JP4747245B2 - Enzymatic construction method of RNAi library - Google Patents

Description

The present invention relates to a method for constructing an RNAi library based on a DNA expression vector, and more particularly, to a method for enzymatically constructing a systematic library for a target sequence.
The present invention relates to a screening method for selecting an appropriate iRNA expression construct capable of silencing a target gene or the like using the enzymatically constructed RNAi library.

The reverse genetics approach that determines the function of each gene by loss of function using the large amount of genomic data currently available (Non-Patent Documents 1-5) completely elucidates the relationship between genes and functions. It has become an important technology. Techniques for implementing a reverse genetics approach include the creation of knockout animals by gene targeting, or antisense techniques.
Gene targeting technology based on homologous recombination (Non-Patent Document 6) is widely used to determine the functions of various genes. However, since the process requires labor, it cannot be applied to simple and wide use. In addition, antisense oligonucleotides can be used more conveniently, but their introduction often results in toxicity, instability, and non-specific effects (Non-Patent Document 7).
On the other hand, RNA interference (RNAi), which is a gene suppression phenomenon induced by double-stranded RNA (Non-Patent Document 8), is usually specific and gene suppression with unprecedented speed in a wide range of organisms. Since the body can be obtained, it has become an attractive alternative to gene targeting technology and antisense technology (Non-Patent Documents 7 and 9). RNAi has also been applied to genome-wide reverse genetics in C. elegans (Non-patent Document 10). However, in higher vertebrates, the use of RNAi was initially limited to invertebrates, as long (> 30 nucleotides) double-stranded RNAs elicit harmful interferon responses. This problem was overcome by the finding that short (21-23 nucleotides) double-stranded RNA, called short interfering RNA (siRNA), induces RNAi without causing adverse events in mammals (Non-Patent Document 11). ). The transient nature of the oligo siRNA effect has been overcome by the development of DNA vectors that express siRNA or short hairpin double-stranded RNA (shRNA) mimicking siRNA in cells (non- Patent Document 12-17).

Consortium, TCeS Genome sequence of the nematode C. elegans: a platform for investigating biology.Science 282, 2012-2018 (1998). Adams, MD et al. The genome sequence of Drosophila melanogaster. Science 287, 2185-2195 (2000). Waterston, RH et al. Initial sequencing and comparative analysis of the mouse genome.Nature 420, 520-562 (2002). Lander, ES et al. Initial sequencing and analysis of the human genome.Nature 409, 860-921 (2001). Venter, JC et al. The sequence of the human genome.Science 291, 1304-1351 (2001). Capecchi, MR The new mouse genetics: altering the genome by gene targeting.Trends Genet. 5, 70-76 (1989). Opalinska, JB & Gewirtz, AM Nucleic-acid therapeutics: basic principles and recent applications. Nat. Rev. Drug Discov. 1, 503-514 (2002). Fire, A. et al. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans.Nature 391, 806-811 (1998). Dykxhoorn, DM, Novina, CD & Sharp, PA Killing the messenger: short RNAs that silence gene expression.Nat. Rev. Mol.Cell Biol. 4, 457-467 (2003). Kamath, RS & Ahringer, J. Genome-wide RNAi screening in Caenorhabditis elegans.Methods 30, 313-321 (2003). Elbashir, SM et al. Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells.Nature 411, 494-498 (2001). Brummelkamp, TR, Bernards, R. & Agami, R. A system for stable expression of short interfering RNAs in mammalian cells.Science 296, 550-553 (2002). Miyagishi, M. & Taira, K. U6 promoter-driven siRNAs with four uridine 3 'overhangs efficiently suppress targeted gene expression in mammalian cells. Nat. Biotechnol. 20, 497-500 (2002). Yu, JY, DeRuiter, SL & Turner, DL RNA interference by expression of short-interfering RNAs and hairpin RNAs in mammalian cells.Proc. Natl. Acad. Sci. USA 99, 6047-6052 (2002). Sui, G. et al. A DNA vector-based RNAi technology to suppress gene expression in mammalian cells.Proc. Natl. Acad. Sci. USA 99, 5515-5520 (2002). Lee, NS et al. Expression of small interfering RNAs targeted against HIV-1 rev transcripts in human cells. Nat. Biotechnol. 20, 500-505 (2002). Paul, CP, Good, PD, Winer, I. & Engelke, DR Effective expression of small interfering RNA in human cells. Nat. Biotechnol. 20, 505-508 (2002).

Despite the development of RNAi techniques, there are still no general rules for designing siRNA / shRNA expression constructs with effective gene silencing activity, so it takes time to identify a suitable construct. It costs money.
In addition, RNAi was considered to be specific, but depending on the designed sequence, a phenomenon called “off-target effect” that suppresses specific genes other than the target has been detected. Therefore, in order to avoid the off-target effect, it has become important to design siRNA or shRNA for which sequence in the target gene.

  An object of the present invention is to provide a method for producing a library of shRNA expression constructs capable of performing systematic screening, and a screening method capable of selecting an expression construct having a desired silencing activity using this library. . For this purpose, the inventors have developed a technique called EPRIL (enzymatic production of RNAi library). That is, a target gene or the like is cut at random, and adapters are connected to both ends of the random fragment. Among these, by making any one adapter into a hairpin structure, it becomes a fragment having a hairpin type structure with a random fragment at the center. By extending the fragment converted into the hairpin structure with a polymerase having strand displace activity, an iRNA expression construct capable of mainly expressing the hairpin double-stranded RNA is produced.

In addition, the present inventors fused a target gene with a reporter gene, and used a screening system capable of selecting target gene silencing using reporter gene silencing as an indicator, and also a thymidine kinase gene, which is a negative selection marker. Thus, a technology capable of selecting an shRNA expression construct having the most effective silencing activity from the RNAi library has been developed. Furthermore, the present inventors have also shown that the library preparation method of the present invention can be applied to the preparation of an RNAi library directly from a cDNA library. That is, the present invention is as follows.
1. A method for producing an RNAi library from a desired target DNA, comprising the following steps (1) to (3):
(1) a step of randomly cutting a target DNA to generate a DNA fragment;
(2) connecting a hairpin adapter to one end of the DNA fragment to generate a hairpin DNA fragment;
(3) A method for producing an RNAi library, comprising a step of performing primer extension on a hairpin type DNA fragment using a polymerase having Strand-displacing activity to generate an iRNA expression construct encoding an interference RNA.
2. including a step of connecting a double-strand break type adapter to the other end of the DAN fragment before or after the step (2);
Either one of the double-strand break adapter or the hairpin adapter is provided with a restriction enzyme recognition site of an outside cutter,
The adapter having the restriction enzyme recognition site is first connected to the DNA fragment, and the other adapter is connected to the end formed by adjusting the length of the DNA fragment by the restriction enzyme of the outside cutter. The method according to 1 above.
3. The method according to 2 above, wherein the hairpin adapter is provided with a restriction enzyme recognition site of an outside cutter.
4. The method according to 2 above, wherein the double-strand adapter is provided with a restriction enzyme recognition site for an outside cutter.
5. The method for producing an RNAi library according to 1 above, further comprising a step of connecting the iRNA expression construct to an expression vector after primer extension.
6). 6. The method for producing an RNAi library according to 5 above, wherein the expression vector can function in a mammal.
7). 7. The method for producing an RNAi library according to 5 or 6 above, wherein the expression vector has an integration activity into the host chromosome.
8). 8. The method for producing an RNAi library according to any one of 1 to 7 above, wherein the target DNA is any one of a cDNA of a specific gene, a cDNA library, or a cDNA after subtraction.
9. 2. The method for producing an RNAi library according to 1 above, wherein in the step (1), the target DNA is fragmented using a DNA digestion enzyme that can be used for shotgun cloning.
10. The method for producing an RNAi library according to any one of 1 to 9 above, wherein the restriction enzyme of the outside cutter cleaves at least 19 bp away from the recognition site in the first adapter.
11. 11. The method for producing an RNAi library according to any one of 1 to 10 above, wherein the polymerase having strand-displacing activity is heat resistant.
12 12. The method for producing an RNAi library according to any one of 1 to 11, wherein the polymerase having strand-displacing activity is either Bst polymerase or Vent polymerase.
13. 13. An RNAi library produced by the method according to any one of 1 to 12 above.
14. A method for screening an siRNA expression construct having a desired RNAi activity from an RNAi library produced by the method according to any one of 1 to 12 above,
Introducing the RNAi library into cells expressing the target DNA;
Measuring the expression of the target DNA.
15. 14. The screening method according to 14 above,
As a fusion gene with target DNA fused to a reporter gene,
A screening method, wherein in the step of measuring the expression of the target DNA, the expression of the target DNA is measured using the activity of the reporter protein as an index.
16. A method for screening an siRNA expression construct having a desired RNAi activity from an RNAi library produced by the method according to any one of 1 to 12 above,
Introducing the RNAi library into a cell expressing a fusion gene obtained by fusing a target DNA and a negative selection marker gene;
Screening with a step of performing selection with a negative selection marker and selecting only cells having an RNAi effect.
17. A system for producing an RNAi library,
Including a hairpin adapter and a double-strand adapter,
A system in which any of the adapters is provided with a restriction enzyme recognition site for an outside cutter.
18. 18. The system according to 17 above, further comprising a restriction enzyme of an outside cutter and / or a polymerase having a strand-displacing property.

