5'-Phosphoramidates and 5'-diphosphates of 2'-O-allyl-β-D- arabinofuranosyl-uracil, -cytosine, and -adenine: Inhibition of ribonucleotide reductase

Article abstract of DOI:10.1021/jm9807095

Continuing our studies on ribonucleotide reductase (RNR) mechanism-based inhibitors, we have now prepared the diphosphates (DP) of 2'-O-allyl-1-β-D- arabinofuranosyl-uracil and -cytosine and 2'-O-allyl-9-β-D-arabinofuranosyl- adenine and evaluated their i

Full text of DOI:10.1021/jm9807095

J . Med. Chem. 1999, 42, 3243-3250
3243
5′-P h osp h or a m id a tes a n d 5′-Dip h osp h a tes of 2′-O-Allyl-â-D-a r a bin ofu r a n osyl-
u r a cil, -cytosin e, a n d -a d en in e: In h ibition of Ribon u cleotid e Red u cta se
Stefano Manfredini,*,† Pier Giovanni Baraldi,† Elisa Durini,† Silvia Vertuani,† J an Balzarini,‡ Erik De Clercq,‡
Anna Karlsson,§ Valentina Buzzoni,†, and Lars Thelander
Department of Pharmaceutical Sciences, Ferrara University, Italy, Rega Institute for Medical Research, Katholieke Universiteit
Leuven, Leuven, Belgium, Karolinska Institute, Huddinge University Hospital, Stockholm, Sweden, and Department of Medical
Biochemistry and Biophysics, Umea˚ University, Sweden
Received December 17, 1998
Continuing our studies on ribonucleotide reductase (RNR) mechanism-based inhibitors, we
have now prepared the diphosphates (DP) of 2′-O-allyl-1-â-D-arabinofuranosyl-uracil and
-cytosine and 2′-O-allyl-9-â-D-arabinofuranosyl-adenine and evaluated their inhibitory activity
against recombinant murine RNR. 2′-O-Allyl-araUDP proved to be inhibitory to RNR at an
IC50 of 100 µM, whereas 2′-O-allyl-araCDP was only marginally active (IC50 1 mM) and 2′-O-
allyl-araADP was completely inactive. The susceptibility of the parent nucleosides to phos-
phorylation by thymidine kinase and 2′-deoxycytidine kinase was also investigated, and all
nucleosides proved to be poor substrates for the above-cited kinases. Moreover, prodrugs of
2′-O-allyl-araU and -araC monophosphates, namely 2′-O-allyl-5′-(phenylethoxy-L-alanyl phos-
phate)-araU and -araC, were prepared and tested against tumor cell proliferation but proved
to be inactive. A molecular modeling study has been conducted in order to explain our results.
The data confirm that for both the natural and analogue nucleoside diphosphates, the principal
determinant interaction with the active site of RNR is with the diphosphate group, which forms
strong hydrogen bonds with Glu623, Thr624, Ser625, and Thr209. Our findings indicate that
the poor phosphorylation may represent an explanation for the lack of marked in vitro cytostatic
activity of the test compounds.
In tr od u ction
2′,2′-difluoro-2′-deoxycytidine, dFdC).3 HU has been
used for a long time as a cancer chemotherapeutic agent,
and dFdC has been recently approved by the FDA for
the treatment of pancreatic cancer. We have recently
reported the results of our studies directed to the design
of new nucleoside analogues potentially active as RNR
mechanism-based inhibitors.4,5
Continuing our efforts, we have now prepared the
diphosphates of 2′-O-allyl-1-â-D-arabinofuranosyl de-
rivatives5 and tested them for their inhibitory activity
against the purified recombinant murine RNR. The
susceptibility of the parent nucleosides to phosphory-
lation by thymidine and 2′-deoxycytidine kinases has
been investigated, and a molecular modeling study has
been conducted. The pronucleotide approach has also
been evaluated by preparing the prodrug of 2′-O-allyl-
araU and -araC monophosphate, namely 2′-O-allyl-5′-
(phenylethoxy-L-alanyl phosphate)-araU and -araC. In
this paper, we also describe the experimental procedures
for the synthesis of the parent 2′-O-allyl-araU, -araC,
and -araA compounds.
Ribonucleotide reductase (RNR) is an essential en-
zyme in DNA synthesis: it is responsible for the de novo
synthesis of all deoxyribonucleoside diphosphates (dNDP)
in prokaryotic (E. coli) and eukaryotic (mammalian)
cells.1 The enzyme comprises two homodimers termed
R1 and R2: the site of reduction is contained in the R1
subunit (i.e. Cys462 and Cys225 in E. coli), whereas the
radical is on the R2 subunit (i.e. diferric-tyrosyl cofactor,
Tyr122, in E. coli) maintained by a cofactor. The
enzymatic activity requires subunit association and
transfer of the free radical from R2 onto R1 subunit.
