WO2005078089A1

WO2005078089A1 - Method for producing sirna molecules and sirna molecules produced therewith

Info

Publication number: WO2005078089A1
Application number: PCT/EP2005/001079
Authority: WO
Inventors: Matthias Truss; Christian Hagemeier
Original assignee: Charite Universitätsmedizin - Berlin
Priority date: 2004-02-11
Filing date: 2005-02-03
Publication date: 2005-08-25

Abstract

The present invention relates to the production of siRNA molecules and, in particular, the an improved siRNA molecule comprising a 5'-protruding end that corresponds to the consensus sequence of the nucleotides 0 to +5 of the natural promoter of a DNA-dependent RNA-polymerase. In preferred embodiments, the siRNA molecule according to the invention contains the natural promoter of T3, T7 or SP6 RNA polymerase. Furthermore, DNA molecules are provided that encode for an siRNA molecule according to the invention, as well as respective expression vector constructs.

Description

METHOD FOR PRODUCING SIRNA MOLECULES AND SIRNA MOLECULES PRODUCED THEREWITH

The present invention relates to the production of siRNA molecules and, in particular, an improved siRNA molecule comprising a 5 '-protruding end that corresponds to the consensus sequence of the nucleotides 0 to +5 of the natural promoter of a DNA-dependent RNA- polymerase. In preferred embodiments, the siRNA molecule according to the invention contains the natural promoter of T3, T7 or SP6 RNA polymerase. Furthermore, DNA molecules are provided that encode for an siRNA molecule according to the invention, as well as respective expression vector constructs.

Background of the invention

RNA-interference (RNAi) designates a mechanism for the inhibition ofthe genetic expression by the selective RNA-degradation that is conserved in all multicellular organisms. This mechanism that has been discovered only very recently is designated as the "immune system of the cell" and plays an important physiological role, amongst others, in the suppression of viral infections. This process is triggered by short double-stranded RNA-molecules (small interfering RNAs, siRNAs). These associate in the cell with a protein complex (RISC; RNA induced silencing complex) having ribonuclease activity. Thus, the complex is enabled to specifically bind and cleave such mRNA-molecules that contain a sequence motive that is homologous to the siRNA-sequence. Not all siRNA-sequences appear to be equally active. In systematic tests the observed reduction of the gene expression varied between 0 and by far more than 90 %.

In the last two years RNAi has developed into a indispensable tool in biotechnological and pharmaceutical research and development. Market researchers predict for the next five years an increase ofthe current market size from estimated 38 Mio US$ to 185 Mio US$. Currently, three different product groups are marketed, by which a targeted RNA-degradation can be achieved: synthetic double-stranded RNA-oligonucleotides, siRNA kits, wherein the siRNA is generated enzymatically in vitro, and vector systems (plasmids and viruses) as transport vehicles for the intracellular expression of short double-stranded RNA. Currently, the market for RNAi is dominated by the synthetic siRNA-oligonucleotides that are introduced transiently into cells or animals, and lead to the immediate and short-lived (up to 4 days) reduction of the genetic expression. According to estimates of Eric Ladder, Quiagen NN (Nenlo, The Netherlands), the sales of RNA-oligonucleotides increased during the course of only one single year from 0 to 15 Mio US$. The advantage of the comparably high specificity and sensitivity of these oligonucleotides is opposed by the drawback of very high production costs and market prices. Thus, synthetic siRNAs are up to now primarily used for the so-called "knock-down" of selected candidate genes and not for e.g. screening methods.

DNA-vector systems combine the RNAi-technology with tested delivery-systems and provided the basis for long-term knock-down strategies. The market share of the recently released vector systems continuously increases and is estimated to reach 25-30% in 2008 (Achema Trendbericht Nr. 20: Biotechnologie-RNA-Interferenz, 2003).

Until now, kits for the enzymatic synthesis of siRNAs do not play a big role in biotechnological research. According to the inventors' own analysis of the scientific literature, the amount of publications that have been generated by using siRNA kits represent only a very small percentage. The major part is represented by kits that rely on the synthesis of large double-stranded RNA molecules that are subsequently cleaved by means of double- strand-specific ribonucleases. The method requires a laborious purification of the reaction products. In addition, this method is only suitable for use within tightly defined analyses of genetic function. By using this technique it is, due to the use of siRNA mixes, impossible to identify active siRNA sequences which can be used for subsequent applications (vector systems; synthesis of defined siRNAs for e.g. animal testing). These drawbacks have been overcome with the use of the Gene silencer™ kit (Ambion, Inc., Austin, TX) wherein the direct synthesis of siRNA-molecules by in vttro-transcription using DNA-dependent polymerases is performed. This in principle attractive approach nevertheless requires a substantial timely effort and leads only to a low cost-advantage in comparison with synthetic siRNA-molecules, suffers from highly variable synthesis yields and often results in cytotoxic side-effects ofthe products ofthe synthesis. Due to these drawbacks, the system could not be successfully introduced into the market, yet. Yu et al (in 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. 2002 Apr 30;99(9): 6047-52. Epub 2002 Apr 23) describe the in vitro transcription of siRNAs with T7 RNA polymerase. It was found that hairpin siRNAs or siRNAs expressed from an RNA polymerase III vector based on the mouse U6 RNA promoter effectively inhibited gene expression in mammalian cells. It was furthermore observed that mismatches within hairpin siRNAs can increased the strand selectivity of a hairpin siRNA, which may reduce self- targeting of vectors expressing siRNAs. Furthermore, the use of hairpin siRNA expression vectors for RNAi gene therapy is mentioned.

Donze and Picard (in RNA interference in mammalian cells using siRNAs synthesised with T7 RNA polymerase. Nucleic Acids Res. 2002 May 15;30(10):e46) describe an alternative method to obtain cheap and large amounts of siRNAs using T7 RNA polymerase. With multiple transfection procedures, including calcium phosphate co-precipitation, silencing of both exogenous and endogenous genes was demonstrated.

Katoh et al (Katoh T, Susa M, Suzuki T, Umeda N, Watanabe K, Suzuki T Simple and rapid synthesis of siRNA derived from in vitro transcribed sliRNA. Nucleic Acids Res Suppl. 2003 ;(3). -249-50) describe a method for preparing active siRNA derived from short hairpin (sh) RNA which is transcribed from a single-stranded synthetic DNA template using T7 RNA polymerase.

