and the measured spectra of other exosomes could be discriminated from the results of this plot. This means that the surface molecules of exosomes derived from X-ray-irradiated normal cells are significantly different from the surface molecules of other exosomes.

[0139] Exosomes derived from X-ray-irradiated cells contain fragmented DNA, and it is known that the proliferation of these DNA fragments within cells can cause the cells to become cancerous. Therefore, detecting exosomes containing fragmented DNA (X-ray-irradiated normal cells in Figure 12) can reveal the tendency of cells to become cancerous. This can be useful in preventive medicine, as it can diagnose changes in the body before they become cancerous.

[0140] <Raman spectrum measurement of silica fine particles> Silica microparticles, which are inorganic microparticles of PM2.5, were used as the microparticle sample. Using the same substrate as used in measuring the exosome sample, the Raman spectrum of silica microparticles (manufactured by Corefront Co., Ltd.) with a diameter of 100 nm was measured in the same manner, and spectral data consisting of a bundle of 100 measured spectra was obtained. In other words, according to the present invention, it was found that spectral data for analysis using microparticles with lengths of 5 μm or less can be generated from each individual microparticle. Figure 13 shows the results of overlapping the measured Raman spectra of multiple (N=100) silica microparticles. Figure 13 shows that the highly uniform silica microparticle sample showed a peak at 478 cm in the spectra of all microparticles. -1 ±5cm -1 Near and 506 cm -1 ±5cm -1 It can be seen that signals with maximum values ​​around this area are observed. Because these peaks are derived from silica (Non-Patent Document 6), it can be seen that the fine particles can be measured correctly, and highly reproducible data can be obtained. [Non-Patent Document 6] Kazunori Matsui, et al., "Raman Spectra of Silica Gel Prepared from Triethoxysilane and Tetraethoxysilane by the Sol-Gel Method", Journal of the Ceramic Society of Japan, Vol. 106, pp. 528-530.

[0141] By using the spectral data of silica microparticles obtained by the production method of the present invention as a new database obtained by the present invention, or by combining it with spectral data of other PM2.5 particles obtained by similarly applying the production method of the present invention to other PM2.5 particles and performing the above-mentioned principal component analysis, it is possible to identify the type of PM2.5 in the atmosphere and apply it to the analysis of air pollution conditions.

[0142] <Fluorescence spectrum measurement of gold nanoparticles> The particle samples were gold nanoparticles with a diameter of 20 to 100 nm, each of which had a gold coating on the surface of inorganic nanoparticles. Using the same substrate as used for measuring the exosome sample described above, the fluorescence spectra of gold nanoparticles trapped within the through-holes were measured using the same procedure (excitation wavelength: 532 nm). Figure 14 shows the measured fluorescence spectrum of one gold nanoparticle outside the through-hole and one gold nanoparticle inside the through-hole. Figure 14 demonstrates that measuring the fluorescence spectrum within the through-holes of the substrate according to the present invention increases the fluorescence from the gold nanoparticles. Note that the peak fluorescence wavelength of gold nanoparticles is below 800 nm, so this graph shows the long-wavelength tail of the fluorescence spectrum.

[0143] This application claims priority from Japanese Patent Application No. 2021-188402, filed November 19, 2021. The entire contents of the specification and drawings of that application are incorporated herein by reference. [Industrial Applicability]

[0144] The present invention is useful in various fields, such as determining the type of particulate matter, determining the possibility of disease, and analyzing air pollution conditions. [Explanation of symbols]

[0145] 1 board 2 through holes 3 Opening 4. Silicon 5. Silicon nitride film 6 Gold (Au) layer 7 First liquid tank 8 Second liquid tank 9 electrodes 10 Immersion objective lens 17 First reservoir filled with electrolyte 18 Second reservoir filled with electrolyte 20 Analysis methods

Claims

1. A method for generating spectral data for a particulate sample containing a plurality of particulates, comprising: a step of passing the plurality of particles through the through-holes of the substrate one by one while irradiating the through-holes with light, and acquiring a measurement spectrum from each of the plurality of particles; the through-hole has an inclined structure in which the width continuously decreases from one surface of the substrate to the other surface, At least a part of the inner surface of the through hole is made of a metal that exhibits plasmon resonance, The spectral data is a bundle of a plurality of the measured spectra. Generation method.