Because RNR is the rate-limiting enzyme in dNDP
synthesis in eukaryotic cells and because RNR gene
overexpression has been associated with malignant
transformation and metastatic potential,2 RNR has been
considered as an attractive target for antitumor/
antiviral chemotherapy, and a number of potent inhibi-
tors have been discovered and proposed as potential
antitumor/antiviral drugs. As a result of these studies
three main classes of inhibitors have emerged: (a)
compounds that scavenge the tyrosyl radical on the R2
subunit and/or inhibit its transfer to the R1 subunit (i.e.
hydroxyurea, HU); (b) compounds that prevent R1 and
R2 association (i.e. HSV R2-C-terminal amino acid
peptidomimetics); (c) mechanism-based inhibitors, mainly
modified nucleosides that must enter the active site at
the R1 subunit to block the enzyme (i.e. gemcitabine,
Ch em istr y
Although the 2′-O-allylation of ribonucleosides has
been reported without protecting the 3′- and 5′-hydroxy
groups, in the case of our arabinofuranosides we pro-
ceeded with protected precursors, the reason being that
a mixture of 2′- and 3′-O-allyl derivatives is usually
obtained with unprotected nucleosides.6,7 Thus, 2,2′-
anhydro-1-â-D-arabinofuranosyl-uracil (1)8 was chosen
as a common starting material for the preparation of 4
and 7. This intermediate is particularly useful for our
† Ferrara University.
‡ Katholieke Universiteit Leuven.
§ Karolinska Institute.
Umea˚ University.
10.1021/jm9807095 CCC: $18.00 © 1999 American Chemical Society
Published on Web 08/11/1999

3244 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17
Manfredini et al.
Sch em e 1a
a
(i) DHP, TsOH‚H2O, CH3CN, (ii) KOH, EtOH; (iii) allyl bromide, NaH (60%), THF; (iv) TsOH‚H2O, MeOH; (v) 1,2,4-triazole, POCl3,
Et3N, CH3CN; (vi) 30% NH4OH/H2O, dioxane; (vii) Dowex 1X2 resin (OH- form), H2O.
Sch em e 2a
A: Taking advantage of the procedure described by
Kova´cs and O¨ tvo¨s12 for the synthesis of dNTPs, the
starting nucleoside 7 (Scheme 3) was treated with
POCl3/trimethyl phosphate solution to give the mono-
phosphate dichloride 11, which was in turn reacted with
tri-n-butylammonium orthophosphate (0.5 M solution)
to give, after purification and displacement of the
butylammonium ion, 2′-O-allyl-1-â-D-arabinofuranosyl-
cytosine 5′-diphosphate (12) as the sodium salt in 13%
yield.
a
(i) Allyl bromide, NaH (50%), THF; (ii) NH4F/MeOH.
Meth od B: We adapted the procedure described by
Poulter et al.;13 2′-O-allyl-5′-O-tosyl-1-â-D-arabinofura-
nosyl-uracil (16a ) and -adenine (16b) were reacted with
the tris(tetra-n-butylammonium)hydrogen pyrophos-
phate to displace the tosyl group and to obtain, after
purification and displacement of the butylammonium
ion, the expected diphosphates 17a and 17b (19% and
38% yield, respectively). Both 5′-tosyl nucleosides 16a
and 16b were prepared by conventional procedures from
the corresponding starting nucleosides in four steps.
Thus, 2′-O-allyl-3′,5′-O-TBDMS-araU (13a ), obtained
from compound 4 by reaction with tert-butyldimethyl-
silyl chloride, and 2′-O-allyl-3′,5′-O-TBDMS-araA (9)
were deprotected at the 5′-position by treatment with
80% acetic acid to give the corresponding 2′-O-allyl-3′-
O-TBDMS nucleosides 14a and 14b in 54% and 67%
yield, respectively. Next, reaction of these intermediates
with tosyl chloride in the presence of 4-dimethylami-
nopyridine (DMAP) gave the 2′-O-allyl-3′-O-TBDMS-5′-
O-tosyl derivatives 15a and 15b in 84% and 80% yield,
respectively. Removal of the silyl protecting group with
2% HF gave 2′-O-allyl-5′-O-tosyl-araU (16a ) and 2′-O-
allyl-5′-O-tosyl-araA (16b) in 89% and 59% yield, re-
spectively.
purpose because it can be easily protected in a regiospe-
cific manner at the 5′- and 3′-hydroxy groups and it can
also be converted to the corresponding cytosine deriva-
tive by known procedures.9 Briefly, 2,2′-anhydro-1-â-D-
arabinofuranosyl-uracil (1) was protected as the 3′,5′-
O-THP derivative, and the crude residue was treated
with KOH in ethanol to give 3′,5′-O-THP-araU (2)
(Scheme 1) which was finally allylated at the 2′-position
with allyl bromide and sodium hydride, as the base, in
THF to give compound 3 (58% overall yield, as compared
to 1). This latter was deprotected to give the expected
2′-O-allyl-araU (4) by treatment with toluenesulfonic
acid (TsOH) in 63% yield or converted into the corre-
sponding 3′,5′-O-THP-2′-O-allyl-araC (6) in two steps.
Next treatment with TsOH gave the deprotected 7 in
27% yield.
In the case of araA, better results were achieved if
the 3′- and 5′-hydroxyl moieties were protected with tert-
butyldimethylsilyl groups (TBDMS). Allylation of 810
with allyl bromide and sodium hydride gave the ex-
pected 9 in 32% yield, with concomitant formation of
mixtures of byproducts alkylated at the N6- and/or 2′-
O-positions (Scheme 2). The TBDMS groups were then
removed by treatment with ammonium fluoride/metha-
nol (NH4F/MeOH) to give the final 10 in 82% yield.11
Syn th esis of Nu cleosid e Dip h osp h a tes. Meth od
Syn th esis of Nu cleosid e P h osp h or a m id a tes. Pro-
nucleotides of 2′-O-allyl-1-â-D-arabinofuranosyl-uracil
and -cytosine, namely compounds 18 and 19, were

Inhibition of Ribonucleotide Reductase
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17 3245
Sch em e 3a
a
For 7, 11, 12: (i) (Me3O)3PO, Proton Sponge, POCl3, 0 °C; (ii) n-Bu3NH+H2PO4-, n-Bu3N, TEAB (0.2 M); (iii) Dowex AG 50 W-X2
(Na+ form). For 9, 13a , 14a ,b-17a ,b: (i) AcOH (80%), 60 °C; (ii) tosyl chloride, 4-DMAP, 0 °C; (iii) 2% HF/H2O, CH3CN; (iv)
[(n-But)4N]3HP2O7, CH3CN; (v) Dowex AG 50W-X2 (Na+ form).
