| Nucleic Acids Research | Pages |
©1999 Oxford University Press |
Chiral and steric effects in the efficient binding of [alpha]-anomeric deoxyoligonucleoside N-alkylphosphoramidates to ssDNA and RNA
Introduction
Materials And Methods
P-chirally enriched [alpha]-dinucleoside phosphoramidates
Phosphoramidites of [alpha]-dinucleoside phosphoramidates 9F and 9S, 10F and 10S and 11F and 11S
Protected 2[prime]-deoxy-[alpha]-ribonucleoside-3[prime]-O-H-phosphonates 12, 13 and 14 (Scheme 3)
Oligonucleotides synthesis
Melting experiments
Results And Discussion
Preparation of oligonucleotides containing a single chirally enriched phosphoramidate linkage
Evaluation of the chirality effect on hybridization properties of [alpha]-oligonucleoside N-alkylphosphoramidates
Effect of steric hindrance of the P substituent on hybridization properties
Specificity of hybridization of [alpha]-oligonucleoside N-alkyl-phosphoramidates
Influence of salt concentration on the thermal stability of hybrids
Acknowledgements
Supplementary Material
References
Chiral and steric effects in the efficient binding of [alpha]-anomeric deoxyoligonucleoside N-alkylphosphoramidates to ssDNA and RNA
Received August 3, 1999; Revised and Accepted September 14, 1999
ABSTRACT We report hybridization properties of new phosphate-modified [alpha]-oligonucleoside analogs with non-ionic or cationic internucleotide linkages such as methoxy-ethylphosphoramidate (PNHME), phosphoromorpholi-date (PMOR) and dimethylaminopropylphosphor-amidate (PNHDMAP). First we evaluated the chirality effect of the phosphorus atom on the affinity of [alpha]- or [beta]-dodecanucleoside phosphodiesters containing one chirally enriched N-alkylphosphoramidate linkage located in the middle of the sequence d(TCTT-AA*CCCACA). As for P-substituted [beta]-oligonucleo-tides, a difference in binding behavior between the two diastereoisomers (difference in [Delta]Tm) exists in the hybridization properties of [alpha]-analogs when DNA was the target but this effect was not detrimental to duplex stability. This effect was considerably reduced when RNA was the target. Secondly we studied the effect of steric hindrance around phosphorus on the affinity of fully modified [beta]- and [alpha]-oligonucleoside N-alkylphosphoramidates for their DNA and RNA targets. This effect was very weak with [alpha]-analogs whereas it was more pronounced with [beta]-oligos. PNHME-modified [alpha]-oligonucleosides formed more stable duplexes with DNA ([Delta]Tm +9.6°C) and RNA ([Delta]Tm +1.4°C) targets than the `parent' phosphodiester. Finally, base pairing specificity of these [alpha]-oligonucleoside N-alkylphosphoramidates for their targets was found to be as high as for natural oligonucleoside phosphodiesters.
INTRODUCTION
The high affinity and specificity of Watson-Crick hybridization has made oligonucleotides (ODNs) attractive agents for diagnostic and therapeutic applications via the antisense or antigene mode of action (1-5). Much effort has focused on the chemical modifi-cation of oligonucleotides (6-9) to improve their hybridization affinity, their biostability and their cellular uptake. Among these modifications, several phosphate-modified oligonucleotides, in particular ionic phosphorothioate (PS) (10) and non-ionic methylphosphonate (MP) ODNs (11), have been investigated. These compounds are resistant to nucleases and have been used in various in vitro and in vivo systems as antisense agents (5,12). Phosphoramidate ODNs in which a non-bridged oxygen is substituted by primary or secondary amino groups have also been studied (13-22). The replacement of one non-bridging oxygen atom by any substituent induces the appearance of a new chiral center at the phosphorus. The acquired chirality of the new internucleoside linkages leads to a mixture of diastereo-isomers, only a few of which possess the optimal combination of physical and chemical properties supporting the desired biological effects (23). It was recognized early that the stability of complexes formed between oligonucleotide analogs and complementary oligo- and polynucleotides depends on the stereochemistry of P-chiral internucleoside linkages. Most of the melting temperature (Tm) data were collected for complexes involving oligonucleotides with a single phosphoramidate internucleoside linkage (14), or oligomers with alternating phosphodiester/phosphoramidate linkages (16,20,21). Although oligonucleoside phosphoramidates have been studied, data on chirally pure diastereoisomers are limited (18,19). In addition to stereochemical effects, the steric hindrance of the P substituent also has a detrimental impact on the overall stability of the duplex (23).
Some years ago, our group developed nuclease-resistant [alpha]-anomeric oligonucleotides which form stable parallel-stranded duplexes with complementary natural [beta]-DNA or [beta]-RNA strands (24,25). Interestingly, we found that PS [alpha]-ODNs which combine two structural modifications, one located in the sugar moiety and the second in the internucleoside link, hybridize to their DNA and RNA targets more efficiently than do their PS [beta]-homologs (26,27). The same behavior was observed with [alpha]- and [beta]-dodecathymidylates containing non-ionic primary phosphoramidate (PNH2) internucleoside linkages (28). Surprisingly, the non-ionic PNH2 [alpha]-dT12 hybridizes to its targets even better than does the parent [beta]-oligonucleoside phosphodiester (PO). More recently, we reported that non-ionic MP [alpha]-ODNs form duplexes with DNA and RNA targets more stable than MP [beta]-ODNs. In addition, MP [alpha]-ODNs bind to complementary DNA more tightly than do the corresponding natural PO [beta]-oligomers (29).
