Is the anomeric effect an important factor in the rate of adenosine deaminase catalyzed hydrolysis of purine nucleosides? A direct comparison of nucleoside analogues constructed on ribose and carbocyclic templates with equivalent heterocyclic bases selected to promote hydration.

Article abstract of DOI:10.1016/S0968-0896(02)00099-8

The aglycone of (North)-methanocarbadeoxyadenosine [(N)-MCdA, (5)], a relatively weak substrate for adenosine deaminase (ADA)-relative rate of deamination ca. 100 times lower than adenosine-was modified with substitutions at positions 6 (6-fluoro, compound 6) and 8 (8-aza, compound 7) with the intent to improve the level of hydration and hence hydrolysis by ADA. In these substrates the fused cyclopropane moiety constrains the cyclopentane ring to mimic the conformation of a furanose sugar in the North hemisphere of the pseudorotational cycle, which matches the conformation of the ribose ring of adenosine in complex with ADA. The order of susceptibility to ADA hydrolysis was adenosine>>(N)-MCdA (5) approximately equal to(N)-6F-MCdP (6)>(N)-8-aza-MCdA (7). Despite the known fact that 8-azaadenosine is hydrolyzed twice as fast as adenosine, the corresponding carbocyclic analogue 7 was hydrolyzed at approximately half the rate of the parent 5. These results argue in favor of the critical role of the O(4') oxygen atom and its associated anomeric effect in assisting hydrolysis by ADA.

Full text of DOI:10.1016/S0968-0896(02)00099-8

Bioorganic & Medicinal Chemistry 10 (2002) 2723–2730
Is the Anomeric Effect an Important Factor in the Rate of
Adenosine Deaminase Catalyzed Hydrolysis of Purine
Nucleosides? A Direct Comparison of Nucleoside Analogues
Constructed on Ribose and Carbocyclic Templates with
Equivalent Heterocyclic Bases Selected to Promote Hydration
Susana Hernandez, Harry Ford, Jr. and Victor E. Marquez*
Laboratory of Medicinal Chemistry, Center for Cancer Research, National Cancer Institute at Frederick,
376 Boyles St., Frederick, MD, 21701, USA
Received 5 December 2001; accepted 14 February 2002
Abstract—The aglycone of (North)-methanocarbadeoxyadenosine [(N)-MCdA, (5)], a relatively weak substrate for adenosine deami-
nase (ADA)—relative rate of deamination ca. 100 times lower than adenosine—was modified with substitutions at positions 6 (6-
fluoro, compound 6) and 8 (8-aza, compound 7) with the intent to improve the level of hydration and hence hydrolysis by ADA. In
these substrates the fused cyclopropane moiety constrains the cyclopentane ring to mimic the conformation of a furanose sugar in
the North hemisphere of the pseudorotational cycle, which matches the conformation of the ribose ring of adenosine in complex
with ADA. The order of susceptibility to ADA hydrolysis was adenosine> >(N)-MCdA (5)ꢀ(N)-6F-MCdP (6)>(N)-8-aza-MCdA
(7). Despite the known fact that 8-azaadenosine is hydrolyzed twice as fast as adenosine, the corresponding carbocyclic analogue 7
was hydrolyzed at approximately half the rate of the parent 5. These results argue in favor of the critical role of the O(40) oxygen
atom and its associated anomeric effect in assisting hydrolysis by ADA. # 2002 Elsevier Science Ltd. All rights reserved.
Introduction
continued search for more effective and less toxic inhi-
bitors of ADA. An alternative way to achieve a similar
goal is to employ effective adenosine analogues that are
resistant to ADA deamination. Along these lines, our
laboratory has recently reported the synthesis of two
classes of conformationally locked North and South
carbocyclic 20-deoxyadenosine analogues [(N)-MCdA (5)
and (S)-MCdA (8)] that are, respectively, poorly dea-
minated or almost impervious to the action of ADA.7
Remarkably, these compounds can be recognized by
various adenosine receptors as bona fide adenosine ana-
logues, sometimes even displaying higher binding affinity
than ribosides themselves.8,9 Relative to their resistance to
deamination, our initial observations suggest that replac-
ing the ribose or deoxyribose with a carbocyclic moiety
significantly lessens the rate of deamination, thus point-
ing to the important role of the O(40) oxygen in the
mechanism of enzymatic deamination by ADA.7 The
role of this oxygen appears to be strictly mechanistic
since ADA crystal structures complexed with a transi-
tion-state inhibitor (20-deoxycoformycin) or a reaction
coordinate analogue [(6S)-hydroxyl-1,6-dihydropurine
Adenosine deaminase (ADA, EC 3.5.4.4) is a critical
enzyme of the purine metabolic pathway that catalyzes
the irreversible hydrolytic deamination of adenosine
(Ado, 1) and 20-deoxyadenosine (dAdo, 2)—as well as
other exogenous substrates—to their hypoxanthine
derivatives 3 and 4, and ammonia (Scheme 1).1,2 The
importance of ADA in maintaining purine nucleotide
pools is evident in several hereditary conditions asso-
ciated with either overproduction or deficiency of the
enzyme. Excess of ADA function has been associated
with hemolytic anemia,3 while ADA deficiency results in
impaired B- and T-cell-based immunity.4 The complete
lack of T lymphoblasts characterized by ADA deficiency
provided the rationale for the use of ADA inhibitors in
the treatment of T-lymphoblastic leukemias.5,6 How-
ever, the limited success of this approach justifies the
*Corresponding author. Tel.: +1-301-496-8065; fax: +1-301-402-
2275; e-mail: marquezv@dc37a.nci.nih.gov
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00099-8

2724
S.Hernandez et al./ Bioorg.Med.Chem.10 (2002) 2723–2730
Scheme 1.
