Nucleic Acids Research

INTRODUCTION

Selective control of gene function is currently becoming assessed using oligodeoxynucleotides (ODNs) targeted toward either DNA or RNA (1). The selectivity of such ODNs should be controlled by Watson-Crick base pairing interaction between the two complementary sequences. RNAs can be written as single strands, but because of self pairing, they show branched structures. A typical feature is a hairpin consisting of a double- stranded stem region bound to single-stranded regions at the foot of the stem. Francois et al. (2) found enhanced duplex stability due to coaxial stacking when single-stranded DNA was hybridized next to the stem to one of the sides of a DNA hairpin foot. Bruice and Lima (3) similarly reported about preferred binding of antisense DNA to Ha-ras mRNA stem-loop fragment using combinatorial screening. They found preferred 5[prime]-single strand binding to the 5[prime]-loop site allowing coaxial stacking between the hybridizing DNA and the stem. Another way of recognizing the structured nucleic acids could be hybridization to both sides of the foot of the stem by forming a three-way junction (TWJ). Leontis et al. (4) have studied the TWJ and their data showed that the TWJ could be stabilized when unpaired bulged bases were added to one of the strands at the branch point. Moreover, TWJ has been studied by NMR on complexes with two unpaired bases at the branch point and a preferred coaxial stacking interaction at the branch point was found (5,6). Cload and Schepartz (7) found that a TWJ-forming ODN was an inhibitor of Rev response element of human immunodeficiency virus genome. Even formation of a four-way junction has been reported by Cload et al. (8) as a possible way of recognizing RNA using tethered ODN.

To enhance the penetration of ODN into the cells, it could be an advantage to synthesize the ODN carrying hydrophobic (9,10) and lipophilic (11) moieties. Especially, it would be of interest to synthesize new ODNs with lipophilic conjugates which could improve both cell uptake and hybridization properties. The problem, however, is to find a proper way of increasing duplex stability using lipophilic moieties. Casale and Mclaughlin (12), studying the hybridization of double-stranded DNA, observed destabilization of the duplex when the modified ODN contained a polyaromatic hydrocarbon substituted at N² on guanosine. We found it more promising to improve the hybridization properties by conjugating a lipophilic intercalating moiety to a junction region of nucleic acids. From thermal spectroscopy of TWJ in DNA, improved stabilization was found for the DNA TWJ when 2[prime]-deoxy-5-methyl-N⁴-(1-pyrenylmethyl)cytidine was inserted into the junction region of a TWJ-forming ODN which was hybridized across the foot of a hairpin (13). In this investigation we synthesize 2[prime]-deoxy-5-methyl-N⁴-(4-phenoxyphenyl)cytidine for incorporation into the TWJ.

MATERIALS AND METHODS

NMR spectra were recorded at 250 MHz for ¹H-NMR, 62.9 MHz for ¹³C-NMR and 101.3 MHz for ³¹P-NMR on a Bruker AC-250FT spectrometer; [delta]-values are in p.p.m. relative to tetramethylsilane as internal standard (¹H- NMR and ¹³C-NMR), relative to 85% H₃PO₄ as external standard in ³¹P-NMR. Positive FAB mass spectra were recorded on a Kratos MS 50 RF spectrometer. Analytical silica gel TLC was performed on Merck precoated 60 F₂₅₄ plates. The silica gel (0.040-0.063 mm) used for column chromatography was purchased from Merck. Melting experiments were carried out on a Perkin-Elmer UV-VIS spectrometer Lamda 2 fitted with a PTP-6 Peltier temperature programming element. The absorbance of 260 nm was increased by 1°C/min in a 1 cm cuvette. DNA syntheses were performed on a Pharmacia Gene Assembler Special^® DNA synthesizer. Desalting of oligonucleotides was accomplished using disposable NAP-10 columns (Pharmacia).