It is a figure which shows the outline | summary of EPRIL. (A) Illustrated protocol for enzymatic induction of cDNA into shRNA expression library. (B) PAGE of DNAs in the process of enzymatic induction. M, 25 bp ladder DNA marker. Lanes 1, 2 and 3 show the product after ligation and MmeI digestion of the first adapter, the adapter 2-ligation fragment, and the fragment after the polymerase reaction. For each lane, the product is indicated by the tip of the arrow. FIG. 5 shows silencing of GFP expression by RNAi library prepared from cDNA encoding GFP. (A) Histogram showing RNAi efficiency distribution. Bars represent the number of shRNA expression constructs. Circles indicate standardized cumulative frequency. (B) The amount per fraction of shRNA expression constructs with greater efficiency as indicated by the fold reduction value is plotted against the fold reduction. (C) Location and efficiency of individual shRNA expression constructs. The vertical axis represents the relative decrease and the horizontal axis represents the position of the GFP sequence. Each short horizontal bar represents the position and its RNAi efficiency. The orientation of the shRNA expression construct is also indicated by the color of the bar, and the gray and black bars indicate the shRNA expression construct where the guide sequence is 5 ′ and 3 ′ to the inverted repeat, respectively. (D) Plot the average relative reduction for each position of the GFP sequence (mean ± SD). The numbers in parentheses indicate the number of data for each group. FIG. 6 shows RNAi efficiency profiles by various transduction operations. (A) Compare the efficiency of GFP silencing by transfection of plasmid (left vertical axis, black circle) or PCR amplified expression cassette (right vertical axis, gray circle) with the efficiency of retroviral transduction (horizontal axis) ). (B) The efficiency of GFP silencing by transfection of in vitro transcribed shRNA with and without Dicer digestion (grey circles) and with black circles is plotted against the efficiency of plasmid transfection (horizontal axis) . IP ₃ is a diagram showing the silencing of type 1 IP ₃ R expression by RNAi library derived from R cDNA. (A) RNAi efficiency estimated based on GFP fluorescence intensity. The relative decrease value of GFP fluorescence intensity is shown. The square box represents the IP ₃ R coding region and the dots indicate the target location of each shRNA expression construct. Data for shRNA expression constructs, SI2A5 and SI3G6. (B) no shRNA expression construct (control), GFP-silenced with shRNA expression vector (shGFP), or a shRNA expression constructs an IP ₃ R targeting (SI2A5 and SI3G6), it was infected with retrovirus Western blotting of A7r5 cells. Actin has a role as a loading control. The edited normalized IP ₃ R expression level is shown in the lower panel (mean ± SEM, n = 3). (C) shows the intracellular calcium response of A7r5 cells induced by vasopressin when not transducing the IP ₃ R targeting shRNA expression construct and when introducing SI2A5 or SI3G6. Bars indicate AVP application. FIG. 2 shows a thymidine kinase based system for efficient RNAi construct selection. (A) Scheme of selection strategy. The upper figure shows a selectable marker gene construct. TK-puro represents mRNA encoding a fusion protein of thymidine kinase and puromycin N-acetyltransferase. GCV is converted to a toxic derivative of GCV (GCV-ppp), leading to cell death. (B) RNAi effect with a mixture of reconstituted shRNA expression constructs in the RNAi library after GSV selection (left panel) and without selection (right panel). The continuous curve in the right panel represents data from the parent shRNA library. The dashed curve represents the GFP fluorescence distribution without transduction of the shRNA expression construct. (C) Analysis of individual RNAi library clones with or without GCV selection. Individual clones were classified into four according to their RNAi efficiency; weak (fold reduction value <1.5), moderate (1.5-2.5), strong (2.5-4.5) and very strong (> 4.5). The normalized population represents the number of clones in each classification normalized to the total number of clones (n = 121 and 101 for clones with and without GCV selection, respectively).

siRNA Expression Constructs The present invention provides a method for producing an RNAi library from a desired target DNA. The RNAi library produced in the present invention is a library that can be applied to genetics of mammals, plants, insects, yeasts and the like. Therefore, in this document, “iRNA” is a short double-stranded RNA that is applied to mammalian cells (generally called “siRNA”, and the term siRNA is used synonymously in this document), a short hairpin-type double-stranded RNA Compared to RNA (commonly referred to as “shRNA” in this document, the term is used interchangeably in this document) and siRNA that is applied to gene suppression in nematodes, insects, plants, yeast, etc. Used to include.
In the RNAi library production method of the present invention, firstly, a step of randomly cleaving a target DNA into fragments is included. Here, the target DNA may be a specific gene, a plurality of genes, a gene group contained in an individual, a genome, or a cDNA library. Here, the “gene” may be a genomic sequence including an intron, or may be cDNA or the like. The species from which these genes and libraries are derived is not limited to mammals such as humans, plants, insects, and bacteria as described above.

  In this first step, a desired target DNA is prepared and randomly fragmented. The fragmentation length may be at least the length encoding RNA capable of inducing RNAi, for example, 10 or more bp. The appropriate length depends on the cell type that you want to induce silencing. For example, in mammalian cells, it can be 19 to 400 bp, preferably 19 to 200 bp, more preferably 19 to 50 bp. On the other hand, in plants, fungi and the like, gene silencing can be induced even by a long iRNA of about 500 bp, for example. Therefore, the enzyme that can be used in this step is not particularly limited as long as it can generate a fragment having a length or longer depending on the cell type for which silencing is to be induced. . Examples of enzymes that can be used include, but are not limited to, DNaseI and restriction enzymes that can be used for shotgun cloning, such as CvilJI, HaeIII, RsaI, AluI, and HpaI.

