Janus-AT bases: Synthesis, self-assembly, and solid state structures

Article abstract of DOI:10.1021/jo061840k

The high yielding synthesis of heterocycles with defined H-bond accepting and donating capabilities provides for the design of self-assembling structures and specific recognition of biological targets. Herein we report the syntheses and solid-state structures of three self-complementary uracil/thymine derivatives where each presents the standard ADA face inherently complementary to adenine and a synthetically appended DAD face complementary to uracil/thymine. These heterocycles, which have never before been reported or characterized, represent diaminopurine-uracil/thymine hybrids that, in two of the three cases, relate to previously reported heterocyclic hybrids of G and C. All three heterocycles crystallized to afford the first X-ray crystal structures of self-complementary heterocycles capable of ADA-DAD pairing. The potential use in DNA and RNA recognition are briefly discussed.

Full text of DOI:10.1021/jo061840k

Janus-AT Bases: Synthesis, Self-Assembly,
and Solid State Structures
Ali Asadi, Brian O. Patrick, and David M. Perrin*
Department of Chemistry, UniVersity of British Columbia, 2036 Main Mall,
VancouVer, British Columbia V6T-1Z1
dperrin@chem.ubc.ca
ReceiVed September 6, 2006
The high yielding synthesis of heterocycles with defined H-bond accepting and donating capabilities
provides for the design of self-assembling structures and specific recognition of biological targets. Herein
we report the syntheses and solid-state structures of three self-complementary uracil/thymine derivatives
where each presents the standard ADA face inherently complementary to adenine and a synthetically
appended DAD face complementary to uracil/thymine. These heterocycles, which have never before
been reported or characterized, represent diaminopurine-uracil/thymine hybrids that, in two of the three
cases, relate to previously reported heterocyclic hybrids of G and C. All three heterocycles crystallized
to afford the first X-ray crystal structures of self-complementary heterocycles capable of ADA-DAD
pairing. The potential use in DNA and RNA recognition are briefly discussed.
Introduction
and a semiconservative templated synthesis of DNA and RNA.
Furthermore, nucleic acids associate into various regular su-
pramolecular architectures including double and triple helices,
G-quartets, and i-motifs which may compete with the standard
Whereas the specificity characteristic of life in an aqueous
world is often attributed to hydrogen bonding, water effectively
competes for hydrogen bonds. Thus the specificity of interac-
tions attributed to H-bonding is further mediated by a delicate
interplay between hydrophobic forces and electrostatic interac-
tions. Such an interplay is typified in nucleic acids, wherein
correct H-bonding along with hydrophobic base stacking and
electrostatic interactions between phosphates and countercations
ensures sequence specificity for complementary hybridization
1
double helical structure.
Synthesis of unnatural molecules capable of defined H-
bonding interactions has led to the creation of artificial self-
assembling structures and more recently supramolecular non-
covalentpolymersthathavepotentialapplicationsinnanotechnology,
(1) Lane, A. N.; Jenkins, T. C. Curr. Org. Chem. 2001, 5, 845-869.
10.1021/jo061840k CCC: $37.00 © 2007 American Chemical Society
4
66
J. Org. Chem. 2007, 72, 466-475
Published on Web 12/21/2006

Janus-AT Bases
rosettes,2
6,30-32,38
regular noncovalent polymer arrays, and
4,5
molecular tectonics, and crystal engineering, particularly in
organic media where H-bonding is considerably stronger, and
in the solid phase where X-ray diffraction provides visualization
of H-bonded arrays.2-10
1
8,39-42
irregular crinkled heteromorphic composites.
When self-complementary monomers are designed to be
simultaneously complementary to two nucleobases capable of
Watson-Crick base pairing (i.e., G and C, or A and T), there
is the intriguing potential for sequence specific DNA recognition
that would compete with self-association. Such monomers have
been nicknamed Janus-bases after the two-faced Roman god.43
Nevertheless, the first example of such a J-base was actually a
non-self-complementary pyrimidine (a JGA-base) that that had
been designed to simultaneously recognize C and U, which
neither form a Watson-Crick base pair, nor readily associate
in solution. Consequently, the simultaneous association of the
Although H-bond mediated assembly of two different mono-
mers presenting complementary recognition sites (e.g., melamine
and cyanuric or barbituric acid) has been used extensively to
create a variety of architectures,11-28 there are far fewer
antecedent examples in which self-complementary monomers
have been examined for self-assembly.2
9-36
The consequence
of monomeric self-complementarity affords the possibility of
3
5,37
several multimeric material forms including trimers,
1
JGA-base with both C and U could be cleanly monitored by H
(
2) Brunsveld, L.; Folmer, B.; Meijer, E.; Sijbesma, R. Chem. ReV. 2001,
01, 4071-4097.
3) Fournier, J.; Maris, T.; Wuest, J. J. Org. Chem. 2004, 69, 1762-
775.
NMR titration that gave rise to new peaks which demonstrated
1
43,44
the expected three-way interaction.
Recently, the same JGA-
(
base was incorporated onto an octameric PNA homopolymer
third strand that afforded recognition of heteroduplex DNA
1
(
4) Gong, H.; Krische, M. J. Am. Chem. Soc. 2005, 127, 1719-1725.
4
5
(5) Ikeda, M.; Nobori, T.; Schmutz, M.; Lehn, J. Chem.sEur. J. 2005,
containing C-T mismatches.
1
1, 662-668.
6) Laliberte, D.; Maris, T.; Wuest, J. J. Org. Chem. 2004, 69, 1776-
787.
Although there have been several reports on the synthesis
and self-assembly of several classes of JGC-bases which self-
(
1
(
7) Steiner, T. Angew. Chem. 2002, 41, 48-76.
associate via a DDA-AAD interaction, and which could
(8) Su, D.; Wang, X.; Simard, M.; Wuest, J. Supramol. Chem. 1995, 6,
30-32,35,45
potentially recognize a G-C base pair,
a solution-phase
1
71-178.
9) ten Cate, A.; Kooijman, H.; Spek, A.; Sijbesma, R.; Meijer, E.
J. Am. Chem. Soc. 2004, 126, 3801-3808.
10) Lawrence, D. S.; Jiang, T.; Levett, M. Chem. ReV. 1995, 95, 2229-
260.
11) Archer, E.; Goldberg, N.; Lynch, V.; Krische, M. J. Am. Chem.
1
H NMR experiment that could demonstrate simultaneous
recognition of C and G by a JGC-base, analogous to that which
was used to demonstrate C and U recognition by the JGA-base,
would be difficult if not impossible to perform cleanly, and to
date no such experiment has been reported. To date, these JGC-
heterocycles have been characterized by various techniques that
verify only self-association whereas the potential for DNA
recognition has remained entirely unexplored.
(
(
2
(
Soc. 2000, 122, 5006-5007.
(
(
12) Archer, E.; Krische, M. J. Am. Chem. Soc. 2002, 124, 5074-5083.
13) Arduini, M.; Crego-Calama, M.; Timmerman, P.; Reinhoudt, D.
J. Org. Chem. 2003, 68, 1097-1106.
(14) Beijer, F.; Sijbesma, R.; Vekemans, J.; Meijer, E.; Kooijman, H.;
Spek, A. J. Org. Chem. 1996, 61, 6371-6380.
In order to eventually use J-bases to recognize both bona fide
Watson-Crick base pairs, i.e., GC and AT, in contrast to various
mismatches, one would need JAT-bases in addition to the
aforementioned JGC-bases. Surprisingly, the synthesis of a JAT-
base has never been reported, either with an eye to DNA
recognition, or more simply, in terms of an investigation of self-
association via potential DAD-ADA H-bonding interactions.46
Herein, we describe the design, high yielding syntheses, and,
to the best of our knowledge, the first X-ray structures of solid-
state supramolecular arrays of three JAT-bases capable of DAD-
ADA interactions (technically these are not AT-hybrids but
closer to a thymine-diaminopurine or a uracil-diaminopurine
hybrid). The monomers are shown in Figure 1. Several
(
(
(
15) Bernhardt, P. Inorg. Chem. 1999, 38, 3481-3483.
