DISCONTINUOUS OLIGONUCLEOTIDE LIGANDS
20230332148 · 2023-10-19
Inventors
Cpc classification
A61K31/713
HUMAN NECESSITIES
C12N15/111
CHEMISTRY; METALLURGY
C12N15/117
CHEMISTRY; METALLURGY
C12N2310/51
CHEMISTRY; METALLURGY
C12N15/113
CHEMISTRY; METALLURGY
International classification
C12N15/113
CHEMISTRY; METALLURGY
C12N15/11
CHEMISTRY; METALLURGY
C12N15/117
CHEMISTRY; METALLURGY
Abstract
The present invention relates to an oligonucleotide conjugate K of the structure RNA1-B-RNA2 RNA3 RNA4 or a pharmaceutically active salt thereof, wherein each RNA1, RNA2, RNA3 and RNA4 is a strand of a ribonucleic acid or of an analogue or of a derivative thereof, wherein B is a divalent linker that covalently bonds the 5′ terminus of RNA1 to the 5′ terminus of RNA2 or the 3′ terminus of RNA1 to the 3′ terminus of RNA2, and wherein RNA3 and RNA4 are not covalently bonded to each other. The invention further relates to the medical and non-medical use of such an oligonucleotide conjugate K and to corresponding manufacturing methods.
Claims
1-20. (canceled)
21. A method of treating or precluding a tumor or a viral infection, or both, in a subject, wherein the method comprises administering an effective amount of an oligonucleotide conjugate K comprising the structure ##STR00009## or a pharmaceutically compatible salt thereof to the subject, wherein: a) RNA1 is a first ribonucleic acid or an analog or derivative thereof of at least six nucleotides in length; b) RNA2 is a second ribonucleic acid or an analog or derivative thereof of at least six nucleotides in length; c) RNA3 is a third ribonucleic acid or an analog or derivative thereof of at least six nucleotides in length, which forms at least five complementary base pairs with RNA1; d) RNA4 is a fourth ribonucleic acid or an analog or derivative thereof of at least six nucleotides in length, which forms at least five complementary base pairs with RNA2; and e) B is a bivalent phosphodiester linker having a molecular weight of not more than 1500 Da that covalently bonds the 5′ terminus of RNA1 to the 5′ terminus of RNA2 or the 3′ terminus of RNA1 to the 3′ terminus of RNA2, wherein RNA3 and RNA4 are not covalently bonded to one another, wherein the oligonucleotide conjugate K is an activator of the cytosolic helicase retinoic acid-inducible gene I (RIG-I), and wherein RNA1 and RNA3, and RNA2 and RNA4, each have no overhang of the 5′-terminal nucleotide residues and an overhang of the 3′-terminal nucleotide residues of not more than five nucleotides.
22. The method of claim 21, wherein the sequence of RNA2 is identical to the sequence of RNA1.
23. The method of claim 21, wherein the sequence of RNA3 is identical to the sequence of RNA4.
24. The method of claim 21, wherein B is a bivalent linker that a) covalently bonds the 5′ terminus of RNA1 to the 5′ terminus of RNA2, and wherein RNA3 and RNA4 each have, at their 5′ termini, triphosphate residues, triphosphate analog residues or free hydroxyl groups; or b) covalently bonds the 3′ terminus of RNA1 to the 3′ terminus of RNA2, and wherein RNA1 and RNA2 each have, at their 5′ termini, triphosphate residues, triphosphate analog residues or free hydroxyl groups.
25. The method of claim 21, wherein RNA1, RNA2, RNA3 and RNA4 each have a length of between 10 and 50 nucleotides.
26. The method of claim 25, wherein RNA1, RNA2, RNA3 and RNA4 each have a length of between 15 and 40 nucleotides.
27. The method of claim 26, wherein RNA1, RNA2, RNA3 and RNA4 each have a length of between 19 and 30 nucleotides.
28. The method of claim 27, wherein RNA1, RNA2, RNA3 and RNA4 each have a length of between 20 and 25 nucleotides.
29. The method of claim 21, wherein RNA1 and RNA3, and RNA2 and RNA4, each have no overhang of the 5′-terminal nucleotide residues and an overhang of the 3′-terminal nucleotide residues of not more than four nucleotides.
30. The method of claim 21, wherein RNA1 and RNA3, and RNA2 and RNA4, each have no overhang of the 5′-terminal nucleotide residues and an overhang of the 3′-terminal nucleotide residues of not more than three nucleotides.
31. The method of claim 21, wherein RNA1 and RNA3, and RNA2 and RNA4, each have no overhang of the 5′-terminal nucleotide residues and an overhang of the 3′-terminal nucleotide residues of not more than two nucleotides.
32. The method of claim 21, wherein RNA1 and RNA3, and RNA2 and RNA4, each have no overhang of the 5′-terminal nucleotide residues and an overhang of the 3′-terminal nucleotide residues of not more than one nucleotide.
33. The method of claim 21, wherein RNA1 and RNA3, and RNA2 and RNA4, each have no overhang of the 5′-terminal nucleotide residues and no overhang of the 3′-terminal nucleotide residues.
34. The method of claim 21, wherein RNA1 is fully complementary to RNA3 and RNA2 is fully complementary to RNA4.
35. The method of claim 21, wherein the sequence of RNA2 is identical to the sequence of RNA1 and the sequence of RNA3 is identical to the sequence of RNA4, wherein each of the sequences of RNA1, RNA2, RNA3, and RNA4 are of equal length, and wherein RNA1 is fully complementary to RNA3 and RNA2 is fully complementary to RNA4.