Sch em e 4a
Å. The resulting conformation was further optimized for
1500 steps of SD and 2000 steps of PRCG, and the
structure of the appropriate substrate was modified at
the 2′-position, on the glycosylic moiety, by substitution
of the â-hydrogen with the â-O-allyl group and of the
R-hydroxy group with an R-hydrogen. The structure was
finally optimized until convergence. These calculations
indicated high stability of the substrate-enzyme com-
plexes and close similarity between the structures of the
enzyme complexed with the studied synthetic molecules
and the X-ray structure. This confirms that the calcula-
tions utilized for the structural analysis are able to
reproduce the real situation very closely. For the opera-
tions above-described, MacroModel (5.5) software and
the Amber force field were used.
a
(i) Phenylethoxy-L-alanyl phosphorochloridate, 1-methylimi-
dazole, THF.
prepared by reaction of nucleosides 4 and 7 with
phenylethoxy-L-alanyl phosphorochloridate, adapting a
procedure described by McGuigan et al.14 (Scheme 4).
The compounds were obtained in 20% and 18% yield,
respectively. All the compounds were characterized by
1H and 31P NMR, mass spectroscopy, and HPLC.
Biology
RNR In h ibition . Murine recombinant RNR proteins
R1 and R2 were obtained and purified as previously
reported.16,17 RNR activity of recombinant murine R1
and R2 protein was measured using the [3H]CDP
reaction assay as described earlier.18 ADP reduction was
determined after separation of the product from the
substrate by boronate affinity chromatography as de-
scribed by Shewach.19 The degree of inhibition of RNR
is shown in Figure 1 with the best IC50 (100 µM) for
2′-O-allyl-araUDP. No clear evidence for a time-depend-
ent inhibition was obtained, and therefore, the simplest
explanation is that the inhibition is competitive.
P h osp h or yla tion of Nu cleosid es by Kin a ses. The
susceptibility of the parent nucleosides to phosphory-
lation by cellular and viral-encoded kinases has been
investigated against purified cytosolic thymidine kinase
(TK-1) derived from human T-lymphocyte CEM cells
and cytosolic deoxycytidine kinase (dCyd) and mito-
Mod elin g P a r a m eter s
The enzyme was studied with the three natural
substrates (CDP, ADP, and UDP) docked in the binding
site of the R1 subunit, by modifying the structure of the
natural substrate GDP contained in the crystallographic
structure of the enzyme.15 After placement of the
substrate in the binding site, the substrate, along with
the amino acid residues of the binding site in the range
of 8 Å from the substrate itself, was optimized for 500
steps of steepest descent (SD) and 2000 steps of polar
Ribiere conjugate gradient (PRCG). A distance-depend-
ent dielectric constant was used in conjunction with a
12 Å cutoff in all calculations. The substrate was then
subjected to 25 ps of molecular dynamics (MD) at 300
K and reoptimized using the conjugate gradient algo-
rithm until the gradient norm fell below 0.01 kcal/mol/

3246 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17
Manfredini et al.
an IC50 of 100 µM, whereas 2′-O-allyl-araCDP was only
marginally active (IC50 ∼ 1 mM) and 2′-O-allyl-araADP
was completely inactive. The susceptibility of the parent
2′-O-allyl nucleosides to phosphorylation by thymidine
and 2′-deoxycytidine kinases was investigated, and they
were shown to be very poor substrates for the above-
cited kinases. Since for many nucleoside analogues the
first phosphorylation step is, in general, the rate-
limiting step in the activation of the drug, we have
therefore prepared the 2′-O-allyl-araU and -araC mono-
phosphate prodrugs, namely 2′-O-allyl-5′-(phenylethoxy-
alanyl phosphate)-araU (18) and -araC (19). In particu-
lar, 2′-O-allyl-araU was chosen because it was inactive
in cell culture but the most inhibitory to RNR. However,
when the pronucleotides were evaluated for their effects
on cell proliferation, they also proved inactive. Although
the phosphoramidate pronucleotide approach does not
necessarily guarantee a release of the monophosphate,
absence of biological activity can also be explained by
the lack of further phosphorylation of the drug to its
5′-di(and 5′-tri)phosphate.
Finally, the first crystallographic structure of the R1
subunit complexed to the natural substrate GDP has
been recently obtained by Eklund et al.15 Starting from
these data the three synthetic nucleotides 2′-O-allyl-
araUDP (17a ), -araCDP (12), and -araADP (17b) have
been placed into the active site and their interactions
have been studied through molecular mechanics calcu-
lations. Briefly, the order of the observed inhibitory
activity (2′-O-allyl-araUDP > -araCDP . -araADP)
could be explained by the more hindered conformation
adopted by 2′-O-allyl-araADP into the active site. An-
other important observation emerging from this study
is that for both natural and analogue diphosphates the
principal interaction with the active site concerns the
diphosphate group, which forms strong hydrogen bonds
with Glu623, Thr624, Ser625, and Thr209 (Figure 2).