RESULTS AND DISCUSSION

Synthesis of the building block

The protected nucleoside 3[prime]-O-acetyl-2[prime]-deoxy-5[prime]-O-(dimethoxytrityl)thymidine (1) (13) was reacted overnight at room temperature with 2,4,6-trimethylbenzenesulfonyl chloride in the presence of triethylamine and 4-dimethylaminopyridine (DMAP) in dichloromethane to give the O⁴-thymidine derivative 2 in 86% yield after purification by column chromatography (Scheme 1). The N⁴-cytidine derivative 3 was synthesized in 73% yield by refluxing 4-phenoxyaniline with the O⁴-thymidine derivative 2 in tetrahydrofuran in the presence of N,N-diisopropylethylamine for 36 h followed by chromatographic column purification. The alternative triazole activation (13) by reaction of 1 with a triazole phosphorus oxychloride reagent was found insufficient in our hands for the reaction with an aromatic amine. Deacetylation of compound 3 was carried out with sodium methoxide in methanol for 1 h to give the free 3[prime]-OH nucleoside 4 in 100% yield. Subsequent deprotection using the detritylation solution Cl₃CCOOH/C₂H₄Cl₂ at room temperature for 5 min gave the deprotected N⁴-substituted cytidine derivative 5 in 92% yield. Phosphitylation of 4 to give 6 was achieved in 83% yield by reaction of 2-cyanoethyl-N,N-diisopropylchlorophosphoramidite with the nucleoside 4 in the presence of N,N-diisopropylethylamine and dichloromethane for 2 h. The purity of the phosphoramidite 6 was 100% according to ³¹P-NMR spectroscopy.

Scheme 1.

Figure 1. ODNs used for hybridization studies. X is the diphenyl ether conjugated nucleoside 5.

According to the phosphoramidite methodology (17) both modified and unmodified ODNs were synthesized on a Pharmacia Gene Assembler Special DNA synthesizer in a 0.2 µmol scale. The coupling efficiencies for the modified phosphoramidite 6 (6 min couplings) and commercial ones (2 min couplings) were 99%. The efficiency of each coupling step was monitored by release of dimethoxytrityl cation after each step. Removal from the solid support and deprotection was carried out at room temperature for 4 days in 25% ammonia. All ODNs were desalted using the Pharmacia NAP-10 columns. The concentration of the ODNs strands was quantitated by UV absorption at 260 nm using the molar extinction coefficients for the modified nucleoside 5 and for the unmodified ones (18). The ratio of the molar extinction coefficient for 2-deoxycytidine:5 was 8.7:1 in methanol at 260 nm.

DNA TWJs and DNA duplexes

The hybridization properties of the modified ODNs (Fig. 1) were studied for formation of DNA TWJs (Table 1) and DNA duplexes (Tables 2 and 2). The TWJs and duplexes were performed in medium salt buffer, 1 mM EDTA, 10 mM Na₂HPO₄ and 140 mM NaCl at pH 7.0, using equimolar amounts (3 µM) of each strand. Before each experiment, all samples were heated at 90°C in a water bath for 5 min and then cooled slowly to 0°C. The increase in the UV absorbance at 260 nm as a function of time was recorded while the temperature was raised gradually (1°C/min) from 15-70°C.

For our investigations of the effect of a conjugated intercalator at the TWJ branch point, we selected the target A (Fig. 1) which by spectroscopic measurements and chemical reactions (2) has been shown to form a TWJ when hybridized with a 17mer at the flanks of the hairpin. The target A was hybridized with the targeting C-17mer and X-17mer (Fig. 1) affording a TWJ bulged with C or the conjugated nucleoside 5 at the branch point. When introducing an intercalator at the branch point there is a risk that the TWJ is not formed and that the melting properties are merely due to duplex formation to the side with the highest ratio of C:G base pairs. Therefore, target B was selected (half of target A) assuming that the effect of the dangling end would nearly be the same as for the dangling end of target A if only hybridized at the upstream flank without forming the hairpin and the TWJ.

