Zwitterionic oligonucleotides with 2'-O-[3-(N,N-dimethylamino)propyl]- RNA modification: Synthesis and properties

Article abstract of DOI:10.1016/S0040-4039(00)00703-6

A novel 2'-modification, 2'-O-[3-(N,N-dimethylamino)propyl] or 2'-O- DMAP, has been incorporated into oligonucleotides and compared to the known 2'-O-(3-aminopropyl) or 2'-O-AP modification for antisense properties. The 2'-O-DMAP modified oligonucleotides exhibit very high nuclease resistance like the 2'-O-AP modification due to the 'charge effect' and maintain high binding affinity to target RNA relative to known modifications when a few 2'- O-DMAP residues are dispersed throughout the oligonucleotide. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00703-6

Tetrahedron Letters 41 (2000) 4855±4859
Zwitterionic oligonucleotides with
20-O-[3-(N,N-dimethylamino)propyl]-RNA modi®cation:
synthesis and properties
Thazha P. Prakash, Muthiah Manoharan,* Allister S. Fraser, Andrew M. Kawasaki,
Elena A. Lesnik and Stephen R. Owens
Department of Medicinal Chemistry, Isis Pharmaceuticals, 2292 Faraday Ave, Carlsbad, CA 92008, USA
Received 7 March 2000; revised 26 April 2000; accepted 27 April 2000
Abstract
A novel 20-modi®cation, 20-O-[3-(N,N-dimethylamino)propyl] or 20-O-DMAP, has been incorporated into
oligonucleotides and compared to the known 20-O-(3-aminopropyl) or 20-O-AP modi®cation for antisense
properties. The 20-O-DMAP modi®ed oligonucleotides exhibit very high nuclease resistance like the 20-O-
AP modi®cation due to the `charge e€ect' and maintain high binding anity to target RNA relative to
known modi®cations when a few 20-O-DMAP residues are dispersed throughout the oligonucleotide. # 2000
Elsevier Science Ltd. All rights reserved.
Cationic oligonucleotides1 (oligonucleotides with positively charged backbones) and zwitter-
ionic oligonucleotides2 4 (oligonucleotides with positively charged tethers) are interesting classes
of nucleic acids because of their intrinsically favorable binding properties towards RNA and single-
stranded as well as double-stranded DNA. High anity binding is due to charge neutralization
and potentially fast on-rates of hybridization. In addition, these oligonucleotides are expected to
have good cellular permeation properties for potential therapeutic applications.
The 20-position of the carbohydrate moiety has proven to be a prime position for oligonucleotide
modi®cations for antisense technology. Many of these 20-modi®cations exhibit high binding anity
to target RNA, enhanced chemical stability and nuclease resistance, as well as increased lipo-
0
0
5
0
philicity. All high binding anity 2 -modi®cations have a C3 -endo sugar pucker. Among the 2 -
modi®cations reported in the literature, the 20-O-(3-aminopropyl) modi®cation (Fig. 1), 20-O-AP,3
o€ers a 1.0ꢀC increase in binding anity/modi®cation as a diester (20-O-AP/PˆO) compared to the
®rst generation of 20-deoxyphosphorothioate (20-H/PˆS) compounds when the positive charge is
dispersed in the oligonucleotide chain. This modi®cation as a phosphodiester linkage exhibits
nuclease resistance (measured as the half-life of the full-length oligonucleotide, t1/2) approximately
6 to 8 times better than a 20-deoxyphosphorothioate modi®cation. Thus, 20-O-aminopropyl
* Corresponding author. E-mail: mmanoharan@isisph.com
0040-4039/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00703-6