In our opinion, these data are of relevance in the design
of new RNR mechanism-based inhibitors.
F igu r e 1. Inhibition of RNR by 2′-O-allyl-araCDP (9), 2′-O-
allyl-araADP ([), or 2′-O-allyl-araUDP (b). In the experiments
with araCDP- and araUDP-O-allyl derivatives the concentra-
tion of [3H]CDP was 0.5 mM and dGTP was used as a positive
effector. With the araADP-O-allyl derivative the concentration
of [14C]ADP was 0.02 mM formed after 30-min incubation at
37 °C, where 100% is the activity in the absence of drug.
Ta ble 1. Cytostatic Activity of Compounds 4, 7, 10, 18, and 19
IC50 (µM)a
compd
L1210
Molt4/C8
CEM
FM3A
4
7
10
18
19
>500
>500
>500
>500
38 ( 8.4
181 ( 20
>500
61 ( 6.0
79 ( 1.7
325 ( 32
316 ( 28
285 ( 30
>200
>200
>200
>200
>200
>200
>200
>200
a
50% inhibitory concentration, or compound concentration
required to inhibit tumor cell proliferation by 50%.
chondrial TK-2 thymidine kinase, both derived from
human liver source. None of the compounds served as
substrates for the three nucleoside kinases (IC50 > 1000
µM).
Exp er im en ta l Section
Ch em istr y. Melting points were determined in open capil-
lary tubes and are uncorrected. Reaction courses were rou-
tinely monitored by thin-layer chromatography (TLC) on silica
gel precoated F254 Merck plates with detection under a 254-
nm UV lamp and/or by spraying the plates with 10% H2SO4/
MeOH and heating and/or by spraying the plates with
ammonium molybdate reagent. Nuclear magnetic resonance
spectra were determined in DMSO-d6, CDCl3, CD3OD, or D2O
solution with a Bruker AC-200 spectrometer, and chemical
shifts are presented in ppm from internal tetramethylsilane
as a standard; 31P NMR spectra were determined in DMSO-
d6, CDCl3, or D2O solution with a Bruker AM-200 spectrom-
eter, and chemical shifts are presented in ppm from internal
85% H3PO4/H2O solution as a standard. Ultraviolet spectra
were recorded on a Kontron UVIKON 922 spectrometer.
Matrix-assisted laser desorption ionization time-of-flight (MAL-
DI-TOF)spectra were obtained on a Hewlett-Packard HPG2025A
mass spectrometer operating in a positive linear mode. HPLC
(analytical and preparative) separations were performed on a
Waters 600E chromatographic system; reverse-phase Waters
C18 columns (150 × 4.6 mm, 150 Å) were used. Column
chromatography was performed with Merck 60-200 mesh
silica gel. Room temperature varied between 22 and 25 °C;
all drying operations were performed over anhydrous magne-
sium sulfate. Microanalyses, unless stated, were in agreement
with calculated values (0.4%.
Cytosta tic Activity. The cytostatic activity of the
nucleosides was evaluated against murine leukemia
L1210 and mammary FM3A cells and human T4-
lymphocyte Molt and CEM cells. The uracil derivative
4 was not cytostatic at 500 µM, the cytosine derivative
7 was cytostatic between 37 and 325 µM, and the
adenine derivative 10 was cytostatic between 80 and
316 µM (Table 1). The pronucleotides 2′-O-allyl-5′-
(phenylethoxy-L-alanyl phosphate)-1-â-D-arabinofura-
nosyl-uracil (18) and -cytosine (19) were virtually
inactive (IC50 > 200 µM).
Discu ssion
We started the present study in order to evaluate
whether the activity observed in preliminary in vitro
evaluation conducted on some 2′-O-allyl-â-D-arabino-
furanosyl derivatives,5 previously designed by us as
potential RNR mechanism-based inhibitors, was directly
correlated to an interaction with the enzyme. The 2′-
O-allyl-araC, -araU, and -araA diphosphates were pre-
pared and tested for their inhibitory effect on recombi-
nant murine RNR, R1 and R2 subunits. Under the
conditions tested, 2′-O-allyl-araUDP was endowed with
3′,5′-Bis-O-tetr a h yd r op yr a n yl-a r a U (2). 2,2′-Anhydro-
araU8 (1) (500 mg, 2.21 mmol) was dissolved in CH3CN (30

Inhibition of Ribonucleotide Reductase
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17 3247
F igu r e 2. Substrate-binding cleft (stereoview) of the R1 subunit with the natural substrate GDP (orange) and the three synthetic
diphosphates 2′-O-allyl-araCDP (12, pink), 2′-O-allyl-araADP (17a , red), and 2′-O-allyl-araUDP (17b, green). The â-phosphate
groups strongly bind to the hydrogens of residues Glu623, Thr624, Ser625, and Thr209.
mL), and dihydropyrane (DHP; 2 mL, 221 mmol) and p-
toluenesulfonic acid monohydrate (TsOH; 42 mg, 0.22 mmol)
were added. After 3 h at room temperature, TLC (CH2Cl2/
MeOH, 9:1) indicated complete reaction, the solvent was
evaporated, and the residue was dissolved in CH2Cl2 (10 mL)
and washed with saturated NaHCO3 (1 × 10 mL). The organic
layer was dried and evaporated to give a solid, which was
dissolved in 0.1 M KOH/EtOH (15 mL). The mixture was
heated at reflux condictions for 4 h (TLC: EtOAc/hexane, 8:2).
