JP2006294014A - Analysis program, protein chip, method for manufacturing protein chip and antibody cocktail - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To efficiently proceed with proteomics analysis. <P>SOLUTION: Raw data including at least an electrophoretic image and a mass spectroscopic signal pattern by a mass analyzer having various different characteristics obtained by a proteome analysis method using tissues, cells and cultured cells derived from patients and disease model animals showing various clinical conditions, a protein ID identified based on the raw data and quantitative data are stored in a local DB and a proteomics DB. The proteomics analysis system converts the pieces of data stored in the proteomics DB into one language, fuses them together, retrieves other opened DBs accessible via a network to extract a specific protein group and a specific intracellular signal network. The clinical conditions are analyzed and information to the most sure medical treatment and drug design is obtained by predicting a function from these obtained results and performing verification by a unique protein chip. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention is most suitable for the disease onset mechanism and each patient by analyzing the biological personality of various disease subjects by pathological proteomics, and the detailed and accurate analysis at the protein level of the pathological condition and progress at each time point. It is related to an analysis program that is optimal for establishing reasonable therapeutic drugs and prevention methods. In particular, this information processing system is suitable for use in fields such as medicine, clinical examination, and drug discovery.

  Furthermore, the present invention relates to a protein chip and a method for diagnosing the progress of a disease state and a method for determining drug sensitivity using the protein chip.

  The concept of proteome analysis so far is mainly based on differential display detection and database search for proteins that have been separated, analyzed, and searched by appropriate comprehensive protein analyzers or DNA array methods such as DNA array methods. A common method is to identify and target proteins in a stand-alone manner, and to search for pathological conditions and progress in various diseases at the protein level.

  In addition, differential display analysis using two-dimensional electrophoresis and differential display analysis using LC-shot gun method in proteomics research cannot cover protein information in all tissue cells only by each result.

  Furthermore, in mass spectrometers that can obtain the amino acid sequence information of peptides used in the current analysis, various ionization methods (MALDI method, ESI method, etc.) have characteristics that are notable for identification depending on the peptide type. ) And different separation / detection methods, and only one analysis method using one mass spectrometer, or a single analysis methodology cannot cover all tissues and cells. It is necessary to analyze using an analyzer. The results output from analyzers in comprehensive analysis of proteins in tissues and cells are output in large quantities, and there are a wide variety of data types such as image data and Internet files such as HTML. Using these different languages Comprehensive analysis with high throughput was a major barrier to realization.

  With conventional technology, data from multiple analyzers cannot be integrated and analyzed, thus reproducing the most important functions of proteomics such as intracellular protein networks, post-translational modifications, and intracellular protein degradation products. In addition, mRNA information obtained by DNA array analysis, real-time PCR analysis, etc. used in transcriptome research is greatly related to protein information, but proteomic information and mRNA from the same sample Since it was difficult to integrate and analyze information, it was derived on a trial and error basis by investigating researcher's empirical knowledge and long-term literature search.

  Moreover, after obtaining the results of searching for molecules, it is necessary to confirm reproducibility in vivo again.

  In addition, in order to develop applications such as detection / diagnosis of specific gene mutations that are related to susceptibility to specific diseases and side effects on medicines, it is necessary to prepare content that indicates what probe is installed. There are many events that can be analyzed for the first time by a combination of multiple probes instead of a single one.

On the other hand, a brain tumor is mentioned as a tumor with a very poor prognosis. The present inventors have conducted analysis using various neuroproteomic techniques for brain tumors for the purpose of developing pathological mechanisms, therapeutic methods, therapeutic / preventive drugs, and clinical markers for predicting prognosis (for example, non-proteomics). Patent Document 1, Non-Patent Document 2). An oligoplastic glioma (anaplastic origodendroglioma / astrosytoma: AOG) is known to be sensitive to anticancer agents in malignant gliomas with a poor prognosis. An interesting finding of the commonly known AOG is the loss of a single allele on the short arm of chromosome 1 (1p36) and the long arm of chromosome 19 (19q13) (Loss
of Heterozygosity: LOH) (60% or more in grades 2 to 3) and those not showing 1p, 19qLOH. Although both have the same pathological findings, the group with LOH is highly sensitive to anticancer drug therapy and radiotherapy and generally has a good prognosis, while the type without LOH is not sensitive to these therapies. It is characterized by sensitivity and very poor prognosis. To date, there has been no report on the molecular proteins that cause differences in prognosis, as well as on the marker proteins that show the difference between the two types at the molecular level, and on the mechanism of drug resistance.
"Proteomic analysis of human brain identifies -enolase as anovel autoantigen in Hashimoto's encephalopathy", Ochi H, Horiuchi I, Araki N et al., FEBSLetters, 528 (1-3), pp197-202, 2002 “New signal transduction pathways revealed by pathological proteomics in brain tumors” by Rekie Araki, “New mechanisms of vital functions seen in proteomics” Experimental Medicine vol23 (7), Yodosha, pp1050-1058, 2005 Year

  As described above, the concept of conventional proteome analysis is to search the database for proteins and peptides separated by 2D-PAGE, 2D-LC, etc. using various mass spectrometers using the peptide mass fingerprint method or sequence tag method. By this, identification and quantification of proteins involved in pathological conditions are performed. However, since the analysis of the analysis device differs depending on the individual device, the obtained data is independent, and it is difficult to evaluate the analysis data of a plurality of devices in an integrated manner. For this reason, even if more than one piece of information is analyzed using different devices and more information about the protein is obtained, it is difficult to organically link it and interpret it. It has not reached.

  Therefore, there are difficulties in evaluating post-translational modifications required for proteomics analysis and the evaluation and reproducibility of interaction functions such as degradation products of intracellular proteins. In addition, even when a new conceptual method or apparatus is developed, it is difficult to combine and analyze the data.

  An object of this invention is to provide the analysis program which can advance a proteomics analysis efficiently.

  Furthermore, the present invention uses molecular information obtained by the above-mentioned proteomic analysis system and method to list molecular groups that can be used for monitoring disease progress, drug sensitivity, etc. Solved the problem of providing a system and method for profiling the expression pattern of the above-mentioned molecular groups on a natural micro-biological tissue / cell protein chip of a serum protein and creating a database to monitor the progress of the disease state and drug sensitivity more reliably. It was an issue that should be done.

  The present invention relates to a variety of analytical devices used in the field of proteomics research, protein information data of a large variety of data types obtained from analysis software, and protein molecular data derived from the field of genomics and transcriptome research. Is integrated into the proteomics database, and the results of functional analysis by data mining are stored in the post-functional analysis database. It is possible to distinguish high RAW data and perform molecular function analysis in a large amount and in a shorter time than before while ensuring data reliability.

An object of the present invention is to provide an electrophoretic image obtained by a proteome analysis method using tissues, cells and cultured cells derived from patients and disease model animals exhibiting various pathological conditions, and mass spectrometry using mass spectrometers having various characteristics. In a computer comprising a raw data including at least a signal pattern, a protein ID identified based on the raw data, and a database storing quantitative data, the computer includes:
Converting and merging data stored in the database into the same language;
Searching other publicly accessible databases accessible over the network;
It is achieved by an analysis program characterized in that a specific protein group is extracted, and these functions are predicted to analyze a disease state.

In one embodiment, in a computer with an MS spectrum database storing MS spectrum data obtained by applying a plurality of LC-MS shotgun methods, the computer includes:
Identifying a protein;
Storing the analysis result obtained by the identification of the protein in a database;
The step of merging the analysis result stored in the database with the protein interaction analysis system and the biological information integration platform by linking them via a network is executed.

In another embodiment, in a computer comprising a database storing logical spectra, said computer includes:
The raw data obtained by applying the stable isotope-labeled LC-MS shotgun method is based on the similarity between the peptide peak list created under the existing conditions and the theoretical spectrum stored in the database. Identifying a peptide; and
Determining a relative amount of protein containing the peptide from the difference in intensity of the pair of stable isotope-labeled pairs;
And classifying a protein group that specifically appears in a specific tissue / cell.

In yet another embodiment, mRNA quantitative expression analysis data by analysis of different properties, including DNA arrays and / or real-time PCR, and two-dimensional electrophoresis, LC-ESI-MS / MS shotgun method, And / or a computer comprising a database storing quantitative data by proteomics using different techniques, including the LC &#8212; MALDI-MS / MS shotgun method,
Each of the data stored in the database is subjected to an analysis using the same algorithm including a differential display.

In a more preferred embodiment, the computer further comprises:
The analysis data, which is the result of the analysis performed by linking the database with another different database and stored in the database on the computer, is based on the information stored in the other linked different database. Run the step to analyze.

In still another embodiment, in a computer comprising image data including 2D-DIGE, including image data including spot information and protein information including post-translational modification structures, the computer includes ,
The step of associating the spot in the image data with the protein information including the post-translational modification structure from the database and storing the association in the database is executed.

In a preferred embodiment, the computer is caused to reversely input the information of the created integrated database to the analysis platform of the analyzer and execute a reanalysis step. In a more preferred embodiment, the mRNA is a DNA array and / or Alternatively, real-time PCR, solubilized proteins can be performed simultaneously using a fluorescence-labeled two-dimensional electrophoresis including 2D-DIGE and / or a stable isotope labeling LC-MS shotgun method including iTRAQ. Centralize and analyze data from quantitative differential displays.

In another embodiment, in a computer having a database storing a molecular network, the computer includes:
Examining a molecular network in the database based on molecular integration data;
Extracting a pathological tissue / cell-specific signal cascade such as specifically activated or specifically deactivated in a pathological condition.

In yet another preferred embodiment, a computer comprising an Accession No identified based on raw data including mass spectrometry signal patterns from mass analyzers having different characteristics, and a database storing at least the intensity ratio, the computer includes: ,
Extracting an intensity ratio for the same Accession No. based on different mass analyzers stored in the database;
Integrating the values of the intensity ratio, and storing the integrated values in the database in association with the AccessionNo.

More preferably, the database further includes an Accession No based on data derived from the DNA array, and an Intensity ratio,
Extracting an integrated intensity ratio for the same Accession No and an intensity ratio for a DNA array stored in the database;
And normalizing the intensity ratio value and storing the normalized value in association with the AccessionNo in the database.

In another preferred embodiment, in a computer comprising a database storing image data including 2D-DIGE and including spot information, the computer includes:
Each spot information and a database connected to the computer, the other information database storing the known protein spot information, the spot information and the known protein information When the spot information matches, giving the spot information of the known protein as the spot information, and storing in the database,
A step of identifying a protein applied to the spot based on the second information of the spot, which is obtained by another method including analysis by a mass spectrometer and stored in the database when there is no match .

  In a preferred embodiment, the step of identifying a protein based on the second information is a further database connected to the computer, and stores information of the protein identified by the other method. Searching the database.

  The present invention also provides a method for simultaneously preparing mRNA and protein from the same solubilized tissue from specific sites of pathological tissues / cells.

  Furthermore, the present invention provides a very small protein 2D map, which is characterized in that a protein extracted from a living tissue or cell is directly supplied to micro two-dimensional electrophoresis (2DE). Furthermore, the present invention provides a natural micro-tissue cell protein chip that covers all proteins synthesized in tissue cells, characterized in that the prepared 2D map is electrically transferred to a probe and immobilized.

  In addition, all proteins on the natural micro-tissue cell protein chip, post-translationally modified proteins such as sugar chains and phosphorylation are pre-labeled with a fluorescent dye such as Cy, or stained with a specific fluorescent staining reagent, A method for imaging all protein spot information on a microchip is provided.

  Furthermore, a method for detecting antigens against autoantibodies in patient sera by reacting with patient sera showing various pathological conditions using natural micro-tissue / cell protein chips, and patient autoantibodies on the chip based on these results Providing a method for profiling positive spots and creating a database to monitor the progress of disease states and drug sensitivity.