  In the second step, a hairpin type DNA is generated from the above DNA fragment using an adapter (hereinafter referred to as “hairpin type adapter”) made of a hairpin-shaped oligonucleotide. Here, the “hairpin adapter” functions as a linker that connects at least one end of the double-stranded DNA fragment to the hairpin shape, and finally the sense RNA strand of the iRNA expressed from the iRNA expression construct obtained by this method. It encodes an RNA linker that connects the antisense RNA strand to the hairpin shape. The hairpin adapter may be of an effective length for inducing RNA interference, and may be, for example, 5 to 50 base, preferably 6 to 20 base. However, a hairpin adapter longer than the length shown here can also be used. By trimming and adjusting the length in the process of constructing the iRNA expression construct described later, the length can be adjusted to the appropriate length when the iRNA expression construct is generated. It can also be designed such that a long hairpin RNA portion can be trimmed in the cell to generate an siRNA having an appropriate length by coding tRNA in the hairpin portion or combining with a hammerhead ribozyme. The sequence of the hairpin adapter is not particularly limited, and may be an artificial sequence or a sequence derived from microRNA. The hairpin adapter is preferably composed of DNA, but may be composed of RNA.

  When a hairpin adapter is added to the random DNA fragment, the hairpin adapter generally binds to both ends of the DNA fragment unless the terminal structure of the DNA fragment is controlled. When an adapter binds to both ends of the DNA fragment, the operation of the primer extension described later is hindered. Therefore, in order to form a hairpin DNA in which a hairpin adapter is bound to only one end of the double-stranded DNA, the double-stranded portion of the DNA fragment that has been circularized with the hairpin adapter bound to both ends of the DNA fragment is cut. Thus, it is conceivable to produce two hairpin DNAs. In order to produce a hepin type DNA from such a circularized DNA fragment, for example, cleaving with a restriction enzyme that randomly cleaves a double-stranded region derived from the DNA fragment can be mentioned. However, in order to control the efficiency of hairpin DNA generation and the length of the hairpin DNA, the hairpin adapter has an outside cutter restriction enzyme recognition site inside, and the double-stranded DNA region is cleaved by this restriction enzyme. It is preferable to use the technique to do. Regarding the control of the length of the double-stranded DNA region of hairpin DNA, it is applied to mammals because double-stranded RNA of about 19 to 21 bp can induce effective RNA interference in mammals. When preparing the RNAi library, the restriction enzyme of the outside cutter is preferably an enzyme that cleaves at least 19 bp away from the recognition site. An example of such an enzyme is MmeI. For example, MmeI cleaves a double strand at a site 20 or 21 bases away from the recognition site. Therefore, by providing an MmeI recognition site at the end of the hairpin adapter, the DNA fragment to which the hairpin adapter is connected is digested with MmeI and the hairpin adapter is connected to one end of a 20 to 21 base DNA fragment. Generated hairpin DNA. Therefore, MmeI can be said to be a preferred restriction enzyme that generates a hairpin type DNA encoding an iRNA having an effective length for silencing a target gene in a mammalian cell. In addition to MmeI, an enzyme that cleaves at a position 20 to 21 bases away from the recognition site as in MmeI is effective in this step. In the case of a restriction enzyme that cleaves 20 to 21 bases or more, the length of the DNA fragment encoding iRNA is desired by adjusting the length of the adapter and the location of the restriction enzyme recognition site in the adapter. Can be adjusted.

  In the above, it was explained that trimming is performed in any of the steps of generating an iRNA expression construct when a long hairpin adapter is used. This trimming technique is based on whether the adapter is DNA-based or RNA-based. It depends on what. When the adapter is DNA-based, the length of the adapter is adjusted by providing a restriction enzyme recognition site or cleavage site for trimming in the adapter, digesting with this restriction enzyme, and then reconnecting by ligation. be able to. The restriction enzyme for trimming is not particularly limited as long as it is a restriction enzyme that can shorten the length of the adapter. As shown in the examples described later, BcgI that cuts in both directions from the recognition site can be mentioned as an example of a restriction enzyme. By providing a recognition site for BcgI in the adapter, the adapter can be cleaved at two locations with one restriction enzyme BcgI, so that trimming can be performed easily. It is of course possible to adjust the length of the adapter by combining two restriction enzymes. When an RNA-based adapter is used, an RNA / DNA hybrid duplex is formed in the adapter region by the primer extension described below. Such trimming of the RNA / DNA hybrid region can be performed by first digesting the RNA strand with RNase H and then digesting ssDNA with an enzyme that digests single-stranded DNA (ssDNA enzyme).

  In the third step, an iRNA expression construct is generated using a hairpin DNA formed by adding a hairpin adapter to the DNA fragment. By applying a primer extension to the hairpin DNA duplex using a polymerase having Strand-displacing activity, an iRNA expression construct in which a complementary sequence is head-to-tail bonded with an adapter interposed therebetween is generated. Here, the hairpin DNA generated in the above step may be used as it is, and an iRNA expression construct may be generated by the primer extension. Preferably, however, a hairpin DNA adapter is connected before the primer extension is performed. It is preferable to protect the other end with another adapter. Here, as an adapter to be connected for terminal protection, a DNA adapter having both ends having a fragmented structure that does not hinder primer extension operation later (hereinafter referred to as “double-strand-end adapter” or “disruption” for convenience). "End adapter"). The sequence of this adapter is not particularly limited. The length is, for example, 5 to 100 base, preferably 20 to 40 base in consideration of the cost for synthesizing the adapter. However, since DNA ends can be protected even if the length is longer than this, the upper limit of the length is basically not limited as long as it does not interfere with the experimental operation.

  Further, since the stump type adapter is added for the purpose of protecting the DNA fragment, it is preferably removed after the primer extension is completed. In order to facilitate removal of the stump-type adapter after the primer extension, which will be described in detail later, it is preferable that the stump-type adapter is provided with a restriction enzyme site or a cleavage site for excision. There are no particular restrictions on the restriction enzyme that can be used to cut out the truncated adapter, but to prevent the truncated adapter sequence from remaining on the iRNA expression construct obtained by primer extension, use the restriction enzyme from the outside cutter. Is preferred. For example, as shown in an example described later, BpmI cuts a position at a certain distance from the recognition site. Therefore, by designing the BpmI recognition site in the truncated adapter so that this cleavage site is the boundary between the DNA fragment and the truncated adapter, the truncated adapter is almost removed from the iRNA expression construct. Further, the adapter sequence is completely removed by further digesting the sticky ends with an enzyme having ssDNA digestion activity. In addition, this truncated adapter sequence is added to both ends of the iRNA expression construct, but by cutting the both ends with different restriction enzymes, it becomes possible to control the direction of connection to the vector described later. As shown in the examples described below, one of the stump-type adapters can be excised with BpmI and the other with BbsI to remove both-end stump-type adapters and control the structure of both ends of the iRNA expression construct. The stump adapter can be designed as follows.

  The hairpin DNA to which the above-mentioned truncated adapter is connected is subjected to primer extension using a primer that can be annealed to the end of the adapter and a polymerase having Strand-displacing activity. The polymerase used here is required to have at least a strand-displacing activity, but is preferably heat-resistant in order to perform this operation using a PCR apparatus. Examples of the polymerase having strand-displacing activity include Klenow Fragment and phi29, and those having further heat resistance include Bst polymerase and Vent polymerase. The extension conditions can be determined as appropriate depending on the polymerase used. For example, the conditions when Bst polymerase is used are shown in the examples described later. When the hairpin adapter is RNA-based, it is necessary to use an enzyme having reverse transcription activity in addition to the above-mentioned strand-displacing activity.

  By carrying out the primer extension, a double-stranded DNA having a structure in which a DNA fragment derived from a target DNA has a head-tail bond with a hairpin adapter sandwiched between both ends is provided with a truncated adapter sequence. It is formed. This structure removes unwanted terminal truncated adapter sequences to generate an iRNA expression construct. Removal of an unnecessary terminal truncated adapter sequence is performed by cutting with a restriction enzyme that recognizes a site provided on the truncated adapter. When a stump adapter is trimmed by a restriction enzyme, an enormous iRNA expression construct is generated.

