5
5492-5496 Nucleic Acids Research, 1994, Vol. 22, No. 24 ©1994 Oxford University Press Antisense pro-drugs: 5'-ester oligodeoxynucleotides Nikolai N.Polushin and Jack S.Cohen* Cancer Pharmacology Section, Georgetown University Medical Center, Washington DC 20007, USA Received April 11, 1994; Revised and Accepted June 22, 1994 ABSTRACT Oligonucleotides bearing a terminal llpophlllc group attached through a biodegradable ester bond should be useful as antisense pro-drugs with improved cellular uptake. The synthesis of 5-ester oligonucleotides is, however, problematic due to lability of the ester bond during aqueous ammonia treatment that Is commonly used for the deprotectlon of synthetic oligonucleotides. The synthesis of 5'-palmltoyl oligodeoxynucleotides was accomplished In good yield by the use of a combination of base-labile tert-butylphenoxyacetyl amino protecting groups (t-BPA), the oxalyl-CPG anchor group, and ethanolamine (EA) as a deprotectlng reagent. INTRODUCTION The inhibition of gene expression utilizing synthetic oligonucle- otides and their analogs has attracted great interest in recent years [ 1 ]. The promise of antisense therapeutics have been clearly demonstrated [2,3] and several clinical trials are underway [4,5]. There is, however, a major limitation in the application of oligonucleotides as potential universal drugs, that is the low ability of these polyanionic molecules to traverse cell membranes [6]. Because of this limitation hydrophobic nonionic backbone- modified analogs, such as methylphosphonates, were developed, in order to penetrate membrane bilayers due to their lipophilicity [7]. Nevertheless, the problem of cellular uptake does exist in the case of such polyanionic analogs as phosphorothioates and phosphorodithioates, as well as their co-polymers. It has been shown that cellular uptake of native oligonucleotides and polyanionic analogs may be improved by the attachment to these molecules of terminal lipophilic groups, such as cholesterol, that are known to interact specifically with cell membranes [8—16]. But this approach may have a substantial drawback, in that the lipophilic groups of these conjugated oligomers may remain attached to cell membranes, and this anchoring may make the antisense molecules unavailable for binding to their mRNA targets. To address this potential problem MacKellar a al. [8] proposed the attachment of a lipophilic moiety through a tetrathymidine bridge with normal phosphodiester bonds which are degradable by nucleases in cells and serum. Similarly, Oberhauser and Wagner used a biodegradable disulfide bond to attach a cholesteryl moiety on 2'-O-methyl-oligoribonucleotides [14]. Another attractive possibility to realization of the so-called 'pro- drug' approach for antisense olgonucleotide analogs would be to attach the lipophilic moiety through an ester bond, that is known to be easily hydrolyzed by intracellular esterases [17], This is not, however, a trivial synthetic task, as the ester bond is very sensitive to ammonia treatment that is usually used for oligonucleotide deprotection. An alternative is to derivatize an unprotected oligonucleotide with the terminal phosphomonoester group using the carbodiimide method [18]. But this route is definitely impractical from the synthetic viewpoint, since it requires the synthesis of an appropriate precursor containing an ester bond and loses the advantage of direct solid-phase derivatization. We here report a simple approach for the synthesis of oligodeoxynucleotides with a terminal lipophilic (palmitoyl) group attached through a biodegradable ester bond. This novel method employs ethanolamine for the selective deprotection of 5'-ester oligonucleotides with labile amino protecting groups. RESULTS AND DISCUSSION Acylation of 5-OH It is not useful to attempt 5'-esterification with a strong reagent such as acyl chloride, even in the case of protected oligonucleotides, due to possible base modifications. The use of acid anhydride is more appropriate, since 5'-acetylation with acetic anhydride is routinely used during oligonucleotide synthesis to cap unreacted 5'-OH groups. We found, however, that substitution of chlorine with tetrazole gives another useful alternative. This was first checked by palmitoylation of 3'-O- acetyl-thymidine. The acylating solution was prepared from palmitoyl chloride and tetrazole in the presence of N,N- diisopropylethylamine and used in situ. At 5-fold excess of palmitoyl tetrazole with one equivalent of N,N-diisopropyl- ethylamine, which is needed to neutralize released tetrazole, esterification of 3'-O-acetylthymidine is completed in 10 min at room temperature. No side products were observed by TLC analysis. The isolated yield of 5'-O-palmitoyl-3'-O- acetylthymidine was more than 90%. Thus, we believe that acyl tetrazole is a good choice for mild and rapid 5'-esterification. *To whom correspondence should be addressed