Kumar et al (Kumar R, Conklin DS, Mittal V. High-Throughput Selection of Effective RNAi Probes for Gene Silencing. Genome Res. 2003 Oct;13(10):2333-40) describe an approach for the rapid and efficient identification of the most effective siRNA against any gene. The efficacy of siRNA sequences is monitored by their ability to reduce the expression of cognate target-reporter fusions with easily quantified readouts. Using microarray-based cell transfections, the potential of this approach in high-throughput screens for identifying effective siRNA probes for silencing genes in mammalian systems was demonstrated.

Harborth et al. (Harborth J, Elbashir SM, Nandenburgh K, Manninga H, Scaringe SA, Weber K, Tuschl T. Sequence, Chemical, and Structural Variation of Small Interfering RΝAs and Short Hairpin RΝAs and the Effect on Mammalian Gene Silencing. Antisense Nucleic Acid Drug Dev. 2003 Apr;13(2):83-105) investigated the positional variation of siRNA-mediated gene silencing. Fluorescent chromophores did not perturb gene silencing when conjugated to the 5'-end or 3'-end of the sense siRNA strand and the 5'-end of the antisense siRNA strand, but conjugation to the 3'-end of the antisense siRNA abolished gene silencing. RNase- protecting phosphorothioate and 2'-fluoropyrimidine RNA backbone modifications of siRNAs did not significantly affect silencing efficiency. Synthetic RNA hairpin loops were subsequently evaluated for lamin A/C silencing as a function of stem length and loop composition. As long as the 5'-end ofthe guide strand coincided with the 5'-end ofthe hairpin RNA, 19-29 base pair (bp) hairpins effectively silenced lamin A/C, but when the hairpin started with the 5 '-end ofthe sense strand, only 21-29 bp hairpins were highly active.

The commonly used methods (published protocols at www.ambion.com; commercially available Gene Silencer™ Kit from Ambion, Inc.) for the enzymatic synthesis of siRNA by means of in vitro transcription of a DNA template are either based on the separate synthesis of two short single-stranded RNA molecules out of which the active double-strand is generated by hybridisation (Ambion Gene Silencer™ kit) or by the transcription of a longer template to a RNA-molecule that folds itself into a "hairpin". Other protocols for the production and use of siRNAs can be found at www. a axa. com; www.invivogen.com; www.promega.com and/or www.genetherapysystems.com

Both above methods (not only) in the experiments of the present inventors proved to be unsuitable both for the reliable production of single siRNAs, and particularly for the parallel synthesis of a larger number of siRNA molecules. In doing so, the effort that was related with the synthesis and the purification, and in particular the huge variations in the yields of the synthesis were not tolerable. It was furthermore problematic that only a small fraction (<30%) of all templates was efficiently transcribed.

Although certain progresses have been achieved with respect to the production of synthetic siRNAs, these systems, nevertheless, still exhibit different drawbacks that lead to the need of improved possibilities for the production of siRNAs.

Currently, the widespread use of the breakthrough RNAi-technology is inhibited by the high cost factor of synthetic siRNAs and the high timely effort, and the low reliability of systems for enzymatic synthesis. Due to the constantly increasing need of siRNAs, in particular in genomic research that is often based on screening methods, the object ofthe present invention is to provide tools and methods as an adequate alternative for the available siRNA production, and to enable high throughput methods. The inventive siRNA synthesis should, at least partially, fulfil the following criteria in order to provide essential advantages in comparison with already available systems. Furthermore

- robustness

- biocompatibility

- cost effective production

- fast, "point of analysis" (i.e. to be used where needed) and easy to use production

- "biological" interface with vector-based siRNA systems

- suitable for automatisation

This object is achieved according to a first aspect of the present invention by providing an siRNA molecule that comprises a 5 '-protruding end that corresponds to the consensus sequence of the nucleotides 0 to +5 of the natural promoter of a DNA-dependent RNA- polymerase. In a preferred embodiment of the present invention, said natural promoter is selected from the group of T3, T7 or SP6 RNA polymerase.

Without wanting to being bound by this theory, the failure of many DNA-dependent polymerases, in particular the T7-RNA-polymerase, to efficiently transcribe a majority of the template could be either based on an inefficient initiation of the transcription or by the sequence to be transcribed itself. It is known that the transcripts have to start with a guanine. If one further compares several naturally occurring polymerase promoters, in particular T7- polymerase promoters, it is apparent that the 5 basepairs following the guanine downstream of the transcription start-site are conserved and thus appear to be functionally relevant. Until today, this conserved region has never been taken into account, in particular since until now one had to assume that at the 3' end of a functional siRNA molecule an overhang of two basepairs was mandatory. This leads to the fact that the transcribed siRNA-sequences starts with a guanine, but after this immediately continues into the gene specific sequences, such that the following five basepairs are differing in dependency from the target gene sequences from the optimal consensus sequence (Figure 1A). For optimising of the initiation of the transcription, the inventors therefore introduced a "leader-Sequence" that consists out ofthe 6 transcribed bases having nearly perfect, preferably perfect sequence identity to the consensus sequence of a naturally occurring polymerase promoter, in particular a T7-polymerse promoter (Figure IB).

These modifications according to the present invention lead to a modified siRNA molecule according to the present invention that comprises a 5' -protruding end. Thus, in another preferred aspect of the siRNA molecule according to the present invention the 5^Λ -protruding end comprises the sequence GGGAGA or GAAGAG.

As indicated above, the present invention generally employs a modification of the area of the transcript following the guanine downstream of the transcription start-site and a "leader- Sequence" that consists out ofthe 6 transcribed bases having nearly perfect, preferably perfect sequence identity to the consensus sequence of a naturally occurring DNA-dependent polymerase promoter, in particular a T7-polymerse promoter (Figure IB).

In another aspect of the present invention, a DNA molecule is provided that encodes for an siRNA molecule according to the present invention. Said DNA molecule forms the basis for the transcription in order to produce the above siRNA molecule of the present invention. Methods how to design and produce DNA molecules for such purposes include standard methods that are well known to the person of skill in molecular genetic technology. In addition, such methods can be furthermore found in the literature as cited herein.