These data obtained with PS, PNH2 and MP [alpha]-ODNs prompted us to study new phosphate-modified [alpha]-oligonucleoside analogs with non-ionic N-alkylphosphoramidate linkages which are chemically more stable than PNH2 links and allow variation of the steric hindrance around phosphorus as a function of the alkyl groups attached to the nitrogen atom. Here we report hybridization properties of [alpha]-ODNs with non-ionic or cationic internucleotide linkages such as methoxy-ethylphosphoramidate (13,15-17), phosphoromorpholidate (13,18,19) and dimethylaminopropylphosphoramidate (20-22) (Scheme 1). In this paper we focus on the effect of the chirality of the internucleotide linkage and the effect of steric hindrance around the phosphorus on the hybridization properties of these new [alpha]-analogs with single-stranded DNA and RNA targets.
Scheme 1. N-alkylphosphoramidate internucleoside linkages used in this study.
First, to evaluate the effect of chirality, we prepared [alpha]- or [beta]-dodecanucleoside phosphodiesters containing one chirally enriched phosphoramidate modification in the middle of the sequence [beta]-3[prime]-d(TCTTAA*CCCACA)-5[prime] or [alpha]-5[prime]-d(TCTT-AA*CCCACA)-3[prime] complementary to the splice acceptor site of HIV-1 tat RNA. The effect of chirality on the stability of the duplexes formed with complementary single-stranded (ss)DNA or RNA strands was measured by the variation in Tm ([Delta]Tm).
Secondly, we studied the effect of steric hindrance around the phosphorus on the hybridization properties of fully modified phosphorus-substituted dodecamers [beta]-3[prime]- and [alpha]-5[prime]-d(TCTTAACCCACA). Finally, base pairing specificity of these [alpha]-oligonucleoside N-alkylphosphoramidates with their complementary DNA and RNA targets was investigated.
MATERIALS AND METHODS
Except as noted, chemicals were reagent grade or better and used without further purification. Acetonitrile was distilled over calcium hydride, tetrahydrofuran was distilled over LiAlH4 and methylene chloride was distilled over phosphorus pentoxide. 2-Methoxyethylamine (Aldrich) and N,N-dimethyl-3-aminopropylamine (Aldrich) were distilled over calcium hydride whereas morpholine (Aldrich) was used without purification. Silica gel TLC was carried out on Kieselgel 60 F254 plates (Merck), and compounds were visualized by UV shadowing. Silica gel 60 (230-240 mesh; Merck) was used for flash column chromatography. FAB mass spectra were recorded on a JEOL DX300 spectrometer operating with a JMA-DA 5000 mass data system in positive or negative ion mode; MALDI-TOF mass spectra of oligonucleotides were obtained on a Linear Voyager DE instrument (PerSeptive Biosystems) equipped with a 337 nm UV laser. The 31P NMR spectra were recorded with a Bruker AC 250 spectrometer, and chemical shifts were measured relative to 85% H3PO4 as external reference. Reverse phase HPLC was performed on a Waters 600E system equipped with a Model 990 or 996 photodiode array detector and using an EC Nucleosil 5µ C18 column (150 × 4.6 mm; Macherey-Nagel) for analytical purposes (flow rate 1.0 ml/min) and a Waters Delta Pak 15µ C18 100 Å column (300 × 7.8 mm) for preparative work (flow rate 2.0 ml/min). Elution was performed with linear gradients of acetonitrile in 50 mM triethylammonium acetate (TEAAC), pH 7.
P-chirally enriched [alpha]-dinucleoside phosphoramidates
O-[5[prime]-O-(4,4[prime]-dimethoxytrityl)-6-N-benzoyl-[alpha]-2[prime]-deoxyadenosin-
3[prime]-yl]-O-[3[prime]-O-(tertiobutyldimethylsilyl)-4-N-benzoyl-[alpha]-2[prime]-deoxy-
cytidin-5[prime]-yl]-N-alkylphosphoramidates (-). The [alpha]-dinucleoside phosphoramidates 3-5 were prepared from 5[prime]-O-dimethoxytrityl-N-6-benzoyl-2[prime]-deoxy-[alpha]-adenosine-3[prime]-O-(methyl N,N-diisopropylphosphoramidite) 1 synthesized as described previously (30) and 3[prime]-O-t-butyldimethylsilyl-N-4-benzoyl-2[prime]-deoxy-[alpha]-cytidine 2 obtained according to standard procedures (tritylation, silylation and detritylation) (31) with 77% overall yield from [alpha]-dCBz (30).
2. 1H NMR (CDCl3, 250 MHz) [delta] 8.72 (br s, 1H, NH), 8.13 (d, 1H, H-6), 7.95-7.50 (m, 6H, H-5, H ar), 6.32 (dd, 1H, H-1[prime], J = 6.9, 1.4), 4.43 (m, 2H, H-3[prime], H-4[prime]), 3.73 (m, 2H, H-5[prime], H5[prime][prime]), 2.74 (m, H, H-2[prime]), 2.51 (br s, 1H, OH-5[prime]), 2.24 (m, 1H, H-2[prime][prime]), 0.82 [s, 9H, C(CH3)3], 0.04, 0.00 (2s, 6H, 2× CH3). FAB mass spectrum (positive mode, glycerol/thioglycerol 50/50 v/v): 446 [M + H]+.
Compounds 1 (745 mg, 0.91 mmol) and 2 (290 mg, 0.65 mmol) were dried three times by evaporation from dry CH3CN, then the residue was dissolved in dry CH3CN (3.2 ml) and tetrazole (273 mg, 3.9 mmol) was added to the mixture. The reaction was complete after stirring for 30 min at room temperature. A solution of iodine (230 mg, 0.91 mmol) and amine (24.7 mmol) (methoxyethylamine, morpholine or dimethylaminopropylamine) in dry THF (8 ml) was added to the reaction mixture. After 5 min, the mixture was diluted with CH2Cl2 (20 ml), poured into 5% aqueous sodium bisulfite (150 ml) and extracted with CH2Cl2 (3 × 50 ml). The organic layer was washed repeatedly with saturated NaCl, then dried over Na2SO4 and concentrated under reduced pressure. Efforts to separate the two diastereoisomers of compounds 3 and 4 have been unsuccessful. Compound 3 was purified (98% yield) as a diastereoisomeric mixture by silica gel chromatography using a 0-5% MeOH gradient in CH2Cl2 in the presence of 0.5% triethylamine.