16) Binder, W. Monatsh. Chem. 2005, 136, 1-19.
17) Klok, H.; Jolliffe, K.; Schauer, C.; Prins, L.; Spatz, J.; Moller, M.;
Timmerman, P.; Reinhoudt, D. J. Am. Chem. Soc. 1999, 121, 7154-7155.
18) Mascal, M.; Hansen, J.; Fallon, P.; Blake, A.; Heywood, B.; Moore,
M.; Turkenburg, J. Chem-Eur. J. 1999, 5, 381-384.
19) Murray, T. J.; Zimmerman, S. C. J. Am. Chem. Soc. 1992, 114,
010-4011.
20) Perdigao, L.; Champness, N.; Beton, P. Chem. Commun. 2006, 538-
40.
(
(
4
5
2
(
(21) Scherman, O.; Ligthart, G.; Sijbesma, R.; Meijer, E. Angew. Chem.
006, 45, 2072-2076.
(
(
13.
22) Sessler, J. L.; Wang, R. Z. J. Org. Chem. 1998, 63, 4079-4091.
23) Seto, C. T.; Whitesides, G. M. J. Am. Chem. Soc. 1991, 113, 712-
7
(
24) Thalacker, C.; Miura, A.; De Feyter, S.; De Schryver, F.; Wurthner,
F. Org. Biomol. Chem. 2005, 3, 414-422.
25) Yagai, S.; Nakajima, T.; Karatsu, T.; Saitow, K.; Kitamura, A.
J. Am. Chem. Soc. 2004, 126, 11500-11508.
26) Zerkowski, J. A.; Seto, C. T.; Whitesides, G. M. J. Am. Chem. Soc.
992, 114, 5473-5475.
27) Zerkowski, J. A.; Seto, C. T.; Wierda, D. A.; Whitesides, G. M.
(
(37) Zimmerman, S. C.; Duerr, B. F. J. Org. Chem. 1992, 57, 2215-
2217.
(38) Kolotuchin, S.; Zimmerman, S. J. Am. Chem. Soc. 1998, 120, 9092-
9093.
(39) Mathias, J. P.; Seto, C. T.; Simanek, E. E.; Whitesides, G. M.
J. Am. Chem. Soc. 1994, 116, 1725-1736.
(40) Mathias, J. P.; Simanek, E. E.; Whitesides, G. M. J. Am. Chem.
Soc. 1994, 116, 4326-4340.
(
1
(
J. Am. Chem. Soc. 1990, 112, 9025-9026.
(
28) Zhang, X.; Chen, X. Cryst. Growth Des. 2005, 5, 617-622.
(29) Lehn, J. M.; Mascal, M.; Decian, A.; Fischer, J. J. Chem. Soc.,
Perkin Trans. 2 1992, 461-467.
30) Fenniri, H.; Mathivanan, P.; Vidale, K.; Sherman, D.; Hallenga,
K.; Wood, K.; Stowell, J. J. Am. Chem. Soc. 2001, 123, 3854-3855.
31) Mascal, M.; Hext, N.; Warmuth, R.; Arnall-Culliford, J.; Moore,
M.; Turkenburg, J. J. Org. Chem. 1999, 64, 8479-8484.
32) Mascal, M.; Hext, N. M.; Warmuth, R.; Moore, M. H.; Turkenburg,
J. P. Angew. Chem. 1996, 35, 2204-2206.
33) Nagai, K.; Hayakawa, K.; Ukai, S.; Kanematsu, K. J. Org. Chem.
986, 51, 3931-3939.
34) Schall, O. F.; Gokel, G. W. J. Am. Chem. Soc. 1994, 116, 6089-
100.
35) Sessler, J.; Jayawickramarajah, J.; Sathiosatham, M.; Sherman, C.;
Brodbelt, J. Org. Lett. 2003, 5, 2627-2630.
36) Zhu, P.; Kang, H.; Facchetti, A.; Evmenenko, G.; Dutta, P.; Marks,
T. J. Am. Chem. Soc. 2003, 125, 11496-11497.
(41) Mathias, J. P.; Simanek, E. E.; Zerkowski, J. A.; Seto, C. T.;
Whitesides, G. M. J. Am. Chem. Soc. 1994, 116, 4316-4325.
(42) Zerkowski, J. A.; Seto, C. T.; Whitesides, G. M. J. Am. Chem. Soc.
1992, 114, 5473-5475.
(
(
(43) Branda, N.; Kurz, G.; Lehn, J. M. Chem. Commun. 1996, 2443-
2444.
(
(44) Marsh, A.; Nolen, E. G.; Gardinier, K. M.; Lehn, J. M. Tetrahedron
Lett. 1994, 35, 397-400.
(
1
(45) Chen, D.; Meena; Sharma, S.; McLaughlin, L. J. Am. Chem. Soc.
2004, 126, 70-71.
(
6
(46) Zhu et al.36 published on growth of thin films based on the self-
assemply of 5-{4-[2-(4,6-diamino-[1,3,5]triazin-2-yl)-vinyl]-benzylidene}-
pyrimidine-2, 4,6-trione which presents a DAD face (diaminotriazine) along
with two ADA faces (benzylidinebarbituric acid) that are separated by an
extended styrenyl linker.
(
(
J. Org. Chem, Vol. 72, No. 2, 2007 467

Asadi et al.
disulfide51,52 in the presence of potassium carbonate in DMF,
followed by treatment of the resulting ketene dithioacetal
dipotassium salt, 7, with 2 equiv of methyl iodide in water-
acetonitrile to give ketene dithioacetal, 8. Intermediates 9a and
9
b were obtained via dropwise addition of appropriate amine
to an ethanolic solution of 5 at room temperature to afford a
presumed monoamine adduct that was cyclized at reflux to
afford the 5-cyano-6-methylthiouracils 9a and 9b which were
1
13
FIGURE 1. Janus-AT nucleobases 1, 2a, 2b, 3. Arrows indicate
expected H-bond donating or accepting functionalities.
characterized by both H/ C NMR and HRMS. In addition, 9a
was crystallized and characterized by X-ray diffraction to verify
the correct substitution at positions 5 and 6 (Supporting
Information). Compounds 9a and 9b were reacted with guani-
dine (free base generated in situ) in ethanol at reflux to afford
the corresponding bicyclic products 2a and 2b in good to
excellent yields.
considerations arose with regards to the design of JAT-1, JAT-2,
and JAT-3 as well as the choice for their corresponding synthetic
routes. To begin with, we chose to incorporate a diaminopurine-
like face instead of a more natural A-like (aminopurine) face
to afford three H-bonds instead of two with the thymine-like
face. Although the ADA-DAD hydrogen bonding array is not
generally found in regular DNA base pairs, it exists in the
cyanophage S-2L viral genome that contains 2,6-diaminopu-
The procedure for synthesizing 3 (Scheme 3) started with
the condensation of N-cyanoacetylurethane, 4, with triethyl
orthoacetate in acetic anhydride at reflux to afford 10 followed
by addition of butylamine and ring closure in water in high
yield (91%). 6-Methyl 5-cyanouracil, 11, was then condensed
with carbon disulfide in dry THF in the presence of potassium
t-butoxide at room temperature followed by acidic workup to
4
7
rine, a base that has been found to increase duplex stability
and specificity in synthetic systems relative to the more common
adenine and thymine (A-T) base pair.48 In addition, previous
reports on ADA-DAD associated networks involving melamine
guided our designs. From a synthetic perspective, heterocycles
with two exocyclic amines, i.e., JAT-1, JAT-2, and JAT-3, would
be somewhat more synthetically accessible compared to ana-
logues with only one exocyclic amine, particularly for the
synthesis of JAT-1 (Vide infra). JAT-1 was chosen for synthetic
ease despite the considerable rotational flexibility between the
two aryl rings. JAT-2 and JAT-3 were chosen because they are,
respectively, constitutional isomers of the JGC-base reported by
1
afford 12 in 89% yield which was characterized by both H/
C NMR and HRMS. In addition, 12 was crystallized in
13
acetonitrile and characterized by X-ray diffraction to verify the
formation of the novel structure. Compound 12 was treated
with 30% ammonium hydroxide at reflux in a sealed tube to
afford 13 in high yield (93%). A hexyl derivative of 13 was
also synthesized similarly and characterized by X-ray diffrac-
tion (see Supporting Information). Intermediate 14 was obtained
via dropwise addition of methyl iodide to an aqueous solution
of 13 in the presence of NaOH at room temperature in high
yield (90%). Oxidation of methyl sulfide 14 to methyl sul-
fone 15 was achieved through two routes; by using MCPBA in
chloroform at room temperature, or by using hydrogen peroxide
in formic acid at room temperature in high yields (95%).