  In addition, the present invention lists molecular groups that can be used for monitoring progress, drug sensitivity, etc. from the molecular information obtained by, for example, the above claims 1, 2, 3, 8, 9, etc. Using antibody cocktails against molecules, we can profile antibody reaction positive patterns on natural micro-biological tissues / cell protein chips of patient tissues / cells and serum proteins, and create a database for more accurate pathological progress and drug sensitivity. Provide a method to provide for monitoring.

  Another object of the present invention is to subject a protein sample obtained by labeling a protein extracted from living tissue or cells with a fluorescent dye to first-dimensional isoelectric focusing on a strip, and then isoelectric focusing. A micro-sized protein chip that covers all proteins synthesized in living tissue or cells after reduction and SDS conversion of the electrophoretic strip, followed by SDS-polyacrylamide electrophoresis in the second dimension. Is achieved.

  In a preferred embodiment, the protein chip has a size of 10 cm × 10 cm or less.

  In another preferred embodiment, the protein chip is immobilized on a protein sample transferred from an electrophoresis gel to a membrane.

  In a preferred embodiment, the method for producing a protein chip comprises (1) subjecting a protein sample obtained by labeling a protein extracted from a biological tissue or cell with a fluorescent dye to first-dimensional isoelectric focusing on a strip. And (2) a step of reducing the strip after isoelectric focusing and converting it to SDS, and then subjecting it to SDS-polyacrylamide electrophoresis in the second dimension.

  In another preferred embodiment, the method for producing a protein chip comprises (1) a protein sample obtained by labeling a protein extracted from a biological tissue or cell with a fluorescent dye, and a first-dimensional isoelectric point on the strip. Step for subjecting to electrophoresis, (2) Step for subjecting strip after isoelectric focusing to reduction treatment and SDS, and then subjecting it to second-dimensional SDS-polyacrylamide electrophoresis, and (3) SDS-poly for protein samples A step of transferring from the electrophoresis gel after acrylamide electrophoresis to the membrane.

  In another preferred embodiment, the method of imaging the protein sample as protein spot information on a protein chip includes detecting fluorescence derived from a fluorescent dye labeled on the protein sample on the protein chip.

  In the present invention, (1) it can be used for monitoring the progress of a disease state or drug sensitivity from molecular information obtained by causing a computer to execute the program according to claim 1, 2, 3, 8, or 9. Preparing a group of molecules and preparing an antibody cocktail against these molecules, (2) contacting the antibody cocktail with the protein chip according to any one of claims 14 to 16, and (3) a protein A method for monitoring the progress of a disease state or drug sensitivity, comprising profiling antibody reaction positive patterns on a chip and creating a database is provided.

In the present invention, (1) anti-GFAP antibody, anti-CD44 antibody, anti-p53 antibody, anti-prothrombin antibody, anti-α1-antitrypsin antibody, anti-ubiqutine
activating enzyme E1 antibody, anti-Erk antibody, anti-p-Tyr antibody, and a step of preparing an antibody cocktail containing at least two kinds of antibodies selected from anti-zyxin antibodies, (2) extraction from biological tissue or cells of brain tumor patients The protein sample obtained by labeling the obtained protein with a fluorescent dye is subjected to first-dimension isoelectric focusing on the strip, and then the strip after isoelectric focusing is reduced and converted to SDS. Contacting the antibody cocktail with a micro-sized protein chip covering all proteins synthesized in living tissue or cells obtained by subjecting to eye SDS-polyacrylamide electrophoresis, and (3) on the protein chip A method for monitoring the progress of a disease state or drug sensitivity of a brain tumor, comprising the steps of profiling a positive antibody reaction pattern of the Provided.

Furthermore, in the present invention, an antibody cocktail used in the method of claim 21 for monitoring the progress of a disease state or drug sensitivity of a brain tumor, comprising an anti-GFAP antibody, an anti-CD44 antibody, an anti-p53 antibody, Anti-prothrombin antibody, anti-α1-antitrypsin antibody, anti-ubiqutine
An antibody cocktail comprising at least two antibodies selected from activating wnzyme E1 antibody, anti-Erk antibody, anti-p-Tyr antibody, and anti-zyxin antibody is provided.

  ADVANTAGE OF THE INVENTION According to this invention, it becomes possible to provide the analysis system and method which can advance proteomics analysis efficiently.

  Furthermore, according to the present invention, by utilizing the molecular information obtained by the above-described proteomic analysis system and method, a list of molecular groups that can be used for monitoring disease progress, drug sensitivity, etc. is listed, To provide a system and method for profiling the expression pattern of the above-mentioned molecular groups on cells and serum protein natural micro-biological tissues / cell protein chips, creating a database, and monitoring the progress of disease states and drug sensitivity more reliably. Is possible.

  Embodiments of the present invention will be described below with reference to the accompanying drawings. FIG. 1 is a block diagram showing the overall configuration of a proteomic analysis system according to an embodiment of the present invention. As shown in FIG. 1, the proteomic analysis system includes a proteomics-based processing device group 12 and a genomics-based processing device group 14. The proteomics analysis system also performs integrated analysis for controlling data storage in the proteomics DB 16, function classification DB 18, and proteomics DB 16 that stores predetermined information among information from databases (DBs) provided in each processing device group. The system 20 has a data mining system 22 that controls data storage in the function classification DB 18. These are connected to each other via a data communication path 26 such as a bus, LAN, or WAN. In addition, data can be transmitted / received to / from the international DB 24. Although not shown, the proteomics analysis system includes a display device and an input device including a mouse and a keyboard, and can display images based on data read from various DBs included in the system. Necessary processing can be executed in response to an instruction or input from the operator.

  In the present embodiment, the publicly disclosed database of proteins is collectively referred to as “international DB”. As the international DB 24, “SwissnPlot”, “MSDB”, “GeneBank” and the like can be used.

  FIG. 2 is a block diagram showing the configuration of the proteomics-based processing device group in more detail. The first mass spectrometer 30 to the third mass spectrometer 34 are apparatuses that measure mass spectra of individual proteins and peptides such as MALDI-TOF-TOF and LC-MS / MS, respectively. The control / analysis units 40 to 44 connected to the first mass spectrometer 30 to the third mass spectrometer 34 respectively transmit protein control parameters for measurement to the corresponding mass spectrometer. And a function for analysis for obtaining a normal signal from the raw data by specifying a measurement point and a threshold of the raw data (RAW data).

  As a method for deriving a protein from spectrum data, public database information such as an international DB or a unique DB (for example, a local DB described later) stored in-house is used. Each of the first local DB 50 to the third local DB 54 has a function of storing raw data of a spectrum and a data file including data obtained by analysis. In order to be able to store a large amount of data, it may be provided separately from the control / analysis units 40 to 44.

  The 2D-DIGE measuring device 36 is a protein two-dimensional electrophoresis measurement device for proteomic analysis, and identifies a protein from the two-dimensional position of the protein separated on the gel from the characteristics of the molecular weight and isoelectric point of the protein. Used for. A control / analysis unit 46 connected to the 2D-DIG measurement device 36 takes in the image of the control of the 2D-DIG measurement meter and the gel pattern as a digital image, and outputs analysis data after image processing. The fourth local DB 56 stores gel pattern images and analysis data.

  FIG. 3 is a block diagram showing the configuration of the genomics-based processing device group in more detail. The microdissection measuring device 60 selectively cuts out important sites to be used for analysis of human tissue cells with an ultraviolet laser and collects DNA and RNA proteins. The control / analysis unit 70 connected to the microdissection measurement apparatus 60 executes control of a mechanism for collecting a cell slice and cell image analysis. The obtained image data and analysis results are stored in the fifth local DB 80.

  A DNA array is obtained by spotting a specific combination of nucleic acid sequences in a fixed pattern on a glass, nylon or plastic substrate. The DNA array measurement device 62 performs functional genome analysis by hybridizing with a labeled DNA or RNA sample using a DNA array. A control / analysis unit 72 connected to the DNA array measuring device 62 controls the DNA array measuring device 62 and analyzes the spot DNA, and outputs analysis data. Image data and analysis data of the DNA array are stored in the sixth local DB 82.

  The real-time PCR measurement device 64 is a device that monitors the amplification amount of PCR in real time and analyzes DNA. The control / analysis unit 74 controls the connected real-time PCR measurement device 64 and performs DNA analysis. The seventh local DB 84 stores the analysis result.

  The proteomics DB 16 displayed in FIG. 1 is connected to the first local DB 50 to the seventh local DB 84 via the communication path 26, and has a memory for collecting the raw data and analysis data in these DBs. Here, the collection may be realized by a configuration in which the proteomics DB 16 stores all or part of the raw data and analysis data in each local DB, or each of the raw data and analysis data. For the above, a pointer indicating the storage position in the local DB may be realized by a configuration in which the proteomics DB 16 stores the pointer.

  The integrated analysis system 20 can perform the analysis executed by the control / analysis unit of the above-described proteomics-based processing device group 12 or genomics-based processing device group 14, and the raw data stored in the proteomics DB 16 ( RAW data) and the analysis result can be stored in the proteomics DB 16. In addition, the integrated analysis system 20 is a laboratory information management system LIMS (Laboratory Information Management) that manages the deployment addresses of all data obtained in the proteomic analysis system (addresses in the local DB and the proteomics DB 16), the history of measurement data, and the like. It also has a System) function.

  The data mining system 22 can execute data mining to find meaningful correlations and patterns from a huge amount of data stored in the proteomics DB. The result of category classification, function classification, and the like by the data mining algorithm included in the program of the data mining system 22 is stored in the function classification database DB18. The data used for data mining can always be monitored by the LIMS function of the integrated analysis system 20.

  An example of processing executed by the proteomic analysis system configured as described above will be described below. In the drawing, the first local DB to the fifth local DB are also expressed as local DB1 to local DB5, respectively.

  FIG. 4 shows an outline of LC &#8722; a shotgun method using a stable isotope labeling method and a differential display method using two-dimensional electrophoresis among data profiling methods using a proteomics-based processor group 12. FIG.

  Using the third mass spectrometer 34 and the control / analysis unit 44, a plurality of samples are labeled with various isotope labeling reagents having different molecular weights or “in vivo labeling method” of amino acids containing isotopes. Quantitative difference analysis is simultaneously performed by a shotgun method using LC-MS / MS, and identification and quantitative data are stored in the third local DB 54. On the other hand, by using the 2D-DIGE measurement device 46 and the control / analysis unit 46, the same plurality of samples are analyzed using a differential display analysis method using two-dimensional electrophoresis, etc. The identification results such as point, molecular weight, quantitative, protein, post-translational modification) are stored in the fourth local DB 56 including the image data. These data and additional data are added and integrated into one database (proteomics DB 16) to form a proteomics integrated database.

  Regarding the technique using LC-MS / MS, in this technique, two pairs of labeled peptides are measured in a chromatographic format, and then a protein identification process is executed by MS / MS analysis. FIG. 5 is a flowchart showing the protein identification process. As a result of this processing, intensity (expression intensity) information indicated by “Light” and “Heavy” is obtained by the ICAT method and stored in the third local DB 54. The processing of FIGS. 4 and 5 will be described in detail later.

  The peak list and protein information after the noise processing and centroid processing by the control / analysis unit 44 are also stored in both the third local DB 54 and the proteomics DB 16. Alternatively, information is stored only in the third local DB 54, and pointer information of the location stored in the third local DB 54 is stored in the proteomics DB 16, and the search is performed via the proteomics DB 16. Then, the data analyzed by the isotope labeling method may be configured to be backtracked. The storage of these pieces of information in the local DB or the proteomics DB may be executed by the control / analysis unit 44 or may be executed by the integrated analysis system 20.

  The same applies to differential display analysis using two-dimensional electrophoresis. The control / analysis unit 46 dyes normal cells and diseased cells or pattern-synthesizes each two-dimensional image profile divided by the labeling reagent color, the spot address, and protein information on the spot together with the original image data. 4 in the local DB 56. The control / analysis unit 46 or the integrated analysis system 20 performs double buffering by storing the data stored in the fourth local DB 56 in the proteomics DB 16. Alternatively, the proteomics DB 16 stores the pointer information of the location stored in the fourth local DB 56.