4856
modi®ed oligonucleotides5 have shown the greatest nuclease resistance4 of any modi®cation tested
thus far (t1/2>24 h in snake venom phosphodiesterase assays compared to 4 min for unmodi®ed
phosphodiester oligonucleotide). The superior nuclease resistance of the 20-O-aminopropyl mod-
i®cation was attributed to the `charge e€ect' present in the cationic side chain of the 20-O-amino-
propyl modi®ed oligonucleotides, which interacts competitively with the metal ions involved in
the phosphodiesterase activity of the nucleolytic enzymes.3,4
Figure 1.
We have now synthesized the dimethyl analog of the 20-O-(3-aminopropyl), the 20-O-DMAP
modi®cation (Fig. 1) and evaluated its antisense properties. Dialkylation of the amino group as in
the 20-O-DMAP modi®cation would eliminate the need for a protecting group for the amino
functionality in the synthesis of the 20-O-aminopropyl modi®ed oligonucleotides.3 The protecting
groups now used for 20-O-aminopropyl modi®ed oligonucleotides, phthalimido and tri¯uoro-
acetyl, have certain limitations in terms of monomer synthesis and oligonucleotide deprotection
conditions.6 Moreover, we expected the 20-O-DMAP oligonucleotides to be more lipophilic than
the 20-O-AP modi®ed oligonucleotides, a characteristic which could improve the cellular and
protein binding properties and hence the biodistribution of the 20-O-DMAP oligonucleotides
compared to 20-O-AP oligonucleotides.
We recently reported a novel and ecient scheme to synthesize the 20-O-DMAP modi®ed
5-methyluridine nucleoside 1 and the corresponding amidite 27 involving alkylation of the 20-O-
position using a six-membered cyclic sulfate (1,3,2-dioxathiane 2,2-dioxide) followed by nucleophilic
displacement with dimethylamine (Fig. 2). Treatment of N-3-benzyloxymethyl-5-methyluridine
with NaH in DMF at ^45ꢀC followed by addition of the cyclic sulfate a€orded the 20-O-propyl-
sulfate in 50% yield, with a selectivity for 20 over 30 alkylation of about 3:1. The sulfate was then
displaced using dimethylamine as a nucleophile in an autoclave. Reductive cleavage of the N-3-
benzyloxymethyl (BOM) protecting group using catalytic hydrogenation over a palladium
hydroxide catalyst proceeded smoothly to give the 20-modi®ed nucleoside 1 in 67% yield. This
compound was then protected at the 50-hydroxyl with the dimethoxytrityl (DMT) group and then
converted to the phosphoroamidite 2.
The nucleoside 1 was loaded onto controlled pore glass (CPG) following the standard synthetic
procedure8 to obtain the functionalized CPG 3 with 53 mmol/g loading capacity. The amidite 2
was dissolved in anhydrous acetonitrile to 0.1 M and loaded onto an Expedite Nucleic Acid
Synthesis system (Millipore 8909) to synthesize the oligonucleotides. The coupling eciencies

4857
Figure 2.
were more than 98%. For the modi®ed amidite 2, the coupling time was extended to 10 min and
coupling was repeated twice. Oxidation of the phosphite intermediate was carried out using CSO
[1-S-(+)±(10-camphorsulfonyl) oxaziridine].9 All other steps in the protocol supplied by Millipore
were used without modi®cation. After completion of the synthesis, the CPG was suspended in
aqueous ammonia (30%) at room temperature for 2 h to cleave the oligonucleotide from the
CPG. The CPG was ®ltered; the ®ltrate was then heated at 55ꢀC for 6 h to e€ect the complete
removal of the base-labile protecting groups. Ammonia was removed on a speed vac concentrator.
The resulting product was puri®ed by High Performance Liquid Chromatography (HPLC,
Waters, C-4, 7.8Â300 mm, A=50 mM triethylammonium acetate (TEAAc), pH=7, B=aceto-
nitrile, 5 to 60% B in 55 min, ¯ow 2.5 mL/min, l=260 nm). Detritylation with aqueous 80%
acetic acid and evaporation followed by desalting by HPLC on a Waters C-4 column gave the
20-modi®ed oligonucleotides listed in Table 1, which were analyzed by HPLC, capillary gel electro-
phoresis (CGE) and mass spectrometry.
Table 1
Oligonucleotides containing the 20-O-[3-(N,N-dimethylamino)propyl] modi®cation
When the modi®ed oligonucleotides were hybridized to complementary RNA, the change in
Tm was dependent on the placement of the modi®cations (Table 2). When the substitutions were
not adjacent as in ISIS 111710, the modi®ed oligonucleotide showed a +1.2ꢀC increase in Tm for

4858
Table 2
Tm values of 20-O-[3-(N,N-dimethylamino)propyl] modi®ed PˆO oligonucleotides against RNA
each substitution compared to unmodi®ed DNA. This observed di€erence of a 1.2ꢀC per mod-
i®cation translates to a nearly 2ꢀC increase/modi®cation when compared0to the ®rst generation of
antisense compounds in clinical use, the 20-deoxyphosphorothioate (2 -H/PˆS) compounds.10
However, when substitutions were placed adjacent to each other, the enhanced stability was lost
(ISIS 111709). As more modi®cations were incorporated (ISIS 111711), destabilization was
observed, perhaps owing to actual charge repulsion between adjacent cationic groups. This is in
agreement with the reported hybridization pro®le of the 20-O-aminopropyl and 20-O-aminoethyl
oligonucleotides.3,6 The addition of steric bulk because of the two methyl groups does not cause
additional destabilization relative to other 20-O-amino modi®cations, however. In fact, there is
stabilization of almost +0.6ꢀC/modi®cation in the 111710 sequence over the 20-O-AP oligo-
nucleotide with the same sequence and modi®cation pattern. This di€erence might be due to the
relative extents of hydration, which would be less for 20-O-DMAP (one hydrogen bond with
water in the protonated form) than 20-O-AP (three hydrogen bonds with water in the protonated
form).4
The 20-O-DMAP end-capped oligonucleotide ISIS 116168 was digested with snake venom
phosphodiesterase (SVPD) to evaluate the 30-exonuclease resistance of this modi®cation.12 Fig. 3
shows the relative nuclease stability of the 20-O-DMAP modi®ed oligonucleotide compared to the
Figure 3. Nuclease resistance of 20-H, 20-O-AP and 20-O-DMAP modi®ed oligonucleotides