Thus, in yet another aspect the present invention provides an expression vector that comprising a DNA molecule as above. According to the present invention, such an expression vector is preferably selected from in vitro and in vivo expression vectors. Respective hairpin siRNA expression vectors for RNAi gene therapy (i.e. in vivo expression) are known to the person of skill in the art and are mentioned, for example, in Yu et al. (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. 2002 Apr 30;99(9):6047-52. Epub 2002 Apr 23). Likewise, respective hairpin siRNA expression vectors for in vitro production can be found from the literature as cited herein as well. These vectors can be easily modified by the person of skill in the art in order to be applied in the methods of the present invention. The expression vectors usually contain all genetic elements that are necessary for the production of a specific siRNA molecule. In some embodiments of the present invention, the expression vectors according to the present invention can have the form of a "transgene", i.e. an expression element in, for example, a suitable vector that is designed for an expression and particularly an inducible and/or controllable expression in vivo. Accordingly, the transgene comprises nucleic acids ofthe present invention together with certain genetic control elements as discussed further below.

In yet another aspect ofthe present invention, a host cell is provided that has been transfected or infected with an expression vector according to the present invention. Methods for transfection and/or infection of host cells are well known to the person of skill in the art and described in the respective literature (see, for example : Herweijer H, Wolff JA. Progress and prospects: naked DNA gene transfer and therapy. Gene Ther. 2003 Mar; 10(6): 453-8.). Further examples are also described herein in example 2, below. The transfection can also be accomplished by means of oligofectamine (InNitrogen). In a preferred embodiment of the host cell according to the present invention, said cell is a mammalian cell.

Yet another aspect of the present invention is related to a transgenic, non-human animal, one or more of whose cells comprise a transgene according to the present invention, wherein the transgene is expressed in one or more cells of the transgenic animal resulting in the animal exhibiting ribonucleic acid interference (RNAi) ofthe target gene by the expressed siRNA.

Methods for producing transgenic animals are known to the person of skill in the art and well described in the literature, and such methods can be easily modified in order to be suitable for the needs according to the present invention. Preferred is an animal that is a non-human primate, e.g. a rodent, such as a mouse or a rat. More preferred is a transgenic animal according to the present invention, wherein the transgene is expressed with an efficiency of more than 90% compared to commercially available expression systems. More preferably, the transgene in said transgenic animal is expressed in one or more cardiac cells, lymphocytes, liver cells, vascular endothelial cells, or spleen cells. In another preferred embodiment, the expression ofthe transgene is constitutive or inducible and/or tissue specific, i.e. the transgene comprises nucleic acids of the present invention together with certain specific genetic control elements.

Yet another aspect of the present invention is related to a cell that is derived from the transgenic animal according to the present invention. Preferably, said cell is a lymphocyte, a liver cell, a cardiac cell, a vascular endothelial cell, or a spleen cell. Another aspect of the present invention is related to a method for producing a genetic construct for the production of an siRNA molecule according to the present invention. In general, three different embodiments of the above constructs are provided, depending from the way, how the respective siRNA shall be produced, namely

- synthetic double-stranded RNA-oligonucleotides,

- siRNA kits, wherein the siRNA is generated enzymatically in vitro, and

- vector systems (plasmids and viruses) as transport vehicles for the intracellular expression of short double-stranded RNA.

Chemical synthesis: The chemical synthesis is based on a step wise addition (3' to 5') of modified (protected) nucleotides to an oligonucleotide that is covalently attached to a solid support. Details for chemical synthesis (see also Dharmacon Homepage at www.Dharmacon.com): Dharmacon RNA synthesis chemistry is based on a novel protecting group scheme ((a) Scaringe, S. A. Ph.D. Thesis, University of Colorado, 1996. (b) Scaringe, S. A. and Caruthers, M. H. "Silyl Ether Protection of the 5'-Hydroxyl during Solid Phase Oligonucleotide Synthesis," in preparation, (c) Scaringe, S. A. and Caruthers, M. H. "5'-O- Silyl Ethers in Conjunction with Acid-labile 2'-O-orthoesters for the Solid Phase Synthesis of RNA." in preparation.). A new class of silyl ethers is used to protect the 5'-hydroxyl (5'-SIL) in combination with an acid-labile orthoester protecting group on the 2'-hydroxyl (2'-ACE) (Scaringe, S. A., Wincott, F. E. and Caruthers, M. H. "Novel RNA Synthesis Method Using 5 '-Silyl-2 '-Orthoester Protecting Groups," J Am. Chem. Soc, 120, 11820-11821 (1998).). This set of protecting groups is then used with standard phosphoramidite solid-phase synthesis technology ((a) Matteucci, M. D. and Caruthers, M. H. J Am. Chem. Soc. 103, 3185-3191 (1981). (b) Beaucage, S. L. and Caruthers, M. H. Tetrahedron Lett. 22, 1859-1862 (1981)).

For enzymatic synthesis the kit manual from Ambion can be found at: http://www.ambion.com/techlib/prot/bp 1620.pdf Two synthetic, partially complementary DNA oligos are hybridised for each strand ofthe siRNA molecule. After a fill in reaction with Klenow-Polymerase, the double stranded templates are transcribed by T7 RNA polymerase. The resulting single stranded RNA reaction products are then combined, hybridised and processed with a mixture of sequence specific ribonucleases to generate 2bp 3'protuding ends. The reaction products are purified using small spin- columns. Alternatively, one template is used to generate small hairpin RNA molecules (see, for example, Proc Natl Acad Sci U S A. 2002 Apr 30;99(9):6047-52.).

For plasmid based siRNA, DNA oligos coding for a short hairpin sequence (19 to 21bp stem; typically 9bp loop) are cloned behind a HI or an U6 promoter. After transfection into mammalian cells, transcription of these promoters by RNA polymerase III produces small hairpin RNA molecules, that are processed by an intracellular RNAse (DICER). This generates siRNA molecules with 2bp 3 '-protuding ends (see, for example, Science. 2002 Apr 19;296(5567):550-3).

Another aspect ofthe present invention is related to an in v/trø-method for producing siRNA, comprising the steps of a) providing an expression construct according to the present invention, b) admixing said expression construct with a recombinant DNA-dependent RNA- polymerase (e.g. T7 RNA-polymerase), ribonucleotides, suitable buffers and RNAse inhibitor, c) incubating said mixture from step b) at a suitable temperature from 30 minutes to 24 hours for transcription of the template, d) optionally, inactivating the DNA template (e.g. by DNase treatment, and e) purifying the expressed siRNA. The person of skill will realise that this method can be modified in order to be suitable for different DNA-dependent RNA polymerases. Alternatively, a commercial kit can be used based on the manufacturer's recommendations.