  Hereinafter, the protein identification process (FIG. 5) will be described in more detail. As shown in FIG. 5, the mass spectrometer (for example, the third mass spectrometer 34) and the control / analysis unit (for example, the control / analysis unit 44) decompose the protein obtained from the sample with trypsin or the like. And subjected to mass spectrometry to obtain an MS spectrum. Next, when the raw data (RAW data) of the MS spectrum is obtained from the sample, the control / analysis unit 44 stores the MS table for the sample in a local DB (for example, the third local DB 54). In addition, the control / analysis unit 44 sets a threshold in accordance with an operator's input or the like, and generates a peak list (Peak List). The generated peak list file is also stored in the third local DB 54.

  Next, the control / analysis unit 44 selects the international DB 24 in accordance with an operator input or the like. The control / analysis unit 44 searches the RMF / sequence tag etc. using the selected international DB based on the data subjected to the peak list calibration, and quantitatively identifies the protein including the post-translational modification. .

  The control / analysis unit 44 stores the identified protein information and the peptide table in the local DB (third local DB 54). Further, the control / analysis unit 44 or the integrated analysis system 20 stores the same information as the information stored in the third local DB 54 in the proteomics DB 16. Alternatively, the control / analysis unit 44 or the integrated analysis system 20 may store pointer information indicating the location where the protein identification result is stored in the third local DB 54 in the proteomics DB 16. The control / analysis unit 44 and the integrated analysis system 20 may backtrack the MS search result in the third local DB 54, the peak list file, the protein search result using the international DB, and the like from the pointer information of the proteomics DB 16. it can.

  The data storage management according to this embodiment will be supplementarily described below. As described above, the MS spectrum acquired from the mass spectrometer is subjected to noise processing and centroid processing for peak detection, and then output as a peak list file and stored in the local DB. The local DB also stores large-capacity LC-MS / MS measurement raw data (binary data file). When raw data can be processed by the integrated analysis system, the binary data file of the raw data is transferred to the proteomics DB and stored therein. In addition, in order to manage the experimental data from which the final results are derived and the contents of analysis, it knows which host has the binary data file of raw data. The peak list file extracted from the raw data binary file and the protein information derived from the peak list using a public database such as the Mascot database are usually deployed on the host that stores the raw data binary data. Therefore, it is not easy to manage where the analysis result is stored in the local DB or stored in the proteomics DB and where the analysis result that led to the final conclusion is located. Therefore, in the present embodiment, a data storage management table as shown in FIG. 21 is provided so as to maintain the consistency of data scattered at a plurality of locations. The data storage management table stores, for example, information (host name, IP address, database name, holder name) that can specify the name of the data file used to derive the protein and the location where the file is stored. .

  The present invention is a method characterized by performing a plurality of quantitative differential displays and a DNA array method simultaneously on the integrated database, wherein the obtained various data are stored as a proteomics database and stored. Proteins are identified and quantified based on various data, converted to a language for integration on the same platform, integrated, and protein functions that appear specifically from protein interactive function analysis software and international network databases are predicted. An algorithm characterized by systematic analysis of pathological conditions is provided.

  Next, an example of data profiling by a genomics-based processor group will be described with reference to FIG. The microdissection device 60 and the control / analysis unit 70 store normal cells and diseased cells in the fifth local DB 80 together with the sample image information, with the mRNA information obtained using the dissection. Similarly, the real-time PCR measurement device 64 and the control / analysis unit 74 perform gene expression analysis using real-time PCR, and store PCR image information and mRNA expression analysis information in the seventh local DB 84.

  Further, when the DNA array measuring device 62 and the control / analysis unit 72 execute gene expression analysis using the DNA array, the control / analysis unit 72 selects the international DB 24 in accordance with an operator input or the like, An mRNA search by analysis is executed, and the RNA information, protein molecule information, and DNA array image data of the search result are stored in the sixth local DB 82.

  Similar to the storage of data in the local DB and the storage of data in the proteomics DB 16 included in the proteomics-based processor group 12, the protein molecule information obtained by the genomics-based processor group 14 is stored in both the local DB and the proteomics DB 16. It may be stored. Alternatively, only the pointer information indicating the file storage location may be stored in the proteomics DB 16.

  In the present invention, a specific dissection of pathological tissue is performed, and mRNA and protein are simultaneously prepared from the same solubilized tissue. For mRNA, DNA arrays and solubilized proteins are subjected to two-dimensional electrophoresis (for example, Fluorescent labeling method 2D-DIGE method, etc.) and stable isotope labeling (cICAT method, iTRAQ method, etc.) LC-MS shotgun method (for example, MALDI-TOF-TOF, LC-ESI / MALDI-QQ-TOF, triple quadruple) Using a polar LC / MS / MS, an ion trap type or other hybrid type LC / MS / MS, FT-MAS, etc., a plurality of differential displays for heterogeneous analysis can be performed simultaneously in parallel. 4). In addition, based on the results of these analyzes, analysis algorithms fused and linked to proteomics DB 16 and protein interaction analysis software, and data mining algorithms that automatically classify proteins that specifically appear in specific tissue organs (claims) (See Item 6), specific protein molecule extraction, protein function analysis and pathological proteomic analysis can be performed.

  An outline of the data mining process according to this embodiment will be described. Conventionally, it has not been possible to correlate the correlation between detected molecules between protein level and mRNA level. However, as shown in FIG. 8, when applied to the framework of the molecular signal network, the molecular signal linkage pathway becomes visible. In FIG. 8, when the molecular signal network at the protein level on the right side and the molecular signaling network at the mRNA level on the left side are overlapped, as shown in the “protein and mRNA” in the center, they are similar by overlapping. You can find sex.

  The molecular signal network that rises at the protein and mRNA level shown in FIG. 8 is a technique for extracting a signal network that rises in both protein and mRNA in a diseased tissue / cell sample. For example, even if the expression of protein and mRNA does not match, it is assumed that mRNA will be expressed as a protein in the near future, and a specific signal cascade can be extracted by joining both activated molecular networks.

  In this way, by obtaining a molecular signal network, in FIG. 8, it is possible to obtain information on a target network such as extraction of the most upstream molecule or the most downstream molecule indicated by “A”, treatment or drug discovery. Become.

  On the other hand, electrophoretic images such as 2D-DIGE stored in the fourth local DB 56 and the like are positioned in specific properties of individual proteins, but these are image data. Although integration is difficult, data can be unified by linking image data with numerical data of the proteomics DB 16, the function classification DB 18, and the local DB (see claim 6). For example, in the image data stored in the local DB 56, an area defined by a plurality of coordinates of a spot (for example, four coordinates representing a rectangle), a database type (for example, a proteomics DB, a function classification DB, etc.), This can be realized by storing the information indicating the storage location in the database in association with each other. By using the function classification DB 18 and the proteomics DB 16, it is possible to link disease proteomics data as shown in FIG. In this example, if a spot in the profile image of 2D-DAGE is selected, the protein information of the spot can be retrieved from the international database and displayed. As described above, information indicating the association between the coordinates of the spot area and the information in the international database may be stored in the local DB (for example, the fourth local DB 56).

  The data mining system 22 displays the measurement MS data and measurement nanoLC data acquired by the proteomics analysis system visually. This can greatly improve the confirmation effort of protein identification.

  FIG. 9 will be additionally described. When a certain protein spot in the profile image (including 2D-DIGE and post-translational modification special staining and antibody reaction spots) displayed on the display device of the proteomics analysis system is specified and clicked, The name of the protein, the accession number of various international DBs (NCBI, Swiss slot, gene bank, etc.) and these DBs can be linked directly. As described above, this can be realized by associating the spot area with the accession number or the like and storing them in the proteomics DB 16 or the like. In addition, raw signal data (M / Z, intensity, MS / MS information, post-translational modification, molecular weight, isoelectric point, amino acid sequence, etc.) by mass spectrometry of peptides obtained by extraction from gels, stable isotopes Link the raw data of the difference analysis using the body, the result of the difference analysis obtained from a method other than two-dimensional electrophoresis (LC shotgun, etc.) using the same material. In addition, use more detailed post-translational modifications, information on 3D structures and molecular functions such as intracellular localization, information related to pathological conditions, information related to inhibitors and activators of the molecule, and the antibody of the molecule. Intracellular localization image database and pathological pathology images of diseases caused by abnormalities in the molecule (MRI, morphological images of tissue sections, etc.) Is included. Based on this information, it is possible to develop a tissue / cell Natural protein chip or the like obtained by transferring this two-dimensional electrophoresis image to a PVDF membrane or the like.

For example, FIG. 19 is a diagram for explaining a supplement of a tissue / cell Natural protein chip. The “tissue / cell Natural protein chip” is uniquely named by the present inventors. Brain tissue sample tissue and cell natural by two-dimensional electrophoresis
It is an example of analysis of self-antigen using a protein chip. Natural that about 10,000 protein spots were transcribed by separating tissues and cells by two-dimensional electrophoresis and transferring them to a probe such as a PCDF membrane
Protein chips can be created. The patient's serum can be reacted with this, and the spot where the autoantibody reacts can be identified from the database shown in FIG.

  FIG. 20 is a diagram showing profiling of autoantibody-reactive protein spots in patient serum using a Natural protein chip. The example of the molecule | numerator identified using the method shown in FIG. 19 using the produced Natural protein chip | tip is shown. Example of identification and profiling of autoantibody target protein using Natural protein chip from sera of patients with Hashimoto encephalopathy, multiple sclerosis, and amyotrophic lateral sclerosis. This allows profiling of the patient's condition. By miniaturizing this, it can be used for clinical examinations as a simple Natural protein chip.

  Further, in the present invention, the integrated database unifies a plurality of analysis data and enables an exhaustive analysis of the original meaning, thereby making it possible to detect a minute amount that could not be detected or interpreted by the analysis of a single device. Data may be closed up and need to be verified again. To make this possible, the data analyzed in the integrated database is returned to the individual analysis software, and a processing algorithm is provided to perform the analysis again. (See claim 7).

  Also, in the proteomic analysis method characterized by analyzing the intracellular function of a specific tumor-related gene product, the DNA array method or stability of specific cells in which the target gene is deleted or mutated Produces a protein profile using the isotope labeling method and performs differential display with normal cells, but provides an algorithm that unifies the quantitative differential display of multi-factor data to enable quantitative analysis of changing proteins. As described below, when measured with different instruments (for example, a mass spectrometer), each signal information comes out with different parameters. Therefore, work to convert them into a unified language (unification by absolute quantification using immediate internal control, standardization of peptide molecular weight (M / Z), accession of identified molecules, etc.) Plan.

  Proteins identified from a plurality of mass spectrometers (for example, the first mass spectrometer 30 to the third mass spectrometer 34 shown in FIG. 2) are condition setting parameters based on the sensitivity of the apparatus and raw data (RAW data). In general, proteins identified from each mass spectrometry result (threshold setting, public database setting, etc.) do not completely match.

  FIG. 7 is a Venn diagram in which proteins obtained from each of the first mass spectrometer 30 to the third mass spectrometer 34 according to the present embodiment are classified. The protein P + belonging to two or more mass spectrometers is classified as a positive protein and stored in the function classification DB in pairs with intensity.

  In FIG. 7, for example, when cICAT (isotope labeling method 1) and MALDI-TOF-TOF (mass spectrometer 1) are combined, cICAT (isotope labeling method 1) and ESI-QQ-TOF ( When using a combination of mass spectrometers 2), and using a combination of iTRAQ (isotope labeling method 2) and ESI-QQQ (mass spectrometer 3), a differential display of exactly the same sample ( For example, integrated mining is performed on data of molecular groups that specifically increase in cancer tissue in comparison with cancer tissue. In such a case, since the molecular species that can be analyzed by the respective analysis methods are not good, the same molecule identified in common in all methodologies as shown in FIG. There are groups of molecules that are identified in this way, molecules that are specifically identified by a single methodology (this often includes false positives at the time of identification). Among these groups of molecules, the most reliable group of molecules is a group of molecules commonly identified in all methodologies, followed by the group of molecules identified in common by two or more methodologies. It is. For those identified with only one methodology, it is necessary to return to the mass spectrometry raw data (RAW data) to confirm its reliability, but it also includes clear positive findings. In other words, all the molecular groups are covered only by combining all of these methodologies.

