Modified short interfering RNA

The present invention relates to double-stranded, short interfering RNA (siRNA) molecules that down-regulate the expression of target genes, whereby the siRNA molecules are chemically modified. Furthermore, the present invention relates to a support material as well as phosphoramidites suitable for producing respective siRNA molecules. It is known that siRNA molecules (double-stranded (i.e. duplex) short interfering RNA having a length of 21-23 nucleotides (nt) with terminal 3 '-overhangs of 2 nucleotides) can be used to post-transcriptionally silence gene expression in mammalian cells (Elbashir et al, Nature 2001, 411:494-498). Complexing of siRNAs with suitable transfection reagents and application of these complexes onto the cells results in endocytotic uptake of the siRNA- complexes. SiRNA is finally released into the cytoplasm and the siRNA molecules are recognized and incorporated into a complex called RISC (RNA induced silencing complex).. A RISC associated ATP-depended helicase activity unwinds the duplex and enables either of the two strands to independently guide target mRNA recognition (Tuschl, Mol. Interv. 2002, 2:158-167).

The RISC complex then recognizes a region within the mRNA to which the siRNA sequence is complementary and binds to this region in the mRNA. The mRNA is then endonucleolitically cleaved at that position, where the RISC complex is bound. In a final step, the endonucleolitically cleaved mRNA is degraded by exonucleases. This mechanism of siRNA mediated gene silencing (RNA interference, RNAi) is widely used to perform knockdown experiments in eukaryotic cell cultures.

It is also known that not every siRNA duplex is able to knockdown the mRNA level to the same degree. Some siRNAs are able to knockdown the initial level to approx. 10-20%, some others to intermediate levels (20-70%), and still others are not able to knockdown the mRNA level to a measurable level at all. There are certain design rules (so called "Tuschl-rules") for the design of siRNAs, but even siRNAs designed according to these rules show the above described variability. The reason for this observation is not known. It is believed that also the thermodynamic stability of the duplex may play a role in the efficiency of the siRNA to induce silencing. This variability as well as the fact that even "good" siRNAs are not able to completely knockout all mRNAs is a severe disadvantage of the siRNA technology (Gong D., Trensd. Biotechnol. 2004, 22:451-454; Khvorova A, Cell 2003, 115:209-216; Schwarz D.S., Cell 2003, 115:199-208; Reynolds A., Nat. Biotechnol. 2004, 22:326-330).

The search for more potent gene silencing either by enhancement of uptake into the cells or by more potent intracellular siRNA duplexes is therefore central to the field of siRNA research.

Moreover, it is loiown that several chemical modifications have been tested in order to increase the ability of siRNA molecules to induce silencing and also to increase their stability. Chemical modifications of siRNA molecules or oligonucleotides (ODN) are affecting the nuclease stability of these molecules or their affinity to target mRNAs {Manoharan M., Curr. Opin. in Chem. Biol. 2004, 8:570-579). Different types of modifications have been described to improve the stability against serum nucleases and/or to improve the affinity to targets. Described are:

a. 2'-OH modifications with halogen, amine, -O-alkyl, -O-allyl, alkyl-groups;

b. internucleotid linkages, such as orthoester, phosphate ester, phosphodiester, -triester, phosphorothioate, phosphorodithioate, phosphonate, phosphonothioate, phosphorothiotriester, phosphoramidate, phosphorothioamidate, phosphinate, and boronate linkage, ether-, allyl ether-, allyl sulfide-, formacetal/ketal-sulfide-, sulfoxide-, sulfone-, sulfamate-, sulfonamide-, siloxane-, amide-, cationic alkylpolyamine-, guanidyl-, morpholino-, hexose sugar or amide-containing linkage, or a two to four atom linkage;

c. conjugates, such as aminoacids, peptides, polypeptides, protein, sugars, carbonhydrates, lipids, polymers, nucleotides, polynucleotides, as well as combinations thereof, passive delivery with cholesterol conjugates.

Normal (unprotected) siRNAs show a low stability in cell culture media and in body fluids, such as serum. This results in degradation of siRNA during the transfection or the systemic delivery. Accordingly, due to the degradation, less siRNA remains in the cells for efficient silencing, and the degraded (shortened) siRNA may lead to off target effects in cells due to unspecific base pairing. Transfection of siRNA to cells is shown to lead to off target effects. Specifically, high amounts of siRNA may lead to PKR activation (PKR: Protein Kinase R; double stranded RNA-activated protein kinase). Also, the recognition and incorporation of the sense strand in the RISC instead of the antisense strand might lead also to off target effects.

It is therefore the object of the present invention to provide novel and improved siRNA molecules that are protected against degradation so that the gene silencing activity of the siRNA molecules is improved and the amount of siRNA molecules needed for efficient gene silencing is decreased.

This object is solved by the siRNA molecules according to independent claim 1, the method according to independent claim 19, the use according to independent claim 20, the support material according to independent claim 21, the compound according to independent claim 29, and the phosphoramidite according to independent claim 33. Further advantageous aspects, details and features of the invention are evident from the dependent claims, the description, the examples and the figures.

The chemically modified siRNA molecules according to the present invention enhance the stability of these molecules in culture media and body fluids, e.g. serum. The degradation of siRNA is effectively suppressed, and the full length siRNA will be more active and lead to no off target effects. Moreover, the sensitivity of the siRNA according to the present invention is enhanced. Lower amount of siRNA will be needed to achieve a high level of gene silencing. The siRNA may be incorporated into the RISC in the right orientation, so that only the antisense strand will be able to induce silencing, which leads to higher specificity. In addition, the sense strand is inactivated due to modifications, resulting in the elimination of off target effects. The siRNA according to the present invention enables a more efficient unwinding of the siRNA duplex and hybridization with the target mRNA, leading to an improved target binding affinity. Furthermore, the transfection efficiency of the siRNAs is enhanced and the biodistribution and pharmacokinetic properties are improved.

Chemically modified siRNA molecules with the combination of the following groups and linkages at the well defined positions of both the sense and antisense strands of the siRNA have not been described previously to simultaneously improve stability, sensitivity and specificity of the siRNA: 2'-deoxy modified nucleotides;

2'-methoxy modified nucleotides;

two nucleosides linked by a 3' to 5 ' or to 5' formacetal linkage;

nucleotides modified at the 2'-position by a -0-CH ₂ -0-(CH ₂ ) ₂ -OH group; and nucleotides comprising in the 3'-position a -0-CH ₂ -0-(CH ₂ ) ₇ -CH ₃ group.

It is believed that the modified siRNA molecules according to the present invention outperform respective non-modified siRNA molecules with respect to the down-regulation of the expression of a target gene because the modification(s) positively influence the thermostability of the double-stranded (i.e. duplex) RNA. Furthermore, it is assumed that the modification(s) allow for a more efficient opening of the duplex within the RNA-induced silencing complex (RISC), so that the further steps (recognizing of an mRNA sequence, binding to that sequence, and finally cutting same) can be carried out more quickly and with higher efficiency compared to non-modified siRNA.

According to one aspect, the present invention refers to a double-stranded siRNA molecule comprising a sense strand and an antisense strand, whereby the antisense strand is essentially complementary to the sense strand. Each of the sense and the antisense strands comprises at least 17 nucleotides (nt). The siRNA further comprises at least one overhang at the 5' and/or 3' end, wherein the overhang residue or the overhang residues are chemically modified and selected independently from each other from the group consisting of:

(a) 2'-deoxy modified nucleotides;

(b) 2'-methoxy modified nucleotides;

(c) two nucleosides linked by a 3' to 5' or 2' to 5' formacetal linkage;

(d) nucleotides modified at the 2 -position by a -0-CH ₂ -0-(CH ₂ ) ₂ -OH group; and

(e) nucleotides comprising in the 3'-position a -0-CH ₂ -(CH ₂ ) ₇ -CH3 group.

"Independently of each other" means that the overhang can be modified by (a), (b), (c), (d) or (e) alone or by any combination of modifications (a) to (e). Modifications comprising two of the above-mentioned alternatives are e.g. (a) and (b), (a) and (c), (a) and (d), (a) and (e); (b) and (c), (b) and (d), (b) and (e); (c) and (d), (c) and (e); (d) and (e). Modifications comprising three of the above-mentioned alternatives are e.g. (a), (b) and (c), (a), (b) and (d), (a), (b) and (e), (a), (c) and (d), (a), (c) and (e), (a), (d) and (e); (b), (c) and (d), (b), (c) and (e); (c), (d) and (e). Modifications comprising four of the above-mentioned alternatives are e.g. (a), (b), (c) and (d), (a), (b), (c), and (e), (a), (c), (d) and (e); and (b), (c), (d) and (e). The modification comprising five of the above-mentioned alternatives is (a), (b), (c), (d) and (e).

"Independently of each other" further means that the 5' end and/or the 3' end of the sense strand and/or the antisense strand of the siRNA molecule can be modified by any of the aforementioned combinations.

If above and in the following the term "siRNA" is used, this does not mean that the siRNA molecules do contain RNA components exclusively. The siRNA according to the present invention may also contain other components, e.g. modified and/or unmodified DNA moieties as well as modified RNA moieties. More generally, "siRNA" as used herein refers to small interfering nucleic acids (siNA).

The sense and the antisense strands of the siRNA molecules according to the present invention preferably comprise between 15 and 29 nt, more preferably between 17 and 23 nt, and particularly 19 to 21 nt. These figures refer to the double-stranded part of the siRNA molecule. To this double stranded part there is added a so-called "overhang", which means one or more unpaired bases at the 3' and/or 5' end of the sense and/or antisense strands of the siRNA molecule of the present invention. The double-stranded part of the siRNA is also referred to as the "core" region of the siRNA.

According to one specific embodiment, the siRNA molecule of the present invention comprises in the sense strand, particularly at the 5' end of the sense strand, at least one modification (a), (b), or (c).

According to a further specific embodiment, the siRNA molecule according to the present invention comprises in the antisense strand, particularly at the 5' end of the antisense strand, at least one modification (a), (b), or (c). According to a further preferred embodiment, the siRNA according to the present invention has no overhang at the 5' end of the antisense strand and/or the 5' end of the sense strand.

According to a still further embodiment, the siRNA according to the present invention has an overhang at the 3' end of the antisense strand and/or the 3' end of the sense strand. If the siRNA accordmg to the present invention does comprise an overhang at the 3' end of the antisense strand, then the overhang residue or overhang residues comprise preferably at least one modification selected from (a) to (e). Moreover, the siRNA according to the present invention preferably comprises at the 3' end of the sense strand an overhang, the overhang comprising at least one modification (c).

A further preferred embodiment of the siRNA molecule according to the present invention comprises in at least one of the core nucleotides of the sense strand and/or the antisense strand a modification (a), (b) or (d).

Preferably, the overhang of the sense strand and/or the antisense strand of the siRNA molecule according to the present invention comprises two modified nucleosides. According to a still further preferred embodiment, the sense strand of the siRNA molecule according to the present invention has an overhang of two modified nucleotides at the 3' end, said modifications being (c), and the 3' end of the core of the sense strand has at least one modification (a) or (b), and the modification of the overhang at the 5' end of the sense strand is selected from (a) or (b), and the 5' core end of the sense strand has no modification, and the antisense strand has an overhang of two modified nucleotides at the 3' end, the modification being selected from (b), (c), or a combination of (b) and (e), and the antisense strand has no overhang at the 5' end, and the antisense strand has no modification at the 3' end of the core, and the core of the 5' end of the antisense strand comprises a modification (a) and/or (b), preferably (a).

If the siRNA molecule according to the present invention comprises the modification (c), then the nucleobase is preferably selected from the group consisting of adenine, cytosine, guanine, myrnine, and uracil, preferably uracil. If the siRNA molecule according to the present invention comprises the modification (d), then the nucleobase is preferably selected from the group consisting of adenine, cytosine, guanine, and uracil. In a further preferred embodiment of the siRNA molecule according to the present invention the overhang residue according to modification (c) is one according to formula Va or Vb

FORMULA Vb, wherein B is a nucleobase preferably selected from the group consisting of adenine, cytosine, guanine, thymine or uracil and derivatives thereof, and preferably is uracil or a derivative thereof. R ⁸ is selected from the group consisting of H, OH, OCH ₃ , or an alkyloxy group, wherein the alkyl chain in the alkyloxy group may contain one or more heteroatoms or heteroatom groups, said heteroatoms or heteroatom group preferably being selected from the group consisting of O, S, or NH, and preferably is O. R ⁹ is selected from the group consisting of H, OH, OCH ₃ , or an alkyloxy group, wherein the alkyl chain in the alkyloxy group may contain one or more heteroatoms or heteroatom groups, said heteroatoms or heteroatom groups preferably being selected from the group consisting of O, S, or NH, and preferably is O. If in the compounds according to formulas Va and Vb R ⁸ and R ⁹ is H, only one of the groups R ⁸ and R ⁹ is H, thereby forming a 2'- or 3'- deoxy-modified nucleoside. In a further embodiment either R ⁸ or R ⁹ is OCH ₃ and R ⁹ or R ⁸ , respectively, is OH, thereby forming a 2'- or 3'- methoxy modified nucleoside. If R and/or R is an alkoxy group, the allcoxy group can be one having the following chemical structure: 0-(CH ₂ ) _r -(Y-CH ₂ ) _s -CH ₃ , wherein Y is a heteroatom or heteroatom groups selected from O, S, and NH. Preferably, Y is 0. r is an integer in the range from 1 to 12, preferably 1 to 5, most preferably 1 or 2, s is an integer in the range from 0 to 10, preferably 0 to 5 and most preferably is 0.

In a preferred embodiment of the siRNA molecule according to the present invention, R ⁸ is selected from H, OH, and OCH ₃ , and preferably is OH, and R ⁹ is selected from H, OH and OCH ₃ , and preferably is OH.

In yet a further preferred embodiment of the siRNA molecule according to the invention the core nucleotide comprising the modification (d) is a nucleotide having a chemical structure according to formula VI

FORMULA VI wherein B is a nucleobase or derivative thereof, preferably selected from the group consisting of adenine, cytosine, myrnine, guanine, and uracil. In yet another embodiment of the siRNA molecule according to the present invention, the overhang residue according to modification (e) is one according to formula Vila or Vllb

FORMULA Vila,

FORMULA Vllb, wherein B is a nucleobase preferably selected from the group consisting of adenine, cytosine, guanine, thymine or uracil and derivatives thereof, and preferably is uracil or a derivative thereof, R ¹¹ is selected from the group consisting of H, CH ₃ or OR ¹¹ forms an alkyloxy group, wherein the alkyl chain in the alkyloxy group may contain one or more heteroatoms or heteroatom groups, said heteroatoms or heteroatom groups preferably being selected from the group consisting of O, S, or NH, and preferably is O, and R ¹¹ preferably is H. The alkyloxy group can be of the same nature as outlined above with reference to Formulas Va and Vb.

If the siRNA molecule according to the present invention comprises one the modification (e), then the nucleobase is preferably uracil.

The modified siRNAs according to the present invention are stabilized by means of end- modifications of backbone phosphates or sugars. This requires one minimum changes in the core sequence of the sense strand of the siRNA molecule, and as a consequence, there are even less changes in the core of the antisense strand, which avoids inactivation of the antisense strand and avoids interference with the RNAi mechanism, and additional off target effects. The modification on the sense strand inactivates this strand, and therefore there are less off target effects, which are normally caused by the sense strand's "false" activity. According to the present invention, siRNA was chemically modified using different types of modifications. Particularly, the ends (overhang(s) and/or core end(s)) of the siRNA of the present invention were modified. The sense strand of the siRNA carries modifications that stabilize the siRNA against nuclease cleavage and simultaneously inactivate it, in order to minimize off target effects. The antisense strand is also modified for enhanced serum stability, but to a degree, which does not affect the siRNA activity. In addition, the combination of these modifications on sense and antisense strands enhance the activity of the siRNA in the tested cases, probably by partially changing the thermodynamic stability of duplex.

With the modifications according to the present invention, siRNAs were obtained which are more stable against serum nucleases and more efficient in inducing gene silencing at the same time. The longevity of these siRNA is also enhanced by these modifications. Moreover, the modifications on the sense strand lead to inactivation of this strand, following minimization of the induced off target effects. Due to higher effectiveness of these new modified siRNA, less siRNA is necessary for efficient gene silencing, and this minimizes also off target and toxic effects in the cells.