riboside] reveal absolutely no contacts between O(40)
and any amino acid residue at the active site.10
withdrawing groups at C6 (i.e., fluoro or tri-
fluoromethyl) can also produce large enhancements in
hydration.14
Since the propensity to form a covalent hydrate must
depend on the interplay of stereoelectronic forces
between the purine ring and the ribofuranosyl moieties,
we surmised that the carbocyclic analogues were poorer
substrates for ADA because these interacting forces
were nullified by the absence of the O(40) oxygen. The
question we seek to answer in the present study is whe-
ther it is possible to manipulate one component of the
system (i.e., the purine ring) to compensate for the
absence of the other [i.e., the O(40) oxygen]. If we accept
as a premise that the O(40) oxygen is critical for the
formation of the covalent hydrate, our goal is to com-
pensate for its absence by using modified purine bases
linked to an appropriate carbocyclic moiety to over-
come the absence of the electron withdrawing effect of
the sugar. In this report we provide evidence that these
two structural elements act in concert and that effective
covalent hydration of the heteroaromatic ring is
strongly dependent on the presence on the O(40) oxygen,
regardless of the electronic nature of the modified
heteroaromatic ring.
The confirmed mechanism for ADA-catalyzed deami-
nation is via formation of an unstable hydrated inter-
mediate engendered by the addition of water across the
purine N1–C6 double bond (Scheme 1).10,11 Hence, the
efficiency of a substrate in facilitating formation of this
critical covalent hydrate should be directly proportional
to its capacity to undergo hydrolysis. Heteroaromatic
hydration is controlled by multiple factors.12À14 Indeed,
the extent of hydration is dependent on the presence
and position of heteroatoms on the ring, their proto-
nation state, and the position and the nature of sub-
stituents attached to the ring. Hydration of purines is
not a favorable event due to loss of resonance energy
and hence hydrated forms are almost undetectable in
solution.12À14 However, the presence of substituents or
additional heteroatoms appears to facilitate the
process.12À14 For example, the calculated difference in
resonance energy loss between 8-aza-9-methylpurine
and 9-methylpurine is 6.6 Kcal/mol less for the former,
suggesting that the extent of hydration increases by the
addition of the heteroatom.12,13 With regard to ADA,
the additional heteroatom resulted in a more than 2-fold
increase in the rate of deamination (rel. Vmax) of
8-azaadenosine relative to adenosine,15 and in a 400-
fold increase in the inhibitory potency of 8-azapurine
riboside (Ki=4.0Â10À8 M) versus purine riboside (Ki=
1.6Â10À5 M),16 reflecting the important role of the 8-aza
modification in facilitating formation of a covalent
hydrate at the active site. In addition, studies of sub-
stituted 9-methylpurine analogues suggested that electron
In this work, we chose to modify the best of the carbo-
cyclic ADA substrates, (N)-MCdA (5),7 with substitu-
tions at positions 6 (6-fluoro, 6) and 8 (8-aza, 7) on the
purine ring with the intent to improve the level of
hydration and hence hydrolysis by ADA. MCdA (5)
belongs to a class of nucleosides where the fused cyclo-
propane moiety constrains the cyclopentane ring to
mimic the conformation of a furanose sugar in the
North hemisphere of the pseudorotational cycle.7 The
selection of the North analogue was based on the 100-
fold difference in the rate of deamination favoring the
constrained North analogue over its antipodal South
(8).7 This conformational preference for a North, C30-
endo pucker is also in agreement with the crystal struc-
ture of ADA–inhibitor complexes10 and enzymatic
studies on conformationally biased substrates.17
Chemistry
The synthesis of the 6-fluorinated analogue 6 was based
on the recently published work of Harris et al.18 for the
efficient conversion of 6-chloropurine nucleosides to
6-fluoropurine derivatives using tetrabutylammoium
triphenyldifluorosilicate
(TBAT).
Our
requisite

S.Hernandez et al./ Bioorg.Med.Chem.10 (2002) 2723–2730
2725
carbocyclic 6-chloropurine 13 (Scheme 2) was synthe-
sized from the known protected intermediate 1019 pre-
pared earlier in our laboratory by Mitsunobu coupling
of 9 with 6-chloropurine. Removal of the O-benzyl
protection in 10 with boron trichloride without neu-
tralization with triethylamine during the quenching step
produced exclusively the hypoxanthine analogue 11,
which was then transformed into the desired 6-chlor-
opurine 13 by direct halogenation with thionyl chloride
using the method of Robins and Basom.20 More
directly, 13 was obtained as the exclusive product if
triethylamine was included during the quenching step.
The conversion of 13 to 6 proceeded according to the
method of Harris et al.18 in excellent yield.
protection afforded the target carbocyclic 8-azaadeno-
sine analogue 7.
Biological Results and Discussion
As reported earlier, the ADA-mediated relative rate of
hydrolysis for the parent conformationally locked (N)-
MCdA (5) was 100 times slower than that for the natural
substrate adenosine (Table 1).7 The two new, carbocyclic
nucleoside analogues 6 and 7 showed rates that were
either essentially unchanged compared to 5, or 200 times
slower than adenosine (Ado, 1) (Table 1). The order of
susceptibility to ADA hydrolysis was Ado (1)> >(N)-
MCdA (5)ꢀ(N)-6F-MCdP (6)>(N)-8-aza-MCdA (7).