From Table 1, it can be seen that the TWJ (Duplex type I) with the modified ODN (X-17mer) is more stabilized than the natural ones. We observed only a small difference in melting point ([Delta]T_m = 2.8°C) when a bulged C was inserted at the branch point while a large increase in melting temperature ([Delta]T_m = 9.0°C) was observed when the intercalating phenoxyphenyl conjugated cytidine was inserted. Also, we observed a positional effect of the insertion on the stability of the TWJ when the conjugated nucleoside 5 was inserted around the branch point. We observed T_m = 29.8°C ([Delta]T_m = 2.6°C) when X was bulged next to the branch point in the 3[prime]-end direction of the hairpin, and T_m = 31.8°C ([Delta]T_m = 4.6°C) when X was bulged next to the branch point in the other direction. This positional effect has previously been observed on insertion of 7 (13) and may be related to coaxial stacking since a determining factor for this could be sequence of the base pairs immediately flanking the junction region (19). Entry 4 in Table 1 confirms an expected lower stability of a duplex corresponding to one of the arms of the TWJ and this excludes Entry 3 to be of Duplex type II with two long dangling ends.

Table 1. Melting temperatures T_m obtained by hybridization of 17mer, C-17mer or X-17mer with the targets A or B

In Table 2, the thermal melting data are shown for a duplex corresponding to Duplex type I except that the loop stem parthas been deleted from target A. The data show that the bulged Cat the former branch point (Entry 6) destabilize the duplex([Delta]T_m = -9.2°C) and this is in good agreement with the previously reported results (20,21), while the insertion of intercalating 5 (Entry 7) gave less destabilization of the duplex ([Delta]T_m = -6.6°C).

Table 2. Hybridization data (T_m/°C) for duplexes formed with 3[prime]-TGACATA₆GA₂GAGA₃G₂T-5[prime] (deletion of hairpin in A). X is the conjugated diphenyl ether nucleoside 5

It has been reported (22,23) that a TWJ, without bulged nucleotides in the junction region, has a trigonal arrangement of the three intact helical arms. In this case, base pair stacking across the junction is not possible. There are six base pair helix ends in this type of TWJ compared with two helical ends in a duplex with the same number of base pairs, and this may explain the lower stability of the TWJ when compared with the corresponding duplex. It has been shown that unpaired bases at the point of the branch changes the trigonal arrangement into an asymmetrical structure which is consistent with a stacking between two of the arms at the branch point, and an inclined orientation of the third (non-stacked) arm to the other two (stacked) arms (23,24).

One of the complementary strands of this third helical arm contains the strand that has the extra bulged nucleotides. For the latter TWJ there are still additional two helical ends compared with a non-branched two-stranded duplex structure. At first approximation the additional two helical ends in the branched structure would correspond to a decrease in the stacking interactions of a consecutive base pair in the middle of a helical duplex (24). For the TWJ (Type I, Entry 3) with diphenyl ether conjugated to an extra nucleotide, it is now assumed that the extra duplex stability is achieved by placing the large non-polar diphenyl ether as a stacking moiety covering the end of the inclined helical arm at the branch point. The stacking affinity of the diphenyl ether to the inclined helical arm is most likely enhanced because of the diphenyl ether hydrophobicity. The group of Kool (25) has investigated the aromatic stacking affinities of dangling C-nucleosides at the 5[prime]-ends of a self-complementary duplex. When thymine was replaced by pyrene as the base of the dangling nucleotide, the T_m of the duplex was raised by 16°C due to the better stacking ability of the pyrenes which were assumed to cover the ends of the duplex.