In addition, the process until the iRNA expression construct is generated, every time an adapter is added or digested with a restriction enzyme, or at the final stage of purification of the iRNA expression construct, the target fragment is purified to remove excess adapters or The fragment to be removed with a restriction enzyme is preferably removed from the reaction system. As an example of purification, a method of extracting a fragment of a target length or an expected length from a gel after electrophoresis, or a method of purifying a target fragment using a tag or the like can be mentioned.
In the above embodiment, after the target DNA has been fragmented, an example in which a hairpin adapter is added first and then a stump adapter is added is shown, but the reverse operation is also possible.

  The iRNA expression construct generated above is connected to a vector to construct an RNAi library. The vector can be selected depending on the cell type to which this library is to be applied, but an expression vector having a plasmid backbone that can be amplified by bacteria such as Escherichia coli can be preferably used. Examples of plasmid backbones that can be amplified in Escherichia coli include M13 vectors, pUC vectors, pBR322, pBluescript, pCR-Script, and the like. Since such bacterially derived plasmids can be amplified in large amounts, it is easy to prepare a large amount of a library, and it is convenient when performing treatments such as trimming of hairpin adapter sequences. The bacterial plasmid preferably carries a drug selection marker such as Amp, an auxotrophic gene or the like as necessary.

  In addition, such a plasmid backbone is preferably provided with an expression cassette corresponding to the host to be experimented. The promoter in the expression cassette may be any suitable for host cells, but in the case of mammalian cells, it is preferable to use an RNA polymerase III-driven promoter in order to express short RNAs such as siRNA / shRNA. Examples of the RNA polymerase III-driven promoter include a mouse U6 gene-derived promoter, a tRNA promoter, an adenovirus VAl promoter, a 5S rRNA promoter, a 7SK RNA promoter, a 7SL RNA promoter, an H1 RNA promoter, and the like. In order to incorporate the iRNA expression construct into the chromosome and stably express the iRNA, it is preferable to use a retrovirus expression cassette and incorporate the iRNA expression construct into this cassette. As an example using a retrovirus expression cassette, it shows in the Example mentioned later.

  In order to express iRNA and perform forward or reverse genetics in animal cells, plant cells, etc., a vector corresponding to the host cell may be selected. For example, mammalian-derived expression vectors such as pcDNA3 (manufactured by Invitrogen), pEGF-BOS (Nucleic Acids. Res. 1990, 18 (17), p5322), pEF, pCDM8 and other insect cell-derived expression vectors For example, “Bac-to-BAC baculovairus expression system” (manufactured by Gibco BRL), pBacPAK8, etc., plant-derived expression vectors, eg, pMH1, pMH2, etc. Animal virus-derived expression vectors, eg, pHSV, pMV, pAdexLcw, etc. , Expression vectors derived from yeast such as pZIPneo, such as “Pichia Expression Kit” (manufactured by Invitrogen), pNV11, SP-Q01, etc., such as pPL608 , PKTH50 and the like. The vast number of iRNA expression constructs are inserted into individual vectors to generate RNAi libraries.

  The RNAi library generated here can be directly used for forward and reverse genetics. In addition, an oligo RNAi library may be synthesized from the RNAi library of the present invention using an in vitro transcription system, and this oligo RNAi library may be generated and used for forward and reverse genetics research. .

Screening method The present invention provides a screening method for selecting a clone carrying an iRNA expression construct capable of suppressing the expression of a target gene from the RNAi library. Specifically, the screening method of the present invention includes a step of introducing an RNAi library prepared from the target DNA into the cell in which the target DNA is expressed, and a step of measuring the expression of the target DNA. It is.

  The method for introducing the RNAi library into the cell in which the target DNA is expressed can be appropriately selected depending on the vector used when the RNAi library is constructed. For example, when a virus vector having an infectivity is used, the RNAi library is introduced into the cells by the infectivity of the virus. In the case of a plasmid vector, a method using a cationic ribosome DOTAP (manufactured by Boehringer Mannheim), an electroporation method, a lipofection method, a gene gun method, a calcium phosphate method, a DEAE dextran method, etc. may be used as appropriate. it can.

  Target DNA expression may be measured based on the expression level of the protein using the target DNA antibody. If the activity of a specific gene has been identified, the target DNA expression may be measured. May be. For example, when it is known that the downstream gene is regulated as the activity of the target DNA, the expression of the target DNA may be indirectly measured by measuring the expression of the downstream gene. .

  If the function of the target DNA is unknown or it is difficult to measure the activity, a transformant holding the fusion gene with the reporter gene connected to the target DNA is prepared and screened using this. It is preferable.

  For example, an enzyme such as a fluorescent protein (luciferase, GFP, CFP, YFP, RFP, etc.), aminoglycoside transferase (APH), thymidine kinase (TK), dihydrofolate reductase (dhfr) can be used as the reporter. By connecting the target DNA to a gene encoding such a reporter, mRNA of the fusion gene is expressed in the cell. The iRNA against the target DNA expressed from the RNAi library suppresses the expression of this fusion gene, and the reporter activity decreases. Therefore, by using a fusion gene in which such a target DNA is fused with a reporter gene, it becomes possible to easily measure the silencing activity of each clone of the RNAi library against the target DNA using the reporter activity as an index. For example, when used for a reporter gene such as the above fluorescent protein, the silencing activity of each clone in the RNAi library can be measured based on the decrease in the expression of the fluorescent protein. On the other hand, when a negative selection marker such as TK is used, clones that do not have silencing activity in the RNAi library cannot suppress TK activity, but cells are killed by the addition of ganciclovir, whereas silencing In clones having activity, the activity of TK is suppressed, and cells can survive even by the addition of ganciclovir. In this way, when a negative selection marker is used as a reporter, only cells into which a clone having silencing activity has been introduced selectively survive, so that a clone carrying an iRNA expression construct having silencing activity can be obtained efficiently. It is done. By recovering the vector retaining the iRNA expression construct from the cells selected here, it becomes possible to identify whether it is preferable to target any region in the target DNA for the iRNA.

Kit The present invention provides an RNAi library kit for carrying out the above-described enzymatic construction method of RNAi library. The kit may contain the above-described hairpin adapter, stump adapter, primer extension primer and enzyme, an enzyme used for trimming, if necessary, an enzyme for purifying a random DNA fragment, and the like. A vector for inserting an iRNA expression construct may also be included. Further, a protocol for carrying out the above-described enzymatic construction method of RNAi library may be attached. Thus, the kit for the material for carrying out the library construction method of the present invention makes it possible to carry out the RNAi library construction method of the present invention in a more simple and familiar manner.

result
Preparation of RNAi library
EPRIL includes several steps of enzyme treatment to create an shRNA expression vector library from the cDNAs of interest (Figure 1a). First, double-stranded DNAs are fragmented almost randomly with DNaseI (Anderson, S. Nucleic Acids Res. 9, 3015-3027 (1981)). Next, a hairpin-shaped adapter containing a recognition sequence for MmeI is ligated to the fragment. MmeI has been reported to cleave the upper and lower strands 20 and 18 bases away from the recognition sequence, respectively (Boyd, AC et al. Nucleic Acids Res. 14, 5255-5274 (1986) ), We found that MmeI also cleaves DNA at bases 21 and 19. Thus, MmeI digestion yields short 3′-overhanging DNA fragments with sequences of 20 or 21 bases in length from the target cDNAs. Polyacrylamide gel electrophoresis (PAGE) shows MmeI digested DNA as a -40 bp band (Figure 1b). After digestion with MmeI, a second adapter is ligated to the generated fragment. This adapter has two degenerate bases at the 3 ′ end of one strand so that the 3 ′ overhang of the MmeI digest is blocked. After ligation of the second adapter, a primer extension reaction is performed to convert the single stranded hairpin DNA into double stranded DNA with inverted repeats joined via a loop sequence (FIG. 1). The primer extension product is digested with an appropriate restriction endonuclease to remove extra sequences outside the inverted repeat and inserted into the plasmid vector described below. Excess sequences in long loops flanked by inverted repeats are removed by appropriate restriction endonucleases followed by recircularization.
For library construction and shRNA expression, we used a plasmid with a retroviral vector containing an RNA polymerase III driven promoter from the mouse U6 gene. By using a retroviral vector to introduce the shRNA expression cassette, a stable RNAi effect is ensured (Paddison, PJ & Hannon, GJ Cancer Cell 2, 17-23 (2002)).