Antisense pro-drugs: 5'-ester oligodeoxynucleotides · 2017-04-16 · Antisense pro-drugs: 5'-ester oligodeoxynucleotides Nikolai N.Polushin and Jack S.Cohen* Cancer Pharmacology

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Antisense pro-drugs: 5'-ester oligodeoxynucleotides · 2017-04-16 · Antisense pro-drugs: 5'-ester oligodeoxynucleotides Nikolai N.Polushin and Jack S.Cohen* Cancer Pharmacology

5492-5496 Nucleic Acids Research, 1994, Vol. 22, No. 24 ©1994 Oxford University Press

Antisense pro-drugs: 5'-ester oligodeoxynucleotides

Nikolai N.Polushin and Jack S.Cohen*Cancer Pharmacology Section, Georgetown University Medical Center, Washington DC 20007, USA

Received April 11, 1994; Revised and Accepted June 22, 1994

ABSTRACT

Oligonucleotides bearing a terminal llpophlllc groupattached through a biodegradable ester bond shouldbe useful as antisense pro-drugs with improved cellularuptake. The synthesis of 5-ester oligonucleotides is,however, problematic due to lability of the ester bondduring aqueous ammonia treatment that Is commonlyused for the deprotectlon of synthetic oligonucleotides.The synthesis of 5'-palmltoyl oligodeoxynucleotideswas accomplished In good yield by the use of acombination of base-labile tert-butylphenoxyacetylamino protecting groups (t-BPA), the oxalyl-CPGanchor group, and ethanolamine (EA) as a deprotectlngreagent.

INTRODUCTION

The inhibition of gene expression utilizing synthetic oligonucle-otides and their analogs has attracted great interest in recent years[ 1 ]. The promise of antisense therapeutics have been clearlydemonstrated [2,3] and several clinical trials are underway [4,5].

There is, however, a major limitation in the application ofoligonucleotides as potential universal drugs, that is the low abilityof these polyanionic molecules to traverse cell membranes [6].Because of this limitation hydrophobic nonionic backbone-modified analogs, such as methylphosphonates, were developed,in order to penetrate membrane bilayers due to their lipophilicity[7]. Nevertheless, the problem of cellular uptake does exist inthe case of such polyanionic analogs as phosphorothioates andphosphorodithioates, as well as their co-polymers.

It has been shown that cellular uptake of native oligonucleotidesand polyanionic analogs may be improved by the attachment tothese molecules of terminal lipophilic groups, such as cholesterol,that are known to interact specifically with cell membranes[8—16]. But this approach may have a substantial drawback, inthat the lipophilic groups of these conjugated oligomers mayremain attached to cell membranes, and this anchoring may makethe antisense molecules unavailable for binding to their mRNAtargets.

To address this potential problem MacKellar a al. [8] proposedthe attachment of a lipophilic moiety through a tetrathymidinebridge with normal phosphodiester bonds which are degradableby nucleases in cells and serum. Similarly, Oberhauser and

Wagner used a biodegradable disulfide bond to attach acholesteryl moiety on 2'-O-methyl-oligoribonucleotides [14].

Another attractive possibility to realization of the so-called 'pro-drug' approach for antisense olgonucleotide analogs would beto attach the lipophilic moiety through an ester bond, that isknown to be easily hydrolyzed by intracellular esterases [17],This is not, however, a trivial synthetic task, as the ester bondis very sensitive to ammonia treatment that is usually used foroligonucleotide deprotection. An alternative is to derivatize anunprotected oligonucleotide with the terminal phosphomonoestergroup using the carbodiimide method [18]. But this route isdefinitely impractical from the synthetic viewpoint, since itrequires the synthesis of an appropriate precursor containing anester bond and loses the advantage of direct solid-phasederivatization. We here report a simple approach for the synthesisof oligodeoxynucleotides with a terminal lipophilic (palmitoyl)group attached through a biodegradable ester bond. This novelmethod employs ethanolamine for the selective deprotection of5'-ester oligonucleotides with labile amino protecting groups.