Despite the literature report showing that 8-azaadeno-
sine was hydrolyzed twice as fast as adenosine,15 the
corresponding carbocyclic analogue 7 was in fact hydro-
lyzed at approximately half the rate of the parent carbo-
cyclic adenosine analogue 5. These results argue in favor
of the important role of the oxygen atom in assisting
hydrolysis as we previously suggested for the parent
carbocyclic nucleoside, (N)-MCdA (5).7 We propose
that formation of the requisite covalent hydrate inter-
mediate depends, among other factors, on the critical
role of the anomeric effect, which allows for the efficient
transmission of electronic effects between the sugar and
the base. Chattopadhyaya et al.23 have clearly shown
For the synthesis of the 8-aza analogue 7, a linear
approach was chosen to avoid obtaining mixtures of
products from the coupling of 9 with 8-azapurine.
Hence, the first step involved formation of azide 14
(Scheme 3) with diphenylphosphoryl azide which pro-
ceeded with the expected inversion of configuration.
Reduction of the azide via formation of the iminophos-
phorane as reported by Carrie et al.21 proceeded very
smoothly to give the corresponding carbocyclic amine
15. From this intermediate, the stepwise construction of
the 8-azapurine ring continued along a well established
route using nitrous acid for the ring closing step from
16.22 Ammnolysis of 17 and removal of the O-benzyl
Scheme 2.
Scheme 3.

2726
S.Hernandez et al./ Bioorg.Med.Chem.10 (2002) 2723–2730
Table 1. Relative rate of ADA hydrolysis
Substrate
Kinetic dataa (37 ꢁC)
UV analytical data lmax (nm)d
K (minÀ1 b
)
Relative ratec
Substrate
Product
Ado (1)
—
100
260
261
250
279
250
251
250
256
(N)MCdA (5)
(N)-6F-MCdP (6)
(N)-8-aza-MCdA (7)
0.30
0.27
0.13
0.99
0.90
0.42
aData was fitted to 1st order exponential decay using (A=BeÀkt), where B is initial substrate concentration, A is the concentration at time (t), and k
is the rate constant.
bRate constant normalized to 1 Unit of ADA.
cRelative rates calculated using normalized rate constants.
dUV maxima obtained on-the-fly or with spectrophotometer.
that the purine base of adenosine and its sugar moiety
communicate electronically via the anomeric effect. By
measuring the change of ÁGꢁ for the North/South
equilibrium, they showed that as the pH of the solution
decreases, the protonated purine favors the North con-
former in which the orbital overlap between one of the
O(40)oxygen’s lone pair and the s* antibonding orbital
of the C–N glycosyl bond is maximized (anomeric
effect, Fig. 1). Since protonation weakens the glycosyl
bond via the anomeric effect (increased electron density
in the antibonding orbital), this means that the resulting
increase in negative charge on the purine ring would help
stabilize the protonated base at the active site of ADA.
Such effect would be absent in the case of the carbocyclic
nucleoside since any interplay of stereoelectronic forces
between the purine ring and the ribofuranosyl moiety is
completely absent.
It is clear that structural changes to the purine base that
would otherwise expedite hydrolysis in purine nucleo-
sides fail to improve hydrolysis rates when the same
purine bases are presented to the enzyme as carbocyclic
nucleosides, even when these have the locked North
platform (6 and 7) preferred by ADA.7,10 Additional
evidence of the importance of the anomeric effect in the
generation of the critical covalent hydrate intermediate
will probably require syntheses of compounds with het-
erocyclic bases that deviate from conventional purines,
but which per se facilitate even more the formation of a
covalent hydrate. These compounds, such as 19, 20 and
21, bearing identical heterocyclic bases to the natural
product formycin24 and its deaminoformycin analogue,25
and the recently developed imidotriazine riboside,25 will
ultimately help determine the role of the anomeric effect
in hydration. It is important to highlight that the
anomeric effect could still be an important factor in the
efficient deamination of the C-nucleoside formycin,24
since the transmission of electronic effects in C-nucleo-
sides is mediated by an equivalent p!s* anomeric
effect.26 The syntheses of the corresponding bicylo[3.1.0]
hexane nucleosides bearing these new modified bases
(19–21), which will be devoid of anomeric effect, will
For ADA to catalyze the effective hydration of a hetero-
aromatic base the formation of a stable protonated spe-
cies is essential. This is achieved at the active site of ADA
by Glu217, which is perfectly aligned to donate a proton
to N1, thereby making the C6 carbon more susceptible to
nucleophilic attack and hydration.10
2
2
2
Figure 1. O40–C10–N9 anomeric effect. The lone pair orbitals are sp2 [higher energy nsp(p-type), lower energy nsp(s-type)]. The nO4 !s* overlap
stabilizes the North conformation following protonation of the base (box).23 Orbital interaction in the North conformation is possible due to a near
anti-periplanar orientation.
1
0

S.Hernandez et al./ Bioorg.Med.Chem.10 (2002) 2723–2730
2727
ultimately show whether effective hydration of the base
is possible at the active site of ADA in the absence of
this driving force. The synthesis of these compounds is
currently under investigation.
pre-adsorbed on silica gel, packed on the top of a silica
gel column and chromatographed with EtOAc/10%
MeOH to give 13 (0.21 g, 99%) as an off-white solid: mp
179–180 ꢁC. H NMR (CD3OD) d 9.09 (s, 1H, H-8),
1
8.78 (s, 1H, H-2), 5.26 (d, J=6.6 Hz, 1H, H-40), 4.97 (t,
J=8.5 Hz, 1H, H-20), 4.35 (AB d, J=11.7 Hz, 1H,
CHHOH), 3.41 (AB d, J=11.7 Hz, 1H, CHHOH), 2.19
Experimental
(dd, J=14.8, 8.0Hz, 1H, H-3 0 ), 1.90(ddd, J=15.0, 9.0,
a
All chemical reagents were commercially available.
Reported melting points were determined on a Fisher-
Johnson melting point apparatus and are uncorrected.
