ABSTRACT
2',5'-Linked oligo-3'-deoxyribonucleotides bind selectively to complementary RNA but not to DNA. These oligonucleotides (ODNs) do not recognize double-stranded DNA by Hoogsteen triplex formation and the complexes formed by these ODNs with RNA are not substrates for Escherichia coli RNase H. Substitution of the 2',5'-phosphodiester backbone by phosphorothioate linkages gives 2',5'-linked oligo-3'-deoxynucleoside phosphorothioate ODNs that exhibit significantly less non-specific binding to cellular proteins or thrombin. Incorporation of a stretch of seven contiguous 3',5'-linked oligo-2'-deoxynucleoside phosphorothioate linkages in the center of 2',5'-linked ODNs (as a putative RNase H recognition site) afford chimeric antisense ODNs that retain the ability to inhibit steroid 5[alpha]-reductase (5[alpha]R) expression in cell culture.
A number of recent reports have demonstrated that 2',5'-linked oligonucleotides bind selectively to complementary single-stranded RNA but not to DNA (1 -3 ). This raises the possibility of the utility of 2',5'-linked oligonucleotides in potential antisense and diagnostic applications. Previously, 2',5'-linked oligoadenylates (2',5'-An) were detected in a variety of cells and tissues, including L1210 cells and human lymphocytes (4 ). Studies employing the 2',5'-oligoadenylate system indicate that this unique linkage is involved in the regulation of cell growth, cell differentiation and the observed antiviral effects of interferon (5 ). In the 2',5'-An pathway interferon and double-stranded RNA activate 2',5'-oligoadenylate synthetase to catalyze the formation of 2',5'-linked oligoadenylates from ATP. While these oligoadenylates vary in length from 2 to 15 residues, di-, tri- and tetra-adenylates are the most abundant products formed in the reaction and the amounts of larger oligoadenylate units diminish with increasing chain length (6 ). The 2',5'-An oligomers subsequently bind to RNA and activate a ubiquitous endoribonuclease (RNase L), which is then responsible for cleavage of RNA. RNase L is an enzyme found in many eukaryotic cells (7 -8 ) and non-specific cleavage of messenger and ribosomal RNA is often the result of activation of this endoribonuclease by 2',5'-An (9 ). Additionally, these oligoribonucleotides are also reported to inhibit the activities of HIV-1 reverse transcriptase (10 ) and DNA topoisomerase-I in HIV-1-infected cells (11 ). When conjugated with an antisense 3',5'-linked phosphodiester oligonucleotide, the 2',5'-An portion of the chimeric oligonucleotide can also direct antisense cleavage of target RNA, presumably through activation of RNase L in cell culture (12 ).
All solvents were HPLC grade. Acetonitrile, dichloromethane and pyridine were dried over calcium hydride and distilled. Methanol was dried by refluxing over magnesium methoxide before distillation. [3H]1,2-Benzodithiol-3-one-1,1-dioxide (Beaucage reagent), 2'-deoxy- and 2'-O-t-butyldimethylsilyl 3'-(cyanoethyl)ribonucleoside phosphoramidite monomers and LCAA-CPG bearing deoxyribo- or ribonucleosides were obtained from Chem Genes Corp. (Waltham, MA). Triethylamine trihydrofluoride used for silyl deprotection in RNA was obtained from Aldrich Chemicals (Milwaukee, WI). [[gamma]-32P]ATP, polynucleotide kinase, snake venom phosphodiesterase and shrimp alkaline phosphatase were obtained from Amersham-US Biochemicals Inc. (Cleveland, OH). Stains-Alltm was obtained from Sigma Chemical Co. (St Louis, MO).
2,4-Diaminopurine (15 g, 100 mmol) was suspended in dry pyridine (200 ml) and acetic anhydride (31 ml, 324 mmol) was added. The mixture was stirred with refluxing for 3.5 h, cooled and allowed to stand at room temperature overnight. Ethanol (100 ml) was added and the solid filtered, washed with ethanol and dried. The solid was next stirred with a saturated bicarbonate solution (200 ml) for 1 h, diluted with water (500 ml) and filtered. The residue was washed with water and dried over phosphorus pentoxide to give 15.5 g (66% yield) of the product. 1H-NMR (DMSO-d6) d, 10.16, bs, (D2O exchange), 1H, imidazole N-H; 8.29, s, 1H, H-8; 3.20, bs, 2H, N-H; 2.25, 2.18, 2s, 3H each, Ac-CH3.