In a further aspect the present invention provides a solid support material suitable for the automated synthesis of nucleic acids, preferably oligoribonucleotides, comprising:

a) a solid phase material,

b) a chemical group selected from the groups according to Formulas la, lb, Villa or Vlllb which is bonded to the solid phase material via a linker group L:

FORMULA la,

FORMULA Vinb, and wherein B is a nucleobase group preferably independently selected from the group consisting of adenine, cytosine, thymine, guanine and uracil or derivatives thereof. Nucleobase derivatives are here understood to include especially derivatives in which the nucleophilic NH ₂ groups of adenine, guanine, thymine and cytosine are masked, typically by acylation. For the synthesis of oligoribonucleotides adenine, cytosine, guanine and uracil or derivatives thereof are of course preferred. In formulas la, lb, Villa and Vlllb R ² and R ¹⁰ , respectively, are a protective group selected from the group consisting of 4,4'- dimethoxytrityl, 4-methoxytrityl or silyl. R ⁸ is a group selected from H, OH, OCH ₃ , or an alkyloxy group, wherein the alkyl chain in the alkyloxy group may contain one or more heteroatoms or heteroatom groups, said heteroatoms or heteroatom groups preferably being selected from the group consisting of O, S, or NH, and preferably is O. Regarding the nature of the alkyloxy group, the same applies as outlined above with reference to the alkyloxy group in formulas Va and Vb. Alternatively R ⁸ is OR ³ , wherein R ³ is a protective group selected from the group consisting of acyl groups, silyl groups or trialkylsilyloxymethyl groups. R ⁹ is a group selected from H, OH, OCH ₃ , or an alkyloxy group, wherein the alkyl chain in the alkyloxy group may contain one or more heteroatoms or heteroatom groups, said heteroatoms or heteroatom groups preferably being selected from the group consisting of O, S, or NH, and preferably is O. Again, regarding the nature of the alkyloxy group, the same applies as outlined above with reference to the alkoxy group in formulas Va or Vb. Alternatively, R ⁹ is OR ¹ , wherein R ¹ is a protective group selected from the group consisting of acyl groups, silyl groups or trialkylsilyloxymethyl groups. In formulas la and lb, if R ⁸ or R ⁹ is H, only one of R ⁸ and R ⁹ is H. The use of the protective groups R ¹ and R ³ is of course preferred when the nucleosides incorporated into an oligonucleotide are intended to be 2'- or 3- hdroxyl substituted nucleosides. Accordingly, in a preferred embodiment of the support material of the present invention R ⁸ and R ⁹ are OR ¹ and OR ³ , wherein R ¹ and R ³ are protective groups selected from the group consisting of acyl groups, silyl groups or trialkylsilyloxymethyl groups. Suitable acyl groups for use as R ¹ and/or R ³ are for instance acetyl groups, benzoyl groups etc. Suitable silyl groups for use as R ¹ and R ³ include triethylsilyl or dimethyltertbutylsilyl groups, etc. In principle, all protective groups can be used, which are cleaved off during deprotection of the oligonucleotides with MeNH ₂ or NH ₃ . The nature of the protective groups selected as R ² , R ¹ and R ³ may to some extend depend of the reactivity of each of the respective protective groups. If, for instance, R ² is selected to be a silyl group, then R ¹ and R ³ are preferably not a silyl group.

In a preferred embodiment of this aspect of the invention, the chemical group in the support material is one according to formulas la or lb, R ⁹ is OR ¹ wherein R ¹ is benzoyl (Bz), R ² is 4,4'-(dimethoxy)trityl (DMT), and R ⁸ is OR ³ wherein R ³ is (triisopropysilyl)oxymethyl (TOM). This combination of protective groups on the respective hydroxy groups is especially suitable for carrying out automated nucleic acid synthesis using standard equipment and procedures known in the art, for instance those described by S. Pitsch, P.A. Weiss, L. Jenny, A. State and X. Wu in Helv. Chim. Acta 2001, 84, 3773 - 3795. In a further embodiment of this aspect of the invention, the chemical group in the support material is one according to formulas Villa or Vlllb, and R ¹⁰ is 4,4'(dimethoxy)trityl.

In a preferred embodiment of this aspect of the present invention, the nucleobase group B in any one of the chemical groups according to formulas la, lb, Villa or Vlllb is uracil. The selection of the respective nucleobase group will of course depend on the nucleotide which is to be introduced into the nucleic acid, especially the oligoribomicleotide.

In a further preferred embodiment the solid phase material in the support material according to the invention is controlled pore glass (CPG), preferably having a pore size in the range between 100 to 1500 A, most preferably from 500 to 1000 A. Preferably, the CPG-based support material is functionalized with aminoalkyl groups, which can form a chemical bond, for instance an amide bond, with the linker group L in the chemical groups contained in Formulas la and lb. Other suitable solid phase materials are polystyrene and surfaces.

In yet a further embodiment of the support materia] according to the present invention, the linker group L is a - 0(θΧ<¾) _η -(Ι!(0)0^ουρ, wherein n is an integer from 1 to 10, preferably from 2 to 5, and most preferably is 2 or 5. The OL-moiety bonded to the polysaccharide moiety therefore preferably forms a dicarboxylate group, preferably a succinate group,(-0(0)CCH ₂ CH ₂ C(0)0-) or diheptandioate group, (- 0(0)C(CH ₂ ) ₅ C(0)0-), and most preferably a diheptandioate group (-0(0)C(CH ₂ ) ₅ C(0)0-). A preferred loading of the support material with compounds according to Formulas la lb, Villa or Vlllb on the material is in the range of 10 to 50 μηιοΐ/g, preferably in the range of 30 to 40 μηιοΐ/g.

The support materials according to the present invention provide a means for introducing specific modifications at the 3 '- or 5'- ends of a respective nucleic acid. In the synthesis of the nucleic acid, which is preferably automated, appropriately activated and protected nucleotide building blocks are added stepwise to the chemical groups attached to the solid phase until the desired sequence has been obtained. By using the solid support of the present invention, it is possible to prepare nucleic acids, especially oligoribonucleotides, having modified end groups derived from the chemical groups according to Formulas la lb, Villa or Vlllb comprised in the solid support material, i.e. either 2'-5'-formacetal linked nucleosides, 3'-5'-formacetal linked nucleosides or nucleosides having a 2Ό- or 3'O-octyloxymethyl modified nucleosides. Preferably, the nucleobase in these nucleosides is uracil. In a further aspect, the present invention provides compounds according to Formulas Ila, lib, IXa or IXb as shown below. These compounds can be used for the synthesis of the solid support materials discussed above. In the compounds according to Formulas Ila, lib, IXa or IXb

FORMULA lib,

FORMULA Ixa

FORMULA Kb

In formulas Ila. lib, IXa and LXb, B is a nucleobase group or derivative thereof, preferably independently selected from the group consisting of adenine, cytosine, thymine, guanine and uracil or derivatives thereof, wherein for the synthesis of oligoribomicleotides adenine, cytosine, guanine and uracil or derivatives thereof are preferred. Uracil or derivatives thereof are especially preferred.

R ² is a protective group selected from the group consisting of 4,4'-(dimethoxy)trityl, 4- (methoxy)trityl or silyl. R ⁸ is a group selected from H, OH, OC¾, or an alkyloxy group, wherein the alkyl chain in the alkyloxy group may contain one or more heteroatoms or heteroatom groups, said heteroatoms or heteroatom groups preferably being selected from the group consisting of O, S, or NH, and preferably is O. Regarding the nature of the alkyloxy group, the same applies as outlined with reference to the alkyloxy group in formulas Va and Vb. Alternatively, R ⁸ is OR ³ , wherein R ³ is a protective group selected from the group consisting of acyl groups, silyl groups or trialkylsilyloxymethyl groups. R ⁹ is a group selected from H, OH, OCH3, or an alkyloxy group, wherein the alkyl chain in the alkyloxy group may contain one or more heteroatoms or heteroatom groups, said heteroatoms or heteroatom groups preferably being selected from the group consisting of O, S, or NH, and preferably is O. Again, regarding the nature of the alkyloxy group, the same applies as outlined above with reference to the alkyloxy group in formulas Va and Vb. Alternatively, R ⁹ is OR ¹ , wherein R ¹ is a protective group selected from the group consisting of acyl groups, silyl groups or triallcylsilyloxymethyl groups. The provisio applies that if R ⁸ or R ⁹ is H, only one of R ⁸ and R ⁹ is H

Regarding the selection and nature of the groups R ² , R ⁸ , R ⁹ , R ¹ and R ³ , the same applies as outlined with reference to the support materials according to formulas la and lb.

R ⁴ is selected from the group consisting of H and -C(0)(CH ₂ ) _n C(0)0-R ⁵ , wherein R ⁵ is a group selected nitrophenyl, tetrafluorophenyl, pentafluorophenyl and H, and n is an integer in the range from 1 to 10. n is preferably an integer in the range form 2 to 5 and preferably is 2 or 5. To attach these compounds to the solid phase material, a chemical bond must be provided between the solid phase material and the chemical compounds according to Formulas Ila, lib, IXa or IXb. This bond is achieved either by first attaching a linker group at the 2'-0 or 3'-0 position on the nucleoside carrying the OR ⁴ group, if R ⁴ is -H, and then bonding the linker to the solid phase material which preferable comprises functional groups capable of reacting with the linker group. In a preferred embodiment of these compounds according to the invention, the compound is one according to formula Ila or lib, B is uracil, R ⁹ is OR ¹ wherein R ¹ is benzoyl (Bz), R ² is 4,4'-(dimethoxy)trityl (DMT), R ⁸ is OR ³ wherein R ³ is (triisopropysilyl)oxymethyl (TOM), R ⁴ is -C(0)(CH ₂ ) _n C(0)0-R ⁵ and n is an integer in the range from 2 to 5 and preferably is 2 or 5, wherein R ⁵ is 4-nitrophenyl. In this case, the compounds already comprise the linker group. In an other preferred embodiment, the compound is one according to formula Ila or lib, B is uracil, R ⁹ is OR ¹ wherein R ¹ is benzoyl (Bz), R ² is 4,4'-(dimethoxy)trityl (DMT), R ⁸ is OR ³ , wherein R ³ is (triisopropysilyl)oxymethyl (TOM), and R ⁴ is H. In this case, a linker group is preferably attached to the compound, prior to attaching it to the solid phase material. In a further preferred embodiment of the compounds according to the present invention, the compound is one according to formula IXa or Ixb and B is preferably selected from the group consisting of adenine, cytosine thymine, guanine and uracil, preferably from adenine, cytosine, guanine and uracil, and most preferably B is uracil. In this embodiment, R ² preferably is 4,4'-(Dimethoxy)trityl (DMT), and R4 is preferably selected from H or - C(0)(CH ₂ ) _n C(0)0-R ⁵ , wherein n is an integer in the range from 2 to 5 and preferably is 2 or 5, and wherein R ⁵ is 4-nitrophenyl.

In a further aspect the present invention provides specially modified phosporamidites which can be used for the syntheses of nucleic acids. These phosphoramidites enable the introduction of specially modified nucleosides into nucleic acids, especially into oligoribonucleotides. These phosporamidites are represented by the general formula III below.

FORMULA III

In formula III, B is a nucleobase or derivative thereof, which is preferably selected from the group consisting of adenine, cytosine, thymine, guanine and uracil, wherein adenine, cytosine, guanine and uracil or derivatives thereof are preferred. R ² is a protective group, preferably R ² is a 4,4'-(dimethoxy)trityl group, 4-(methoxy)trityl group or a silyl group. This protective group is selected in such a manner that it can be activated in respective nucleic acid syntheses to enable formation of a 5 '-3 '-linkage to a further nucleotide. R ⁶ is an acyl group, silyl group or trialkylsilyloxymethyl group. As discussed before, suitable acyl groups include acetyl and benzoyl groups, etc. Suitable silyl groups include triethylsilyl or dimethyltertbutylsilyl groups, etc. In preferred embodiments of the phosporamidites according to the invention, B is selected from the group consisting of N ⁶ -Acetlyadenine, N ⁴ -acetylcytosine, N ² -acetylguanine and uracil. R ² preferably is 4,4'-(dimethoxy)trityl. Preferably, R ⁵ is a benzoyl group.

The phosphoramidites according to the present invention can be used in the synthesis of nucleic acids, such as oligoribonucleotides, preferably in automated synthesis using standard procedures and equipment. The phosphoramidites can either be used with conventional solid support materials or with the specially modified solid support material according to the present invention. The phosporamidites can for instance be used in combination with 2'-0- tom-protected ribonucleoside phosporamidites as described by S. Pitsch, P. A. Weiss, L. Jenny, A. Stutz and X. Wu in Helv. Chim. Acta 2001, 84, 3773 - 3795. They allow the introduction of 2'-acetal modified ribonucleotides into oligoribonucleotides. General

Reagents and solvents (highest purity) were from various suppliers and used without further purification, unless otherwise stated. 2'-<9-tom-protected ribonucleoside phosphoramidites were prepared according to S. Pitsch, P. A. Weiss, L. Jenny, A. Stutz, X. Wu, Helv. Chim. Acta 2001, 84, 3773. Work-up implies partitioning of the reaction mixture between CH ₂ C1 ₂ and saturated aqueous aHCOs solution, drying of the organic layer (MgSC>4), and evaporation. TLC: pre-coated silica-gel plates from Merck, stained by dipping into a solution of anisaldehyde (10 ml), H ₂ S0 ₄ (10 ml), and AcOH (20 ml) in EtOH (180 ml), and subsequent heating with a heat gun. CC (column chromatography): silica gel 60 (230-400 mesh) from F!uka. Anion-exchange (AE) HPLC: DNAPAC PA-10 (9.0 X 250 mm; Dionex), flow 2.5 ml/min; eluant ^: 12 raM Tris. HC1 (pH 7.4), 6 M urea; eluant B: 12 mM Tris. HC1 (pH 7.4), 0.5 M NaC10 ₄ , 6 M urea; detection at 260 nm, elution at 85°C. NMR (Bruker instrument): chemical shift δ in ppm relative to external standards (1H and ¹³ C: Me ₄ Si); coupling constants J in Hz. MALDI- MS (linear negative mode): Axima CFR Plus (Kratos/Shimadzu); matrix: 2,4,6-trihydroxyacetophenone, diammonium citrate; m/z (rel. intensity in %). Acrylamide solution (30%): 300 g acrylamide, 8 g N sF- methylendiacrylamide; final volume: 1 1 H ₂ 0. 10X TBE (Tris acid Boric EDTA): 54 g Tris base, 27.5 g Boric acid, 20 ml 0.5 M EDTA pH 8, final volume 500 ml (complete with water). Native 1 TBE 12 % polyacrylamide gel: 6ml Acrylamid solution (30 %), 3 ml 5X TBE, 5 ml H ₂ 0, 150 μ APS and 20 μΐ TEMED.

Multistep-Preparation of support materials

Preferred embodiments of the support material according to the present invention can be prepared according to the synthetic route outlined below in reaction scheme 1.

Reaction scheme 1 : Multi-step preparation of support materials

According to this reaction scheme in a first synthetic step (1→2) the 5'-0-, and 3'-0- protected uridine is deprotected in 5'-0-postion and the previously non-protected 2'-0- position is protected with a suitable protecting group, which can withstand the reaction conditions required for the subsequent reaction steps. In the second synthetic step (2→3) a suitable reactive group for forming an formacetal linkage is introduced at the 5'-0-position of compound 2. This group is preferably a MeSCH ₂ -group. Functionalized compound 3 is subsequently reacted with the 5'-0-protected uridine 4 in step 3 (3 + 4→5 and 6), forming the 3 '-5'- or 2 '-5' formacetal linked compounds 5 and 6. In subsequent synthetic steps, 4a (5→7) and 4b(6→9), preferably carried out after separation of the regioisomers 5 and 6 a linker group is bonded to the non-protected 2'-OH or 3'-OH group of compounds 7 or 9, respectively. Finally, in synthetic steps 5a (7→8)or 5b (9→10), the compounds are bonded to the solid phase material via the linker group.

A preferred embodiment of this general reaction sequence in outlined in the following specific preparation steps.

Synthesis of 2

To a solution of 1 (5.2 g, 7.11 mmol, prepared according to S. Pitsch, P. A. Weiss, L. Jenny, A. Stutz, X. Wu, ffelv. Chim. Acta, 2001, 84, 3773) in pyridine (45 ml) was added benzoyl chloride (989 μΐ, 8.53 mmol). The resulting reaction mixture was stirred at room temperature overnight. After work-up, the crude product was dissolved in CH ₂ C1 ₂ (50 ml) and dic oroacetic acid (5 ml) was added. 5 min later the reaction was complete; 3 ml methanol were then added. After work-up and recristallisation in hexane/CH ₂ Cl ₂ 1 :1, a colorless solid 2 (2.95 g 78 %) was obtained. TLC (AcOEt/hexane 1 :1): R/0.35. 1H-NMR (400 MHz, CDC1 ₃ ): 1.02-1.04 (m, 'PrsSi); 3.94 (dd, J = 2.3, 11.3, H-C(5')); 4.05 (dd, J = 2.3, 1 1.3, H-C(5*)); 4.37 (m, H-C(4')); 4.72 (t, J = 5.4 , H-C(2')); 4.91, 5.08 (2d, J= 5.0. OCH ₂ 0); 5.57 (t, J = 5.2, H- C(3')); 5.78 (d, J= 8.1, H-C(5)); 6.07 ( d, J= 5.1 , H-C(l')); 7.45-7.70 (m, 5 arom. H); 8.08 (d, J= 8.3, H-C(6)); 8.26 (br. s, H-N(3)). Synthesis of 3

2 (2.95 g, 5.52 mmol) was dissolved in a 1 :1 :1 solution of DMSO (7 ml)/Ac ₂ 0 (7 ml)/AcOH (7 ml) and the reaction mixture was stirred 3 days at room temperature. After work-up with hexane/AcOEt 1 :1 as organic layer and CC (50 g of Si0 ₂ , AcOEt hexane 1 :9 → AcOEt/hexane 6:4), a colorless solid 3 (1.85 g, 56 %) was obtained. TLC (AcOEt/hexane 1 :1): R/0.55. 1H-NMR (400 MHz, CDC1 ₃ ): 1.02-1.04 (m, T^Si); 2.24 (s, SCH ₃ ); 3.82 (dd, J = 2.5, 10.7, H-C(5')); 3.91 (dd, J = 2.2, 10.7, H-C(5')); 4.50 (m, H-C(4')); 4.60 (dd, J = 3.7, 5.2, H-C(2')); 4.76 (d, J = 5.5, 2H, OCH ₂ S); 4.87, 5.01 (2d, J = 5.0, OCH ₂ 0); 5.38 (/, J= 5.6 H-C(3')); 5.85 (d, J= 8.2, H-C(5)); 6.39 (d, J= 6.0, H-C(l')); 7.44-7.57 (m, 2 arom. H); 7.60 (m, 1 arom. H); 7.81 (d, J= 8.2, H-C(6)); 8.06 (m, 2 arom. H); 8.48 (br. s, H-N(3)).

Synthesis of 5 and 6

Step J: To a solution of 3 (1.62 g, 2.72 mmol) and cyclohexene (5.44 mmol, 551 μΐ) in CH ₂ C1 ₂ (11 ml) was added at -80 °C S0 ₂ C1 ₂ (2.99 mmol, 242 μΐ). After stirring 35 min at -80°C and then 40 min at room temperature, the reaction mixture was evaporated under reduced pressure and dried 10 min at high vacuum.