RESULTS AND DISCUSSIONAcylation of 5-OHIt is not useful to attempt 5'-esterification with a strong reagentsuch as acyl chloride, even in the case of protectedoligonucleotides, due to possible base modifications. The use ofacid anhydride is more appropriate, since 5'-acetylation withacetic anhydride is routinely used during oligonucleotide synthesisto cap unreacted 5'-OH groups. We found, however, thatsubstitution of chlorine with tetrazole gives another usefulalternative. This was first checked by palmitoylation of 3'-O-acetyl-thymidine. The acylating solution was prepared frompalmitoyl chloride and tetrazole in the presence of N,N-diisopropylethylamine and used in situ. At 5-fold excess ofpalmitoyl tetrazole with one equivalent of N,N-diisopropyl-ethylamine, which is needed to neutralize released tetrazole,esterification of 3'-O-acetylthymidine is completed in 10 min atroom temperature. No side products were observed by TLCanalysis. The isolated yield of 5'-O-palmitoyl-3'-O-acetylthymidine was more than 90%. Thus, we believe that acyltetrazole is a good choice for mild and rapid 5'-esterification.

*To whom correspondence should be addressed

Page 2: Antisense pro-drugs: 5'-ester oligodeoxynucleotides · 2017-04-16 · Antisense pro-drugs: 5'-ester oligodeoxynucleotides Nikolai N.Polushin and Jack S.Cohen* Cancer Pharmacology

Nucleic Acids Research, 1994, Vol. 22, No. 24 5493

1 2 3 4 5 6

XC

BPB

Figure 1. PAGE analysis of crude 5'-palm-<Tp)9T-3' prepared on oxalyl-CPGand deprotected with (2) NH3(lq) for 6 min; (3) NH3(lq) for 2 h; (4) EA for 10min; (5) the mixture N2H4/EA/MeOH (1:1:5, v/v/v) for 3 min. (1) is anauthentic S'-fTp^T-V; (6) represents HPLC-purified

Synthesis of 5'-ester oligonucleotidesAs a model compound, decathymidine (5'-(Tp)9T-3') wassynthesized by the /3-cyanoethyl phosphoramidite method (4 X1/tmol) on oxalyl-CPG (19) and, after removing the terminal DMTgroup, was 5'-esterized with palmitoyl tetrazole. But, in this case,apart from a great excess of acylating reagent, the reaction timewas increased to 30 min to ensure complete derivatization.Efficiency of derivatization was checked by PAGE. A smallportion of protected oligomer was treated with aqueous ammoniafor 6 min at room temperature. This short treatment is enoughto cleave the oxalyl anchor [ 19] and to almost completely remove/3-cyanoethyl phosphate protecting groups, but is not enough toextensively hydrolyze the ester bond. As can be seen from Figure1, lane 2, three distinct bands were observed on theelectrophoresis gel. As we expected the main product migratesslower than authentic 5'-<Tp)9T-3' (Fig. 1, compare lanes 1 and2) that indicates the presence of the palmitoyl group. The upperminor band probably corresponds to the palmitoylated 10-merwith one unremoved /3-cyanoethyl group that reduces the overallnegative charge by one, accordingly decreasing mobility. Insupport of this assumption the upper band completely disappearedafter prolonged ammonia treatment. The relatively small amountof underivatized 5'-(Tp)9T-3' (lower minor band) indicates thehigh efficiency of derivatization.

6 8 *

3496

Figure 2. HPLC analysis of crude 5'-palm-(Tp)9T-3' prepared on oxalyl-CPG and deprotected with (2) NH3<Kl) for 2 h; (3) EA for 10 min; (4) the mixtureN2H4/EA/Me0H (1:1:5, v/v/v) for 3 min. (1) is an authentic 5'-{Tp)<,T-3'; (5) represents HPLC-purified 5'-palm-<Tp)9T-3'.