3. 31P NMR (CDCl3, 101 MHz) [delta] 9.18 and 9.10. FAB mass spectrum (negative mode, thioglycerol): 1221 [M - H]-.
Crude 4. 31P NMR (CDCl3, 101 MHz) [delta] 8.42 and 8.25. FAB mass spectrum (negative mode, thioglycerol): 1233 [M - H]-.
The two diastereoisomers of 5 were separated by silica gel chromatography as the `fast' 5F (25% yield) and `slow' 5S (29% yield) eluted using 4% isocratic EtOH in CH2Cl2 in the presence of 0.1% Et3N.
5F. 31P NMR (CD3CN, 101 MHz) [delta] 10.11.
5S. 31P NMR (CD3CN, 101 MHz) [delta] 10.12.
5F and 5S FAB mass spectra (negative mode, NOBA) 1247 [M - H]-.
O-[5[prime]-O-(4,4[prime]-dimethoxytrityl)-6-N-benzoyl-[alpha]-2[prime]-deoxyadenosin-
3[prime]-yl]-O-(4-N-benzoyl-[alpha]-2[prime]-deoxycytidin-5[prime]-yl)-N-alkylphosphoramidates 6F and 6S, 7F and 7S, and 8F and 8S. Compounds 3, 4, 5F and 5S (1 mmol) were treated with tetrabutylammonium fluoride (2 mmol) in dry THF (25 ml) for 1 h. After evaporation of THF, the residue was dissolved in CH2Cl2 (150 ml) and washed with water (3 × 100 ml). The organic layer was dried over Na2SO4, filtered and evaporated under reduced pressure. Purification of compounds 6, 7, 8F and 8S was performed as follows.
The two diastereoisomers of 6 were separated by silica gel chromatography as the `fast' 6F and `slow' 6S eluting isomers using 2% isocratic EtOH in CHCl3 in the presence of 0.1% Et3N; six consecutive chromatographic steps were necessary to separate the two isomers.
6F. Yield 18%; HPLC RT 20.58 min (91% diastereoisomeric purity) with a 10 min linear gradient of 72-90% CH3CN in 50 mM TEAAC followed by 90% isocratic CH3CN. 31P NMR (CDCl3, 101 MHz) [delta] 8.97. FAB mass spectrum (negative mode, thioglycerol) 1106 [M - H]-.
6S. Yield 27%; HPLC RT 19.15 min (86% diasteroisomeric purity) with a 10 min linear gradient of 72-90% CH3CN in 50 mM TEAAC followed by 90% isocratic CH3CN.31P NMR (CDCl3, 101 MHz) [delta] 9.22. FAB mass spectrum (negative mode, thioglycerol) 1106 [M - H]-.
The two diastereoisomers of 7 were separated by silica gel chromatography as the `fast' 7F and `slow' 7S eluting isomers using 1% isocratic EtOH in CHCl3 in the presence of 0.15% Et3N; five consecutive chromatographic steps were necessary to separate the two isomers.
7F. Yield 35%; HPLC RT 12.68 min (88.5% diastereoisomeric purity) with a 20 min linear gradient of 48-64% CH3CN in 50 mM TEAAC.31P NMR (CD3CN, 101 MHz) [delta] 8.38. FAB mass spectrum (negative mode, thioglycerol) 1118 [M - H]-.
7S. Yield 25%; HPLC RT 12.06 min (80% diasteroisomeric purity) with a 20 min linear gradient of 48-64% CH3CN in 50 mM TEAAC. 31P NMR (CD3CN, 101 MHz) [delta] 8.21. FAB mass spectrum (negative mode, thioglycerol) 1118 [M - H]-.
Diastereoisomers 8F and 8S were purified by silica gel chromatography using a 0-5% MeOH gradient in CH2Cl2 in the presence of 0.5% Et3N.
8F. Yield 17%; HPLC RT 16.28 min (93% diastereoisomeric purity) with a 30 min linear gradient of 48-64% CH3CN in 50 mM TEAAC. 31P NMR (CDCl3, 101 MHz) [delta] 9.45. FAB mass spectrum (positive mode, NOBA) 1134 [M + H]+.
8S. Yield 21%; HPLC RT 17.61 min (85% diasteroisomeric purity) with a 30 min linear gradient of 48-64% CH3CN in 50 mM TEAAC.31P NMR (CDCl3, 101 MHz) [delta] 9.39. FAB mass spectrum (positive mode, NOBA) 1134 [M + H]+.
Phosphoramidites of [alpha]-dinucleoside phosphoramidates 9F and 9S, 10F and 10S and 11F and 11S
A mixture of compound 6F (0.5 mmol) and diisopropylammonium tetrazolide (0.25 mmol) was dried by co-evaporation three times with anhydrous pyridine and then acetonitrile. The dry residue was suspended in freshly distilled CH2Cl2 (10 ml) and O-(2-cyanoethyl)-N,N[prime]-bis(diisopropylamino)phosphine (0.6 mmol) was added under argon. After stirring at room temperature for 16 h, the reaction mixture was diluted with ethyl acetate (200 ml), the solution filtered and washed successively with saturated aqueous NaHCO3 (250 ml) and saturated NaCl (250 ml). The organic layer was dried over Na2SO4 and concentrated under reduced pressure. The residue was purified by silica gel chromatography using a 60-98.5% CH2Cl2 gradient in cyclohexane in the presence of 1.5% Et3N, then a 0-2% MeOHgradientin CH2Cl2 in the presence of 1.5% Et3N. The phosphoramidite derivative 9F was obtained in 43% yield as a white powder after lyophilization from benzene. Isomer 9S was prepared similarly from 6S.
9F. 31P NMR (CD3CN, 101 MHz) [delta] 150.11 and 149.65 (stereoisomeric phosphoramidite groups) and 10.31 and 10.26 (internucleoside phosphoramidate group) in a 1:1 ratio for the two sets of signals.