36. The method of claim 21, wherein: RNA1 or RNA3 have a sequence homology of at least 80% to SEQ ID NO: 1; or each of RNA1 and RNA2 or RNA3 and RNA4 have a sequence homology of at least 80% to SEQ ID NO: 1 and each of RNA3 and RNA4 or RNA1 and RNA2 have a sequence homology of at least 80% to SEQ ID NO: 2.
37. The method of claim 36, wherein each of RNA1 and RNA2 or RNA3 and RNA4 have a sequence homology of at least 80% to SEQ ID NO: 1.
38. The method of claim 21, wherein: RNA1 or RNA3 have a sequence homology of at least 80% to SEQ ID NO: 1and each of RNA1 and RNA2 or RNA3 and RNA4 have a sequence homology of at least 80% to SEQ ID NO: 1 and each of RNA3 and RNA4 or RNA1 and RNA2 have a sequence homology of at least 80% to SEQ ID NO: 2.
39. The method of claim 38, wherein each of RNA1 and RNA2 or RNA3 and RNA4 have a sequence homology of at least 80% to SEQ ID NO: 1.
40. The method of claim 21, wherein the oligonucleotide conjugate K is comprised in a pharmaceutical composition further comprising a pharmaceutically acceptable vehicle.
41. The method of claim 21, wherein the oligonucleotide conjugate K is used in combination with one or more active antiproliferative or antiviral ingredients.
42. A method for increasing secretion of interferon alpha by a cell comprising contacting the cell with an oligonucleotide conjugate K comprising the structure ##STR00010## or a pharmaceutically compatible salt thereof, wherein: a) RNA1 is a first ribonucleic acid or an analog or derivative thereof of at least six nucleotides in length; b) RNA2 is a second ribonucleic acid or an analog or derivative thereof of at least six nucleotides in length; c) RNA3 is a third ribonucleic acid or an analog or derivative thereof of at least six nucleotides in length, which forms at least five complementary base pairs with RNA1; d) RNA4 is a fourth ribonucleic acid or an analog or derivative thereof of at least six nucleotides in length, which forms at least five complementary base pairs with RNA2; and e) B is a bivalent phosphodiester linker having a molecular weight of not more than 1500 Da that covalently bonds the 5′ terminus of RNA1 to the 5′ terminus of RNA2 or the 3′ terminus of RNA1 to the 3′ terminus of RNA2, wherein RNA3 and RNA4 are not covalently bonded to one another, wherein the oligonucleotide conjugate K is an activator of the cytosolic helicase retinoic acid-inducible gene I (RIG-I), and wherein RNA1 and RNA3, and RNA2 and RNA4, each have no overhang of the 5′-terminal nucleotide residues and an overhang of the 3′-terminal nucleotide residues of not more than five nucleotides.
Description
BRIEF DESCRIPTION OF THE FIGURES
[0170]
[0171]
[0172]
[0173]
[0174]
[0175]
EXAMPLES
Material and Methods
Reagents
[0176] The reagents used were purchased from Sigma-Aldrich or Roth and used without further purification.
RIG-I Stimulation Assay in Human Peripheral Blood Mononuclear Cells (PBMCs)
[0177] For stimulation tests with human PBMCs, the cells were isolated and plated out directly before the start of the experiment. The cells were pretreated with chloroquine in order to avoid stimulation of the endosomal TLR receptors and to enable selective analysis of the RIG-I activity. Stimulation with GFP2_as (single-strand RNA) as TLR ligand here controlled suppression of the endosomal receptors. For the analysis of the double-stranded oligonucleotides, the complementary single strands were first hybridized by heating to 72° C. for two minutes and gradually cooling to room temperature. Subsequently, the stimuli were introduced into cells using Lipofectamin 2000 (Invitrogen) in the specified concentration series (the molar concentrations are each based on the amount of the 24mer units). After 20 h, the IFNα secretion was determined by means of ELISA. The stimulation assays were each conducted in a double determination for 2-4 donors.
Oligonucleotides
[0178] 5′—OH—Oligoribonucleotide syntheses were purchased from either of Biomers.net GmbH and Axolabs GmbH. The syntheses of 5′-3′-linked oligonucleotides were effected using DMT-2′-O-TBDMS-rC(ac), DMT-2′-O-TBDMS-rA(bz), DMT-2′-O-TBDMS-rG(ib) and DMT-2′-O-TBDMS-rU-phosphoramidites on the 0.2-1 .Math.molar scale with a CPG loading of 39-44 .Math.mol/g.
[0179] Tetraethylene glycol and propyl linkers were introduced using the corresponding phosphoramidites. For the linear synthesis of 5′-5′- or 3′-3′-linked structures, sub-sequences were synthesized with the aid of the reverse 3′-DMT-rN-5′-CED-phosphoramidites (ChemGenes Corporation).
[0180] The parallel synthesis of 5′-5′- or 3′-3′-linked structures was effected on branching supports (ChemGenes Cooperation), which had a glycerol unit having two DMT-protected hydroxyl groups for the chain extension. For 5′-triphosphorylated sequences, the triphosphorylation and purification were effected as described above (WO2012/130886, Goldeck et al., 2014, Angew. Chem. 126:4782-4786).