Page 3: Antisense pro-drugs: 5'-ester oligodeoxynucleotides · 2017-04-16 · Antisense pro-drugs: 5'-ester oligodeoxynucleotides Nikolai N.Polushin and Jack S.Cohen* Cancer Pharmacology

5494 Nucleic Acids Research, 1994, Vol. 22, No. 24

Due to the great susceptibility of the ester bond toward base,incorporation of such a bond into a synthetic oligonucleotide withthe classical amino protecting groups (benzoyl for A and C, andisobutyryl for G) is impossible, as they are routinely removedin strongly basic conditions (aqueous ammonia, overnightheating). However, protection of nucleoside heterocycles maybe successfully accomplished by utilizing so-called base-labileamino protecting groups, such as phenoxyacetyl [20], isopropoxy-acetyl (IPA) [21] and tert-butyl-phenoxyacetyl (t-BPA) [22]. Forexample, aqueous ammonia completely removes the t-BPA groupfrom the least reactive guanine amino function in only 2 h at roomtemperature [22]. But these conditions are still too harsh for theester bond. Therefore, to incorporate an ester bond in anoligonucleotide conjugates, another protecting or deprotectingstrategy is necessary.

Recently we proposed two new techniques for acceleratedoligo-nucleotide deprotection, based on the use of ethanolamine(EA) and a mixture of EA, hydrazine and methanol [23-26].EA completely removes IPA and t-BPA amino protecting groupsin 10 min at room temperature [25,26]. On the other hand, ittakes almost 3 h to completely cleave the standard succinic esteranchor with this reagent (data not shown). This observation,encouraged us to investigate the use of EA deprotection forpreparing 5'-esterified oligonucleotides.

We used a base-sensitive oxalyl anchor that is completelycleaved within 2 min by both EA and the mixtureN2H4/EA/MeOH [25]. Initial experiments were performed with

decathymidine. For the synthesis of 5'-esterized oligomers,incorporating all nucleosides, the t-BPA amino protecting groupwas chosen (appropriate monomers are commercially availablefrom Millipore). Three small portions of CPG with immobilizedprotected 5 '-palmitoyl decathymidine prepared as described abovewere treated at room temperature with: 1) aqueous ammonia for2 h; 2) ethanolamine for 10 min; and 3) the mixtureN2H4/EA/Me0H (1:1:5, v/v/v) for 3 min. In every case theselected time is the minimum necessary for the full deprotectionof oligonucleotides synthesized with the use of t-BPA aminoprotected groups, /3-cyanoethyl phosphate protecting group, andan oxalyl anchor [25,26].

To prevent prolonged exposure of the ester bond towardsreactive EA and N2H4, they were immediately neutralized withacetic acid after deprotection. The reaction mixtures were desaltedon Sephadex G25 and analyzed by PAGE and reverse phaseHPLC (Figs. 1 and 2). As expected, the ester bond was almostcompletely hydrolyzed during ammonia treatment and, accordingto HPLC data, only 15% of the 5'-palmitoylated decathymidinewas obtained (Fig. 1, lane 3; Fig. 2, profile 2). In contrast, EAdeprotection gave the desirable conjugate with 68% yield (Fig. 1,lane 4; Fig. 2, profile 3). The mixture NzRj/EA/ MeOH wasmore selective than aqueous ammonia but considerably lessselective than EA (34% of derivatized oligomer, Fig. 1, lane 5and Fig. 2, profile 4). EA deprotection was further used for thelarge-scale preparation of 5'-palmitoyl decathymidine. Theconjugate was purified by preparative reverse phase HPLC with

(-CH2-)12

(Palm)

J J J J J I - i - l J - i

Figure 3. 'H-NMR spectrum of HPLC-purified 5'-palm-(Tp)9T-3'. The ratio between the integrals of thymidine l'-protons (m, 10H. 6 18-6.36 ppm) and CH3

protons of the palmitoyl residue (t, 3H, 0.71-0.81 ppm) equals 3.1 that is in good agreement with theoretical value 3.3. Signal A (s, 4.75-4.90 ppm) correspondsto HDO. Signals B (q, 3.00-3.30 ppm, (-CH2->j) and C (t, 1.15-1.40 ppm, (CH3-)3) corresponds to triethylammonium cation. Insert shows 31P-NMR spectrumof the same conjugate. The only signal at -0.07 ppm corresponds to phosphodiester groups.