In yet another aspect, the present invention provides an in vttro-method for producing an siRNA, which in one embodiment in principle comprises the steps ofthe method as described, for example, in Yu et al. (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. 2002 Apr 30;99(9):6047-52. Epub 2002 Apr 23), but wherein a construct according to the present invention is employed. Alternatively, other methods as described in the present state of the art can be easily modified by the skilled artisan, such as, for example, methods as described by Donze and Picard (in RNA interference in mammalian cells using siRNAs synthesised with T7 RNA polymerase. Nucleic Acids Res. 2002 May 15;30(10):e46), Katoh et al (Katoh T, Susa M, Suzuki T, Umeda N, Watanabe K, Suzuki T Simple and rapid synthesis of siRNA derived from in vitro transcribed shRNA. Nucleic Acids Res Suppl. 2003;(3):249-50), and Kumar et al (Kumar R, Conklin DS, Mittal V. High-Throughput Selection of Effective RNAi Probes for Gene Silencing. Genome Res. 2003 Oct;13(10):2333- 40) and the respective references as cited therein. In another aspect of the present, the inventive in- vitro method provides an efficiency of more than about 75%, preferably 90% and most preferred about 95%. The "efficiency" of siRNA sequences is monitored according to the present invention by their ability to reduce the expression of cognate target-reporter fusions with easily quantified readouts. Using microarray-based cell transfections, the potential of this approach in high-throughput screens for identifying effective siRNA probes for silencing genes in mammalian systems was demonstrated. The efficiency can also be monitored as described in the present examples, in particular example 2.

In yet another aspect of the present invention, a method of inducing ribonucleic acid interference (RNAi) of a target gene in a cell in an animal is provided, the method comprising obtaining a transgenic animal according to the present invention comprising a transgene comprising a nucleic acid molecule encoding an engineered siRNA precursor; and expressing the siRNA precursor to form a small interfering ribonucleic acid (siRNA) within the cell, thereby inducing RNAi of the target gene in the animal. Preferably, said method of inducing ribonucleic acid interference(RNAi) of a target gene in a cell comprises obtaining a host cell of the invention; culturing the cell; and enabling the cell to express the siRNA precursor to form a small interfering ribonucleic acid (siRNA) within the cell, thereby inducing RNAi of the target gene in the cell. Related publications are cited in the present specification and can be further found in, for example, Matsukura et al. (Matsukura S, Jones PA, Takai D. Establishment of conditional vectors for hairpin siRNA knockdowns. Nucleic Acids Res. 2003 Aug 1; 31(15): e77.), and Gou et al (Gou D, Jin N, Liu L. Gene silencing in mammalian cells by PCR-based short hairpin RNA. FEBS Lett. 2003 Jul 31; 548(1-3): 113-8.) and the respective references as cited therein.

The methods according to the present invention have a number of essential advantages in comparison with the commercially available siRNA-synthesis kit of Ambion, Inc.:

1.) A clearly increased percentage of efficiently transcribed template in comparison with the concomitant methods (>90% vs. <30%, see also Figs. 2 and 5)

2.) The synthesis occurs cost effectively by using a single gene specific template-oligo as well as with universal T7-promoter-oligos (Figs. 5 and 1)

3.) The synthesis requires only one transcription reaction per siRNA. 4.) The synthesis is linked with a clearly shorter production time (Fig. 5) 5.) The synthesis is easier to handle and can additionally be automated since only one hybridisation reaction, one transcription reaction as well as one optional DNAse I -digest are required.

As discussed, the present invention in accordance with the object thereof can establish a novel way of producing siRNA that fulfils the given criteria and therefore has a broad utility. The siRNAs that are produced enzymatically by this method are not only much cheaper than chemically synthesised siRNA-molecules, but due to the ability to automate the production this method in particular appears to be suited for high-throughput projects in the genomic research and in target gene validation as well as in academic research groups and in pharmaceutical industry. In the sense of a high-throughput strategy, using the synergistic combination from technical feasibility and economic affordability, it is now possible to employ a strategic use for siRNA-technology on a broad basis, that until now was not available in this particular form. In addition, since the synthesis is easy to perform even for inexperienced users, the use of this technique in form of a siRNA-synthesis kit is envisaged and encompassed by the present invention as well.

Since the majority of the synthesis costs (>80%) is related with the DNA-template molecule, but for each synthesis only less than a thousandth of the delivered amount of template is required and the DNA-Templates due to their stability can be stored and shipped without problems, template-set could be individually put together in accordance with the customers needs and wishes that could be delivered lyophilised in 96 Well-plates (routine step at certain providers of DNA-oligos) and then be transcribed by the final user without problems. If wanted, the transcription reaction could be delivered in pre-prepared form. Thus, this technique would be particularly useful for companies having expertise in the production of DNA-oligonucleotides of interest, since the production costs of such a kit could be held particularly low. Furthermore, such a kit due to its easy handling would be useful for a large group of scientists.

One additional advantage of the method according to the invention is that the template oligos can be modified to be directly subcloned into such vectors. This enables a two step strategy, in which a) active siRNA sequences are first identified using enzymatically synthesised siRNA and b) in the second step, the corresponding short hairpin expression vectors are constructed, using a strategy based on the existing DNA-template oligos. The shRNA coding insert will be transcribed by RNA polymerase III. Dependent on the design ofthe leader sequence, the resulting constructs will either express the novel 5 'extended siRNA molecules or conventional, 3 'protuding hairpins.

To retain the 5' protruding end, a variant of the universal T7 primer (GGAGGGATCCGGTAATACGACTCACTATA (Seq ID. No. 1) instead of GGTAATACGACTCACTATA (Seq ID. No. 2)) has to be used. The 9 additional nucleotides (bold letters) at its 5 'end introduce a Bglll restriction site. After hybridising it to the template oligonucleotide, the partially double stranded hybrid has to be converted to full double strands. This is achieved by treatment with DNA-dependent DNA polymerases (Klenow enzyme; Taq-Polymerase), that extend the 3 'end of the universal T7 primer oligonucleotide. The resulting double stranded DNA can be restricted with Bglll and subcloned into shRNA expression vectors with RNA polymerase III promoters. The resulting transcripts generated in mammalian cells will be retain the 5 'protuding end.

