JP2007181453A - Method for obtaining array data, method for extracting target gene by using the same, and method for designing protein - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To further improve efficiency of obtaining data by IVV (in vitro virus) method, which data include several partial sequences of a gene. <P>SOLUTION: The method repeats rounds including (1) a step for producing, from a nucleic acid library, a molecular library of an associating molecule bonded with a protein and a nucleic acid encoding the protein, (2) a step for mixing a target substance with the library of the associating molecule, (3) a step for selecting an associating molecule bonding to the target substance and (4) a step for preparing, by RT-PCR or PCR, a nucleic acid library from the nucleic acid of the associating molecule selected in (3), wherein, in the RT-PCR or PCR, the cycle number is determined when the increase of the amplified products reaches, for the first time, the predetermined amount less than the saturation amount, and after the cycle number reaches to be a constant, one more round is performed and sequence analysis of the nucleic acid library is performed. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a disease-related gene or the like in which DNA sequence or mRNA sequence is changed by alternative splicing, RNA editing, single nucleotide polymorphism (SNP), sequence mutation, sequence fusion, etc., and as a result, protein-protein interaction is changed. And a method for obtaining sequence data useful for bait protein design in protein interaction analysis, and a method for performing the extraction and bait protein design using the sequence data.

  In the extraction of drug-related target genes, that is, disease-related genes in which the DNA sequence or mRNA sequence changes due to alternative splicing, RNA editing, SNP, sequence mutation, sequence fusion, etc., resulting in changes in protein-protein interactions In order to follow changes in DNA sequences and mRNA sequences derived from alternative splicing, RNA editing, SNPs (single nucleotide polymorphisms), etc., it is possible to obtain a lot of sequence information by analyzing the interaction between proteins. is important.

  Known methods for detecting protein-protein interactions include immunoprecipitation, pull-down assay using GST fusion protein, TAP (Tandem Affinity Purification) method, and yeast two-hybrid method. On the other hand, as a method for comprehensive analysis of protein-protein interactions in post-genome functional analysis by applying `` association of genes (genotypes) and proteins (phenotypes) '' that was born as an evolutionary molecular engineering tool, In vitro virus method (Non-patent document 1, Non-patent document 2, Non-patent document 3, Patent document 1, Patent document 2), STABLE method, phage display method, ribosome display method, mRNA-peptide fusion (mRNA display) method, etc. It has been known.

  Moreover, in the protein-protein interaction analysis such as the conventional two-hybrid method and TAP method, it is necessary to design the bait protein by trial and error for each protein. In large-scale analysis, there is a demand for a protein-protein interaction analysis system that enables efficient bait protein design and that can be repeated automatically and continuously as much as possible. In protein design, proteins that are difficult to crystallize for three-dimensional structural analysis are cloned in various lengths of amino acids in various lengths, and their binding ability is determined by pull-down assays using GST fusion proteins. All had to be confirmed experimentally.

According to the in vitro virus method, it is known that data with few false positives can be obtained (Patent Document 3).
Miyamoto-Sato E, et al. Viva Origino 25, 35 (1997) Nemoto N, et al. FEBS Lett. 414, 405 (1997) Miyamoto-Sato, E. et al. Nucleic Acids Res. 31, e78 (2003) International Publication No. WO98 / 16636 Pamphlet International Publication No. WO02 / 46395 Pamphlet International Publication No. WO 2005/050518 Pamphlet

  When searching for drug discovery target genes, it is only possible to analyze which proteins interact with each other in order to follow detailed changes in DNA and mRNA sequences derived from alternative splicing, RNA editing, SNPs, etc. Is not enough. It is important to detect in detail the site of interaction, such as the in vitro virus (IVV) method, and when discovering new functions by overlaying with databases such as variants and SNPs It is important to use data including several partial sequences for one gene.

  In designing a bait protein for protein-protein interaction analysis, it is important to obtain sequence data of various regions in terms of position and length where the interaction is recognized.

  In conventional protein-protein interaction analysis methods such as the two-hybrid method and the TAP method, it is difficult to efficiently obtain the sequence data as described above.

  According to the IVV method, as described in the pamphlet of International Publication No. WO 2005/050518, data with few false positives can be obtained, and the above sequence data can be efficiently obtained in terms of data quality. Although it can be said, improvement is still expected.

  An object of the present invention is to further improve the efficiency of acquiring data including several partial sequences for one gene by the IVV method.

  In the IVV method, the ratio of partial sequences for one gene in the library is increased by repeating a round consisting of creation of a play library, formation of a bait and play complex, and screening of the complex.

  Whether or not a library containing a large number of partial sequences for one gene has been obtained can be determined by sequencing, but after each round, sequencing and determining this can improve efficiency. It is a problem. In view of this, a method was examined in which it was possible to know in advance at what timing the base sequence analysis should be performed by predicting the proportion of partial sequences for one gene in the obtained library. As a result, when the number of RT-PCR cycles per round is monitored, a large number of portions of one gene can be obtained by rotating the round from the point where the number of RT-PCR cycles is constant until a predetermined enrichment rate is obtained. It was found that a library containing the sequence was obtained.

Based on these findings, the present invention has been completed. The present invention provides the following.
[1] (1) A step of producing a library of mapping molecules in which a protein and a nucleic acid encoding the protein are bound from a nucleic acid library, and (2) mixing the library of mapping molecules and a target substance. (3) selecting the mapping molecule bound to the target substance, (4) preparing a nucleic acid library by RT-PCR or PCR from the nucleic acid of the mapping molecule selected in (3) Including repeating rounds,
In the RT-PCR or PCR, the number of cycles in which the amplification amount of the amplification product first reaches a certain amount less than or equal to the saturation amount is obtained, and after the number of cycles reaches a certain value, one or more rounds are passed. A method for obtaining sequence data, characterized in that the sequence analysis of a nucleic acid library is sometimes performed.
[2] A method for obtaining sequence data using a library of mapping molecules in which a protein and a nucleic acid encoding the protein are bound, wherein (a) a bait protein is designed, and (b) the designed bait protein (C) Sequence analysis of the corresponding molecules selected in the selection, acquisition of sequence data, (d) determination of the interaction site based on the acquired sequence data And (e) extracting sequence data based on the determined interaction site, wherein the steps (b) and (c) are performed by the method described in [1].
[3] A protein comprising extracting interaction sites based on sequence data, designing a bait protein based on the extraction result, and obtaining sequence data by the method according to [1] or [2] Design method.
[4] Characteristically, sequence data is obtained by the method according to [1] or [2], and using the obtained sequence data, a gene whose protein-protein interaction changes due to a change in base sequence is extracted. Gene extraction method.

  Sequence data including various partial sequences for one gene can be efficiently obtained. This improvement in efficiency is particularly important in large-scale protein-protein interaction analyses. Further, such sequence data is sequence data suitable for use in extraction of a protein whose interaction between proteins changes due to a change in the base sequence of the interaction site in searching for a drug discovery target gene. This accelerates the functional analysis of genes and the elucidation of disease mechanisms. Furthermore, it is possible to extract a disease-related gene by analyzing the relationship between the change in the base sequence of the interaction site and the disease.

<1> Sequence Data Acquisition Method of the Present Invention The sequence data acquisition method of the present invention comprises (1) a step of producing a library of mapping molecules in which a protein and a nucleic acid encoding the protein are bound, from a nucleic acid library; ) A step of mixing the library of matching molecules and the target substance, (3) a step of selecting the mapping molecule bound to the target substance, (4) RT from the nucleic acid of the mapping molecule selected in (3) -Repeating a round including a step of preparing a nucleic acid library by PCR or PCR, in RT-PCR or PCR, determining the number of cycles that the amplification amount first reaches a certain amount below the saturation amount; and After the number of cycles reaches a certain value, the sequence analysis of the nucleic acid library is performed when one or more rounds are passed.

  Said round may be the same as the screening method using a normal matching molecule (refer patent documents 1-3). Screening methods that use an association molecule in which a protein and a nucleic acid encoding the protein are combined include, for example, amplification of genotype (nucleic acid), linkage of phenotype (protein) and genotype, and affinity selection. These steps are repeated as necessary. A typical example of such a method is a screening method based on the in vitro virus method (see, for example, WO 02/48347, JP-A-2002-176987).

  A mapping molecule (hereinafter also referred to as IVV) by the in vitro virus method is formed by binding a phenotype molecule containing a protein and a genotype molecule containing a nucleic acid encoding the protein. The genotype molecule is formed by binding a coding molecule having a region encoding a protein in such a form that the base sequence of the region can be translated, and a spacer part. For convenience of explanation, a part derived from a phenotype molecule, a part derived from a spacer molecule, and a part derived from a coding molecule in the mapping molecule are referred to as a decoding part, a spacer part, and a coding part, respectively. Moreover, the part derived from the spacer molecule and the part derived from the coding molecule in the genotype molecule are referred to as a spacer part and a coding part, respectively.

