Functionalized nucleoside 5'-triphosphates for in vitro selection of new catalytic ribonucleic acids

Article abstract of DOI:10.1016/S0960-894X(00)00226-2

A series of novel 2'-modified nucleoside 5'-triphosphates was synthesized. The amino, imidazole, and carboxylate functionalities were attached to the 5-position of pyrimidine base of these molecules through alkynyl and alkyl spacers, respectively. Two different phosphorylation methods were used to optimize the yields of these highly modified triphosphates. (C) 2000 Elsevier Science Ltd. All rights reserved.

Full text of DOI:10.1016/S0960-894X(00)00226-2

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1299±1302
Functionalized Nucleoside 50-triphosphates for In Vitro Selection
of New Catalytic Ribonucleic Acids
Jasenka Matulic-Adamic, Andrew T. Daniher, Alexander Karpeisky, Peter Haeberli,
David Sweedler and Leonid Beigelman*
Department of Chemistry & Biochemistry, Ribozyme Pharmaceuticals, Inc., 2950 Wilderness Place, Boulder, CO 80301, USA
Received 6 March 2000; accepted 12 April 2000
AbstractÐA series of novel 20-modi®ed nucleoside 50-triphosphates was synthesized. The amino, imidazole, and carboxylate func-
tionalities were attached to the 5-position of pyrimidine base of these molecules through alkynyl and alkyl spacers, respectively.
Two di€erent phosphorylation methods were used to optimize the yields of these highly modi®ed triphosphates. # 2000 Elsevier
Science Ltd. All rights reserved.
Recently, much attention has been focused on the devel-
opment of functionalized nucleotides suitable for in vitro
selection with the hope of increasing the potential of
nucleic acids for binding and catalysis.1 6 For RNA in
vitro selections modi®cations should be at the nucleotide
level so that they can be incorporated simply and eciently
using RNA polymerase without problematic side reactions
associated with synthetic posttranscriptional modi®cation.
20-O-Methyluridine was 30,50-bis-acetylated using acetic
anhydride in pyridine and then converted to its 5-iodo
derivative 1a using I2/ceric ammonium nitrate reagent8
(Scheme 1). Both reactions proceeded in a quantitative
yield and no chromatographic puri®cations were nee-
ded. Coupling between 1a and N-tri¯uoroacetyl pro-
pargylamine using copper(I) iodide and tetrakis (tri-
phenylphosphine)-palladium(0) catalyst as described by
Hobbs9 yielded 2a in 89% yield. Selective O-deacylation
with aqueous NaOH a€orded 3a, which was phosphory-
lated with POCl3/triethylphosphate (TEP) in the presence
of 1,8-bis(dimethylamino)naphthalene (Proton-Sponge)
(Method A).10 The intermediate nucleoside phosphorodi-
chloridate was condensed in situ with tri-n-butylammo-
nium pyrophosphate. At the end the N-TFA group was
removed with concentrated ammonia. 50-Triphosphate
was puri®ed on Sephadex DEAE A-25 ion exchange
column using a linear gradient of 0.1±0.8 M triethyl-
ammonium bicarbonate (TEAB) for elution. Traces of
contaminating inorganic pyrophosphate are removed
using C-18 RP HPLC to a€ord analytically pure mate-
rial. Conversion into Na-salt was achieved by passing
the aqueous solution of triphosphate through Dowex
50WX8 ion exchange resin in Na+ form to a€ord 4a in
45% yield (see Table 1). When Proton-Sponge was omit-
ted in the ®rst phosphorylation step, yields were reduced
to 10±20%. Catalytic hydrogenation of 3a yielded 5-
aminopropyl derivative 5a, which was phosphorylated
under conditions identical to those described for propy-
nyl derivative 3a to a€ord triphosphate 6a in 50% yield.
When designing monomeric nucleoside triphosphates
for selection of therapeutic catalytic RNAs one has to
take into account nuclease stability of such molecules in
biological sera. A common approach to increase RNA
stability is to replace the sugar 20-OH group with other
groups like 20-¯uoro, 20-O-methyl or 20-amino. For-
tunately such 20-modi®ed pyrimidine 50triphosphates are
shown to be substrates for RNA polymerases.5,7 On the
other hand it was shown that variety of substituents at
the 5-position of the pyrimidine is well tolerated by T7
RNA polymerase,1 most likely because the natural
hydrogen-bonding pattern of these nucleotides is pre-
served. We have chosen 20-¯uoro and 20-O-methyl pyr-
imidine nucleosides as starting materials for attachment
of di€erent functionalities to the 5-position of the base.
Both rigid (alkynyl) and ¯exible (alkyl) spacers were
used. The choice of imidazole, amino and carboxylate
pendant groups is based on their ability to act as general
acids, general bases, nucleophiles and metal ligands, all
of which can improve the catalytic e€ectiveness of
selected nucleic acids.
For the preparation of imidazole derivatized triphosphates
9a,b and 11a,b we developed an ecient synthesis of
*Corresponding author. Tel.: +1-303-449-6500; fax: +303-449-6995;
e-mail: beigel@rpi.com
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00226-2