  Proteins that could only be identified with a single mass spectrometer include false positive protein P #. Therefore, is a case like false positive protein P # positive by adjusting the condition setting parameters of the analysis software? Please search again. The procedure will be described with reference to the processing flow of FIG.

  FIG. 10 shows an algorithm for confirming retrospectively whether the data processing performed when classifying the identified protein in the function classification database is correct.

  In the example of FIG. 10, the identified proteins are subjected to data mining and cluster classification, and when the mining result determination is equal to or less than the criterion, return matching mining according to the present embodiment can be performed. If there is a case where the determination criteria are satisfied by data mining, profiling is performed in the function classification DB 18. Further, when the data of the raw data level (RAW level) is acquired in the re-search, it is additionally stored in the local DB and the proteomics DB 16. Return matching mining will be described later.

  The example of FIG. 10 will be described in more detail. The data mining system 22 shown in FIG. 1 derives a meaningful result as much as possible from protein information derived from each mass spectrometer or the like by a mining technique. The filter parameter information of the derived process is adjusted, and data mining is performed again. FIG. 10 (return matching mining algorithm) shows the process of performing this again by a flow.

  Box 1 “Confirmation determination of identified protein” in the flow is an algorithm entry part that is entered when the operator gives an instruction by operating the input device to the data mining system 22 to perform data mining again. .

  In box 2 “Change the setting conditions of the identified mass spectrometer and search for the protein again” in the flow, the control / analysis units 40 to 44 connected to the mass spectrometers 30 to 34 shown in FIG. Change the input parameter conditions such as public database settings from the previously set contents, and search for proteins again. By changing the threshold value, the contents of the peak list change, and proteins that could not be discovered before can be identified. In addition, a protein that could not be found because there is no information related to the peak list in one international DB (for example, public database A) is registered in another international DB (for example, public database B). May be solved by using.

  In the step of box 3 “Minning until the protein most closely matches those identified by other mass spectrometers” in the flow, the three mass spectrometers 30 to 34 shown in FIG. The parameter information is tuned to derive the protein information so that the total of the output protein becomes the maximum. In this step, false positive proteins can also be detected.

  In step 4 “Record to local DB if optimal protein identification is possible” in the flow step, mass spectrometry that measured and analyzed the analysis results together with the input parameter conditions such as threshold settings and public database settings that were re-searched Store in the local DB related to the meter and the control / analysis unit.

  In the box 5 “add record destination in proteomics DB” step in the flow, necessary processing is performed so that the analysis result and the setting parameter stored in the local DB can be referred to in the proteomics DB.

  In the box 6 “profiling the result of re-identification in the function classification DB” in the flow, if a meaningful function analysis result can be derived as a result of return matching mining, the result is profiled in the function classification DB 18.

  In the function classification DB 18, RAW data and analysis data obtained by running the return matching algorithm are profiled in addition to various data profiled at the time of the initial search.

  The return matching mining algorithm will be further described with reference to FIGS. 22 and 23. FIG. When the identified protein results do not come out as expected, there are cases where the contents of the public database used to draw conclusions and the threshold used for data filtering were judged to have occurred because they were not optimal. is there. The return matching algorithm shown in FIG. 10 is a processing flow when the protein is searched again after returning to level 3 in FIG. The proteomic result obtained by the re-search is stored in the structure tracking table as shown in FIG. 34 (that is, stored in the DB). If a plurality of items are registered in the same hierarchy, the one with the new registration date is given priority, but it can be used as long as past results are not deleted.

  Backtracking can be realized based on the tracking table of FIG. FIG. 22 shows a data flow until a proteomic result is obtained. When experimentation and analysis are performed by trial and error, the data is managed so as to be able to respond to the request of how far back to perform. In this embodiment, for example, the levels are classified into four stages, and a tracking table for managing data for each level is created.

  When analyzing protein expression information using an intermolecular network, attention should be paid to the handling of protein intensity.

  The results of mass spectrometry identify three groups of proteins, but there is no problem if the proteins contained in the common region (region A in the Venn diagram of FIG. 11) have the same intensity, but as shown in the table at the bottom of FIG. If they are different, some normalization processing is required using the correction formula.

  11 and 12 will be described. As a result of analysis by individual mass spectrometers, the intensity with respect to the reference is not an absolute amount, and therefore a normalizing process for comparison is necessary. FIG. 9 shows that the first mass spectrometer 1 (Mg1), the second mass spectrometer (Mg2), and the third mass spectrometer (Mg3) identify proteins from the same disease state sample. ”,“ Protein 2 ”, and“ Protein 3 ”, the normalization calculation is shown. The sum of the intensities of “Protein 1”, “Protein 2”, and “Protein 3” obtained by the mass spectrometer 1 is taken as the intensity total (Mg1_IT). Similarly, the normalization correction equation shown in FIG. 11 is obtained by calculating the other two intensity totals “Mg2_IT” and “Mg3_IT” and taking the arithmetic average value thereof. Since the normalization correction coefficient NI obtained by this equation is obtained for three proteins, α = NI / 3 is a normalization correction coefficient for one protein unit. By dividing the intensity of the protein derived by this coefficient α, the whole proteins found can be compared.

  The intensity of the identified protein is divided by the normalization coefficient NI and unified. If the protein ID and intensity are grouped by intensity interval and imported into molecular signal linkage analysis software (for example, Keymolnet), the protein level network of molecular signal linkage shown in FIG. 8 is constructed. In FIG. 8, a molecule indicated by ○ indicates that tissue cells in a certain pathological state are specifically increased, decreased, or translated after differential display analysis at the “proteins” protein level or “transscriptome mRNA” level. A group of molecules detected as being subjected to modification, and those linked by network analysis are indicated by arrows.

  In addition, mRNA obtained at the genomics / transcriptome level is normalized in the same manner as in proteomics, and imported into the molecular signal linkage analysis software at the mRNA level, the mRNA level network shown in FIG. 8 is constructed. .

  In addition, the proteomic analysis system according to the present embodiment can unify the data output from the heterogeneous analysis apparatus and perform a centralized process, which can be used for comprehensive analysis. Therefore, even when a new analysis device is developed or a new analysis method is developed, the output data can be accepted and incorporated into the integrated database.

In the present invention, information obtained using the present invention is applied as clinical data through the following six stages.
(1) The protein data acquired by the mass spectrometer and the protein data acquired by using other methods such as two-dimensional electrophoresis are unified and stored in the database.
(2) To facilitate image analysis by visualizing stored data.
(3) Merging the data of (1) and (2) (proteomic data) with the data of the DNA array.
(4) Extracting important signals and molecular groups based on the data and the visualized images.
(5) Producing and verifying a protein chip based on (1) to (4) above.
(6) Apply as clinical data based on the verification results.

  The item (1) relates to the matter described with reference to FIGS. 4, 5, 7, 10, 11, 12, 13, 15, 21, 22, and 23. Items (2) to (4) are items described with reference to FIGS. 4, 8, 13, 14, 21, 22 and 23, respectively, and FIGS. 6, 8, 13, 16, 17, 21, 22. And related to the matters described with reference to FIGS. 8, 13, 17, 18, 21, 22, and 23. In addition, the protein chip of item (5) relates to the matter described with reference to FIGS. 4, 9, 14, 18, 19, 21, 22, and 23.

  Hereinafter, the items (1) and (3) will be supplementarily described.

  As described above, even when the same protein is analyzed by a mass spectrometer, the data value is generally different depending on the apparatus. FIGS. 24A and 24B are diagrams showing the results of measuring a certain protein with two different mass analyzers. Data of measurement results obtained by the mass analyzer is stored in a DB (for example, the first local DB 50 shown in FIG. 2). In the examples of FIGS. 24A and 24B, there are two ratios (Intensity ratios). However, the ratio is not limited to this, and there may be three or more ratios.

  The ratio is usually obtained by mixing and analyzing a plurality of samples to be compared using stable isotope modifying reagents (ICAT, ITRAQ), and integrating the intensity ratio of the obtained identification peptides at the protein level and averaging them. , Marked with “accessionNo” of the protein. The above “accession No” is based on the “swiss plot-international protein database”. 25 is a diagram showing an example of data obtained by integrating the data shown in FIGS. 24A and 24B by the data integration processing shown in FIG. 26. FIG. 26 shows an example of the data integration processing. It is a flowchart to show.

  As shown in FIG. 26, in the data integration process, the integrated analysis system 20 reads data as shown in FIGS. 24A and 24B stored in the local DB (step 2601). Hereinafter, M1 [i] is a data record of the i-th row in the local DB storing the measurement result by the mass spectrometer 2, and M2 [i] is a data record in the local DB storing the measurement result by the mass spectrometer 2. The data record in the i-th row is assumed. Mp [i]. accession, “AccessionNo”, Mp [i]. relation1 is “ratio1”, Mp [i]. Let relation2 be “ratio2” in the data record in the i-th row. In this example, p = 1 or 2.

Next, parameters in the loop are initialized, and end conditions are set (steps 2602 to 2606). Unless the parameter j exceeds the number of M2 arrays, that is, the number of records of the measurement result by the mass spectrometer 2 (No in step 2607), the average of ratios is obtained for the mass spectrometer 1 and the mass spectrometer 2. Here, since “Ratio1” and “Ratio2” exist as a ratio in one record, the average of the two ratios (M1 [i] .ratio1 + M2 [j] .ratio1) / 2
(M1 [i] .ratio2 + M2 [j] .ratio2) / 2
Is calculated (step 2609).

  This calculated average R [k]. ratio1 and R [k]. In the ratio 2, accession R [k]. As accession, M1 [i]. An accession is given (see step 2608).

  R [k]. accession, R [k]. ratio1 and R [k]. ratio2 is stored in a database (for example, proteomics DB 16) as an integrated record. As a result, as shown in FIG. 25, for example, for (AccessionNo = P10809, P11142, P61247), data in which the “Intensity ratio” is averaged can be obtained.

  Similar integration is possible for DNA arrays and proteins. A DNA array is given an “Accession No” unique to each array. In the present embodiment, this is integrated with the “Accession No” of the protein. Extraction and normalization from protein data and DNA array data is realized by the following method.

  FIG. 27A is a diagram showing the result of the integration process shown in FIG. 26 performed on the protein data shown in FIGS. 24A and 24B, and is the same as FIG. FIG. 27B is a diagram showing data of a certain DNA array stored in the local DB (for example, the sixth local DB 82). FIG. 29 is a flowchart showing an example of processing for integrating data as shown in FIGS. 27 (a) and 27 (b). FIG. 28 is a diagram illustrating a processing result by the data integration processing.

  As shown in FIG. 29, the integrated analysis system 20 reads a data record for a protein and a data record for a DNA array from the DB, respectively (step 2901). The integrated analysis system 20 extracts the same “AccessionNo” record (step 2902) and normalizes the “Intensity ratio”. Normalization is realized as follows.

  Activated molecule group: The integrated analysis system 20 compares the values obtained as “Intensity ratio” (steps 2903, 2904), and if both are larger than “1” in both the protein and DNA arrays, The values are averaged (step 2905), and the average value is stored in the DB (eg, proteomics DB 16) as an integrated ratio of records relating to the “AccessionNo”. If only the value related to the protein is larger than “1” as a result of the comparison, the integrated analysis system 20 adopts the value of the protein and stores it in the DB as an integrated ratio (step 2906). On the other hand, when only the value of the DNA array is larger than “1”, the integrated analysis system 20 adopts the value of the DNA array and stores it in the DB as an integrated ratio (step 2907).

  Inactive molecule group: The integrated analysis system 20 compares the values obtained as the “Intensity ratio” (steps 2908 and 2909), and when both of the protein and the DNA array are smaller than “1”, both The values are averaged (step 2910), and the average value is stored in the DB as an integrated ratio of records relating to the “AccessionNo”. As a result of the comparison, if only the value relating to the protein is smaller than “1”, the integrated analysis system 20 adopts the value of the protein and stores it in the DB as an integrated ratio (step 2911). On the other hand, when only the value of the DNA array is smaller than “1”, the integrated analysis system 20 adopts the value of the DNA array and stores it in the DB as an integrated ratio (step 2912).