Guanine (15.1 g, 100 mmol) was suspended in dry 1-methyl-2-pyrrolidinone (150 ml) and acetic anhydride (25 ml, 262 mmol) added. The reaction mixture was heated to 150oC for 2 h. The resulting solution was stirred at room temperature for 12 h. The solid was filtered, washed with acetone and dried to give 20.13 g (86% yield) of 2-N,9-diacetyl guanine. 1H-NMR (DMSO-d6) d, 12.19, 11.72, 2 bs (D2O exchangeable), 1H each, N-H; 8.42, s, 1H, H-8; 2.79, 2.19, 2s, 3H each, Ac-CH3.
The bis-acetylated compound (11.76 g, 50 mmol) was suspended in a mixture of dry pyridine (250 ml) and diisopropyl ethylamine (17.4 ml), followed by slow addition of 1.2 equiv. of diphenylcarbamoyl chloride. The color changes to orange immediately. The reaction mixture was stirred for 1 h followed by addition of 25 ml ice-cold water. The mixture was stirred for an additional 10 min and solvent removed in vacuo. The residue was co-evaporated with toluene (3 * 100 ml) and the pink-orange residue resuspended in 50% ethanol (600 ml). The reaction mixture was heated to 100oC for 1.5 h and kept at room temperature for 24 h. The solid was filtered, washed with ethanol and dried to give 18.45 g product (95% yield). 1H-NMR (DMSO-d6) 13.56, bs, (D2O exchangeable), 1H, imidazole N-H; 10.62 bs, (D2O exchangeable),1H, N-H; 8.44, s, 1H, H-8; 7.47-7.0, m, 10H, Ar-H; 2.15, s, 3H, Ac-CH3.
2-N-Acetyl-6-O-diphenylcarbomyl guanine
The protecting groups in
Nucleoside
Compound
3'-Deoxyguanosine
Remaining 3'-deoxynucleosides, e.g. 3'-deoxycytidine, 3'-deoxy-5-methyluridine and 3'-deoxyadenosine, were prepared using reported literature procedures (13 -15 ).
1H-NMR (CDCl3) d, 8.73, bs, (D2O exchange), N-H; 8.45, d, 1H, H-6; 7.87, d, 1H, H-5; 7.63-7.16, m, 14H, Ar-H; 6.86, d, 4H, o-Ar-H to OCH3, 5.79, s, 1H, H-1'; 4.70, m, 1H, H-2'; 4.51, bs, 1H, H-4'; 4.31, bs, (D2O exchange), O-H; 3.80, s, 6H, OCH3; 3.56, 3.31, 2d, 2H, H-5'; 2.22, 2.06, 2m, 2H, H-2'. HRMS: m/e calculated for C37H35O7N3 (M+) 633.2430, observed (M+) 633.2475.
1H-NMR (CDCl3) d, 10.24, bs, (D2O exchange), N-H; 7.80, s, 1H, H-6; 7.53-7.16, m, 9H, Ar-H; 6.83, d, 4H, o-Ar-H to OCH3, 5.76, s, 1H, H-1'; 4.92, m, 1H, H-2'; 4.65, bs, 1H, H-4'; 4.51, bs, (D2O exchange), O-H; 3.78, s, 6H, OCH3; 3.57, 3.28, 2d, 2H, H-5'; 2.27, 2.02, 2m, 2H, H-2'; 1.41, s, 3H, C-CH3. HRMS: m/e calculated for C31H32O7N2 (M+) 544.2209, observed (M+) 544.2211.
1H-NMR (CDCl3) d, 9.12, bs, (D2O exchange), N-H; 8.79, s, 1H, H-8; 8.29, s, 1H, H-2; 7.59-7.20, m, 14H, Ar-H; 6.78, d, 4H, o-Ar-H to OCH3, 5.99, s, 1H, H-1'; 4.90, bs, (D2O exchange),1H, O-H; 4.70, m, 1H, H-2'; 4.64, bs, 1H, H-4'; 3.77, s, 6H, OCH3; 3.42, 3.27, m, 2H, H-5'; 2.34, 2.23, m, 2H, H-2'. HRMS: m/e calculated for C38H35O6N5 (M+) 657.2587, observed (M+) 657.2580.