Table 3. Melting temperatures T_m obtained by hybridization of 9mer, C-9mer or X-9mer with the targets A, B, C or D

Another way to target a stem flank region is to use a short ODN (2) which hybridizes to only one of the flanks, in such a way that coaxial stacking to the loop is possible (see Duplex type III in Table 3). Again, we used the target A which for hybridization with 9mer, C-9mer, and X-9mer (Fig. 1) was compared with the target B (Duplex type VI) which can not contribute to the duplex stability by coaxial stacking. Also target C and target D were included in this investigation in order to see the effect of the long dangling downstream flank of target A. Using the C-9mer in Duplex types III-V, it always resulted in lower melting temperatures than by using the corresponding 9mer. The opposite effect was observed with the X-9mer which is always showing the higher T_m ([Delta]T_m = 2.6-5.8°C), most likely due to the intercalating effect of the phenoxyphenyl conjugate at the branch point. For the C-9mer, the extra C is causing sterical interactions which lower the thermal stability of the duplexes. Similar interactions from the long dangling end, are also most likely to be the cause of the lower stabilities of the Duplex type III when compared with Duplex type V. For Duplex type VI a considerable stabilization was also observed with the X-9mer indicating in this case also the stabilizing effect of the conjugated intercalator. However, the T_m was lower than those found for the Duplex types III-V which have the additional effect of coaxial stacking.

RNase H studies

From the above studies on the ODN duplexes one could think it also possible to stabilize the junction regions in DNA/RNA TWJs. From the literature it is known that antisense ODNs linked to an intercalating acridine derivative can improve translational inhibition of mRNAs, most likely by an RNase H dependent mechanism (26,27). We therefore found it is of interest to investigate how insertion of the nucleoside 5 or 7 (Scheme 2) into an ODN could influence the RNase H cleavage of an RNA at a junction region. The N⁴-pyren-1-ylmethylcytidine analog 7 (13) has previously been shown to stabilize an ODN TWJ and was of interest to be included in this study because it is always possible that different chemical modifications of the targeting ODNs can show variations in RNase H hydrolyses of the DNA-RNA hybrids (28).

Scheme 2.

For our antisense studies we selected hok mRNA from plasmid R1 which has a well-defined secondary structure (Fig. 2) (15,16). The nucleotides C-74 to C-104 constitute a single stem-loop domain flanked with single strands which on hybridization with an antisense ODN could possibly form a TWJ. The natural mRNA does not show any base pairing at A-103 and C-104. From the hybridization studies of Type III ODN duplexes (Table 3) it is evident that there is an improved hybridization when using short ODNs at junctions. Since short antisense ODNs have a distinct advantage of simplified manufacturing (29), we studied the effect of insertion of 5 and 7 on binding of Type III duplexes between an antisense ODN and hok mRNA. We selected the 7mer ODN complementary to the mRNA sequence G-76 to A-82, which would allow coaxial stacking to C-98. Indeed, the natural ODN (Entry 20, Fig. 3) resulted in a strong RNase H activity. The RNase H activity was maintained, although weaker (Entry 21), when the conjugated pyrene nucleoside 7 was inserted between T-5 and A-6 of the complementary ODN. When a non-conjugated C was inserted at the same position no RNase H cleavage was observed (Entry 22). The RNase H cleavage also depends on the conjugated aromatic system, since RNase H activity was absent when the diphenyl ether nucleoside 5 was inserted at the same position (data not shown).

Extension of the ODN from Entry 20 at the 5[prime] end with the conjugated pyrene nucleoside 7 resulted in an increased RNase H cleavage (Entry 25). However, it is not possible to tell whether this effect is due to intercalating effect of the pyrene ring or due to simple base pairing by replacing C-98 in the G-83:C-98 base pair with the modified nucleoside 7. In fact, an improved RNase H cleavage was also observed when the ODN from Entry 20 was extended with C at the 5[prime] end (Entry 23). Further extension of the latter ODN with 7 at the 5[prime] end gave even further improved RNase H cleavage (Entry 24). However, the intercalating effect should be considered important, since RNase H cleavage was completely absent when 5 was used instead of 7 (data not shown). Another interesting feature of the extended ODNs 23 and 24 was an increasing downstream hydrolysis of the hok mRNA on the 3[prime]-side of the stem-loop domain. This effect could be due to cleavage on the other side of the stem-loop region at the single-stranded RNA adjacent to the RNA-DNA duplex as reported by Lima and Crooke (30).