The solvent was then evaporated, and the resulting residue
was purified by silica gel column chromathography (EtOAc/
hexane, 7:3) to give 662 mg of 2 as a yellow oil: yield 76%; 1H
NMR (CDCl3) δ 8.60-8.46 (m, 1H, NH); 7.88 (m, 1H, H6); 6.27
(m, 1H, H1′); 6.09 (m, 1H, OH2′); 5.67 (m, 1H, H5); 4.70-3.50
(m, 11H, H2′, H3′, H4′, H5′, H5”, CHO- and CH2O-THP); 1.80-
1.55 (m, 12H, CH2 × 6, THP).
(ꢀ 2100); 1H NMR (DMSO-d6) δ 11.33 (sbr, 1H, NH); 7.67 (d, J
) 7.9 Hz, 1H, H6); 6.13 (d, J ) 5.0 Hz, 1H, H1′); 5.75-5.50
(m, 3H, Hb-allyl, H5 and OH); 5.15-5.00 (m, 3H, Hc-allyl and
OH); 4.05-3.55 (m, 7H, H2′, H3′, H4′, H5′, H5” and Ha-allyl);
MALDI MS (M + H)+ 285.3 Da. Anal. (C12H16N2O6) C, H, N.
N4-Tr ia zolyl-3′,5′-b is-O-t et r a h yd r op yr a n yl-2′-O-a llyl-
a r a U (5). 1,2,4-Triazole (539 mg, 7.8 mmol) and POCl3 (152
µL, 1.67 mmol) were dissolved in CH3CN (6 mL), and Et3N (1
mL, 7.47 mmol) was added dropwise at 0 °C. Compound 3 (265
mg, 0.58 mmol) dissolved in CH3CN (4 mL) was added to the
above prepared mixture, and the suspension obtained was
stirred at room temperature for 5 h (TLC: EtOAc/hexane, 6:4).
The mixture was then treated with Et3N (0.72 mL, 5.16 mmol)
and H2O (0.25 mL, 13.9 mmol), and after 10 min the solvent
was evaporated and the residue was dissolved in CH2Cl2 (10
mL) and washed with aqueous saturated NaHCO3 solution (1
× 5 mL). The aqueous layer was further extracted with CH2-
Cl2 (2 × 10 mL). The combined organic layers were dried and
evaporated, and the crude residue was purified by silica gel
column chromatography (EtOAc/hexane, linear gradient from
1:1 to 7:3) to give 163 mg of compound 5 as a pale-yellow oil:
3′,5′-Bis-O-tetr a h yd r op yr a n yl-2′-O-a llyl-a r a U (3). Com-
pound 2 (614 mg, 1.49 mmol) was dissolved in THF (10 mL),
and 60% NaH (149 mg, 3.72 mmol) was added under vigorous
stirring under positive argon pressure; after 10 min, allyl
bromide (315 µL, 3.72 mmol) was added and the mixture was
stirred at room temperature for 17 h (TLC: EtOAc/hexane,
7:3). The solvent was then evaporated, and the residue was
dissolved in CH2Cl2 (20 mL), washed with aqueous saturated
NH4Cl solution (1 × 10 mL) and H2O (1 × 10 mL), dried, and
evaporated. The resulting residue was purified by column
chromatography (Et2O/hexane, 8:2) to give 270 mg of 3 as a
white foam: yield 40%; 1H NMR (CDCl3) δ 8.65-8.50 (m, 1H,
NH); 7.77 (m, 1H, H6); 6.25 (m, 1H, H1′); 5.75-5.65 (m, 2H,
H5 and Hb-allyl); 5.25-5.10 (m, 2H, Hc-allyl); 4.80-4.65 (m,
2H, Ha-allyl); 4.25-3.50 (m, 11H, H2′, H3′, H4′, H5′, H5”,
CHO- and CH2O-THP); 1.80-1.55 (m, 12H, CH2 × 6, THP).
1
yield 56%; H NMR (CDCl3) δ 9.29 (s, 1H, H-triazole); 8.50-
8.30 (m, 1H, H6); 8.13 (s, 1H, H-triazole); 7.05-6.95 (m, 1H,
H5); 6.36 (m, 1H, H1′); 5.70-5.60 (m, 1H, Hb-allyl); 5.20-5.05
(m, 2H, Hc-allyl); 4.80-4.70 (m, 2H, Ha-allyl); 4.30-3.50 (m,
11H, H2′, H3′, H4′, H5′, H5”, Ha-THP, CH2O-THP); 1.80-1.55
(m, 12H, CH2 × 6, THP).