TABLE-US-00003 Sequences of some oligonucleotides used and structural units of the oligonucleotide conjugates K No. Based on SEQ ID NO Name Sequence 1 1 GFP2_s 5′-GACGCUGACCCUGAAGUUCAUCUU-3′ 2 2 GFP2_as 5′-AAGAUGAACUUCAGGGUCAGCGUC-3′ 3 1.sup.p ppp-GFP2_s ppp-5′-GACGCUGACCCUGAAGUUCAUCUU-3′ 4 2.sup.p ppp-GFP2_as ppp-5′-AAGAUGAACUUCAGGGUCAGCGUC-3′ 5 1.sup.# GFP2_s2n55 3′-UUCUACUUGAAGUCCCAGUCGCAG-5′-5′ -GACGCUGACCCUGAAGUUCAUCUU 6 3 GFP2_s2n53 5′-GACGCUGACCCUGAAGUUCAUCUU GACGCUGACCCUGAAGUUCAUCUU-3′ 7 2.sup.+ GFP2_as2n33 5′-AAGAUGAACUUCAGGGUCAGCGUC-3′-3′-CUGCGACUGGGACUUCAAGUAGAA-5′ 8 1.sup.M1p ppp-GFP2_sOMe1 ppp-5-gACGCUGACCCUGAAGUUCAUCUU-3′ 9 1.sup.M1 GFP2_sOMe1 5′-gACGCUGACCCUGAAGUUCAUCUU-3′ 10 2.sup.M1 GFP2_asOMe1 5′-aAGAUGAACUUCAGGGUCAGCGUC-3′ 11 1.sup.M2 GFP2_sOMe2 5′-GaCGCUGACCCUGAAGUUCAUCUU-3′ 12 1.sup.M3 GFP2_sOMe3 5′-GAcGCUGACCCUGAAGUUCAUCUU-3′ 13 1.sup.M4 GFP2_sOMe4 5′-GACqCUGACCCUGAAGUUCAUCUU-3′ 14 1.sup.M5 GFP2_sOMe5 5′-GACGcUGACCCUGAAGUUCAUCUU-3′ 15 1.sup.M6 GFP2_sOMe6 5′-GACGCuGACCCUGAAGUUCAUCUU-3′ 16 1.sup.M7 GFP2_sOMe7 5′-GACGCUgACCCUGAAGUUCAUCUU-3′ 17 1.sup.M8 GFP2_sOMe8 5′-GACGCUGaCCCUGAAGUUCAUCUU-3′ 18 1.sup.M9 GFP2_sOMe9 5′-GACGCUGAcCCUGAAGUUCAUCUU-3′ 19 1.sup.M10 GFP2_sOMe10 5′-GACGCUGACcCUGAAGUUCAUCUU-3′ 20 2.sup.FAM GFP2_as5′-FAM 5′-FAM-AAGAUGAACUUCAGGGUCAGCGUC-3′ 21 2.sup.# GFP2_as2n55 3′-CUGCGACUGGGACUUCAAGUAGAA-5′- 5′-AAGAUGAACUUCAGGGUCAGCGUC-3′ 22 2.sup.#L GFP2_as2n55_C3br 3′-CUGCGACUGGGACUUCAAGUAGAA-5′-X1-5′-AAGAUGAACUUCAGGGUCAGCGUC-3′ 23 2.sup.#L GFP2_as2n55_T4EG 3′-CUGCGACUGGGACUUCAAGUAGAA-5′-X2-5′-AAGAUGAACUUCAGGGUCAGCGUC-3′ 24 2.sup.+L GFP2_as2n33_C3br 5′-AAGAUGAACUUCAGGGUCAGCGUC-3′-X1-3′-CUGCGACUGGGACUUCAAGUAGAA-5′ 25 2~.sup.L GFP2_as2n53_C3 5′-AAGAUGAACUUCAGGGUCAGCGUC-3′-X3-5′-AAGAUGAACUUCAGGGUCAGCGUC-3′ 26 1.sup.#L GFP2_s2n55_T3EG 3′-UUCUACUUGAAGUCCCAGUCGCAG-5′-X4-5′-GACGCUGACCCUGAAGUUCAUCUU p with triphosphorylated 5′ terminus # two copies of this sequence linked to one another via a 5′-5′ linkage + two copies of this sequence linked to one another via a 3′-3′ linkage ~ two copies of this sequence 3′-5′-linked to one another via a propyl linker M1 where the nucleotide has a 2′—O—methylation at position 1 M2 where the nucleotide has a 2′—O—methylation at position 2 M3 where the nucleotide has a 2′—O—methylation at position 3 M4 where the nucleotide has a 2′—O—methylation at position 4 M5 where the nucleotide has a 2′—O—methylation at position 5 M6 where the nucleotide has a 2′—O—methylation at position 6 M7 where the nucleotide has a 2′—O—methylation at position 7 M8 where the nucleotide has a 2′—O—methylation at position 8 M9 where the nucleotide has a 2′—O—methylation at position 9 M10 where the nucleotide has a 2′—O—methylation at position 10 FAM 5′-terminally modified with FAM (= 6-carboxyfluorescein) L linked via a bivalent linker B: X1 = glycol linker, X2 = tetraethylene glycol linker, X3 = propyl linker, X4 = triethylene glycol linker
Results
The Linkage of Oligoribonucleotide Duplexes via 5′-5′- or 3′-3′-phosphodiester bonds leads to highly active RIG-I ligands
[0181] It is known that 5′-triphosphate-dsRNAs serve as potent activators of the RIG-I immune sensor. In a comparative study, it was now assessed whether, as well as the triphosphate group, structural modifications also enable modulation of the RIG-I ligand properties of 5′-OH-RNAs. In a first example, the linkage of two 5′—OH—dsRNA 24mer units to give dimers via 5′-5′-, 3′-3′- and 3′-5′-phosphodiester bonds was the subject of a comparative study.
[0182] The model sequence chosen for the studies which follow was the 24mer ‘GFP2’ sequence. 48mer structures were obtained via standard RNA syntheses wherein, for the introduction of the 5′-5′ and 3′-3′ linkages, a sub-sequence was constructed in each case using the reverse RNA amidites (ChemGenes Corporation). Subsequently, the sequences were supplemented with the complementary 24mer opposing strands and checked for RIG-I activity in human PBMCs (tables 2 and 3). For the 5′-3′-linked dimer, as expected for 5′—OH—dsRNAs, very weak RIG-I activity was observed. By comparison, surprisingly, the linkage of the two 24mer units via 5′-5′- or 3′-3′-phosphodiester bonds led to highly active RIG-I ligands. Thus, these results introduce the 5′-5′ or 3′-3′ linkage of dsRNA units as a new RIG-I activity-enhancing structural element.