Figure 2. hok mRNA from plasmid R1.

Figure 3. RNase H cleavage of the 361 nucleotide hok mRNA from plasmid R1. (A) Effect of 7 or C when inserted in the middle of a short coaxially stacked ODN. (B) Using 7 or C for extending a short coaxially stacked ODN at the junction. (C) Insertion of 5 or C at the branch point of a TWJ.

The Type I TWJ using hok mRNA was considered possible by using the complementary ODN to the sequence G-76 to A-82 tethered via 5, 7 or cytidine to the complementary ODN to the sequence C-99 to A-103. The effect of one or two deletions at the 5[prime] end of the latter ODN sequence was also investigated. Except for one case, all these tethered ODNs did not mediate any significant RNase H cleavage despite all having the same sequences as Entry 20, 23 or 25 at the upstream part of the hok mRNA. It is believed that the extra dangling tethered ODN due to steric interactions prevents a sufficiently high stability of the duplex, and is also influencing the RNase H cleavage. Only in one case, by using the conjugated diphenyl ether nucleoside 5 for tethering and with one deletion at the 5[prime] end, was significant RNase H activity observed (Entry 27). In this case, not only was significant RNase H activity observed, but the cleavage was also taking place on the upstream side of the hairpin of hok mRNA when compared with Entry 20. One can speculate why exactly one deletion at the 5[prime] end of the ODN is a prerequisite for RNase H cleavage. However, with this length of the ODN, coaxial stacking from the 5[prime] end of the targeting ODN to U-329 is possible and may induce the base pairings A-103:U-329 and C-104:G-328. In this way, a sufficiently long DNA-RNA duplex is established to accomplish RNase H hydrolysis at this side of the hairpin.

In this work we have shown that conjugated intercalators at a DNA-DNA TWJ improve its stability as measured by its thermal denaturation. It is also clear that short ODNs can increase RNase H cleavage at the flanks of an RNA hairpin. For the conjugated ODNs the results are more ambiguous whether the RNase cleavage can be enhanced. Although a lot of work is needed in order to find the best type of conjugate and linker which again may depend on the specific target, we still think it possible to develop short and lipophilic conjugated ODNs for targeting junctions in RNA.

REFERENCES

1. Ecker,D.J. (1993) In Crooke,S.T. and Lebleu,B. (eds), Antisense Research and Applications. CRC Press Inc., Boca Raton, FL, pp. 387-400.

2. Francois,J.-C., Thuong,N.T. and Héléne,C. (1994) Nucleic Acids Res., 22, 3943-3950. MEDLINE Abstract

3. Bruice,T.W. and Lima,W.F. (1997) Biochemistry, 36, 5004-5019. MEDLINE Abstract

4. Leontis,J.B., Kwok,W. and Newman,J.S. (1991) Nucleic Acids Res., 19, 759-766. MEDLINE Abstract

5. Rosen,M.A. and Patel,D.J. (1993) Biochemistry, 32, 6563-6575. MEDLINE Abstract

6. Rosen,M.A. and Patel,D.J. (1993) Biochemistry, 32, 6576-6587. MEDLINE Abstract

7. Cload,S.T. and Schepartz,A. (1994) J. Am. Chem. Soc., 116, 437-442.

8. Cload,S.T., Richardson,P.L., Huang,Y.-H. and Schepartz,A. (1993)J. Am. Chem. Soc., 115, 5005-5014.

9. Kabanov,A.V., Vinogradov,S.V., Ovcharenko,A.V., Krivonos,A.V., Melik-Nubarav,N.S., Kiselev,V.I. and Severin,E.S. (1990) FEBS Lett., 259, 327-330. MEDLINE Abstract

10. Gryaznov,S.M. and Lioyd,D.H. (1993) Nucleic Acids Res., 21, 5909-5915. MEDLINE Abstract

11. Manoharan,M., Johnson,L.K., Tivel,K.L., Springer,R.H. and Dan Cook,P. (1993) Bioorganic and Medicinal Chem. Letts., 13, 2765-2770.