Compound 15 was reacted with ammonia in methanol at 100
4
9
30
both Lehn and Fenniri and of the JGC-base reported by
Mascal.
3
1,32
Furthermore, the diaminopurine-like face in JAT-3
was installed on a pyridinyl ring system (as opposed to a
pyrimidinyl ring system in JAT-2) to afford the eventual
possibility of regioselective attachment to a (deoxy)ribose
moiety on either the nitrogen of the uracil-like face or on the
carbon of the diaminopurine-like face. Finally, for the current
study we chose high yielding synthetic routes that were
amenable to extensive variation in terms of the alkyl chain that
could be installed on either target, and three alkyl derivatives
were studied in greater detail with regards to self-assembly.
°
C in a sealed tube to afford the bicyclic product 3 in good
yield.
In general terms, JAT-1, JAT-2, and JAT-3 represent three
different, yet related, shape classes of self-complementary
heterocycles whose synthetic elaboration, composition, and full
characterization have heretofore been unknown. Compound 1
was moderately soluble (∼12 mg/mL) in polar solvents (DMSO,
DMF, NMP, methanol) that are known to disrupt hydrogen
bonds whereas it was virtually insoluble in chloroform, dichlo-
romethane, benzene, toluene, nitrobenzene, THF, or water.
Compounds 2a and 2b were practically insoluble in conventional
solvents and manifested low solubility even in DMSO, DMF,
and NMP (<2 mg/mL) but readily dissolved in 98% formic
acid to an appreciable extent. As our synthetic routes allowed
us to prepare a series of alkyl derivatives, we undertook such
in the hope of finding a well-behaved monomer whose self-
association could be monitored in various NMR solvents such
Results
Synthesis of 1 (Scheme 1) began with the facile construction
of 5-cyanouracil, 6,50 that was obtained from condensation of
N-cyanoacetylurethane, 4, with triethylorthoformate in acetic
anhydride at reflux to afford 7 followed by addition of
butylamine and ring closure reaction in water in high yield
(94%). 5-Cyanouracil, 6, was then condensed with dicyandia-
mide in dry DMSO in the presence of KOH at reflux to afford
,6
1
in 80% yield.3
The procedure for synthesizing 2a and 2b (Scheme 2) started
with the condensation of N-cyanoacetylurethane, 4, with carbon
as CDCl wherein H-bonding is observable and strong. Several
3
other analogs of 1 (ethyl, propyl, phenyl, and hexadecyl) and
of 2 (ethyl, propyl, pentyl, and hexyl) were prepared similarly
(data not shown), and none was appreciably more soluble in
(
47) Kirnos, M. D.; Khudyakov, I. Y.; Alexandrushkina, N. I.; Vanyushin,
B. F. Nature 1977, 270, 369-370.
48) Haaima, G.; Hansen, H.; Christensen, L.; Dahl, O.; Nielsen, P.
Nucleic Acids Res. 1997, 25, 4639-4643.
(
(
49) Marsh, A.; Silvestri, M.; Lehn, J. Chem. Commun. 1996, 1527-
(51) Jensen, K. A.; Henrikse, L. Acta Chem. Scand. 1968, 22, 1107-
1109.
(52) Yokoyama, M.; Kumata, K.; Yamada, N.; Noro, H.; Sudo, Y.
J. Chem. Soc., Perkin Trans. 1 1988, 2309-2313.
1
1
528.
(50) Senda, S.; Hirota, K.; Notani, J. Chem. Pharm. Bull. 1972, 20,
380-1388.
468 J. Org. Chem., Vol. 72, No. 2, 2007

Janus-AT Bases
SCHEME 1. Synthesis of JAT-1
SCHEME 2. Synthesis of JAT-2
SCHEME 3. Synthesis of JAT-3
any of the solvents noted. Therefore, this work features more
in-depth studies on either butylated (1, 2a, and 3) or heptylated
is consistent with the finding that the ADA-DAD interaction
2
-1
is significantly weaker even in CHCl (K ∼ 10 M ) than the
3
a
4
-1 19
(
2b) derivatives. Compound 3 was moderately soluble (∼15
DDA-AAD configuration (K ∼ 10 M ). Although H-
a
mg/mL) in polar solvents (DMSO, DMF, NMP, methanol) that
are known to disrupt hydrogen bonds whereas it was virtually
insoluble in chloroform, dichloromethane, benzene, toluene,
nitrobenzene, THF, or water
bonding in general is considerably weaker in DMSO, the
compounds were appreciably more soluble in DMSO compared
to chloroform, and thus, their association was investigated by
1
1
H NMR. The H chemical shifts of all three shape classes did
Given the limited solubility of these compounds, we were
unable to achieve suitably high concentrations in chloroform
where H-bonds would have formed appreciably to be monitored
not vary upon dilution from ∼5 mg/mL down to 0.1 mg/mL in
d6-DMSO (see Supporting Information), and as such, we were
unable to observe solution-phase self-association in a solvent
from which H-bonded networks ultimately crystallized. Nev-
ertheless, the inability to observe self-association in either
chloroform or DMSO might also have been consistent with an
incorrect supposition that we had prepared the said targets.
5
3
by NMR. That self-association in solution was unobservable
(
53) We were unable to readily observe H-bonding by either UV-vis
1
or H NMR; consequently attempts to further assign the structure through
N- H NMR experiments (e.g., flip-back N- H 2D-NOESY) failed to
15
1
15
1
provide any meaningful information. Attempts to increase H-bond strength
in d6-DMSO by addition of CD2Cl2 or CDCl3 also failed; even at 10%
cosolvent, turbidity and ultimately precipitation prevailed before any new
peaks characteristic of intermolecular H-bonds could be reliably observed.
As all three syntheses began with creation of the uracil face
prior to installation of the DAD face containing the two
exocyclic amines, we were reasonably certain that we had indeed
J. Org. Chem, Vol. 72, No. 2, 2007 469

Asadi et al.
FIGURE 2. (A) Schematic of H-bonded ribbon array of JAT-1. (B) Cut-away X-ray diffraction of the extended unit cell showing the lattice of 2
sets of parallel arrays.
prepared the target JAT-bases as indicated. Nevertheless, neither
H NMR nor HRMS would unambiguously differentiate the
intended JAT-base from various constitutional isomers, including
two parallel repeating arrays at an angle of approximately 60°.
Two of the four arrays exhibit a dihedral angle between the
two aromatic rings of approximately 5° with relatively coplanar
H-bonds (Figure 3A) whereas the other two arrays exhibit a
significant dihedral angle of nearly 20° (Figure 3B).