RNAi library from cDNA encoding GFP
To evaluate EPRIL, we generated a shRNA expression library from cDNA encoding green fluorescent protein (GFP). Sequence analysis of individual clones from the library showed that 290 out of 343 clones with sequence data had inverted repeats. 39 of these clones had incomplete inverted repeats, whereas 251 clones were for GFP silencing with an inverted repeat of 19 or more bases in length derived from the GFP sequence Of candidate shRNA expression constructs. Most clones have an inverted repeat length of either 20 or 21 nucleotides, 18 of 251 clones (7.2%), 139 (55.4%), and 94 (37.5%) It was 19, 20, and 21 bases long. Since the first adapter can be attached to both ends of the DNaseI digested fragment, it should be possible to obtain shRNAs with two different orientations. Guide sequences that are complementary to the mRNA sequence are present at approximately the same frequency 5 'or 3' to the shRNA (54.6% vs 45.4%, p> 0.1). Analysis of the target sequence showed that different shRNA expression constructs were generated from different parts of the target gene; non-overlapping from 251 independent clones across the entire 720 bp GFP coding sequence 157 shRNA expression constructs covered 96.3% of the total. Thus, EPRIL enables high-throughput production of a vast array of shRNA expression constructs from the cDNA of interest.

Retroviruses with shRNA expression constructs are produced from individual plasmids in a 96 well plate format, which allows us to simultaneously obtain a vast array of independent viruses. Jurkat T cells expressing GFP were also infected with the virus in the same format. The level of GFP expression in infected cells was determined by flow cytometry to quantify RNAi efficiency. There was considerable variation in RNAi efficiency depending on the shRNA expression construct. Thus, the present inventors analyzed the distribution of RNAi efficiency from the measurement results of 262 non-overlapping constructs (FIG. 2a). About 56% of the constructs had low RNAi activity (less than 1.5 fold reduction). In order to estimate the probability of finding an shRNA expression construct having an RNAi efficiency greater than a predetermined reduction fold (x), we determined the amount per fraction of shRNA expression constructs whose reduction fold exceeds x. Plotted against. Roughly 30% showed more than a 2-fold decrease in RNAi efficiency, whereas only a small percentage had a decrease greater than 8-fold. The overall probability is proportional to the fold reduction with ~ 1.7 power law scaling, which means that if you want to find an RNAi construct with a 5 fold reduction efficiency, for example, screen 15 (5 ^1.7 ) times more candidate constructs It means you have to.

  Next, the inventors analyzed the region dependence of RNAi efficiency. Both efficient and inefficient shRNA expression constructs were distributed throughout the region (Figure 2c). For quantitative analysis, we classified the data according to 10 equally subdivided regions (FIG. 2d) and performed analysis of variance. We did not find significant region dependence (Kruskal-Wallis test; p <0.4). In particular, the results show that despite the previous assumption that the region is a poor substrate for RNAi (Dykxhoorn, DM et al. Nat. Rev. Mol. Cell Biol. 4, 457-467 (2003)) It has been shown that even a region within 72 bp from the codon can serve as a good target. With respect to the orientation of the shRNA expression construct, an overall trend was observed that the guide sequences present on the 3 ′ side were more effective.

The variation pattern of RNAi efficiency is cell type independent, since the 10 representative constructs examined showed similar efficiency profiles in HEK293 or HeLa cells (data not shown). Direct transfection of the plasmid or PCR amplified minimal shRNA expression cassette yielded a profile similar to retroviral transduction (Figure 3a), eliminating the effects of differences in virus titer. Furthermore, RNAi profiles from direct transfection of in vitro transcribed shRNA correlated well with DNA-based expression profiles (FIG. 3b). Similar results were obtained when shRNAs transcribed in vitro were pre-digested with Dicer. These results indicate that the factors that determine the RNAi efficiency profile are downstream of the transcription and dicer-processing stages.

RNAi library from cDNA encoding type 1 IP ₃ receptor We investigated whether the above strategy could be applied to endogenous genes. The present inventors used type 1 inositol 1,4,5-trisphosphate receptor (Mignery, GA et al. J. Biol. Chem. 265, 12679-12685 (1990)) (IP ₃ R) as a target. The encoding DNA was used. As with GFP, we obtained various shRNA expression constructs for IP ₃ R. Of the 256 clones with sequence data, 214 were found to have shRNA expression constructs, including 199 constructs with different target positions and not overlapping each other. In order to rapidly screen for effective shRNA expression constructs, we generated Jurkat T cells carrying reporter constructs expressing GFP and IP ₃ R head-tail bound mRNAs. Target mRNA degradation by shRNA expression constructs can be assessed by monitoring the decrease in GFP fluorescence (Kumar, R et al. Genome Res. 13, 2333-2340 (2003)). ShRNA expression constructs targeting IP ₃ R also altered RNAi efficiency (FIG. 4a). We selected two shRNA expression constructs, SI2A5 and SI3G6 among the most effective constructs, and these clones were effective for IP ₃ R expressed endogenously in the vascular smooth muscle cell line A7r5. It was investigated whether RNAi could be induced (De Smedt, H. et al. J. Biol. Chem. 269, 21691-21698 (1994)). Western blot analysis confirmed that the IP ₃ R expression level was reduced (FIG. 4b). Loss of function was also confirmed by measuring the vasopressin-induced intracellular Ca ²⁺ response (FIG. 4c). These results indicate that EPRIL is useful for searching effective shRNA expression constructs that target endogenous genes.

Intracellular selection of effective shRNA expression constructs To facilitate high-throughput screening, we have developed a novel intracellular selection scheme based on special selectable marker genes for positive selection of efficient shRNA expression constructs. (Fig. 5a). The marker gene consists of two head-tail junctions; the first encodes a fusion protein consisting of thymidine kinase and puromycin N-acetyltransferase (Chen, YT & Bradley, A. Genesis 28, 31 -35 (2000)), the latter encodes the target mRNA. In general, when ganciclovir (GCV) is applied, cells with this marker gene by intracellular accumulation of toxic derivatives of GCV by the action of thymidine kinase (Chen, YT & Bradley, A. Genesis 28, 31-35 (2000)) Is killed. When cells are transduced with shRNAs that have effective RNAi activity with respect to the target gene, the cells will escape cell death by silencing thymidine kinase expression. The present inventors constructed such a marker gene using GFP as a target and introduced it into Jurkat T cells for puromycin selection. Next, the inventors made shRNA-expressing retroviruses as a mixture from shRNAi libraries targeting GFP, and infected marker gene expressing cells with these retroviruses. The infected cells were treated with GCV for 48 hours, and GCV resistant cells were cultured. We recovered shRNA expression constructs by PCR amplification from viable cells and reconstructed them into retroviral expression vectors to obtain selected libraries. To examine whether GCV selection actually enriched for effective shRNA expression constructs, Jurkat T cells expressing GFP were infected with the virus mixture prepared from the reconstructed library. Flow cytometry results show that GFP expression in cells transduced with retroviruses from the GCV-selected library was significantly attenuated than that in cells transfected with retroviruses from the intact or parental library. Indicated. This indicates that GCV selection enriched for more effective shRNA expression constructs (FIG. 5b). Successful enrichment of valid shRNA expression constructs by GCV selection was also confirmed by analysis of individual clones. A significant increase in the population of shRNA expression constructs with high efficiency was observed in the GCV selection library (FIG. 5c). These results indicate that a novel intracellular selection scheme is useful to obtain an effective shRNA expression construct from the library.