Step2: To a solution of 4 (1.76 g, 3.264 mmol) in acetonitrile (13 ml) was added di-tert- butyltin dichloride (1.09g, 3.59 mmol) and diisopropylethylamine (2.23 ml, 13 mmol). After stirring 15 min at room temperature, the crude product fom step 1, dissolved in acetonitrile (5 ml), was added to this mixture and stirred 8 h at room temperature. After workup, using AcOEt as organic layer, the two diastereoisomers were separated by 3 CC (50 g of Si0 ₂ , AcOEt/hexane 5:5 ^→ AcOEt 100 %, two time to column volume per gradient). Combined yield: 41 %. (6/5: 7/3) Data of 5: TLC (AcOEt/hexane 95/5): 2?/ 0.80. 1H-NMR (400 MHz, CDC1 ₃ ): 0.95-1.10 (m, 'PraSi ; 3.57 (dd, J = 2.3, 11.3, H-C(5')) 3.63 (dd, J = 2.2, 11.2, H-C(5')); 3.80 (s, 2 MeO); 3.93-3.99 (m, 2 H-C(5')); 4.10 (m, H-C(2')); 4.50 (m, H-C(2')); 4.49-4.65 (m, H-C(4'), H- C(3')); 4.85, 4.98 {2d, J= 5.1, OCH ₂ 0); 4.88, 5.37 (2d, J= 6.8, OCH ₂ 0); 5.35 (d, J = 6.5, H- C(5)); 5.54 (d, J = 6.9, H-C(5)); 5.94 (d, J = 7.2, H-C(l ^* )); 6.52 (d, J = 7.0, H-C(l')); 6.84- 6.90 (m, 4 arom. H); 7.26-7.40 (m, 12 arom. H); 7.87 (d, J = 8.2, H-C(6)); 7.96-7.99 (m, 2 arom. H); 8.14 (d, J= 8.2, H-C(6)); 10.21 (br. s, H-N(3));10.75 (br. s, H-N(3)). MS: fehlt Data of 6: TLC (AcOEt/hexane 95/5): i?/0.85 (2' isomer). Ή-NMR (400 MHz, CDC1 ₃ ): 0.95-

I .10 (m, ^! Pr ₃ Si); 3.45 (dd, J= 2.3, 11.3, H-C(5')); 3.58-3.67 (m, 2 H-C(5')); 3.80 (s, 2 MeO); 3.87 (dd, J= 3.1, 11.0, H-C(5')); 4.14-4.19 (m, 2 H-C(4*)); 4.20-4.26 (m, 2 H-C(2')); 4.38 (m,

H-C(3')); 4.83 (m, H-C(3')); 4.85, 4.89 (2d, J = 6.8, OCH ₂ 0); 4.91, 4.95 (2d, J = 5.1, 0C¾0); 5.35 (d, J= 8.0, H-C(5)); 5.58 (d, J- 2.5 , H-C(l')); 5.73 (d, J = 7.9, H-C(5)); 5.83 (d, J= 2.4 , H-C(l')); 6.84-6.90 (m, 4 arom. H); 7.26-7.40 (m, 12 arom. H); -8.08, 7.62 (2m, 2 arom. H); 8.02 (d, J= 8.2, H-C(6)); 8.06 (d, J= 8.3, H-C(6)); 9.35 (br. s, H-N(3)); 9.55 (br. s, H-N(3)).

Synthesis of 7 and 9

Pimelic acid-dinitrophenylester (893 mg, 2.22 mmol) was dissolved in pyridine (3.7 ml). DMAP (22.6 mg, 0.18 mmol) and 5 or 6 (400 mg each, 0.37 mmol) were added. After stirring over night at room temperature, the mixture was evaporated and co-evaporated twice with toluene (20 ml). CC (10 g of Si0 ₂ , AcOEt/hexane 3:7→ AcOEt/hexane 6:4) gave 7 (243 mg, 46%) and 9 (250 mg, 48%), respectively. Colorless foams.

Data of 7: TLC (AcOEt/hexane 7:3): R 0.72. Ή-NMR (400 MHz, CDC1 ₃ ): 0.95-1.05 (m, 'PrsSi); 1.47 (m, CH ₂ ); 1.72 (m, CH ₂ ); 1.79 (m. CH ₂ ); 2.42 (t, J = 7.4, CH ₂ ); 2.62 (t, J = 7.3, CH ₂ ); 3.50 (dd, J = 2.3, 10.7, H-C(5')); 3.58 (dd, J ~ 2.2, 10.6, H-C(5')); 3.76 (dd, J = 3.3,

I I.2, H-C(5')); 3.80 (s, 2 MeO); 3.85 (dd, J= 2.4, 11.1, H-C(5')); 4.27 (m, H-C(4')); 4.32 (m, H-C(4')); 4.49 (t, J = 5.4, H-C(3')); 4.53 (t, J= 4.7, H-C(3')); 4.76, 4.78 (2d, J= 6.9, OCH ₂ 0); 4.80, 4.92 (2d, J = 5.0, OCH ₂ 0); 5.31,5.36 (2d, J = 7.4, 2 H-C(5)); 5.46 (r, J= 5.1, H-C(2')i 5.67 (dd, J= 2.1, 8.0, H-C(2'); 6.06 (d, J= 5.1, H-C(l')); 6.14 (d, J= 5.1, H-C(l')); 6.83-6.89 (m, 4 arom. H); 7.26-7.45 (m, 16 arom. H); 7.75 (d, J = 8.1, H-C(6)); 8.05-8.07 (m, 2 arom. H); 8.26 (d, J= 8.3, H-C(6)).

Data of 9: TLC (AcOEt/hexane 7:3): R 0.72. 'H-NMR (400 MHz, CDC1 ₃ ): 0.95-1.05 (m, ^j Pr ₃ Si); 1.46 (m, CH ₂ ); 1.70 (m, CH ₂ ); 1.77 (m. CH ₂ ); 2.42 (t, J = 7.6, CH ₂ ); 2.62 (t, J = 7.5, CH ₂ ); 3.45 (dd, J = 2.3, 1 1.1, H-C(5')); 3.64 (dd, J = 2.3, 11.2, H-C(5')); 3.73 (dd, J = 3.5, 10.8, H-C(5')); 3.80 ( , 2 MeO); 3.95 (dd, J = 2.9, 10.8, H-C(5')); 4.30 (m, H-C(4')); 4.40 (m, H-C(4')); 4.63 (t, J = 4.8, H-C(3')); 4.57 (dd, J = 5.1 , 5.3, H-C(3')); 4.83, 5.02 (2d, J = 6.8, OCH ₂ 0); 4.90, 4.97 (2d, J = 5.0, OC¾0); 5.32,5.35 (2d, J= 7.4 2 H-C(5)); 5.48 (i, J = 5.4, H-C(2'); 5.72 (dd, J = 2.3, 8.2, H-C(2'); 6.10 (d, J= 53, H-C(l')); 6.24 (d, J = 4.3, H-C(l')); 6.84-6.90 (m, 4 arom. H); 7.26-7.40 (TW, 16 arom. H); 7.90 (d, J= 8.20, H-C(6)); 8.05-8.07 (m, 2 arom. H); 8.26 (d, J= 8.3, H-C(6)).

Synthesis of solid supports 8 and 10

To a solution of the active ester 7 or 9 (0.13 mmol each) in DMF (4.16 ml) was added long- chain-alkylamino CPG (1.04 g) and then 'Pr ₂ NEt (0.4 ml). The mixture was shaken for 20 h at room temperature. After filtration, the solid support was washed with DMF and CH ₂ C1 ₂ , dried, suspended in pyridine (2.6 ml) and Ac ₂ 0 (1.56 ml) and shaken for 2 h at room temperature. After filtration, the solid was washed with DMF and CH ₂ C1 ₂ , and dried under high vacuum. Typical loadings of the solid support were 30 μιηοΐ/g with 500- A CPG.

Reaction scheme 1.1 : General synthesis of the 2 -5' coupled dimers

DMT =

TOM =

Bz = Be nzoyl

CPG = Controlled Pore Glass

Daa-dd Eaa-dd R = OC ₆ H ₄ NO ₂ Faa-dd R = NH-CPG

According to this reaction scheme in a first synthetic step (A,a-d-→B,a-d; a-d stands for the four compounds containing the nucleobase derivatives uracil (a). A^-acetylcytidine (b), N ⁶ - acetyladenine (c) and N ² -acetylguanine (d)) the 5'-0-. and 2 ^, -0-protected uridine is deprotected in 5'-0-postion and the previously non-protected 2'-0-position is protected with a suitable protecting group, which can withstand the reaction conditions required for the subsequent reaction steps, followed by the introduction of a suitable reactive group for forming an formacetal linkage at the 5'-0-position. This group is preferably a MeSCH - group. Functionalized compound B,a-d is subsequently reacted with the S'-O-protected nucleosides C,a-d in step 2 (B,a-d + C,a-d→D,a-d), forming preferentially the 2' -5' formacetal linked compounds D,aa-dd (D,aa-dd stands for the 16 possible combinations of dimicleotides, each containing two of the nucleobase derivatives uracil (a), Λ^-acetylcytidine (b), Λ^-acetyladenine (c) and jV ² -acetyl guanine (d); the first descriptor refers to the nucleotide Y at the 5 ^r -end (containing a 5'-0-DMT group) and the second to the nucleotide X at the 3'- end (containing a 3'-0-benzoyl group)). In the subsequent synthetic step 4 (D,aa-dd→E,aa- dd), carried out after isolation of the compound D,aa-dd. a linker group is bonded to the nonprotected 3'-OH group of compounds E,aa-dd. Finally, in synthetic step 5 (E,aa-dd→F,aa- dd). the compounds are bonded to the solid phase material via the linker group.

The reaction conditions for this general reaction sequence in outlined in the following specific preparation steps.

Synthesis of four compounds B,a-d

To a solution of A,a-d (20 mmol., prepared according to S. Pitsch. P. A. Weiss. L. Jenny. A. Stutz, X. Wu, He . Chan. Acta, 2001, 84, 3773) in pyridine (130 ml) was added benzoyl chloride (3 ml, 25 mmol). The resulting reaction mixture was stirred at room temperature overnight. After work-up, the crude product was dissolved in CH ₂ CI ₂ (150 ml) and dichloroacetic acid (15 ml) was added. 5 min later the reaction was complete; 15 ml methanol were then added. After work-up. the crude products were were dissolved in a 1:1 : 1 solution of DMSO (30 ml)/Ac ₂ 0 (30 ml)/AcOH (30 ml) and the reaction mixture was stirred 3 days at room temperature. After work-up with hexane/AcOEt 1 :1 as organic layer and CC (150 g of Si0 ₂ ), compounds B,a-d were obtained as colorless solids

Data of B,a (uracil): Yield 75%. TLC (AcOEt/hexane 1 :1): Rf 0.59. Ή-NMR (400 MHz, CDCI ₃ ): 1.02-1.04 (m, ^rsSi) ; 2.22 (s, SCH ₃ ) 3.88-3.90 (m, H-C(5') ₅ H-C(5")); 4.48 (?», H- C(4 ^* )) ; 4.69 {dd, J - 5.2, 7.1, H-C(2') ; 4.76 (dd, J = 11.7, 18.1, OC¾S) ; 4.92 (m, OCH ₂ 0) : 5.56 (dd, J = 2.1, 5.2, H-C(.V)) ; 5.79 (del J = 2.1, 8.2, H-C(5)) ; 6.34 (d, J = 7.2, H-C(F)) ; 7.50 (m, 2 arom. H) ; 7.63 (;;;, 1 arom. H) ; 7.83 (d, J = 8.3. H-C(6» ; 8.12 (;;?. 2 arom. H) : 8.45 (br, s, H-N(3)).

Data of B,b (V-acetylcytosine): Yield 60%. TLC (CH ₂ Cl ₂ / eOH 19:1): i?/ ^' 0.62. Ή-NMR (400 MHz, CDC1 ₃ ): Ή-RMN (400 MHz, CDCI ₃ ) : 1.02-1 .04 (m. 'PraSi) ; 2.23 (s, SCH ₃ ) : 2.35 (s, N-CH ₃ ) ; 3.96 (m, H-(C-5')) ; 3.01 (m, H-C(5 ⁵ )) ; 4.57 (m, H-C(4')) ; 4.66 (r, J = 5.0, H-C(2')) ; 4.76 (dd, J = 7.5, 17.2, OC¾S) ; 5.05 (dd, J = 4.8, 8.7, OCH ₂ 0) ; 5.48 (t, J = 5.2, H-C(3 ^f )) ; 6.33 (d _: J = 5.2, H-C(l ')) ; 7.45 (»;, 3 arom. H) : 7.62 (w. H-C(5)) ; 8.12 (m. 2 arom. H) 8.32 (d, J = 7.5, H-C(6)).

Data of B,c (Λ^-acetylademne): Yield 55%. TLC (C¾Cl ₂ /MeOH 19:1): Rf 0.67. 'H-NMR (400 MHz, CDCI ₃ ): [WIRD NACHGELIEFERT]

Data of B,d (N ² -acetylguanine): Yield 25%. TLC (CH ₂ Cl ₂ /MeOH 19: 1): R/ 0.59. 'H- MR (400 MHz. CDC1 ₃ ): 1.03-1.05 (m, ^s Pr ₃ Si) ; 2.23 (s, SCH ₃ ) ; 2.32 (5, N-C¾) ; 3.85 (dd, J = 3.3, JO.7, H-(C-5 ⁵ )) ; 3.96 (dd, J = 3.6, 10.7. H-C(5')) ; 4.57 (m, H-C(4')) ; 4.76 (d, J = 4.5, OC¾S) ; 4.85, 4.90 (2d, J = 5.1, OCH ₂ 0) ; 5.17 (dd, J = 5.1, 6.9, H-C(2')) ; 5.74 (dd, J = 2.1, 5.1, H-C(3')) ; 6.10 (d, J = 6.9, H-C(l ')) : 7.51 (m, 2 arom. H) ; 7.65 (m, 1 arom. H) ; 8.02 (s, H-C(8)) ; 8.13 (/;?, 2 arom. H). Synthesis of sixteen compounds D,aa-dd

Stepl: To a solution of B,a-d (8 mmol) and cyclohexene (16 mmol, 1.65 ml) in CH ₂ C1 ₂ (33 ml) was added at -80 °C S0 ₂ C1 ₂ (8 mmol, 0.6 ml). After stirring 35 min. at -80°C and then 40 min at room temperature, the reaction mixture was evaporated under reduced pressure and dried 10 min at high vacuum.

Step2: To a solution of C,a-d (10 mmol) in 1 ,2-dichlorethane (39 ml) was added di-n-butyltin dicMoride (3.2 g. 10.8 mmol) and diisopropylethylamine (3.7 ml, 39 mmol). After stimng 15 min at room temperature, the crude product from step 1, dissolved in 1,2-dichlorethane (15 ml), was added to this mixture and stirred 8 h at room temperature. After workup, using AcOEt as organic layer, the products were isolated by CC (150 g of Si0 , AcOEt hexane or CH ₂ Cl ₂ /MeOH gradients). Data of compounds D,aa-D,dd

1) AcOEt/hexane 19:1 ; 2) CH ₂ Cl ₂ /MeOH 9:1

Synthesis of sixteen compounds E,aa-E,dd Pimelic acid-dinitrophenylester (2.7 g, 6.5 mmol) was dissolved in pyridine (10 ml). DMAP (70 mg, 0.55 mmol) and compounds D,aa-D,dd (1 mmol each) were added. After stirring over night at room temperature, the mixture was evaporated and co-evaporated twice with toluene (50 ml). CC (30 g of SiQ , AcOEt/hexane gradients) gave compounds E,aa-E,dd in yields between 65% and 75%. Colorless foams.

Synthesis of sixteen solid supports F,aa-F,dd

To a solution of the active esters E,aa-E,dd (0.4 mmol each) in DMF (12 ml) was added long-cham-alkylamino CPG (3 g. 500-A) and then ^! Pr ₂ NEt (1.2 ml). The mixture was shaken for 20 h at room temperature. After filtration, the solid support was washed with DMF and CH ₂ C1 ₂ , dried, suspended in pyridine (8 ml) and Ac ₂ 0 (4.5 ml) and shaken for 2 h at room temperature. After filtration, the solid was washed with DMF and CH ₂ C1 ₂ , and dried under high vacuum. Loadings of the sixteen solid supports F,aa-F,dd were between 30 and 50 μηιοΐ/g.

Multistep-Preparation of Phosphoramidites

Preferred embodiments of the phosphoramidites according to the present invention prepared according to the synthetic route outlined below in reaction scheme 2.