1
the corresponding JGC-base. Although there was little likelihood
of a rearrangement5
4-56
that would have transformed the
intended JAT-base target into any number of constitutional
isomers, we sought additional confirmation of structure and thus
turned to X-ray diffraction to provide unambiguous structural
proof of compounds 1, 2b, and 3. Beyond absolute proof of
composition and connectivity, such structures would provide
information as to whatever H-bonded associations these com-
pounds would assume in the solid phase. One question we asked
was whether these compounds would form hexameric rosettes
as described for the JGC-bases or extended arrays characteristic
of melamine based materials.
For compound 1, a single crystal suitable for X-ray diffraction
was finally obtained by dissolving 1 in pure and very dry DMSO
at elevated temperature whereupon after several days colorless
prismatic crystals deposited after standing at room temperature.
Crystals of 1 diffracted in a P 2h 1/c space group in which the
asymmetric cell comprised four monomers and no solvent. The
asymmetric cell was propagated using Mercury version 1.4.1
to identify a repeating matrix containing four slightly different
H-bonded arrays of the type shown in Figure 2A,B.
In one of the latter two arrays, a similar distortion angle is
seen within the H-bonded faces (Figure 3C) where the three
H-bonds deviate quite substantially from coplanarity. This
structure nicely indicates the conformational flexibility of this
biaryl JAT-base as well as the extent to which the DAD-ADA
H-bonding motif itself can vary in the same crystal structure.
Similar variability in H-bonding is seen for Watson-Crick bases
pairs within duplex B-DNA and is described as roll and propeller
twist. To further demonstrate that the DAD and ADA faces
remain in the intended tautomeric states in solution, a 15N- H
HSQC NMR experiment was performed in d6-DMSO on 1. This
demonstrated that two sets of two protons correlated with two
nitrogens and not with three or four nitrogens, as would have
been the case for other tautomers (see Supporting Information).
1
Attempts at obtaining single crystals of 2a and 2b for X-ray
diffraction, in dry DMSO were unsuccessful as were numerous
attempts with a variety of mixed solvents applied in conjunction
with either vapor diffusion or refluxing at cloud point. Never-
theless, we were delighted to find that compound 2b crystallized
as colorless plates of two-component twin crystals via slow
evaporation of a dilute solution of 2b in 98% formic acid (2a
also crystallized but provided twin crystals that did not cleanly
diffract). X-ray diffraction identified a P 1h space group with a
unit cell that comprised one molecule of 2b and two molecules
Quite strikingly, the crystal packing of these arrays shows
two repeating parallel arrays that are interdigitated with another
(54) Graboyes, H.; Jaffe, G. E.; Pachter, I. J.; Rosenbloom, J. P.; Villani,
A. J.; Wilson, J. W.; Weinstock, J. J. Med. Chem. 1968, 11, 568-573.
(
(
55) Ueda, T.; Fox, J. J. J. Org. Chem. 1964, 29, 1762-1769.
56) Ueda, T.; Fox, J. J. J. Org. Chem. 1964, 29, 1770-1772.
470 J. Org. Chem., Vol. 72, No. 2, 2007

Janus-AT Bases
FIGURE 3. (A) Viewed along the array planes (blue ellipses), the dihedral angle of two arrays is roughly 5° or less. (B) The dihedral angle of the
other two arrays is roughly 15°. (C) Shown is the propeller twist of a distorted DAD-ADA H-bonding interaction in one of the four arrays.
of formic acid. Extension of the unit cell indicated a formate-
Significance and Conclusions
ion mediated cocrystal as shown in Figure 4A-C.
This work is, to the best of our knowledge, the first report of
high-resolution X-ray structural data for supramolecular solids
derived from three self-complementary monomers capable of
DAD-ADA H-bonding recognition whose compositions and
syntheses have not been known. In the case of compound 1,
four different H-bonded networks cocrystallize to exemplify the
microscopic diversity of such DAD-ADA hydrogen bonds that
assume different dihedral angles of considerable variance within
the same crystal lattice. The observation of such variability in
the crystal lattice may have implications for the design of
melamine-barbituric/cyanuric acid based structures that are
presumed to be rigid and planar but may indeed be more
polymorphic or flexible than thought.
In the case of compounds 2b and 3, no crystals could be
grown in our hands that would have suggested direct association,
and the only lattices that could be formed were obtained in the
presence of formate. For 2b we obtained a formate-bridged
lattice of an absolutely flattened array that is not actually a
DAD-ADA H-bonded lattice but rather a DDD motif interfaced
with a DA-AD interaction mediated by formate anion in which
the alkyl tails of 2b were in the same side of the formed linear
ribbon. For 3 we obtained a formate-bridged network in which
the alkyl tails of 3 were on the opposite sides of the zigzagged
ribbon. In this case, we observed a DDD-D H-bonding motif
of diaminopurine face aligned with an ADA H-bonding pattern
of thymine face mediated by AA H-bonding motif of formate
anion.
As can be seen in the X-ray crystal structure of 2b, formic
acid mediates the association by forming 3 complementary
H-bonds within the polymeric array as it (1) protonates the DAD
face, (2) accepts a proton from one of the exocylic amines, and
(3) intercepts the N3H of the thymine-like face. The formate
salt essentially masks the DAD face and provides for a DDD-
AD pairing interaction with the thymine-like face. This unique
asymmetrical hydrogen-bonding pattern between the pyrimi-
dinium cation face and the thymine-like face mediated by
formate anion, along with an AA H-bonding motif, gave rise
to formation of an infinite hydrogen-bonded molecular ribbon
lattice. In addition, a second molecule of formic acid serves to
bridge one lattice with another.
As 2b could only be crystallized from 98% formic acid (pH
∼
2), such acidic conditions obviated the possibility of observing
the expected ADA-DAD interaction. In order to verify that
b indeed retained both the ADA and DAD faces in solution,
we investigated its solution-phase tautomeric state. Thus, an
2
1
5
1
N- H HSQC NMR experiment was performed in d6-DMSO
to demonstrate that, at low concentrations, two exchangeable
protons correlated with a single nitrogen and the other correlated
with the other nitrogen. Thus, 2b also adopted the expected
ADA and DAD tautomeric faces (see Supporting Information).
To further demonstrate self-association in the absence of formic
acid, an ESI experiment (positive ion mode) was run on
compounds 2a and 2b. Analysis of 2a gave evidence of several
In no case did we observe rosette-like structures for any of
the heterocycles described herein. Taken together, these obser-
vations are consistent with the generally accepted relative
weakness of the DAD-ADA H-bonding scheme compared to
the isosteric DDA-AAD scheme. It is perhaps not surprising
that a crystal structure of a DAD-ADA H-bonded supramo-
lecular solid has remained elusive until now.
+
multimeric sodiated species, (Ma + Na) where a ) 1-6, as
well as putative (Mb + 2Na)2+ where b ) 7, 9, and 11 (see
Supporting Information). ESI analysis of 2b gave several
+
multimeric protonated species (Mc + H) where c ) 4-6, as
well as putative (Md + 2H)2+ where d ) 8-11.
Likewise, 3 crystallized as colorless plate crystal via slow
evaporation of a dilute solution of 3 in 98% formic acid in the
presence of methylene chloride. On standing, compound 3
crystallized as formate salt in which the DAD face provides
DDD-AD pairing interaction with the thymine-like face due
to the protonation of the nitrogen in diaminopyridine ring.
Similar to the extended H-bonding network of 2b, this unique
asymmetrical hydrogen-bonding pattern between the pyridinium
cation face and the thymine-like face mediated by formate anion,
afforded formation of a zigzagged hydrogen-bonded molecular
ribbon association. Extension of the unit cell indicated a formate-
ion mediated cocrystal as shown in Figure 5A,B.