RNAi library derived from cDNA library
EPRIL offers the opportunity to construct shRNAi libraries from complex mixtures of cDNAs such as cDNA libraries rather than a single cDNA source. Such shRNAi libraries should be extremely valuable for comprehensive searches of genes involved in specific cell functions. For this reason, the present inventors investigated whether or not the present technology can be realized for such a purpose. EPRIL was performed on a cDNA library prepared from mRNA expressed in mouse bone marrow progenitor cells FL5.12 cells to create a shRNAi library. When randomly selected clones were sequenced, it was found that 215 of the 240 sequences obtained contained inverted repeats (Table 1). Among these, inverted repeats from 35 clones included sequences consisting of more than 10 bases of polyA or T, possibly derived from the polyA tail of mRNAs. Therefore, a BLAST search was performed on the remaining 180 clones. The sequence of 165 clones matched the cDNA sequence and / or ESTs (Table 2). We classified these clones derived from gene transcripts according to unigene clusters (Build 126) and found that 146 out of 165 clones belong to at least one of the unigene clusters. Among these, several clones suggested that these clones were derived from the same gene, corresponding to the same cluster. These include genes encoding heat shock protein 8, ubiquitin B and ribosomal protein L41, all of which are highly expressed in many organs. Since the size of the unigene cluster is an approximate estimate of the relative expression level, the size of the cluster was compared with the number of appearances of the same gene. The size of clusters with more than one repeat sequence was significantly larger than clusters without repeat sequences (p <0.003, Mann-Whitney U-test). These features suggest that the RNAi library represents the original expression profile. We further analyzed the target location of the transcript. 53% corresponded to the coding region, 44% corresponded to the 3′-untranslated region, and 3% corresponded to the 5 ′ untranslated region. Thus, shRNA expression constructs were obtained from various parts of the transcript. The results show that EPRIL can be applied even to cDNA libraries consisting of complex mixtures of various genes.

Cell culture
Jurkat T cells were cultured in RPMI 1640 medium (Invitrogen) containing 10% fetal calf serum (FCS), penicillin and streptomycin. GP293, HEK293, A7r5, and HeLa cells were maintained in Dulbecco's modified Eagle medium (Sigma or Invitrogen) containing 10% FCS. FL5.12 cells (donated by Dr. Inaba, Hiroshima University) were maintained in RPMI 1640 (Sigma) containing 10% FCS and 1 ng / ml IL-3 (Wako, Japan).

Plasmid construction
All plasmids were constructed using standard molecular biology techniques, including plasmid pNAMA-U6 with shRNA expressing retroviral vectors. Briefly, the U6 promoter and termination signal-encoding DNA obtained by PCR amplification from pSilencer1.0-U6 (Ambion) were transferred from the pMX retroviral vector (donated by Dr. Kitamura, University of Tokyo). It was inserted into the NheI site of the plasmid with LTR. For further PCR amplification into the plasmid using the primer pair, 5'-CGCGGATCCGAATGCTTTTTTTAATTCCTGCAGCCCG-3 '(SEQ ID NO: 1) and 5'-CGCGGATCCGAAGACCCCAAACAAGGCTTTTCTCCAAG-3' (SEQ ID NO: 2) to insert the shRNA-encoded DNA fragment A BbsI / BsmI site was formed, resulting in pBsk-U63-3LTR. To construct pNAMA-U6, a HindIII / SalI fragment with a 3 ′ LTR containing a U6 promoter at the NheI site was excised from pBsk-U63-3LTR and the DsRed2 gene derived from pDsRed2-N1 (Clontech) ) And the 3 ′ LTR-deficient pMX vector was inserted into the HindIII / SalI site of pda5LTR-DeRed2-M4.

  Plasmid pMS240-PNS encoding a retrovirus that expresses a thymidine kinase-puromycin-N-acetyltransferase fusion protein (Chen, YT & Bradley, A. Genesis 28, 31-35 (2000)) under the control of the SV40 promoter. Was constructed from PCR amplified fragments encoding thymidine kinase and puromycin acetyltransferase. A DNA fragment containing thymidine kinase and puromycin N-acetyltransferase gene was subcloned into the self-inactivating retroviral vector pMS240 to produce pMS240-PNS. pMS240-PNS has a BamHI / NotI site for cloning the DNA fragment encoding the target gene. To select an effective shRNA expression construct targeting GFP, a GFP coding fragment from pd2EGFP-1 (Clontech) was inserted at that site.

To construct pMX-d2EGFP-IP ₃ R, a plasmid encoding a retroviral vector for monitoring IP ₃ R silencing, DNA encoding a destabilized GFP variant was prepared from pd2EGFP-1 This was inserted into pMX. DNA encoding IP ₃ R from pBS-IP ₃ R was inserted into a site downstream of D3EGFP.

EPRIL
To create a shRNA expression construct targeting GFP, a BamHI / NotI DNA fragment encoding GFP was obtained from pEGFP-1 (Clontech). For IP ₃ R, the fragment encoding the entire coding region of the rat type 1 IP ₃ R was prepared by EcoRI / NotI digestion of pBS-IP ₃ R1. To create a shRNA expression construct derived from a cDNA library, a Super SMART PCR cDNA synthesis kit (Clontech) was used to create a double-stranded cDNA library from mRNA prepared from FL5.12 cells and subcloned. Used directly for further induction without. Next, EPRIL was applied using these DNA fragments according to the following six-step protocol:

Stage 1. Preparation of random DNA fragments
The DNA fragment was limited digested with DNase I in the presence of 1 mM MgCl ₂ , 0.1 mg / ml BSA, and 50 mM Tris-HCl (pH 7.5) to give 0.1 U / μl T4 DNA polymerase (Takara) , Japan). The concentration of DNase I (Takara, Japan) was determined empirically for each preparation, resulting in digests having an average size of about 100-200 bp on PAGE.

Stage 2. Ligation and MmeI cleavage of the first adapter The digest is digested with T4 DNA ligase (DNA ligation kit version 2, Takara, Japan), adapter 1, 5'-GTCGGACAATTGCGACCCGCATGCTGCGGGTCGCAATTGTCCGAC-3 '(SEQ ID NO: Ligated to 3). Hairpin-shaped adapter 1 was purified by native PAGE from oligonucleotides chemically synthesized prior to use. The nick between the 5 ′ end of adapter 1 and the 3 ′ end of digested DNA is 0.1 mM NAD, 1.2 mM EDTA, 10 mM (NH ₄ ) ₂ after treatment with 1.0 or 1.25 U / μl T4 polynucleotide kinase. Repaired by 0.006 U / μl E. coli ligase (Takara, Japan) in the presence of SO ₄ , 4 mM MgCl ₂ , and 30 mM Tris-HCl (pH 8.0). A short DNA fragment to which the adapter was bound was generated by cleavage with MmeI. The resulting DNA fragment was detected as a single band migrating at ~ 40 bp on native PAGE. The gel band was excised and the MmeI-cleavage product was obtained by phenol-chloroform extraction and ethanol precipitation.