2. The method of claim 1, wherein the length of the microparticles is between 10 nm and 5 μm.

3. The production method described in claim 1, wherein in the process of acquiring the measurement spectrum, the microparticles dispersed in the liquid are moved into the through-hole by one or more methods selected from the group consisting of electrophoresis, dielectrophoresis, optical tweezers, Brownian motion, and Coulomb interaction.

4. The generation method according to any one of claims 1 to 3, wherein the measured spectrum is a Raman spectrum.

5. The generation method according to any one of claims 1 to 3, wherein the measured spectrum is a fluorescence spectrum.

6. A method for analyzing particles, comprising the step of performing statistical analysis of the spectral data acquired by the generation method according to claim 1.

7. The step of performing statistical analysis of the spectral data includes: forming a set of peaks having a high correlation coefficient among the plurality of measured spectra of the spectrum data; a step of comparing the obtained set of peaks with the spectrum of a known substance to identify at least one component contained in the microparticle; The analysis method according to claim 6 , comprising:

8. The step of performing statistical analysis of the spectral data includes: performing multivariate analysis on the plurality of measured spectra of the spectral data; a step of comparing the spectrum obtained by the multivariate analysis with a spectrum of a known substance to identify at least one component contained in the microparticles; The analysis method according to claim 6 , comprising:

9. A method for identifying unidentified particles, comprising: A step of acquiring a measurement spectrum of each of a plurality of particles A whose type is known and a measurement spectrum of each of a plurality of particles B whose type is known by the generation method according to claim 1; a step of performing a principal component analysis on spectral data including the measured spectra of the plurality of particles A and the measured spectra of the plurality of particles B, and determining an index for discriminating between the measured spectra of the particles A and the measured spectra of the particles B from the scores of two or more principal components; a step of obtaining a measured spectrum from each of one or more unidentified particles in a particle sample by the generation method according to claim 1; calculating scores of the two or more principal components for the measured spectrum of the unclassified particle; a step of comparing the scores of the unclassified particles with the index to perform classification; Including, The length of the unidentified fine particles is 10 nm to 5 μm. Discrimination method.

10. the unidentified microparticles are exosomes, The microparticle A is an exosome derived from a cancer cell, The microparticle B is an exosome derived from a normal cell, In the step of performing the discrimination, it is determined whether or not the exosomes as the undistinguished fine particles are derived from cancer cells. The method according to claim 9 .

11. A method for determining the presence or absence of exosomes derived from cancer cells in a sample derived from a body fluid containing exosomes, comprising: A step of generating spectral data consisting of a plurality of measurement spectra obtained from each of a plurality of exosomes in the sample by the generation method according to claim 1; Between the multiple measurement spectra, 1087 cm －１ ±5cm －１ A signal with a maximum at 1435 cm －１ ±5cm －１ calculating a correlation coefficient with a signal having a maximum value at determining that cancer cell-derived exosomes are present in the sample when the correlation coefficient is equal to or greater than a certain value; A determination method including:

12. An apparatus for acquiring spectral data of a particulate sample containing a plurality of particulates, comprising: a device having a substrate including a through hole; a guide portion for passing the fine particles one by one through the through holes; A light source and a detection unit that measures light generated when the light from the light source is irradiated onto the particle in the through-hole and acquires a measurement spectrum; Including, the through-hole has an inclined structure in which the width continuously decreases from one surface of the substrate to the other surface, At least a part of the inner surface of the through hole is made of a metal that exhibits plasmon resonance, When the light source irradiates light into the through-hole of the substrate, the guide unit passes the plurality of particles through the through-hole one by one, and the detection unit acquires a measurement spectrum from each of the plurality of particles; The spectral data is a bundle of a plurality of the measured spectra. Device.

13. The apparatus of claim 12 , wherein the measured spectrum is a Raman spectrum.

14. The apparatus of claim 12 , wherein the measured spectrum is a fluorescence spectrum.

15. A method for generating the material described in claim 1, wherein the upper and lower openings of the through hole have a rectangular shape.

16. A method for generating the structure described in claim 1, wherein the shape of the through hole is a truncated quadrangular pyramid.

17. The device described in claim 12, wherein the upper and lower openings of the through hole have a rectangular shape.

18. The device described in Claim 12, wherein the shape of the through hole is a truncated quadrangular pyramid.

Images