2′-O-Allyl-a r a C (7). Compound 5 (150 mg, 0.3 mmol) was
dissolved in dioxane (1 mL), and 30% NH4OH (1 mL) was
added; the mixture was then stirred at room temperature
(TLC: CH2Cl2/MeOH, 9:1). After 20 h, the solvent was
evaporated and the crude residue (compound 6, 100 mg, about
0.22 mmol) was dissolved in MeOH (5 mL). TsOH monohy-
drate (80 mg, 0.42 mmol) was added to the solution, which
was then vigorously stirred at room temperature for 3 h
(TLC: CH2Cl2/MeOH, 9:1). After this time, the reaction was
complete, the solvent was evaporated and the residue was
purified by column chromatography (CH2Cl2/MeOH, 9:1) and
preparative HPLC with CH3CN/H2O (linear gradient from 1:9
to 6:4). This gave, on freeze-drying of the appropriate fractions,
compound 7 toluenesulfonic acid salt as a white solid. Further
purification on Dowex 1-X2 resin (200-400 mesh, OH- form)
gave, after freeze-drying of appropriate fractions, 23 mg of
2′-O-Allyl-a r a U (4). Compound 3 (276 mg, 0.6 mmol) was
dissolved in MeOH (5 mL), and TsOH monohydrate (221 mg,
1.16 mmol) was added. The mixture was vigorously stirred at
room temperature until complete conversion to the deprotected
compound (TLC: CH2Cl2/MeOH, 9:1). After evaporation, the
crude residue was purified by column chromatography (CH2-
Cl2/MeOH, linear gradient from 9.5:0.5-9:1) and preparative
HPLC using CH3CN/H2O (linear gradient from 1:9 to 6:4). This
gave, on freeze-drying of the appropriate fractions, 110 mg of
compound 4 as a white solid: yield 63%; mp 168-170 °C
(MeOH/Et2O); UV (MeOH) λmax 264 nm (ꢀ 9700), λmin 232 nm

3248 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17
Manfredini et al.
compound 7 (free base) as a gray solid: yield 27%; mp 87-93
°C (MeOH/Et2O); UV (MeOH) λmax 273 nm (ꢀ 7300), λmin 253
1H, Hb-allyl); 5.20-4.80 (m, 3H, Hc-allyl and H5); 4.24-4.10
(m, 1H, H3′); 4.08-3.90 (m, 6H, Ha-allyl, H2′, H4′, H5′, H5”);
31P NMR (DMSO-d6) δ -7.87 (d, J p,p ) 20.33 Hz, 1P, PR);
-15.36 (d, J p,p ) 20.2 Hz, 1P, Pâ); MALDI MS (M + Na)+
466.2 Da.
1
nm (ꢀ 5500); H NMR (DMSO-d6) δ 7.58 (d, J ) 7.4 Hz, 1H,
H6); 7.15 (m, 2H, NH2); 6.14 (d, J ) 5.0 Hz, 1H, H1′); 5.75-
5.50 (m, 3H, Hb-allyl, H5 and OH); 5.15-4.95 (m, 3H, Hc-
allyl and OH); 4.15-3.50 (m, 7H, H2′, H3′, H4′, H5′, H5” and
Ha-allyl); MALDI MS (M)+ 283.8 Da. Anal. (C12H17N3O5) C,
H, N.
3′,5′-Bis-O-ter t-bu tyl-d im eth ylsilyl-a r a A (8). Compound
8 was prepared following and adapting the procedure reported
by Baker et al.10 Ara-A (250 mg, 0.94 mmol) was dissolved in
anhydrous DMF (5 mL) under positive argon pressure, and
freshly distilled Et3N (0.65 mL, 4.7 mmol) and TBDMS-Cl (531
mg, 3.52 mmol) were then added. After 72 h at 60 °C (TLC:
CH2Cl2/MeOH, 9:1) the mixture was diluted with EtOAc (10
mL) and washed with H2O (10 mL). The organic layer was
dried and evaporated, and the crude residue was purified by
silica gel column chromatography (CH2Cl2/MeOH, 9.5:0.5) to
give 348 mg of 8: yield 74%; mp 175-178.10
3′,5′-Bis-O-ter t-bu tyld im eth ylsilyl-2′-O-a llyl-a r a A (9).
Compound 8 (174 mg, 0.35 mmol) was dissolved in THF (6
mL), and NaH 60% (35 mg, 0.88 mmol) was added under
vigorous stirring; after 15 min allyl bromide (75 µL, 0.88 mmol)
was added and the mixture was stirred under positive argon
pressure at room temperature for 16 h (TLC: EtOAc/hexane,
8:2). When the reaction was complete, the solvent was evapo-
rated and the residue was dissolved in CH2Cl2 (10 mL), washed
with aqueous saturated NH4Cl solution (2 × 10 mL), dried,
and evaporated. The resulting solid was purified by silica gel
column chromatography (EtOAc/hexane, 6:4) to give 61 mg of
9: yield 32%; white foam; 1H NMR (CDCl3) δ 8.34 (s, 1H, H2);
8.20 (s, 1H, H8); 6.52 (d, J ) 5.1 Hz, 1H, H1′); 6.00 (sbr, 2H,
NH2); 5.62-5.45 (m, 1H, Hb-allyl); 5.38-4.92 (m, 2H, Hc-allyl);
4.52-4.49 (m, 1H, H3′); 4.05-3.65 (m, 6H, H2′, H4′, H5′, H5”
and Ha-allyl); 0.91 and 0.93 (s, 18H, 2 × tBut-Si); 0.14, 0.09,
0.07 and 0.02 (s, 12H, 2 × Me2-Si).
2′-O-Allyl-3′,5′-bis-O-ter t-bu tyldim eth ylsilyl-a r a U (13a ).
Compound 4 (200 mg, 0.7 mmol) was dissolved in anhydrous
DMF (5 mL); freshly distilled Et3N (0.66 mL, 5.5 mmol) and
TBDMS-Cl (351 mg, 3.85 mmol) were then added, and the
mixture was heated at 60 °C under vigorous stirring and under
positive argon pressure for 20 h (TLC: CH2Cl2/MeOH, 9.5:
0.5). The solvent was evaporated; the residue was dissolved
in CH2Cl2 (10 mL), washed with saturated NaHCO3 (1 × 10
mL) and brine (1 × 10 mL), dried, filtered, and evaporated.