Stage 3. Ligation of the second adapter Adapter 2 with both ends truncated by annealing oligonucleotide pair, 5'-CAAAGAGTCTTCGGCCCTCCAGACCGTGAGTC-3 '(SEQ ID NO: 4) and 5'-GCCGAAGACTCTTTGNN-3' (SEQ ID NO: 5) And then purified using PAGE. The latter oligonucleotide contains two degenerate bases at the 3 ′ end called “N”, which represents A, C, G, or T. Adapter 2 was ligated to the MmeI digested DNA fragment from step 2 using T4 DNA ligase. After purification by PAGE, nicks on the adapter 2-ligation fragment were repaired by treatment with T4 polynucleotide kinase (Takara, Japan) and T4 DNA ligase.

Stage 4. Primer extension Next, the product from step 3 was subjected to a primer extension reaction, along with a primer oligonucleotide, 5'-GACTCACGGTCTGGAGGGCCGAA-3 '(SEQ ID NO: 6), 0.1% Triton X-100, 0.2 mM dNTPs, 10 After incubating for 135 seconds at 94 ° C. in a reaction buffer containing mM KCl, 10 mM (NH ₄ ) ₂ SO ₄ , 2 mM MgSO ₄ , and 20 mM Tris-HCl (pH 8.8), the mixture was cooled to 62 ° C. Next, 0.02 U / μl Bst DNA polymerase large fragment (NEB) was added to the reaction mixture to initiate the primer extension reaction. The reaction was carried out at 65 ° C. for 300 seconds and stopped by cooling to 4 ° C. The reaction product was purified by PAGE.

Step 5. Subcloning into plasmid The product from step 4 was digested with BpmI, blunted with Klenow fragment (Takara, Japan) and then digested with BbsI. pNAMA-U6 was digested with BsmI, similarly blunt ended, digested with BbsI, and then treated with bacterial alkaline phosphatase (Takara, Japan). The digested fragment was gel purified and ligated at a molar ratio of about 3: 1. The ligation mixture was purified and then introduced into ElectroMAX DH5a-E competent cells (Invitrogen) by electroporation. Transformed cells were seeded on 500 cm ² LB agar plates containing 100 μg / ml carbenicillin. After overnight incubation, turf-grown bacteria were collected with a spatula and plasmid DNA was prepared using a plasmid purification kit (plasmid MIDI kit, Qiagen).

Step 6. Excess linker cleavage The plasmid purified from step 5 was digested with BcgI, blunted with T4 DNA polymerase and recircularized by self-ligation. This treatment removes most of the sequence from adapter 1, but retains the short linker sequence, 5'-TTGTCCGAC-3 '(SEQ ID NO: 7). In order to eliminate as much as possible contamination of the incompletely digested plasmid, the DNA was digested with MfeI whose recognition sequence was present in the BcgI excised portion of the sequence from adapter 1. ElectroMAX DH5a-E competent cells were transformed with the recircularized plasmid and selected on LB-agar plates containing 100 μg / ml carbenicillin. After overnight culture on the plate, a plasmid library stock was obtained.

Recovery of shRNA expression constructs from the genome
shRNA expression constructs were recovered by PCR amplification from 100 ng genomic DNA prepared from transduced FL5.12 cells. For PCR, Vent DNA polymerase (NEB) was used with a primer pair, 5'-GTACGAGCGCACCGAGGGCCGCCACC-3 '(SEQ ID NO: 8) and 5'-GGCGTTACTTAAGCTAGCGATCCGACGCCGCCATC-3' (SEQ ID NO: 9). PCR amplified DNA was digested with NotI and AflII and subcloned into pBsk-3LTR. After transformation and purification, DNA encoding the 3 ′ LTR containing the recovered shRNA expression construct was excised with HindIII and NotI and subcloned into pda5LTR-DsRed2-M4. The resulting plasmid is structurally identical to pNAMA-U6 and can be packaged to produce retroviruses with recovered shRNA expression constructs.

Preparation of shRNA in vitro
shRNAs were synthesized by using in vitro transcription based on the T7 promoter. To prepare a template for in vitro transcription, two chemically synthesized oligonucleotides, T7 promoter-encoding oligonucleotide, 5'-taatacgactcactataG-3 '(SEQ ID NO: 10) and shRNA-encoding oligonucleotide 5'-AAAAN _20-21 GTCGGACAAN _'20-21 ctatagtgagtcgtatta-3' (SEQ ID NO: 11, N _20-21 and N _'20-21 shows a sequence corresponding to the basic structure of the shRNA, lowercase, complementary to the T7- promoter encoding oligonucleotide Were incubated at 94 ° C. for 135 seconds, followed by annealing at 45 ° C. for 30 seconds. Next, the annealed oligonucleotide was converted into a double-stranded DNA template by extension reaction at 50 ° C. for 600 seconds using a large Bst DNA polymerase fragment (0.08 U / μl). shRNA was generated from the purified template using the CUGA7 in vitro transcription kit (Nippon Gene, Japan) according to the manufacturer's protocol. The transferred shRNA was purified using a gel filtration spin column (Microspin G-25, Amersham Bioeciences). When making Dicer digested shRNA, the shRNA was treated with recombinant human Dicer (Gene Therapy Systems) according to the manufacturer's protocol.

N _20-21 sequences of different shRNA-encoded oligonucleotides are
TCTCGGCATGGACGAGCTGT (shGFP1) (SEQ ID NO: 12)
CTTCACCTCGGCGCGGGTCT (shGFP5) (SEQ ID NO: 13)
CTACAAGACCCGCGCCGAGG (shGFP7) (SEQ ID NO: 14)
CCCCATCGGCGACGGCCCCGT (shGFP10) (SEQ ID NO: 15)
AGATCCGCCACAACATCGAGG (shGFP12) (SEQ ID NO: 16)
TCAGCTCGATGCGGTTCACC (shGFP13) (SEQ ID NO: 17)
CCACAAGTTCAGCGTGTCCGG (shGFP14) (SEQ ID NO: 18)
CTTCAAGTCCGCCATGCCCGA (shGFP17) (SEQ ID NO: 19)
TGCTGGTAGTGGTCGGCGAGC (shGFP18) (SEQ ID NO: 20)
TCGGCGCGGGTCTTGTAGTTG (shGFP19) (SEQ ID NO: 21) and
TTGTAGTTGCCGTCGTCCTT (shGFP20) (SEQ ID NO: 22).
The N ′ _20-21 sequence was complementary to the N _20-21 sequence.

Retrovirus production and transduction For high-throughput screening, the generation and transduction of retroviruses with shRNA expression constructs were performed in a 96-well plate format. Plasmids were prepared in 96 well plate format using the QIA well 96 ultra plasmid kit (Qiagen) according to the manufacturer's protocol. The GP293 packaging cell line (Clontech) was transfected with ~ 200 ng of retroviral vector plasmid and 17 ng of VSG-G encoding plasmid using Lipofectamine 2000 (Invitrogen) for each well in a 96 well plate. Two days after transfection, a culture medium containing retrovirus was obtained. Retroviral transduction was performed by adding 50 μl of medium containing retroviral particles to 50 μl of Jurkat T cell suspension (1.0 × 10 ⁵ cells / ml). For production of retrovirus at moderate scale, DNA was purified by plasmid MIDI kit and grown in 10 cm culture dishes by transfection with 24 μg of retroviral vector plasmid and 2 μg of VSG-G encoding plasmid. Retroviral packaging was performed at If necessary, retroviral particles were concentrated by centrifugation overnight and then resuspended in an appropriate culture medium.

Assessment of flow cytometry and RNAi efficiency Relative GFP expression levels were estimated based on fluorescence intensity analyzed using a FACscan flow cytometer (BD). The fold decrease in GFP fluorescence (control fluorescence intensity divided by the fluorescence intensity of the shRNA-transduced cells) was used as a measure of RNAi efficiency. A series of internal control shRNA expression constructs were included in each 96-well plate to correct for batch-to-batch differences in RNAi efficiency due to variations in retroviral titers.