In order to generate expression vectors that produce conventional shRNA molecules with 3 ' protruding ends, the leader sequences GGGAGA or GAAGAG have to be modified to include a restriction site or the leader sequence has to be extended by additional nucleotides to generate a restriction site. This can be achieved by either modifying the leader sequence to GGGATC or by extending it to GGGAGATC to introduce a Sau3A restriction site. After converting the partially double stranded hybrid of T7 universal primer with the template oligonucleotide into complete double strand form by primer extension (see above), a universal TTTT-Hindlll linker has to be added. Following digestion with Sau3A and Hindlll, the resulting cleavage product can be subcloned into shRNA expression vectors. The shRNA transcripts will display conventional 3 'protuding ends.

As mentioned above, in vitro synthesised siRNAs are primarily used for the transient application in cell culture and animal models, since the effect of transfection levels off after a few days. In order to achieve a long-term effect that is particularly needed in animal models vector systems have recently been introduced on the market that permanently or inducibly express siRNAs after (stable) transfection into cells. The transfer of the information of promising siRNA target sequences from the system of chemically synthesised siRNAs to hairpin molecules produced in vivo is currently possible only to a very limited extent. Also in this case, the inventive system offers an essential advantage. Due to the successful in vivo testing of the inventive siRNAs produced in vitro using the inventive system, the positively evaluated hairpin sequences can be introduced 1:1 into vector systems. Envisaged by the invention is the introduction of restriction sites into the oligonucleotide templates said sites allowing the direct subcloning of the templates into respective vectors. Thus, the presented systems provides a direct "biological interface" between siRNAs produced in vitro and in vivo.

Preliminary studies for an automatisation have already been successfully performed. For this, 69 different siRNA-molecules were synthesised in parallel by means of manually operated multi-channel pipettes in a 96-well-plate. The synthesis products were then detected in polyacrylamid gels (Figure 2). It could be shown that nearly all templates were transcribed with the same efficiency. The transfer of this synthesis approach into a robot-operated system is therefore very realistic, and it should pose no further problems for the person of skill to amend the present methods accordingly.

The invention shall now be further described in the following examples with reference to the accompanying figures. The example are provided for illustration and not limitation. In the Figures:

Figure 1: Comparison of the structure of common or optimised DNA-templates for enzymatic siRNA-synthesis.

(A) Common synthesis: bases 1-19: T7 Promotor; Pos. 20: start of transcription.

(B) Inventive synthesis: bases 1-19 T7 Promotor; Pos. 20: start of transcription. 20-25 Leader- sequence; NNN = target gene specific sequence segment.

Figure 2A: Comparison of the transcription efficiencies of conventional DNA-templates (A) and the inventive optimised systems (B).

Figure 2B: Comparison of the transcription efficiencies of conventional DNA-templates shl- sh20 and the inventive optimised systems sh.81-100.

Figure 3: Three independent assays for a comparison of the activity of commercial chemically synthesised siRNAs (E2F4C), and siRNAs (E2F4 E) that are enzymatically produced in accordance with the inventive method having an identical E2F-4 mRNA specific sequence.

Figure 4: Analysis ofthe biocompatibility ofthe synthesis products.

Figure 5: Comparison of the performance of the synthesis kit of the company Ambion with the present inventive method.

Examples

Materials and Methods

T7-template and siRNA oligos

All DNA template oligos were obtained form Metabion, Germany. Equimolar ratios of T7- template oligos (see below) and a universal T7 oligonucleotide (5'GGTAATACGACTCACTATA3' (Seq ID. No. 2)) were hybridised in lOmM Tris-HCl pH 8.0; 50mM NaCl by heating to 95°C for 5 minutes and subsequently, samples were allowed to cool down to room temperature. This yields partial (19bp) double stranded DNA templates for in vitro transcription. The sequence of the template oligos for conventional in vitro transcriptions is:

3 ' CCATTATGCTGAGTGATATNNNNNNNNNNNNNNNNNNaagttctctNNNNNNNNNNNNNN NNNNAA 5 ' (Seq ID. No. 3) and for the inventive optimised system:

3 ' CCATTATGCTGAGTGATATccctctNNNNNNNNNNNNNNNNNNaagttctctNNNNNNNN NNNNNNNNAA 5 ' (Seq ID. No. 4)

The 19 bp sequence stretches coding for the siRNA (sense and antisense) are indicated by rows of Ns.

The oligonucleotide sequences used are listed below. siRNA sequences (5'-> 3') as transcribed in the conventional system (Fig. 2 A (A)): GGAGCTGGTGCCCAGCATC (Seq ID. No. 5), GCACCTCATCGACTACATC (Seq ID. No. 6), GATGGAAATCCTGCAGCAC (Seq ID. No. 7), GATGTGCATGGGCAATTTC (Seq ID. No. 8), GAGGGATATAACTGGTGCC (Seq ID. No. 9), GCCTCTTGTCATCAACAGC (Seq ID. No.10), GAAAGCACCTGCTATGTTC (Seq ID. No.11), GACTACCGATGGTTACTTG (Seq ID. No. 12), GGAGCTTCATGGTGAAGGC (Seq ID. No. 13), GAAAGTTGTGTGGAAGTTC (Seq ID. No.14), GGCAGCTGCAATGGAAAAC (Seq ID. No. 15), GTGATTTGTACCTCAGAGC (Seq ID. No. 16), GATGATGGAGGAAGCCACC (Seq ID. No.17), GGGCTGTGAAGCTGGAAAC (Seq ID. No.18), GAATCAGGACACATCTTCC (Seq ID. No. 19), GAGAGATGTGGAATAACAC (Seq ID. No. 20), GTTCTTGTTTGGGCACAGC (Seq ID. No.21), GCACCCAGATGCTTCAGTC (Seq ID. No. 22), GGTCAGCTGGTTAAGATGC (Seq ID. No.23), GCCCATACCTTTATCCACT (Seq ID. No.24), GACTTGGGAACAGAAAGCC (Seq ID. No.25), GACTCATTCTCACAGTTCC (Seq ID. No. 26), GTTCCTTTAGAGGTTCCTC (Seq ID. No. 27), GCAAATCTTGGTTGCCTTC (Seq ID. No.28).