TABLE-US-00004 Comparison of the RIG-I activity of 5′-5′- and 5′-3′-linked 5′—OH—dsRNA dimers RNA (nM) IFNα (ng/ mL) GFP2_as (single strand, Seq. No. 2) Medium ppp-GFP2 (Seq. No. 3/2) GFP2_s2n55 (Seq. No. 5/2) GFP2_s2n35 (Seq. No. 6/2) MW SD MW SD MW SD MW SD MW SD 50.00 0.017 0.00 0.017 0.00 2.19 0.553 8.681 1.342 0.752 0.06 5.00 1.761 0.107 6.366 0.295 0.718 0.036 0.50 0.078 0.012 2.444 0.269 0.13 0.012
[0183] The compounds were introduced into chloroquine-treated human PBMCs in a titration series (50, 5, 0,5 nM) and the stimulatory activity was detected via measurement of the IFNα secretion after 20 h. The molar concentrations are based on the content of 24mer units. What are shown are the mean and standard deviation from the double determination for a representative donor (n = 4).
TABLE-US-00005 Comparison of the RIG-I activity of 5′-5′- and 3′-3′-linked 5′—OH—dsRNA dimers RNA (nM) IFNα (ng/ mL) GFP2_as (single strand, Seq. No. 2) Medium ppp-GFP2 (Seq. No. 3/2) GFP2_s2n55 (Seq. No. 5/2) GFP2_as2n33 (Seq. No. 7/1) MW SD MW SD MW SD MW SD MW SD 17.00 0.00 0.00 0.02 0.03 0.64 0.01 1.13 0.10 0.68 0.00 5.00 0.34 0.07 1.23 0.11 0.42 0.12 0.50 0.05 0.00 0.56 0.00 0.06 0.03
[0184] The RNA duplexes were introduced into chloroquine-treated human PBMCs in a titration series (17, 5, 0.5 nM) and the stimulatory activity was detected via measurement of the IFNα secretion after 20 h. The molar concentrations are based on the content of 24mer units. What are shown are the mean and standard deviation from the double determination for a representative donor (n = 4).
The Immune Recognition of 5′-5′- and 3′-3′-linked dsRNAs by RIG-I
[0185] The novel symmetric 5′-5′- and 3′-3′-linked dimer ligands was examined in detail hereinafter with reference to structure-activity analyses, in order to ensure and give a more detailed characterization of recognition by the RIG-I receptor. For the recognition of short RNA duplexes by RIG-I, it is common knowledge that, at first, the 5′-end is recognized by the C-terminal domain. Then the helicase domain is bound to the RNA duplex, and there is a change in conformation and activation through release of the CARD domains which mediate the further signal transmission. It was possible in each case to clarify the binding details of the RIG-I-CTD to 5′-triphosphate-dsRNA and also to 5′—OH—dsRNA from crystal structures (Wang et al. 2010, Nature Structural & Molecular Biology, Nat Struct Mol Biol. 2010 Jul;17(7):781-787; Lu et al. 2010, Nucleic Acids Res. 2011 Mar;39(4):1565-1575). In both cases, a stacking interaction of the 5′-terminal base pair with the phenylalanine residue F853 and a hydrogen bond between the 2′-hydroxyl group of the first nucleotide and the histidine residue H830 constitute essential binding contacts. For the recognition both of 5′-ppp-dsRNA and of 5′—OH—dsRNA, as well as an intact base pair at the 5′ end, a free 2′—OH group at the terminal 5′ nucleotide is thus essential. Therefore, these known binding contacts were subjected to a comparative study for substitution studies with ppp-dsRNA and the novel 5′-5′- and 3′-3′-linked 5′—OH—dsRNA dimers (table 4).
[0186] It was possible to confirm that a 2′—OMe substitution on the first nucleotide at the 5′ triphosphate terminus of ppp-dsRNA ligands (pppGFP2_sOMe1) prevents RIG-I CTD attachment by steric hindrance of the interaction with H830 and suppresses the interferon response in hPBMCs. At the opposite, inner 5′—OH end of ppp-dsRNA, a 2′—OMe substitution at the terminal nucleotide (pppGFP2_asOMe1) has only a minor influence on the stimulatory activity. This confirms that, in the RIG-I recognition of ppp-dsRNA, the opposite end of the duplex from the triphosphate group is not bound by the CTD domain. The analysis of the 2′—OMe substitution effects was thus suitable for interrogating ultimate duplex recognition by RIG-I, and both the effect of the substitution of the outwardly directed ends (GFP2_s2n55_asOMe1) and methylation on the 5′-terminal nucleotides of the ‘inner’ phosphate-bonded ends (GFP2_as2n33_sOMe1) was examined in the 5′-5′- and 3′-3′-linked dimer. Complete absence of the interferon response on substitution of the outer ends and substantial tolerance at the inner ends gave a clear suggestion that the immune response of the dimer structures is mediated exclusively by RIG-I and the RIG-I activation is initiated by the RIG-I CTD attachment at the outer 5′—OH ends.