The spacer molecule can bind to the peptide by, for example, a donor region capable of binding to the 3 ′ end of the nucleic acid, a PEG region mainly composed of polyethylene glycol bound to the donor region, and a peptide transfer reaction bound to the PEG region. And a peptide acceptor region containing a group.
There may be no PEG area.

  The donor region that can bind to the 3 ′ end of a nucleic acid usually consists of one or more nucleotides. The number of nucleotides is usually 1-15, preferably 1-2. The nucleotide may be ribonucleotide or deoxyribonucleotide.

  The sequence at the 5 ′ end of the donor region affects the ligation efficiency. In order to ligate the coding part and the spacer part, it is necessary to contain at least one residue. For an acceptor having a poly A sequence, at least one residue of dC (deoxycytidylic acid) or 2 The residue dCdC (dideoxycytidylic acid) is preferred. The base type is preferably C> U or T> G> A.

  The PEG region is mainly composed of polyethylene glycol. Here, the main component means that the total number of nucleotides contained in the PEG region is 20 bases or less, or the average molecular weight of polyethylene glycol is 400 or more. Preferably, it means that the total number of nucleotides is 10 bases or less, or the average molecular weight of polyethylene glycol is 2000 or more.

The average molecular weight of polyethylene glycol in the PEG region is usually 400 to 30,000, preferably 1,000 to 10,000, and more preferably 2,000 to 8,000. Here, when the molecular weight of polyethylene glycol is lower than about 400, when a genotype molecule containing a spacer part derived from this spacer molecule is associated and translated, a post-process of associated translation may be required ( Liu, R., Barrick, E., Szostak, JW, Roberts, RW (2000) Methods in Enzymology,
vol. 318, 268-293), and using a PEG having a molecular weight of 1000 or more, more preferably 2000 or more, high-efficiency association can be achieved only by association translation, so that post-translation processing is not necessary. In addition, when the molecular weight of polyethylene glycol increases, the stability of genotype molecules tends to increase, especially when the molecular weight is 1000 or higher, and when the molecular weight is 400 or lower, the DNA spacer may not be so characteristic and unstable. .

  The peptide acceptor region is not particularly limited as long as it can bind to the C-terminus of the peptide. For example, puromycin, 3'-N-aminoacylpuromycin aminonucleoside (3'-N-aminoacylpuromycin aminonucleoside, PANS-amino acid) For example, PANS-Gly whose amino acid part is glycine, PANS-Val of valine, PANS-Ala of alanine, and other PANS-all amino acids corresponding to all amino acids can be used. In addition, 3'-Aminoacyladenosine aminonucleoside (AANS-amino acid, 3'-Aminoacyladenosine aminonucleoside, formed by dehydration condensation of the amino group of 3'-aminoadenosine and the carboxyl group of amino acid as a chemical bond For example, AANS-Gly having an amino acid part of glycine, AANS-Val of valine, AANS-Ala of alanine, and other AANS-all amino acids corresponding to all amino acids can be used. Further, a nucleoside or a nucleoside and an amino acid ester-bonded one can be used. In addition, nucleoside or a substance having a chemical structure skeleton similar to nucleoside and an amino acid or a substance having a chemical structure skeleton similar to amino acid can be used as long as they can be chemically bonded.

  The peptide acceptor region is preferably composed of puromycin or a derivative thereof, or puromycin or a derivative thereof and one or two residues of deoxyribonucleotide or ribonucleotide. Here, the derivative means a derivative capable of binding to the C-terminus of the peptide in the protein translation system. The puromycin derivatives are not limited to those having a complete puromycin structure, but also include those lacking a part of the puromycin structure. Specific examples of the puromycin derivative include PANS-amino acid and AANS-amino acid.

  The peptide acceptor region may be composed of puromycin alone, but preferably has a base sequence consisting of DNA and / or RNA of 1 residue or more on the 5 ′ end side. The sequence is dC-puromycin, rC-puromycin, etc., more preferably dCdC-puromycin, rCrC-puromycin, rCdC-puromycin, dCrC-puromycin, etc., and the 3 ′ end of aminoacyl-tRNA is A mimicked CCA sequence (Philipps, GR (1969) Nature 223, 374-377) is suitable. The base type is preferably C> U or T> G> A.

  The spacer molecule preferably includes at least one function-imparting unit between the donor region and the PEG region. The function-imparting unit is preferably a functional modification of at least one residue of deoxyribonucleotide or ribonucleotide base. For example, a substance into which various separation tags such as a fluorescent substance, biotin, or His-tag are introduced as a function modifying substance can be used.

  The coding molecule includes, for example, a 5 ′ untranslated region containing a transcription promoter and a translation enhancer, an ORF region encoding a protein bound to the 3 ′ end of the 5 ′ untranslated region, and a 3 ′ end of the ORF region. A nucleic acid having a bound poly A sequence and containing an affinity tag sequence 5 ′ upstream of the poly A sequence.

  The coding molecule may be DNA or RNA. In the case of RNA, the 5 'end may or may not have a Cap structure. The coding molecule may be incorporated into any vector or plasmid.

  The 3 ′ end region may comprise an affinity tag sequence and a poly A sequence downstream thereof. As a factor affecting the ligation efficiency between the spacer molecule and the coding molecule, the polyA sequence in the 3 ′ end region is important, and the polyA sequence may be a mixture of dA and / or rA having at least 2 residues or more. The poly A continuous chain is preferably 3 or more residues, more preferably 6 or more, and even more preferably 8 residues or more.

  Elements affecting the translation efficiency of the coding molecule include a 5 ′ UTR consisting of a transcription promoter and a translation enhancer, and a combination of a 3 ′ terminal region containing a poly A sequence. The effect of the poly A sequence in the 3 ′ end region is usually exerted with 10 residues or less. T7 / T3 or SP6 can be used as the 5 ′ UTR transcription promoter, and there is no particular limitation. SP6 is preferable, and SP6 is particularly preferable when an omega sequence or a sequence containing a part of the omega sequence is used as an enhancer sequence for translation. The translation enhancer is preferably part of the omega sequence, and part of the omega sequence may be part of the TMV omega sequence (O29; Gallie DR, Walbot V. (1992) Nucleic Acids Res., Vol. 20, 4631 -4638, and WO 02/48347 (see FIG. 3) and O ′ (SEQ ID NO: 37) are preferred.

  The affinity tag sequence is not particularly limited as long as it is a sequence for using any means capable of detecting a protein such as an antigen-antibody reaction. Preferably, a Flag-tag sequence or a His-tag sequence, which is a tag for affinity separation analysis by antigen-antibody reaction.

The ORF region may be any sequence consisting of DNA and / or RNA. A gene sequence, exon sequence, intron sequence, random sequence, or any natural or artificial sequence is possible, and there is no sequence limitation. A nucleic acid library having such a sequence can be used. Further, by setting the 5′UTR of the coding molecule as SP6 + O29 and the 3 ′ terminal region as, for example, Flag + XhoI + A _n (n = 8), each length is about 60 bp in 5′UTR, The length of the 3 ′ end region is about 32 bp, and it can be incorporated as an adapter region into a PCR primer. For this reason, a coding molecule having a 5 ′ UTR and a 3 ′ terminal region can be easily prepared by PCR from any vector, plasmid, or cDNA library. In the coding molecule, translation may be done beyond the ORF region. That is, there may not be a stop codon at the end of the ORF region.

  The coding molecule was bound to the 5 'untranslated region containing the transcription promoter and translation enhancer, the ORF region encoding the protein bound to the 3' end of the 5 'untranslated region, and the 3' end of the ORF region. Or a nucleic acid containing a 3 ′ terminal region containing a poly A sequence.

  The genotype molecule is obtained by converting the above coding molecule into a form in which the base sequence of the protein coding region can be translated (for example, after transcription), if necessary, and the 3 ′ end of the coding molecule and the spacer molecule. The donor region can be produced by binding by a normal ligase reaction. The reaction conditions usually include conditions at 4 to 25 ° C. for 4 to 48 hours. When adding polyethylene glycol having the same molecular weight as the polyethylene glycol in the PEG region of the spacer molecule containing the PEG region, to the reaction system It is also possible to shorten to 0.5 to 4 hours at 15 ° C.

  The combination of spacer molecule and coding molecule has an important effect on ligation efficiency. In the 3 ′ terminal region of the coding part corresponding to the acceptor, there should be a DNA and / or RNA polyA sequence of at least 2 residues, preferably 3 residues, more preferably 6-8 residues, The UTR translation enhancer is preferably a partial sequence of an omega sequence (O29 or O '), and the spacer region donor region is at least one dC (deoxycytidylic acid) or two residues dCdC (dideoxy). Cytidylic acid) is preferred. By using RNA ligase, the problem of DNA ligase can be avoided and the efficiency can be maintained at 60 to 80%.