  In this way, for the protein data and DNA array data to which the same “AccessionNo” is attached, the “Intensity ratio” can be integrated and stored in the DB. In FIG. 8 and FIG. 17, the signal linkage is analyzed and merged with respect to the change in the expression of the molecule at the mRNA level and the change in the expression at the protein level. By applying the above normalization, it is possible to automate the merging of protein data and DNA array data.

  As shown in FIG. 30, the integrated analysis system 20 (or may be the control / analysis unit 46. The same applies hereinafter) is a spot list based on 2D-DIGE data stored in the fourth local DB 56 or the like. Is generated (step 3001) and matched with known spot information stored in the DB (for example, the international DB 24) (step 3002). For example, standard data by SwissPlot can be used as the known spot information. Next, the integrated analysis system 20 determines whether or not the protein spot is known, that is, whether or not it matches the reference spot (step 3003). A spot is a large number of individual proteins separated by two-dimensional electrophoresis, and known spot information includes an isoelectric point, a molecular weight, and an intensity (protein amount) obtained from the migration pattern of the spot. Therefore, the fact that the spots are matched means that the mobility of two-dimensional electrophoresis determined by the isoelectric point and the molecular weight is matched.

  If it is determined as Yes in step 3003, if the protein name or registration number of the reference spot is known, the information is acquired using SwissPlot (step 3004). The acquired information can be stored in, for example, the proteomics DB 16.

On the other hand, when the protein spot is not known (No in step 3003)
First, an analysis by a mass spectrometer is performed, and the data is acquired (step 3005 to step 3007). Each of the data obtained by mass spectrometry is a value obtained by measuring the mass of a plurality of peptides obtained by degrading a protein with trypsin, and the amino acid sequence of the peptide is estimated from information on the mass of the peptide. The integrated analysis system 20 combines the amino acid sequences of a plurality of peptides and searches the international DB 24 to identify proteins (step 3008).

  Depending on the individuality of the mass spectrometer, there are some that cannot identify amino acids of each peptide. Therefore, it is meaningful to perform protein identification advantageously by measuring with a plurality of mass spectrometers, integrating the measurement results. At this time, there is a method of obtaining the final data by unifying the data at the peptide level and a method of obtaining the data at the protein level. In this embodiment, the identification of the final protein is performed by unifying the data at the peptide level. Is going to be effective.

  FIG. 8 and FIG. 17 show examples in which the signal linkage is obtained for the expression change of the molecule at the mRNA level and the expression change at the protein level, and the obtained signal linkage is analyzed and merged. This will be supplemented with reference to FIG. First, mRNA of normal tissue and pathological tissue is taken out, and the expression level of mRNA is obtained using a DNA array (step 3101). The mRNA expression level data (for example, spot intensity) is stored in the sixth local DB 82.

  In the original data obtained from the DNA array, the molecular name is displayed with a unique AccessNo. Therefore, in order to merge with the data in the proteomic DB 16, it is necessary to convert it into an Accession No of an international DB such as SwissProt. The integrated analysis system 20 generates a list including the mRNA expression level of each molecule obtained from the DNA array and the unique Access No of each molecule changed to the Access No of an international DB such as SwissProt. (Step 3102).

  Further, the integrated analysis system 20 inputs the expressed protein level to the protein node of the genome molecule signal network (step 3103).

  On the other hand, for protein, the expression level of normal tissue and pathological tissue is derived from the MS level (step 3104), and the information is obtained by searching for the protein by, for example, Mascot search using MS or MS / MS. (Step 3105). The acquired protein information can be stored in the first local DB 50, for example. The integrated analysis system 20 extracts the protein information stored in the DB, and inputs the expressed protein level to the protein node of the signal network for proteomics (step 3106). In this way, a signal network based on the expression level of mRNA and a signal network based on the expression level of the protein can be obtained.

  The integrated analysis system 20 superimposes the signal networks (step 3107). Here, the integrated analysis system 20 can execute the processing shown in FIG. 29 to merge the expression change of the molecule at the mRNA level and the expression change at the protein level. That is, by referring to the intensity ratio of the corresponding AccessionNo, it is possible to merge by normalizing, and store the merged new value (normalized intensity ratio) in a DB (for example, proteomics DB 16).

  Next, the above item (2) facilitates image analysis by visualizing the stored data, and (4) extracts important signals and molecular groups based on the data and the visualized images. This will be explained supplementarily. FIG. 32A is a diagram illustrating an example of protein information obtained from a 2D-DIGE image and stored in a DB (for example, the fourth local DB 56). In the example of FIG. 32 (a), Protein1- * represents a protein (spot) obtained from normal DIGE, and Protein2- * represents a protein (spot) obtained from DIGE in a pathological state. In FIG. 32A, “Mr” is the molecular weight, “pl” is the isoelectric point, and “D” is the concentration. Prior to data mining, the integrated analysis system refers to protein data stored in the DB and calculates a Euclidean measure between the respective proteins. The Euclidean scale U is obtained according to the following formula, for example.

^{_{U = {(Mr i -Mr j}} ) 2 + (pl i -pl j) 2 + (D i -D j) 2} 1/2
FIG. 32B is a distance table showing the obtained Euclidean scale.

At the time of data mining, the integrated analysis system 20 executes, for example, the following processing to find a specific protein.
(1) Focusing on the first horizontal row of the distance table, a pair having a minimum distance is searched for. In the example of FIG. 32 (b), the protein with the shortest distance (the closest distance) to “Protein 1-1” is found. For example, if the value between “Protei1-1” and “Protein2-1” is the minimum value, a pair of (Protei1-1, Protein2-1) is created, and the pair of Euclidean measures (here, the smaller one) To be adopted).
(2) Next, paying attention to the second horizontal row of the distance table, a pair with the smallest distance is searched. For example, if a value between “Protein 1-2” and “Protein 2-2” becomes the minimum value, a pair of (Protein 1-2, Protein 2-2) is formed, and the Euclidean scale of the pair is determined.
(3) By repeating such processing and proceeding to the last row, all pairs and the scale of the pair are determined.
(4) Since the pair determined by the above-described processing is a comparison in a row, it is not always the minimum pair with other proteins. Therefore, for example, for (Protein1-1, Protein2-1) in (1), the Euclidean scale between one and the other “Protein *-*” is compared, and the other “Protein *-*” is compared. Search for “Proteinx-x” that minimizes the Euclidean scale.

  If the Euclidean scale of the pair of (Protein1-1, Protein2-1) is smaller than the pair including the third searched “Proteinx-x”, (Protein1-1, Protein2-1, Proteinx-x) ) Is newly grouped, and the number of groups is three. By this grouping, the protein paired with “Proteinx-x” is released from the pair with “Proteinx-x” and is isolated.

  In this way, a protein that does not belong to any pair or group is found by repeating the grouping with the group number 2 or 3 by processing similar to the generation of a dendrogram (dendrogram).

  FIG. 33 is a diagram schematically showing the extraction of the specific protein. For example, in the 2D-DIGE image of a normal tissue, there are two protein “Protein 1-1” and “Protein 1-2” spots, and in the 2D-DIGE image of a diseased tissue, three proteins “Protein 2-1”. ~ Consider that there is a spot of “Protein 2-3”. Information on these spots is stored in the DB. As shown in the dendrogram, the integrated analysis system 20 generates a pair of (Protein1-1, Protein2-1) and a group of 2 (Protein1-2, Protein2-2). On the other hand, if it is determined that “Protein2-3” does not belong to a pair or group, the integrated analysis system 20 determines that “Protein2-3” is a specific protein, and the information is stored in a DB (for example, proteomics). DB16).

  Further, the integrated analysis system 20 refers to the known spot information stored in the DB, compares the position and concentration of the position spot with the position and concentration of the protein determined to be a specific protein, and matches them. Then, the specific protein is identified. Thereby, the information of the identified protein can be added to DB.

  Furthermore, the present invention relates to a protein chip and a method for diagnosing the progress of a disease state and a method for determining drug sensitivity using the protein chip.

The analysis of the present invention is performed in four steps according to a protocol as shown in FIG. First, an example of a brain tumor (glioma: AOG) is shown as a pathological sample. After surgical removal of the tumor tissue, pathological examination is performed.
Classification based on the presence or absence of 19q LOH (anticancer drug sensitive / insensitive). If normal tissue is removed from the vicinity of the tumor sample, it is classified as a control. For these samples, analysis by two types of proteomics and analysis by a DNA chip using the same sample are performed. As a methodology for proteome analysis, 2D-DIGE method using two-dimensional electrophoresis, cICAT method using LC-MASS without electrophoresis, and iTRAQ method are performed simultaneously, and mass analysis of molecular groups that differ at the protein level is performed. Identify by analysis. At the same time, the mRNA level data obtained with the DNA chip is compared, and all the obtained data are integrated and used for network analysis using molecular analysis software. In addition, all data is made into a database. After performing various functional analyzes including molecular cluster analysis and network analysis from these integrated data, and extracting and verifying the most important molecular groups, application to clinical markers, treatment, prevention, drug development, etc. Connect. In this strategy, the naturally-developed natural proteins shown below were cocktailed with the proteins that are particularly characteristic of the extracted molecules and the antibodies against post-translational modification structures such as phosphorylation.
Profiling was performed using protein chip.

  ADVANTAGE OF THE INVENTION According to this invention, the protein chip of the micro size which covered all the protein synthesize | combined in the biological tissue or the cell is provided. In the protein chip of the present invention, a protein sample obtained by labeling a protein extracted from biological tissue or cells with a fluorescent dye is subjected to first-dimensional isoelectric focusing on a strip, and then isoelectric focusing It is characterized in that it is obtained by subjecting the subsequent strip to reduction treatment and SDS, and then subjecting it to a second-dimensional SDS-polyacrylamide electrophoresis. The minute size mentioned here is preferably a size having a size of 10 cm × 10 cm or less, for example, a size of about 5 cm × 5 cm. The protein chip of the present invention is characterized in that all the proteins synthesized in living tissue or cells are covered on the above-described micro-sized chip. The conditions of isoelectric focusing and SDS-polyacrylamide electrophoresis are not particularly limited as long as all proteins synthesized in living tissue or cells are covered on a microchip as described above. However, those skilled in the art can appropriately adopt the conditions described in the following examples or conditions equivalent thereto.

  In addition, the protein chip of the present invention may use the gel after two-dimensional electrophoresis as it is, but it is preferable to use a protein sample in which the protein sample in the gel is transferred to a membrane and immobilized. The type of membrane used here is not particularly limited, and a PVDF membrane or the like can be used. In the present invention, it is particularly preferable to use one having low fluorescence, low protein adsorption and high resolution.

  In the protein chip of the present invention, a protein extracted from a living tissue or cell is labeled with a fluorescent dye such as CyDye (Cy 5 or the like). Therefore, by detecting the fluorescence derived from the fluorescent dye, it is possible to image the protein on the protein chip as protein spot information on the protein chip.

  Furthermore, in the present invention, a molecular group that can be used for monitoring the progress of a disease state or drug sensitivity is listed from the molecular information obtained by the proteomic analysis method according to the present invention described above in the present specification, and these molecules are listed. An antibody cocktail against can be prepared. Next, the antibody cocktail can be brought into contact with the protein chip described above in this specification, and the antibody reaction positive pattern on the protein chip can be profiled and databased. This makes it possible to monitor the progress of the disease state or drug sensitivity.

Examples of the present invention include anti-GFAP antibody, anti-CD44 antibody, anti-p53 antibody, anti-prothrombin antibody, anti-α1-antitrypsin antibody, anti-ubiqutine
Activating wnzyme E1 antibody, anti-Erk antibody, anti-p-Tyr antibody, and antibody cocktail containing at least two kinds of antibodies selected from anti-zyxin antibodies, and immobilizing proteins extracted from living tissue or cells of brain tumor patients The antibody cocktail can be brought into contact with a micronized protein chip and the antibody reaction positive pattern on the protein chip can be profiled and databased. Thereby, it is possible to monitor the progress of the disease state or drug sensitivity of the brain tumor.