Western blot analysis
A7r5 cells were infected with retrovirus carrying the shRNA expression construct. Four days later, the cells were collected by trypsinization, solubilized, and subjected to SDS-PAGE. After the SDS-PAGE product is transferred to a PVDF membrane, the protein corresponding to IP ₃ R is used as a primary antibody and a secondary antibody, respectively, as rabbit IgG anti-type 1 IP ₃ R (Alomone) and HRP-conjugated anti-rabbit IgG. Probed by (MBL, Japan) and detected by chemiluminescence. Quantification of immunoblot signal intensity was performed using IPLab spectral software (Scananalytics, Inc). The IP ₃ R signal was normalized by an actin immunoblot signal probed with an anti-actin antibody (Santa cruz).

Intracellular calcium measurement In order to measure intracellular calcium, A7r5 cells seeded on a cover glass were infected with shRNA-expressing retrovirus, and Ca ²⁺ indicator substance Fura-2 was loaded 4 days after the infection. Changes in intracellular Ca ²⁺ concentration, as previously described (Hirose, K. et al. Science 284, 1527-1530 (1999)), are inverted microscopes equipped with a CCD camera (Photometrics). And evaluated by fluorescence measurement by ratio measurement. Cells were stimulated by adding 1 nM arginine vasopressin (AVP) in the presence of nicardipine (10 μM), an L-type Ca ²⁺ channel inhibitor.

  As described above, the present inventors have established a technique “EPRIL” that can enzymatically derive an shRNA expression library from a cDNA template. When the present invention is applied to a single cDNA source, the library can provide a vast array of candidate shRNA expression constructs for various regions on the cDNA. Combining EPRIL with high-throughput screening and intracellular selection schemes provides a generalizable platform for creating a large collection of the best shRNA expression constructs for any gene. Such collections will generally contribute significantly to high throughput reverse genetics based on RNAi in mammals.

The inventors have attempted to identify efficient RNAi sequence selectivity. We observed a weak but general trend that shRNA expression constructs with a U at the 4th nucleotide from the 5 ′ end of the guide sequence are effective for GFP and IP ₃ R, but against other positions The selectivity was not always consistent. For example, G at the second nucleotide tended to be favorable for IP ₃ R, but no such trend was observed for GFP. Clearly, more data from various genes is needed to determine overall sequence selectivity. EPRIL will greatly facilitate such research.

EPRIL can generate large shRNAi libraries from a cDNA library consisting of a complex mixture of cDNAs. We estimated that the library contained 3 × 10 ⁵ to 4 × 10 ⁶ independent cDNA-derived shRNA expression constructs. Thus, the methods of the present invention allow genes to be identified by selection based on absolute phenotypic criteria such as cell morphology, adhesion, cell death, expression levels of key molecules and other functional indicators Forward genetics based on RNAi could be performed. shRNA sequences can perfectly match 20 and 21-base long tags with random sequences (Saha, S. et al. Nat. Biotechnol. 20, 508-512 (2002)) probability 9.1 × 10 ⁻¹³ and 2.3 × 10 ⁻¹³ can serve as reliable tags for gene identification because they are small enough considering the size of the mouse or human genome ˜3 × 10 ⁹ bp.

  Forward genetics using the RNAi library of the present invention may be applied at the animal level. Because lentivirus-mediated transduction can generate mice with shRNA expression constructs (Rubinson, DA et al. Nat. Genet. 33, 401-406 (2003)), we target different genes A huge array of mutant animals can be efficiently produced by the shRNA expression construct. When systematically screening based on phenotype, genes involved in a specific phenotype can be easily identified. This feature is in contrast to the features of a large mutagenesis project ongoing in mice using ethylnitrosourea (ENU) (Kile, B.T. et al. Nature 425, 81-86 (2003)). ENU-mutagenesis creates mutant animals efficiently, but requires a tedious and tedious process to determine the locus. Furthermore, variations in the RNAi efficiency of shRNA expression constructs may lead to the discovery of unique phenotypes due to variations in the degree of gene silencing. Thus, EPRIL will provide a new style of forward genetics in mammals and, together with its role in reverse genetics based on RNAi, will contribute significantly to elucidating the relationship between genes and functions at the genome level. .

Claims (16)

A method for producing an RNAi library from a desired target DNA comprising the following steps (1) to (3), wherein double-strand breaks are attached to the other end of the DNA fragment before or after the step (2): A step of connecting a type adapter, the restriction enzyme recognition site of the outside cutter is provided in either one of the double-strand type adapter or the hairpin type adapter described in the step (2), An adapter having a restriction enzyme recognition site is first connected to the DNA fragment, and the other adapter is connected to the end formed by adjusting the length of the DNA fragment by the restriction enzyme of the outside cutter. .
(1) a step of randomly cutting a target DNA to generate a DNA fragment;
(2) connecting a hairpin adapter to one end of the DNA fragment to generate a hairpin DNA fragment;
(3) A step of generating an iRNA expression construct encoding an interference RNA by performing primer extension on a hairpin type DNA fragment using a polymerase having Strand-displacing activity.   The method according to claim 1, wherein the hairpin adapter is provided with a restriction enzyme recognition site of an outside cutter.   The method according to claim 1, wherein the double-strand-end adapter is provided with a restriction enzyme recognition site of an outside cutter.   The method for producing an RNAi library according to claim 1, further comprising a step of connecting the iRNA expression construct to an expression vector after primer extension.   The method for producing an RNAi library according to claim 4, wherein the expression vector can function in a mammal.   The method for producing an RNAi library according to claim 4 or 5, wherein the expression vector has an integration activity into the host chromosome.   The method for producing an RNAi library according to any one of claims 1 to 6, wherein the target DNA is any one of a cDNA of a specific gene, a cDNA library, or a cDNA after subtraction.   The method for producing an RNAi library according to claim 1, wherein in the step (1), the target DNA is fragmented using a DNA digestion enzyme that can be used for shotgun cloning.   The method for producing an RNAi library according to any one of claims 1 to 8, wherein the restriction enzyme of the outside cutter cleaves at least a position separated by 19 bp or more from the recognition site in the first adapter.   The method for producing an RNAi library according to any one of claims 1 to 9, wherein the polymerase having strand-displacing activity is thermostable.   The method for producing an RNAi library according to any one of claims 1 to 10, wherein the polymerase having strand-displacing activity is either Bst polymerase or Vent polymerase. A method for screening an iRNA expression construct encoding an interference RNA from an RNAi library produced by the method according to any one of claims 1 to 11,
Introducing the RNAi library into cells expressing the target DNA;
Measuring the expression of the target DNA. The screening method according to claim 12 ,
As a fusion gene with target DNA fused to a reporter gene,
A screening method, wherein in the step of measuring the expression of the target DNA, the expression of the target DNA is measured using the activity of the reporter protein as an index. A method for screening an iRNA expression construct encoding an interference RNA from an RNAi library produced by the method according to any one of claims 1 to 11,
Introducing the RNAi library into a cell expressing a fusion gene obtained by fusing a target DNA and a negative selection marker gene;
Screening with a negative selection marker, and selecting only cells that had an RNAi effect. A kit for producing an RNAi library by the method according to any one of claims 1 to 11,
Including the hairpin adapter and the double-strand break adapter,
A kit, wherein any one of the adapters is provided with a restriction enzyme recognition site for an outside cutter. Furthermore, the kit of Claim 15 provided with the polymerase which has the restriction enzyme of an outside cutter and / or Strand-displacing property.