Column chromatography was performed on silica gel
60, 230–240 mesh (E. Merck), and analytical TLC was
performed on Analtech Uniplates silica gel GF. Routine
IR and 1H, 13C and 19F NMR spectra were recordered
using standard methods. Specific rotation were measured
in a Perkin-Elmer model 241 polarimeter. Low resolution,
positive-ion fast atom-bombardment mass spectra
(FABMS) were obtained on a VG 7070-EHF mass spec-
trometer operated under control of a MSS MASPEC-II32
data system, while accurate mass FABMS determinations
were conducted on a JEOL SX102 instrument. Ele-
mental analyses were performed by Atlantic Microlab,
Inc., Norcross, GA, USA.
6.6 Hz, 1H, H-30b), 1.79 (dd, J=8.3, 3.4 Hz, 1H, H-50),
1.15 (dd, J=5.8, 3.9 Hz, 1H, H-6endo), 0.88 (dd, J=8.0,
6.1 Hz, 1H, H-6exo); 13C NMR (CD3OD) d 152.8, 151.1,
146.7, 132.5, 71.6, 63.6, 57.3, 38.8, 37.4, 26.7, 10.7;
DEPT (CD3OD) d 152.6 (CH), 146.5 (CH), 71.3 (CH),
63.3 (CH2), 57.0(CH), 38.5 (CH 2), 26.4 (CH), 10.3
(CH2); FABMS m/z (relative intensity) 281 (MH+,
100), 155 (b+2H, 53). Anal. calcd for C12H13ClN4
.
O2 0.75H2O: C, 48.98; H, 4.96; N, 19.04. Found: C,
49.26; H, 4.89; N, 18.76.
(1S,2S,4S,5R)-9-[4-Hydroxy-5-(hydroxymethyl)bicyclo
[3.1.0]hex-2-yl]hydropurin-6-one (11). This compound
was obtained using the same protocol as for compound
13, except that quenching of the reaction was performed
without Et3N; mp 132–134 ꢁC; [a]D25 À6.5ꢁ (c 0.32,
1
MeOH). H NMR (DMSO) d 12.3 (br s, 1H, NH, D2O
exchanged), 8.43 (s, 1H, H-2), 8.07 (s, 1H, H-8), 5.04
(app t, 1H, OH, D2O exchanged), 4.90(d, J=6.3 Hz,
1H, H-20), 4.77 (br d, 1H, OH, D2O exchanged), 4.66 (t,
J=8.3 Hz, 1H, H-40), 4.09 (AB d, J=11.5 Hz, 1H,
CHHOH), 3.20(AB d, J=11.7 Hz, 1H, CHHOH), 1.88
Adenosine deaminase (ADA) reaction
In a 1.5 mL Eppendorf vial, 14.2 mL of an aqueous solu-
tion of the nucleoside (5–7, final concentration 50 mM)
and 965.8 mL of phosphate buffer (pH 7.4, 50mM) was
incubated at 37 ꢁC. To this solution, 20 mL diluted ADA
was added to reach a final concentration of enzyme in the
reaction vessel of ꢂ0.3 U/mL. Aliquots (50 mL) were
removed at predetermined time intervals and mixed
with the stopping solution, which contained 450 mL of
water and 2 mL of 20-deoxycoformycin at a concentra-
tion of 0.5 mg/mL. The ADA-mediated hydrolysis was
monitored by HPLC at 255 nm. Chromatography and
on-the-fly UV scans were obtained using a Thermo
Separations Products HPLC system with a Perkin-
Elmer LC235 photodiode array detector. ADA reaction
mixtures were separated on a LC-18S 4.6Â250mm col-
umn with a linear gradient of 0–15% acetonitrile over
20min, with a 5 min re-equilibration time. Due to peak
interference, an isocratic system of 92:8 (water/aceto-
nitrile) was used for compound 6. The flow rate was
1 mL/min and a 50 mL injection from each aliquot was
applied to follow the disappearance of substrate and
formation of product.
(dd, J=14.4, 8.0Hz, 1H, H-3 0 ), 1.73 (ddd, J=15.1, 8.8,
a
6.6 Hz, 1H, H-30b), 1.57 (dd, J=8.3, 3.4 Hz, 1H, H-10),
0.94 (dd, J=5.1, 4.0Hz, 1H, H-6 0endo), 0.69 (dd, J=8.0,
5.9 Hz, 1H, H-6exo); 13C NMR (DMSO-d6) d 156.5,
147.6, 145.1, 137.9, 123.8, 69.1, 61.4, 54.2, 37.8, 35.8,
24.9, 9.3. FAB MS m/z (relative intensity) 263 (MH+,
.
100), 137 (b+2H, 67). Anal. calcd for C12H14N4O3
0.75H2O: C, 50.34; H, 5.87; N, 19.57. Found: C, 50.19;
H, 5.59; N, 19.40.