  (a) a 5 ′ untranslated region containing a transcription promoter and a translation enhancer, an ORF region encoding a protein bound to the 3 ′ end of the 5 ′ untranslated region, and a 3 ′ end of the ORF region The 3 'end of the coding molecule, which is RNA containing the 3' end region containing the poly A sequence, and (b) the donor region of the spacer molecule consisting of RNA constitutes the PEG region in the spacer molecule It is preferable to bind with RNA ligase in the presence of free polyethylene glycol having the same molecular weight as polyethylene glycol.

By adding polyethylene glycol having the same molecular weight as the PEG region of the spacer part including the PEG region during the ligation reaction, the ligation efficiency is improved to 80 to 90% or more regardless of the molecular weight of the polyethylene glycol of the spacer part. The separation step can also be omitted.
The mapping molecule in this embodiment is linked to a phenotype molecule that is a protein encoded by the ORF region in the genotype molecule by translating the genotype molecule in a cell-free translation system. Can be obtained. The cell-free translation system is preferably that of wheat germ or rabbit reticulocytes. The conditions for translation may be those normally employed. For example, the conditions are 15 to 240 minutes at 25 to 37 ° C. In addition, the nucleic acid portion of the mapping molecule of this embodiment can be a hybrid of RNA and DNA by reverse transcription after translation.

  Target substances include peptides (which may be chemically synthesized, isolated from nature, or partially digested proteins), proteins, nucleic acids (DNA or RNA), saccharides, various low molecular compounds, metals All compounds and substances such as metal compounds are applicable.

The target substance used in the present invention is bound to a solid phase such as a resin or a bead depending on the embodiment, and the timing of binding may be before or after mixing with the corresponding molecule, and is selected according to the binding method. . Examples of the method for binding to a solid phase include a method of binding via a modifying substance and a method of binding using a moiety other than that.

  The modifier used for binding via a modifier is usually a molecule that specifically binds to a specific polypeptide (hereinafter sometimes referred to as a “ligand”) and is attached to the solid surface. Binds a specific polypeptide that binds to the ligand (hereinafter sometimes referred to as "adapter protein"). Adapter proteins also include binding proteins, receptor proteins that constitute receptors, antibodies, and the like. Examples of adapter protein / ligand combinations include biotin-binding proteins such as avidin and streptavidin / biotin, maltose-binding protein / maltose, G protein / guanine nucleotide, polyhistidine peptide / nickel or cobalt or other metal ions, glutathione-S -Transferase / glutathione, DNA binding protein / DNA, antibody / antigen molecule (epitope), calmodulin / calmodulin binding peptide, ATP binding protein / ATP, or various receptor proteins such as estradiol receptor protein / estradiol / ligands thereof It is done.

  Among these, adapter protein / ligand combinations include biotin-binding proteins such as avidin and streptavidin, maltose-binding proteins / maltose, polyhistidine peptides / metal ions such as nickel or cobalt, glutathione-S-transferase / glutathione, The combination of streptavidin / biotin is most preferable. These binding proteins are known per se, and the DNA encoding the protein has already been cloned.

  The adapter protein can be bound to the solid phase surface by a method known per se. Specifically, for example, tannic acid, formalin, glutaraldehyde, pyruvaldehyde, bis-diazotized benzidone, toluene- A method utilizing a 2,4-diisocyanate, an amino group, a carboxyl group that can be converted into an active ester, or a hydroxyl group or amino group that can be converted into a phosphoramidide can be used.

  When binding to the solid phase by a moiety other than the modifying substance, known methods commonly used for binding proteins, nucleic acids, sugar chains, low molecular weight compounds to the solid phase, specifically, for example, tannic acid, formalin, Use a method that uses glutaraldehyde, pyruvic aldehyde, bis-diazotized benzidone, toluene-2,4-diisocyanate, amino group, carboxyl group that can be converted to active ester, hydroxyl group or amino group that can be converted to phosphoramidide, etc. be able to.

  The solid support is preferably agarose beads or agarose beads in which a magnetic material is embedded.

  The library of the mapping molecule and the target substance may be mixed under the condition that the target protein of the mapping molecule interacts with the target substance. This condition is appropriately selected according to the interaction to be detected and the type of target substance. When the target substance is a protein, RNA encoding the protein may be prepared and synthesized together during the synthesis of the corresponding molecule (cotranslation). Such a method is described in WO03 / 048363.

The process of selecting the mapping molecule bound to the target substance is a process of separating the mapping molecule bound to the target substance and the mapping molecule that does not bind to the target substance. Usually, the target substance is fixed to the solid phase. By separating or immobilizing, separation can be performed by washing the solid phase on which the target molecule after mixing with the corresponding molecule is immobilized. Washing conditions are appropriately selected according to the interaction to be detected and the type of target substance. Here, immobilization on the solid phase means that the conjugate of the mapping molecule and the target substance can be separated from the non-bonded molecule. For example, when the target substance is a membrane protein, the cell membrane of the cell Membrane proteins expressed in the above or proteins embedded in artificial membranes are also included in the target substance immobilized on the solid phase.

  The step of preparing a nucleic acid library based on the nucleic acid of the selected mapping molecule includes preparing DNA by RT-PCR or PCR. In the case of PCR, reverse transcription is performed and then DNA is used. RT-PCR is advantageous in terms of ease of operation. The prepared DNA is used for sequence analysis, or used as a nucleic acid library for the synthesis of mapping molecules.

  A specific example of the screening method will be described with reference to FIG. This screening method repeats a round consisting of (I) transcription, (II) cell-free cotranslation, (III) IVV selection, and (IV) RT-PCR. In the co-translation, the production of the mapping molecules and the mixing of them with the bait are performed at the same time. Therefore, the steps (I) and (II) correspond to the production step (1) and the mixing step (2). Steps (III) and (IV) correspond to the selection step (3) and the preparation step (4), respectively.

  In step (I), the bait DNA and prey cDNA library are transcribed into bait protein mRNA and prey IVV mRNA (puromycin bound to the 3 ′ side of mRNA via a spacer). In step (II), bait protein mRNA and prey IVV mRNA are translated in a cell-free translation system, and a complex of IVV and bait is formed. In step (III), a two-stage selection is performed. In step (IV), DNA is amplified by RT-PCR based on the nucleic acid portion of IVV. The amplified DNA is used as the next round library or subjected to sequence analysis.

  In RT-PCR or PCR, the reaction becomes saturated as the number of cycles is increased. When a library is used as a template, the product amount will not reflect the original template amount when the reaction is saturated. This is because the template sequence and amount are different, so that even if the amplification of some DNA reaches saturation, the other amplification does not stop. Bias caused by such a phenomenon is not preferable in screening.

  Therefore, in the screening, the amplification product before reaching saturation is generally used for the next round. Usually, the amplification amount is 90% or less of the saturation amount. The lower limit of the amplification amount is not limited as long as RT-PCR or PCR is possible, and generally the amplification amount is 40% or more of the saturation amount. The amount of amplification can be easily measured by real-time PCR, and the amount of saturation can be specified by reacting under the same conditions until the reaction is saturated. As an amplification amount index, an intensity value such as fluorescence measured by real-time PCR can be used as it is.

  In the present invention, this cycle number is determined as the number of cycles that first reaches a certain amount of amplification amount below the saturation amount. A certain amount is selected from a saturation amount, usually in the range of 59 to 100%, preferably in the range of 70 to 85%. If the amount is too small, it may not be possible to obtain a sufficient amount of DNA for the next round or cloning. If a certain amount is too large, the distribution of sequences obtained by selection may change. This specific cycle number was found to be related to the quality of the library obtained in each round (percentage of partial sequences for one gene), as shown in the Examples below.

  In the present invention, after the number of specific cycles reaches a certain value, one or more rounds are further performed to perform sequence analysis of the nucleic acid library. Here, “constant” means that it is substantially constant in consideration of experimental errors and the like, and usually, fluctuations within 2 times are preferably allowed within 1 time. When the number of specific cycles reaches a constant, the number of partial sequences for one gene in the sequence data starts to increase significantly. After the number of specific cycles reaches a certain level, the upper limit of the number of rounds to be carried out may be selected from an economic point of view, but if too many, the variation of the partial sequence tends to decrease and the quality of the sequence data tends to decrease There is.

  The number of rounds to be performed after the number of specific cycles reaches a certain value is preferably performed so that a concentration rate of 1000 times or more is obtained.

Here, the enrichment rate is a value obtained by a ratio between the abundance of the indicator gene in the initial cDNA library and the abundance of the indicator gene after the first round. The indicator gene is FOS / JUN, and is the ratio when Jun of the interaction partner is concentrated in the prey cDNA library using bait Fos. The FOS / JUN interaction strength is a physicochemical value (Kd = 10 ^-7 to 10 ^-8 ), and the interaction strength is universal, but the concentration rate depends on the type of beads used and It is affected by the purification tag and the number of purification stages.

  In the conditions of Examples described later, the ratio of the abundance of the indicator gene in the initial cDNA library and the abundance of the indicator gene in the first stage of the 1st round is 100 (concentration rate 100 times), and in the second stage of the 1st round The ratio of the abundance of the indicator gene is 10 (concentration rate 10 times), the concentration rate of the first stage of each round is 100 times, and the concentration ratio up to the second stage is 1000 times. Therefore, after the number of specific cycles reaches a certain value, the round using this one-stage selection may be performed twice, and the two-stage selection may be performed once.