4859
unmodi®ed oligonucleotide and one with 20-O-AP modi®cations. Incorporation of 20-O-DMAP T
1
residue provides very high exonuclease resistance with t > 24 h and it is similar to that of 20-O-
2
AP modi®cation
High binding anity to RNA, favorable lipophilicity, synthetic convenience, and extremely
high nuclease resistance as diester (PˆO) linkages all warrant further in vivo pharmacokinetic
and biodistribution evaluation of 20-O-DMAP modi®ed oligomers in antisense and triplex-
forming strategies. Furthermore, this modi®cation is expected to improve the lifetime and ecacy
of oligomers, having a favorable impact on the frequency of administration and on the oral
absorption potential of therapeutic oligonucleotides.
Acknowledgements
We thank Dr. P. Dan Cook (Isis Pharmaceuticals) and Professor Mike Jung (UCLA) for
stimulating discussions.
References
1. Letsinger, R. L.; Singman, C. N.; Histand, G.; Salunkhe, M. J. Am. Chem. Soc. 1988, 110, 4470±4471.
2. (a) Guinosso, C. J.; Hoke, G. D.; Freier, S.; Martin, J. F.; Ecker, D. J.; Mirabelli, C. K.; Crooke, S. T.; Cook,
P. D. Nucleosides Nucleotides 1991, 10, 259±262; (b) Manoharan, M.; Guinosso, C. J.; Cook, P. D. Tetrahedron
Lett. 1991, 32, 7171±7174; (c) Hashimoto, H.; Nelson, M. G.; Switzer, C. J. Am. Chem. Soc. 1993, 115, 7128±7134;
(d) Manoharan, M.; Tivel, K. L.; Andrade, L. K.; Cook, P. D. Tetrahedron Lett. 1995, 36, 3647±3650; (e)
Chaturvedi, S.; Horn, T.; Letsinger, R. L. Nucleic Acids Res. 1996, 24, 2318±2323; (f) Horn, T.; Chaturvedi, S.;
Balasubramaniam, T. N.; Letsinger, R. L. Tetrahedron Lett. 1996, 37, 743±746; (g) Manoharan, M.; Ramasamy,
K. S.; Mohan, V.; Cook, P. D. Tetrahedron Lett. 1996, 37, 7675±7678; (h) Cuenoud, B.; Casset, F.; Husken, D.;
Natt, F.; Wolf, R. M.; Altmann, K.-H.; Martin, P.; Moser, H. E. Angew. Chem., Int. Ed. 1998, 37, 1288±1291.
3. Gri€ey, R. H.; Monia, B. P.; Cummins, L. L.; Freier, S.; Greig, M. J.; Guinosso, C. J.; Lesnik, E.; Manalili, S. M.;
Mohan, V.; Owens, S.; Ross, B. S.; Sasmor, H.; Wancewicz, E.; Weiler, K.; Wheeler, P. D.; Cook, P. D. J. Med.
Chem. 1996, 39, 5100±5109.
4. Teplova, M.; Wallace, S. T.; Tereshko, V.; Minasov, G.; Symons, A. M.; Cook, P. D.; Manoharan, M.; Egli, M.
Proc. Natl. Acad. Sci. USA 1999, 96, 14240±14245.
5. (a) Manoharan, M. Biochim. Biophys. Acta 1999, 1489, 117±130; (b) Cook, P. D. Ann. Rev. Med. Chem. 1998, 33,
313±325.
6. Manoharan, M.; Prakash, T. P.; Barber-Peoc'h, I.; Bhat, B.; Vasquez, G.; Ross, B. S.; Cook, P. D. J. Org. Chem.
1999, 64, 6468±6472.
7. Fraser, A. S.; Kawasaki, A. M.; Jung, M. E.; Manoharan, M. Tetrahedron Lett. 2000, 41, 1523±1526.
8. (a) Kumar, P.; Sharma, A. K.; Sharma, P.; Garg, B. S.; Gupta, K. C. Nucleosides Nucleotides 1996, 15, 879±888;
(b) Bayer, E.; Bleicher, K.; Maier, M. A. Z. Naturforsch. 1995, 50b, 1096±1100.
9. Manoharan, M.; Lu, Y.; Casper, M. D.; Just, G. Org. Lett. 2000, 2, 243±246.
10. The 20-deoxyphosphorothioates cause ^0.8ꢀC/modi®cation in RNA hybridization compared to DNA.
11. Freier, S. M.; Altmann, K. H. Nucleic Acids Res. 1997, 25, 4429±4443.
12. Cummins, L. L.; Owens, S. R.; Risen, L. M.; Lesnik, E. A.; Freier, S. M.; McGee, D.; Guinosso, C. J.; Cook. P. D.
Nucleic Acids Res. 1995, 23, 2019±2024.