[0187] In the context of further sedition experiments on the 5′-terminal nucleotides, the influence of steric hindrance was examined. Substitution by 5′-fluorescein as a bulky radical at the outer ends (GFP2_s2n55_asFAM) led to a complete loss of activity (table 4). On the other hand, to increase the RIG-I affinity of the outer ends, a 5′-triphosphate group was introduced (GFP2_s2n55s_as-ppp). This further potentiated the activity of the GFP2 dimer, and even in low concentrations maximum amounts of IFNα were measured (table 5). These observations thus further support the hypothesis of the necessity of a RIG-I CTD attachment to the outer ends of the dimer for the stimulatory activity.
[0188] After these studies relating to the first binding contact between the RIG-I CTD domain and the 5′ terminus of the ligands, in the next step, the further RIG-I activation with binding of the helicase domains was examined by substitution analysis on the first 10 nucleotides. In a comparative analysis, 2′—O—methylation in the 5′—OH—dsRNA 24mer and in the 5′-5′-linked 48mer ligands was analyzed, with substitution at the outer ends of the duplex in the case of the dimer (tables 6 and 7). In accordance with the observations so far, the 24mers generally showed much lower activities than the corresponding dimer compounds. Both compounds correspondingly showed a relative decrease in the interferon response in the case of 2′—O methylations at nucleotides 4-9. These results support a common recognition mechanism of the 5′—OH—dsRNA 24mer compounds and the 5′-5′-linked dimer compounds by RIG-I with comparable binding contacts to the RIG-I CTD domain and also to the duplex-recognizing helicase domain.
TABLE-US-00006 Substitution effects of 2′—O—methylation at the terminal nucleotides RNA (nM) IFNα (ng/ mL) GFP2_as (single strand, Seq. No. 2) Medium pppGFP2 (Seq. No. 3/2) pppGFP2 _asOMe1 (Seq. No. 3/10) pppGFP2 _sOMe1 (Seq. No. 8/2) GFP2 _s2n55 (Seq. No. 5/2) GFP2_s2n 55 _asOMet (Seq. No. 5/10) GFP2 _as2n33 (Seq. No. ⅐) GFP2 _as2n33_s OMe1 (Seq. No. 9/7) MW SD MW SD MW SD MW SD MW SD MW SD MW SD MW SD MW SD 17.00 0.00 0.00 0.00 0.00 3.36 0.15 1.43 0.12 0.06 0.03 3.25 0.29 0.00 0.00 3.14 0.18 2.46 0.20 5.00 1.49 0.22 0.45 0.17 0.03 0.04 2.82 0.21 0.00 0.00 2.38 0.10 1.52 0.00 1.70 0.01 0.01 0.00 0.00 0.01 0.01 0.39 0.55 0.00 0.00 0.01 0.01 0.03 0.04
[0189] 2′—OMe substitutions were introduced at the 5′-terminal nucleotides of the 5′-triphosphate-dsRNA ligand pppGFP2 and the 5′-5′- and 3′-3′-linked duplexes GFP2_s2n55 and GFP2_as2n33, and the stimulatory activity was introduced into chloroquine-treated PBMCs in a titration series. IFNα production was determined by means of ELISA after 20 h. The molar concentrations are based on the content of 24mer units. What are shown in each case are mean values and standard deviation from the double determination for a representative donor (n = 4).
TABLE-US-00007 Effect of steric hindrance on the outer 5′-terminal nucleotides of 5′-5′-linked dimer ligands on the RIG-I activity RNA (nM) IFNα (ng/ mL) GFP2_as (single strand, Seq. No. 2) Medium ppp-GFP2 (Seq. No. 3/2) GFP2_s2n55_asFAM (Seq. No. 5/20) MW SD MW SD MW SD MW SD 50.00 0.07 0.02 0.10 0.03 20.62 2.57 0.12 0.02 5.00 9.57 0.27 0.09 0.04 0.50 0.42 0.17 0.04 0.00
[0190] The hybridization of the 5′-5′-linked dimer GFP2_s2n55 with a 5′-fluorescein-substituted complementary opposing strand gave a ligand variant with steric hindrance at the outer dimer ends. RIG-I activity was detected in chloroquine-treated PBMCs by determination of the IFNα production. The molar concentrations are based on the content of 24mer units. What are shown in each case are mean values and standard deviation from the double determination for a representative donor (n = 4).
TABLE-US-00008 Effect of 5′-triphosphorylation at the outer 5′-terminal nucleotides of 5′-5′-linked dimer ligands on RIG-I activity RNA (nM) IFNα (ng/ mL) GFP2_as (single strand, Seq. No. 2) Medium ppp-GFP2 (Seq. No. 3/2) GFP2_s2n55 (Seq. No. 5/2) GFP2_s2n55_as-ppp (Seq. No. 5/4) MW SD MW SD MW SD MW SD MW SD 50.00 0.11 0.01 0.05 0.01 28.16 3.11 24.44 8.93 36.52 2.29 5.00 3.76 0.02 13.06 1.75 35.83 1.54 0.50 0.07 0.00 0.13 0.01 29.85 1.43
[0191] The 5′-5′-linked dimer GFP2_s2n55 was hybridized with a 5′-triphosphorylated complementary opposing strand and introduced into chloroquine-treated PBMCs. After 20 h, IFNα production was detected. The molar concentrations are based on the content of 24mer units. What are shown in each case are mean values and standard deviation from the double determination for a representative donor (n = 4).