siRNA sequences (5'-> 3') as transcribed by the inventive optimised system (Fig.2A (B)): GCTGTGCATCTACACCGAC (Seq ID. No.29), GTTCATTTCCAATCCGCCC (Seq ID. No. 30), GCTTCATTCTCCTTGTTGT (Seq ID. No. 31), GTCACCTCTTGGTTACAGT (Seq ID. No.32), GAAGAGGTCTTCCCTCTGG (Seq ID. No.33), GTCCCATCTGCAACTCCTG (Seq ID. No. 34), GCATGCTCAGACCTTCATT (Seq ID. No. 35), GCACCTTTGCTCTCAGCCA (Seq ID. No.36), GATGCTGGCTTACTGGATG (Seq ID. No. 37), GCCCTGTTGACACAGGTCT (Seq ID. No. 38), ACTGTGCATCTACACCGAC (Seq ID. No.39), GCCGCACATGCGGAAGATG (Seq ID. No.40), GATAGCACGATCTTGTCAG (Seq ID. No. 41), GCTCAGTAACTGGGAACCA (Seq ID. No. 42), GTGGTTTCCAAGGTGTGAG (Seq ID. No.43), GAACAGAATGGCTCTCTTT (Seq ID. No. 44), GCCTGTTCAGGACTTTCCT (Seq ID. No.45), GGGCTGCATCATGAAGCAG (Seq ID. No.46), ATCCAGGCCACTATGGAGG (Seq ID. No.47), CACCTATCCCTCCTTCCTT (Seq ID. No. 48), GAACCCAGACCTCGAGTTT (Seq ID. No. 49), TGATCTGCAAGAACCCAGA (Seq ID. No.50), GTTCACAAGACCAGACCAC (Seq ID. No. 51), GATAGAGGAGAGGTTGCAG (Seq ID. No. 52), GGAATGGTCCACTTCCATT (Seq ID. No.53), GATCAGTCACCAGAGAGAT (Seq ID. No.54), GAGGATGGAGAGGAGCTCA (Seq ID. No. 55), GGACAACAAGGTTGCTGTA (Seq ID. No. 56), GAAGTACATCATCTACGCC (Seq ID. No.57), GACTTTTGCCCGCTACCTT (Seq ID. No. 58), GGATTTTGTGGCCTCCATT (Seq ID. No. 59), GGATGAGGAGAACAATCCC (Seq ID. No.60), GAAGGGAGGCTACACCTCT (Seq ID. No.61), GCAATCCTCTGCTTGCTCT (Seq ID. No. 62), GTGGATCTGCAGTCTGACG (Seq ID. No. 63), GACATTGACCAGCTCATCA (Seq ID. No.64), GGCTGGGATCATCTGTCAG (Seq ID. No. 65), GGACTACATTGCCTACGCG (Seq ID. No.66), GAACAGTATCCAGTGGAAG (Seq ID. No.67), GTTGTGGCTACAGCAAAGC (Seq ID. No.68), GAGTCACTCAGGGCCTATC (Seq ID. No. 69), GCCAGACTTTCCAGCAGAC (Seq ID. No. 70), GAGGCTCCTGAGAAGAAGG (Seq ID. No.71), GAGGCTCCTGAGAAGAAGG (Seq ID. No. 72), GCGCAGACACAGACAGTCC (Seq ID. No.73), GACAAGAAGGACACAGGTG (Seq ID. No.74), GGAGGTTTTGGAGGAGGGG (SeqID. No.75), GCTGTGGTCGAGCAAATAG (Seq ID. No. 76), GGAGCTGGACCTGTACCTG (Seq ID. No. 77), GTTGTGAGTCTCTTCCCCA (Seq ID. No.78), GGAGAAGGAGCAGGAGAAT (Seq ID. No. 79), GCCCACAAGCTCTCTTTCC (Seq ID. No.80), GGACAATGCCCTGCTGACG (Seq ID. No.81), GCCAAAAAGCCGTCCCACA (Seq ID. No.82), GGAGACCTTTCTGCACAGT (Seq ID. No. 83), GCAAGTGAGCGCTGTCACC (Seq ID. No. 84), GGGAGAGAAGGGCATTGGT (Seq ID. No.85), GACCAGAACTCAAAAGCAG (Seq ID. No. 86), GATCATCTTGTCAGATGAG (Seq ID. No.87), GCTCATCACCCTCCCACGG (Seq ID. No.88), GGTCATCACCCTCCCACGG (Seq ID. No.89), GCACTGCAGAGACATGGAA (Seq ID. No. 90), CTGGACTTCCAGAAGAACA (Seq ID. No. 91), CGGCAGGACTCCGGGCCGA (Seq ID. No.92), TGAGAAGTCTCCCAGTCAG (Seq ID. No. 93), CTCTGGAGGAAAAGAAAGT (Seq ID. No.94), CATACTGGCCTGGACTGTT (Seq ID. No.95), AGCTGATATTGATGGACAG (Seq ID. No.96), CAGTTGTGGTTAAGCTCTT (Seq ID. No.97) .

siRNA sequences (5'-> 3') as transcribed in Figure 2B Sh 1-20:

GGAGCTGGTGCCCAGCATC (Seq ID. No.98), GCACCTCATCGACTACATC (Seq ID. No. 99), GATGGAAATCCTGCAGCAC (Seq ID. No.100), GATGTGCATGGGCAATTTC (SeqID. No. 101), GAGGGATATAACTGGTGCC (Seq ID. No.102), GCCTCTTGTCATCAACAGC (Seq ID. No. 103), GAAAGCACCTGCTATGTTC (Seq ID. No. 104), GACTACCGATGGTTACTTG (Seq ID. No.105), GGAGCTTCATGGTGAAGGC (Seq ID. No. 106), GAAAGTTGTGTGGAAGTTC (Seq ID. No.107), GGCAGCTGCAATGGAAAAC (Seq ID. No.108), GTGATTTGTACCTCAGAGC (Seq ID. No.109), GATGATGGAGGAAGCCACC (Seq ID. No. 110), GGGCTGTGAAGCTGGAAAC (Seq ID. No. Ill), GAATCAGGACACATCTTCC (Seq ID. No.112), GAGAGATGTGGAATAACAC (Seq ID. No. 113), GTTCTTGTTTGGGCACAGC (Seq ID. No.114), GCACCCAGATGCTTCAGTC (Seq ID. No.115), GGTCAGCTGGTTAAGATGC (Seq ID. No.116), GCCCATACCTTTATCCACT (Seq ID. No.117).