Synthesis of 13 from 11
A magnetically stirred suspension of 11 (0.074g, 0.28
mmol) in anhydrous CH2Cl2 (5mL) was cooled to 0 ꢁC
(ice bath) under an argon atmosphere. Trifluoroacetic
anhydride (0.8 mL) was added and the mixture was
stirred at 0 ꢁC for 1 h. The reaction vessel was removed
from the cold bath and the mixture was further stirred
at room temperature for 1 h until starting material was
consumed. All volatiles were removed under high
vacuum and the glassy residue was taken up in anhy-
drous CH2Cl2 (5 mL) and maintained under argon. A
solution of SOCl2 (0.5 mL, 6.9 mmol) in anhydrous
DMF (0.5 mL, 6.4 mmol) was added via syringe, and the
mixture was heated to 55 ꢁC for 4 h with monitoring by
TLC (1:1 hexane/EtOAc) until completion. After cool-
ing to rt, the solution was poured into a vigorously stir-
red, ice-cold solution of NaHCO3 and stirring continued
for 15 min until the phases separated. The aqueous layer
was extracted with CH2Cl2 (2Â5 mL) and the combined
organic phases were dried (MgSO4) and concentrated
under reduced pressure. The crude solid was dissolved
in cold MeOH (2 mL) and this solution was passed
(1R,2S,4S,5S)-4-(6-Chloropurin-9-yl)-1-(hydroxymethyl)
bicyclo[3.1.0]hexan-2-ol (13). A stirred solution of 1019
(0.29 g, 0.67 mmol) in anhyd CH2Cl2 (5 mL) was cooled
to À78 ꢁC under a blanket of argon and treated with a
solution of BCl3 (1 M/CH2Cl2, 7 mL). The reaction
mixture was stirred for 5 h at À78 ꢁC, allowed to warm
up to À20 ꢁC, and further stirred for 1 h. After cooling
back to À78 ꢁC, the reaction was quenched by the slow
addition of a 1:1 mixture of MeOH/Et3N (12 mL) and
stirred for 15 min. Note: when the reaction was quenched
with just methanol, the inosine analogue 11 was obtained
as the only product of the reaction (vide infra). After
evaporation of the volatiles, the off-white residue was

2728
S.Hernandez et al./ Bioorg.Med.Chem.10 (2002) 2723–2730
through a short column of neutral alumina packed in
MeOH. After further elution with MeOH, the solution
was evaporated and the residue was purified by column
chromatography with the identical system as described
above to provided an off-white solid having the same
mp and all spectral characteristics of 13.
2.17 (dd, J=14.4, 7.5 Hz, 1H, H-30a), 1.56–1.67 (m, 2H,
H-30b, H 50) 0.96 (dd, J=5.6, 4.1 Hz, 1H, H-60endo), 0.78
(dd, J=8.3, 5.8 Hz, 1H, H-60exo); 13C NMR (CDCl3) d
138.7, 138.6, 128.5, 128.4, 127.8, 127.7, 127.6, 79.6, 72.7,
72.5, 71.4, 62.3, 35.2, 32.4, 26.7, 10.1. FAB MS m/z
(relative intensity) 350(MH +, 10), 91 (100); HRMS
(FAB, MH+) calcd for C21H24N3O2 350.1869; found
350.1874.
(1R,2S,4S,5S)-4-(6-Fluoropurin-9-yl)-1-(hydroxymethyl)-
bicyclo[3.1.0]hexan-2-ol (6). A stirred solution of 13
(0.085 g, 0.30 mmol) in dimethoxyethane (3 mL) and
anhyd DMF (1 mL) was maintained at rt under argon
and treated via cannula with condensed Me3N. The
solution was stirred at room temperature and mon-
itored by TLC (10% EtOH in EtOAc) until completion
(ca. 1.75h) after which time the trimethylammonium salt
precipitated. All volatiles were removed under high
vacuum to give the intermediate salt as a white solid. A
solution of tetrabutylammonium triphenyldifluorosilicate
(TBAT, 0.32 mg, 0.60 mmol) dissolved in anhydrous
DMF (5 mL) was added to the above salt and the reac-
tion was stirred at room temperature for 3 h. The vola-
tiles were removed under reduced pressure and traces of
DMF were removed after several cycles of evaporation
with toluene. The residue was purified by flash column
chromatography on silica gel using 100% EtOAc, fol-
lowed by elution with a mixture of EtOAc/10% EtOH to
afford 6 (0.065 g, 82%) as a white solid: mp 120–121 ꢁC;
[a]D25+17.1ꢁ (c 0.32, MeOH). 1H NMR (CD3OD) d 9.05
(s, 1H, H-8), 8.68 (d, J2,F=0.7 Hz, 1H, H-2), 5.28 (d,
J=6.8 Hz, 1H, H-40), 4.97 (t, J=8.3 Hz, 1H, H-20), 4.35
(AB d, J=11.7 Hz, 1H, CHHOH), 3.41 (AB d, J=
11.7 Hz, 1H, CHHOH), 2.18 (dd, J=14.8, 8.0Hz, 1H,
H-30a), 1.91 (ddd, J=15.3, 9.0, 6.6 Hz, 1H, H-30b), 1.79
(dd, J=08.3, 3.4 Hz, 1H, H-50), 1.15 (dd, J=5.8, 3.9 Hz,
(1S,2S,4S,5R)-4-(Phenylmethoxy)-5-[(phenylmethoxy)-
methyl)bicyclo [3.1.0]hex-2-ylamine (15). A solution of
azide 14 (0.26 g, 0.74 mmol) in anhydrous THF (7 mL)
maintained under argon was treated with triphenylpho-
sphine (0.49 g, 1.85 mmol) and stirred overnight at room
temperature. Water (0.1 mL, 5.5 mmol) was added and
the solution was heated to reflux for 4 h. After cooling
to room temperature, the solvent was removed under
reduced pressure and the crude material was dissolved
in Et2O, dried (MgSO4) and concentrated. The oily
residue was chromatographed on silica gel using first
EtOAc, followed by 15% MeOH in CHCl3 to afford
amine 15 (0.19 g, 80%) as a colorless oil. 1H NMR
(CDCl3) d 7.39 (br s, 10H, Ph), 4.54–4.69 (m, 5H, H-40,
2ÂPhCH2O), 4.00 (AB d, J=10.2 Hz, 1H, PhCH2O
CHH), 3.42 (d, J=5.8 Hz, 1H, H-20), 3.28 (AB d,
J=10.2 Hz, 1H, PhCH2OCHH), 1.66 (br s, 1H, NH2),
1.65–1.93 (m, 2H, H-30a), 1.55 (ddd, J=14.4, 8.7,
6.3 Hz, 1H, H-30b), 1.29–1.34 (m, 1H, H-10), 0.91
(apparent t, J=5.1 Hz, 1H, H-60endo), 0.63 (dd, J=8.0,
5.8 Hz, 1H, H-60exo); 13C NMR (CDCl3) d 139.0, 138.7,
128.3 (Â2), 127.7, 127.6, 127.5 (Â2), 79.3, 72.8, 72.1
(Â2), 52.0, 38.4, 32.0, 30.8, 10.7. FAB MS m/z (rela-
tive intensity) 324 (MH+, 79), 91 (100); HRMS
(FAB, MH+) calcd for C21H26NO2 324.1964; found
324.1952.