Reaction scheme 2: Preparation of phosphoramidites l

In synthetic step 1 (11→12) one of the two hydroxyl groups of ethylene glycol 11 is protected by a suitable protective group while the second hydroxyl group is reacted to give a functional thioacetal group. In subsequent step 2 (12→13), the thioacetal group reacted to yield chloromethylether 13. Subsequently, in step 3 (13 + 14→15a-15d), 5'-0-protected nucleosides 14a-14d (prepared according to S. Pitsch, P. A. Weiss, L. Jenny, A. Stutz, X. Wu, Helv. Chim. Acta, 2001, 84, 3773) are alkylated in the 2'-0-postion to yield 2'-0-formacetal substituted nucleosides 15a-15d. Finally, in step 4 (15a-15d→16a-16d), nucleosides 15a-15d are reacted with diisopropylphosphoramidochlorodite in the 3'-0 position to yield phosphoramidites 16a-16d. Synthesis of 12

Benzoyl chloride (2.3 ml, 20 mmol) was added dropwise to a solution of ethylene glycol (2.2 ml, 40 mmol) in pyridine (80 ml). After lh of stirring the major part of the pyridine was evaporated. After usual work up, the residue was co-evaporated with 3x 10 ml of toluene. The residue was dissolved in DMSO (20 ml), treated with AcOH (20 ml) and Ac ₂ 0 (20 ml). After 1 day of stirring, the reaction mixture was poured in a well stirred mixture of 400 ml (AcOEt Hexane 1 :1) and 400 ml of ¾0. The organic phase was then extracted with 400 ml of sat. NaHC0 ₃ , dried with MgS0 ₄ and evaporated. FC (100 g of Si0 ₂ , AcOEt/hexane 5:95 to 15:85): 1 (2.47g, 55%) as colorless oil. R _f 0.58 (AcOEt/Hexane 2:8). 1H -NMR (400 MHz, CDC1 ₃ ) : 2.15 (s, MeS); 3.89 (m, SCH ₂ ); 4.50 (m, C¾0); 4.71 (s, SCH ₂ 0); 7.44 (m, 2H arom.); 7.56 (m, 1H arom.); 8.06 (m, 2H arom.). MS: 249.5 (100, [M+ Na ).

Synthesis of crude 13

At -78°, S0 ₂ C1 ₂ (0.39 ml, 4.9 mmol) was added dropwise to a stirred solution of 12 (1.00 g, 4.4 mmol) and cyclohexene (0.89 ml, 8.8 mmol) in CH ₂ C1 ₂ (17.5 ml). After 30 min, the reaction mixture was allowed to warm up to room temperature and stirred for another 45 min. After evaporation of the solvent, the residue was dried under high vacuum for 15 min, dissolved in 1 ,2-dichloroethane and directly used for the alkylation reactions.

General procedure for the synthesis of nucleosides 15a-d

To a solution of nucleosides 14a-d (1 eq) in 1,2-dichloroethane (0.25 M) were added successively ('PrfeNEt (3.5 eq) and Bu ₂ SnCl ₂ (1.2 eq). The reaction mixtures were stirred at room temperature for 30 min, heated to 70°, treated with the chloromethylether 13 (1.2 eq) and stirred for 30 min at 70°. The reaction mixtures were then subjected to usual work up. The organic phase was filtered through Celite, dried with MgS0 ₄ and evaporated. The residue was purified by FC.

Synthesis of 15a

According to the general procedure with 14a (1.20 g, 2.0 mmol) and crude 13. FC (25 g of Si0 ₂ , AcOEt/hexane 5:5 to AcOEt then to AcOEt MeOH 97:3): 15a (442 mg, 27%) as white foam. R _f 0.38 (AcOEt). ¾ -NMR (400 MHz, CDC1 ₃ ) : 2.60 (s, MeCO); 2.90 (d, J = 5.4, HO- (C3')); 3.44 (dd, J = 4.0, 10.7, H-(C5')); 3.52 (dd, J = 3.0, 10.7, H-(C5')); 3.79-3.84 (m, MeO, 1H CH ₂ 0); 3.88-3.94 (m, 1H CH ₂ 0); 4.26 (m, H-(C4'); 4.39-4.42 (m, CII ₂ OBz); 4.57 (m, H- (C3')); 4.93-4.96 (m, OCH ₂ 0, H-(C2')); 6.26 (d, J = 4.4, H-(C1 ')): 6.79-6.83 (m, 4H arom.); 7.20-7.44 (m, 11H arom); 7.52-7.57 (m, 1H arom); 7.99-8.02 (m, 2H arom); 8.18 (s, H-(C2)); 8.62 (s, H-(C8)); 8.82 (br, s, NHAc). MS: 789 (100.. [Mf); 81 1 (16, [M + Na]*).

Synthesis of 15b According to the general procedure with 14b (1.17 g, 2.0 mmol) and crude 13. FC (25 g of Si0 _2, AcOEt/hexane 5:5 to AcOEt then to AcOEt/MeOH 97:3): 15b (680 mg, 45%) as white foam. R _f 0.57 (CH ₂ Cl ₂ /MeOH 9:1). 1H -NMR (400 MHz, CDC1 ₃ ) : 2.21 (s, MeCO); 2.91 (d, J = 9.2, HO-(C3')); 3.50 (dd, J = 2.4, 11.2, H-(C5')); 3.55 (dd, J = 2.0, 11.3, H-(C5')); 3.81 (s, 2 MeO); 3.92 (m, CH ₂ 0); 3.99-4.08 (m, C¾0, H-(C4')); 4.27 (d, J = 4.27, H-(C2')); 4.35- 4.49 (m, H-(C3')); 4.47-4.50 (m, CH ₂ OBz); 5.03, 5.22 (2d, J = 6.6, OCH ₂ 0); 5.98 (s, II- (Cl ⁵ )); 6.84-6.88 (m, 4H arom.); 7.08 (d, J =7.5; H-(C5)); 7.24-7.42 (m, 11H arom.); 7.49- 7.53 (m, 1H arom.); 8.01-8.05 (m, 2H arom.); 8.45 (d, J = 7.5; H-(C6)); 9.18 (br, s, NHAc). MS: 765 (100, [Mf ); 787 (37, [. + Na ) 803 (20, [M + †). Synthesis of 15c

According to the general procedure with 14c (1.25 g, 2.0 mmol) and crude 13. FC (25 g of Si0 ₂ , AcOEt hexane 5:5 to AcOEt then to AcOEt/MeOH 97:3): 15c (840 mg, 54%) as white foam. R _f 0.42 (CH ₂ Cl ₂ /MeOH 9:1). ¾ -NMR (400 MHz, CDC1 ₃ ) : 1.92 (s, MeCO); 3.24- 3.30 (m, H-(C5'), HO-(C3'); 3.49 (dd, J = 2.1 , 10.7, H-(C5')); 3.74 (s, 2 MeO); 3.86 (m, CH ₂ 0); 4.20 (m, H-(C4'); 4.40-4.55 (m, CH ₂ OBz, H-(C3'); 4.87 (m, OCH ₂ 0, H-(C2')); 6.01 (d, J = 4.6, H-(C1 ')); 6.74-6.79 (m, 4H arom.); 7.14-7.54 (m, 12H arom); 7.85 (s, H-(C5)); 7.97 (m, 2H arom); 9.29 (br, s, NHAc); 12.02 (br, s H-(N1)). MS: 805 (100, [ ] ⁺ ); 827 (24,

[M+ †).

Synthesis of 15d

According to the general procedure with 14d (1.09 g, 2.0 mmol) and crude 13. FC (25 g of Si0 ₂ , AcOEt hexane 2:8 to AcOEt): 15d (540 mg, 37%) as white foam. R _f 0.72 (AcOEt). ¾ -NMR (400 MHz, CDC1 ₃ ) :2.72 (d, J = 8.3, OH-(C3')); 3.49 (dd, J = 2.3, 11.0, H-(C5')); 3.53 (dd, J - 2.2, 11.0, H-(C5')); 3.79 (s, 2 MeO); 3.90-4.05 (ra, CH ₂ 0, H-(C4'); 4.31 (dd, J = 2.4, 5.2, H-(C2'); 4.45-4.55 (m, CH ₂ OBz, H-(C3'); 4.98, 5.09 (2d, J = 6.8, OCH ₂ 0); 5.29 (d, J = 8.1, H-(C5)); 6.02 (d, J = 2.4, H-(C1 ')); 6.85-6.89 (m, 4H arom.); 7.24-7.45 (m, 11H arom); 7.54-7.58 (m, 1H arom); 7.95 (d, J = 8.1, H-(C6)); 8.05-8.08 (m, 2H arom); 8.47 (br. s, H- (N3)). MS (Pos. Mode): 746 (100, [Af + Naf) ; 762 (13, [M+ K] ⁺ )

General procedure for the synthesis of phosphoramidites 16a-d To a solution of the nucleosides 15a-d (1 eq) in CH ₂ C1 ₂ (0.25 M) were added successively ('Pr) ₂ NEt (2.5 eq) and diisopropylphosphoramidochloridite (1.2 eq). The reaction mixtures were stirred for 14 h at room temperature and subjected to FC.

Synthesis of 16a

According to the general procedure with 15a (393 mg, 0.50 mmol). FC: (8 g of Si0 ₂ , AcOEt/hexane 2:8 (+2% NEt ₃ ) to AcOEt/hexane 8:2 (+2% NEt ₃ )): 16a (390 mg, 79%) as white foam (1 :1 mixture of diastereoisomers). R _f 0.70 (AcOEt). ¾ -NMR (400 MHz, CDC1 ₃ ) : 1.06 (d, 3H, J - 6.8, NCH( e) ₂ ); 1.16-1.19 (m, 9H, NCH( ?) ₂ ); 2.39 (t, J =6.4, CH ₂ CN); 2.59 (s, MeCO); 2.63 (t, J - 6.4, CH ₂ CN) 3.37 (m, POCH ₂ ); 3.50-3.94 (m, 2H-(C5'), MeO, CH ₂ 0, NCH(Me) ₂ , POCH ₂ ); 4.27-4.44 (m, H-(C4'), C¾OBz); 4.70 (m, H-(C3'); 4.81-4.95 (m, OCH ₂ 0); 5.11 (m, H-(C2')); 6.24, 6.26 (2d, J = 6.4, H-(Cl'); 6.79-6.82 (m, 4H arom.); 7.18-7.43 (m, 11H arom); 7.52-7.56 (m, 1H arom); 7.96-7.98 (m, 2H arom); 8.18, 8.19 (2s, H- (C2)); 8.58 (br, s, NHAc); 8.60, 8.61 (2s, H-(C8)). MS: 990 (100 _r [M+Uf).

Synthesis of 16b

According to the general procedure with 15b (579 mg, 0.76 mmol). FC: (11 g of Si0 ₂ , AcOEt/hexane 2:8 (+2% NEt ₃ ) to AcOEt/hexane 8:2 (+2% NEt ₃ )): 16b (662 mg, 90%) as white foam (1:1 mixture of diastereoisomers). R _f 0.60 (AcOEt). ). H -NMR (400 MHz, CDCI ₃ ) : 1.00 (d, 3H, J = 6.8, NCH( ?) ₂ ); 1.14-1.17 (m, 9H, NCH( e) ₂ ); 1.76, 1.85 (2s, MeCO); 2.32 (t, J =6.3, CH ₂ CN); 2.69 (dt, J = 3.0, 6.6, CH ₂ CN); 3.34-3.62 (m, 2H-(C5'), NCH( e) ₂ , 2H POCH ₂ ); 3.702, 3.712, 3.715 (3s, MeO); 3.83-4.02 (m, CH ₂ 0); 4.16-4.22 (m, H-(C4')); 4.27 (m, H-(C2'), 4.35-4.48 (m, H-(C3'), CH ₂ OBz); 4.92-5.02 (m, OCH ₂ 0); 5.99 (s, H-(Cl '); 6.73-6.78 (m, 4H arom.); 6.82, 6.90 (2d, J = 7.5, H-(C5)); 7.16-7.34 (ra, 11H arom); 7.41-7.45 (m, 1H arom); 7.92-7.96 (m, 2H arom); 8.34, 8.43 (2d, J = 7.5, H-(C6)); 9.39, 9.51 (br, 2s, NHAc). MS: 966 (100, [Α +Η] ⁺ ) 988 (25, [Af+Na] ⁺ ); 1005 (15, [M+K) ⁺ ). Synthesis of 16c

According to the general procedure with 15c (594 mg, 0.74 mmol). FC: (8 g of Si0 ₂ , AcOEt/hexane 5:5 (+2% NEt ₃ ) to AcOEt (+2% NEt ₃ )): 16c (590 mg, 80%) as white foam (1 :1 mixture of diastereoisomers). R _f 0.66 (AcOEt). ¾ -NMR (400 MHz, CDC1 ₃ ) : 1.00 (d, J = 6.8, NCH( ?) ₂ ); 1.14-1.17 (m, NCH( e) ₂ ); 1.76, 1.85 (2s, MeCO); 2.31 (t, J =6.3, CH ₂ CN); 2.69 (dt, J = 2.9, 6.6, CH ₂ CN); 3.24 (m, POCH ₂ ); 3.49-4.00 (m, 2H-(C5'), MeO, CH ₂ 0, NCH(Me) ₂ , P0C¾); 4.27 (m, 0.5 H-(C4')); 4.34-4.56 (m, 0.5 H-(C4'), 0.5 H-(C3'), C¾OBz); 4.62 (m, 0.5 H-(C3')); 4.78, 4.82, 4.85, 4.91 (4d, J = 7.1 OCH ₂ 0); 5.07 (m, II- (C2')) ₅ 5.95, 6.01 (2d, J = 5.2, 5.9, H-(Cl '); 6.78-6.82 (m, 4H arom.); 7.16-7.59 (m, 13H arom); 7.864, 7.866 (2s, H-(C8)); 8.00-8.02 (m, 2H arom); 8.46, 8.76 (br, 2s, NHAc). MS: 1007 (100, [ +H] ^'1' ); 1028 (11, [M+Na\ ^l ).

Synthesis of 16d According to the general procedure with 15d (497 mg, 0.69 mmol). FC: (10 g of Si0 ₂ , AcOEt hexane 2:8 (+2% NEt ₃ ) to AcOEt/hexane 6:4 (+2% NEt ₃ )): 16d (575 mg, 91%) as white foam (1 :1 mixture of diastereoisomers). R _f 0.87 (AcOEt). 1H -NMR (400 MHz, CDC1 ₃ ) : 1.02 (d, 3H, J = 6.8, NCH( e) ₂ ); 1.14-1.17 (m, 9H, NCH( ?) ₂ ); 2.43 (t, J =6.3, C¾CN); 2.64 (dt, J = 2.4, 6.1, C¾CN); 3.41-3.77 (m, 2H-(C5'), NCH(Me) ₂ , 2H POCH ₂ ); 3.791, 3.800, 3.803 (3s, MeO); 3.87-4.02 (m, CH ₂ 0); 4.19-4.28 (m, H-(C4')); 4.42-4.63 (m, H- (C2'), H-(C3'), CH ₂ OBz); 4.92, 5.03 (2d, J = 6.8, OCH ₂ 0); 4.96 (s, OCH ₂ '0); 5.20, 5.24 (2d, J = 8.2, H-(C5)); 6.06 (d, J = 3.1, H-(Cl '); 6.82-6.86 (m, 4H arom.); 7.22-7.44 (m, 11H arom); 7.53-7.58 (m, 1H arom); 7.94, 7.99 (2d, J = 8.2, H-(C6)); 8.03-8.06 (m, 2H arom); 8.33 (br, s, NHAc). MS: 925 ( _. 100, [ +Hf).

A preferred embodiment of the synthesis of a support material comprising chemical groups according to Formulas Villa and/or Vlllb as defined above and in the claims is described in the following with reference to reaction scheme 3. Reaction scheme 3: Synthesis of support material comprising 2Ό- or 3Ό- octyloxymethyl modified nucleosides.

20 = OC ₆ H4N02 DMT = 4,4'-(Dimethoxy)trityl C 21 = 0C ₆ H ₄ N0 ₂

CPG = Controlled Pore Glass

22 R = NH-CPG 23 R = NH-CPG

Synthesis of 18 and 19

Stepl: A mixture of octane- l-ol (390 mg, 3 mmol) and para-formaldehyde (90 mg, 3 mmmol) in CH ₂ C1 ₂ (11 ml) was dissolved in dichloromethane (18 ml). A flow of HCl(g) was bubbled 15 min through this reaction mixture. The resulting organic layer was separated from the aqueous layer, evaporated to dryness and finally dried 10 min under high vaccum.

Step2: To a solution of 5'-0-(4,4 -dimethoxytrityl) uridine 1 (2g g, 3.658 mmol) in 1,2- dichloroethane (12 ml) was added Bu ₂ SnCl ₂ (1.33g, 4.39 mmol) and diisopropylethylamine (2.17 ml, 13.8 mmol). The reaction mixture was stirred 15 min at room temperature and then to 70 °C. The crude product fom step 1, dissolved in dichlorometliane (3 ml), was added slowly to the previous mixture and stirred 30 min at 70°C. After workup, the two diastereoisomers 18 (lg, 40 %) and 19 (490 mg, 20% ) were separated by CC (50 g of Si0 ₂ , AcOEt/hexane 5:5 to AcOEt 100 %). Global yield: 59 %. TLC (AcOEt/hexane 95/5): Rf 0.S0 (18); 0.68 (19). 1H-NMR (400 MHz, CDC1 ₃ ); 18: 2.83 (d, J = 7.2, OH-(C3')); 3.52-3.68 (m, H-(C5'), H-(OCH ₂ ) _n ); 3.80 (s, 2 MeO); 4.11 (m, H-(C4')); 4.34 (dd, J = 3.6, 5.1, H-(C2'); 4.49 (m, H-(C3')); 4.73, 4.78 (2d, J = 6.9, OCH ₂ 0); 5.31 (d, J = 8.2, H-(C5)); 6.04 (d, J = 3.1, H-(C1')); 6.85-6.89 (m, 4H arom.); 7.24-7.45 (m, 9H arom); 7.95 (d, J = 8.2, H-(C6)); 8.23 (br. s, H-(N3)). 19: 3.39-3.60 (m, OH-(C2'), H-(C5'), H-(OC¾) _n ); 3.80 (s, 2 MeO); 4.26- 4.36 (m, H-C(2'), H-C(3'), H-C(4')); 4.73, 4.78 (2d, J = 6.8, OCH ₂ 0); 5.37 (d, J = 8.1, H- (C5)); 5.97 (d, J = 3.1, H-(C1 ')); 6.85-6.89 (m, 4H arom.); 7.24-7.45 (m, 9H arom); 7.85 (d, J = 8.2, H-(C6)); 8.9 (br. s, H-( 3)).