  The concentration rate can be evaluated by the method of Genome Research vol.15 pp710-717 (2005) (see Fig. 1B of the same document). Specifically, a 100 nM jun template is translated in the presence and absence of a 400 nM fos template (bait) under the conditions described in Examples below, and serial dilutions of translation products are subjected to electrophoresis, etc. It can be obtained by quantifying with.

  This enrichment rate is based on the ratio of the concentration of the indicator gene in the initial cDNA library to the amount of the indicator gene after the first round, and the degree of enrichment is evaluated. In the second round, the concentration does not actually have to be performed at this concentration rate.

  In this way, if the number of specific cycles was used as an index and selection with a known enrichment ratio was used in advance, sequence data including several partial sequences for one gene was obtained by sequence analysis for each round. Since it is not necessary to confirm whether or not, the target sequence data can be obtained efficiently.

The sequence analysis of the nucleic acid library can be performed as follows.
The vector sequence is removed from the base sequence obtained by sequencing, the cDNA sequence is extracted, the database is searched if necessary, the protein frame is confirmed, and at the same time, the IVV frame is confirmed.

  Sequence sequences detected by IVV selection are waveform data of commercially available sequencers (ABI, BECKMAN, etc.) and sequence data based on the analysis thereof.

  The removal of the vector sequence is to remove the vector sequence used for cloning for the sequence and to remove the extra sequence so that only the cDNA library sequence is obtained.

  Extraction of a cDNA sequence is to remove a sequence common to the library and to use only the cDNA sequence in this case for database search.

  Database search includes nt as a nucleic acid database, nr as a protein database, and other databases such as refseq and fantom2, and a genome database.

  The confirmation of the protein frame is to check whether the frame translated by the cell-free translation system matches the frame with the ORF of the protein registered in the database. If the frames do not match, clarify whether it is a frame shift or a 3'UTR, 5'UTR array.

  Confirmation of the IVV frame is to confirm whether the sequence can be translated by a cell-free translation system. That is, whether or not a plurality of stop codons appear or whether the location is on the N-terminal side is determined.

  Alignment can be performed by a normal method. Software for performing such an alignment (for example, ClustalX) is available.

<2> Application of Sequence Data Acquisition Method of the Present Invention According to the sequence data acquisition method of the present invention, sequence data of a library including a large number of partial sequences for one gene can be efficiently acquired. For applications where the nature of such sequence data is important, the sequence data acquisition method of the present invention can be used to acquire sequence data. Hereinafter, an example will be described.

<2-1> Extraction of target genes To track changes in DNA sequences and mRNA sequences derived from alternative splicing, RNA editing, SNPs, etc., which proteins interact with which proteins in the protein-protein interaction analysis method It is not enough to be able to analyze what to do. As in the IVV method, it is important to detect in detail what kind of base sequence site interacts, and in addition to knowing the base sequence site, it can also be overlaid with databases such as variants and SNPs. When finding a new function, it is better to have more IVV sequence data to be superimposed (hereinafter also referred to as IVV sequence). If many partial sequences are obtained for one gene, as shown in FIG. 2, in the case of variant analysis, it is possible to analyze which exon of a certain gene contributes to the function. Further, the fusion gene can be analyzed in the same manner as in FIG. 2 as shown in FIG. Furthermore, as shown in FIG. 4, in the case of SNPs analysis, it is possible to analyze which SNPs of a certain gene contribute to the function. In addition, RNA editing and cancer-derived mutations can be analyzed in the same manner as SNPs, as shown in FIG.

  According to the sequence data acquisition method of the present invention, such sequence data can be obtained efficiently. Therefore, the sequence data is acquired by the sequence data acquisition method of the present invention, and the interaction between proteins changes due to the change of the base sequence. It is advantageous to extract the genes that do. Therefore, according to the present invention, there is provided a gene extraction method for acquiring sequence data by the sequence data acquisition method of the present invention and using the sequence data to extract a gene whose protein-protein interaction changes due to a change in base sequence. .

  FIG. 1 is a configuration diagram of “extraction of a gene whose protein-protein interaction is changed by a change in base sequence” of the present invention using the sequence data acquisition method of the present invention. (1) Experiment of IVV method, (2) Public database or combination set of public database, (3) Analysis of IVV data by public database, extraction of genes whose interaction changes due to base sequence change and creation of DB (4) Furthermore, by extracting genes whose base sequence changes depending on the disease and making them into a DB, genes whose interaction changes due to the change of the base sequence related to the disease can be extracted and accumulated. By proceeding with this operation on a large scale, a new database is constructed, and the elucidation of gene functions and disease mechanisms is accelerated.

A public database, or a combination set of public databases, may be a variant database, such as a public database such as ASD (Alternative Splicing Database), a local data set such as an alternative splicing database obtained through experiments, or a combination thereof. It doesn't matter. The database of SNPs may be a public database such as dbSNP, a local dataset such as a database of SNPs obtained through experiments, or a combination thereof. RNA editing, mutant genes, fusion genes, etc. can also be local data sets such as public databases, sets extracted from public database annotations, databases of SNPs obtained through experiments, or combinations thereof. Analysis based on information such as trans-splicing, start codon variation, stop codon read-through, etc. and a database is also possible. Moreover, in the analysis with the structural analysis database, the analysis result of the interaction site by the structural analysis is collated with the result by the IVV method, and it can be confirmed that the interaction site has been accurately extracted.

  By extracting genes whose base sequence changes depending on the disease and making them into a database, it is possible to extract and accumulate genes whose interaction changes due to changes in the base sequence related to the disease. By proceeding with this operation on a large scale, a new database of disease target genes will be constructed, and gene function and disease mechanism elucidation will be accelerated.

<2-2> Protein Design Sequence data including a large number of partial sequences for one gene is useful for determining an interaction site with high accuracy. When performing bait design for efficient large-scale analysis of protein-protein interactions, including the determination of interaction sites, and protein tertiary structure analysis, such sequence data can be acquired efficiently. is important.

  Therefore, if the sequence data acquisition method of the present invention is used, it is possible to design a bait protein by accurately determining the interaction site from the prey protein database (primary DB) obtained by protein interaction analysis using the prepared bait protein. By creating a database (secondary DB) of necessary region information and using the DB information, it is possible to provide a system in which bait protein design is performed quickly and can be continuously and repeatedly analyzed. In addition, it is possible to efficiently acquire region information necessary for protein design in the system. Furthermore, it is possible to efficiently extract a functional part of a protein that is difficult to crystallize for three-dimensional structural analysis and the functional part.

The present invention relates to a sequence data acquisition method using a library of mapping molecules in which a protein and a nucleic acid encoding the protein are bound to each other, comprising: (a) designing a bait protein; and (b) a designed bait protein. (C) Perform sequence analysis of the corresponding molecule selected in the selection, obtain sequence data, and (d) select the interaction site based on the obtained sequence data. And (e) extracting the sequence data based on the determined interaction site, wherein the steps (b) and (c) are performed by the sequence data acquisition method of the present invention. To do.
Steps (a), (d) and (e) can be performed by a usual method.

  The present invention also includes extracting an interaction site based on sequence data, designing a bait protein based on the extraction result, and obtaining sequence data by the sequence data acquisition method of the present invention. Provide a design method. The designed proteins include bait proteins used for screening, X-ray analysis of protein structures including interaction regions (functional regions), proteins required for NMR analysis, and proteins required for protein crystallization technology. is there.

Hereinafter, an efficient large-scale analysis of protein-protein interactions will be described as an example.
As shown in Fig. 11, the large-scale analysis system for protein-protein interaction designs bait proteins, performs IVV selection, performs sequence analysis, and creates a pre-protein database (DB) that interacts with baits. It is obtained as a primary DB, and based on the primary DB, an interaction site such as a binding site is determined, and a function information database such as “common interact region” and “full length interact region” is obtained as a secondary DB. In this example, a bait protein is designed using a secondary DB.

  FIG. 11 is a configuration diagram and a flow of a “large-scale analysis system for protein-protein interaction”. The system consists of (1) (bait) protein design, (2) bait preparation and selection, (3) sequence analysis, (4) primary DB, (5) determination of interaction site, (6) secondary DB By repeating this flow, large-scale analysis is performed efficiently, and the accumulation of data is also promoted in each DB.

  (1) (Bait) protein design is structured to be efficient by using information of (6) secondary DB. Specific examples of utilization include bait protein design for protein-protein interaction analysis, protein design for protein structure analysis by X-ray and NMR, and drug discovery (drug design) In addition, it is possible to provide applications such as protein design necessary for protein crystallization techniques for the three-dimensional structure analysis of complexes interacting with proteins and ligands.

  (2) Bait production and selection are important components for establishing the system of FIG. Here, bait creation and selection in a completely in vitro system is required. A specific method for producing the bait is shown in FIG. The selection method can be used as long as it can be used for the in vitro virus method. The species used for the library and bait can be mouse, human, or any other species.