Not only are these the first crystal structures of compounds
with both ADA and DAD faces, but also these structures are
featured herein as absolute proof of the constitutional con-
nectivity of three unique JAT-base compositions which have not
been previously reported in any form. Such characterization is
important if one is to contemplate their future use in developing
oligomers and bioconjugates capable of DNA and RNA duplex
recognition. In that regard, whereas 1 crystallized nicely with
the expected H-bonded tautomers characteristic of A-T base
pairs, 2b and 3 could only be crystallized from 98% formic
acid (pH ∼ 2) as formate salts. Nevertheless, it is unlikely that
at physiological pH (7.4) compounds 2b or 3 will remain
J. Org. Chem, Vol. 72, No. 2, 2007 471

Asadi et al.
FIGURE 4. (A) Scheme of formate anion bridging the array. (B) X-ray single-crystal structure of supramolecular ribbon of 2b. (C) The extended
structure with interlattice formate bridges.
protonated as formate salts (pKa’s of diaminopyrmindines,
diaminopyridines ∼5.5) and thus should favor the intended
H-bonding recognition patterns as seen for compound 1, and
as confirmed independently by solution phase 15H- H HSQC
NMR (2b) and by ESI (2a and 2b). On the basis of the similarity
or (ii) difficulties in preparing soluble oligomers of JGC that
will have the suitable length and consequent energetics to permit
stable recognition of stretches of nucleic acid sequences.
If such a strategy is to be fully realized, it will necessar-
ily require the synthesis and definitive structural proof of a
JAT-base (e.g., 2b, 3), as reported herein, that is isosteric for,
and which may be used in conjunction with previously reported
JGC-bases. The robust syntheses of 2a, 2b, and 3 extend the
repertoire for using this shape-class of J-bases in the context of
an oligomeric scaffold (e.g., PNA, ODN) to potentially recog-
nize any nucleic acid sequence. Although to date there is no
analogous JGC-base corresponding to 1, such is also readily
envisioned to provide for a different shape-class of JAT and JGC
bases. The DNA recognition properties of the heterocycles
reported herein await full investigation, along with the same
for JGC heterocycles.
1
1
between the chemical shifts of 2b and 3 in the H NMR spectra,
along with the structural proof of 3 afforded by X-ray crystal-
lography, we did not further investigate 3 by 15H- H HSQC
NMR.
1
The potential for using J-bases to recognize both Watson-
Crick bases pairs (GC and AT), and thus any DNA sequence,
is an attractive idea that has yet to be fully elaborated. Curiously,
despite several previous reports of JGC-bases, the potential for
GC base pair recognition has never been investigated. This may
be due to (i) an inability to cleanly monitor simultaneous
recognition of monomeric G and monomeric C in CDCl3 against
a backdrop of both the self-association of the JGC-base and an
independent association of monomeric G with monomeric C,
5
7
(57) Dervan, P. B. Bioorg. Med. Chem. 2001, 9, 2215-2235.
472 J. Org. Chem., Vol. 72, No. 2, 2007

Janus-AT Bases
FIGURE 5. (A) Scheme of formate anion bridging the hydrogen bonded array of 3. (B) X-ray single-crystal structure of supramolecular crinkled
ribbon of 3 with formate bridging.
Although alkylated monomers such as 1, 2b, and 3 are rather
insoluble in water and thus unlikely to find immediate applica-
tion for DNA or RNA recognition, solubility itself is a
complicated issue and the lack thereof in water can be used
advantageously to enforce H-bonds in an aqueous medium. For
example, both guanine and guanosine are very insoluble in
water, yet when properly oligomerized on a water-soluble
backbone (e.g., deoxyribophosphodiester), find use in DNA
recognition and nanostructured architectures.58 The synthetic
routes we report herein should enable a diverse functionalization
of JAT1-3 for use in bioconjugation and oligomerization.
Although at this juncture we have not reported on the synthesis
of a derivative that can be oligomerized, such is easily
contemplated by using suitably protected (e.g., NRBoc-COOtBu)
lysine instead of an alkylamine.
rosette structure on an internal JAT base that clearly has
crystallized out as a tape.
The syntheses presented herein will enable us, and others, to
evaluate the properties of self-association, nanostructured as-
sembly, and DNA and RNA recognition. As AT-rich sequences
are found at critical regions of both ribozymes and DNA
promoters, properly functionalized JAT-bases that recognize such
sequences may prove useful for studying the biological processes
governed by these sequences. Such recognition will be reported
in due time.
Experimental Section
General experimental procedures and methods are given in the
Supporting Information.
(Z)-2-Cyano-3-ethoxyacrylethyl carbamate (5). A stirring
mixture of N-cyanoacetylurethane, 4 (1.56 g, 10 mmol), triethyl
orthoformate (1.64 mL, 10 mmol), and acetic anhydride (4 mL)
was refluxed at 110 °C for 45 min. The reaction mixture was then
allowed to cool to room temperature. The colorless crystals of
product were filtered and washed with light petroleum ether (20
In terms of targeting DNA, J-base target recognition may be
43
complicated by self-association. Nonetheless, the formation
of hexameric J-base rosettes will be a fifth or possibly sixth
order process whereas duplex DNA recognition will be first or
second order. As such, low concentrations may be identified to
kinetically disfavor higher order self-associated structures and
favor duplex formation. That notwithstanding, self-association
by ligands designed to target DNA and RNA is neither a new
phenomenon nor a reason not to contemplate their use: G-rich
oligonucleotides offer high Tm’s with respect to target recogni-
tion, yet form self-associated G-quartets in competition with
target strand recognition.
1
mL) and diethyl ether (5 mL) (yield 95%). H NMR (300 MHz,
CDCl
3
, 25 °C) δ ) 1.24 (t, J ) 7.14 Hz, 3H, CH
3
), 1.38 (t, J )
7.14 Hz, 3H, CH ), 4.17 (q, J ) 7.14 Hz, 2H, CH ), 4.36 (q, J )
7.14 Hz, 2H, CH ), 7.91 (br s, 1H, NH), 8.17 (s, 1H, CH).
3
2
1
3
2
C
NMR (300 MHz, CDCl
3
) δ 14.2, 15.3, 62.5, 74.7, 81.5, 87.0, 104.7,
+
1
13.3, 150.1, 159.2, 161.3, 174.2. ESI-MS (m/z) 235 (M + Na) .
+
HRMS (ESI ) calcd for C
9 12 2 4
H N O Na 235.0695, found 235.0697.
1
-Butyl-1,2,3,4-tetrahydro-2,4-dioxopyrimidine-5-carboni-
trile (6). To a stirring solution of 5 (2.12 g, 10mmol) in water (10
mL) was added butyl amine (0.99 mL, 10mmol), and the reaction
mixture was heated at 90 °C for 25 min. The reaction mixture was
then allowed to cool to room temperature and acidified by 5 N
HCl. The colorless crystals were filtered and washed with cold water
(10 mL). (Colorless crystals recrystallized from water, yield 93%.)
Obviously, the other edge of this sword is self-association,
which underlies the design of nanostructured assemblies unre-
lated to DNA recognition. Hexameric rosettes offer interesting
properties unto themselves as underscored in a recent report of
a hexameric rosette comprising three monomeric deoxyribo-
sylbarbituric acids and three monomeric deoxyribosyl-triami-
nopyrimidines.59 It will be interesting to see if two JGC bases
which sandwich a JAT base (e.g., JGC-JAT-JGC) can enforce a
1
H NMR (300 MHz, d -DMSO, 25 °C) δ ) 0.87 (t, J ) 7.36 Hz,
6
3H, CH
3
), 1.22-1.30 (m, 2H, CH
2
2
), 1.55-1.60 (m, 2H, CH ), 3.70
(
t, J ) 7.32 Hz, 2H, CH
2
), 8.69 (s, 1H, CH), 11.94 (br s, 1H, NH).
-DMSO, 25 °C) δ ) 13.5, 19.0, 30.3, 48.6,
1
3
C NMR (300 MHz, d
6
8
7.3, 113.9, 114.4, 149.6, 154.7, 160.6. ESI-MS (m/z) 216 (M +
(
58) Seeman, N. Acc. Chem. Res. 1997, 30, 357-363.