  As a result of diligent research, the present inventors have conducted a verification experiment of a novel analysis method by the following means, and found that the above problems are solved and the object is achieved. That is, the present invention is characterized in that a protein extracted from a living tissue or cell is directly supplied to micro two-dimensional electrophoresis (2D) to create a very small protein 2D map, which is electrically transferred to a probe. Thus, a biological tissue protein chip that covers all proteins synthesized in tissue cells by solid-phase formation is provided.

  In addition, using a biological tissue protein chip, it is possible to easily search for a molecule to which an antibody, a biological protein, a synthetic peptide, a drug, or the like binds.

  Furthermore, after obtaining a series of results using a biological tissue protein chip, it is possible to reproduce interaction functions such as post-translational modifications and intracellular protein degradation products without confirming reproducibility in vivo. It is possible to do that.

  In general, the more accurate the reproducibility of interaction required for a protein chip, the more important it is to reproduce the interaction of degradation products of intracellular proteins, and the present invention has been made based on the above circumstances. The purpose is not only to replace artificially synthesized protein chips completely with biological tissue protein chips, but also to pathological research using proteomics, that is, comprehensive research of cellular tissues at the protein level corresponding to genomic research. It is characterized by providing an excellent method.

  In this demonstration experiment, we created a proteome map of a normal human brain tissue solubilized protein, and at the same time, to search for the antigen protein of an autoantibody contained in the serum of an encephalitis patient with unknown cause associated with an autoimmune disease. A part of the tissue is transferred to a PVDF membrane after two-dimensional electrophoresis, and the biological tissue protein chip according to claim 1 is prepared and reactive to autoantibodies contained in serum by immunoblotting with patient serum. The target protein was screened and identified.

  In this demonstration experiment, at least 7 targets were detected in addition to the autoantibody target protein discovered to date.

  Specifically, unexplained encephalitis associated with autoimmune diseases has been reported in various ways, such as multiple sclerosis (MS), characterized by inflammatory demyelination of the central nervous system. Despite the focus on the relationship between immunity and the pathogenesis mechanism, exhaustive analysis of the target protein that can be an antigen of the patient's autoantibody and nervous system tissue cells is not common. In this experiment, in parallel with the creation of a proteome map of a normal human cerebral tissue solubilized protein, a part of the protein was simultaneously transferred to a PVDF membrane after two-dimensional electrophoresis, and self-immunized by immunoblotting with patient serum containing human autoantibodies. Screening and identification of target proteins reactive with antibodies were performed.

  In this system, to date, myelin protein has been discovered as the target for autoantibodies in MS patients. However, in the above experiment, at least 7 targets were detected, and the muscle ring atrophic lateral cord was detected. It was also discovered that four autoantibody-reactive proteins were present in cerebral white matter and gray matter tissue in the sera of patients with sclerosis (amyotrophic lateral sclerosis, ALS).

  Furthermore, recently in the sera of patients with Hashimoto's encephalopathy, which has been suggested to have anti-neuroantibodies, three target proteins were detected, and three “alfa-enolases” with different isoelectric points were identified as antigens. . As a result of analysis of Hashimoto encephalopathy, Hashimoto's disease without encephalopathy, encephalitis patients with or without autoimmune disease, and 100 normal human sera, autoantibodies against afa-enolase were detected in 100% of Hashimoto encephalopathy patients sensitive to steroid treatment This suggests that it may be used as a marker for early steroid therapy.

  In addition, since it is clear that the structure including post-translational modification is an antigen, this screening method is particularly effective compared to the conventional cDNA library expression screening (SELEX) method using E. coli. found.

Furthermore, in the proteome analysis experiment of cancer, for the purpose of elucidating various onset mechanisms of brain tumors, developing markers, etc., brain tumors were analyzed with particular attention to “anaplastic origo dendro gloma”.

  The "anaplastic origo dendro glioma" is generally known to have a type having a short allelic deletion LOH typical of the short arm of chromosome 1 and the long arm of chromosome 19, LOH There is a type that does not show the same, and pathologically shows exactly the same findings, but the type with LOH is sensitive to anticancer drugs and radiotherapy and has a very good prognosis, and conversely, the type without LOH It is insensitive to anticancer drugs and has a very poor prognosis.

  To date, there has been no information on the molecular groups that cause these prognostic differences and the mechanisms of drug resistance, as well as marker proteins that show these differences at the molecular level. Therefore, the following analysis was performed using these brain tumor samples.

  Brain tumor tissues were classified according to the presence or absence of LOH and anticancer drug sensitivity, and proteins and mRNA were simultaneously extracted from these samples, and analyzed by two types of proteomics and differential analysis using a DNA chip. After identifying all the molecular groups with differences, protein and mRNA expression level data were compared and subjected to signal linkage analysis using molecular network analysis software (see FIG. 13).

Hereinafter, FIG. 13 will be described. FIG. 13 is a diagram for explaining a differential display strategy for obtaining molecular integration information from a disease state sample. After surgical removal of brain tumor tissue, the grade and progress are analyzed from a pathological point of view, “genetic mutation”
Classification is based on the presence or absence of defects, deletions, sensitivity to anticancer agents, and insensitivity. Simultaneous extraction of protein and mRNA from these samples, differential analysis by several types of proteomics (fluorescent reagent labeling method, stable isotope labeling method, two-dimensional electrophoresis method, LC shotgun method, etc.) and DNA array measurement And differential analysis of mRNA according to real-time PCR. After identifying all the molecular groups with differences, protein and mRNA expression level data are integrated and mined, and subjected to signal network analysis using molecular analysis software. Make all data into a database. From these integrated data, various functional analyzes including molecular cluster analysis and network analysis are performed, and the most important molecular groups are extracted and connected to clinical markers, treatment, prevention, drug development, and other applications.

  In proteome differential analysis by 2D-DIGE, profiling of post-translationally modified proteins is mainly possible. Two solubilized proteins are labeled with a fluorescent reagent, mixed, then provided to two-dimensional electrophoresis, analyzed, and a total of about 3,500 spots, about 900 expression-enhancing protein spots in the insensitive sample About 800 reduced protein spots.

  Furthermore, when the phosphorylated protein was analyzed on the same gel using PRO-Q (registered trademark) diamond, which is a phosphorylated protein staining solution, and the same analysis was performed on a sensitive sample, the increased protein spots were 369 spots, 628 that decreased, and 12 spots that were particularly insensitive to phosphorylation were detected (see FIG. 14).

  FIG. 14 is a diagram showing a differential display by multiple fluorescence labeling two-dimensional electrophoresis from a plurality of disease state samples and a special fluorescence staining method for post-translational modification sites. A brain tumor tissue (anti-cancer drug insensitive) showing an example of a brain tumor sample (labeled with Cy5) and a normal part (labeled with Cy2) of the same patient are subjected to differential display by 2D-DIGE, and further phosphorylated protein Were stained with ProQ Diamond which specifically stains. The enclosed spots are proteins whose phosphorylation is enhanced or dephosphorylated in the tumor. The number of these detected spots is shown in the figure.

  In the identification of the detected molecule group, four spots including HSP90 were subjected to specific phosphorylation in the insensitive sample, and “Neurofilament-L” was regarded as subject to dephosphorylation. As a representative, five spots were identified.

  At the same time, analysis was performed by a derivatization method (cICAT (registered trademark) method), which is a proteomic analysis method that does not use two-dimensional electrophoresis. In the cICAT method, each of two samples to be compared is labeled with a cysteine residue of a protein by two labeling reagents of “light” and “heavy”, each having a molecular weight of 9 daltons, using stable isotopes. After labeling with different labeling reagents, both are mixed, fragmented with trypsin, and analyzed using mass spectrometry. Two pairs of labeled peptides that are always fractionated into the same fraction are distinguished by mass spectrometry due to the difference in molecular weight between “Heavy” and “light”, and the difference is quantified. At the same time, peptides are sequenced and proteins are identified by MS / MS analysis. Brain tumor samples were fractionated into 12 fractions starting from 100 ug each, and finally MS / MS analysis of about 7000 peptides was performed.

  A comparison between the insensitive and normal samples detected a group of proteins as markers that were reported to be elevated in glioma. When the difference from the sensitivity sample was observed, protein groups related to “cell cycle” and “proteolysis” were detected (see FIG. 15).

  FIG. 15 is a diagram showing an example of a differential display (brain tumor sample) by an LC-shot gun using a stable isotope labeling reagent for a plurality of disease state samples. Comparison of anticancer drug-insensitive brain tumor tissue with a sample of the same pathology but anticancer-sensitive tissue. The differential display by LC-shotgun using cICAT or iTRAQ shows the number of molecular groups specifically increased in the anticancer drug-insensitive brain tumor tissue, the number of decreased molecular groups, and part of the molecular names.

  Next, analysis using a DNA chip was performed using the same sample. About 20,000 kinds of human cDNA oligo chips by “CodeLink” were detected, and 287 increase and 217 decrease compared to normal, and 412 increase and 238 decrease compared to sensitivity were detected in insensitive samples. It was. As a result of comparing the protein group specifically identified by the proteomics method with the expression pattern of mRNA, the expression pattern of both was greatly different. For example, in the comparison of insensitive and normal samples, 103 elevated proteins On the other hand, 287 mRNAs, 41 molecules in common, and about 30% of the identified protein molecules were found to cross the mRNA molecules even in the reduced molecular group ( (See FIG. 16).

  Therefore, all the data of the molecular groups that increased at the protein and mRNA levels were input into the existing molecular network search software KeyMonet (registered trademark), and the linkage analysis of the identified molecular groups was performed (upper left and upper right in FIG. 17: Protein level: upper left, mRNA level: upper right, some of the analysis results for all networks are shown).

  FIG. 16 shows anti-cancer agent-insensitive brain tumor tissue vs normal tissue (MS vs N) of the same patient, anti-cancer agent-insensitive brain tumor tissue vs. the same pathology but anti-cancer 13 sensitive tissue (MS vs HL). FIG. 6 is a diagram showing the number of molecules identified at the protein and mRNA level.

  A group of proteins identified as those whose expression was specifically increased and decreased by the proteomic technique was compared with the expression pattern of mRNA detected from the same sample. The pattern of mRNA expression and protein expression are very different. Only about 30% of the molecules identified in proteins cross in common with mRNA expression molecules. Many of these molecules were also decreased at the mRNA level although they were elevated at the protein level.

  The reason for the difference is considered to reflect the time lag between the expression and stability of proteins and mRNAs in cells. The increase in the expression of mRNA is terminated when it is sufficiently reflected as a protein, and decreases. When the protein is cleared and insufficient, mRNA expression is induced and increased. In other words, if you cut only a certain time axis in the cell, some proteins are rising, mRNA is decreasing, some proteins are decreasing, but mRNA is rising. Is not always going on. Time course experiments that solve the mystery of this time difference, such as synchronizing the cells in the sample in minute amounts, are not possible.

  Whether the time cap can be filled can be explained by the following mechanism of increased molecular expression. That is, assuming that all mRNAs are expressed as proteins, if mRNA is excessively expressed on a certain time axis, the protein will increase in the near future even if the protein decreases at that time. Then, it should be possible to extract a specific activation signal network with the width of the time axis in the pathological sample by integrating and mining all of the rising molecular groups ignoring the temporal lag. FIG. 8 shows a schematic diagram of the simple concept. By performing such a signal network extraction operation, it is possible to determine the most upstream factor (A) and the most downstream factor (G) of the activation signal that could not be found only by analyzing each of the protein and mRNA alone. I can do it. Therefore, all the data of the molecular groups that increased at the protein and mRNA levels were input to Key Molnet (registered trademark), which is an existing molecular network search software, and the linkage analysis of the identified molecular groups was performed.