1300
J. Matulic-Adamic et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1299±1302
Scheme 1. Synthesis of 5-[3-aminopropynyl(propyl)]uridine 50-triphosphates and 4-imidazoleacetic acid conjugates: Reagents and conditions: (i) N-
TFA propargylamine, CuI, tetrakis(Ph3P)Pd(0), Et3N, DMF, 16 h; (ii) aq NaOH, pyr, MeOH, 0 ꢀC, 1 h; (iii) POCl3, Proton-Sponge, (EtO)3PO, 2 h;
(iv) n-Bu3N PPi, MeCN, 15 min; (v) 1M Et3NH+HCO3 , then NH4OH, 16 h; (vi) H2, 5% Pd-C, 24 h, 40 psi; (vii) 40% MeNH2, 3 h; (viii) NH4OH,
4 ꢀC, 16 h; (ix) ImAADPC, EDCꢁ HCl, DMF,16 h.
N-diphenylcarbamoyl 4-imidazole-acetic acid (ImAADPC).
Transient protection of the carboxyl group as a TMS-ester
using TMS-Cl/pyridine followed by DPC-Cl allowed for a
clean and quantitative conversion of 4-imidazoleacetic
acid (ImAA) to its N-DPC protected derivative.
Complete deacylation of 2a a€orded 5-(3-aminopropynyl)
derivative 8a, which was condensed with 4-imidazoleace-
tic acid in the presence of 1-(3-dimethylaminopropyl)-3-
ethylcarbodiimide (EDC) to a€ord 9a in 68% yield. Cat-
alytic hydrogenation of 8a yielded 5-(3-aminopropyl)
derivative 10a which was condensed with ImAADPC to
yield conjugate 11a in 32% yield. Yields in these cou-
plings were greatly improved when 50-OH was protected
with DMT group (not shown) thus eciently preventing
undesired 50-O-esteri®cation. Both 9a and 11a failed to
yield triphosphate products in reaction with POCl3/
TEP/Proton-Sponge.
Scheme 2. Synthesis of 5-[3-(N-4-imidazoleacetyl)-aminopropynyl
(propyl)]uridine 50-triphosphates: Reagents and conditions: (i) DMT-
Cl, pyr, 16 h; (ii) Ac2O, pyr, 2 h; (iii) 3%TCA, CH2Cl2, 2 h; (iv) 2-Cl-
4H-1,3,2-benzodioxaphosphorin-4-one, pyr, dioxane, 30 min; (v) n-
Bu3N PPi, DMF, 30 min; (vi) I2, pyr-H2O, 20 min; (vii) NH4OH, 2 h.
On the contrary, phosphorylation of 30-O-acetylated
derivatives 12a and 13a using 2-chloro-4H-1,3,2- benzo-
dioxaphosphorin-4-one followed by pyrophosphate addi-
tion and oxidation (Method B,11 Scheme 2) a€orded the
desired triphosphate 14a and 15a in 57% yield, respec-
tively. 20-Deoxy-20-¯uoro nucleoside 50-triphosphates
containing amino- (4b and 6b) and imidazole- (14b and
15b) linked groups were synthesized in a manner analo-
gous to that described for the preparation of 20-O-
methyl nucleoside 50-triphosphates (Schemes 1 and 2).
Again, only Ludwig±Eckstein's phosphorylation11
worked for the preparation of 4-imidazoleacetyl deriva-
tized triphosphates.
It is worth noting that when `one-pot, two-steps' phos-
phorylation reaction (Method A)10 of 5b was quenched