9S. Yield 64%, 31P NMR (CD3CN, 101 MHz) [delta] 150.21, 149.69 and 10.49, 10.40 (phosphoramidite/phosphoramidate, 1:1).
9F and 9S FAB mass spectrum (positive mode, NOBA) 1308 [M + H]+ and (negative mode, NOBA) 1306 [M - H]-.
The other phosphoramidite dinucleoside blocks 10F, 10S, 11F and 11S were prepared in the same manner. Each exhibited signals for phosphoramidite and phosphoramidate in a 1:1 ratio.
10F. Yield 31%, 31P NMR (CD3CN, 101 MHz) [delta] 150.12, 149.77 and 8.50, 8.44.
10S. Yield 21%, 31P NMR (CD3CN, 101 MHz) [delta] 150.15, 149.78 and 8.42, 8.35.
11F. Yield 65%, 31P NMR (CD3CN, 101 MHz) [delta] 150.16, 149.69 and 10.39, 10.34.
11S. Yield 58%, 31P NMR (CD3CN, 101 MHz) [delta] 150.36, 149.75 and 10.50, 10.41.
Protected 2[prime]-deoxy-[alpha]-ribonucleoside-3[prime]-O-H-phosphonates 12, 13 and 14 (Scheme 3)
2[prime]- Deoxy-[alpha]-ribonucleosides-3[prime]-O-H-phosphonate ([alpha]-dT 12, [alpha]-dCBz 13 and [alpha]-dABz 14) were prepared from base-protected 5[prime]-O-dimethoxytrityl-2[prime]-deoxy [alpha]-nucleosides (30) according to a published procedure (32).
5[prime]-O-(4,4[prime]-dimethoxytrityl)-[alpha]-thymidine-3[prime]-O-H-phosphonate (triethylammonium salt) 12. Yield 70%. 31P NMR (CDCl3, 200 MHz) [delta] 4.46, 1JP-H = 626 Hz, 3JP-H = 8.9 Hz. FAB mass spectrum (negative mode, thioglycerol) 607 [M - H]-.
5[prime]-O-(4,4[prime]-dimethoxytrityl)-4-N-benzoyl-2[prime]-deoxy-[alpha]-cytidine-3[prime]-O-H-phosphonate (triethylammonium salt) 13.Yield 70%. 31P NMR (CDCl3, 101 MHz) [delta] 3.62, 1JP-H = 626 Hz, 3JP-H = 8.9 Hz. FAB mass spectrum (negative mode, thioglycerol) 696 [M - H]-.
5[prime]-O-(4,4[prime]-dimethoxytrityl)-6-N-benzoyl-2[prime]-deoxy-[alpha]-adenosine-3[prime]-O-H-phosphonate (triethylammonium salt) 14.Yield 76%. 31P NMR (CDCl3, 101 MHz) [delta] 3.69, 1JP-H = 623 Hz, 3JP-H = 8.7 Hz. FAB mass spectrum (negative mode, thioglycerol) 720 [M - H]-.
Oligonucleotides synthesis
[beta]- and [alpha]-oligonucleoside phosphodiesters 15, 15a, 15b, 15c, 16, 16a, 16b, 16c, 17 and 18. Protected deoxy-2[prime]-[beta]-ribonucleoside-3[prime]-O-(N,N-diisopropyl)cyanoethylphosphoramidites were purchased from PerSeptive Biosystems GmbH and deoxy-2[prime]-[alpha]-ribonucleoside-3[prime]-O-(N,N-diisopropyl)cyanoethylphosphoramidites ([alpha]-dT, [alpha]-dCBz and [alpha]-dABz) were prepared according to published procedures (30). The RNA target oligomer r(AGA-AUUGGGUGU) 15 was a gift from ISIS Pharmaceuticals (Carlsbad, CA). The DNA target oligomer d(AGAATTGGGTGT) 16 and the complementary phosphodiesters [beta]-d(ACACCCAATTCT) 17 and [alpha]-d(TCTTAACCCACA) 18 were prepared on the 1 µmol scale with an Applied Biosystems Model 381A DNA synthesizer. The protected oligomers were cleaved from the support and deprotected upon treatment with concentrated ammonium hydroxide for 5 h at 55°C. The oligomers were purified by preparative reverse phase HPLC using a linear gradient of 5-50% acetonitrile in 50 mM TEAAC.
[beta]- and [alpha]-oligonucleoside phosphodiesters containing a single chirally enriched phosphoramidate linkage 19F, 19S, 20F, 20S, 21F, 21S, 22F, 22S, 23F, 23S, 24F and 24S. Oligonucleotides were prepared on the 1 µmol scale following standard procedures of phosphoramidite chemistry except that the coupling step of the chirally enriched dinucleotide was repeated three times with extended (3 min) coupling times. The oligomers were purified by preparative reverse phase HPLC using a linear gradient of 7.5-25% acetonitrile in 50 mM TEAAC.
MS [negative mode MALDI-TOF, matrix: 2,4,6-trihydroxyacetophenone (THAP) and ammonium citrate]: 19F m/z calculated 3605.5, found 3603.2; 19S m/z calculated 3605.5, found 3606.4; 20F m/z calculated 3632.5, found 3631.8; 20S m/z calculated 3632.5, found 3632.1; 21F m/z calculated 3617.5, found 3618.5; 21S m/z calculated 3617.5, found 3618.6; 22F m/z calculated 3605.5, found 3603.1; 22S m/z calculated 3605.5, found 3606.4; 23F m/z calculated 3632.5, found 3629.4; 23S m/z calculated 3632.5, found 3631.3; 24F m/z calculated 3617.5, found 3617.7; 24S m/z calculated 3617.5, found 3616.7.