  (3) Sequence analysis can be used as long as the sequence can be analyzed. Specific examples include dideoxy method, epifluorescence imaging method, optical fiber method, nanopore method, sequence analysis by real-time DNA sequencing, or sequence analysis by microarray method or tiling array method.

  (4) The primary DB is a database composed of (3) sequence information determined by assigning genes or proteins for data whose sequences have been clarified by sequence analysis. As a method for determining a gene or protein for each sequence, there is a method using a public database (Nt, Locus link, refseq, etc.). There is also a method using an IVV analysis system (IWAS; Genome Research vol.15 pp710-717 (2005)).

  (5) The determination of the interaction site is an important means for establishing the system of FIG. As shown in FIG. 12, (4) using the primary DB, the amino acid sequence of each sequence information of each gene (protein) is used as an “interaction region” of each gene, and they are aligned (in FIG. 12, bait FOS A common amino acid region obtained by performing an alignment of prey JUN obtained by the IVV method using a common amino acid region is obtained as a highly accurate binding site. At the same time, information such as “full length interact region” can be obtained. Furthermore, using the gene (protein) information of the sequence, it is searched whether or not those genes (proteins) are registered in the PDB of the public database. If so, as shown in FIG. 14, the PDB data Confirm the consistency by comparing the interaction region calculated from the above and the “common interact region”, and confirm that it is a highly accurate binding site.

  Here, the “IVV common interact region” is a sequence in which a plurality of IVV sequences are aligned to determine a common nucleic acid (amino acid) sequence. In addition, the “IVV full length interact region” is the most from the beginning of a non-common sequence of 5 ′ sequences including a common nucleic acid (amino acid) sequence by aligning a plurality of IVV sequences to the end of a non-common sequence of 3 ′ sequences. The long area is determined.

(6) The secondary DB is composed of (5) “common interact region”, “full length interact region” and the like created by determining the interaction site. Using this information, (1) (bait) protein design can be realized. In addition, it is possible to design proteins for three-dimensional structural analysis and NMR analysis using “common interact region”, “full length interact region” and the like.

  Next, a description will be given of an example of determination of an interaction area using a primary DB of play, a method of creating a secondary DB, and use thereof.

  As an example of a method for determining the binding site based on the primary DB, there is a method of obtaining by aligning IVV sequences of genes (proteins) obtained as a prey (FIG. 12: If a plurality of IVV sequences are obtained in the same gene) High accuracy can be determined).

  In the case of bait protein design, the IVV sequences are aligned to determine an “IVV common interact region”, and the longest IVV sequence including the IVV sequences is selected and used for bait design.

  In the case of crystallization protein design for 3D NMR analysis and X-ray analysis, etc., up to “IVV full-length interact region” including “IVV common interact region” (covering all interaction regions of IVV sequence) Designing for crystallization using as a region.

  As an alignment method, a gradual multiple sequence alignment may be performed using a program such as ClustalW. This program aligns IVV sequences detected as fragments of the same gene to determine a common nucleic acid (amino acid) sequence.

  The following is a method for determining the “IVV common interact region” and “IVV full length interact region” using the play DB as the primary DB, and the interaction region “PDB interact region” in which this “IVV common interact region” is predicted by the PDB. An example of how to confirm that they match is shown.

[Select basic data to use]
1. Apply the IVV sequence to NT with BLAST, identify the gene from which the IVV sequence is derived from the Accession ID of the homologous sequence in Nt and the positional information in that sequence, and obtain the RefseqID.
2. Get the PDB ID from RefseqID in 1. The correspondence between PDB and RefseqID is taken from the information in Ensembl. If multiple PDB IDs exist in one RefseqID, all PDB IDs are acquired.
3. Obtain the full-length sequence (base) from RefseqID, then obtain the full-length amino acid sequence of the gene, and specify the reading frame.

[Identify PDB area]
4. Get the PDB file from the PDB ID of 2.
5. The PDB ID obtained in 4 is searched by PQS, and the PDB ID existing in the PQS is flagged.
6. Extract the PDB amino acid sequence of the region determined from the PDB file of 4. When multiple amino acid sequences are registered in one PDBID, all amino acid sequences are extracted separately. 7. Align 6 PDB amino acid sequences with 3 full-length amino acid sequences to identify the region of 6 PDB amino acid sequences. When a plurality of amino acid sequences exist in one PDB ID, an amino acid sequence having a longer region and higher homology is used as the PDB amino acid sequence. When a plurality of PDB IDs correspond to one full-length amino acid sequence, a region is specified for all PDB amino acid sequences.

[Identify IVV interact area]
8. Specify the region of the base sequence of the gene from which the IVV sequence is derived from the BLAST result of 1.
9. From the amino acid sequence of the full length of 3, the region of the amino acid sequence corresponding to the region specified in 8 is identified as the “IVV interact region”. When there are a plurality of IVV sequences, the region of the amino acid sequence from the most upstream to the most downstream is collectively referred to as a “full length IVV interact region”. A region of an amino acid sequence shared by all the plurality of IVV sequences is referred to as a “common IVV interact region”.
10. Compare 7 PDB amino acid region with 9 “IVV interact region” and identify the PDB ID where the two regions overlap. When a plurality of PDB IDs correspond, a PDB ID including a longer PDB amino acid sequence is selected. If the overlapping regions have the same length, select a PDB ID that has a longer PDB amino acid sequence and is flagged with 5.

[Identification of PDB interact area]
11. From the PDB IDs obtained in 10 above, select the PDB ID in which multiple amino acid sequences are registered.
12. Calculate the interatomic distance between all carbon, nitrogen, and oxygen atoms (excluding hydrogen atoms) registered in the PDB ID obtained in 11.
13. Among them, identify atoms that are on different amino acid sequences and at a distance of 4 angstroms or less.
14. The amino acid containing the atom specified in 13 is defined as “PDB interact region”.

[Visualization of interact area]
15. The PDB file with the PDB ID obtained in step 10 is saved to molecular graphics such as RasMol.
Load the program to visualize the 3D structure.
At that time, the “IVV interact region”, “IVV full length interact region”, “IVV common interact region” obtained in 9 and “PDB interact region” obtained in 14 are displayed so that they can be identified.

  Moreover, according to the IVV method, the selected play can be efficiently diverted to a bait. Thereby, verification of the mode of interaction (direct or indirect, etc.) by exchanging bait and play and expansion of the interaction network can be performed efficiently.

An example of the procedure for diversion from play to bait will be described with reference to FIG.
A primer set (1st 5) prepared by PCR (1st PCR) using a plasmid containing the prey gene selected by the IVV method as a template (see Genome Research vol.15 pp710-717 (2005)). Using “primer and 1st 3 ′ primer”, DNA fragments are prepared by adding enhancer + T7tag on the 5 ′ side of each gene and a part of calmodulin-binding protein (CBP) on the 3 ′ side. Further, a DNA having a CBP + zz domain (CBPzz fragment) is prepared by PCR (2nd PCR) via the 3′-side tag of the DNA fragment. Both fragments are ligated by overlap PCR (3rd PCR). Perform PCR (4th PCR) to add SP6 promoter to the 5 'end of the 2nd PCR product, and prepare a template DNA for transcription. Perform transcription reaction using the prepared template DNA to prepare mRNA.

As described above, according to the present invention, since bait can be produced from the obtained prey rather than analysis of the prey protein from the bait protein, efficient large-scale analysis of protein-protein interaction and thereby Accumulate data in DB and improve data reliability. In addition, since the large-scale analysis system can efficiently accumulate data in the DB, it can provide a large-scale DB of region information necessary for bait protein design and a method for creating the same. Large-scale database usage includes not only bait protein design for protein-protein interaction analysis, but also X-ray analysis of the three-dimensional structure of proteins including interaction regions (functional regions), protein design necessary for NMR analysis, and In the three-dimensional structure analysis for drug discovery (drug design), the three-dimensional structure analysis of the complex that interacts with proteins and ligands is important, and the protein design necessary for protein crystallization technology for that purpose, etc. Application to can be provided. That is, a protein having a long amino acid length or a structure that is unstable at the full length and difficult to crystallize is often crystallized by cloning only a functional region and partially translating it. In the conventional methods, it has been necessary to clone various regions of the full length of amino acids with various lengths and experimentally confirm their binding ability. In addition, the protein that forms the dimer is stabilized by interacting with the paired protein and may be successfully crystallized, and in some cases, identification of the paired protein is also necessary. The sequence obtained by the IVV method indicates an interaction functional site of the protein. If there are a plurality of IVV sequences, the prediction of the interaction functional site becomes more accurate. Furthermore, a pair of proteins that interact with the site have already been identified as baits and can be used for crystallization as they are. The protein coding region specified using the IVV sequence is used as a region used for crystallization, which enables efficient three-dimensional structural analysis.