Synthesis of 20 and 21 Pimelicacid-dinitrophenylester (697 mg, 1.73 mmol) was dissolved in pyridine (2.89 ml). DMAP (17.7 mg, 0.14 mmol) and 18 or 19 (200 mg, 0.29 mmol each), respectively, were added. After stirring over night at room temperature, the mixture was evaporated and co- evaporated twice with toluene (20 ml). CC (10 g of Si0 ₂ , AcOEt/hexane 3:7 to AcOEt/hexane 6:4) gave 20 (121 mg, 46%) and 21 (118 mg, 45 %). Colorless foam. TLC (AcOEt hexane 7:3): RfO O (20); 0.67 (21). ¾- MR (400 MHz, CDC1 ₃ ); 20: 1.46 n, CH ₂ ); 1.70 (m, CH ₂ ); 1.77 m. CH ₂ ); 2.42 (t, J = 7.6, CH ₂ ); 2.62 (t, J = 7.5, CH ₂ ); 3.52-3.68 (m, H-(C5'), H- (CH ₂ ) _n ); 3.80 (s, 2 MeO); 4.11 (m, H-(C4')); 4.34 (dd, J = 3.6, 5.1, H-(C2'); 4.49 (m, H- (C3')); 4.73, 4.78 (2d, J = 6.9, OCH ₂ 0); 5.31 (d, J = 8.2, H-(C5)); 6.04 (d, J = 3.1, H-(C1')); 6.85-6.89 (m, 4H arom.); 7.24-7.45 (m, 9H arom); 7.95 (d, J = 8.2, H-(C6)); 8.23 (br. s, H- (N3)). 21: 1.46 (»?, CH ₂ ); 1.70 (ra, CH ₂ ); 1.77 (m. CH ₂ ); 2.42 (t, J= 7.6, C¾); 2.62 (t, J= 7.5, CH ₂ ); H-(C5'), H-(CH ₂ )„); 3.80 (s, 2 MeO); 4.26-4.36 (m, H-C(2') _> H-C(3'), H-C(4')); 4.73, 4.78 (2d, J = 6.8, OCH ₂ 0); 5.37 (d, J = 8.1, H-(C5)); 5.97 (d, J = 3.1, H-(C1')); 6.85-6.89 (m, 4H arom.); 7.24-7.45 (m, 9H arom); 7.85 (d, J = 8.2, H-(C6)); 8.9 (br. s, H-(N3)).

Preparation of the solid support 22 and 23

To a solution of the active ester (0.05 mmol) 20 or 21 in DMF (1.6 ml) was added long-chain- alkylamino CPG (400 g) and then 'Pr ₂ NEt (0.15 ml). The mixture was shaken for 20 h at room temperature. After filtration, the solid support 22 or 23 was washed with DMF and CH ₂ C1 ₂ , dried, suspended in pyridine (1 ml) and Ac ₂ 0 (0.64 ml) and shaken for 2 h at room temperature. After filtration, the solid was washed with DMF and CH ₂ C1 ₂ , and dried under high vacuum. Typical loadings of the solid support was 30 μηιοΐ/g with 500-A CPG. Reaction scheme 4: Preparation of phosphoramidites

In synthetic step 1 (24→25) the hydroxy 1 group of phthalimide derivative 24 is reacted to give a functional thioacetal group. In subsequent step 2. the thioacetal group of 25 is reacted in situ to yield the corresponding cliloromethylether. Subsequently, in step 2 (25 + 14a- d→15a-15d), 5'-0-protected nucleosides 34a~14d (prepared according to S. Pitsch, P. A. Weiss, L. Jenny, A. Stutz, X. Wu, Helv. Chim. Acta, 2001, 84, 3773) are alkylated in the 2'- O-postion to yield 2 ^! -0-formacetal substituted nucleosides 26a-26d. Finally, in step 3 (26a- 26d→27a-27d), nucleosides 26a-26d are reacted with diisopropylphosphoramidochlorodite in the 3 ^" -0 position to yield phosphoramidites 27a-27d.

Synthesis of 25

24 (9.6 g, 50 mmol) was dissolved in DM80 (75 nil), treated with AcOH (75 ml) and Ac ₂ 0 (75 ml). After 2 days of stirring the reaction mixture was poured in a well stirred mixture of 400 ml (AcOEt hexane 1 :1) and 400 ml of H 0. The organic phase was then extracted with 400 ml of sat. NaHC0 ₃ , dried with MgS0 ₄ and evaporated. The residue was dissolved in hot AcOEt (20 ml) and hexane (180 ml) was added. The mixture was stirred overnight at 4°C and tire crystalised product 25 was filtered, washed with AcOEt Hexane 1 :9. and dried under vacuum: 25 (9.3g, 74%) as a white powder. R _f 0.58 (AcOEt). ^! H-NMR (400 MHz, CDClj) : 2.06 (s, MeS); 3.83, 3.84 (2t, J = 5.5, SCH ₂ , CH ₂ N); 4.65 (s, SCH ₂ 0); 7.73-7.75 (m, 2H arom.); 7.86-7.89 (m, 2H arom.). General procedure for the synthesis of nucleosides 26a-d

To a solution of nucleosides 14a-d (1 eq) and thiomethylether 25 in 1,2-dichloroethane (0.25 M) were added successively ('Pr^NEt (3.5 eq) and Bu ₂ SnCl ₂ (1.2 eq). The reaction mixtures were stirred at room temperature for 30 min, heated to 70°. treated with N-c orosuccinimide (2 eq) and stirred for 30 min at 70°. The reaction mixtures were then subjected to usual work up. The organic phase was filtered through Celite. dried with MgS0 ₄ and evaporated. The residue was purified by CC.

Synthesis of 26a

According to general procedure with 14a (2.44 g, 4.0 mmol). CC (50 g of Si0 ₂ , AcOEt/hexane 7:3 to AcOEt then to AcOEt/MeOH 97:3): 26a (785 mg, 24%) as a foam. R _r 0.45 (AcOEt MeOH 95:5). ¾ NMR (400 MHz, CDC1 ₃ ) : 2.65 (s, MeCO), 2.78 (d, J = 5.4 Hz, HO-(C3')), 3.43 (dd, J = 10.6, 4.2 Hz, H-(C5 ^? )), 3.49 (dd, J = 10.6, 3.2 Hz, H-(C5')), 3.73-3.88 (m, MeO, CH ₂ 0, CH ₂ N), 4.24 (q, J = 4.2 Hz, H-(C4') _; 4.53 (q, J = 5.1 Hz, H- (C3 ^r )); 4.84-4.90 (m, OCH ₂ 0, H-(C2')), 6.20 (d. ./ - 4.5 Hz, H-(CF)), 6.79-6.85 (m, 4H arom.), 7.21-7.34 (m, 7H arom.), 7.40-7.45 (m, 2H arom.), 7.68-7.72 (m, 2H arom.), 7.79- 7.83 (m, 2H arom.), 8.18 (s, H-(C2)), 8.50 (br, s, NHAc), 8.60 (s, H-(C8)). ESI-MS (Pos. Mode): 815.4 (100, [M+ H] ⁺ ). Synthesis of 26b

According to general procedure with 14b (2.38 g, 4.0 mmol. CC (50 g of Si 0 ₂ . AcOEt hexane 7:3 to AcOEt then to AcOEt/MeOH 97:3): 26b (582 mg 19%) as light yellow foam. Rf 0.43 (AcOEt/MeOH 95:5). Ή -NM.R (400 MHz, CDC1 ₃ ) : 2.23 (s, MeCO), 2.96 (d, J = 9.3 Hz, HO-(C3 ')), 3.53 (dd, J= 11.2, 2.6 Hz, H-(C5 ^: )), 3.58 (dd, J= 1 1.5, 2.2 Hz, H-(C5 ^' )), 3.84 (d, J = 1.0 Hz, MeO), 3.88-3.98 (m, CH ₂ 0, CH ₂ N), 4.07 (m. H-(C4')), 4.21 (d, J = 5.1 Hz, H- (C2')), 4.42 (td, J = 9.3, 5.4 Hz, H-(C3')), 4.99 (d, J = 6.7 Hz, OCH ₂ 0), 5.17 (d, J = 6.4 Hz, OC¾0), 5.94 (s, H-(C1')), 6.87-6.91 (m, 4H arom.), 7.09 (d, J = 7.5 Hz, H-(C5)), 7.26-7.37 (m, 7H arom.), 7.41 -7.45 (m, 2H arom.), 7.68-7.72 (m, 2H arom.), 7.81-7.85 (m, 2H arom.), 8.46 (d, J = 7.5 Hz, H-(C6)), 9.11 (br, s, NHAc). ESI-MS (Pos. Mode): 791.3 (100, [M + H] ⁺ ).

Synthesis of 26c

According to general procedure with 14c (2.51 g, 4.0 mmol). CC (25 g of Si0 ₂ , AcOEt/hexane 8:2 to AcOEt then to AcOEt/MeOH 95:5): 26c (1950 mg, 60%) as yellow foam. R _f 0.38 (AcOEt/MeOH 95:5). ^! H -NMR (400 MHz, CDC1 ₃ ) : 1.99 (s, MeCO). 3.30 (dd, J = 10.9, 4.2 Hz, H-(C5')), 3.51 (dd, J = 10.9, 2.2 Hz, H-(C5 ^r )), 3.73-3.99 (m, MeO, CH ₂ 0, CH ₂ N), 4.17 (m, H-(C4 ^? )). 4.45 (q. J = 5.1 Hz, H-(C3 ^T )) ₅ 4.79-4.88 (m, OCH ₂ 0, H- (C2 5.96 (d, J = 4.2 Hz, H-(C1 ')), 6.79-6.85 (m, 4H arom.), 7.19-7.38 (m, 7H arom.), 7.46- 7.50 (m, 2H arom.), 7.70-7.74 (m, 2H arom.), 7.78-7.82 (m, 2H arom.), 7.86 (s, H-(C8)), 8.96 (br, s, NHAc). 12.02 (br, s H-(N1)). ESI-MS (Pos. Mode): 831.3 (100, [ + Hf).

Syntliesis of 26d

According to general procedure witli 14d (2.38 g, 4.0 mmol). CC (55 g of Si0 ₂ , AcOEt/hexane 2:8 to AcOEt): 26d (840 mg, 28%) as white foam. R _f 0.75 (AcOEt/MeOH 95:5). ¹ H ~NMR (400 MHz, CDC1 ₃ ) : 2.75 (d, J = 7.4 Hz, OH-(C3')), 3.51 (dd, J = 11 .2, 2.6 Hz, H-(C5')), 3.55 (dd, J = 11.2, 2.2 Hz, H-(C5')), 3.82 (s, MeO), 3.87-3.98 (m, CH ₂ 0, CH ₂ N), 4.05 (m, H-(C4')), 4.24 (dd, J= 5.1 , 2.6 Hz, H-(C2')), 4.44 (m, H-(C3')), 4.91 (d, J= 6.7 Hz, OCH ₂ 0), 5.00 (d, J= 6.7 Hz, OCH ₂ 0), 5.29 (dd, J = 8.3, 1.9 Hz, H-(C5)), 5.96 (d, J = 2.6 Hz, H-(Cl ')), 6.84-6.90 (m, 4H arom.), 7.23-7.42 (m. 9H arom,), 7.70-7.74 (m, 2H arom.), 7.83-7.87 (m, 2H arom.), 7.94 (d, J= 8.0 Hz, H-(C6)), 8.13 (br. s, H-(N3)).

General procedure for the synthesis of phosphoramidites 27a-d

To a solution of the nucleosides 26a-d (1 eq) in CH ₂ C1 ₂ (0.25 M) were added successively ('Pr) ₂ NEt (2.5 eq) and diisopropylphosphoramidochloridite (1.2 eq). The reaction mixtures were stirred for 14 h at room temperature and subjected to CC. Synthesis of 27a

According to general procedure with 26a (870 mg, 1.07 mmol). CC: (17 g of Si0 ₂ . AcOEt/hexane 2:8 (+2% NEt ₃ ) to AcOEt (+2% NEt ₃ )): 27a (970 mg, 89%) as white foam (1 :1 mixture of diastereoisomers). R _r 0.50 (AcOEt). Ή -NMR (400 MHz, CDC1 ₃ ) : 1.07 (d, J = 6.7 Hz, NCH(ikfe) ₂ ), 1.16-1.20 (m, NCH( ?) ₂ ), 2.40 (t J= 6.4 Hz, C¾CN), 2.63-2.69 (m _s CH ₂ CN, MeCO), 3.36 (m, 1H POCH ₂ ), 3.49-3.97 (m, 2H-(C5'), MeO, CH ₂ N, CH ₂ 0, 1H POCH ₂ ). 4.36 (m, H-(C4')) _f 4.69 (m, H-(C3 ^? )) _; 4.76 (d, J= 7.0 Hz, 1H OCH ₂ 0), 4.80 (d, J = 7.0 Hz, 0.5H OCH ₂ 0), 4.87 (d, J= 6.7 Hz, 0.5H OC¾0), 5.08 (m, H-(C2')), 6.19 (d. J= 5.1 Hz, 0.5H H-(Cr)), 6.22 (d, J = 4.8 Hz, 0.5H H-(C1 ')), 6.78-6.83 (m, 4H arom.), 7.19-7.44 (m, 9H arom.), 7.68-7.74 (m, 2H arom.) _f 7.77-7.83 (m, 2H arom.), 8-22 (s, H-(C2)) _S 8.50 (br, s, NHAc), 8.60 (s, H-(C8)). ³³ P -NMR (161 MHz, CDC1 ₃ ) : 151.67; 151.77. ESI MS (Pos. Mode): 1015.37 (100, [ + H )

Synthesis of 27b

According to general procedure with 26b (550 mg, 0.70 mmol). CC: (11 g of Si0 ₂ , AcOEt/hexane 2:8 (+2% NEt ₃ ) to AcOEt/hexane 8:2 (+2% NEt ₃ )): 27b (540 mg, 78%) as white foam (1 :1 mixture of diastereoisomers). Rf 0.43 (AcOEt). ). ^! H -NMR (400 MHz, CDCI ₃ ) : 0.99 (d, J= 7.0 Hz, NCH( ?) ₂ ), 1.11 (d, J = 6.7 Hz, NCH( ₂ ), 1.12 (d, J= 7.0 Hz, NCH(M?) ₂ ), 1.16 (d, J = 6.7 Hz, NCH( e) ₂ ), 2.22, 2.23 (2s, MeCO), 2.43 (t, J = 6.4 Hz, CH ₂ CN), 2.66 (m, CH ₂ CN), 3.42-3.78 (m, 2H-(C5 ^, ) ₃ NCH( e) ₂ , 2H POCH ₂ ), 3.79-4.07 (m, MeO, CH ₂ N, CH ₂ 0), 4.24-4.38 (m, H-(C4'), H-(C2')), 4.49 (m, H-(C3')). 4.95 (d, J= 6.7 Hz, 0.5H OCH ₂ 0), 4.97 (d, J= 6.7 Hz, 0.5H OCH ₂ 0), 5.02 (d, J = 6.7 Hz, 0.5H OCH ₂ 0), 5.02 (d, J = 6.4 Hz, 0.5H 0CH ₂ 0) ₅ 6.05 (d, j = 1.0 Hz, 0.5H H-(C1 ')), 6.06 (d, J = 1.3 Hz, 0.5H H- (CI ')), 6.85-6.90 (m, 4H arom.), 6.92 (d, J= 7.7 Hz. 0.5H H-(C5)) _r 7.00 (d, J = 7.7 Hz, 0.5H H-(C5)), 7.26-7.46 (m, 9H arom.), 7.68-7.73 (m, 2H arom,), 7.82-7.86 (m 2H arom.), 8.44 (d, J = 7.4 Hz, 0.5H H-(C6)), 8.50 (d, J = 7.4 Hz, 0.5H H-(C6)). 8.94, 9.07 (br, 2s, NHAc). ^{3 I} P - NMR (161 MHz, CDC1 ₃ ) : 150.88; 152.42. MALDl-TOF MS (Pos. Mode): 1013.9 (100, [M + Na] ⁺ ) ; 1029.9 (49, [M+ K] )

Synthesis of 27c

According to general procedure with 26c (1940 mg, 2.33 ramol). CC: (35 g of Si0 ₂ , AcOEt/hexane 5:5 (+2% NEt ₃ ) to AcOEt (+2% NEt ₃ )): 276c (1810 mg, 80%) as white foam (1 :1 mixture of diastereoisomers). R _f 0.44 (AcOEt). Ή -NMR (400 MHz, CDC1 ₃ ) : 1.00 (d. J = 7.0 Hz, NCH( e) ₂ ), 1.15 (d, J= 6.7 Hz, NCH( ?) ₂ ), 1.15 (d, J= 6.7 Hz, NCH( ?) ₂ ), 1.87, 1.96 (2s, MeCO), 2.32 (t, j = 6.1 Hz, CH ₂ CN), 2.68 (td, J = 6.4, 2.2 Hz, CH ₂ CN), 3.27 (m, 1H POCH ₂ ), 3.48-3.99 (m, 1H POC¾ 2H-(C5'), MeO, CH ₂ 0, CH ₂ N, NCH(Me) ₂ ), 4.30 (m, H-(C4 ^? )) _; 4.51 (m, 0.5 H-(C3')), 4.61 (m 0.5 H-(CT)), 4.73-4.82 (m, 1.5H OCH ₂ 0), 4.89 (d, J= 7.0 Hz, 0.5H, OCH ₂ 0), 4.97 (m, H-(C2')), 5.93 (d, J = 4.5 Hz, 0.5H H-(CF)), 5.98 (d, J= 5.1 Hz, 0.5H H-(C1 ')), 6.79-6.86 (m. 4H arom.), 7.19-7.55 (m, 9H arom.), 7.70-7.77 (m, 2H arom.), 7.79-7.84 (m, 2H arom.), 7.88, 7.91 (2 s, H-(C8)). 8.62, 8.87 (br, 2s, NHAc). 11.93 (br, s, H-(N1)). ³¹ P -NMR (161 MHz, CDCI3) : 150.93; 151.60. MALDl-TOF MS (Pos. Mode): 1031.9 (57, [M+ Na ) 1053.9 (100, [ + Na ) ; 1069.9 (53, [M+ K] ⁺ ).