Page 4: Antisense pro-drugs: 5'-ester oligodeoxynucleotides · 2017-04-16 · Antisense pro-drugs: 5'-ester oligodeoxynucleotides Nikolai N.Polushin and Jack S.Cohen* Cancer Pharmacology

Nucleic Acids Research, 1994, Vol. 22, No. 24 5495

50% isolated yield. The integrity of the product was checkedby PAGE and reverse phase HPLC (Fig. 1, lane 6; Fig. 2, profile5) and its structure was confirmed by 'H- and 31P-NMR(Fig. 3).

The above strategy was also successfully applied for thepreparation of 5'-palmitoylated 17-mer (Palm-dCACCAACTT-CTTCCACA) antisense to the coding region of the /3-globin gene[27] and some other 5'-derivatized oligonucleotides containingG. The isolated yields were 40-50%. It is important to pointout that phosphorothioate conjugates may also be prepared sinceit was shown that EA does not affect PS bonds [26].

CONCLUSION

We found that die combination of labile N-protecting groups andthe easily cleavable oxalyl anchor, along with application ofethanolamine as a deprotecting agent, allows efficientincorporation of the ester bond into oligonucleotide conjugates.This should have the advantage of a pro-drug, releasing the intactantisense oligomer as a result of intracellular esterase activityafter it permeates the cell membrane.

MATERIALS AND METHODS

The following reagents were purchased from commercial sources:t-BPA protected nucleoside N,N-diisopropyl-2-cyanoethylphos-phoroamidites (Millipore); palmitoyl chloride; lH-tetrazole; N,N-diisopropylethylamine; 3'-O-acetylthymidine; ethanolamine;hydrazine (Aldrich), acetonitrile; dichloromethane; chloroform;methanol (Fisher Scientific); long chain alkylamine CPG (Sigma);Kieselgel 60 F-254 TLC plates (Merck). N,N-Diisopropylethylamine was stored over sodium hydroxide, EAwas distilled before use.

Silica gel plates were developed in chloroform-methanol (9:1,v/v). Reverse-phase HPLC was done with a Waters 600E SystemController and a Waters 486 tunable absorbance detector usinga Hamilton PRP-1 column (10 tun, 7.0x305 mm) with flow rateof 2 ml per min for analysis of oligonucleotides, and a preparativeHamilton PRP-1 column (25x305 mm) widi a flow rate of 8ml per min for isolation of 5'-palmitoylated oligonucleotides. A10-60% gradient of acetonitrile (1.67% per min) in 0.1 Mtriediylammonium acetate (pH 7.5) was used as a mobile phase.

Electrophoretic gels were 20% acrylamide/7 M urea and wererun at 500-1000 V using TBE buffer. The gels were visualizedand photographed by UV shadowing over a fluorescent TLCplate. NMR chemical shifts are reported in ppm (6 units)downfield from TMS for 'H- and from H3PO4 for 31P-NMRspectra. Spectra were recorded on a Varian XL-300 spectrometerat 300 and 121 MHz respectively.

The derivatization of CPG through oxalyl anchor wasperformed as described [19]. Oligonucleotides were synthesizedby the phosphoramidite method on an ABI 38OB DNA/RNASynthesizer using ABI reagents. In the syntheses utilizing t-BPAnucleoside phosphoramidites, tert-butylphenoxy-acetic anhydride(Millipore) was used instead of acetic anhydride in the cappingreactions [22].

5'-O-Palmitoyl-3'-O-acetylthymidinePreparation of the acylating mixture. lH-Tetrazole (420 mg, 6.0mmol) was co-evaporated with acetonitrile to remove traces ofwater. The dry residue was suspended in anhydrous CH2C12 (22

ml), and palmitoyl chloride (1.37 g, 5.0 mmol) was added. Thesuspension was cooled on ice and N,N-diisopropylediylamine(1.04 ml, 6.0 mmol) was added widi stirring until die sedimenthad dissolved forming a slighdy colored solution. This solution,which had the concentration of palmitoyl tetrazole ca. 0.2 M,has been used for the 5'-derivatization of both 3'-O-acetyl-thymidine and oligonucleotides.