siRNA sequences (5'-> 3') as transcribed in Figure 2B Sh81-100:

GCTGTGCATCTACACCGAC (Seq ID. No.118), GTTCATTTCCAATCCGCCC (Seq ID. No. 119), GCTTCATTCTCCTTGTTGT (Seq ID. No.120), GTCACCTCTTGGTTACAGT (Seq ID. No. 121), GAAGAGGTCTTCCCTCTGG (Seq ID. No. 122), GTCCCATCTGCAACTCCTG (Seq ID. No.123), GCATGCTCAGACCTTCATT (Seq ID. No. 124), GCACCTTTGCTCTCAGCCA (Seq ID. No.125), GATGCTGGCTTACTGGATG (Seq ID. No. 126), GCCCTGTTGACACAGGTCT (Seq ID. No. 127), ACTGTGCATCTACACCGAC (Seq ID. No.128), GCCGCACATGCGGAAGATG (Seq ID. No. 129), GATAGCACGATCTTGTCAG (Seq ID. No.130), GCTCAGTAACTGGGAACCA (Seq ID. No. 131), GTGGTTTCCAAGGTGTGAG (Seq ID. No. 132), GAACAGAATGGCTCTCTTT (Seq ID. No.133), GCCTGTTCAGGACTTTCCT (Seq ID. No. 134), GGGCTGCATCATGAAGCAG (Seq ID. No.135), ATCCAGGCCACTATGGAGG (Seq ID. No.136), CACCTATCCCTCCTTCCTT (Seq ID. No.137).

siRNA sequences (5 '-> 3 ') as transcribed in Figure 2C

GCGGCGGATTTACGACATT (Seq ID. No.138), CCGGCGGATTTACGACATT (Seq ID. No. 139), GGGGCGGATTTACGACATT (Seq ID. No.140), GCCGCGGATTTACGACATT (Seq ID. No. 141), GCGCCGGATTTACGACATT (Seq ID. No. 142), GCGGGGGATTTACGACATT (Seq ID. No.143), GCGGCCGATTTACGACATT (Seq ID. No. 144), GCGGCGCATTTACGACATT (Seq ID. No.145), GCGGCGGTTTTACGACATT (Seq ID. No.146), GCGGCGGAATTACGACATT (Seq ID. No.147), GCGGCGGATATACGACATT (Seq ID. No. 148), GCGGCGGATTAACGACATT (Seq ID. No. 149), GCGGCGGATTTTCGACATT (Seq ID. No.150), GCGGCGGATTTAGGACATT (Seq ID. No. 151), GCGGCGGATTTACCACATT (Seq ID. No. 152), GCGGCGGATTTACGTCATT (Seq ID. No. 153), GCGGCGGATTTACGAGATT (Seq ID. No. 154), GCGGCGGATTTACGACTTT (Seq ID. No.155), GCGGCGGATTTACGACAAT (Seq ID. No. 156), GCGGCGGATTTACGACATA (Seq ID. No.157).

The siRNA transfected in Figure 3 was a chemically synthesised siRNA from Eurogentec; EGFP(C) upper strand: GGTGAACTTCAAGATCCGCdTdT (Seq ID. No. 158); lower strand: GCGGATCTTGAAGTTCACCdTdT (Seq ID. No. 159);

E2F4 (C) upper strand: GCGGCGGATTTACGACATTdTdT (Seq ID. No. 160); lower strand: AATGTCGTAAATCCGCCGCdTdT (Seq ID. No. 161); or an enzymatically produced siRNA

EGFP (E): transcribed sequence: GGTGAACTTCAAGATCCGC (Seq ID. No. 162);

E2F4 (E): GCGGCGGATTTACGACATT (Seq ID. No. 163).

In vitro transcription of T7-templates

In vitro transcription of the T7 templates was either performed using the AmpliScribe T7 Flash Transcription Kit from EPICENTRE (50ng template; 20 μl reaction volume; according to the manufacturers instructions) or using recombinant T7 RNA-polymerase, ribonucleotides and RNAse inhibitor from MBI Fermentas. (50-100 ng template; 50-500 μl volume; 200- 600U T7 RNA polymerase; 2mM NTP; 10 to 40 U RNAse inhibitor).

The reaction was allowed to proceed for 30 minutes to over night. DNA templates were degraded using DNase I for 30 minutes at 37°C. DNase I was inactivated by heating samples to 95°C for 5 minutes. Integrity of siRNA molecules was verified by acrylamide gel electrophoresis (20% acrylamide; lx TBE). siRNAs were visualised by EtBr staining. Quantification was performed using a raytest ccd video scanner and Diana developer software.

siRNA transfections were performed with Lipofectamine 2000 (HCT116 (ATCC CCL-247)) and Oligofectamine (K562 (ATCC CCL 243)) and HL60 (ATCC: CCL 240)) according to the manufacturers recommendations. Cells were harvested 72h after transfection and expression of proteins was analysed by western blotting.

Example 1: Comparison of the transcription efficiencies of conventional DNA-templates (Figure 2A) and the inventive optimised system (Figure 2B).

50 ng from each of 24 conventional and 69 optimised templates were transcribed with an in vz^'tro-transcription kit of Company Promega (RiboMAX Large Scale RNA Production System-T7) in a 96 well microtiter-plate and a reaction volume of 20 μl for 4 hours at a temperature of 37°C. One microliter of each reaction sample was separated in a native 20% polyacrylamid gel in lx TBE buffer. After staining of the RNA by means of ethidium bromide the documentation took place with a Raytest-video camera and the respective Data- Developer Software. Since gel analyses are presented that have been performed one after another using subsequent ethidium bromide staining, the intensities of the bands between different gels can only be compared to a limited extent.

Comparison ofthe transcription efficiencies of conventional DNA-templates shl-sh20 and the inventive optimised system sh81-100. - Each of 20 conventional (sh 1-sh 20) and 20 optimised templates (sh 81-shl 00) were transcribed in vitro in 2 or 3 parallel reactions in a 96-well microtiter plate using T7 polymerase (lOOng template; 25ul reaction; 4 hours at a temperature of 37°C). One microliter of each of the reaction mix was separated in a native, 20% polyacryl amide gel in lx TBE buffer. After the staining of the RNA by means of ethidium bromide the documentation took place with a Raytest-video camera and the respective Diana-Developer Software.

