R. Krishnamurthy et al.
FULL PAPERS
the mass analysis did not show signals corresponding to the
desired hexamer. Another set of experiments was carried
out to compare results between 1) the full palladium-cata-
lyzed deprotection protocol; 2) the full palladium-catalyzed
deprotection protocol, but without the sodium diethyldithio-
carbamate washing step; and 3) full deprotection protocol,
Various deprotection conditions were tested (NH2NH2 at
58C or MeNH2 (44% aq)/EtOH (3:1) at RT). MeNH2 treat-
ment showed a sharp signal at 19 minutes in the HPLC
trace, and correspondingly the expected mass for the unpro-
tected 4-mer in the mass spectrum (see the Supporting In-
formation, Figure S98); thus, this treatment became the
method of choice. We then attempted the synthesis of a gly-
but with a different catalyst, [PdACHTNUTRGNEUNG(PPh3)4], for the removal of
the allyloxycarbonyl groups. In all three cases the same
signal patterns were observed in the HPLC traces with very
little of the signal (UV-active material) that would corre-
spond to the retention time of the desired hexamer. We also
attempted a SN2’ reaction with mercaptoethanol because of
a literature precedent for the removal of the allyl protecting
group at the phosphate;[21] but this approach was not suc-
cessful in our hands.
We checked the efficiency of the palladium-catalyzed de-
protection of the monomeric compound 17 and found that
the removal of the allyloxycarbonyl group under identical
conditions proceeded smoothly to afford 8. This observation
suggested that the problems we were facing were at the oli-
gomeric level and indicated the instability of the acyclic
phosphate triester backbone under these conditions. Com-
pound 17 was also subjected to treatment with aqueous
NH4OH, MeNH2, or N2H4 to check for the removal of the
allyloxycarbonyl groups by nucleophilic substitution (as car-
ried out with the benzoyl groups previously). However,
these experiments gave a mixture of compounds that could
only be accounted for as urea-type derivatives formed at the
amino groups at C2, C4, or both; the best result was achieved
by using hydrazine to give 50% yield of the desired depro-
tected product 8 and 50% of a mono-semicarbazide deriva-
tive.
(
5-aPNO)6-mer capped with d(T) starting with phosphorami-
dite 35. The trityl assay showed 91% average efficiency per
coupling step. We employed the methylamine deprotection
protocol to obtain the free oligomer. HPLC analysis showed
that after 4 hours at RT, 3 sharp signals appeared at 21, 22
and 23 minutes (see the Supporting Information, Fig-
ure S99d); after purification and desalting, these were as-
signed to the NO-4mer (gly-(5-aPNO)4-d(T)), NO-5mer (gly-
(
5-aPNO)5-d(T)), and the desired NO-6mer (gly-(5-aPNO)6-
d(T)).
To improve on these results and minimize the fragmenta-
tion of the oligomer, we explored the following conditions:
1) an aqueous solution of NH3 at longer reaction times,
2) pretreatment with a solution of Et3N in anhydrous
CH3CN, 3) an aqueous solution of Et3N, and 4) tert-butyla-
mine (see the Supporting Information, Table S5). The best
results were obtained with Et3N in anhydrous acetonitrile,
which removes selectively the cyanoethyl group to afford
the phosphodiester-linked oligomer, followed by exposure
to methylamine at RT for 4 hours (see the Supporting Infor-
mation, Figure S100). The crude oligomer was purified by
ion-exchange HPLC and desalted over Sephadex by size-ex-
clusion chromatography. The MALDI-TOF mass of this oli-
gomer could only be acquired if the sodium ions were com-
pletely exchanged by treatment with ion-exchange resin
(IR-120) in the triethylammonium form, as documented in
Figure 1.
We have not been able to prepare an phosphoglyceric
acid oligomer tagged with 2,4,5-triaminopyrimidine. This
failure is attributed mainly to the instability of the oligomer
under the various conditions employed for the removal pro-
tecting groups on the 2,4,5-triaminopyrimidine heterocycle.
This instability may be a function of the 2,4,5-triaminopyri-
midine heterocycle because the corresponding 2,4-dioxo-5-
aminopyrimidine-tagged oligomers have been shown to be
isolable under comparable deprotection conditions.
Base-Pairing Studies
The base-pairing properties of these oligomers (see the Sup-
porting Information, Table S6) were investigated by the use
of temperature-dependent UV and circular dichroism (CD)
absorption spectroscopy in 10 mm Na2HPO4 (pH 7.0) buffer
containing 1m NaCl and 0.1 mm Na2EDTA (EDTA=ethyle-
nediaminetetraacetic acid).
4-Oxo-2,5-diaminopyrimidine (5-aPNO) Series gly-(5-aPNO)n
Oligomer synthesis was first attempted by using the cya-
noethyl derivative 35 (Scheme 9). We intended to study the
modes of deprotection and stability of the oligomer under
different conditions. We synthesized a gly-(5-aPNO)-4mer
with the DMT in place (to avoid strand scission that may be
caused by the terminal OH). The oligomer synthesis pro-
ceeded with greater than 90% average efficiency per step.
The solid-support-containing oligomer was deprotected (aq
NH3/EtOH, 3:1) at room temperature; HPLC traces after
22 hours showed no further change. MALDI-TOF analysis
showed not only the signal for the desired oligomer without
the final DMT group (under mass measurement conditions
the DMT group falls off), but also other signals correspond-
ing to partial removal of the acetate and cyanoethyl groups.
2,4-Dioxo-5-aminopyrimidine-Tagged Hexamer Series 3’-gly-
(
2ꢀ
5-aPOO)6-2’-OPO3
We compared the temperature-dependent UV absorption
behavior of the hexamer alone and in the presence of the
corresponding complementary DNA and RNA sequences
and determined that 260 nm was the appropriate wavelength
that should be used to measure the melting temperature, Tm
(Table 1 and Figure 2). Base pairing with both complemen-
tary adenine-containing poly-(dA) and poly-(rA) sequences
was observed, as evidenced by the sigmoidal UV melting
curves. However, we were surprised by the magnitude of the
strength of cross-pairing, as indicated by the UV-Tm melting
values, which translated to about 108C per base pair with
1258
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2011, 6, 1252 – 1262