J. Matulic-Adamic et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1299±1302
1301
with 40% aqueous methylamine instead of TEAB or
H2O, the g-amidate 7b was generated as the only detect-
able product. Similar reaction was reported recently for
the preparation or g-amidate of pppA20p50A20p50A.12
A carboxylate group was introduced at the 5-position of
uridine both on the nucleoside level and post-synthetically
(Method C, Scheme 3).
5-Iodo-20-deoxy-20-¯uorouridine (16) was coupled with
methyl acrylate using modi®ed Heck reaction13 to yield 17
in 85% yield. 50-O-Dimethoxytritylation, followed by in
situ 30-O-acetylation and subsequent detritylation a€or-
ded 30-protected derivative 18. Phosphorylation using 2-
chloro-4H-1,3,2-benzodioxaphosphorin-4-one followed
by pyrophosphate addition and oxidation11 a€orded the
desired triphosphate 19 in 54% yield.14 On the other
hand, 5-(3-aminopropyl)uridine 50-triphosphate 6b was
coupled with N-hydroxysuccinimide ester of Fmoc-Asp-
OFm to a€ord, after removal of Fmoc and Fm groups
with diethylamine, the desired aminoacyl conjugate 20
in 50% yield.
Scheme 3. Synthesis of carboxylate tethered uridine. Reagents and
conditions: (i) methyl acrylate, Ph3P, Pd(II)acetate, Et3N, dioxane, 30
min, re¯ux; (ii) DMT-Cl, pyr, 16 h; (iii) Ac2O, pyr, 3 h; (iv) 3% TCA,
CH2Cl2, 1 h; (v) 2-Cl-4H-1,3,2-benzodioxaphosphorin-4-one, pyr,
dioxane, 30 min; (vi) n-Bu3N PPi, DMF, 30 min; (vii) I2, pyr-H2O, 20
min; (viii) 1N NaOH, 5 h; (ix) Fmoc-Asp-OFm NHS-ester, DMF-
0.1M Na2B4O7, 16 h, then Et2NH, 3 h.
Cytidine derivatives comprising 3-aminopropyl and 3(N-
succinyl)aminopropyl groups were synthesized accord-
ing to Scheme 4. Peracylated 5-(3-aminopropynyl)uracil
derivative 2b is reduced using catalytic hydrogenation
and then converted in seven steps and 5% overall yield
into 30-acetylated cytidine derivative 25. This synthesis
was plagued by poor solubility of intermediates and for-
mation of the N4-cyclized byproduct during ammonia
treatment of the 4-triazolyl intermediate. Phosphoryla-
tion of 25 as described in ref. 11 yielded triphosphate 26
and N4-cyclized product 27 in 1:1 ratio. They were
easily separated on Sephadex DEAE A-25 ion exchange
colum using 0.1±0.8 M TEAB gradient. It appears that
under basic conditions the free primary amine can dis-
place any remaining intact 4-NHBz group leading to the
cyclized product. This is similar to displacement of 4-
triazolyl group by primary amine as mentioned above.
We reasoned that utilization of N4-unprotected cytidine
would solve this problem. This lead to an improved
synthesis of 26: Iodination of 20-deoxy-20-¯uorocytidine
(28) provided the 5-iodo derivative 29 in 58% yield.
Scheme 4. Synthesis of 5-(3-aminoalkyl) and 5-[3(N-succinyl)aminopropyl] functionalized cytidine. Reagents and conditions: (i) H2, 5% Pd-C,
MeOH, 24 h, 40 psi; (ii) POCl3, 1,2,4-triazole, Et3N, MeCN, 16 h; (iii) NH4OH, dioxane, 16 h; (iv) CF3COOEt, Et3N, MeOH, re¯ux, 3 h; (v) Bz2O,
EtOH, re¯ux, 5 h; (vi) DMT-Cl, pyr, 16 h; (vii) Ac2O, pyr, 3 h; (viii) 3% TCA, CH2Cl2, 3 h; (ix) HIO3, I2, AcOH, CCl4, H2O, 45 ꢀC, 4 h; (x) N-TFA
propargylamine, CuI, tetrakis(Ph3P)Pd(0), Et3N, DMF, 16 h; (xi) H2, 5% Pd-C, MeOH, 72 h, 40 psi; (xii) POCl3, Proton-Sponge, (MeO)3PO, 2 h;
(xiii) n-Bu3N PPi, MeCN, 15 min; (xiv) NH4OH, 4 ꢀC, 16 h; (xv) succinic anhydride, DMF-0.1 M Na2B4O7 1:1, 16 h.