Comparison of the structure of common or optimised inventive DNA-template in case of enzymatic siRNA-synthesis (Figure 1). This led to the fact that the fraction of efficiently transcribed templates could be increased to more than 90% (Figure 2). For this, in typical synthesis reactions median yields in the range of two-digit micrograms were achieved (30- 50μg/sample). Thus, the first object of the present invention could be achieved, namely to generate a robust system, i.e. to successfully transcribe a particularly high percentage of the given template in comparable amounts.

In the functional tests of the invention, the siRNA-molecules that have been produced in accordance with this methods, despite the fact that their ends differed from the ideal structure, proved to be active. In several independent comparative experiments, no significant differences in the biological activity between the siRNAs that have been produced using commercial chemically synthesised oligos having 3 '-protruding ends and the siRNA- molecules that have been produced by in v/tro-transcription according to the invention could be found (Figure 3, see below).

Example 2: Comparison of the activity of commercial, chemically synthesised siRNA (E2F4C) with molecules produced according to the present invention.

Three independent assays for a comparison the activity of commercially-based, chemically synthesised siRNAs (E2F4C) and the enzymatically produced siRNAs (E2F4 E) with identical, E2F-4 mRNA specific sequence were performed. Same amounts (3μl of a 20uM siRNA solution were transfected by means of oligofectamine (InNitrogen [sic]) in 12 well- samples into the colon carcinoma cell line HCT116. 72 hours after transfection the cells were harvested and whole cell extracts were prepared. Same amounts of protein were then separated on a 15%) SDS-acrylamid gel, transferred onto a PNDF-membrane and the E2F4- Expression was determined with a monoclonal antibody against E2F4.

Example 3: Biocompatibility of the enzymatically synthesised siRΝA:

Analysis of the biocompatibility of the synthesis products. Same amounts of chemically synthesised siRNAs (RGFP-C; Cdkl-C; E2F6-C) and siRNAs synthesised according to the present invention (E2F4-E1; E2F4-E2; E2F4-E3) were transfected by means of oligofectamine (InVitrogen) into the cell lines HL60 and K562. 72 hours after transfection the cells were harvested and whole cell extracts were prepared. Same amounts of protein of all samples were then separated on a 14%) SDS -polyacrylamid gel and transferred onto a PNDF- membrane. The detection of the expression of the proteins E2F4, E2F6, E2F3 as well as CDC6 took place by specific antibodies and ECL (Amersham).

The commonly available enzymatic methods in accordance with the instructions of the manufacturers require phenolic extractions with subsequent ethanol precipitation or, as in the case of the Gene-silencer™ kit of Ambion, a purification of the transcripts by a glass fibre filter. Such intermediate steps harbour essential drawbacks: first, they are expensive, second, they usually lead to partial losses of the transcript, and third, they are unsuited for automatable production processes. Thus, the synthesis parameters were not only optimised in accordance with an efficient synthesis, but also the components of the synthesis sample were tested both for their necessity and biocompatibility. The goal was to be able to use the transcript without a further purification of the synthesis sample directly for transfections. For this, siRNAs according to the present invention were again tested with chemically produced and commercially available siRNAs as gold standard directly in HL60 and K562 (human leukaemia) cell lines. It could be shown that after optimisation the non purified synthesis products in both cell lines in view of their activity and specificity did not differ from the commercial, chemically synthesised molecules, and, in addition, no cytotoxic effects occurred (Figure 4). In all transfections only the target-gene E2F4 was repressed by the enzymatically synthesised siRNAs, whereas the expression ofthe closely related family members E2F3 and E2F6 as well as the expression ofthe control protein CDC6 were not affected. The above examples can be easily modified by the skilled artisan in order to be applied for other DNA-dependent RNA polymerases, in particular T3, T7 or SP6 RNA polymerase.

Claims

1. An siRNA molecule comprising a 5 '-protruding end that corresponds to the consensus sequence of the nucleotides 0 to +5 of the natural promoter of a DNA-dependent RNA-polymerase.

2. siRNA molecule according to claim 1, wherein the natural promoter is selected from T3, T7 or SP6 RNA polymerase.

3. siRNA molecule according to claim 1 or 2, wherein the 5" -protruding end comprises the sequence GGGAGA or GAAGAG.

4. DNA molecule, encoding for an siRNA molecule according to any of claims 1 to 3.

5. Expression vector, comprising a DNA molecule according to claim 4.

6. Expression vector according to claim 5 that is selected from in vitro and in vivo expression vectors.

7. Host cell, transfected or infected with an expression vector according to claim 5 or 6.

8. Host cell of claim 7, wherein the cell is a mammalian cell.

9. A transgene comprising the nucleic acid of claim 4 and/or claim 5.

10. Transgenic, non-human animal, one or more of whose cells comprise a transgene of claim 9, wherein the transgene is expressed in one or more cells of the transgenic animal resulting in the animal exhibiting ribonucleic acid interference(RNAi) of the target gene by the expressed siRNA.

11. The transgenic animal of claim 10, wherein the transgene is expressed with an efficiency of more than 90% compared to commercially available expression systems.

12. The transgenic animal of claim 10 or 11, wherein the transgene is expressed in one or more cardiac cells, lymphocytes, liver cells, vascular endothelial cells, or spleen cells.

13. The transgenic animal according to any of claims 10 to 12, wherein the expression is constitutive or inducible and/or tissue specific.

14. The transgenic animal according to any of claims 10 to 13, wherein the animal is a non-human primate, e.g. a rodent, such as a mouse or a rat.

15. A cell derived from the transgenic animal according to any of claims 10 to 14.

16. The cell of claim 15, wherein the cell is a lymphocyte, a liver cell, a cardiac cell, a vascular endothelial cell, or a spleen cell.

17. In vttro-method for producing an siRNA, comprising the steps of a) providing an expression construct according to claim 5 or 6, b) admixing said expression construct with a recombinant DNA-dependent RNA- polymerase, ribonucleotides, suitable buffers and RNAse inhibitor, c) incubating said mixture from step b) at a suitable temperature from 30 minutes to 24 hours for transcription, d) optionally, inactivating the DNA template, and e) purifying the expressed siRNA.

18. In v/^'trø-method for producing an siRNA according to claim 17, wherein the efficiency ofthe transcription is higher than about 90%, compared to commonly used methods.