12. Casale,R. and Mclaughlin,L.W. (1990) J. Am. Chem. Soc., 112, 5264-5271.

13. Abdel-Rahman,A.A.-H., Ali,O.M. and Pedersen,E.B. (1996) Tetrahedron, 52, 15311-15324.

14. Franch,T. and Gerdes,K. (1996) Mol. Microbiol., 21, 1049-1060. MEDLINE Abstract

15. Gultyaev,A.P., Franch,T. and Gerdes,K. (1997) J. Mol. Biol., 273, 26-37. MEDLINE Abstract

16. Franch,T., Gultyaev,A.P. and Gerdes,K. (1997) J. Mol. Biol., 273, 38-45. MEDLINE Abstract

17. Cauthers,M.H. (1991) Acc. Chem. Res., 24, 278-284.

18. Brown,T. and Brown,D.J. (1991) In Eckstein,F. (ed.), Oligonucleotides and Analogous. A Practical Approach. Oxford University Press, Oxford, p. 20.

19. Leontis,N.B., Hills,M.T., Piotto,M., Ouporov,I.V., Malhotra,A. and Gorenstein,D.G. (1995) J. Biophys., 68, 251-265.

20. Lonfellow,C.E., Kierzk,R. and Turner,D.H. (1990) Biochemistry, 29, 278-285.

21. LeBlanc,D.A. and Morden,K.M. (1991) Biochemistry, 30, 4042-4047. MEDLINE Abstract

22. Welch,J.B., Duckett,D.R. and Lilley,D.M.J. (1993) Nucleic Acids Res., 22, 4548-4555. MEDLINE Abstract

23. Stühmeier,F., Welch,J.B., Murchie,A.I.H., Lilley,D.M.J. and Clegg,R.M. (1997) Biochemistry, 36, 13530-13538. MEDLINE Abstract

24. Stühmeier,F., Lilley,D.M.J. and Clegg,R.M. (1997) . Biochemistry, 36, 13539-13551. MEDLINE Abstract

25. Guckian,K.M., Schweitzer,B.A., Ren,R.X.-F., Sheils,C.J., Paris,P.L., Tahmassebi,D.C. and Kool,E.T. (1996) J. Am. Chem. Soc., 118, 8182-8183.

26. Saison-Behmoaras,T., Tocqué,B., Rey,I., Chassignol,M., Thuong,N.T. and Héléne,C. (1991) EMBO J., 10, 1111-1112. MEDLINE Abstract

27. Cazenave,K., Loreau,N., Thuong,N.T., Toulmé,J.-J. and Héléne,C. (1987) Nucleic Acids Res., 15, 4717-4736.

28. Donis-Keller,H. (1979) Nucleic Acids Res., 7, 179-192. MEDLINE Abstract

29. Wagner,R.W., Matteucci,M.D., Grant,D., Huang,T. and Froehler,B.C. (1996) Nature Biotechnol., 14, 840-844.

30. Lima,W.F. and Crooke,S.T. (1997) J. Biol. Chem., 272, 27513-27516. MEDLINE Abstract

---

*To whom correspondence should be addressed. Tel: +45 6 557 2555; Fax: +45 6 615 8780; Email: ebp@chem.ou.dk

---

This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 17 Oct 1998
Copyright©Oxford University Press, 1998.
CiteULike    Connotea    Del.icio.us    What's this?

This Article Abstract Print PDF (112K) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (6) Request Permissions Commercial Re-use Guidelines for Open Access NAR Content Google Scholar Articles by Ali, O. M. Articles by Pedersen, E. B. Search for Related Content PubMed PubMed Citation Articles by Ali, O. M. Articles by Pedersen, E. B. Social Bookmarking          What's this?