Synthesis of 27d

According to general procedure with 26d (473 mg, 0.63 mmol). CC: (10 g of Si0 ₂ , AcOEt/hexane 2:8 (+2% NEt ₃ ) to AcOEt/hexane 9:1 (+2% NEt ₃ )): 27d (446 mg, 75%) as white foam (1:1 mixture of diastereoisomers). R _f 0.80 (AcOEt). Ή -NMR (400 MHz, CDC.I3) : 1.02 (d, J = 6.7 Hz, NCH( ) ₂ ), 1.13-1.18 (m, NCH( g) ₂ ), 2.44 (t, J = 6.1 Hz, CH ₂ CN), 2.70 (m, CH ₂ CN), 3.39-3.84 (m, POCH ₂ , 2H-(C5 ^; ), MeO, NCH(Me) ₂ ), 3.88-4.01 (m, CH ₂ 0, CH ₂ N), 4.20 (m, H-(C4')), 4.40 (m, H-(C2')), 4.55 (m, H-(C3')), 4.83 (d, J = 6.7 Hz, 0.5H OCH ₂ 0), 4.87 (s, 1H OCH ₂ 0). 4.99 (d, J = 6.7 Hz, 0.5H OCH ₂ 0), 5.21 (d, J= 8.3 Hz, 0.5H H-(C5)), 5.28 (d, J= 8.3 Hz, 0.5H H-(C5)), 6.01 (d, J= 3.2 Hz, 0.5H H-(C1 ')), 6.03 (d, J = 3.5 Hz, 0.5H H-(Cl ')), 6.83-6.88 (m, 4H arom.), 7.23-7.43 (m, 9H arom,), 7.71-7.76 (m, 2H arom.), 7.83-7.87 (m, 2H arom.), 7.91 (d, J= 8.3 Hz, 0.5H H-(C6)), 7.94 (d, J= 8.3 Hz, 0.5H H-(C6)), 8.09 (br, s, H-(N3)). ^{3 !} P -NMR (161 MHz, CDC1 ₃ ) : 151.14: 152.27. MALDI-TOF MS (Pos. Mode): 972.7 (85 _r [Λ/+ Na ) : 988.8 (100, [M+ ] ⁺ ).

Oligonucleotide Synthesis and Deprotection

The RNA oligonucleotides were prepared using a Gene Assembler Plus {Pharmacia) with 40 mg of solid support (either conventional solid supports or solid supports according to the present invention) (loading 30 μηιοΐ/g) and 2'-0-tom-protected ribonucleoside phosphoramidites, DNA phosphoramidites (obtained from the Glen Research Corporation, Sterling, VA, USA), 2'OMe ribonculesoide phosphoramidites (obtained from the Glen Research Corporation, Sterling, VA, USA) or phosphoramidites according to the present invention utilizing standard procedures as for instance outlined by S. Pitsch, P. A. Weiss, L. Jenny, A. Stutz, X. Wu, Helv. Chim. Acta 2001, 84, 3Π3. The selection of the solid supports and the selection of the respective phosphoramidites used in the syntheses depend on the nature of the modification of the RNA oligonucleotides. Additionally used modifications in the nucleosides included 2'-deoxy-modified nucleotides, 2 ' -methoxy-modified nucleotides as well as nucleotides comprising in the 3 '-position a -0-CH ₂ -0-(CH ₂ )7-CH3 group. An example of such a modified nucleotide is shown below in Formula IV.

FORMULA IV

The solid support was subsequently removed from the cartridge and treated with a 1 :1 mixture of 12 M MeNH ₂ in H ₂ 0 and 8 M MeNH ₂ in EtOH (1 ml) for 5 h at 35 °C. By centrifugation, the supernatent solution was separated from the solid support and evaporated, and the residue was dissolved in 1 M Bu ₄ F-3H ₂ 0 soln. in THF (1 ml). After 14 h at 25°C, 1 M 7 zs.HCl buffer (pH 7.4, 1ml) was added. The solution was concentrated to 1ml and desalted on NAP- 10 column (Pharmacia) according to the manufacturer's instructions. The crude RNA sequence was purified by AE-HPLC (20-60 % B in A, see below, in 45 min, 3 injections). The fractions containing the pure sequence were pooled (30 ml), treated with 1 M aqueous Et ₃ N. AcOH (pH 7, 5 ml) and applied to Sepak cartridges (Waters). After elution of the salts with 0.1 M aqueous Et ₃ N.AcOH (pH 7, 10 ml), followed by H ₂ 0 (20 ml), the sequences was eluted with MeCN/¾0 1:1 (5 ml) and finally evaporated to dryness. The residue was redissolved in 1ml water, quantified and analysed by MALDI TOF MS.

For an example of a R A oligonucleotide according of the present invention Figure 1 demonstrates different purification steps of the RNA oligonucleotide.

Oligonucleotide hybridization and duplex purification

The two single strands (sense and antisense strand) in H ₂ 0 were mixed in a 1 :1 ratio (total double strand concentration ca. 20 nmol), heated 1 min. to 80°C and slowly cooled to 25°C. After lyophilization, the duplex was dissolved in 50 μΐ, mixed with 50 μΐ formamide blue (10 ml formamide, 200 μΐ 0.5M EDTA, bromphenol blue) and then subjected to a polyacrylamid gel electrophoresis. (12 % acrylamide, 1:40 bisacrylamide, 250 V). After UV revelation the duplex was separated from the gel in a electrolution cell (45 min, 70 V). The crude product was purified by AE-HPLC (20-60 % B in A, see below, in 25 min, 1 injection). The fractions containing the pure duplex were pooled, treated with 1 M aqueous Et ₃ N. AcOH (pH 7) and applied to Sepak cartridges (Waters). After elution of the salts with 0.1 M aqueous Et ₃ N.AcOH (pH 7; 10 ml), followed by H ₂ 0 (20 ml), the duplex was eluted with MeCN/H ₂ 0 1 :1 (5 ml) and finally evaporated to dryness.

Stability tests Stability in culture medium

2 μg siRNA were incubated at 37 °C in 300 μΐ DMEM/10 % FCS (Dulbecco's Modified Eagle Medium (D-MEM) (IX), Invitrogen-GIBCO). The resulting mixture was analyzed on AE-HPLC (20-60 % B in A, see below, in 25 min, 1 injections) at 10, 4 h, 24 h, 48 h and 3 d. Stability in human serum

3 μ& siRNA were pipetted in 100 μΐ 100 % human serum. The siRNA was incubated in serum at 37°C and at different time points 25 μΐ aliquots were collected and frozen until analysis.

Stability in cell culture medium after complexation with transfection reagent

2 g siRNA were incubated for 15 min with RNAiFect (obtained from QIAGEN GmbH. Hilden, Germany) in 1:6 ratio siRNA : μΐ transfection reagent) to form transfections complexes at real transfection conditions. After complex formation the mixture was pipetted in 300 μΐ cell culture medium with 10 % FCS (fetal calf serum). The transfection complexes were incubated at normal cell growth conditions (i.e. 37°C, 5 % C0 ₂ ) and at different time points aliquots were collected and frozen until analysis. Analysis of siRNA stability

PAA gels

12% PAA-TBE mini gels (Mini-PROTEAN 3 Electrophoresis System, from Bio-Rad) were used for analysis of the treated siR As. 11 μΐ of each aliquot were mixed with 4 μΐ Loading Buffer (2.5xTBE, 50 % Formamid, Orange G, 20 % Glycerol) and loaded onto the gel. The samples were separated for 2,5 h at 100 V.

HPLC analysis

HPLC analysis was performed under denaturing conditions using the following protocol: Column:

DNAPac PA100 Analytical, 4*250 mm

Eluants:

A: 0.0 M NaC10 , 5 M urea, 12.5 mM Tris HC1 (pH= 7.4), final volume: 1 1 (Aqua Braun H ₂ 0)

B: 0.5 M NaC10 ₄ , 5 M urea, 12.5 mM Tris HC1 (pH= 7.4), final volume: 11 (¾0)

Gradient: linear gradient from 20 to 60 % B in A

Time: 25 min. Flow rate: 1 ml/min

Temperature: 80 to 85°C

Inject volume: 500 μΐ (300 μΐ sample + 200 μΐ buffer A)

Column wash: 20 % MeOH, 50 mM NH ₄ C1, final volume: 1 1 H20

Column Storage: 20 % MeOH, 50 mM NH ₄ C1, final volume: 11 H20

The following is a brief description of the drawings:

Fig. 1 shows the purification steps of an RNA oligonucleotide according to the present invention;

Fig. 2 shows the gene silencing efficiency in HeLAS3 cells transfected with 100 nM modified siRNAs according to the present invention (Examples 1 to 95), 48 h after the transfection;

Fig. 3 shows the gene silencing efficiency in HeLAS3 cells transfected with 100 pM modified siRNAs according to the present invention (Examples 27, 30, 33, 35,

43, and 53), with HiPerFect being used as a transfection reagent; Fig. 4 shows the gene silencing efficiency in different cell types (HeLaS3, MCF7, and

T98G) transfected with different amounts of modified siRNA according to the present invention (100 nM, 25 nM, 10 nM, 5 nM);

Fig. 5 shows the gene silencing efficiency of low performance siRNA compared to modified siRNA according to the present invention (Examples 27 and 30) at different concentrations (100 pM, 1 nM, 5 nM);

Fig. 6 shows the gene silencing efficiency in HeLaS3 cells transfected with 100 nM modified siRNA according to the present invention and of T98G cells transfected with 10 nM modified siRNA according to the present invention after 2 days and 6 days, respectively;

Fig. 7 shows the gene silencing efficiency in HeLaS3 cells transfected with 100 pM and 50 pM modified siRNA according to the present invention, respectively, with HiPerFect being used as a transfection reagent;

Fig. 8 shows the HPLC spectrum of siRNA modified according to the present invention, incubated in medium with 10 % serum (Example 58); Fig. 9 shows the stability of siRNA (non-modified versus modified according to the present invention) over 72 h; from which it is evident that transfection agents can only slightly prevent the degradation of non-modified siRNA; Fig. 10 shows a PAGE which shows that modified siRNAs are stable in 100% human serum for at least 8h, while non-modified siRNA is degraded within minutes.

Examples 1 to 95

The siRNA molecules tested did predominantly comprise 19 nt as the core sequence. In the following Table 1, the core sequence is given between the two straight lines. Left and right of the two straight lines, the 3'- and 5' overhangs are indicated. Moreover, the Table indicates the strand name and the name of the siRNA duplex. The sense strand of each paragraph forms a duplex with one of the antisense strands of the same paragraph. The top row of each paragraph denotes the sense strand (e.g. J8-S, MAPK2S etc.), and the further rows of each paragraph represent the antisense strands (e.g. J8-AS, MAPK2AS etc.). The siRNA duplex name is abbreviated with "exp 1" to "exp 95". The following further symbols are used in

Table 1:

A, C, G, U: nucleotides with nucleobases adenine, cytosine, guanine, uracil

dA, dC, dG, dT, dU: 2 '-deoxy-modified nucleotides mA, mC, mG, mT, mU: 2'-methoxy-modified nucleotides

3 'to 5'formacetal linkage

U ^2'A5' U : 2'to 5 ^' formacetal linkage

Ac2, Cc2, Gc2, Tc2, Uc2: 2 '-hydroxyethylmethyl-modified nucleotides

Uc8: 3 '-octyloxymethylmodified nucleotides - MAPK2S and MAPK2AS: nonmodified MAPK2 sense and antisense respectively

U ^{3 5} U represents a 3' to 5' formacetal linkage as shown below in formula A:

FORMULA A represents a 2' to 5' formacetal linkage as shown below in formula B:

FORMULA B Ac2, Cc2, Gc2, Tc2, Uc2 represent 2'-hydroxyethylmethyl- modified nucleotides according to formula C shown below:

FORMULA C

(B = nucleobase: adenine, cytosine, thymine, guanine, uracil, respectively) Uc8 represents 3'-octyloxymethyl modified nucleotides according to formula D shown below:

FORMULA D The compound according to formula D is connected via a phosphodiester to the 5 '-oxygen of the nucleoside

overhang core sequence overhang strand siRNA

name duplex SEQ

name ID*

1 (s/m1)

5 ^' - U G C U G A C U C C A A A G C U C U G u ^3' " ^5' u -3 ^' J8-S

1 (a/m2)

3 ^' - U ^{3 '} " ^{5 '} U A C G A C U G A G G U U U C G A G A C -5 ^' J8-AS exp 2

3 ^' - dU dU A C G A C U G A G G u U U C G A G A C -5 ^' APK2AS exp 5 1 (a/m3) m4)

5 ^' - U G C U G A C U C C. A A A G C U C U G υ ^2'Λ5' υ 1 (s/

-3 ^' J7-S

1 (a/m5)

3 ^' - υ ^2'Λ5' υ A C G A C U G A G G U U U C G A G A C -5 ^' J7-A exp 1

MAPK2A 1 (a/m6)

3 ^' - dU dU A C G A C U G A G G U U U C G A G A c -5 ^' S exp 3

5 ^' - U G C U G A C U C C A A A G C U C U G dU dU -3 ^' AP 2S 1 (s/m7) 5 ^' υ 1 (a/m8)

3 ^' - υ ^2'Λ A C G A C U G A G G U U U C G A G A C -5' J7-AS exp 4

3 ^' - υ ^3'Λ5 A C G A C U G A G G U U u 1 (a/m9)

C G A G A C -5 ^' J8-AS exp 6

3 ^' - υ ^3'Λ5 1 (a/m1 mA C G A C U G A G G U U u C G A G A C -5 ^' J9-AS exp 9

1 (s/m11

5 ^' - U G C U G A C U C C A A A G C U C U G υ ^3'Λ5' υ -3 ^' J9-S

1 (a/m

3 ^' - U ^{3 '} " ^{5 '} U mA C G A C U G A G G U U U C G A G A C -5 ^' J9-AS exp 7

MAP 2A 1 (a/m1

3'- dU dU A C G A C U G A G G U U U C G A G A C -5 ^' S exp 8

1 (s/m1

5'- mA mA U G C U G A C U C C A A A G C U C U G υ ^3'Λ5' υ -3 ^' J10-S

3 ^' - υ ^3'Λ5' υ 1 (a/m1

A C G A C U G A G G U U U C G A G A C -5 ^' J8-AS exp 10

3 ^' - mU mU A C G A C U G A G G U U U C G A G A C -5' J15-AS exp 11 1 (a/m1

3 ^' - mU U A C G A C u G A G G U U U C G A G A C -5 ^' J16-AS exp 12 1 (a/m1

APK2A 1 (a/m1

3 ^" - dU dU A C G A C u G A G G U U u C G A G A C -5 ^' S exp 13

1(s/m18

5'- mA mA U G C U G A C U C C A A A G C U C U mG υ ^3'Λ5' υ -3' J13-S

overhang core sequence overhang strand siRNA

name duplex SEQ

name ID*

1(a/m19

3 ^' - U ^{3 '} " ^{5 '} U A C G A C U G A G G U U U C G A G A C -5 ^' J8-AS exp 4

3 ^' - mU mU A C G A C U G A G G U U U C G A G A C -5 ^' J15-AS exp 5 1(a/m20

3 ^' - mU U A C G A C U G A G G u u U C G A G A C -5' J16-AS exp 16 1(a/m21

ΜΑΡΚ2Α 1(a/m22

3 ^' - dU dU A C G A C U G A G G u u U C G A G A C -5' S exp 17

5'- mA mA U G C U G A C U C C A A A G C U C U mG mU mU -3' J11-S 1(s/m23)

1 (a/m24

3'- U ^{3 '} " ^{5 '} U A C G A C U G A G G U U U C G A G A C -5' J 8- AS exp 8

3 - mU mU A C G A C U G A G G U U U C G A G A C -5' J 15- AS exp 19 1(a/m25

3 ^' - mU U A C G A C U G A G G U U U C G A G A C -5 ^' J16-AS exp 20 1 (a/m26

MAPK2A 1 (a/m27

3'- dU dU A C G A C U G A G G U U U C G A G A C -5 ^' S exp 21

5 ^' - mA mA U G C U G A C U C C A A A G C U C U mG mU mU -3' J17-S 1(s/m28)

1 (a/m29

3 ^' - U ^{3 'A5 '} U A C G A C U G A G G U U U C G A G A C -5' J8-AS exp 22

3 ^' - mU mU A C G A C U G A G G U U U C G A G A C -5 ^' J15-AS exp 23 1 (a/m30

3'- mU U A C G A C U G A G G U U U C G A G A C -5' J16-AS exp 24 1 (a/m31

MAPK2A 1(a/m32

3 ^' - dU dU A C G A C U G A G G U U U C G A G A C -5' S exp 25

U C C U C C 33)

5 ^' - mA mA G U G A A A A G C dU dC dU mG υ ^3'Λ5' υ 1

-3' J18-S

' ^Λ5' υ 1(a/m34

3 ^' - υ ³ A C G A C U G A G G U U U C G A G A mC -5' J22-AS exp 26

1 (a/m35

3'- U ^{3 '} " ⁵ U A C G A C U G A G G U U U C G A G A dC -5' J23-AS exp 27

υ ^3'Λ5' υ 1 (a/m36

3'- A C G A C U G A G G U U U C G dA dG dA dC -5' J26-AS exp 28

3'- mU mU A C G A C U G A G G U u U C G A G A mC -5 ^' J27-AS exp 29 1 (a/m37

3'- mU mU A C G A C U G A G G u u U c G A G A dC -5' J28-AS exp 30 1 (a/m38

3'- mU mU A C G A C U G A G G u u u c G dA dG dA dC -5' J29-AS exp 31 1(a/m39

3 ^' - Uc8 mU A C G A C U G A G G u u u c G A G A mC -5 ^' J24-AS exp 32 1(a/m40

3 ^' - Uc8 mU A C G A C U G A G G u u u c G A G A dC -5 ^' J25-AS exp 33 1(a/m41)

