Table 1. N-Nitrothymidine-Containing Oligonucleotide-Resins
Table 2. Modification of Resins 5a-c with Primary Aminesa,b
Prepareda-c
oligonucleotide-resin
2
5a
5b
5c
T-T-T-TNO -T-T-T-resin
T-TNO -T-T-T-T-T -T-resin
T-GiBu-ABz-TNO -CBz-ABz-T-resin
NO2
2
2
a TentaGel-NH2 resin (Rapp Polymere GmbH) was used as the solid
support. The first 5′-DMT-protected nucleoside was attached to the resin
by means of a standard succinyl linker, and the initial substitution degree
was in the range of 40-190 µmol/g. b Oligonucleotide elongation was
carried out at 2-5 mmol scale following standard phosphite triester
procedures: 0.1-0.15 M solutions of 4 or the commercially available
(Glen Research Corp.) 3′-phosphoramidite derivatives (5′-DMT-Nu-
P[O(CH2)2CN]NiPr2, Nu ) TNO , T, ABz, CBz, GiBu) in anhydrous ACN
2
and 0.5-0.6 M solutions of tetrazole in anhydrous ACN were used for the
coupling steps (15 min). A 0.1 M solution of tBuOOH in CH2Cl2 was used
in the oxidation steps (1.5 min). All syntheses were performed in an Applied
Biosystems 380B DNA synthesizer. c Average coupling yields, calculated
from the UV absorbance at 498 nm of the DMT cation generated in the
deprotection steps, ranged between 99.5% (5a) and 98.2% (5c).
resin
modification conditions
product
5a
5a
5a
5b
5c
5c
5c
5c
MeNH2 (33% in H2O), 1 h, rt
BnNH2 (neat), 1 h, rt
6a
6b
6c
7
8a
8b
8c
into the corresponding oligomers using exactly the same
conditions as those employed for other phosphoramidite
derivatives, with no significant difference in the coupling
yields. In addition, no evidence of degradation of the
N-nitrothymidine moiety was found during the subsequent
elongation cycles.
H2N(CH2)3NH2 (neat), 1 h, rt
MeNH2 (33% in H2O), 1 h, rt
BnNH2 (0.25 M in ACN), 4 days, rt
H2N(CH2)3NH2 (0.1 M in ACN), 24 h, rt
HO(CH2)5NH2 (0.14 M in ACN), 48 h, rt
PhNH2 (neat), 2 h, rt
After assembly of the three oligonucleotide-resins
5a-c, we turned our attention to their reaction with primary
amines. As shown in Table 2, various oligonucleotide
analogues were prepared. The modification reaction was
straightforward and could be carried out efficiently under a
variety of conditions. Either the neat amine, concentrated
aqueous solutions, or dilute solutions of the amine in
acetonitrile (ACN) afforded the target product, provided that
the reaction times were long enough so as to allow for the
completion of the reaction.
Treatment of the oligonucleotide-resins with the primary
amines to some extent also cleaved the oligonucleotides from
the resin and removed protecting groups. In any case,
concentrated aqueous ammonia was added to the reaction
mixture, and the reaction was left to proceed overnight at
55 °C, to ensure completeness of all of the deprotection
processes (cleavage of the nucleoside-resin bond and
removal of the phosphate and nucleobase protecting groups).
This additional treatment was also found to reduce the
amount of byproducts present in the final crude mixtures.
As summarized in Table 2, good results were obtained
with all of the alkylamines, but reaction of resin 5c with
aniline led to a complex reaction mixture in which no
modified nucleoside-containing product could be identified.
This result indicates that aromatic amines are not nucleophilic
enough to complete the reaction and establishes the lower
limit to the scope of the proposed methodology.
a In all cases, the analytical reverse-phase HPLC chromatograms of the
crude products showed a single major peak (80-93%), which corresponded
to the expected oligomers 6-8, and very minor peaks associated with other
nucleosidic byproducts. As expected, retention times of the modified
oligonucleotides 6-8 (R * H) were always higher than those of the
corresponding unmodified ones (R ) H). Reverse-phase HPLC analyses
were performed on either Spherisorb (ODS-2, 25×0.4 cm, 5 µm) or
Phenomenex (C18, 25×0.46 cm, 10 µm) columns using linear gradients of
0.05 M aqueous triethylammonium acetate and ACN/H2O (1:1). b Identities
of all of the major products were established either by electrospray or
MALDI-TOF (2,4,6-trihydroxyacetophenone) mass spectrometry (negative
mode) of the HPLC-purified material. Identities of 8a-c were also
established by enzymatic digestion of the HPLC-purified materials with
snake venom phosphodiesterase and alkaline phosphatase. Samples of the
modified nucleosides were used as standards.
carried one and two N-nitrothymidine residues, respectively.
These results clearly point out that several modifications can
be simultaneously introduced into oligonucleotide-resins
bearing multiple N-nitrothymidine residues by means of a
single final reaction with primary amines. Moreover, the
modification reaction is compatible with the presence of
additional free amine and hydroxyl groups in the modifica-
tion reagent. Successful isolation of the target modified
oligonucleotides 8a and 8b, after treatment of the oligo-
nucleotide-resin 5c with either 5-aminopentanol or 1,3-
propanediamine (see Table 2), illustrates the fact that
protection of the second functional group in the modification
reagent is not necessary.
In summary, our results show that the phosphoramidite
derivative 4 is a good building block for the synthesis of
oligonucleotide-resins containing N-nitrothymidine residues.
The N-nitrothymidine residues behave as “convertible nucleo-
sides” that can be transformed into a range of N3-modified
thymidines by reaction with primary alkylamines.
It is worth noting that no remarkable differences were
found in the reactions involving resins 5a and 5b, which
(5) (1) Tetrazole, ACN, 24 h, rt; (2) Ac2O/2,6-lutidine, THF, 24 h, rt;
(3) N-methylimidazole, THF, 24 h, rt; (4) I2, H2O/py/THF, 24 h, rt; (5)
t-BuOOH 3 M, toluene, 24 h, rt.
Org. Lett., Vol. 4, No. 11, 2002
1829