The resulting solid was purified by silica gel column chroma-
tography (EtOAc/hexane, 3:7) to give 183 mg of compound
13a : yield 51%; white foam; 1H NMR (CDCl3) δ 8.50 (sbr, 1H,
NH); 7.80 (d, J ) 8 Hz, 1H, H6); 6.37 (d, J ) 4.6 Hz, 1H, H1′);
5.94-5.60 (m, 2H, Hb-allyl and H5); 5.40 (dd, J ) 6.6 Hz and
J ) 1 Hz, 2H, Hc-allyl); 4.38-4.00 (m, 3H, H2′, H3′ and H4′);
3.95-3.60 (m, 4H, Ha-allyl, H5′ and H5”); 0.93 and 0.91 (s,
18H, 2 × tBut-Si); 0.12 and 0.1 and 0.08 and 0.03 (s, 12H, 2 ×
Me2-Si); MALDI MS (M + Na)+ 535.9 Da; (M + K)+ 552.0
Da.
2′-O-Allyl-3′-O-ter t-bu tyld im eth ylsilyl-a r a U (14a ). Com-
pound 13a (676 mg, 1.32 mmol) was dissolved in 80% acetic
acid (10 mL), and the solution was stirred at 60 °C for 6 h
(TLC: EtOAc/hexane, 1:1). The mixture was then coevaporated
with EtOH (4 × 10 mL), and the residue was dissolved in (CH2-
Cl2), washed with aqueous saturated NaHCO3 solution (1 ×
10 mL) and brine (1 × 10 mL), dried, and evaporated. The
resulting crude solid was purified by silica gel column chro-
matography (EtOAc/hexane, linear gradient from 3:7 to 8:2)
to give 283 mg of 14a : yield 54%; white foam; 1H NMR (CDCl3)
δ 8.25 (sbr, 1H, NH); 7.70 (d, J ) 6 Hz, 1H, H6); 6.10 (d, J )
4 Hz, 1H, H1′); 5.80-5.50 (m, 2H, H5 and Hb-allyl); 5.10 (dd,
J ) 1 Hz and J ) 6.6 Hz, 2H, Hc-allyl); 4.90 (sbr, 1H, OH);
4.30-4.10 (m, 3H, H2′, H3′ and H4′); 3.85-3.75 (m, 2H, H5′
and H5”); 3.78-3.45 (m, 2H, Ha-allyl); 0.80 (s, 9H, tBut-Si);
0.03 and 0.018 (s, 6H, Me2-Si); MALDI MS (M + Na)+ 422.1
Da; (M + K)+ 438.6 Da.
2′-O -Ally l-3′-O -t er t -b u t y ld im e t h y ls ily l-5′-O -t o s y l-
a r a U (15a ). Compound 14a (283 mg, 0.71 mmol) was dissolved
in anhydrous CH2Cl2 (6 mL), and 4-dimethylaminopyridine
(DMAP) (174 mg, 1.42 mmol) was added under positive argon
pressure. Tosyl chloride (203 mg, 1 mmol) was dissolved in
freshly distilled CH2Cl2 (2 mL), and the solution obtained was
added dropwise at the reaction mixture under vigorous stirring
at 0 °C. After 20 h at 4 °C and 30 min at room temperature,
TLC (EtOAc/hexane, 4:6) indicated complete reaction; the
solvent was evaporated, and the residue was purified by silica
gel column chromatography (EtOAc/hexane, 3:7); 330 mg of
compound 15a was obtained: yield 84%; white foam; 1H NMR
(CDCl3) δ 8.84 (sbr, 1H, NH); 7.75 (d, J ) 8.4 Hz, 2H, Ph);
7.30 (d, J ) 8.4 Hz, 2H, Ph); 7.25 (d, J ) 6 Hz, 1H, H6); 6.15
(d, J ) 3.8 Hz, 1H, H1′); 5.58-5.50 (m, 2H, Hb-allyl and H5);
5.06 (dd, J ) 1 Hz and J ) 6.6 Hz, 2H, Hc-allyl); 4.40-4.10
(m, 3H, H2′, H3′ and H4′); 3.94-3.88 (m, 2H, H5′ and H5”);
3.84-3.77 (m, 2H, Ha-allyl); 2.39 (s, 3H, Me-Ph); 0.80 (s, 9H,
tBut-Si); 0.03 and 0.01 (s, 6H, Me2-Si); MALDI MS (M + Na)+
576.2 Da; (M + K)+ 592.5 Da.
2′-O-Allyl-a r a A (10). The protected compound 9 (57 mg,
0.11 mmol) was dissolved under positive argon pressure in dry
MeOH (2 mL), and NH4F (47 mg, 1.27 mmol) was added. The
mixture was heated at reflux conditions until complete conver-
sion to deprotected compound 10 (TLC: CH2Cl2/MeOH, 9:1).