5 ^" - dA dA U G C U G A C U C C A A A G C dU dC dU mG υ ^3'Λ5' υ 1(s/m42)

-3' J19-S

overhang core sequence overhang strand siRNA

name duplex SEQ

name ID*

1(a/m43

3 ^' - υ ^{3 'Λ5'} υ A C G A C U G A G G U U U c G A G A mC -5' J22-AS exp 34

1 (a/m44

3'- U ^{3 'A5'} U A C G A C U G A G G u U U c G A G A dC -5' J23-AS exp 43

1 (a/m45

3 ^' - υ ^3'Λ5' υ A C G A C U G A G G u u U c G dA dG dA dC -5 ^' J26-AS exp 38

3'- mU mU A C G A C U G A G G u u u c G A G A mC -5' J27-AS exp 39 1(a/m46

3'- mU mU A C G A C U G A G G u u u c G A G A dC -5 ^' J28-AS exp 40 1(a/m47

3'- mU mU A C G A C U G A G G u u u c G dA dG dA dC -5 ^' J29-AS exp 41 1(a/m48

3'- Uc8 mU A C G A C U G A G G u u u c G A G A mC -5 ^' J24-AS exp 51 1 (a/m49

3 ^' - Uo8 mU A C G A C U G A G G u u u c G A G A dC -5 ^' J25-AS exp 37 1 (a/m 50

1 (s/m51)

5 ^' - mA mA U G C U G A C U C C A A A G C dU dC dU dG υ ^3'Λ5' υ -3 ^' J20-S

1(a/m52

3 ^' - U ^{3 'A5'} U A C G A C U G A G G u U u C G A G A mC -5 ^' J22-AS exp 42

(a/m53

3 ^' - U ^{3 'A5'} U A C G A C U G A G G u U u c G A G A dC -5' J23-AS exp 35 1

1 (a/m54

3 ^' - U ^{3 'A5 '} U A C G A C U G A G G u u u c G dA dG dA dC -5 ^' J26-AS exp 46

3 ^' - mU mU A C G A C U G A G G u u u c G A G A mC -5 ^' J27-AS exp 47 1 (a/m55

3 ^' - mU mU A C G A C U G A G G u u u c G A G A dC -5 ^' J28-AS exp 48 1(a/m56

3'- mU mU A C G A C U G A G G u u u c G dA dG dA dC -5 ^' J29-AS exp 49 1(a/m57

3 ^' - Uc8 mU A C G A C U G A G G u u u c G A G A mC -5 ^' J24-AS exp 44 1(a/m58

3'- Uc8 mU A C G A C U G A G G u u u c G A G A dC -5 ^' J25-AS exp 45 1(a/m59

1(s/m60)

5'- dA dA U G C U G A C U C C A A A G C dU dC dU dG υ ^'Λ5' υ -3 ^' J21-S

1(a/m61

3'- U ^{3 ' A5 '} U A C G A C U G A G G U U U C G A G A mC -5 ^' J22-AS exp 50

1(a/m62

3 ^' - U ^{3 ' 5 '} U A C G A C U G A G G u U U C G A G A dC -5 ^' J23-AS exp 36

1 (a/m62

3'- U ^{3 'A5 '} U A C G A C U G A G G u U u c G dA dG dA dC -5 ^' J26-AS exp 54

3 ^' - mU mU A C G A C U G A G G u U u c G A G A mC -5 ^' J27-AS exp 55 1(a/m63

3 ^' - mU mU A C G A C U G A G G u u u c G A G A dC -5 ^' J28-AS exp 56 1(a/m64

3'- mU mU A C G A C U G A G G u u u c G dA dG dA dC -5' J29-AS exp 57 1 (a/m65

3 ^' - Uc8 mU A C G A C U G A G G u u u c G A G A mC -5 ^' J24-AS exp 52 1 (a/m66

3 ^' - Uc8 mU A C G A C u G A G G u u u c G A G A dC -5 ^' J25-AS exp 53 1(a/m67

5'- mA mA U G C U G A C U C C A A A G C U C mU mG υ ^3'Λ5' υ 1(s/m68)

-3' J30-S

overhang core sequence overhang strand siRNA

name duplex SEQ

name ID* υ ^3'Λ5' υ 1 (a/m69)

3 ^' - A C G A C U G A G G U U U c G A G A mC -5 ^' J22-AS exp 58

3 ^' - mU mU A C G A C u G A G G U U U c G A G A mC -5' J27-AS exp 59 1(a/m7Q)

3'- mU mU A c G A C u G A G G U U U c G A G mA mC -5 ^' J35-AS exp 60 1 (a/m71)

1{a/m72)

3 ^' - υ ^3'Λ5' υ A c G A C u G A G G U U U c G A G mA mC -5 ^' J36-AS exp 73

1 (a/m73)

3'- υ ^3'Λ5' υ mA c G A C u G A G G U U U c G A G A mC -5 ^' J37-AS exp 74

3'- υ ^3'Λ5' υ mA c G A C u G A G G U U U c G A G mA mC -5' J38-AS exp 75 1(a/m74)

1(sfrn75)

5'- mA mA dU dG dC dU dG dA dC dU dC dC dA dA dA dG dC dU dC mU mG u ^3' " ^5' u -3 ^' J31-S

1 (a/m76

3'- υ ^3'Λ5' υ A C G A C U G A G G U U U C G A G A mC -5 ^' J22-AS exp 61

3'- mU mU A C G A C U G A G G U U U c G A G A mC -5 ^' J27-AS exp 62 1(a/rr,77)

3'- mU mU A C G A c U G A G G U U U c G A G mA mC -5 ^' J35-AS exp 63 1 (a/m78

5'- mA mA mU dG dC dU dG dA dC dU dC dC dA dA dA dG dC dU mC mU mG υ ^3'Λ5' υ 1

-3 ^' J32-S (8^79) Λ ^5' υ 1 (a/m80)

3'- υ ^' A C G A C u G A G G U U U C G A G A mC -5 ^' J22-AS exp 64

3'- mU mU A C G A C u G A G G U U U C G A G A mC -5 ^' J27-AS exp 65 1(a/m81)

3'- mU mU A C G A C u G A G G U U U C G A G mA mC -5 ^' J35-AS exp 66 1 (a/m82)

1 (s/m83)

5'- mA mA dU dG dC dU dG dA dC dU dC dC dA dA dA dG dC dU mC mil mG υ ^3'Λ5' υ -3' J33-S

υ ^3'Λ5' υ 1(a/m84)

3 ^' - A C G A C U G A G G U U U C G A G A mC -5 ^' J22-AS exp 67

3 ^' - mU mU A C G A C U G A G G U U U C G A G A mC -5 ^' J27-AS exp 68 1(a/m85)

3'- mU mU A C G A C U G A G G U U U c G A G mA mC -5' J35-AS exp 69 1(a/m86)

5'- mA mA mU 1(s/m87) dG dC dU dG dA dC dU dC dC dA dA dA dG dC dU dC mU mG u ^3' ~ ^5' u -3 ^' J34-S

3'- υ ^3'Λ5' υ A C 1 (a/m88)

G A C U G A G G U U U C G A G A mC -5' J22-AS exp 70

3 ^' - mU mU A C G A C u G A G G U U U C G A G A mC -5 ^' J27-AS exp 71 1(afrn89)

3'- mU mU A C G A C u G A G G U U U C G A G mA mC -5 ^' J35-AS exp 72 1(a/m90)

5 ^' - mA mA U G C U G A C U C C A A A G C Uc2 Cc2 Uc2 Gc2 υ ^3'Λ5' υ -3' J47-S 1 (s/m91)

3'- υ ^3'Λ5' υ A C G A C U G A G G U U U C G A G A Cc2 -5' J49-AS 1(ami92) exp 88

overhang core sequence overhang strand siRNA

name duplex SEQ

name ID*

1(a/m93)

3 ^' - U ^{3 ' A5 '} U A C G A C U G A G G U U u c G A G A dC -5 ^' J23-AS exp 90

3'- υ ^3'Λ5 mA C G A C u G A G G U U u c G A G A dC -5' J46-AS exp 94 1(a/m94)

5 ^' - mA mA U G C U G A C U C C A A A G C Uc2 Cc2 Uc2 mG υ ^3'Λ5' υ 1(s/m95)

-3 ^' J48-S

1(a/m96)

3 ^' - υ ^{3 'Λ5'} υ A C G A C U G A G G U U U C G A G A Cc2 -5 ^' J49-AS exp 89

1(a/m97)

3 ^' - υ ^3'Λ5 A C G A C U G A G G U U U C G A G A dC -5 ^' J23-AS exp 91

1(a/m98)

3 ^' - U ³ '" ^{5 '} U mA C G A C U G A G G U U U C G A G A dC -5' J46-AS exp 95

5 ^' - dA dA U G C U G A C U C C A A A G C dU dC 1(s/m97) dU dG υ ^3' ~ ^5' υ -3' J45-S

υ ^3'Λ5 1(a/m98)

3 ^' - mA C G A C U G A G G U U U C G A G A dC -5 ^' J46-AS exp 92

3 ^' - υ ^3'Λ5' υ A C G A C U G A G G U U U C G A G A dC -5' J23-AS exp 93 1(a/m99)

Table 1

* : SEQ ID: In this field, the number in front of the parenthesis indicates the SEQ ID NO as listed in field <210> of the sequence listing. The letters "s" or "a" indicate whether the strand is a sense strand (s) or the respective antisense strand (a). The letter "m" in combination with the respective number represents a consecutive numbering of the respective modification of this sense or antisense strand as outlined in field <223> of the sequence listing.

From Table 1, the modifications of siRNA duplexes can be seen. Long chains or other major modification, for example a 2'-acetal modification (modification (d)). on the 5 '-end of the antisense strand, are not tolerated by the siRNA. On the 5 'end of sense strand, any kind of modification has been tolerated. On the 3 'end of the sense and the antisense strand, all kinds of modifications (a) to (e) were also tolerated. Deoxy-modifications on the whole sense strand in combination with the methoxy-modifications on 5'- and 3 '-end of this strand (see e.g. Exp. 61- 72) can lead to siRNA inactivation. One or more methoxy modification(s) of the 5'-end of the antisense strand in combination with methoxy-modification(s) on the 3 '-end of the antisense strand can decrease siRNA activity (see Exp. 60, Exp. 73-75).

While Table 1 refers to siRNA sequences for the target MAPK2, Table 2 shows siRNA sequences for the targets Lamin AC and CDC2, respectively. Otherwise, what was said with regard to Table I, also applies to Table 2.

Table 2

siRNA sequences for the targets Lamin AC and CDC2. overhang core sequence overhang strand siRNA SEQ

name duplex ID*

name

5 ^' - mA mA C U G G A C U U C C A G A A G dA dA dC mA υ ^3'Λ5' υ 2(s/m1)

-3' J45-S

Lamin 2(a/m2)

3 ^' - υ ^3'Λ5' υ G A C C U G A A G G U C U U C U U G dU -5' J46-AS exp27

Lamin

3 ^' - mU mU G A C C U G A A G G U C U U C U U G dU -5 ^' J23-AS exp30 2(a/m3)

5 ^' - mU mG G G G U C A G C U C G U U A C dU dC dA mA υ ^3'Λ5' υ -3 ^' J45-S 3(s/m1)

3CDC2

3'- mA mC C C C A G U C G A G C A A U G A G U dU -5 ^' J46-AS exp30 3(a(m2)

4(s/m1)

5 ^' - mU mG G G G A C A C C U C G U A A C dU dC dA mA υ ^3'Λ5' υ -3 ^' J45-S

3CDC2 mut3

3'- mA mC C C C U G U G G A G C A U U G A G U dU -5 ^' J46-AS exp30 4(a/m2)

5 ^' - mA mA U G G C A C U G A A U C A U C dC dA dU mA υ ^3'Λ5' υ -3 ^' J45-S 5(s/m1)

1 CDC2 5(a/m2)

3 ^' - U ^{3 ' A5 '} U A C C G U G A C U U A G U A G G U A dU -5 ^' J46-AS exp27

1 CDC2

3 ^' - mU mU A C C G U G A C U U A G U A G G U A dU -5 ^' J23-AS exp30 5(a/m3)

*: SEQ ID: In this field, the number in front of the parenthesis indicates the SEQ ID NO as listed in field <210> of the sequence listing. The letters "s" or "a" indicate whether the strand is a sense strand (s) or the respective antisense strand (a). The letter "m" in combination with the respective number represents a consecutive numbering of the respective modification of this sense or antisense strand as outlined in field <223> of the sequence listing.

With the modified siRNA molecules (exp 1 to exp 95), tests were carried out to determine their ability to perform RNAinterference (RNAi). The ability of siRNA coupled with peptide sequences to induce silencing of the corresponding gene was tested compared to standard siRNA transfections using reagents based on the lipid technology. For this purpose LaminA/C, MAP 2/ERK2 (NM_002745.2 ), and CDC2 (NM_001786) genes were used as target to perform the silencing experiments. LaminA/C-, MAPK2(ERK2)-, and CDC2- siRNAs were modified and transfected to different cells lines. The silencing results were compared to silencing effects induced by the corresponding non-modified siRNAs. Further controls were non-treated cells and cells transfected with non-specific siRNA (GFP22).

The following cell lines were used for the transfections: Human cervix carcinoma; HeLaS3, human breast carcinoma; MCF7, human glioblastoma; T98G. Human embryonic kidney; 293 cell-Line. Fig. 2 shows the transfection of HeLAS3 cells with 100 nM modified siRNAs 48 hours after transfection (RNAiFect ^® was used).

The culture conditions prior to transfection were as follows. 24 h before transfection cells were seeded in a density of 5-6 x 10 ⁴ cells/well. HeLaS3, T98G and 293 cell lines were incubated in standard DMEM complemented with 10 % FCS (Fetal Calf Serum), at 37°C and 5 % C0 ₂ atmosphere. MCF7 cells were incubated in RPMI complemented with 10 % FCS, insulin, glutamine and sodium pyruvate, otherwise under the same conditions as the other cell lines.

Transfection with RNAiFect ^® was carried out as follows (RNAiFect ^® is a trademark of QIAGEN GmbH, Hilden, Germany). 0.15-0.6 g siRNA was diluted in complete medium to a final volume of 100 μί and mixed by a Vortex mixer ("vortexing"). RNAiFect was added in a ratio of 1:6 (RNA ^g) : Reagent (μΐ)) to the mixture, was mixed by vortexing and incubated for 15 min at room temperature. While complex formation was taking place, the medium was aspirated from the plate and 300μ1 growth medium with serum was added onto the cells. After the complex formation was completed, the transfection complexes were added drop by drop onto the cells. The plate was swirled gently and cells were incubated for 48h under normal growth conditions. No medium change was required.

Transfection with HiPerFect was carried out as follows (HiPerFect is a transfection reagent available from QIAGEN GmbH, Hilden, Germany). 0.375-37.5 ng siRNA (dependent on the transfected cell line) were diluted in 100 μΐ siRNA Dilution Buffer (100 niM potassium acetate, 30 mM HEPES-KOH, 2 mM Mg(OAc) _2; pH = 7.4) and mixed by vortexing. 3 μΐ HiPerFect were added to the mixture, further mixed and incubated for 15 min at room temperature on a plate. While complex formation was taking place, the medium was aspirated from the plate and 500 μΐ growth medium with 10 % serum was added onto the cells. After the complex formation was complete, the complex was added drop wise onto the cells. The plate was swirled gently and cells were incubated for 48 h or longer under normal growth conditions (37°C, 5 % C0 ₂ ). The modifications to the siRNA molecules were also monitored for possible toxic effects on the cells. The cell toxicity test was carried out as follows. The Roche LDH cytotoxicity detection kit was used for this purpose. The culture supernatants were collected before cell lysis and lactate dehydrogenase (LDH) was measured. The amount of LDH activity released from the cytosol of damaged cells into the supernatant was used to monitor cell toxicity.

In addition, the transfection efficiency was analyzed. 48 h after the transfection the medium was aspirated from the plates, the cells were lysed in 350 μΐ RLT buffer and the total RNA was prepared using R easy 96 plate and RNeasy 96 protocol ("RNeasy 96" are products available from QIAGEN GmbH, Hilden, Germany).

2 μΐ RNA were amplified in a one tube RT-PCR using TaqMan Primer and Probes ("TaqMan probe ^® " is a trademark of Applied Biosystems; Foster City, CA, USA). Amplification was performed in parallel for the target gene and GAPDH. Target expression was analyzed as quotient of the GAPDH expression. The absolute target expression of the treated cells using modified siRNA was compared to the absolute target expression of the treated cells using standard (i.e. non-modified) siRNA. The targets used were LaminA/C , MAPK2 and CDC2.