[beta]- and [alpha]-oligonucleoside phosphodiesters containing a single chirally enriched methylphosphonate linkage 25f and 25s and 26f and 26s. Oligonucleotides were prepared on the 1 µmol scale using the appropriate protected nucleoside phosphoramidite and methylphosphonamidite synthons. The oligomers 25 and 26 were deprotected and cleaved from the support by sequential treatment with ammonium hydroxide and ethylenediamine (33). The [beta]-oligomer 25 was then purified by reverse phase HPLC using an EC Nucleosil 5µ C18 EC column (150 × 4.6 mm; Macherey-Nagel) and a linear gradient of 11-13% acetonitrile in 50 mM TEAAC, pH 7. The diastereoisomers of the [beta]-oligomer 25 were resolved into two peaks and each diastereoisomer was collected separately as 25f for the `fast' eluting isomer on reverse phase C18 (HPLC RT 33.07 min, 100% diastereoisomeric purity) and 25s for the `slow' eluting isomer on reverse phase C18 (HPLC RT 34.76 min, 100% diastereoisomeric purity). The diastereoisomers of the [alpha]-oligomer 26 were separated into two peaks by reverse phase HPLC using a Lichrospher 5µ C18 EC column (150 × 4.6 mm; Phase Sep) and a linear gradient of 7-9.5% acetonitrile in 50 mM ammonium acetate, pH 5.6.
The two diastereoisomers were collected separately as: 26f, HPLC RT 28.77 min (100% diastereoisomeric purity); 26s, HPLC RT 32.32 min (90% diastereoisomeric purity). MS (negative mode MALDI-TOF, matrix: THAP and ammonium citrate) m/z calculated 3546.4, 25f found 3545.7, 25s found 3545.3, 26f found 3547.6, 26s found 3544.0.
[beta]- and [alpha]-oligonucleoside N-alkylphosphoramidates 27-30. These were synthesized via hydrogen phosphonate chemistry (34). Amidative oxidation of H-phosphonate diesters was performed manually using a 10% solution of either 2-methoxyethylamine or morpholine in an anhydrous CCl4/pyridine solution (50:50 v/v). The ODNs were removed from the solid support and deprotected with concentrated NH4OH at 55°C for 5 h. Purification of ODNs was performed by preparative reverse phase HPLC using a 50 min linear gradient of 8-40% acetonitrile in 50 mM TEAAC. They were characterized by MALDI-TOF mass spectrometry. MS (negative mode MALDI-TOF, THAP and ammonium citrate): 27 m/z calculated 4176.4, found 4174.8; 28 m/z calculated 4308.5, found 4307.2; 29 m/z calculated 4176.4, found 4175.6; 30 m/z calculated 4308.5, found 4306.2.
[beta]- and [alpha]-oligonucleoside methylphosphonates. These were synthesized according to published procedures (29,33) on the 1 µmol scale using either [beta]-methylphosphonamidite or [alpha]-methyl-phosphonamidite synthons. Their purification and characterization were described in a previous publication (29).
Melting experiments
The concentration of each separated oligonucleotide was determined spectrophotometrically at 260 nm and at 80°C assuming that the molar extinction coefficient of the [alpha]- and [beta]-oligomers is the sum of the molar extinction coefficients of the constitutive deoxynucleotides. Optical measurements were performed on a UVIKON 931 spectrophotometer (Kontron). The temperature of the cell was controlled by a Huber PD415 temperature programmer connected to a refrigerated ethyleneglycol/water bath (Huber Ministat). Cuvettes were 1 cm path lengh quartz cells. The cell compartment was continuously flushed with dry nitrogen when the temperature was below room temperature. Prior to the experiments, the oligonucleotides were mixed together each at a final concentration of 3 µM in 100 mM sodium chloride, 10 mM sodium cacodylate (pH 7) in the presence of 0.3% dioxane and allowed to incubate at 90°C for 30 min. During the hybridization (melting) experiments the cooling (heating) rate was 0.5°C/min. Digitized absorbance and temperature values were stored in a computer for subsequent plotting and analysis. Tm values were determined from the maxima of the first derivative plots of absorbance versus temperature.
RESULTS AND DISCUSSION
Preparation of oligonucleotides containing a single chirally enriched phosphoramidate linkage
To examine the effect of chirality on binding interactions, [alpha]- and [beta]-oligodeoxynucleotides containing a single chirally enriched phosphoramidate linkage in the middle of the sequence (Table 1) were prepared on a solid support using the appropriately protected nucleoside phosphoramidite synthons. These ODNs were complementary to the splice acceptor site AGAAUUGGGUGU of the mRNA coding for HIV-1 tat protein, a target 15 already used in our previous studies (26,29). The central phosphoramidate linkage was introduced through a 3[prime]-phosphoramidite dinucleoside building block containing a chirally enriched phosphoramidate internucleoside linkage. The fully protected [alpha]-dinucleoside phosphoramidates 3-5 were prepared according to a procedure described in the literature for the synthesis of [beta]-analogs (21) (Scheme 2). After removal of the TBDMS group, separation from the mixture of each diastereoisomer was achieved by chromatography on silica gel. The absolute configuration of the phosphoramidate linkage in the dinucleotide was not assigned. The two diastereoisomers of each dinucleotide were identified as `fast' (or F) and `slow' (or S) isomers according to their chromatographic mobility on a silica gel column. Subsequent phosphitylation afforded the phosphoramidite synthons 9F, 9S, 10F, 10S, 11F and 11S employed for the synthesis of oligonucleotides 22F and 22S, 23F and 23S, and 24F and 24S, respectively. Similarly, the [beta]-oligonucleotides 19F, 19S, 20F, 20S, 21F and 21S were synthesized.
Scheme 2. Synthesis of 3[prime]-phosphoramidite [alpha]-dinucleoside N-alkylphosphoramidates. (a) Tetrazole, CH3CN, 30 min, then I2, 2-methoxyethylamine or morpholine or dimethylaminopropylamine, THF, 5 min. (b) TBAF 1 M in THF. (c) P[O(CH2)2CN][(i-Pr)2N]2, diisopropylammonium tetrazolide, CH2Cl2, 20 h.