+
+
(59) Rakotondradany, F.; Palmer, A.; Toader, V.; Chen, B.; Whitehead,
Na) . HRMS (ESI ) calcd for C
216.0751.
9 11 3 2
H N O Na 216.0749, found
M.; Sleiman, H. Chem. Commun. 2005, 5441-5443.
J. Org. Chem, Vol. 72, No. 2, 2007 473

Asadi et al.
5
-(4,6-Diamino-1,3,5-triazin-2-yl)-1-butylpyrimidine-2,4(1H,-
tylamine (0.74 mL, 5 mmol) in 10 mL absolute ethanol was added
dropwise over 30 min. The flask was fitted with a reflux condenser,
and the reaction mixture was heated under reflux at 100 °C for 16
h. The reaction mixture was then allowed to cool to room
temperature and the solvent evaporated under reduced pressure with
3
H)-dione (1). A stirring mixture of cyanouracil, 6 (0.97 g, 5
mmol), dicyandiamide (0.46 g, 5.5 mmol), and powdered potassium
hydroxide (0.61 g, 11 mmol) in dry DMSO (5 mL) was heated at
1
00 °C for 5 h. The reaction mixture was then allowed to cool to
room temperature, and water (20 mL) was added. The formed
solution was acidified by 5 N HCl, and the suspended solid product
was filtered and rinsed with water (25 mL) and diethylether (50
a rotary evaporator. The resulting white powder was recrystallized
1
in methanol-water (1:1) (yield 87%). H NMR (300 MHz, d
6
-
2
),
DMSO) δ ) 0.85 (t, J ) 7.3 Hz, 3H, CH
1.6-1.8 (m, 2H, CH ), 2.8 (s, 3H, S-CH
N-CH ), 8.5 (br s, 1H, NH). C NMR (300 MHz, d
) 14.1, 18.1, 19.26, 22.9, 25.0, 27.9, 28.8, 31.8, 45.7, 92.0, 114.4,
3
), 1.1-1.4 (m, 8H, CH
), 4 (t, J ) 7.8 Hz, 2H,
-DMSO) δ
1
mL) to afford pure product 1 (white powder, yield, 80%). H NMR
2
3
13
(
300 MHz, d
6
-DMSO) δ ) 0.89 (t, J ) 7.41 Hz, 3H, CH
.32 (m, 2H, CH ), 1.55-1.60 (m, 2H, CH ), 3.73 (t, J ) 7.3 Hz,
H, N-CH ), 6.69 (br s, 4H, NH ), 8.15 (s, 1H, CH), 11.98 (br s,
H, NH). 13C NMR (400 MHz, d
-DMSO, 25 °C,) δ ) 13.6, 19.1,
0.7, 47.6, 139.7, 147.3, 150.4, 160.6, 166.9, 168.3. HRMS (ESI )
278.1365, found 278.1366.
3
), 1.24-
2
6
1
2
1
3
2
2
+
+
2
2
148.5, 159.9, 167.5. ESI-MS (m/z) 281 (M ). HRMS (ESI ) m/z
calcd for C13 S 281.1198, found 281.1202.
6
19 3 2
H N O
+
5,7-Diamino-1-butylpyrimido [4,5-d] Pyrimidine-2, 4(1H,3H)-
dione (2a). A mixture of guanidinium hydrochloride (0.52 g, 5.5
mmol) and sodium ethoxide (0.37 g, 5.5 mmol) in 5 mL absolute
ethanol was heated at 45 °C for 5 min, and a precipitate formed
(NaCl) from which the solvent and guanidine free base was filtered
off. The filtrate was added to a solution of 9a (1.2 g, 5 mmol) in
absolute ethanol (20 mL). The reaction mixture was heated under
reflux at 100 °C for 16 h. The reaction mixture was then allowed
to cool to room temperature, and a white precipitate formed, which
was filtered and washed with water (2 × 25 mL). The product was
16 7 2
calcd for C11H N O
Potassium 2-Cyano-3-(ethoxycarbonylamino)-3-oxoprop-1-
ene-1,1-bis-thiolate (7). Compound 4 (1.56 g, 10 mmol) and
powdered potassium carbonate (1.38 g, 10 mmol) in dry DMF (25
mL) was vigorously stirred at room temperature for 2 h. Carbon
disulfide (1.3 mL, 20 mmol) was added all at once to the suspension,
and stirring continued for 4 h. Absolute ethanol (100 mL) was added
to the mixture, and the precipitate that formed was filtered and
washed with diethyl ether (150 mL), then dried under reduced
1
1
pressure (pale yellow powder, yield 95%). H NMR (400 MHz,
then crystallized from formic acid (yield 87%). H NMR (400 MHz,
d
6
-DMSO, 25 °C, TMS) δ ) 14.9 (s, 1H; NH), 1.1 (t, J ) 13 Hz,
d
6
-DMSO) δ 0.88 (t, J ) 7.5 Hz, 3H, CH
CH ), 1.49-1.56 (m, 2H, CH ), 3.94 (t, J ) 7.28 Hz, 2H, N-CH
6.86 (br s, 2H, NH ), 7.40 (br s, 1H, NH), 8.01 (br s, 1H, NH),
11.06 (br s, 1H, NH). C NMR (600 MHz, d
19.5, 29.6, 48.6, 82.9, 150.4, 159.4, 162.3, 163.6, 163.7. ESI-MS
3
), 1.22-1.31(m, 2H,
13
3H; CH
3
), 4 (q, J ) 7 Hz, 2H; CH
2
). C NMR (400 MHz, d
6
-
2
2
2
),
DMSO, 25 °C, TMS) δ ) 14.5, 59.5, 97.5, 125.6, 152.4, 164.9,
2
+
13
2
22.1. ESI-MS (m/z) 308 (M ). HRMS (ESI) m/z calcd for
308.9172, found 308.9171.
-Cyano-3,3-bis (methylthio)acrylethyl carbamate (8). A
solution of 7 (3.08 g, 10 mmol) in a water-acetonitrile (40 mL,
:3) mixture was stirred at room temperature. Methyl iodide (1.3
6
-DMSO) δ ) 13.8,
7 7 2 3 2 2
C H N O S K
+
2
15 6 2
(m/z) 251 (M + H) . HRMS (EI) m/z calcd for C10H N O
251.1256, found 251.1258.
7
5,7-Diamino-1-heptylpyrimido[4,5-d]pyrimidine-2,4(1H,3H)-
dione (2b). A mixture of guanidinium hydrochloride (0.52 g, 5.5
mmol) and sodium ethoxide (0.37 g, 5.5 mmol) in 5 mL absolute
ethanol was heated at 45 C for 5 min, and a precipitate formed
(NaCl) which was filtered off. The filtrate was added to a solution
of 9b (1.4 g, 5 mmol) in absolute ethanol (20 mL). The reaction
mixture was heated under reflux at 100 °C for 16 h. The reaction
mixture was then allowed to cool to room temperature, and a white
precipitate formed, which was filtered and washed with water (2
mL, 22 mmol) in 10 mL acetonitrile was added dropwise to the
solution, and the mixture was stirred for 0.5 h at room temperature.