1H, H-6 endo), 0.89 (dd, J=8.0, 6.3 Hz, 1H, H-60exo); 13
C
NMR (CD3OD) d 160.6 (d, J=257.5 Hz), 156.5 (d,
J=11.5 Hz), 152.6 (d, J=14.1 Hz), 146.3 (d, J=3.0Hz),
120.9 (d, J=28.7 Hz), 71.5, 63.6, 57.4, 38.9, 37.3, 26.7,
10.6; DEPT (CD3OD) d 152.4 (d, J=14.1 Hz, CH),
146.1 (d, J=2.9 Hz, CH), 71.3 (CH), 63.3 (CH2), 57.1
(CH), 38.6 (CH2), 26.4 (CH), 10.3 (CH2); 19F NMR
(CD3OD) d 0.20. FABMS m/z (relative intensity) 265
(MH+, 100), 139 (b+2H, 58); HRMS (FAB, MH+)
calcd for C12H14FN4O2: 265.1101; found 265.1097.
(1S,2S,4S,5R)-(5-Amino-6-chloropyrimidin-4-yl){4-(phe-
nylmethoxy)-5 [(phenylmethoxy)methyl]bicyclo[3.1.0]hex-
2-yl}amine (16). A mixture of freshly prepared 15
(0.11 g, 0.35 mmol), 5-amino-4,6-dichloropyrimidine
(0.08 g, 0.5 mmol), triethylamine (2 mL) and n-butanol
(3 mL) was refluxed under a blanket of argon for 48 h.
After removal of all volatiles, the residue was pre-
adsorbed on silica gel, packed on the top the column
and chromatographed with a 1:1 mixture of EtOAc/
hexanes to afford the desired product 16 (0.13 g, 81%).
1H NMR (CDCl3) d 8.10(s, 1H, H-2), 7.32–7.45 (m,
10H, Ph), 5.22 (d, J=7.0Hz, 1H, H-2 0), 4.48–4.65 (m,
5H, H-40, 2PhCH2O), 4.30(AB d, J=10.0 Hz, 1H,
PhCH2OCHH), 3.21 (AB d, J=10.0 Hz, 1H, Ph
CH2OCHH), 2.19 (dd, J=13.9, 7.3 Hz, 1H, H-30a), 1.65
(ddd, J=14.4, 8.7, 5.8 Hz, 1H, H-30b), 1.45 (dd, J=8.7,
4.3 Hz, 1H, H-10), 1.15 (dd, J=5.8, 4.4 Hz, 1H, H-
60endo), 0.81 (dd, J=8.0, 6.1 Hz, 1H, H-60exo); 13C NMR
(CDCl3) d 153.5, 149.1, 142.0, 138.7, 138.3, 128.6, 128.5,
128.0( Â2), 127.7, 122.0, 78.6, 73.3, 72.2, 72.1, 52.2,
35.2, 32.7, 27.3, 11.0; DEPT (CDCl3) d 148.2 (CH),
127.7 (CH), 127.5 (CH), 127.1 (CH), 127.0(CH), 126.8
(CH), 77.6 (CH), 72.3 (CH2), 71.2 (CH2), 71.1 (CH2),
51.2 (CH), 34.1 (CH2), 26.8 (CH), 9.9 (CH2). FABMS
m/z (relative intensity) 453 (37Cl MH+, 33), 451 (MH+,
100), 91 (97). HRMS (FAB, MH+) calcd for C25H28
ClN4O2 451.1901, found 451.1894.
(1R,2S,4S,5S)-[4-Azido-2-(phenylmethoxy)bicyclo[3.1.0]-
hexyl] (phenylmethoxy)methane (14). A stirred solution
of 97 (0.50 g, 1.61 mmol) in anhydrous DMF (8 mL) was
maintained at rt under argon and treated dropwise with
diphenylphosphoryl azide (DPPA, 0.85 mL, 4.02 mmol).
Immediately after, Et3N (0.65 mL, 4.35 mmol) was
added via syringe and the resulting mixture was heated
to 60 ꢁC for 2.5 days. After allowing it to reach rt, the
reaction mixture was extracted with Et2O (4Â10mL),
washed with water (2Â5 mL), dried (MgSO4), and con-
centrated under reduced pressure to give a yellow oil.