Oligonucleotides were synthesized on a Perseptive Biosystems Expeditetm Nucleic Acid Synthesizer system using standard phosphoramidite chemistry. For 2',5'-linked oligodeoxynucleotides, commercially available controlled pore glass (lcaa-CPG-500) loaded with protected ribonucleosides was used as solid support. The average coupling reaction yield was 98.0-99.5%, as determined by absorbance of the dimethoxytrityl cation liberated following treatment with 3% trichloroacetic acid in methylene chloride. For thiolated oligonucleotides, Beaucage reagent was used as the sulfur transfer reagent. Protected oligodeoxynucleotides were cleaved from the solid support (concentrated ammonium hydroxide, 4 h, 25oC) and the protecting groups removed (by heating at 55oC for 18 h). Ammonium hydroxide solution was evaporated in vacuo and the crude mixture purified by HPLC and desalted. By this protocol oligonucleotides of >97% purity were obtained.
High pressure liquid chromatography was carried out on a Beckman System Gold Chromatograph equipped with a diode array UV-Visible detector operating at 260 nm wavelength. Ion exchange chromatography of trityl-off thiolated oligonucleotides was carried out on a Whatman Partisphere Wax column (4.6 * 125 mm) employing a 0-40% gradient of 100-500 mM ammonium sulfate in 20 mM potassium phosphate buffer, pH 6.4, acetonitrile (3:1) over 40 min at a flow rate of 1 ml/min. Reversed phase chromatography of trityl-on phosphodiester and trityl-on thiolated oligonucleotides was carried out on a Beckman Ultrasphere 5 [mu] column (4.6 * 250 mm) employing a gradient of 0-60% CH3CN in 50 mM triethyl ammonium bicarbonate buffer, pH 7.2, at a flow rate of 1 ml/min. The purified dimethoxytritylated oligonucleotides were treated with 80% acetic acid for 0.5 h to remove the 5'-terminal dimethoxytrityl protecting group. Oligoribonucleotides were deprotected from the solid support with a solution of saturated ammonia in dry ethanol (dried and distilled over magnesium ethoxide) for 18 h at 55oC followed by treatment with a solution of TEA-HF for 24 h at 37oC. Butanol was added directly to this mixture to precipitate oligoribonucleotides at -70oC. This crude mixture of oligoribonucleotides was purified by ion exchange chromatography on a Dionex Nucleopac column (4 * 250 mm) employing a 0-40% gradient of 20-600 mM lithium chlorate in 20 mM sodium acetate buffer, pH 7.6, at a flow rate of 1 ml/min. Purified oligoribonucleotides were desalted on Sep-Paktm reversed phase columns, precipitated with ethanol in the presence of 3 M sodium chloride and lyophilized. The final purity of oligonucleotides was checked by HPLC and denaturing polyacrylamide gel electrophoreasis, which was carried out on 20% polyacrylamide slabs containing 7 M urea. Wet gels containing 32P-labeled oligonucleotides were autoradiographed at -80oC. Unlabeled oligonucleotides were visualized with a solution of Stains-Alltm (0.25% in 50% aqueous formamide).
Base composition analysis of the oligonucleotides was performed by incubation of 1.0 A260 oligonucleotide with 4 U snake venom phosphodiesterase and 4 U shrimp alkaline phosphatase in 10 mM Tris-HCl buffer, pH 8.2, containing 2 mM MgCl2 at 37oC for 18 h. The mixture was heated to 90oC for 2 min, allowed to cool to room temperature and then injected onto a reversed phase HPLC column. Elution with buffer for 6 min followed by a gradient of 1-20% acetonitrile in 10 mM triethyl ammonium acetate buffer, pH 6.2, over 20 min at a flow rate of 1 ml/min afforded the cleaved products. Ratio and retention time of the eluted peaks was compared with the HPLC retention times of authentic nucleosides injected under identical conditions.
UV thermal denaturation studies of ODNs were performed on a Varian Cary 3 UV-Visible spectrophotometer equipped with a variable temperature controller assembly. Extinction coefficients of the oligonucleotides were determined by recording the absorbance of oligonucleotide samples (0.5 A260 units) before and after digestion with venom phosphodiesterase (16 ). For thermal denaturation studies, samples were dissolved in 10 mM sodium phosphate buffer, pH 7.0, containing 150 mM NaCl at a concentration of 3 [mu]mol each strand. The samples were heated to 90oC for 2 min and allowed to slowly equilibrate to room temperature. The samples were next heated from 0 to 90oC and back to 0oC at a rate of 1oC/min and the change in UV absorbance recorded at 260 nm.