Derivatization of 3 '-O-acetylthymidine. 3'-O-Acetyldiymidine(114 mg, 0.4 mmol) was co-evaporated widi CH2Cl2to removetraces of water. The acylating mixture (10 ml, 2.0 mmol ofpalmitoyl tetrazole) and N,N-diisopropylemylamine (104 yX, 0.6mmol) were added to the residue with stirring. The reaction wasmonitored by TLC. In 10 min die starting material (Rf= 0.33)was completely converted into the product (Rf= 0.57). Thereaction mixture was diluted with CH2C12 and extracted widi10% aqueous NaHCC>3 (2x50 ml). The organic phase wasdried over Na2SO4 and evaporated to dry ness. The residue waschromatographed over silica gel using a gradient of 0—5%MeOH in CHC13 to yield 5 '-O-Palmitoyl-3' -O-acetylthymidineas a white solid (194 mg, 0.37 mmol, 93 %). 'H NMR (CHC13)5 7.31 (s, 1H, H6 (Thym)), 6.30-6.36 (m, 1H, HI'),5.19-5.23 (m, 1H, H3'), 4.27-4.45 (m/d, 2H, CH25'),4.22-4.27 (m, 1H, H4'), 2.42-2.52 (m, 1H, H2" (down)),2.32-2.41 (t, 2H, CH2CO (Palm)), 2.09-2.19 (s + m, 4H,H2' (up), Ac), 1.93 (s, 3H, CH3 (Thym)), 1.58-1.69 (m, 2H,CH2CH2CO (Palm)), 1.18-1.38 (m, 24H, (CHJ^ (Palm)),0.84-0.93 (t, 3H, CH3 (Palm)). FABMS m/z 521.4" (M -H+), 255.3

5-Palmitoylation of oligonucleotidesThe 5'-terminal DMT group was removed on the DNAsynthesizer. The CPG support widi immobilized oxalyl-boundprotected oligonucleotide (1 /xmol) was dried under vacuum andmen treated wim the acylating mixture (1 ml, 0.2 mmol ofpalmitoyl tetrazole, see above) and N,N-diisopropyl-ethylamine(10 nl, 57 /imol) for 30 min at room temperature. The reactionvessel was shaken several times during this procedure. The liquidwas filtered out, CPG was washed with CH3CN and CH2C12

and dried under vacuum.

Deprotection of 5'-esterified oligonucleotidesWith EA. 5'-Palmitoylated support-bound oligonucleotide (1—2/tmol) was treated with EA (200 /xl) for 10 min at roomtemperature. The reaction mixture was diluted with EtOH (400III), cooled on dry ice and neutralized with acetic acid (190 /d).The pH was neutralized by adding 1M TEAB (1 ml) and thefinal solution was desalted on Sephadex G25 (Pharmacia NAP-25column may also be used). All work up must be performed asquickly as possible to avoid considerable de-esterification. For0.2 fimol syndiesis the amount of reagents should be decreasedtimes diree.

With the mixture N^i^EAIMeOH. Deprotection with themixture N^/EA/MeOH (1:5:5, v/v/v) was performed as withEA, except the deprotection time was decreased to 3 min.With ammonia. The deprotection widi aqueous ammonia(25%,w/w) was performed at room temperature. After a specifictime the reaction mixture was diluted widi water and the solutionwas direcdy desalted on Sephadex G25.

Page 5: Antisense pro-drugs: 5'-ester oligodeoxynucleotides · 2017-04-16 · Antisense pro-drugs: 5'-ester oligodeoxynucleotides Nikolai N.Polushin and Jack S.Cohen* Cancer Pharmacology

5496 Nucleic Acids Research, 1994, Vol. 22, No. 24

ACKNOWLEDGEMENTS

We thank Mridul Ghosh, Nasser Farschtschi, and Tilak Raj forhelpful discussions in the early stages of this project. We arealso very grateful to Alan Morocho for his technical assistanceand to Pat Faustino for recording the NMR spectra. This workwas supported by a research grant from the Cystic FibrosisFoundation.