Purification by flash column chromatography on silica
gel with 1:4 EtOAc/hexanes as eluant afforded 14
(0.42 g,74%) as a pale-yellow oil: IR (neat) 2926, 2095
(N3), 1099cmÀ1. 1H NMR (CDCl3) d 7.32–7.51 (m, 10H,
Ph), 4.57–4.71 (m, 5H, H-20, 2ÂPhCH2O), 4.00 (d, J=
6.1 Hz, 1H, H-40), 3.94 (AB d, J=10.7 Hz, 1H, PhCH2O
CHH), 3.44 (AB d, J=10.7 Hz, 1H, PhCH2OCHH),

S.Hernandez et al./ Bioorg.Med.Chem.10 (2002) 2723–2730
2729
(1R,2S,4S,5S)-[4-(7-chloro(1,2,3-triazolo[5,4-d]pyrimidin-
3 - yl)) - 2 - (phenylmethoxy)bicyclo[3.1.0]hexyl](phenylme-
thoxy)methane (17). Pyrimidine 16 (0.128 g, 0.28 mmol)
was dissolved in a mixture of water (1 mL) and acetic
acid (2 mL). The solution was cooled to 0 ꢁC, and then a
solution of sodium nitrite (0.028 g, 0.4 mmol) in water
(0.5 mL) was added dropwise via syringe. The ice bath
was removed, and the milky suspension was stirred at
room temperature for 3 h. After removal of the volatiles,
the crude mixture was pre-adsorbed on silica gel, packed
on the top of a silica gel column and chromatographed
with 1:1 EtOAc/hexanes as eluant to yield azapurine 17
anhydrous CH2Cl2 (5 mL), cooled to À78 ꢁC and main-
tained under argon with stirring. To this solution, BCl3
(1 M/CH2Cl2, 1.5 mL) was added and the reaction mix-
ture was stirred for 5 h. After allowing the temperature
to reach À20 ꢁC, stirring was continued for an addi-
tional 1 h. MeOH (5 mL) was added and the solution
was evaporated to dryness. The resulting crude product
was neutralized with saturated methanolic ammonia
and reduced to dryness again. The residue was purified
by reverse phase chromatography (Baker octadecyl
C-18) using 15% MeOH in water to give 7 (0.025 g,
75%) as an off-white solid: mp 178 ꢁC; [a]D25 À80ꢁ (c
1
1
(0.075 g, 57%). H NMR (CDCl3) d 8.95 (s, 1H, H-2),
0.15, DMSO). H NMR (DMSO-d6) d 8.41 (br s, 1H,
7.31–7.44 (m, 10H, Ph), 5.56 (d, J=7.3 Hz, 1H, H-40),
5.07 (t, J=8.0Hz, 1H, H-2 0), 4.51–4.77 (m, 4H,
PhCH2O), 3.88 and 3.75 (AB q, J=10.2 Hz, 2H, Ph
CH2OCH2), 2.69 (dd, J=14.9, 7.8 Hz, 1H, H-30a), 2.08–
2.21 (m, apparent pentuplet, 1H, H-30b), 1.77 (dd,
J0=8.3, 3.9, 1H, H-50), 1.26 (dd, J=5.8,04.1 Hz, 1H, H-
6 endo), 1.08 (dd, J=8.3, 6.1 Hz, 1H, H-6 exo); 13C NMR
(CDCl3) d 155.2, 154.1, 149.3, 138.6, 138.4, 134.3, 128.4,
128.3, 127.9, 127.7, 127.5 (Â2), 81.1, 72.8, 72.7, 71.2,
59.8, 35.7, 33.8, 27.6, 10.3; DEPT (CDCl3) d 155.3
(CH), 128.5 (CH), 128.4 (CH), 127.9 (CH), 127.8 (CH),
127.6 (Â2CH), 81.1 (CH), 72.8 (CH2), 72.7 (CH2), 71.2
(CH2), 59.7 (CH), 35.7 (CH2), 27.5 (CH), 10.2 (CH2).
FABMS m/z (relative intensity) 462 (MH+, 36); 91
(100). HRMS (FAB, MH+) calcd for C25H25ClN5O2
462.1697, found 462.1697.
NHH, D2O exchanged), 8.32 (s, 1H, H-2), 8.10(br s,
1H, NHH, D2O exchanged), 5.21 (d, J=7.0Hz, 1H, H-
40), 4.88 (t, J=8.3 Hz, 1H, H-20), 4.73 (d, J=5.8, 1H,
OH, D2O exchanged), 4.44 (dd, J=6.3, 4.8, 1H, OH,
D2O exchanged), 3.81 (dd, J=11.4, 6.1 Hz, 1H,
CHHOH) 3.46 (dd, J=11.4, 4.3 Hz, 1H, CHHOH),
2.34 (dd, J=14.8, 7.8 Hz, 1H, H-30a), 1.82–1.94 (m, 1H,
apparent pentuplet, 1H, H-30b), 1.57 (dd, J=8.3, 3.4,
1H, H-50), 0.92 (dd, apparent t, J=4.8 Hz, 1H, H-
60endo), 0.73 (dd, J=8.0, 5.1, 1H, H-60exo); 13C NMR
(DMSO-d6) d 156.2, 156.1, 148.0, 124.0, 71.3, 61.2, 58.0,
36.6, 35.9, 25.5, 8.3. HRMS (FAB, MH+) calcd for
C11H15N6O2 263.1256, found 263.1253.
References and Notes
1. Wolfenden, R.; Kaufman, J.; Macon, J. B. Biochemistry
1969, 8, 2412.
2. Evans, B. E.; Wolfenden, R. Biochemistry 1973, 12, 392.
3. Valentine, W. N.; Paglia, D. E.; Tartaglia, A. P.; Gilsanz,
F. Science 1977, 195, 783.
4. Martin, D. W.; Gelfand, E. W. Annu.Rev.Biochem. 1981,
50, 845.
5. Grever, M. R.; Siaw, M. F. F.; Jacob, W. F.; Neidhart,
J. A.; Miser, J. S.; Coleman, M. S.; Hutton, J. J.; Balcerzak,
S. P. Blood 1981, 57, 406.
6. Koller, C. A.; Mitchell, B. S. Cancer Res. 1983, 43, 1409.
7. Marquez, V. E.; Russ, P.; Alonso, R.; Siddiqui, M. A.;
Hernandez, S.; George, C.; Nicklaus, M. C.; Dai, F.; Ford, H.,
Jr. Helv.Chim.Acta 1999, 82, 2119.
8. Jacobson, K. A.; Ji, X.; Li, A.; Melman, N.; Siddiqui,
M. A.; Shin, K. J.; Marquez, V. E.; Ravi, R. G. J.Med.Chem.
2000, 43, 2196.