TABLE-US-00009 2′—O—Methyl substitution effects at nucleotides 1-10 in RIG-I recognition of 5′—OH—dsRNA ligands RNA (nM) IFNα (ng/ mL) GFP2_as (single strand, Seq. No. 2) Medium GFP2 (Seq. No. ½) GFP2_sOMe1 (Seq. No. 9/2) GFP2_sOMe2 (Seq. No. 11/2) GFP2_sOMe3 (Seq. No. 12/2) GFP2_sOMe4 (Seq. No. 13/2) MW SD MW SD MW SD MW SD MW SD MW SD MW SD 50.00 0.00 0.00 0.00 0.00 6.61 1.15 0.25 0.09 0.61 0.18 4.75 0.35 0.97 0.25 5.00 0.08 0.02 0.00 0.00 0.00 0.00 0.08 0.02 0.00 0.00
TABLE-US-00010 IFNα (ng/ mL) GFP2_sOMe5 (Seq. No. 14/2) GFP2_sOMe6 (Seq. No. 15/2) GFP2_sOMe7 (Seq. No. 16/2) GFP2_sOMe8 (Seq. No. 17/2) GFP2_sOMe9 (Seq. No. 18/2) GFP2_sOMe10 (Seq. No. 19/2) MW SD MW SD MW SD MW SD MW SD MW SD 1.04 0.05 0.47 0.02 0.22 0.04 0.83 0.10 1.08 0.15 4.75 0.24 0.00 0.00 0.00 0.00 0.02 0.02 0.00 0.00 0.00 0.00 0.02 0.02
[0192] 5′-OH-GFP2 compounds with 2′—O methylations at nucleotides 1-10 were hybridized with the complementary 24mer opposing strand and used for stimulation of chloroquine-treated human PBMCs. Transfection was effected with RNA concentrations of 50 and 5 nM (based on the amount of 24mer units) and, after 20 h, the amounts of IFNα in the supernatant were quantified by ELISA. What are shown are the mean and standard deviation from the double determination for a representative donor (n = 2).
TABLE-US-00011 2′—O—Methyl substitution effects at nucleotides 1-10 in RIG-I recognition of 3′-3′-linked 5′—OH—dsRNA dimer ligands RNA (nM) IFNα (ng/ mL) GFP2_as (single strand, Seq. No. 2) Medium GFP2_as2n 55 (Seq. No. 1/21 GFP2_as2n 55_sOMe1 (Seq. No. 9/21) GFP2_as2n 55_sOMe2 (Seq. No. 11/21) GFP2_as2n 55_sOMe3 (Seq. No. 12/21) GFP2_as2n 55_sOMe4 (Seq. No. 13/21) MW SD MW SD MW SD MW SD MW SD MW SD MW SD 50.00 0.00 0.00 0.00 0.00 15.33 9.45 0.00 0.00 13.02 2.90 22.54 8.59 17.26 3.19 5.00 11.32 1.87 0.00 0.00 6.13 0.87 24.17 0.39 8.95 0.42 IFNα (ng/ mL) GFP2_as2n5 5_sOMe5 (Seq. No. 14/21) GFP2_as2n5 5_sOMe6 (Seq. No. 15/21) GFP2_as2 n55_sOMe 7 (Seq. No. 16/21) GFP2_as2n5 5_sOMe8 (Seq. No. 17/21) GFP2_as2n5 5_sOMe9 (Seq. No. 18/21) GFP2_as2n5 5_sOMe10 (Seq. No. 19/21) MW SD MW SD MW SD MW SD MW SD MW SD 11.88 3.04 13.44 0.25 8.61 1.01 12.26 1.62 14.51 0.20 19.06 1.82 8.16 0.37 10.76 0.30 4.13 0.27 7.94 0.22 15.49 1.67 21.01 2.80
[0193] 5′-OH-GFP2 compounds with 2′—O methylations at nucleotides 1-10 were hybridized with the complementary 5′-5′-linked 48mer opposing strand and used for stimulation of chloroquine-treated human PBMCs. Transfection was effected with RNA concentrations of 50 and 5 nM (based on the amount of 24mer units) and, after 20 h, the amounts of IFNα in the supernatant were quantified by ELISA. What are shown are the mean and standard deviation from the double determination for a representative donor (n = 2).
Variation of the RNA Sequence in RNA-5′-p-5′-RNA Recognition
[0194] In the first examples, dimer formation of the GFP2 model sequence at the otherwise triphosphorylated 5′ terminus (5′-GACG..., GFP2_s2n55 and GFP2_as2n33) was examined, in which the opposite ends (5′-AAGA...) in the dimer are directed outward and hence mediate recognition by RIG-I. As the first sequence variation, by virtue of the alternative linkage of the GFP2 duplexes via the 5′ ends of the antisense strand (GFP2_2n55as), the opposite 5′ termini (5′-GACG...) were directed outward. On analysis of the RIG-I activity in human PBMCs, it was possible to observe an equally strong interferon response analogously to the triphosphate compounds irrespective of the sequence for the dimer compounds (table 8). For this dimer too, it was possible by additional triphosphorylation of the outer ends (pppGFP2_as2n55) to further potentiate the RIG-I activity. These results suggest general applicability of the activity-promoting 5′-5′ and 3′-3′ structure motifs to further RNA dimer sequences.