  Hereinafter, examples of the present invention will be specifically described. However, the following examples should be regarded as an aid for obtaining specific recognition of the present invention, and the scope of the present invention is limited by the following examples. It is not something.

<Reference Example 1>
The example which examined the selective exon using the library which included several partial arrangement | sequences about one gene is shown.

As an index of the partial sequence ratio for one gene in the library, the sequence analysis is performed for the partial sequences in the library using a sequencer, and the calculation is performed as follows for the mouse genes registered in the Entrez gene DB. Single gene rate was used.
Single gene rate: total number of sequences that hit a single gene / total number of sequences that hit a gene

  Mouse Fos was used as a bait protein, and the library was subjected to IVV selection using mRNA extracted from mouse tissue, and the base sequence was analyzed by a sequencer. Specifically, it was performed as described in Example 1 below.

  Subsequently, gene analysis was performed using an IVV analysis system (In Vitro Virus Analysis System: IWAS) (Genome Research vol.15 pp710-717 (2005)).

  As selection conditions, experiment conditions (1) 4 cycles, 1 stage selection, experiment conditions (2) 4 cycles, 2 stages selection experiment conditions, sequence analysis under each condition, IVV sequence data single gene rate Asked. The results are shown in FIG. Under experimental condition (1), 241 sequences were analyzed and the single gene rate was 42%. In condition (2), 131 sequences were analyzed and the single gene rate was 24%.

  With respect to the obtained IVV sequence data, variant analysis of alternative splicing was performed as follows using a public database.

 (a) From the ASD (Alternative Splicing Database), only selectively changing exon regions were extracted and saved as FASTA format files.

(b) The homology search of the IVV sequence was performed by BLAST against the selective exon of (a).
In this case, the threshold was set to an expected value of 0.001 or less, a homology of 90% or more, and a region length of 60 bp or more. As a result, correspondence between IVV sequences and selective exons was obtained.

(c) The duplicated ones existing in the CDS region were integrated, and the gene in which the IVV sequence and the corresponding selective exon existed was extracted.

  The results for Eefld are shown in FIG. Under the experimental condition (1), the sequence of Eef1d is one and the interaction site cannot be determined accurately. However, under the experimental condition (2), multiple Eef1d are obtained, so the interaction site is very accurate. Can be determined so that selective exons can be identified.

<Reference Example 2>
An example in which SNP was examined using a library containing several partial sequences for one gene is shown.

  As in the case of Reference Example 1, the conditions were set using the single gene rate as an index, using human Fos instead of mouse Fos as the bait and mRNA extracted from human brain tissue as the library instead of mRNA extracted from mouse brain tissue. It was investigated. However, the single gene rate was calculated for human genes registered in Entrez gene DB.

  As selection conditions, experiment conditions (3) 2 cycles, 1 stage selection, experiment conditions (4) 4 cycles, 1 stage selection, experiment conditions (5) 5 cycles, 2 stages selection experiment conditions, sequence analysis under each condition The single gene rate was determined in the IVV sequence data. The results are shown in FIG. Under experimental condition (3), 72 sequences were analyzed and the single gene rate was 79%. Under experimental condition (4), 74 sequences were analyzed, and the single gene rate was 44%. Under experimental condition (5), 57 sequences were analyzed, and the single gene rate was 33%.

  About the obtained IVV sequence data, SNPs analysis was performed as follows using a public database.

 (a) Homology search of IVV sequence was performed by BLAST against RefSeq sequence. In this case, the threshold was set to 10 or less for the expected value and 90% or more for the homology. From this, a correspondence table between IVV sequences and RefSeq IDs was created.

 (b) Since the corresponding RefSeq ID is assigned to the SNP sequence of dbSNP, the data was extracted and a correspondence table of SNP and RefSeq ID was created.

 (c) RefSeq IDs that matched both correspondence tables were extracted, and multiple IVV sequences and multiple SNPs might correspond to one RefSeq ID, so they were integrated.

(d) Next, the peripheral sequence (FASTA format) of the corresponding SNP was extracted from dbSNP. IVV sequence and SNP
Peripheral sequences were aligned by SIM4, and the actual positional relationship was confirmed.
As a result, it was confirmed that SNP was located on the IVV sequence.

 (e) Among them, those existing in the coding region (cSNP) were selected, and those that change the amino acid (Non-synonymous cSNP) were selected.

  The results for JUN are shown in FIG. Under experimental condition (3), no JUN sequence was detected. In experimental condition (4), one was detected, so interaction analysis was possible, but the exact interaction site could not be extracted and SNPs could not be narrowed down. In the experimental condition (5), since multiple lines were detected, it was confirmed that there was one non-synonymous cSNP corresponding to the interaction site with JUN's FOS (Fig. 9).

From the above, it is possible to efficiently obtain a library containing various partial sequences for one gene, such as changes in DNA sequence and mRNA sequence by selective splicing, changes in DNA sequence and mRNA sequence derived from SNPs, etc. It turns out to be important to follow.

<Example 1>
Mouse c-Fos was used as a bait protein, and the library was subjected to IVV selection using mRNA extracted from mouse brain tissue (Genome Research vol.15 pp710-717 (2005)). Specifically, it was as follows.

<1-1> Preparation of cDNA library
According to the method of Chechik et al. 1996 BioTechniques 21: 526-534, from the poly (A) + mRNA library derived from mouse brain (Clontech) or the poly (A) + mRNA library derived from human brain (Cosmo Bio), primers ( Reverse transcription (RT) and amplification by random priming using TCATCGTCCTTGTAGTCAAGCTTNNNNNNNNN (SEQ ID NO: 1)) was performed to prepare a cDNA library. 5 'phosphorylated dsDNA was obtained using SuperScript II Double Strand cDNA Synthesis Kit (Invitrogen). In 5′-phosphorylated dsDNA, the adapter-bound cDNA library is used as a primer 5 ′ by utilizing the fact that a synthetic double-stranded adapter (GGAATTCG and GAACAACAACAACAACAAACAACAACAAAATGGCTAGCATGACTGGTGGACAGCAAATGGCGAATTCC (SEQ ID NO: 2)) binds to the 5 ′ end of the sense cDNA strand. Amplification was carried out by PCR using IVVlib (GGAAGATCTATTTAGGTGACACTATAGAACAACAACAACAACAAACAACAACAAAATG (SEQ ID NO: 3)) and 3′IVVlib (ttttttttcttgtcgtcatcgtccttgtagTCAAGC (SEQ ID NO: 4)).

  Subsequently, transcription of the cDNA library and ligation with a PEG Puro spacer were performed (Miyamoto-Sato et al. 2003 Nucleic Acids Res. 31: e78).

<1-2> Preparation of Bait RNA Template The DNA template for bait Fos (parts of amino acids 118-211) was obtained from pCMV-FosCBPzz (Miyamoto-Sato et al. 2003 Nucleic Acids Res. 31: e78). Primer 5'FosCBPzz (gaatttaggtgacactatagaaACAATTACTATTTACAATTACAatggctagcatgactggtggacag (SEQ ID NO: 5)) and primer 3'FosCBPzz (GGATCTCCATTCGCCATTCA (SEQ ID NO: 6) using the sequence (Sleat et al. 1988 Nucleic Acids Res. 16: 3127-3140) , 1 min, and (98 ° C., 20 sec; 55 ° C., 1 min; 72 ° C., 4 min) were prepared by PCR under conditions of 15 cycles. This template contains part of Fos and calmodulin binding protein (CBP) and protein A (z domain) for the TAP method.

The DNA template for bait Mdm2 (amino acids 1-188) was prepared from human brain poly (A) + mRNA library (Cosmo Bio) by cloning. That is, after reverse transcription, PCR (94 ° C
2 min; (94 ° C 30 sec, 62 ° C 30 sec, 73 ° C 2 min) 15 cycles; 73 ° C 15 min) with primers human_5'T7 + mdm2 (atggctagcatgactggtggacagcaaatgggtcgcggatccATgTgCAATACCAACATgTCTgTAC (SEQ ID NO: 7)) and human_3'mdm2-564 ( actatcagatttgtggcgttttctttgtcgttcacc (SEQ ID NO: 8)). Further, from pCMVFosCBPzz, PCR (same conditions as in the above PCR) was performed with primers human_5'mdm2 + CBP (ggtgaacgacaaagaaaacgccacaaatctgatagtctcgagctcaagagaagatggaaaaagaatttcatag (SEQ ID NO: 9)) and 3'FosCBPzz. Using both amplified fragments, overlap PCR (same conditions as the above PCR) using primers human_5'T7 + mdm2 and 3'FosCBPzz, primer 5'SP6 (O ') T7 (GAATTTAGGTGACACTATAGAAacaattactatttacaattacaATGGCTAGCATGACTGGTGGACag (SEQ ID NO: 10)) and 3′FosCBPzz were used for further PCR (same conditions as in the above PCR).

  From these DNA templates, mRNA templates were prepared by the method described in Miyamoto-Sato et al. 2003 Nucleic Acids Res. 31: e78.