Table 1. Oligonucleotides synthesized
| No. | Oligonucleotide | Sequence | Backbone |
| 15 | RNA target | 5[prime]-AGA AUU GGG UGU-3[prime] | PO |
| 15a | RNA target (CC mismatch) | 5[prime]-AGA AUU CGG UGU-3[prime] | PO |
| 15b | RNA target (CU mismatch) | 5[prime]-AGA AUU UGG UGU-3[prime] | PO |
| 15c | RNA target (CA mismatch) | 5[prime]-AGA AUU AGG UGU-3[prime] | PO |
| 16 | DNA target | 5[prime]-AGA ATT GGG TGT-3[prime] | PO |
| 16a | DNA target (CC mismatch) | 5[prime]-AGA ATT CGG TGT-3[prime] | PO |
| 16b | DNA target (CT mismatch) | 5[prime]-AGA ATT TGG TGT-3[prime] | PO |
| 16c | DNA target (CA mismatch) | 5[prime]-AGA ATT AGG TGT-3[prime] | PO |
| 17 | [beta]-oligo DNA | 5[prime]-ACA CCC AAT TCT-3[prime] | PO |
| 18 | [alpha]-oligo DNA | 5[prime]-TCT TAA CCC ACA-3[prime] | PO |
| 19F and 19S | [beta]-oligo with 1 modification | 5[prime]-ACA CCC *AAT TCT-3[prime] | PO/PNHME |
| 20F and 20S | [beta]-oligo with 1 modification | 5[prime]-ACA CCC *AAT TCT-3[prime] | PO/PNHDMAP |
| 21F and 21S | [beta]-oligo with 1 modification | 5[prime]-ACA CCC *AAT TCT-3[prime] | PO/PMOR |
| 22F and 22S | [alpha]-oligo with 1 modification | 5[prime]-TCT TAA* CCC ACA-3[prime] | PO/PNHME |
| 23F and 23S | [alpha]-oligo with 1 modification | 5[prime]-TCT TAA* CCC ACA-3[prime] | PO/PNHDMAP |
| 24F and 24S | [alpha]-oligo with 1 modification | 5[prime]-TCT TAA* CCC ACA-3[prime] | PO/PMOR |
| 25f and 25s | [beta]-oligo with 1 modification | 5[prime]-ACA CCC *AAT TCT-3[prime] | PO/MP |
| 26f and 26s | [alpha]-oligo with 1 modification | 5[prime]-TCT TAA* CCC ACA-3[prime] | PO/MP |
| 27 | [beta]-oligo fully modified | 5[prime]-ACA CCC AAT TCT-3[prime] | PNHME |
| 28 | [beta]-oligo fully modified | 5[prime]-ACA CCC AAT TCT-3[prime] | PMOR |
| 29 | [alpha]-oligo fully modified | 5[prime]-TCT TAA CCC ACA-3[prime] | PNHME |
| 30 | [alpha]-oligo fully modified | 5[prime]-TCT TAA CCC ACA-3[prime] | PMOR |
F, S for `fast' and `slow' isomers according to their chromatographic mobility on a silica gel column of the P-modified dinucleotide incorporated in the ODN (see Materials and Methods).
f, s for `fast' and `slow' isomers according to their chromatographic mobility on a reverse phase HPLC C18 column.
Evaluation of the chirality effect on hybridization properties of [alpha]-oligonucleoside N-alkylphosphoramidates
For each phosphoramidate modification, two [alpha]- (F and S) and two [beta]-oligonucleotides (F and S) were hybridized to their complementary RNA 15 and DNA 16 strands. These interactions were investigated by thermal denaturation and renaturation analyses using change in absorbance at 260 nm versus temperature. No significant differences were observed between thermal association and dissociation curves. Melting temperatures (Tm) were determined at the inflexion point of the sigmoid curves (Table 2). The thermal stability of duplexes having a single phosphoramidate linkage (F or S) was compared to that of the corresponding all-phosphodiester duplexes under identical conditions. The difference in Tm ([Delta]Tm) reported in Figure 1 accounts for the effect due to the stereoselective introduction of one phosphoramidate linkage in the various duplexes. Additionally, data obtained with [alpha]- and [beta]-oligonucleotides 26f and 26s and 25f and 25s bearing a single chirally enriched methylphosphonate linkage are reported.
Figure 1. Variation in Tm induced by the stereoselective introduction of a single phosphoramidate or methylphosphonate linkage in duplexes formed between [alpha]- or [beta]-oligonucleotides and their DNA target 16 or RNA target 15. Position of the phosphoramidate or methylphosphonate linkage is indicated in Table 1.