TABLE-US-00012 Comparison of the dimer compounds GFP2_2n55s and GFP2_2n55as. RNA (nM) IFNα (ng/ mL) GFP2_as (single strand, Seq. No. 2) ppp-GFP2 GFP2_ppp-as (Seq. No. 3/2) GFP2_s2n55 (Seq. No. 5/2) GFP2_s2n55_as-ppp (Seq. No. 5/4) GFP2_as2n55 (Seq. No.1/21) pppGFP2_as2n55 (Seq. No. 3/21) MW SD MW SD MW SD MW SD MW SD MW SD MW SD 5.00 0.16 0.00 14.08 1.00 17.88 1.85 19.10 1.11 22.45 0.54 24.86 2.58 25.73 1.50 1.60 9.73 0.53 13.18 0.19 17.31 1.50 24.35 1.54 19.05 0.19 25.05 1.46 0.50 4.04 0.45 2.13 0.91 7.44 0.42 19.76 2.81 6.94 4.40 20.76 2.92 0.17 0.27 0.06 0.20 0.06 0.41 0.14 11.47 1.08 0.65 0.37 14.59 3.81 0.05 0.16 0.00 0.16 0.00 0.12 0.06 8.74 0.82 0.12 0.06 9.88 2.01 0.01 0.08 0.11 0.20 0.17 0.20 0.06 5.39 0.48 0.04 0.06 5.68 0.24
[0195] The dimer compounds GFP2_s2n55 and GFP2_as2n55 were hybridized with the complementary OH-RNA opposing strand and the pppRNA opposing strand and used in comparison with pppGFP2 and GFP2_as-ppp for stimulation of chloroquine-treated human PBMCs. The transfection was effected as a titration series (5, 1.6, 0,5, 0.16, 0.05, 0.005 nM, based on the concentration of 24 units) and, after 20 h, the amounts of IFNα in the supernatant were quantified by ELISA. What are shown are the mean and standard deviation from the double determination for a representative donor (n = 2).
The Introduction of Linker Structures at the 5′-3′, 5′-5′ and 3′-3′ Bond Between the dimer units is tolerated by RIG-I
[0196] In the further study of ligand requirements, the effect of different linker structures on the 5′-5′ or 3′-3′ linkage of the dimers was analyzed, using the compounds GFP2_as2n55 and GFP2_as2n33 as example sequences. Firstly, the introduction of a C3 unit was examined. The synthesis approach chosen here was the parallel synthesis of the two 24mer units on a CPG-bound glycerol linker (branching 3′-Icaa-CPG, ChemGenes Corporation). The support material enables, proceeding from the linker unit, either 5′-5′ linkage using the reverse RNA amidites for the RNA synthesis or 3′-3′ linkage using the standard RNA amidites. As example compounds, the compounds GFP2_as2n55_C3br and GFP2_as2n33_C3br were thus synthesized. In addition, a tetraethylene glycol unit was inserted into the GFP2_as2n55 compound as an example of a longer linker structure (GFP2_as2n55_T4EG). The synthesis here was implemented as a continuous RNA synthesis using the appropriate tetraethylene glycol phosphoramidite, employing the reverse RNA amidites. As well as these 5′-5′ and 3′-3′ linkages, the effect of the introduction of linker structures was also examined in the context of a regular 5′-3′-linked dimer. As an example synthesis for this purpose, a propanediol linker was introduced during the continuous RNA synthesis (GFP2_as2n53_C3). On testing in human PBMCs, all the linker-containing dimer compounds synthesized showed high RIG-I activity, and a comparable to slightly weaker interferon response was achieved compared to the 5′-5′- and 3′-3′-phosphate-linked dimers (tables 9 and 10). It was possible to further increase the activity of the linker-linked dimers by an additional introduction of 5′-triphosphate groups. These results show that the introduction of linker structures into the 5′-5′- and 3′-3′-linked RNA dimers with retention of the high RIG-I activity is possible, and even the interruption of continuous 5′-3′-linked dimers by linker structures can entail an increase in activity. Thus, the RIG-I activity-promoting structural element can extend to the group of the linker-bridged 5′-3′, 5′-5′ and 3′-3′ linkages. This enables flexible and versatile synthesis of the corresponding ligands, with enablement not only of continuous 48mer synthesis but also of parallel synthesis of the 24mer units at branching CPG units as a further synthesis approach.
TABLE-US-00013 The activation of RIG-I by 5′-5′-linked RNA-p linker-p-RNA compounds RNA (nM) IFNα (ng/ mL) GFP2_as (single strand, Seq. No. 2) Medium ppp-GFP2 (Seq. No. 3/2) GFP2_as2n 55 (Seq. No. 5/21) GFP2_as2n5 5_C3br (Seq. No. 1/22) GFP2_as2n55 _T4EG (Seq. No. 1/23) pppGFP2_as 2n55 (Seq. No. 3/21) pppGFP2_as 2n55_C3br (Seq. No. 3/22) pppGFP2_as 2n55_T4EG (Seq. No. 3/23) MW SD MW SD MW SD MW SD MW SD MW SD MW SD MW SD MW SD 50.00 0.02 0.03 0.00 0.00 27.93 2.07 23.81 5.00 21.80 1.78 21.80 2.07 22.38 4.33 22.38 4.33 22.52 1.35 5.00 21.50 2.02 27.52 3.42 14.32 2.93 12.27 3.72 26.02 1.40 26.02 1.40 24.66 0.72 0.50 1.70 0.57 2.78 0.90 0.16 0.07 0.07 0.05 24.49 2.21 24.49 2.21 20.99 7.26 0.05 0.04 0.00 0.02 0.03 0.00 0.00 0.02 0.03 19.69 0.53 19.69 0.53 13.16 0.91
[0197] The GFP2_as2n55 dimers linked via a glycol linker (C3br), via a propane linker (C3) or via a tetraethylene linker (T4EG) were hybridized with the complementary 5′—OH and 5′-triphosphate opposing strands. Reference substances used were the pppGFP2 duplex and the phosphodiester-linked GFP2_s2n55 duplex. The compounds were used for stimulation of chloroquine-treated human PBMCs. Transfection was effected in a titration series (50, 5, 0.5, 0.005 nM) and, after 20 h, the amounts of IFNα in the supernatant were quantified by ELISA. The concentrations specified are based on the content of 24mer units. What are shown are the mean and standard deviation from the double determination for a representative donor (n=2).