<1-3> Cell-free cotranslation and IVV selection Cell-free cotranslation is performed using 200 nM mouse or human brain IVV mRNA library (prey), 400 nM fos or mdm2 mRNA (bait) prepared above, 1 h at 26 ° C. with Wheat Germ Extract (Promega) in the presence or absence of 400 nM DNA (5 repeat TGACTCA sequence (Chinenov and Kerppola 2001) that binds to Jun / Fos AP1 complex). Two-stage or one-stage purification (selection) was performed according to the TAP method (Rigaut et al. 1999 Nat. Biotechnol. 17: 1030-1032).

  The IVV library obtained by the selection was subjected to RT-PCR with primers 5′IVVlib and 3′IVVlib using a one-step RT-PCR kit (QIAGEN) to amplify the library. RT-PCR products were purified with CHROMA SPIN-1000 Columns (Clontech) and cloned using PCR Cloning Kit (QIAGEN).

  For sequencing, colony PCR was performed using vector-specific primers (M13R: gAAACAgCTATgACCATgATTACg (SEQ ID NO: 11) and M13F: gTTTTCCCAgTCACgACgTTg (SEQ ID NO: 12)) using Qpix (GENETIX) and GenePCR System 2700 (Applied Biosystems). . The purification of the colony PCR product was performed using BioRobot 8000 (QIAGEN), and the sequence of the product was determined using ABI PRISM 3100 Genetic Analyzer (Applied Biosystems).

  For each round, the RT-PCR conditions (number of cycles) for amplifying the library template were examined. When the initial number of cycles is 59% or more when the amount of DNA template amplified by RT-PCR is 100% saturation (Table 1), as shown in FIG. In the third round (stars in FIG. 10), the number of cycles became constant. In the selection after the number of cycles has become constant, the single gene rate starts to decrease significantly, especially when the selection is performed twice, a sufficient part of one gene that is suitable for accurately extracting the interaction site. It was found that a library containing the sequence was obtained. In addition, in the two-stage selection (protein A and calmodulin), when the second-stage selection is performed once again from the third round (the asterisk in FIG. 10) where the number of cycles is constant, the interaction site is accurately extracted. It was found that a library containing sufficient partial sequences for one gene was obtained.

  The results when human MDM2 is used are shown in Table 2. Human MDM2 is used as a bait protein, and the library uses mRNA extracted from human brain tissue to perform one-stage selection of IVV, and after two cycles of selection after the number of cycles is constant, as in FIG. It was found that a library including a sufficient partial sequence for one gene was obtained.

<Reference Example 3>
Taking “FOS / JUN” as an example, “IVV common interact region” and “IVV full length interact region” were determined.

Using FOS as a bait, IVV selection was performed using mRNA extracted from human brain tissue as a library (Genome Research vol.15 pp710-717 (2005)). As a result, the following IVV (1) to IVV (9) were obtained as the IVV sequence for the prey gene JUN (Refseq ID: NM — 002228.2) obtained as a result (FIG. 12). By aligning IVV (1) to IVV (9), “IVV common interact region” and “IVV full length interact region” were obtained. This “IVV common interact region” is an interaction region predicted by PDB (amino acid numbers 286 to 318 (286, 287, 290, 291, 293, 294, 297, 298, 300, 301, 304, 305, 307, 308, 311, 312, 314, 315, 318)).

Play JUN (Refseq ID: NM_002228.2) "IVV interact area":
IVV (1): 050517_5_D1_C05_5_D1C5_034.seq; 284-327
IVV (2): 050517_5_D1_H05_5_D1H5_044.seq; 257-326
IVV (3): 050519_m5_2_G09_m5_2_G9_068.seq; 282-326
IVV (4): D1_A12.seq; 257-326
IVV (5): D1_D12.seq; 284-327
IVV (6): D1_E02.seq; 277-331
IVV (7): H1_M17.seq; 284-327
IVV (8): d1_5th_c04.seq; 277-331
IVV (9): d1_5th_f12.seq; 257-326

Play JUN's “IVV Common Interact Area”: 282-226
Play JUN's “IVV Full Length Interaction Region”: 257-331

<Reference Example 4>
Using c-Fos as a bait protein and mRNA extracted from mouse brain tissue as a library, IVV selection was performed and sequence analysis was performed using a sequencer. Subsequently, gene analysis was performed with an IVV analysis system (IWAS) to obtain the primary DB of play (Genome Research vol.15 pp710-717 (2005)).

Play (Jun / Tro1 / Tro2 / Eef1d / Ship1 / Opt / JunD / C130020M04Rik / Psmc5 / FLJ32000 / Snapc5 / Rit2) that confirmed the interaction with bait c-Fos (Figure 13A: Pull-down play and bait c-Fos Based on the primary DB, the “IVV common interact region” was determined, and the longest IVV sequence containing it was selected.
Selected sequences are shown below.

1. Jun
DNA sequence (SEQ ID NO: 13)
GCGGATCAAGGCAGAGAGGAAGCGCATGAGGAACCGCATTGCCGCCTCCAAGTGCCGGAAAAGGAAGCTGGAGCGGATCGCTCGGCTAGAGGAAAAAGTGAAAACCTTGAAAGCGCAAAACTCCGAGCTGGCATCCACGGCCAACATGCTCAGGGAACAGGTGGCACAGCTTAAGCAGACAAGGTCATGAACCGCTGACAG
Amino acid sequence (SEQ ID NO: 14)
RIKAERKRMRNRIAASKCRKRKLERIARLEEKVKTLKAQNSELASTANMLREQVAQLKQKVMNHVNSGCQLMLTQQ

2. Tro1
DNA sequence (SEQ ID NO: 15)
GATTGAGGCATCTAACAGACAGATTGGGGCATCTAACAGACAAACTGAGGCATCTAACAGACAGATTGGGCATCTAACAGACAGACTGAGGCATCTAACAGACAGATTGGGGCATcTAACAGACAGACTGATGCATCTAACAGACAGACTGATGCATCTAACAGACAGACTGAGGCATCTAGCAGACAGACAGAGGCATG
Amino acid sequence (SEQ ID NO: 16)
IEASNRQIGASNRQTEASNRQIGHLTDRLRHLTDRLGHLTDRLMHLTDRLMHLTDRLRHLADRQRHLADRL

3. Tro2
DNA sequence (SEQ ID NO: 17)
GAGGCATCTAACAGACAGATTGGGGCATCTAACAGACAGACTGAGGCATCTAACAGACAGATTGGGGCATCTAACAGACAGACTGATGCATCTAACAGACAGACTGAGGCATCTAGCAGACAGACAGAGGCATCTAGCAGACAGGCAGAGGCAC
Amino acid sequence (SEQ ID NO: 18)
RHLTDRLGHLTDRLRHLTDRLGHLTDRLMHLTDRLRHLADRQRHLADRQRH

4. Eef1d
DNA sequence (SEQ ID NO: 19)
CCACAGTGAGCTCATTGTGAGGATTACCAGTCTGGAAGTGGAGAATCAGAACCTTCGAGGCGTGGTGCAAGATTTGCAGCAGGCCATTTCCAAGTTGGAGGCCCGGCTGAGCTCTCTAGAGAAGAGTTCACCTACTCCCCGAGCCACGGCCCCACAGACCCG
Amino acid sequence (SEQ ID NO: 20)
HSELIVRITSLEVENQNLRGVVQDLQQAISKLEARLSSLEKSSPTPRATAPQTR

5. Schip1
DNA sequence (SEQ ID NO: 21)
CCCACACATGCCTCACATAAGCGAATGTTTGATGAAAAGAAGCTTAAAGCCCACCGACCTGAGAGACATGACTATCGGGCAGCTACAAGTGATCGTCAATGACCTCCACTCCCAGATTGAGCGTTTGAATGAAGAGTTGGTCCAGCTGCTCCTTATTCGAGATGAGCTGAGCACATGAGCGATGAGATTCCATGCTGGA
Amino acid sequence (SEQ ID NO: 22)
PHMPHISECLMKRSLKPTDLRDMTIGQLQVIVNDLHSQIERLNEELVQLLLIRDELHTEQDAMLVDIEDLTRHAEKEQ

6. Optinurin
DNA sequence (SEQ ID NO: 23)
GCTGAAGACCCAGGTGGAGCAGGAAGTGGAGCATCTGAAGATCCAGGTGATGCGCCTTCGGGCTGAAAAGGCAGACCTGCTGGGCATCGTCTCAGAACTGCAGCTCAAACTCAACTCCGGCGGCTCCTcGGAAGACTCCTTcGTTGAGATCAGGATGACCGAAGGAGAGACTGAAGGGGCAATGAAGGAGATGAAGAGCTGCCCTACACCCaCAAGAACAGACCCCATCAGCTTGAGCAACTGTACAGAGGATGCCAGGAGTTGTGCGGAGTTTGAAGAACTGACTGTGAGCCAGCTTcTGCTTTGCCTAAGGGAAGGAAACCAA
Amino acid sequence (SEQ ID NO: 24)
LKTQVEQEVEHLKIQVMRLRAEKADLLGIVSELQLKLNSGGSSEDSFVEIRMTEGETEGAMKEMKSCPTPTRTDPISLSNCTEDARSCAEFEELTVSQLLLCLREGNQ