Table 2. Thermal stability (Tm and [Delta]Tm) of a duplex formed between [alpha]- or [beta]-oligonucleotides containing one chirally enriched N-alkylphosphoramidate or methylphosphonate linkage and their DNA or RNA targets
| Oligonucleotide | DNA target 16 | RNA target 15 | |||
| Tm (°C) | [Delta]Tm (°C) | Tm (°C) | [Delta]Tm (°C) | ||
| [beta]-PO | 17 | 45.4 | 44.5 | ||
| [beta]-PNHME | 19F | 44.4 | -1.0 | 42.0 | -2.5 |
| [beta]-PNHME | 19S | 43.4 | -2.0 | 40.7 | -3.8 |
| [beta]-PNHDMAP | 20F | 44.2 | -1.2 | 42.6 | -1.9 |
| [beta]-PNHDMAP | 20S | 43.6 | -1.8 | 41.3 | -3.2 |
| [beta]-PMOR | 21F | 43.4 | -2.0 | 41.6 | -2.9 |
| [beta]-PMOR | 21S | 42.9 | -2.5 | 40.5 | -4.0 |
| [beta]-MP | 25f | 44.7 | -0.7 | 43.5 | -1.0 |
| [beta]-MP | 25s | 43.7 | -1.7 | 41.6 | -2.9 |
| [alpha]-PO | 18 | 43.4 | 41.5 | ||
| [alpha]-PNHME | 22F | 42.9 | -0.5 | 41.4 | -0.1 |
| [alpha]-PNHME | 22S | 44.9 | +1.5 | 42.2 | +0.7 |
| [alpha]-PNHDMAP | 23F | 42.8 | -0.6 | 41.3 | -0.3 |
| [alpha]-PNHDMAP | 23S | 46.0 | +2.6 | 41.2 | -0.2 |
| [alpha]-PMOR | 24F | 42.4 | -1.0 | 40.6 | -0.9 |
| [alpha]-PMOR | 24S | 43.4 | 0 | 41.0 | -0.5 |
| [alpha]-MP | 26f | 45.3 | +1.9 | 41.8 | +0.3 |
| [alpha]-MP | 26s | 43.4 | 0 | 41.1 | -0.4 |
With DNA target 16 (Fig. 1), regarding [beta]-oligonucleotides, whatever the backbone modification the two diastereoisomers (designated F and S) formed duplexes less stable than the reference PO [beta]-ODN 17 (-2.5°C < [Delta]Tm < -0.7°C). As expected, one isomer bound to DNA target 16 less tightly than the other. With [alpha]-ODNs, a difference between the two isomers was also observed (from 1.0 to 3.2°C) and it was even more pronounced than with [beta]-ODNs (from 0.5 to 1.0°C). It is noteworthy that with modified [alpha]-ODNs, one isomer had no effect or a stabilizing effect on the duplex (0°C < [Delta]Tm < 2.6°C) whereas the other isomer had a slightly destabilizing effect (-1.0°C < [Delta]Tm < 0°C).
With RNA target 15 (Fig. 1), the backbone modification introduced in [beta]-oligonucleotides destabilized duplexes more dramatically than with DNA target 16 (-4.0°C < [Delta]Tm < -1.0°C). In contrast, the modifications of the internucleoside linkage in [alpha]-ODNs induced a weaker effect whatever the P substituent (-0.9°C < [Delta]Tm < 0.7°C) and the stereoisomeric conformation (from 0.1 to 0.8°C) of the modified linkage. This suggests that the chirality effect in duplexes formed between phosphate-modified [alpha]-ODNs and the RNA target is reduced to such an extent that a stereoselective synthesis of these ODN analogs would not be required for efficient binding to the target.
Effect of steric hindrance of the P substituent on hybridization properties
To study the steric hindrance effect, fully modified [beta]- and [alpha]-oligonucleoside N-alkylphosphoramidates (PNHME or PMOR) 27 and 28 and 29 and 30 with the same sequence as mentioned abovewere synthesized. These analogs were prepared in a non-stereocontrolled way, via H-phosphonate chemistry using the appropriately protected nucleoside H-phosphonate synthons (Scheme 3) and following a standard procedure (34). The coupling yields averaged 96-97% per cycle. The H-phosphonate diester linkages were then oxidized by CCl4 in the presence of 2-methoxyethylamine or morpholine. After deprotection, oligonucleoside N-alkylphosphoramidates were purified by reverse phase HPLC. The ODN N-alkylphosphoramidates obtained as diastereoisomeric mixtures were hybridized to their complementary DNA and RNA strands. The Tm values presented in Figure 2 (versus DNA target and RNA target) were compared to that obtained with PO [beta]- and [alpha]-oligonucleosides 17 and 18, and to [beta]- and [alpha]-dodecanucleoside methylphosphonates previously studied (29).
Scheme 3. Protected 2[prime]-deoxy-[alpha]-ribonucleoside-3[prime]-O-H-phosphonates used for the synthesis of fully modified oligonucleoside phosphoramidates.
Figure 2. Thermal stability of fully phosphate-modified [beta]- and [alpha]-ODNs with their DNA target 16 or RNA target 15. Experimental conditions as indicated in Table 2.
First, in agreement with previous studies (13), the introduction of non-ionic internucleotide linkages into the [beta]-ODNs greatly decreased the thermal stability of the hybrids formed with either complementary DNA or RNA strands according to the bulkiness of the phosphorus substituent. When PO linkages were replaced by PMOR linkages in [beta]-ODNs, the Tm value of the duplex formed with DNA decreased by 32.4°C, whereas the [Delta]Tm between PO and PNHME was 15.9°C. This destabilization induced by the pendant group was more pronounced regarding hybrids formed between [beta]-ODNs and their RNA target (MP [Delta]Tm 18.3°C, PNHME [Delta]Tm 29.6°C, PMOR [Delta]Tm 39.6°C). In contrast, the phosphorus substituent affects to a very low extent the hybridization of [alpha]-ODNs with DNA as well as with RNA targets. Surprisingly, the bulky morpholino group in the [alpha]-series slightly destabilized the hybrid with the DNA target compared to the parent PO [beta]-ODN-DNA duplex ([Delta]Tm -3.4°C). When RNA was the target, the destabilization was more pronounced ([Delta]Tm -9.1°C).
Secondly, the results indicate that whatever the backbone modification, the [alpha]-oligonucleotides formed much more stable duplexes with the DNA target than did their [beta]-homologs. This difference in stability of the hybrids was more pronounced with ODN N-alkylphosphoramidates (25.5°C for PNHME, 29.0°C for PMOR) than with MP oligos (8.0°C for MP). The same effect was observed when RNA was the target (31.0°C for PNHME, 30.5°C for PMOR, 12.5°C for MP) (Fig. 2).
Finally, MP [alpha]-ODN and PNHME [alpha]-ODN formed more stable duplexes with the DNA target than did the natural PO [beta]-ODN ([Delta]Tm +2.6°C and +9.6°C, respectively). Among all the oligonucleotides studied, the [alpha]-oligonucleoside methoxyethylphosphoramidate has the best affinity for its DNA ([Delta]Tm +9.6°C ) and RNA ([Delta]Tm +1.4°C) targets.