TABLE-US-00014 The activation of RIG-I by 3′-3- and 3′-5′-linked RNA-p linker-p-RNA compounds RNA (nM) IFNα (ng/ mL) GFP2_as (single strand, Seq. No. 2) Medium ppp-GFP2 (Seq. No. 3/2) GFP2_as2n3 3 (Seq. No. ⅐) GFP2_as2n3 3_C3br (Seq. No. 1/24) GFP2_as2n5 3_C3 (Seq. No. 1/25) MW SD MW SD MW SD MW SD MW SD MW SD 17.00 0.00 0.00 0.00 0.00 16.15 0.95 19.66 1.15 15.06 2.22 17.92 0.08 5.00 12.94 1.52 12.98 3.76 11.00 2.96 14.66 4.85 1.70 9.86 1.10 5.00 0.95 5.00 0.95 4.35 0.50
[0198] For the 3′-3′ and 3′-5′ linkage of dimers, the effect of a glycol linker (C3br) and of a propane linker (C3) on RIG-I activity was examined. The dimers were hybridized with the complementary opposing strands and used for stimulation of chloroquine-treated PBMCs. After 20 h, the amounts of IFNα in the supernatant were quantified by ELISA. The concentrations reported are based on the content of 24mer units. What are shown are the mean and standard deviation from the double determination for a representative donor (n=2).
The Synthesis of Dimeric RIG-I Ligands by Linker-mediated 5′-5′ Linkage Of support-bound RNA sequences
[0199] As well as the parallel synthesis of the sub-sequences on branching support materials, the possibility of the linkage of support-bound 5′-OH-RNA sequences constitutes an attractive synthesis strategy for dimeric RIG-I ligands. For this purpose, the phosphitylation and hydrolysis of support-bound 5′-OH-RNA sequences to give the H-phosphonate derivatives was followed by the reaction with bifunctional linkers in the presence of pivaloyl chloride and oxidation of the phosphite to phosphate groups. This synthesis approach was applied successfully, for example, to the 5′-5′ linkage of the GFP2 sequence via a triethylene glycol linker (GFP2_2n55_T3EG) (
[0200] The fully protected, support-bound 5′-OH-GFP2 model sequence (1 .Math.mol) was first washed with an anhydrous pyridine/dioxane solution (1:3, v/v, 4 mL). Subsequently, a freshly made-up 50 mM 2-chloro-4H-1,3,2-benzodioxaphosphorin-4-one solution (2 mL, 100 .Math.mol) in dry dioxane/pyridine (3:1, v/v) was passed through the column with the aid of plastic syringes. The solution was repeatedly contacted with the synthesis column by means of the plastic syringe by drawing it up and expelling it for a reaction time of 30 minutes. Subsequently, the synthesis column was washed with acetonitrile (3 mL). In the next step, the synthesis column was incubated with an aqueous 1 M triethylammonium bicarbonate solution (1 mL) for 30 min and hence the support-bound 5′—H—phosphonate-RNA was obtained. Thereafter, the synthesis column was washed with acetonitrile (3 mL) and acetonitrile/pyridine (1:1, v/v, 3 mL). For the linker-mediated linkage, a solution of pivaloyl chloride (12 .Math.L, 0.1 mmol) and triethylene glycol (0.27 .Math.L, 2 .Math.mol) was made up in acetonitrile/pyridine (1:1, v/v, 1 mL) and passed through the synthesis column. An incubation time of 5 h was followed by a further wash step with acetonitrile (3 mL) and oxidation with a 0.1 M iodine solution in THF/pyridine/water (2 mL) for 10 min. Subsequently, the synthesis product was washed with acetonitrile (3 x 3 mL), dried in an argon stream and cleaved from the support material and deprotected under standard conditions. The 5′-5′-linked dimer product was purified via HPLC chromatography on an anion exchange column (Source15Q 4.6/100, GE Healthcare) in a sodium perchlorate gradient (20 mM Tris-HCl, pH8, 1 mM EDTA, 80-320 mM NaClO.sub.4) and desalinated by means of a HiTrap column (GE Healthcare). The integrity of the product obtained was confirmed via RP-LC/MS analysis.
[0201] The RIG-I activity of the GFP2_s2n55_T3EG product obtained was ascertained in a stimulation experiment in human PBMCs. Slightly higher interferon responses compared to ppp-GFP2 and interferon responses comparable to GFP2_s2n55 were measured (
[0202] In summary, it was thus possible to demonstrate, for the synthesis of RIG-I ligand structures bonded via 5′-5′- and 3′-3′-phosphodiester bonds and via 5′-5′, 3′-3′ or 3′-5′ linker structures, the continuous synthesis and parallel synthesis of the dimer units at branching units or in combination with a post-synthetic linkage (
TABLE-US-00015 The RIG-I activity of a triethylene glycol linker-linked 5′-5′ dimer RNA (nM) IFNα (ng/ mL) GFP2_as (single strand, Seq. No. 2) Medium ppp-GFP2 (Seq. No. 3/2) GFP2_s2n5 5 (Seq. No. 5/2) GFP2_s2n5 5_T3EG (Seq. No. 26/2) MW SD MW SD MW SD MW SD MW SD 50.00 0.00 0.00 0.00 0.00 7.37 0.25 7.86 0.98 6.38 0.93 17.00 5.69 0.00 8.43 0.00 8.05 1.47 5.00 5.94 0.00 7.72 1.67 6.23 0.09 1.70 2.23 0.38 5.71 1.30 3.20 1.44
[0203] The GFP2_s2n55_T3EG dimer 5′-5′-linked via a triethylene glycol linker was hybridized with the complementary opposing strand and examined for RIG-I activity in chloroquine-treated PBMCs by comparison with the ppp-GFP2 duplex and with the GFP2_s2n55 dimer duplex directly 5′-5′-linked via a phosphodiester bond. The concentration figures are based on the amount of 24mer units. What are shown are the mean values and standard deviations of a double determination for a representative donor (n = 2).