7. JunD
DNA sequence (SEQ ID NO: 25)
CAAGAGGCTGCGCAACCGCATCGCCGCCTCCAAATGCCGCAAGCGCAAGCTGGAGCGTATCTCGCGCCTGAAGGAGAAAGTCAAGACCCTCAAAAGCCAGAACACCGAGCTGGCGTCCACCGCCAGCCTGCTGCGCGAGCAGGTGGCGCAGCTCAAACAGAAAGTCCTCAGCCACGTCAACAGCGCCTGCCAGC
Amino acid sequence (SEQ ID NO: 26)
KRLRNRIAASKCRKRKLERISRLKEKVKTLKSQNTELASTASLLREQVAQLKQKVLSHVNSGCQLLPQHQA

8. C130020M04Rik
DNA sequence (SEQ ID NO: 27)
CGATGCCGGCAAAAGCGGAAACTGTGGGTGTCCTCCCTGGAAAAGAAGGCAGAAGAACTTACTTCTCAGAACATTCAGCTGAGTAATGAAGTCACATTACTACGCAATGAGGTGGCTCAGCTGAAGCAGCTACTGTTAGCTCATAAAGATTGTCCAGTCACCGCACAA
Amino acid sequence (SEQ ID NO: 28)
RCRQKRKLWVSSLEKKAEELTSQNIQLSNEVTLLRNEVAQLKQLLLAHKDCPVTAQ

9. Psmc5
DNA sequence (SEQ ID NO: 29)
GCCAGAATCTCCGTAGACTGCAGGCACAGAGGAATGAGCTGAATGCAAAAGTTCGCCTGTTGCGGGAGGAGCTGCAGCTGTTGCAGGAACAGGGCTCCTACGTTGGAGAAGTCGTGAGGGCAGA
Amino acid sequence (SEQ ID NO: 30)
RQNLRRLQAQRNELNAKVRLLREELQLLQEQGSYVGEVVRAE

10. FLJ32000
DNA sequence (SEQ ID NO: 31)
GGAAGAAGTTCAGCAGCCTGCAGAAGGAGATCGAAGAGCGCACCAACGACATCGAGCTCCTCAAGTCGGAGCGGATGAAGCTGCAGGGCATCATCAGATCCCTGGAGAAAGACATCCAAGGGCTCAAGAGAGAGATCCAGGAGAGGGACGAGACCATTCAAGACATGGAGAAGCTGAACTG
Amino acid sequence (SEQ ID NO: 32)
RKKFSSLQKEIEERTNDIELLKSERMKLQGIIRSLEKDIQGLKREIQERDETIQDMEKLDYKDDYNSNLEIKL

11. Snapc5
DNA sequence (SEQ ID NO: 33)
CGCCCTAAGGAAGTGGAAGGGGACGTTGAGTCGACTGCAGGAGCTCCGCAAGGAGGTGGAAACCCCGCTGCGTcTAAAGGCGGCTCTACACGACCAACTGAACCGCCTCAAGGTTGAAGAATTAGCCCTTCAATCCATGATAAATTCTCGAGGAAGGACCGAGCAATGGTCG
Amino acid sequence (SEQ ID NO: 34)
ALRKWKGTLSRLQELRKEVETPLRLKAALHDQLNRLKVEELALQSMINSRGRTETLSSQPAPEQLCDMSLHVDNEVTINQTRL

12. Rit2
DNA sequence (SEQ ID NO: 35)
TGCAGCCCTGCGATTCGGTATCGATGATGCTCTTCAAGGCTTAGTGAGAGAAATTCGCAGGAAGGAATCCATGCTGCCCTTGGTGGAAAGGAAATTGAAGAGGAAGGACAGCCTGTGGAAGAAGATAAAAGCCTCCCTGAAGAAGAAGAGG
Amino acid sequence (SEQ ID NO: 36)
AALRFGIDDALQGLVREIRRKESMLPLVERKLKRKDSLWKKIKASLKKKR

  A bait template was created from the prey gene having the IVV sequence thus selected (FIG. 14), and the function as a bait was verified by pull-down. As a result, it was confirmed that 11 out of 12 cells could function as baits except for one clone (C130020M04Rik) that did not express (FIG. 13B: interaction between each bait and pre-c-Fos by pull-down) Confirmation). This shows that the bait design by this method is very efficient.

Details of each procedure were as follows.
1. Preparation of bait protein mRNA By PCR (1st PCR) using ExTaq (Takara) using a plasmid with the selected clone as a template (see Genome Research vol.15 pp710-717), enhancer is placed on the 5 'side of each gene. A DNA fragment in which calmodulin-binding protein (CBP) was partially added to the 3 ′ side was prepared and purified by QIAquick PCR Purification Kit (QIAGEN). DNA having a CBP + zz domain (CBPzz fragment) was bound by overlap PCR (2nd PCR) via the 3′-side tag of the DNA fragment, and treated in the same manner as 1st PCR. PCR to add SP6 promotor to the 5 'end of the 2nd PCR product (3rd
PCR) and a template DNA for transcription was prepared by processing in the same manner as 1st PCR.

  Using the prepared template DNA and Cap analoge (Invitrogen), transcription reaction was performed by RiboMAX Large Scale RNA Production System-SP (Promega) to prepare mRNA. mRNA was purified with RNeasy Mini Kit (QIAGEN).

2. Preparation of pre-c-FOS mRNA PCR using ExTaq (Takara) with plasmid encoding bait c-FOS
A DNA fragment encoding c-FOS was amplified by (4th PCR, primer set: Table 4) and purified by QIAquick PCR Purification Kit (QIAGEN). PCR using 4th PCR product as template
(5th PCR) was performed and purified with the QIAquick PCR Purification Kit (QIAGEN). Transcription was carried out using this 5th PCR product, RiboMAX Large Scale RNA Production System-SP (Promega) and Cap analoge (Invitrogen) to prepare mRNA. mRNA was purified with RNeasy Mini Kit (QIAGEN).

3. Pull-down experiment Bait protein and prey protein from the prepared mRNA in wheat germ extract (Promega)
Prepared by in vitro translation. At this time, the prey protein was fluorescently labeled at the C-terminus. Bait protein and prey protein were mixed, and the bait protein was fixed to IgG beads (SIGMA). After washing the immobilized beads, the bond between the beads and the bait protein-prey protein complex was cleaved, and the prey protein bound to the bait protein was detected by electrophoresis.

It is a block diagram of extraction of the gene from which the interaction between proteins changes by the change of a base sequence. It is explanatory drawing of a variant analysis. It is explanatory drawing of a fusion gene analysis. It is explanatory drawing of SNPs analysis. It is explanatory drawing of a mutated gene and RNA editing analysis. The condition examination result of the reference example 1 is shown. The analysis result of the selective exon in the interaction site of Eef1d is shown. The condition examination result of the reference example 2 is shown. The analysis result of SNP in the interaction site of JUN and FOS is shown. The relationship between the change in the single gene rate and the number of RT-PCR rounds is shown. It is explanatory drawing of a protein interaction large-scale analysis system. It is an example of the coupling | bonding area | region using primary DB. The result of a pull-down experiment is shown. It is explanatory drawing of the procedure of template production for bait from a play gene. It is explanatory drawing of an example of a screening method.

Claims (4)

(1) a step of producing a correspondence molecule library in which a protein and a nucleic acid encoding the protein are bound from a nucleic acid library, (2) a step of mixing the correspondence molecule library and a target substance, (3) Repeat the round including the step of selecting the mapping molecule bound to the target substance, (4) the step of preparing the nucleic acid library by RT-PCR or PCR from the nucleic acid of the mapping molecule selected in (3) Including
In the RT-PCR or PCR, the number of cycles in which the amplification amount of the amplification product first reaches a certain amount less than or equal to the saturation amount is obtained, and after the number of cycles reaches a certain value, one or more rounds are passed. A method for obtaining sequence data, characterized in that the sequence analysis of a nucleic acid library is sometimes performed. A method of acquiring sequence data using a library of mapping molecules in which a protein and a nucleic acid encoding the protein are bound, comprising: (a) designing a bait protein; and (b) interacting with the designed bait protein (C) Sequence analysis of the corresponding molecules selected in the selection, obtain sequence data, (d) determine the interaction site based on the obtained sequence data, ( e) extracting sequence data based on the determined interaction sites, wherein steps (b) and (c) are performed by the method of claim 1. A method for designing a protein, comprising extracting an interaction site based on sequence data and designing a bait protein based on the extraction result, wherein the sequence data is obtained by the method according to claim 1 or 2. A method for extracting a gene, comprising obtaining sequence data by the method according to claim 1 or 2, and using the obtained sequence data to extract a gene whose protein-protein interaction changes due to a change in base sequence .

Images