1302
J. Matulic-Adamic et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1299±1302
Table 1. 5-Functionalized nucleoside 50-triphosphates
Compound
Method
Yield (%)
31P NMR (d, j [Hz])
Pb
Pa
Pg
4a
4b
6a
6b
A
A
A
B
A
B
B
B
B
45
36
50
54
28
57
17
57
14
54
25
37, (22)
22
36
10.50 (d, J=20)
10.50 (d, J=20)
10.35 (d, J=20)
7.02 (d, J=20)
0.48 (d, J=20)
10.26 (d, J=20)
7.12 (d, J=20)
10.35 (d, J=20)
6.51 (d, J=21)
12.34 (d, J=21)
8.60 (d, J=219)
11.56 (d, J=20)
12.28 (d, J=21)
12.20 (d, J=21)
11.48 (d, J=20)
12.97 (d, J=20)
11.66 (d, J=20)
11.24 (d, J=20)
11.55 (d, J=20)
11.59 (d, J=20)
11.47 (d, J=22)
11.66 (d, J=20)
11.40 (d, J=21)
13.53 (d, J=21)
11.41 (d, J=19)
13.59 (d, J=20)
13.47 (d, J=21)
13.46 (d, J=21)
22.75 (t, J=20)a
22.38 (t, J=20)b
22.83 (t, J=20)a
21.84 (t, J=20)a
22.56 (t, J=20)a
22.82 (t, J=20)a
22.13 (t, J=20)a
22.83 (t, J=20)a
21.90 (t, J=20)a
25.61 (t, J=21)b
22.38 (t, J=19)a
24.12 (t, J=20)b
25.42 (t, J=20)b
25.37 (t, J=21)b
7b
14a
14b
15a
15b
19
20
26
27
32
B
Cc
A, (B)
B
Cc
aNa+ salt, recorded in D2O.
bTEA+, salt recorded in CD3OD.
cPost-synthetic coupling.
References and Notes
This compound was then smoothly converted into 5-(3-
aminopropynyl) derivative 30. Hydrogenation a€orded
5-(3-aminopropyl) derivative 31, which was phosphory-
lated directly with POCl3/PPi to a€ord 26 in 37% yield.
Reaction of the 50-triphosphate 26 with succinic anhy-
dride yielded succinylated derivative 32 in 36% yield.
1. Tarasow, T. M.; Eaton, B. E. Biopolymers 1998, 48, 29.
2. Eaton, B. E.; Pieken, W. A. Annu. Rev. Biochem. 1995, 64,
837.
3. Eaton, B. E. Curr. Opin. Chem. Biol. 1997, 1, 10.
4. Dewey, T. M.; Mundt, A. A.; Crouch, G. J.; Zyzniewski, M.
C.; Eaton, B. E. J. Am. Chem. Soc. 1995, 32, 8475.
5. Aurup, H.; Williams, D. M.; Eckstein, F. Biochemistry
1992, 31, 9637.
All new nucleoside triphosphates gave correct molecular
peaks when analyzed by MALDI TOF mass spectro-
metry.15
6. Sakthivel, K.; Barbas III, C. F. Angew. Chem., Int. Ed. Engl.
1998, 37, 2872.
In conclusion, 20-modi®ed uridine and cytidine tripho-
sphates containing reactive functionalities tethered to
the 5-position were synthesized. Their ability to serve as
substrates for RNA polymerases and subsequent utili-
zation in in vitro selection of new nuclease-resistant
phosphodiesterases will be reported in due course.
7. Padilla, R.; Sousa, R. Nucleic Acids Res. 1999, 27, 1561.
8. Asakura, J.; Robins, M. J. J. Org. Chem. 1990, 55, 4928.
9. Hobbs, F. W., Jr. J. Org. Chem. 1989, 54, 3420.
10. Kovacz, T.; Otvos, L. Tetrahedron Lett. 1988, 29, 4525.
11. Ludwig, J.; Eckstein, F. J. Org. Chem. 1989, 54, 631.
12. Nyilas, A. Tetrahedron Lett. 1997, 38, 2517.
13. Dyer, R. L.; Jones, A. S.; Walker, R. T. In: Nucleic Acid
Chemistry; Townsend, L. B.; Tipson, R. T., Eds.; John Wiley
& Sons: New York, 1991; p 79.
14. The same strategy was used to prepare 20-O-Me derivatives
(data not shown).
Acknowledgements
We thank Lori Andrews for MALDI-TOF mass spec-
trometry analysis of modi®ed nucleoside triphosphates.
15. Burgess, K.; Russell, D. H.; Shitangkoon, A.; Zhang, A. J.
Nucleosides Nucleotides 1996, 15, 1719.