  When an mRNA expression molecule group is inserted into the skeleton of the protein increase signal map shown in the upper left of FIG. 17, there are very few molecule groups that show an increase in mRNA expression in common, and conversely, decreased molecules are occasionally observed. The protein and mRNA expression levels were found to be significantly different.

Therefore, the methodology for extracting a signal network that is elevated in both protein and mRNA in an anticancer agent insensitive sample was mainly considered as shown in the figure. For example, protein expression is up upstream of the signal and the downstream signal cannot be identified as down or altered. Conversely, mRNA expression has already sunk upstream of the signal, but is rising downstream. Even if these expressions do not match, it is suggested that a specific signal cascade can be extracted by merging the two activation molecules, and that common upstream and downstream factors can be determined. (See Figure 8)
As an example of network search results for molecules specifically elevated by protein and mRNA, the complement system, the cell cycle by cyclin and CDK, and the network of retinoic acid nuclear receptors were extracted (FIG. 17).

In the anticancer drug insensitive sample, when listing signal networks that are specifically increased compared to normal and anticancer drug sensitivity at the level of protein and mRNA including phosphorylation, Vascular neoplasia, vascular permeability, cell cycle, adhesion factor-related increase in cascades, and activation of specific nuclear receptor groups, transcription factors and their downstream molecules were observed (FIG. 18). )
In addition, molecular cascades that may become specific signal networks and therapeutic targets related to pathological conditions can be obtained only by comparing post-translational modifications by proteomics and changes in protein levels with changes in mRNA expression levels. It was found that can be extracted (FIG. 8).

  It is possible to detect specific clinical markers related to pathological conditions and autoantibody reactive molecules as therapeutic targets from sera of patients with autoimmune diseases with encephalitis by pathological proteomic analysis using biological tissue protein chips It became clear (FIG. 19, FIG. 20).

  FIG. 17 shows an example of a signal network in which molecular groups extracted by performing various differential display analyzes using the same tissue cell solubilized sample are similarly increased at both the protein level and the mRNA level. FIG.

Based on the hypothesis as described with reference to FIGS. 8 and 16, the data are integrated and mined, mainly for signal networks that are similarly elevated at both protein and mRNA levels in insensitive samples. An example is shown. An analysis example relating to expression control through retinoic acid receptor (RAR) is shown in the figure. bCatenin, MAPK, RAR, tTG, integrin (Int), collagen, laminin and the like are increased at the protein level, and mRNA is upregulated as ErbB2, e-cadherin, DHTR, MAPK, tTG, Int Furthermore, it can be seen that the downstream factors LF, PEPCK, p21CIp1, HOXD, CRBP1 and the like of RAR are also significantly increased. When both the protein and mRNA are elevated and the cascades before and after the core are viewed, an activation signal as indicated by bold arrows in the figure can be extracted. These findings are novel as one methodology for finding a new disease state-related signal cascade using a databased signal network based on experimental facts so far. FIG. 18 shows an anticancer drug for a brain tumor patient tissue. FIG. 5 shows a signal network that is similarly elevated at both protein and mRNA levels in an insensitive sample.

  Based on the hypotheses as described with reference to FIGS. 8 and 16, the data were integrated and mined, mainly to show signal networks that are similarly elevated at both protein and mRNA levels in insensitive samples. Listed up. In particular, in relation to drug sensitivity, enhancement of apoptosis-related cascades (AKT, Bad signal, etc.), vascular hyperplasia and vascular permeability, complement coagulation factor / vascular neoplastic factor system, cyclin Of signal cycle by G1 / S and G2 / M related molecule groups centering on the cell, adhesion factor group via integrin / cadherin, enhancement of degradation system, and retinoic acid receptor and androgen receptor (DHTR) related signal Signal enhancement via a transcription factor system such as a nuclear receptor such as was observed. In addition, specific signal networks related to pathological conditions and molecular cascades that are therapeutic targets can be extracted only by proteomics and post-translational modification, and by analyzing changes in protein levels with changes in mRNA expression levels. Recognize.

In this demonstration experiment, in the extraction of specific molecules, the protein function analysis and the pathological proteomics analysis, the simultaneous differential display of three modes described in claim 3 is performed, and the claim 6 is described. In addition, the integrated search data mining algorithm is linked to the proteomics database and protein interaction analysis software, and all the proteins expressed in the cell are comprehensively analyzed along the time axis, and how the total function is controlled. Elucidate what biological activity is carried out, and what changes in what intracellular protein group, when, where, and how normal cellular activity is disrupted, complicated diseases such as cancer Is organized in a systematic and systematic manner in each tissue / cell. That together with information of genomic analysis to obtain an exhaustive information to help in the development of rational therapies, cancer markers and prophylaxis, i.e. found a method that enables proteome analysis including all these.
Example 1:
(Method)
The brain tumor site of the brain tumor patient was treated as a brain tumor tissue sample, and the normal site near the tumor was treated as a normal tissue sample (several millimeters square), which was frozen in liquid nitrogen during the operation and then stored at -80oC. After confirming the tumor grade by pathological findings, these tissues were classified into the presence or absence of LOH on the first chromosome (1P) and chromosome 19 (19q), anticancer drug sensitivity and anticancer drug resistance.
Lysis those tissue samples
buffer (9.8M urea, 4% CHAPS, 0.5 mM EDTA, 1 mM DTT, 1 mM PMSF, 1 mM PEPSTATIN, 1 mM
leupeptin, 10 mM NaF, 2 mM Na ₃ VO ₄ , 1.5 mM Okadaic acid) and homogenated. The lysate was centrifuged (15,000
rpm, 10 min, 4 ° C.), and the insoluble matter removed was used as a tumor or normal brain homogenate sample.

About 100 μg of the sample is 2D
After desalting in a clean up kit (GE Healthcare Bio-Sciences Corp.) or by acetone precipitation, CyDye (Cy 5, GE
Healthcare Bio-Sciences Corp.) labeled the protein in the homogenate sample at a minimum. Dilute about 10 μg of Cy5-labeled protein with swelling buffer to 125
Add μL to the strip holder, add DryStrip (pH 3-11NL, 7 cm. Invitrogen) onto the solution, and perform first-dimension isoelectric focusing (1 &#12316; 3 hours) It was. After isoelectric focusing, DryStrip 0.1
M DTT-containing SDS equilibration buffer (50 mM Tris-HCl, 6M Urea, 30% glycerol, 2% SDS, 0.02%
It was immersed in Bromophenol Blue, pH 8.8) for 10 minutes for reduction treatment and SDS conversion. Furthermore, both ends of the Drystrip were cut into 5 cm length with a cutter, and SDS-polyacrylamide electrophoresis (Phast-gel
System, Pharmacia, and 5cmx5cmPAGE AIST developed micro high resolution gel, 20mA / gel, energized for about 1 hour). This is Natural
A protein gel chip was used.

The Natural Protein gel chip was electrically converted into a 5 cm x 5 cm low-fluorescence PVDF membrane (millipore,
immobilon-FL 0.45 μm) and used as a Natural Protein membrane chip. That is, soak the low-fluorescence PDVF membrane in methanol for about 10 seconds, and then transfer the buffer.
(25 mM Tris, 200 mM glycine, 10% methanol) was fully equilibrated. Using a semi-dry transfer device (Biocraft Co., Ltd.), the protein in the SDS-polyacrylamide gel is electrically transferred to the membrane equilibrated with the transfer buffer (approx. 140
mA, 50 minutes energized). Immediately after the transfer, the protein on the membrane was detected using a fluorescence scanner (Typhoon 9400, GE Healthcare Bio-Sciences Corp.).

After detection, immerse the low-fluorescence PVDF membrane in a blocking solution (0.5% BSA / 0.03% Tween20 / PBS), shake at room temperature for 10 minutes, and then wash with a washing solution (0.3% Tween20 / PBS). High sensitivity and high resolution PVDF membrane was washed (5 minutes x 3 times). Antibody cocktail solution (anti
GFAP pAb 1 / 500x Santa Cruz, anti CD44 pAb 1 / 1000x Ancell, anti p53 mAb 1 / 1000x
Santa Cruz, anti prothrombin mAb 1/1000 times Abcam, anti α1-antitrypsin mAb 1/1000 times
Abcam, anti ubiqutine activating wnzyme E1 mAb 1 / 500x Upstate, anti Erk pAb
1/1000 times Santa Cruz, anti p-Tyr mAb 1/1000 times Upstate, anti zyxin pAb 1/1000 times
abcam) (where pAb represents a rabbit polyclonal antibody and mAb represents a mouse monoclonal antibody) a low-fluorescence PVDF membrane was soaked, shaken at room temperature for 1 hour, and washed with a washing solution (5 minutes × 3 times) Later, the secondary antibody solution {anti
rabbit Ig-HRP linked F (ab ') ₂
Fragment (from donkey) 1 / 20,000 times or anti mouse Ig-HRP linked F (ab ') ₂ fragment (from sheep) 1 / 20,000 times, GE Healthcare Bio-Sciences Corp. Shake for 30 minutes. After washing with washing liquid (5 minutes x 3 times), ECL
Plus Western Blotting Detection Reagents (GE Healthcare Bio-Sciences Corp.) 1
It was immersed in mL and reacted at room temperature for 1 minute. Low fluorescence PVDF membrane sandwiched between transparent films and Hyperfilm (GE Healthcare Bio-Sciences
Corp.) was allowed to adhere (about 2 minutes) and exposed to light, and then the Hyperfilm was developed. At the same time, the fluorescent scanner (Typhoon 9400, GE Healthcare
Bio-Sciences Corp.) and imaged by staining antibody reactive fluorescent spots with pseudo colors and merging them onto Cydye detection images.
(result)
The results of profiling using antibodies of brain tumor tissue samples and normal tissue samples using Natural Protein membrane chips are shown in FIGS. FIG. 38 shows an actual sample pattern, and FIG. 39 shows a summary of patterns of each patient.

GFAP and p53 increased with phosphorylation in tumor samples, and ubiquitinase E-1, CD44, antitrypsin, and prothrombin were detected as patterns with obvious quantitative variations. When using tyrosin-phosphorylated antibody and serine / threonic acid staining method, a clear difference in response pattern between normal and tumor was observed. In addition, 50 and 40 positive spots were compared by anti-cancer drug susceptibility and insensitivity profiling with 5 types of antibody cocktails (GFAP, CD44, tyrosin phosphorylated antibody, Erk, Zyxin) using each patient sample. As a result, a common 30 pattern was detected. Type this
As a pattern, it is possible to list antibodies of proteins that are currently considered to be related to anticancer drug sensitivity, and to perform patterning using these antibody cocktails.

  From the above results, it was shown that this Natural Protein (membrane) chip can be used to distinguish between normal and tumor, anticancer drug resistance and anticancer drug sensitivity. In addition, since several types of antibodies can be used simultaneously, and the degree of reaction (quantitative change) can be quantified, it can be obtained only by quantification of tissue immunostaining and cancer markers used for cancer diagnosis by ELISA. A large number of molecular information including post-translational modification changes that could not be obtained in a few hours can be a useful diagnostic method applicable to all disease-related samples regardless of brain tumors. In addition, this method is useful as a clinical test methodology because the final data can be obtained by shortening the method that took two days or more to 5 hours at the shortest.