A 32P-labeled RNA transcript was prepared from pBS 76-1 (17 ) containing the cDNA for 5[alpha]-reductase type II (5[alpha]R-II). For synthesis the plasmid was linearized by restriction digestion with BamHI. Transcription was carried out using T7 RNA polymerase according to the manufacturer's protocols (Life Technologies, Gaithersburg, MD). The RNA transcript synthesized from the cut vector is 258 nt long (Fig. 1 ). RNA (at 100 000 c.p.m.) was incubated in 20 mM Tris-HCl, pH 7.5, 100 mM KCl, 10 mM MgCl2, 1.4 mM dithiothreitol, 1 [mu]g/ml bovine serum albumin and 0.1 mM EDTA in the presence or absence of 32 mM oligonucleotide. Samples were preheated to 55oC for 5 min, cooled slowly to 37oC and 3 U E.coli RNase H were subsequently added. Following incubation at 37oC for 30 min, samples were mixed with 2* TBE (200 mM Tris-HCl, pH 8.5, 90 mM boric acid and 1 mM EDTA) containing 9 M urea and tracking dye. Samples were analyzed by electrophoresis on 5-20% polyacrylamide exponential gradient gels containing 7 M urea. Following electrophoresis, bands were visualized by autoradiography for the full-length and cleavage products of the RNA transcript.
Human prostate carcinoma cell (American Type Culture Collection no. HTB 81) lysates, prepared using NP-40 RIPA lysis buffer (150 mM NaCl, 50 mM Tris-HCl, pH 7.5, 1 mM EDTA, 2% NP-40, 0.5% deoxycholate and 100 [mu]g/ml aprotinin), were fractionated by differential centrifugation to yield crude cytosolic and nuclear fractions. Fraction aliquots, 20 [mu]g total protein, were incubated for 30 min at 37oC with 1 nM oligonucleotide, containing 100 000 c.p.m. [32P]PNK-labeled oligonucleotide. Proteins were then separated by native gel electrophoresis on a non-denaturing 5-20% exponential gradient polyacrylamide gel and visualized by autoradiography. For assessment of thrombin binding, 1 nM oligonucleotide, containing 100 000 c.p.m. labeled oligonucleotide, was incubated in NP-40 RIPA buffer with 1 [mu]g purified thrombin (Sigma, St Louis, MO) at 37oC for 30 min. Samples were analyzed by native PAGE on a 4% gel and then visualized by autoradiography. Silver staining showed that thrombin runs as a single band (data not shown).
Chinese hamster ovary cells transfected with the human 5[alpha]R-II gene (CHO 1829) were plated at 8 * 106 cells/well in 6-well dishes in DMEM/F12 (1:1) and 5% fetal bovine serum (19 ). Cells were dosed with ODNs (5 [mu]g/ml Lipofectintm; Life Technologies Inc., Gaithersburg, MD) in serum-free Opti-MEM for 3 h at 37oC and then returned to serum-containing medium. Dosing was carried out for for three consecutive days due to the long half-life of the 5[alpha]R protein. Cells were harvested by detachment in phosphate-buffered saline with 5 mM EDTA, pelleted and lysed in NP40-RIPA buffer. Cellular proteins (5 [mu]g) were then separated by SDS-PAGE on a 12% acrylamide gel before transfer to a PVDF membrane. Specific antibodies for 5[alpha]R-II (provided by Dr David W.Russell) and actin (anti-actin A2066; Sigma, St Louis, MO) were used for detection following Western blotting by ECL-Western (Amersham, Arlington Heights, IL). Actin was included as an internal control and all results are presented as the ratio of 5[alpha]R type-II relative to actin.
We have synthesized 2',5'-linked oligo-3'-deoxyribonucleotides and evaluated their chemical and hybridization properties. 3'-Deoxypyrimidine oligonucleotides were synthesized by the reported procedure (13 -15 ). Earlier, 3'-deoxyguanosine was synthesized by radical deoxygenation of a suitably protected guanosine derivative (13 ). We have adopted a convergent route starting with the known 3'-deoxyribose synthon (14 ) as the versatile synthon (Fig. 1 ). Thus, diaminopurine was acetylated with acetic anhydride and 1-methyl-2-pyrrolidinone at 150oC to give the bis-acetylated compound
2',5'-Linked oligodeoxynucleotides were synthesized on a controlled pore glass (500 Å) solid support loaded with protected ribonucleosides. After cleavage from this support and removal of the protecting groups with concentrated ammonium hydroxide, 2',5'-linked ODNs were furnished containing a terminal 3'-hydroxyl group at the 2'-end. The presence of ribose sugar at the 2'-end of the 2',5'-linked oligomer does not interfere in the thermal stability of its complex with RNA. After purification by HPLC these oligonucleotides were found to be homogeneous by HPLC and migrated as a single band under denaturing PAGE analysis. 2',5'-Linked oligodeoxynucleotides are substrates for polynucleotide kinase and can be phosphorylated at the 5'-end with [32P]ATP. Base composition analysis of the oligonucleotides confirmed the fidelity of the oligomers.