9. Lee, K.; Ravi, R. G.; Ji, X.; Marquez, V. E.; Jacobson,
K. A. Bioorg.Med.Chem.Lett. 2001, 11, 1333.
10. Wang, Z.; Quiocho, F. A. Biochemistry 1998, 37, 8314.
11. Jones, W.; Kurz, L. C.; Wolfenden, R. Biochemistry 1989,
28, 1242.
12. Albert, A. J.Chem.Soc.(B) 1966, 427.
13. Bunting, J. W.; Perrin, D. D. J.Chem.Soc.(B) 1966, 433.
14. Erion, M. D.; Reddy, M. R. J.Am.Chem.Soc. 1998, 120,
3295.
(1R,2S,4S,5S)-1-(Phenylmethoxy)methyl-2-(Phenylme-
thoxy)bicyclo[3.1.0]hex-4-yl)-1,2,3-triazolino[4,5-d]pyri-
midine-7-ylamine (18). Chloroazapurine 17 (0.075 g,
0.16 mmol) was dissolved in a minimum amount of
MeOH and added to a saturated NH3/MeOH solution
(10mL). The resulting solution was heated in a sealed
pressure Pyrex tube at 80 ꢁC for 22 h. After cooling, the
solution was evaporated and co-evapored twice with
MeOH (10mL). The crude mixture was pre-adsorbed
on silica gel, packed on top of a silica gel column and
chromatographed with EtOAc as eluantꢁ to give 18
(0.064g, 90%) as an off-white solid: mp 141 C; 1H NMR
(CDCl3) d 8.50(s, 1H, H-2), 7.28–7.44 (m, 10H, Ph), 6.96
(br s, 2H, NH2), 5.43 (d, J=7.3 Hz, 1H, H-40), 5.10(t,
J=8.0Hz, 1H, H-2 0), 4.52–4.78 (m, 4H, 2ÂPhCH2O),
3.90and 3.75 (AB q, J=10.5 Hz, 2H, PhCH2OCH2),
2.68 (dd, J=14.4, 7.5 Hz, 1H, H-30a), 2.03–2.16 (m,
apparent pentuplet, 1H, H-30b), 1.76 (dd, J=8.3, 3.9,
1H, H-50), 1.22 (dd, J=5.6, 4.3 Hz, 1H, H-60endo), 1.03
(dd, J=8.3, 6.3 Hz, 1H, H-60exo); 13C NMR (CDCl3) d
156.4, 156.0, 148.8, 138.8, 138.6, 128.4 (Â2), 127.8,
127.7, 127.6, 127.5, 124.7, 81.3, 72.7, 72.6, 71.6, 58.5,
35.7, 33.6, 27.9, 10.2; DEPT (CDCl3) d 156.5 (CH),
128.5 (CH), 128.4 (CH), 127.9 (CH), 127.7 (CH), 127.6
(Â2CH), 81.3 (CH), 72.7 (CH2), 72.6 (CH2), 71.6 (CH2),
58.5 (CH), 35.7 (CH2), 27.8 (CH), 10.2 (CH2). FABMS
m/z (relative intensity) 443 (MH+, 100); 91 (97); HRMS
(FAB, MH+) calcd for C25H27N6O2 443.2195, found
443.2184.
15. Simon, L. N.; Bauer, R. J.; Tolman, R. L.; Robins, R. K.
Biochemistry 1970, 9, 573.
16. Shewach, D. S.; Krawczyk, S. H.; Acevedo, O. L.; Town-
send, L. B. Biochem.Pharmacol. 1992, 44, 1697.
17. Ford, H., Jr.; Dai, F.; Mu, L.; Siddiqui, M. A.; Nicklaus,
M. C.; Anderson, L.; Marquez, V. E.; Barchi, J. J., Jr. Bio-
chemistry 2000, 39, 2581.
18. Gurvich, V.; Kim, H.-Y.; Hodge, R. P.; Harris, C. M.;
Harris, T. M. Nucleosides Nucleotides 1999, 18, 2327.
19. Siddiqui, M. A.; Ford, H., Jr.; George, C.; Marquez, V. E.
Nucleosides Nucleotides 1996, 15, 235.
(1R,2S,4S,5S)-4-(7-Amino(1,2,3-triazolo[4,5-d]pyrimi-
din-3-yl))-1-(hydroxymethyl)bicyclo[3.1.0]hexan-2-ol (7).
Compound 18 (0.060 g, 0.13 mmol) was dissolved in

2730
S.Hernandez et al./ Bioorg.Med.Chem.10 (2002) 2723–2730
20. Robins, M. J.; Basom, G. L. In Nucleic Acid Chemistry
Part II; Townsend, L. B., Tipson, R. S., Eds.; John Wiley &
Sons: New York, 1978; p 601.
21. Vaultier, M.; Knouzi, N.; Carrie, R. Tetrahedron Lett.
1983, 24, 763.
22. Marquez, V. E.; Lim, B. B.; Driscoll, J. S.; Snoek, R.;
Balzarini, J.; Ikeda, S.; Andrei, G.; De Clercq, E. J.Hetero-
cycl.Chem. 1993, 30, 1393.
23. Luyten, I.; Thibaudeau, C.; Chattopadhyaya, J. J.Org.
Chem. 1997, 62, 8800.
24. Wierzchowski, J.; Shugar, D. Z. Naturforsch 1983, 38C,
67.
25. Bojack, G.; Earnshaw, C. G.; Klein, R.; Lindell, S. D.;
Lowinski, C.; Preuss, R. Org.Lett. 2001, 3, 839.
26. Luyten, I.; Thibaudeau, C.; Sandstrom, A.; Chatto-
padhyaya, J. Tetrahedron 1997, 53, 6433.