The flask was fitted with a reflux condenser, and the reaction
mixture was heated under reflux at 95 °C for 3 h. The reaction
mixture was then allowed to cool to room temperature and
concentrated to 25 mL under reduced pressure with a rotary
evaporator. The product was extracted with ethyl acetate (3 × 50
mL). Combined organic phases were washed with brine (100 mL),
dried over sodium sulfate, and concentrated on a rotary evaporator
to give a viscous oil (pale yellow), which solidified on standing at
× 25 mL). The product was then crystallized from formic acid
1
(yield 85%). H NMR (300 MHz, d
6
-DMSO, 25 °C, TMS) δ )
1
room temperature under reduced pressure (yield 85%). H NMR
11.1 (s, 1H, NH), 8.1(s, 1H, NH), 7.4 (s, 1H, NH), 6.9 (s, 2H,
(
300 MHz, CDCl
CH ), 2.62 (s, 3H, S-CH
Hz, 2H, CH ), 7.95 (br s, 1H, NH). C NMR (300 MHz, CDCl3,
5 °C, TMS) δ ) 14.40, 19.6, 21.1, 29.9, 31.1, 62.8, 98.8, 116.9,
3
, 25 °C, TMS) δ ) 1.27 (t, J ) 7.1 Hz, 3H,
NH
1.4(m, 8H, CH
MHz, d
2
), 3.9 (t, J ) 7.28 Hz, 2H, CH
2
), 1.5-1.6 (m, 2H, CH
2
), 1.2-
13
3
3
), 2.8 (s, 3H, S-CH ), 4.24 (q, J ) 7.08
3
2
), 0.8 (t, J ) 6.8 Hz, 3H, CH
3
). C NMR (400
13
2
6
-DMSO) δ ) 14.3, 22.2, 26.5, 27.6, 28.9, 31.6, 83.1, 151.8,
+
2
1
159.8, 162.8, 163.4, 164.2. EI-MS (m/z) 293 (M + H) . HRMS
(EI) m/z calcd for C13 293.1647, found 293.1656.
+
31.3, 150.6, 159.2, 183.0, 188.6. EI-MS (m/z) 260 (M ). HRMS
260.02894, found 260.02905.
-Butyl-1,2,3,4-tetrahydro-6-(methylthio)-2,4 dioxopyrimidine-
-carbonitrile (9a). A solution of 8 (1.3 g, 5 mmol) in absolute
21 6 2
H N O
(EI) m/z calcd for C
H
9 12
N O S
2 3 2
(2-Cyano-3-ethoxy-but-2-enoyl)carbamic Acid Ethyl Ester
(10). A stirring mixture of Compound 4 (1.56 g, 10 mmol), triethyl
orthoacetate (1.82 mL, 10 mmol), and acetic anhydride (4 mL) was
refluxed at 110 °C for 40 min. The reaction mixture was then
allowed to cool to room temperature. The colorless crystals of
product were filtered and washed with light petroleum ether (20
1
5
ethanol (30 mL) was stirred at room temperature. Butylamine (0.49
mL, 5 mmol) in 10 mL absolute ethanol was added dropwise over
a period of 30 min. The flask was fitted with a reflux condenser,
and the reaction mixture was heated under reflux at 100 °C for 16
h. The reaction mixture was then allowed to cool to room
temperature and the solvent evaporated under reduced pressure with
a rotary evaporator. The resulting white powder was recrystallized
in ethanol-water (1:1) (yield 83%) (ORTEP diagram and X-ray
1
mL) and cold diethyl ether (5 mL) (yield 90%). H NMR (300
MHz, CDCl
J ) 7.3 Hz, 3H, CH
CH ), 4.39 (q, J ) 7.3 Hz, 2H, CH
NMR (300 MHz, CDCl ) δ 15.2, 15.6, 19.9, 62.5, 69.9, 93.5, 118.0,
152.3, 160.1, 178.2. ESI-MS (m/z) 235 (M + Na) . HRMS (ESI )
calcd for C Na 235.0695, found 235.0697.
3
, 25 °C) δ ) 1.27 (t, J ) 7.3 Hz, 3H, CH
), 2.49 (s, 3H, CH ), 4.2 (q, J ) 7.3 Hz, 2H,
), 9.14 (br s, 1H, NH).
3
), 1.53 (t,
3
3
1
3
2
2
C
3
1
+
+
crystal structure data in the Supporting Information). H NMR (300
MHz, d
6
-DMSO) δ ) 0.95 (t, J ) 7.3 Hz, 3H, CH
m, 2H, CH ), 1.58-1.66 (m, 2H, CH ), 2.92 (s, 3H, S-CH
t, J ) 7.8 Hz, 2H, N-CH ), 8.6 (br s, 1H, NH). C NMR (300
-DMSO) δ ) 13.5, 19.2, 19.3, 30.3, 46.3, 92.0, 114.6,
3
), 1.33-1.4
9 12 2 4
H N O
(
(
2
2
3
), 4.13
1-Butyl-6-methyl-2,4-dioxo-1,2,3,4-tetrahydro-pyrimidine-5-
carbonitrile (11). To a stirring solution of 10 (2.26 g, 10mmol) in
water (10 mL) was added butyl amine (0.99 mL, 10mmol), and
the reaction mixture was heated at 90 °C for 20 min. The reaction
mixture was then allowed to cool to room temperature and acidified
to pH 5 by the addition of 5 N HCl. The colorless crystals were
filtered and washed with cold water (10 mL) and cold diethyl ether
(10 mL). Colorless crystals were recrystallized from water-
13
2
MHz, d
49.1, 159.4, 166.3. ESI-MS (m/z) 262 (M + Na) . HRMS (ESI )
m/z calcd for C10 NaS 262.0626, found 262.0625.
-Heptyl-6-(methylthio)-2,4-dioxo-1,2,3,4-tetrahydropyrimi-
6
+
+
1
13 3 2
H N O
1
dine-5-carbonitrile (9b). A solution of 8 (1.3 g, 5 mmol) in
absolute ethanol (30 mL) was stirred at room temperature. Hep-
474 J. Org. Chem., Vol. 72, No. 2, 2007

Janus-AT Bases
1
1H, CH), 8.35 (br s, 1H, NH), 8.54 (br s, 1H, NH). 13C NMR (400
MHz, d -DMSO, 25 °C) δ ) 14.0, 15.1, 20.8, 30.2, 43.3, 91.4,
methanol (1.87 g, yield 91%). H NMR (400 MHz, d
C) δ ) 0.99 (t, J ) 7.6 Hz, 3H, CH ), 1.36-1.45 (m, 2H, CH
), 2.61 (s, 3H, CH ), 3.9 (t, J ) 7.6 Hz,
), 9.92 (br s, 1H, NH). 13C NMR (400 MHz, d
-DMSO,
5 °C) δ ) 13.8, 19.6, 20.1, 30.7, 46.1, 90.6, 113.9, 149.6, 159.8,
6
-DMSO, 25
°
3
2
),
6
1
2
2
1
.61-1.70 (m, 2H, CH
2
3
94.5, 150.1, 151.5, 160.5, 164.7, 167.1. ESI-MS (m/z) 281 (M +
+
+
H, CH
2
6
H) . HRMS (ESI) calcd for C12
17 4 2
H N O S 281.1072, found
281.1071.
+
+
63.5. ESI-MS (m/z) 208 (M + H) . HRMS (ESI ) calcd for
208.1086, found 208.1088.
-Amino-1-butyl-7-thioxo-1,7-dihydro-thiopyrano[4,3-d]py-
5-Amino-1-butyl-7-methanesulfonyl-1H-pyrido[4,3-d]pyrimi-
10 14 3 2
C H N O
5
dine-2,4-dione (15). Method A. Compound 14 (5 mmol, 1.4 g)
was dissolved in formic acid 88% (10 mL), and hydrogen peroxide
rimidine-2,4-dione (12). To a stirring solution of 11 (2.07 g, 10
mmol) in dry THF (100 mL) was added 35 mmol of potassium
t-butoxide. The reaction mixture was stirred for 10 min at room
temperature followed by dropwise addition of carbon disulfide (25
mmol, 1.6 mL) in THF (10 mL) over period of 10 min. The reaction
mixture was stirred vigorously at room temperature until full
conversion of starting material (3 h). Upon completion, the solvent
was evaporated under reduced pressure. The crude product was
dissolved in water (30 mL), and the solution was acidified to pH
3
0% (15 mmol, 1.7 mL) was added dropwise to the solution at 0
°
C over 5 min. The reaction mixture was stirred for 5 h at room
temperature. Upon completion, the solvent was evaporated under
reduced pressure. The crude product was washed with water (10
mL) and diethyl ether (5 mL) to afford 1.48 g (95%) of a pale
yellow solid.