The affinity of 2',5'-linked oligodeoxynucleotides for DNA and RNA was determined by thermal denaturation studies. 2',5'-Linked oligodeoxynucleotides form stable duplexes with RNA but not with single-stranded DNA (Table 1 ). Comparison of the thermal denaturation of the complexes shows that, in sharp contrast to the hybridization of previously reported 2',5'-linked oligoribonucleotides (1 ), 2',5'-linked oligodeoxynucleotides form relatively more stable duplexes with RNA. Thus, earlier reports indicate that 2',5'-linked oligoribonucleotide ODNs show a decrease of -1.3oC/2',5' linkage in their thermal stability with RNA (1 ). With 2',5'-linked oligodeoxynucleotides containing a phosphodiester backbone, however, this decrease corresponds to -0.5oC/2',5' linkage (compare ODNs 1 and 2). A phosphorothioate backbone in 2',5'-linked oligodeoxynucleotides results in a drop of only -0.22oC/2',5' linkage relative to the 3',5'-linked phosphorothioate oligonucleotide (compare ODNs 3 and 5). Furthermore, 2',5'-linked oligodeoxynucleotides with phosphodiester and phosphorothioate backbone do not hybridize to complementary DNA (ODNs 1 and 5). Thus it is clear that phosphorothioate substitution in 2',5'-linked oligodeoxynucleotides is less destabilizing in comparison with 3',5'-linked oligodeoxynucleotides when hybridized to complementary RNA, as reflected by changes in Tm.
One possible mechanism by which an antisense oligodeoxynucleotide prevents translation of mRNA to protein is by activation of RNase H. Cellular RNase H cleaves the RNA portion of the RNA:DNA heteroduplex. The cleavage event irreversibly inactivates the RNA and thereby prevents translation of mRNA to protein (22 ). Potentially, each antisense oligonucleotide can inhibit multiple copies of the target RNA. This feature may be very important in amplifying antisense-mediated inhibition of protein synthesis.
That 2',5'-linked oligodeoxynucleotide ODNs do not support RNase H activity was determined by cleavage of a partial 5[alpha]R-II transcript prepared by T7 RNA polymerase transcription of plasmid pBS 76-1. The RNA was obtained as a 258 nt run-off chimeric transcript containing 31 nt of the polylinker and the first 227 nt from the 5'-end of the 5[alpha]R-II RNA (Fig. 2 A). Antisense oligonucleotide sequences (21mers) complementary to nt 40-61 of the chimeric RNA transcript containing 3',5'-phosphodiester linkages (ODN 1), 3',5'-phosphorothioate linkages (ODN 3), 2',5'-phosphorohthioate linkages (ODN 5) or chimeric 2',5'/3',5'/2',5'-phosphorothioate linkages (ODN 7) were tested for their ability to support E.coli RNase H activity. The data indicate that 3',5'-linked ODN 1 (lane 3) and ODN 3 (lane 4) activate RNase H, resulting in cleavage of the RNA transcript (Fig. 2 B). Incubation of the transcript with an RNase H-supporting oligonucleotide results in cleavage of the transcript to products that are ~40 and ~195 nt long. In contrast, the 2',5'-linked phosphorothioate oligonucleotide (ODN 5) does not activate RNase H (lane 6). However, incorporation of seven 3',5'-linked phosphorothioates in the center of ODN 5 generates a chimeric oligonucleotide (ODN 7) that can support RNase H activity (lane 5).
An additional band is evident in the three lanes where RNase H was activated. That this extra band is generated by RNase H in all three lanes suggests that there is an alternative binding site for the antisense oligonucleotide sequence and that this alternative binding site must be contained within the 3',5' portion of chimeric ODN 7. Nucleotide sequence analysis of the transcript suggests that this band may arise through oligonucleotide binding to an alternate 6 nt binding site located in the 5[alpha]R portion of the transcript. This putative 6 base binding site is contained within the 7 base 3',5'-phosphorothioate sequence located in chimeric ODN 7.