FIG. 1 is a block diagram showing the overall configuration of a proteomic analysis system according to an embodiment of the present invention. FIG. 2 is a block diagram showing the configuration of the proteomics-based processing device group in more detail. FIG. 3 is a block diagram showing the configuration of the genomics-based processing device group in more detail. FIG. 4 shows an outline of LC &#8722; a shotgun method using a stable isotope labeling method and a differential display method using two-dimensional electrophoresis, among data profiling methods using a group of proteomics-based processors. FIG. FIG. 5 is a flowchart showing protein identification processing according to the present embodiment. FIG. 6 is a diagram illustrating an example of data profiling performed by a genomics-based processing device group. FIG. 7 is a Venn diagram in which proteins obtained from each of the first mass spectrometer to the third mass spectrometer according to the present embodiment are classified. FIG. 8 is a diagram showing an example of a protein level network of molecular signal linkage. FIG. 9 is a diagram for explaining cooperation of disease proteomic data in the function classification DB. FIG. 10 is a diagram illustrating an example of a return matching mining algorithm. FIG. 11 is a diagram for explaining the normalization processing of protein intensity. FIG. 12 is a diagram for explaining the normalization processing of protein intensity. FIG. 13 is a diagram for explaining a differential display strategy for obtaining molecular integration information from a disease state sample. FIG. 14 is a diagram showing a differential display by multiple fluorescence labeling two-dimensional electrophoresis from a plurality of disease state samples and a special fluorescence staining method for post-translational modification sites. FIG. 15 is a diagram showing an example of a differential display (brain tumor sample) by an LC-shot gun using a stable isotope labeling reagent for a plurality of disease state samples. FIG. 16 shows anti-cancer agent-insensitive brain tumor tissue vs normal tissue (MS vs N) of the same patient, anti-cancer agent-insensitive brain tumor tissue vs. the same pathology but anti-cancer 13 sensitive tissue (MS vs HL). FIG. 6 is a diagram showing the number of molecules identified at the protein and mRNA level. FIG. 17 shows an example of a signal network in which molecular groups extracted by performing various differential display analyzes using the same tissue cell solubilized sample are similarly increased at both the protein level and the mRNA level. FIG. FIG. 18 is a diagram showing a signal network that is similarly elevated at both protein and mRNA levels in an anticancer agent-insensitive sample of brain tumor patient tissue. FIG. 19 is a diagram for explaining supplementary explanation of the tissue / cell Natural protein chip. FIG. 20 is a diagram showing profiling of autoantibody-reactive protein spots in patient serum using a Natural protein chip. FIG. 21 is a diagram for explaining the data storage management table. FIG. 22 is a diagram for explaining the return matching algorithm. FIG. 23 is a diagram for explaining the return matching algorithm. FIGS. 24A and 24B are diagrams showing the results of measuring a certain protein with two different mass analyzers. FIG. 25 is a diagram illustrating an example of data obtained by integrating the data illustrated in FIGS. 24A and 24B by data integration processing. FIG. 26 is a flowchart showing an example of the data integration process according to the present embodiment. FIG. 27 (a) is a diagram showing the result of the integration processing shown in FIG. 26 on the protein data shown in FIGS. 24 (a) and (b), and FIG. 27 (b) is a diagram showing the result of FIG. These are figures which show the data of a certain DNA array memorize | stored in local DB. FIG. 28 is a diagram illustrating an example of data obtained by integrating the data illustrated in FIGS. 27A and 27B by data integration processing. FIG. 29 is a flowchart showing an example of data integration processing for integrating data as shown in FIGS. 27 (a) and 27 (b). FIG. 30 is a flowchart showing an example of a protein identification procedure according to the present embodiment. FIG. 31 is a flowchart illustrating an example of merging of protein and mRNA according to the present embodiment. FIG. 32A is a diagram illustrating an example of protein information obtained from a 2D-DIGE image, and FIG. 32B is a diagram illustrating an example of Euclidean scale information between proteins. FIG. 33 is a diagram schematically showing extraction of a specific protein according to the present embodiment. FIG. 34 shows tumor-specific protein profiling by two-dimensional Western blotting. FIG. 35 shows tumor-specific protein profiling by two-dimensional Western blotting. FIG. 36 shows tumor-specific protein profiling by two-dimensional Western blotting and Pro-Q diamond staining. FIG. 37 shows tumor-specific protein profiling by two-dimensional Western blotting using an antibody cocktail. FIG. 38 shows the membrane chip pattern of individual patient tissues with antibody cocktail. FIG. 39 shows a representative chip pattern of patient samples of insensitive and sensitive AOG brain tissue using GCTEZ cocktail.

Explanation of symbols

12 Proteomics-based processor group 14 Genomics-based processor group 16 Proteomics DB
18 Function classification DB
20 Integrated analysis system 22 Data mining system 24 International DB
30, 32, 34 Mass spectrometer 36 2D-DIGE measuring device 40, 42, 44, 46 Control / analysis unit 50, 52, 54, 56 Local DB
60 Microdissection device 62 DNA array measurement device 64 Real-time PCR measurement device 70, 72, 74 Control / analysis unit 80, 82, 84, Local DB

Claims (22)

Including at least electrophoretic images obtained by proteome analysis methods using tissues, cells and cultured cells derived from patients and disease model animals showing various pathological conditions, and mass spectrometry signal patterns by mass analyzers having various different characteristics, In a computer provided with a database storing raw data, a peptide / protein ID identified based on the raw data, and quantitative data, the computer includes:
Converting and merging data stored in the database into the same language;
Searching other publicly accessible databases accessible over the network;
An analysis program characterized by: extracting a specific protein group, predicting these functions, and analyzing a disease state. In a computer having an MS spectrum database for storing MS spectrum data obtained by applying a plurality of LC-MS shotgun methods, the computer includes:
Identifying a protein;
Storing the analysis result obtained by the identification of the protein in a database;
An analysis program comprising: executing a step of linking and integrating an analysis result stored in a database, a protein interaction analysis system and a biological information integration platform through a network. In a computer comprising a database storing logical spectra, the computer includes:
The raw data obtained by applying the stable isotope-labeled LC-MS shotgun method is based on the similarity between the peptide peak list created under the existing conditions and the theoretical spectrum stored in the database. Identifying a peptide; and
Determining a relative amount of protein containing the peptide from the difference in intensity of the pair of stable isotope-labeled pairs;
And a step of classifying a protein group that specifically appears in a specific tissue / cell. MRNA quantitative expression analysis data by analysis of different properties, including DNA arrays and / or real-time PCR, and two-dimensional electrophoresis, LC-ESI-MS / MS shotgun method, and / or LC &#8212; In a computer comprising a database storing quantitative data by proteomics using different methods, including the MALDI-MS / MS shotgun method, the computer includes:
An analysis program for causing each of data stored in the database to execute a step of performing analysis by the same algorithm including a differential display. Further, the computer
The analysis data, which is the result of the analysis performed by linking the database with another different database and stored in the database on the computer, is based on the information stored in the other linked different database. The analysis program according to claim 4, wherein the analysis step is executed. In a computer comprising image data including 2D-DIGE, including image data including spot information and protein information including post-translational modification structures, the computer includes:
An analysis program for associating a spot in the image data with protein information including a post-translational modification structure from the database and storing the association in the database. In the computer,
The analysis program according to any one of claims 1 to 5, wherein the step of reverse-inputting the information of the created integrated database to the analysis platform of the analyzer and executing the re-analysis is executed.   mRNA is DNA array and / or real-time PCR, solubilized protein is fluorescence-labeled two-dimensional electrophoresis including 2D-DIGE and / or LC-MS shotgun method of stable isotope labeling including ICAT and iTRAQ 5. The analysis program according to claim 4, wherein the analysis is performed by unifying the data by the quantitative differential display of the method at the same time. In a computer comprising a database storing a molecular network, the computer includes:
Examining a molecular network in the database based on molecular integration data;
And a step of extracting a pathological tissue / cell-specific signal cascade that is specifically activated or specifically deactivated in a pathological condition. In a computer comprising an Accession No identified based on raw data including mass spectrometry signal patterns by mass analyzers having different characteristics, and a database storing at least the intensity ratio, the computer includes:
Extracting an intensity ratio for the same Accession No. based on different mass analyzers stored in the database;
An analysis program, comprising: integrating the values of the intensity ratio, and storing the integrated values in the database in association with the AccessionNo. Further, the database includes an Accession No based on data derived from a DNA array, and an Intensity ratio,
Extracting an integrated intensity ratio for the same Accession No and an intensity ratio for a DNA array stored in the database;
The analysis program according to claim 10, wherein the step of normalizing the value of the intensity ratio and storing the normalized value in association with the AccessionNo in the database is executed. In a computer comprising image data including 2D-DIGE and storing image data including spot information, the computer includes:
Each spot information and a database connected to the computer, the other information database storing the known protein spot information, the spot information and the known protein information When the spot information matches, giving the spot information of the known protein as the spot information, and storing in the database,
A step of identifying a protein applied to the spot based on the second information of the spot, which is obtained by another method including analysis by a mass spectrometer and stored in the database when there is no match An analysis program characterized by that.   The step of identifying a protein based on the second information is a step of searching a database in which the information of the protein identified by the other method is stored, which is still another database connected to the computer. The analysis program according to claim 12, comprising: A protein sample obtained by labeling a protein extracted from biological tissue or cells with a fluorescent dye is subjected to first-dimensional isoelectric focusing on the strip, and then the strip after isoelectric focusing is subjected to reduction treatment and SDS. A small-sized protein chip that covers all proteins synthesized in living tissue or cells, obtained by subjecting to second-dimensional SDS-polyacrylamide electrophoresis. The protein chip according to claim 14, which has a size of 10 cm × 10 cm or less. The protein chip according to claim 14 or 15, wherein the protein sample is transferred from the electrophoresis gel to a membrane and immobilized. (1) a step of subjecting a protein sample obtained by labeling a protein extracted from biological tissue or cells with a fluorescent dye to first-dimensional isoelectric focusing on a strip; and (2) isoelectric focusing. The method for producing a protein chip according to claim 14 or 15, comprising a step of subjecting the subsequent strip to reduction treatment and SDS, and then subjecting the strip to a second-dimensional SDS-polyacrylamide electrophoresis. (1) a step of subjecting a protein sample obtained by labeling a protein extracted from biological tissue or cells with a fluorescent dye to first-dimensional isoelectric focusing on a strip; (2) after isoelectric focusing The step of reducing the strip and converting it to SDS, then subjecting it to the second-dimensional SDS-polyacrylamide electrophoresis, and (3) transferring the protein sample from the electrophoresis gel after SDS-polyacrylamide electrophoresis to the membrane. The manufacturing method of the protein chip of Claim 16 containing. 17. Detecting fluorescence derived from a fluorescent dye labeled on the protein sample on the protein chip according to any one of claims 14 to 16, and imaging the protein sample as protein spot information on the protein chip. Method. (1) A list of molecules that can be used for monitoring the progress of a disease state or drug sensitivity is listed from molecular information obtained by causing a computer to execute the program according to claim 1, 2, 3, 8, or 9. Preparing an antibody cocktail against these molecules, (2) bringing the antibody cocktail into contact with the protein chip according to any one of claims 14 to 16, and (3) a positive antibody reaction pattern on the protein chip. A method for monitoring the progress of a disease state or drug sensitivity, which comprises the steps of profiling and preparing a database. (1) Anti-GFAP antibody, anti-CD44 antibody, anti-p53 antibody, anti-prothrombin antibody, anti-α1-antitrypsin antibody, anti-ubiqutine
activating enzyme E1 antibody, anti-Erk antibody, anti-p-Tyr antibody, and a step of preparing an antibody cocktail containing at least two kinds of antibodies selected from anti-zyxin antibodies, (2) extraction from biological tissue or cells of brain tumor patients The protein sample obtained by labeling the obtained protein with a fluorescent dye is subjected to first-dimension isoelectric focusing on the strip, and then the strip after isoelectric focusing is reduced and converted to SDS. Contacting the antibody cocktail with a micro-sized protein chip covering all proteins synthesized in living tissue or cells obtained by subjecting to eye SDS-polyacrylamide electrophoresis, and (3) on the protein chip A method of monitoring the progress of a disease state or drug sensitivity of a brain tumor, comprising the step of profiling the antibody reaction positive pattern of the above and preparing a database. An antibody cocktail used in the method of claim 21 for monitoring the progress of a disease state or drug sensitivity of a brain tumor, comprising an anti-GFAP antibody, an anti-CD44 antibody, an anti-p53 antibody, an anti-prothrombin antibody, an anti-α1- antitrypsin antibody, anti-ubiqutine
An antibody cocktail comprising at least two kinds of antibodies selected from activating wnzyme E1 antibody, anti-Erk antibody, anti-p-Tyr antibody, and anti-zyxin antibody.