Method B. Compound 14 (5 mmol, 1.4 g) was dissolved in
chloroform (50 mL) along with 3-chloroperoxybenzoic acid 77%
(15 mmol, 2.5 g). The reaction mixture was stirred for 4 h at room
5
using glacial acetic acid. The orange precipitate was filtered off
and washed with petroleum ether (10 mL) and cold diethyl ether
10 mL). Further purification was obtained by reprecipitating in
methanol (2.52 g, yield 89%). Single orange crystals for X-ray
temperature. Upon completion, the solvent was evaporated under
reduced pressure. The crude product was washed with diethyl ether
(
1
(50 mL) to afford 1.2 g (81%) of a pale yellow solid. H NMR
(400 MHz, d
6
-DMSO, 25 °C) δ ) 0.92 (t, J ) 7.6 Hz, 3H, CH
), 1.52-1.59 (m, 2H, CH ), 3.24 (s, 3H,
), 3.97 (t, J ) 7.6 Hz, 2H, CH ), 6.9 (s, 1H, CH), 7.97 (br s,
3
),
1
6
analysis were obtained from acetonitrile. H NMR (300 MHz, d -
1
CH
.3-1.39 (m, 2H, CH
2
2
DMSO, 25 °C) δ ) 0.94 (t, J ) 7.5 Hz, 3H, CH
3
), 1.3-1.38 (m,
), 3.94 (t, J ) 7.5 Hz, 2H, CH ),
.73 (s, 1H, CH), 9.84 (br s, 1H, NH), 10.57 (br s, 1H, NH), 11.65
-DMSO, 25 °C) δ ) 15.1,
0.8, 30.3, 44.5, 92.9, 111.3, 151.0, 151.3, 163.9, 173.5, 188.7.
3
2
2H, CH
2
), 1.48-1.56 (m, 2H, CH
2
2
13
1
(
9
H, NH), 8.41 (br s, 1H, NH), 11.85 (br s, 1H, NH). C NMR
400 MHz, d -DMSO, 25 °C) δ ) 15.0, 20.7, 30.3, 40.7, 95.3,
6.7, 151.2, 152.4, 161.3, 162.2, 164.5. EI-MS (m/z) 312 (M).
S 312.08923, found 312.08917.
,7-Diamino-1-butyl-1H-pyrido[4,3-d]pyrimidine-2,4-dione (3).
6
(
6
br s, 1H, NH). 13C NMR (300 MHz, d
6
2
16 4 4
HRMS (EI) calcd for C12H N O
-
-
ESI-MS (m/z) 282 (M) . HRMS (ESI ) calcd for C11
82.0371, found 282.0374.
-Amino-1-butyl-7-thioxo-6,7-dihydro-1H-pyrido[4,3-d]pyri-
12 3 2 2
H N O S
5
2
Compound 15 (2 mmol, 0.62 g) was suspended in ammonia in
methanol (7 N, 10 mL), and the mixture was placed in a 150 mL
Teflon-sealed thick-walled glass tube equipped with a magnetic
stirrer and heated while stirring to 100 °C (24 h) (Caution ! use
blast shield). The mixture was cooled to room temperature, and
the excess ammonia was evaporated under reduced pressure. The
precipitate was washed with water, methanol, and diethyl ether to
afford 0.45 g (92%) of a white solid. 1H NMR (400 MHz,
5
midine-2,4-dione (13). Compound 12 (0.89 g, 3 mmol) and 30%
ammonium hydroxide (30 mL) were placed in a 150 mL Teflon-
sealed thick-walled glass tube equipped with a magnetic stirrer and
was heated while stirring to 100 °C (12 h) (Caution! use blast
shield). The mixture was cooled to room temperature, and the excess
ammonia was evaporated under reduced pressure. The formed tan
precipitate was filtered off and washed with water (10 mL) and
cold diethyl ether (20 mL) (0.74 g, yield 93%, tan powder). A single
d
6
-DMSO, 25 °C) δ ) 0.93 (t, J ) 7.6 Hz, 3H, CH
m, 2H, CH ), 1.48-1.58 (m, 2H, CH ), 3.75 (t, J ) 7.6 Hz, 2H,
CH ), 5.47 (s, 1H, CH), 6.45 (br s, 2H, NH ), 6.72 (br s, 1H, NH),
.09 (br s, 1H, NH), 10.79 (br s, 1H, NH). 13C NMR (400 MHz,
-DMSO, 25 °C) δ ) 13.7, 19.5, 28.6, 41.7, 79.9, 84.8, 149.2,
3
), 1.29-1.38
(
2
2
crystal for X-ray crystallography analysis was obtained by using
1
2
2
9
8% formic acid as crystallization solvent. H NMR (300 MHz,
8
d
1
d
6
-DMSO, 25 °C) δ ) 0.9 (t, J ) 7.26 Hz, 3H, CH
), 1.43-1.61 (m, 2H, CH ), 3.82 (t, J ) 7.32 Hz, 2H,
), 6.23 (s, 1H, CH), 7.1 (br s, 1H, NH), 8.35 (br s, 1H, NH),
1.33 (br s, 1H, NH), 12 (br s, 1H, NH). 13C NMR (300 MHz,
3
), 1.25-1.33
6
(m, 2H, CH
2
2
-
-
50.3, 160.7, 162.5, 162.6. ESI-MS (m/z) 248 (M) . HRMS (ESI)
248.1147, found 248.1149.
CH
2
14 5 2
calcd for C11H N O
1
d
1
6
-DMSO, 25 °C) δ ) 14.0, 15.12, 20.8, 30.3, 43.3, 91.4, 94.5,
+
Acknowledgment. This work was supported by NSERC.
D.M.P. is the recipient of a Junior Career Scholar Award of
the Michael Smith Foundation for Health Research in British
Columbia, and A.A. is the recipient of a Gladys Estella Laird
predoctoral fellowship. We thank Dr. Nick Burlinson for
instruction on N NMR experiments, Dr. Marcel Hollenstein
for Tm Analysis of the Interaction of DNA with 2b, and
Professor Marco A. Ciufolini for critical reading of this
manuscript.
50.1, 151.5, 160.5, 164.7, 167.1. ESI-MS (m/z) 267 (M + H)
.
+
HRMS (ESI ) calcd for C11
H N
15 4
O
2
S 267.0916, found 267.0914.
5-Amino-1-butyl-7-methylsulfonyl-1H-pyrido[4,3-d]pyrimidine-
2,4-dione (14). Compound 13 (2.66 g, 10 mmol) was added to an
aqueous solution of NaOH (2.4%, 50 mL). The reaction mixture
was stirred for 5 min at room temperature followed by dropwise
addition of methyl iodide (10 mmol, 0.62 mL) over 3 min. The
reaction mixture was stirred at room temperature (1.5 h); upon
completion of reaction, the mixture was acidified to pH 6 by dilute
HCl and the formed white precipitate filtered off and washed with
water (20 mL). Further purification was obtained by reprecipitating
1
5
Supporting Information Available: H and 13C NMR spectra,
1
1
in methanol (2.65 g, yield 90%), forming white powder. H NMR
1
15
H- N HSQC data, and CIF files of the X-ray structures. This
(
1
CH
400 MHz, CDCl
.39-1.48 (m, 2H, CH
3
, 25 °C) δ ) 1.01 (t, J ) 7.6 Hz, 3H, CH
), 1.62-1.70 (m, 2H, CH ), 2.54 (s, 3H,
), 3.93 (t, J ) 7.6 Hz, 2H, CH ), 5.52 (br s, 1H, NH), 6.17 (s,
3
),
materialisavailablefreeofchargeviatheInternetathttp://pubs.acs.org.
2
2
3
2
JO061840K
J. Org. Chem, Vol. 72, No. 2, 2007 475