Table 1
Antisense activity of the 2',5'-linked oligodeoxynucleotide phosphorothioate ODNs was evaluated by studying the ability of these oligonucleotides to inhibit expression of 5[alpha]R in cultured cells. Steroid 5[alpha]R catalyzes reduction of a major circulating androgen, testosterone, to its 5[alpha]-reduced form, dihydroxytestosterone (24 ). This androgen is reportedly involved in regulation of male differentiation during fetal development, prostatic hyperplasia, male pattern baldness, acne and hirsuitism (25 ). To evaluate if antisense phosphorothioate oligonucleotides have the ability to suppress synthesis of steroid 5[alpha]R, an in vitro cell culture model was designed. Using Chinese hamster ovary cells transfected with the human 5[alpha]R type-II gene, a number of 3',5'-linked phosphorothioate oligonucleotides were screened. A 21mer phosphorothioate (ODN 3, Table 1 ) targeted against the 5'-untranslated region of 5[alpha]R mRNA was found to reduce 5[alpha]R expression in a dose-dependent manner (data not shown). Compared with vehicle controls (Fig. 4 , lane 1), ODN 3 (lane 2) reduced the levels of 5[alpha]R protein at 100 nM oligonucleotide concentration by 70%, as determined by ECL-Western analysis. ODN 3 demonstrated a potent dose-response effect with an IC50 of ~20 nM (data not shown). In contrast, a two base mismatch sequence (ODN 4, lane 3) with lower thermal stability against target RNA and a scrambled sequence (not shown) did not show antisense activity.
2',5'-Linked 3'-oligodeoxynucleotide ODNs demonstrate selective binding with target RNA but not DNA. However, once bound to target RNA, 2',5'-linked oligodeoxynucleoside ODNs do not activate RNase H degradation of the bound RNA. Furthermore, in contrast to the antisense 3',5'-linked phosphorothioate ODNs, isomeric 2',5'-linked ODNs fail to inhibit steroid 5[alpha]R-II expression in cell culture. The antisense activity of the 2',5'-linked oligomer was restored by incorporating a 7 base 3',5'-linked phosphorothioate segment (as a putative RNase H active site) in the chimeric ODN. That this site can support RNase H activity is suggested by the results with E.coli RNase H. The 2',5'-linked oligonucleoside phosphorothioate ODNs show less non-specific binding with cellular proteins and thrombin in comparison with the 3',5'-linked phosphorothioate ODNs. This non-sequence specific binding of phosphorothioate oligonucleotides has been associated with various undesirable side effects in vivo (28 ). A reduction in these non-sequence specific protein interactions with the 2',5'-linked phosphorothioate motif may produce fewer of these side effects in vivo. Thus, chimeric 2',5'-linked oligo-3'-deoxynucleoside ODNs hold promise as a new generation of antisense ODNs for the control of unwanted gene expression. Research geared to introduce additional modifications at the 3' position of the sugar and in the 2',5'-backbone itself is currently in progress.
Part of this study was funded by a SBIR grant from the National Institutes of Health (R44 GM49581-02). The authors would like to thank Dr David W.Russell (University of Texas, Southwestern) for providing the 5[alpha]R vector, cell lines and 5[alpha]R antibodies. Preliminary accounts of these findings were presented at the International Conference on Therapeutic Oligonucleotides, From Cell to Man, April 5-7, 1995, Seillac, France
ODN no.
Oligonucleotide
Tm (oC)a
DNA
RNA
Phosphodiester ODNs
1
5'-CATCGCGCCGTGTTCCTCGCC-3'
74.2
67.0
2
5'-CATCGCGCCGTGTTCCTCGCC-2'
57.0
Phosphorothioate ODNs
3
5'-CATCGCGCCGTGTTCCTCGCC-3'
68.5
60.5
4
5'-CAT
49.5
42.0
5
5'-CATCGCGCCGTGTTCCTCGCC-2'
56.0
6
5'-CAT
35.5
Chimeric 2'-5'/3'-5'/2'-5' ODNs
7
5'-CAT CGC Gcc gtg ttC CTC GCC-2'
58.0
8
5'-CAT
38.0
9
5'-CAT CGC Gcc gtg ttC CTC GCC-2'(PO)
60.0
REFERENCES