Specificity of hybridization of [alpha]-oligonucleoside N-alkyl-phosphoramidates
To examine whether the high affinity of the PNHME [alpha]-ODN for complementary strands was detrimental to the base pairing specificity, we measured the influence of a single mismatch in the DNA 16 or RNA 15 targets on the stability of the hybrids (Table 3). Introduction of one CC, CT or CA mismatch instead of the regular match CG in the hybrids formed with PNHME [alpha]-ODN 29 and DNA targets 16a, 16b and 16c decreased the melting temperature by 18.8, 19.5 and 18.5°C, respectively. These values compare well with those obtained when the same mismatches were introduced into the corresponding duplexes formed with unmodified PO [beta]-oligomer 17 (19.5, 17.2 and 20.2°C, respectively). When PMOR [alpha]-ODN 30 was hybridized to DNA targets 16a, 16b and 16c, an average decrease of 19.3°C was found, similar to that (18.9°C) observed with the PNHME [alpha]-ODN and [beta]-PO ODN. The specificity of hybridization of [alpha]-oligonucleoside N-alkylphosphoramidates was also studied with RNA targets and similar results were obtained. Destabilization of the duplex relative to the perfectly matched duplex was not dependent upon the bulkiness of the P-substituent. These data indicate that the Watson-Crick base pairing specificity of [alpha]-ODN phosphoramidates is as high as that of natural oligonucleoside phosphodiesters despite their relatively high affinity.
Table 3. Influence of a single mismatch on the Tm of duplexes with N-alkylphosphoramidates [alpha]-oligonucleosides
| ODN | Tm (°C) | ||||||||
| DNA targets | RNA targets | ||||||||
| 16 CG | 16a CC | 16b CT | 16c CA | 15 CG | 15a CC | 15b CU | 15c CA | ||
| [beta]-PO | 17 | 45.4 | 25.9 | 28.2 | 25.2 | 44.6 | 24.8 | 23.3 | 28.7 |
| [alpha]-PNHME | 29 | 55.0 | 36.2 | 35.5 | 36.5 | 46.0 | 25.0 | 22.5 | 26.3 |
| [alpha]-PNMOR | 30 | 42.0 | 23.0 | 22.5 | 22.5 | 35.5 | 15.0 | 14.0 | 15.9 |
Influence of salt concentration on the thermal stability of hybrids
Experiments at 0.1 M NaCl or in the absence of NaCl showed that the Tm values for the hybrids formed between fully modified PNHME [alpha]-ODN 29 and DNA 16 or RNA 15 targets were not modified by changes in the ionic strength of the medium (Table 4). This behavior reflects the absence of charge repulsion between the non-ionic oligonucleotide and the negatively charged phosphodiester backbone of the target strand as observed for neutral backbone [beta]-analogs (13,35). In the absence of NaCl, PNHME [alpha]-ODN 29 formed much more stable duplexes with DNA 16 ([Delta]Tm 26.7°C) and RNA 15 ([Delta]Tm 13.8°C) than PO [beta]-ODN 17.
Table 4. Salt concentration dependence of the Tm (°C) of the fully modified phosphoramidate oligomer 29 with RNA 15 and DNA 16 targets compared with the natural duplex formed with [beta]-PO 17
| ODN | Tm (°C) | ||||
| DNA target 16 | RNA target 15 | ||||
| 0.1 M NaCl | No NaCl | 0.1 M NaCl | No NaCl | ||
| [beta]-PO | 17 | 45.4 | 28.3 | 44.6 | 31.2 |
| [alpha]-PNHME | 29 | 55.0 | 55.0 | 46.0 | 45.0 |
| [Delta]Tm (°C) | +9.6 | +26.7 | +1.4 | +13.8 | |
In conclusion, we have demonstrated that inversion of the anomeric configuration (from [beta] to [alpha]) of the sugar moieties in fully modified oligonucleoside N-alkylphosphoramidates increases their affinity for ssDNA and ssRNA targets. An enhancement of affinity for DNA and RNA induced by the combination of two structural modifications was previously observed with non-ionic oligonucleoside PNH2 (28) and MP ODNs (29) and, to a lesser extent, with ionic PS ODNs (27), and this finding seems to be a general phenomenon for backbone-modified oligodeoxynucleotides in the [alpha]-anomeric configuration. The high binding of 2-methoxyethylphosphoramidate ODNs to DNA is promising for triplex formation and we are currently studying the affinity of these backbone-modified [alpha]-oligonucleotides for dsDNA. Furthermore, [alpha]-oligonucleoside N-alkylphosphoramidates offer several advantages over phosphodiester ODNs and phosphorothioate analogs that include increased affinity, high specificity and ability to hybridize under low salt conditions. These properties make them attractive for antisense purposes as well as for use as diagnostics probes.
ACKNOWLEDGEMENTS
The authors are grateful to Jean-Charles Bologna for synthesis of chirally enriched [beta]-dinucleoside N-alkylphosphoramidates and the corresponding oligonucleotides used as references in the study of the chirality effect on hybridization properties. The authors also thank Albert Meyer for synthesizing [alpha]-nucleosides. This work was supported by the Agence Nationale de Recherche sur le Sida (ANRS, France). A.L. thanks the Association pour la Recherche contre le Cancer (ARC, France) for a research studentship.
SUPPLEMENTARY MATERIAL
See Supplementary Material available in NAR online (61 KB PDF file).
REFERENCES
*To whom correspondence should be addressed. Tel: +33 4 67 14 38 98; Fax: +33 4 67 04 20 29; Email: debart{at}univ-montp2.fr
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  T. Michel, C. Martinand-Mari, F. Debart, B. Lebleu, I. Robbins, and J.-J. Vasseur Cationic phosphoramidate {alpha}-oligonucleotides efficiently target single-stranded DNA and RNA and inhibit hepatitis C virus IRES-mediated translation Nucleic Acids Res., September 15, 2003; 31(18): 5282 - 5290. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||