REFERENCES

1. Crooke.S.T. and LeWeu.B- (ed.) {\Wi) Antisense Research and Applications.CRC Press, Boca Raton, FL.

2. Ratajczak.M.Z., KanU.A., Luger.S.M., Huiya,N., Zhang.J., Zon,G. andGerwim.A.M. (1992) Proc. Natl. Acad. Sd. USA, 89, 11823-11827.

3. Calabretta.B , Sims,R.R., Valtien.M., Caracciolo.D., Szczylik,C,VentureUi,D. and Gewirtz.A.M. (1991) Proc. Natl. Acad. Set. USA, 88,2351.

4. Agrawal.S. and TangJ.Y. (1993) Antisense Res. Devel, 2, 261-266.5. Bayever.E., lversen,P., Smith,L., SpinoloJ. and Zon.G. (1992) Antisense

Res. Devel., 2, 109-110.6. JaroszewskiJ.W. and CohenJ.S. (1991) Adv. Drug Delivery Revs. 6,

235-250.7. Miller,P.S. (1989) in Cohen.J.S. (ed.) Ohgodeoxynucleotides: Antisense

Inhibitors of Gene Expression. CRC Press, Boca Raton, FL, pp 79-95.8. MacKellar.C, Graham.D., Will.D.W., Burgess,S. and Brown.T. (1992)

Nucleic Acids Res., 20, 3411-3417.9. Boutorin.A.S., Gus'kova.L.V., Ivanova.E.M., Kobetz.N.D., Zarytova.V.F.,

Ryte,A.S., Yurchenko.L.V. and Vlassov.V.V. (1989) FEBS LETTERS, 254,129-132.

10. Gmeiner,W.H., Luo,W., Pon.R.T. and LownJ.W. (1991) Bioorg. and Med.Otem. Lett., 1, 487-490.

11. Letsinger.R.L., Zhang,G., Sun.D.K., Ikeuchi.T. and Sarin.P.S. (1989) Proc.Natl. Acad. Sci. USA, 86, 6553-6556.

12. Stein.C.A., Pal.R., De Vico.A.L., Hoke.G., Mumbauer.S., Kinstler,O.,Samgadharan.M G. and Letsinger.R.L. (1991) Biochemistry, 30,2439-2444.

13. Reed.M.W., AdamsAD., NelsoiU.S. and Meyer.R.B. (1991) BioconjugateChem., 2, 217-225.

14. Oberhfliuer.B. and Wagner.E. (1992) Nucleic Acids Res., 20, 533-538.15. Will.D.W. and Brown,T. (1992) Tetrahedron Lett., 33, 2729-2732.16. Shea.R.G., Marsters.J.C. and Bischofberger.N (1990) Nucleic Acids Res.,

18, 3777-3783.17. Ghosh.M. and Mitra.A.K. (1991) Pharm. Res., 8, 771-775.18. Ivanovskaya.M.G., Gottikh.M.B. and Shabarova.Z.A. (1987) Nucleosides

& Nucleotides, 6, 913-934.19. Alul.R.H., Singman.C.N., Zhang.G. and Letsinger.R.L. (1991) Nucleic

Acids Res., 19, 1527-1532.20. SchulhoU.C, Molko.D. and Teoule,R. (1987) Nucleic Acids Res., 15,

397-416.21. Uznanslri,B., Grajkowski.A. and WUk.A. (1989) Nucleic Acids Res , 17,

4863-4871.22. Sinha.N.D., Davis,P., Usman.N., Perez,J., Hodge,R., Kremsky.J. and

Casale,R. (1993) Biochimie, 75, 13-23.23. Polushin.N.N., Pashkova.I.N. and Efimov.V.A. (1991) Bioorgan. Khimiya,

17, 1145-1148.24. Polushin.N.N., Pashkova,I.N. and Efimov.V.A. (1991) Nucleic Adds Res.,

Symp. Ser., NJ24, 49-50.25. Polushin.N.N., Pashkova.I.N., Chakhmakhcheva, O G. and Efimov.V.A.

(1993) Bioorgan. Khimiya, 19, 318-326.26. Potushin.N.N., Morocho.A.M.. Chen,B.-C. andCohenJ.S. (\994)Nudeic

Acids Res., 22, 639-645.27. Ghosh,M., Ghosh,K. and ConenJ.S. (1993) Anti-cancer Drug Des., 8,

15-32.