Primers and Probes used in QuantiTect Duplex RT-PCR analysis ("QuantiTect" is an assay for gene expression analysis using quantitative, real-time RT-PCR available from QIAGEN GmbH, Hilden; Germany; "F" and "For" denotes "Forward", "R" and "Rev" denotes "Reverse"):

Primers Sequence

LaminA_776F GGC GGG TGG ATG CTG AGA ACA (SEQ ID NO:6) LaminA_881R TGT CAA TCT CCA CCA GTC GGG (SEQ ID NO:7)

MAPK2_F CCT TCC AAC CTG CTG CTC AAC AC (SEQ ID O:8) MAPK2JR. GCC ACA TAT TCT GTC AGG AAC CC (SEQ ID NO:9) CDC2_F AAT AAG CCG GGG ATC TAG CAT AC (SEQ ID NO: 10) CDC2 R TTT CAT GGC TAG CAC TTG ACC TG (SEQ ID NO: 11) hGAPDH-TM-For GAA GGT GAA GGT CGG AGT (SEQ ID NO: 12) hGAPDH-TM-Rev GAA GAT GGT GAT GGG ATT TC (SEQ ID NO: 13)

TaqMan Probes: Sequence:

LaminA 838TM 5 '-FAM-ATCTACAGTGAGGAGCTGCGTGAGA-3 '-BHQ

(SEQ ID O:14)

MAPK2 TM 5 -FAM-TGGCCCGTGTl GCAGATCCAGAC-3 '-BHQ

(SEQ ID NO: 15)

CDC2 TM 5 ^' -FAM-ATGGAGTTGTGTATAAGGGTAGAC-3 '-BHQ

(SEQ ID NO: 16)

hGAPDH-TM 5 '-HEX-CAAGCTTCCCGTTCTCAGCCT-3 '-BHQ

(SEQ ID NO: 17)

(BHQ = Black Hole Quencher; FAM and HEX = Fluorophores; oligonucleotides are synthesised and modified with fluorophores by Operon) As a next step the gene silencing efficiency of the siRNA molecules according to the present invention were tested. For this purpose, a screening of modified siRNAs was carried out. Fig. 2 shows the result of this screening experiment. In Fig. 2, the MAPK2 (ERK2) expression in HeLaS3 cells is shown. The "No" bar in Fig. 2 represents untreated cells. "GFP" and "ERK2" represent control transfections with nonspecific siRNA and non modified siRNA, respectively. Exp 1-95 represent the residual MAPK2 expression in cells transfected with different modified MAPK2-siRNAs (the type of the modification can be seen from Table 1). Each of the bars for Exp. 1 to Exp. 95 represents the mean value of four measurements. The line indicated at the top of each bar represents the standard deviation for each of Exp. 1 to 95. The cells were tested also for cytotoxicity. None of the tested siRNAs was found to be cytotoxic.

As can be seen from Fig. 2, the residual MAPK2 expression in HeLaS3 cells transfected with the different modified MAPK2-siRNAs according to Exp. 27, 30, 33, 35, 43, and 53 (see Table 1 for exact structures) show a convincing gene silencing effect. Taking furthermore into account the stability of the respective siRNAs against nuclease attack, Exp. 27, 30, 33, 35, 43, and 53 have shown the best performance of the modified siRNAs tested.

It can be seen from the above that siRNAs with end modifications do not negatively affect the gene silencing mechanism. Depending on the amount and type of modifications, siRNAs are obtained that enhance the siRNA activity. The screening was performed also with 25 nM siRNA to visualize and specify slight differences between the modified siRNAs (data not shown). The data with 25 nM siRNA confirmed the data obtained with 100 nM. In the following , the best performing siRNAs were used to test sensitivity and longevity of siRNA.

Fig. 3 shows the results of experiments in which transfection of HeLAS3 cells with MAPK2- siRNAs in the HiPerFect transfection system was carried out with lower siRNA concentrations than for the experiments shown in Fig. 2 (100 pM in Fig. 3 compared to 100 nM in Fig. 2). As can be seen, the difference in performance (gene silencing) between modified and non-modified siRNA is becoming greater at very low siRNA concentrations (100 pM). The modified siRNA according to the present invention induces high gene silencing levels even at concentrations in the pmolar range.

In addition, the gene silencing efficiency (potency/sensitivity) of the siRNA according to the present invention has been investigated. Different cell lines were transfected using RNAiFect or HiPerFect and different siRNA amounts. Very low amounts of siRNA were used to estimate the siRNA sensitivity. Specifically, the gene silencing efficiency of the best performing modified siRNAs (Exp. 27, 30, 33, 35, 43, and 53) has been tested further. Fig. 4 represents the MAPK2 expression in HeLaS3, MCF7, and T98G cells 48 h after transfection with different concentrations of modified and non modified siRNAs using RNAiFect or HiPerFect. The amount of siRNA, which was sufficient to induce silencing, and which was used for the transfections was estimated for each of the cell lines HeLaS3, MCF7, and T98G.

In Fig. 4, the "no treatment" bar represents untreated cells. "Non specific siRNA" and "non modified siRNA" represent control transfections with nonspecific siRNA and non modified siRNA, respectively. Exp. 27, Exp. 30, Exp. 33, Exp. 35, Exp. 43, and Exp. 53 represent the remaining MAPK2 expression in cells transfected with different modified MAPK2-siRNAs. The modified siRNA shown on these graphs (Exp. 27, 30, 33, 35, 43, and 53) are those, which have shown the best performance in terms of gene silencing and stability against nucleases. From the results shown in Fig. 4 it can be concluded that the chosen modified siRNAs (Exp. 27, 30, 33, 35, 43, and 53) perform better than non modified, regardless of transfection reagent or cell line. Again it can be seen that the difference in performance between modified and non modified siRNA is getting greater at very low siRNA concentrations. Modified siRNA induce high silencing levels even with concentrations in the pmolar range. This is also evident from the test results shown in Fig. 3. This Fig. 3 refers to HeLaS3 cells transfected with MAPK2-siRNA using the HiPerFect transfection support system. As can be seen from Fig. 5, the modified siRNA according to the present invention (Exp. 27, Exp. 30, Exp. 33, Exp. 35, Exp. 43, and Exp. 53) are very effective in an amount of 100 pM only compared to a respective amount of unmodified MAPK2 specific siRNA. E.g., Exp. 30 and Exp. 43 were more than 10 times more effective than the respective non-modified siRNA.

Fig. 5 shows that an siRNA modified according to the present invention can change a non- potent siRNA into a potent one at reasonable concentrations between 100 pM and 5 nM. The gene expression can be down-regulated with the modified siRNAs according to the present invention to about 10 % of the value obtained with non-treated siRNAs (10 % for 5 nM of the siRNA modified according to Example 27 and 12 % for 5 nM of the siRNA modified according to Example 30).

As a next step, the gene silencing efficiency of the inventive siRNA has been tested in view of the siRNA longevity. The duration of the silencing effects of the best performing modified siRNAs (Exp. 27, 30, 33, 35, 43, and 53) were tested in HeLaS3 and T98G cells. HeLaS3 cells were transfected with RNAiFect and 100 nM MAPK2-siRNA and HiPerFect and 10 nM MAPK2-siRNA, respectively. MAPK2 -down-regulation was measured 2 and 6 days after transfection. The modified siRNAs, which were used to induce long term silencing, are compared to non-modified siRNA. The results are shown in Fig. 6. As can easily be seen from these results, siRNAs modified in accordance with the present invention perform better in terms of efficiency and longevity independent of the transfection system used.

Furthermore, the gene silencing efficiency of the inventive siRNAs have been tested using different targets. Fig. 7 represents gene silencing of Larnin A/C and CDC2 in HeLaS3 cells. LaminA/C- and CDC2-siRNA were modified using the best performing modifications (Exp. 27, Exp. 30). Again "No treatment" bars represent untreated, normal growing cells. "Non specific" and "non modified" represent control transfections with nonspecific siRNA and non modified siRNA, respectively. Exp. 27 and Exp. 30 represent the remaining Lamin/AC or CDC2 expression in cells, which were transfected with the corresponding modified siRNA. The conclusion that can be drawn from the results in Fig. 7 is that the modifications used improved the gene silencing performance of active siRNAs and slightly enhanced the siRNA efficiency of moderate active siR As. Mutated-negative control siRNA still remains inactive, a false activity due to modifications is excluded.

Furthermore, the stability of the siRNA according to the present invention in cell culture medium has been tested. For this purpose, the stability in 10 % Fetal Calf Serum (FCS) has been determined by HPLC. In the following Table 3, the amount of siRNA remaining after t = 4 h, 24 h, 48 h, and 72 h in percent is given (t = 0 h is 100 %).

Table 3

From the results shown in Table 3 above as well as the data given in Figs. 8 and 9, it can be seen that the modifications of the siRNA used dramatically enhance the siRNA stability in cell culture media (containing 10 % serum). Fig. 8 shows an HPLC analysis of traces of the modified siRNA (Exp. 58) in cell culture medium with 10 % serum after 4 h, 24 h, 36 h, 48 h, and 72 h (= 3 d). From Fig. 9, the stability of siRNA in cell culture medium before and after complexing with transfection reagent is shown (modified according to Exp. 28; transfection reagent RNAifect ^® trademark of the QIAGEN GmbH, Hilden, Germany; "w/" with transfection agent, "w/o" without transfection agent). Complexing with transfection reagents slightly prevents siRNA degradation, but still a high an unacceptable amount of unmodified siRNA is cleaved by serum nucleases (Fig. 10, unmodified MAPK2 w/RNAifect). The modifications according to the present invention assure almost full length siRNAs even after three days incubation in cell culture media (see Figs. 8 and 9), thereby preventing off-target effects due to shortened (and therefore more unspecific) siRNA.

Finally, the stability of the siRNAs according to the present invention in 100 % human serum has been tested. Specifically, siRNAs with different sequences were tested for their stability in pure human serum. Examples of siRNAs with modifications according to Exp. 30, Exp. 27, Exp. 84, Exp. 90 and Exp. 94 were compared with non modified siRNA. The results are shown in Fig. 10. It can be seen from the Fig. 10 that the stability of siRNA is strongly depended on the sequence of the tested siRNA. A defined cleavage has been obtained after incubation in 100 % human serum. This cleavage has been dramatically slowed down due to the chemical modifications accordmg to the present invention. After 24 h of incubation, full length modified siRNA is still present, while non-modified siRNA is being completely degraded within minutes.

Screening of additional modified siRNAs

Additional chemical modifications were tested for their effects in the performance of the siRNA in terms of siRNA potency, longevity and stability. The sequences and modifications are listed in the following table. Additional tested modifications

siRNA duplexes (top row of each paragraph: sense-strand, followed by antisense strands)

Legend: dA, dC, dG, dT, dU: 2'-deoxy-modified nucleotides

mA, mC, mG, mT, mU: 2"-met oxy-modifIed nucleotides

U ³ ' ^AS 'U : 3 'to 5 'formacetal linkage

U≠ ^'*5' u : 2 ^' to 5 'formacetal linkage

Ac2, Cc2, Go2, Tc2, Uc2: 2'-acetal-modified nucleotides

Ac2n, Cc2n, Gc2n, Tc2r», Uc2n: 2'-acetaI-amfne-modified nuolotides

t n Dl

ffi Φ

c C

S o

& α>

3 < 3 < «£ < ¾ U

3 3 3 < < 3 < 3 < 3 -< ϋ CD ο D CD Β O CD α CD O CD CD O CD ο

CD Ο CD Ο Ο ϋ ο ο a o

CD

< -3 < 3 < 3 S 3 < 3 < 3 < 3

< 3 < =3 < 3 § 3 < < < 3

< 3 < 3 < 3 < 3 < 3 < 3 < 3 < < 3 ο ο ϋ CD o ο ο CD O CD ο Ο CD ϋ ο ϋ Ο O CD

3 < 3 => < 3 < 3 < 3 < < 3 <

Ο CD ϋ Ο ϋ c? ϋ CO CD ϋ (!) ο t?

< 3 < =ι < 3 < = < 3 < 3 < < 3 < 3 CD Ο CD ο CD ϋ CD ϋ CD O CD O CD ϋ CD O

3 < 3 < 3 < 3 < < < 3 <

Ο CD Ο CD O CD O CD ϋ ο a a

CD CD

CD Ο CD Ο CD O CD O ο ο ο ϋ te g Τ3 if

in In <n in « to « a.

x

CM

U

a a u o ffl g

CO

in

-5

eg

3 .3

< r> < s 3 <

¾

o © O ©

Q

-3 < ¾ 3 <

O o ©

< <

3 <

3 < < ϋ C3 u a O CD

_-> < 3 < a © ϋ ©

© o CO o

< =1 <

CD o CD > < <

CO O o OP O © o

© O © o

U S

C5

< <

¾

3 < ¾ < Screening of modified siRNAs (Exp 96-121) Silencing efficiency: potency

The silencing efficiency of the siRNA was tested by transfecting cells with the siRNAs and quantifying the expression of the specific silenced targets by quantitative RT-PCR as describe in material and methods. Fig.11 represents the MAPK2 (ERK2) expression in HeLaS3.

The No treatment bar represents untreated cells. GFP and ERK2 represent control transfections with nonspecific siRNA and non modified siRNA respectively. Exp 96-121 represent the MAPK2 rest expression in cells transfected with different modified MAPK2- siRNAs (see table 1). The cells were tested also for cytotoxicity. None of the tested siRNAs was found to be cytotoxic (data not shown).

Conclusion

The additional tested chemically modified siRNAs perform better than non modified siRNAs. Silencing of higher degree is obtained at very low concentrations of siRNA compared non modified siRNA.

If the modification design, that was used for the sense strand is applied to modify the antisense strand this leads to inactivation of the siRNA (e.g. Expl02neg, Expl04neg and Exp 120).

Thus the choice of the given design can be used to inactivate the sense strand.

It is known that the sense strand may lead to sequence specific off target effects. Inactivation of the sense strand by the given modifications may lead to elimination of sequence specific off target effects.

Silencing efficiency: siRNA longevity

The duration of the silencing effects of the modified siRNAs was tested in HeLaS3 cells (Fig.12). HeLaS3 cells were transfected with HiPerFect and 250pM MAPK2-siRNA. MAPK2-downregulation was measured 1, 4 and 6 days after transfection. The modified siRNAs (Exp96-114), which were used to induce long term silencing, are compared to non modified siRNA indicated by (ERK2 250pM). No treatment and GFP represent controls of untreated cells and cells treated with non specific siRNA respectively.

Conclusion:

Modified siRNAs Exp 27, Exp 97-98, Exp 101-104, Expl l l and Exp 114 induce a more efficient and prolonged silencing compared to non modified siRNA.

Silencing efficiency using different targets in different cells

Fig.13 represent silencing of Lamin A/C in HeLaS3 and MCF7 cells. LaminA/C siRNA was modified using the best performing modifications.

Again No treatment bars represent Lamin expression in untreated, normal growing cells and GFP non specific in cells treated with control non specific siRNA.

Lamin non modified represent control transfections with non modified siRNA and Lamin Exp27-104 represent the Lamin/AC rest expression in cells, which were transfected with the corresponding modified siRNA.

Conclusion

Chemical modifications enhance the performance of the Lamin siRNA in HelaS3 and MCF7cells

Fig.14 represents silencing of CDC2 in HeLaS3 and MCF7 cells. 3CDC2 siRNA was modified using the best performing modifications.

No treatment bar represents CDC2 expression in untreated, normal growing cells and GFP non specific in cells treated with control non specific siRNA.

3CDC2 non modified represent control transfections with non modified siRNA and 3CDC2 Exp27-102 represent the CDC2 rest expression in cells, which were transfected with the corresponding modified siRNA. Conclusion

Chemical modification further enhance the performance of the already very good performing 3CDC2 siRNA in HelaS3 and MCF7cells.

MCF7 cells were transfected with 5nM of different siRNAs against CDC2 (Fig.15).

3CDC2 siRNA is a high active siRNA against CDC2 and 1CDC2 indicates a siRNA of low potency. The No treatment and GFP bar represent CDC2 expression in untreated, normal growing cells and in cells treated with control non specific siRNA respectively.

3CDC2 represents the control transfection for high efficient silencing with non modified siRNA, and 1CDC2 the expression of CDC2 in cells transfected with the non modified siRNA of low potency. 1CDC2 Expl02-101 represent the CDC2 rest expression in cells, which were transfected with the corresponding modified siRNA.

Conclusion

Chemical modifications improved the performance of a less active siRNA.

Silencing efficiency: siRNA longevity using siRNA against different targets

HeLaS3 cells were transfected with 250pM chemically modified or non modified siRNA against the MAPK2 target. The silencing analysis took place 1, 4 and 6 days after transfection (Fig.16).

"MAPK2 250pM non modified" represents the MAPK2 rest expression in HeLaS3 cells, that were treated with the non modified siRNA. Exp 101-104 represent the MAPK2 rest- expression in cells treated with the corresponding modified siRNAs.

HeLaS3 cells were transfected with 250pM chemically modified or non modified Lamin- siRNA. The Silencing analysis took place similar to the MAPK2 target 1, 4 and 6 days after Transfection (Fig.16). Lamin 250pM non modified represent the Lamin rest expression in HeLaS3 cells, that were treated with the non modified siRNA. ExplOl-104 represent the Lamin rest-expression in cells treated with the corresponding modified siRNAs.

Conclusion

Chemical modifications prolong the duration of silencing in transfected cells, avoiding in this way repeated transfections, which affects the normal cell growth and the significance (accuracy) of the silencing experiment.

siRNA stability

Modified siRNAs were incubated in inactivated human serum in order to test the stability against nucleases activity.

Stability analysis was performed for each modified siRNA separately. For this purpose siRNA was incubated in human serum and at several time points an aliquots was collected and analyzed by PAGE.

All siRNAs, that were chemically modified according the design of Exp27 show very similar stability in human serum and culture medium.

Fig.17 (A,B,C) represent some examples of modified siRNA and their extended stability in human serum.

Modified siRNAs are indicated by the name of the siRNA and the abbreviation Exp with the number of the corresponding modification.

Conclusion

The stability of the siRNA in serum is strongly sequence-depended, but in general unmodified siRNA is getting degraded very fast. Chemical modifications stabilize both stable and less stable siRNAs in presence of serum nucleases. More stable siRNA could be stabilized for over 24h, while non modified siRNA is being degraded within 15 minutes to 1 hour. The stability of unstable siRNAs could be prolonged for several hours as well, while the corresponding non modified ones were cleaved within the very first minutes of incubation.