After evaporation, the crude residue was purified by silica gel
column chromatography (CH2Cl2/MeOH, 9.5:0.5) to provide 29
mg of 10: yield 82%; white solid; mp 219-224 °C (MeOH/
1
Et2O); UV (MeOH) λmax 260 nm (ꢀ 12500); H NMR (DMSO-
d6) δ 8.22 (s, 1H, H2); 8.13 (s, 1H, H8); 7.27 (sbr, 2H, NH2);
6.39 (d, J ) 5.7 Hz, 1H, H1′); 5.70-5.45 (m, 2H, Hb-allyl and
OH); 5.15-4.70 (m, 3H, Hc-allyl and OH); 4.5-4.3 (m, 1H,
H3′); 4.30-3.55 (m, 6H, H2′, H4′, H5′, H5” and Ha-allyl);
MALDI MS (M)+ 307.6 Da. Anal. (C13H17N5O4) C, H, N.
2′-O-Allyl-a r a C 5′-Dip h osp h a te (12). Compound 7 (80 mg,
0.28 mmol) and Proton Sponge (150 mg, 0.7 mmol) were
dissolved in anhydrous trimethyl phosphate (1.4 mL), and
freshly distilled POCl3 (52 µL, 0.56 mmol) was added dropwise
under positive argon pressure at 0 °C. After 4 h (TLC: i-PrOH/
NH4OH/H2O, 11:7:2) a solution of 0.5 M tri-n-butylammonium
orthophosphate (n-But3NH+H2PO4-) (2.8 mL, 1.4 mmol) and
tributylamine (0.28 mL) was added to the stirred suspension
followed by a 0.2 M solution of triethylammonium bicarbonate
(TEAB) (pH 7.5, 15 mL). The solvent was then evaporated,
maintaining bath temperature at about 40-45 °C, and the
residue was purified by a DEAE Sephadex A-25 (2 × 30 cm)
column with TEAB (linear gradient from 0.01 to 1 M), pH 7.5,
at a flow rate of 33 mL/h over 10 h. Tetra-n-butylammonium
cation was exchanged for sodium by passing the solution
through a Dowex AG 50W-X2 column (50-100 mesh, Na+
form) and eluting with 10 column volumes of deionized water.
After freeze-drying of appropriate fractions, the solid mass
obtained was further purified by HPLC with CH3CN/H2O
(linear gradient from 1:9 to 6:4) to give, on freeze-drying, 19
mg of compound 12: retention time ) 5.90; yield 13%; white
solid; mp > 300 °C (MeOH/Et2O); UV (H2O) λmax 278 nm (ꢀ
10200) and 242 nm (ꢀ 9500); 1H NMR (CD3OD) δ 7.95 (d, J )
7.6 Hz, 1H, H6); 6.27 (d, J ) 4.6 Hz, 1H, H1′); 5.90-5.65 (m,
2′-O-Allyl-5′-O-tosyl-a r a U (16a ). Compound 15a (322 mg,
0.58 mmol) was dissolved in CH3CN (15 mL), and 2% HF/H2O
(3 × 15 mL) was added at room temperature every 12 h (TLC:
CH2Cl2/MeOH, 9.5:0.5). The mixture was neutralized with
saturated NH4HCO3, and the solvent was evaporated to give,
on freeze-drying, a crude residue, which was dissolved in a
CH2Cl2/MeOH (9:1) solution, filtered, and purified by column
chromatography (CH2Cl2/MeOH, linear gradient from 9.8:0.2
1
to 9.5:0.5) to give 226 mg of 16a : yield 89%; white foam; H
NMR (CDCl3) δ 8.80 (sbr, 1H, NH); 7.73 (d, J ) 8.4 Hz, 2H,
OTs); 7.32 (d, J ) 8.4 Hz, 2H, OTs); 7.20 (d, J ) 6 Hz, 1H,
H6); 6.10 (d, J ) 3.8 Hz, 1H, H1′); 5.55-5.50 (m, 3H, Hb-allyl,
H5 and OH); 5.05 (dd, J ) 1 Hz and J ) 6.6 Hz, 2H, Hc-allyl);

Inhibition of Ribonucleotide Reductase
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17 3249
4.43-4.13 (m, 3H, H2′, H3′ and H4′) 3.95-3.88 (m, 2H, H5′
and H5”); 3.84-3.77 (m, 2H, Ha-allyl); 2.40 (s, 3H, Me-Ph);
MALDI MS (M + H)+ 439.8 Da; (M + Na)+ 462.0 Da; (M +
K)+ 478.0 Da.
38%; white solid; mp 101-103 °C (MeOH/Et2O); UV (H2O) λmax
259 nm (ꢀ 13600), λmin 226 nm (ꢀ 3900); 1H NMR (D2O) δ 8.52
(s, 1H, H2); 8.29 (s, 1H, H8); 6.59 (d, J ) 6 Hz, 1H, H1′); 5.51-
5.43 (m, 1H, Hb-allyl); 5.01-4.91 (m, 2H, Hc-allyl); 4.68-4.57
(m, 1H, H3′); 4.35-4.20 (m, 4H, H2′, H4′, H5′ and H5”); 4.18-
3.70 (m, 2H, Ha-allyl); 31P NMR (D2O) δ -6.96 (d, J p,p ) 14.8
Hz, 1P, PR); -10.25 (d, J p,p ) 24 Hz, 1P, Pâ); MALDI MS (M
+ H)+ 468.4 Da; (M + Na)+ 490.3 Da; (M + K)+ 506.1 Da.
2′-O-Allyl-a r a U 5′-d ip h osp h a te (17a ). Compound 16a
(226 mg, 0.51 mmol) was dissolved in dry CH3CN (300 µL),
and tris(tetra-n-butylammonium)hydrogen pyrophosphate7 (698
mg, 0.77 mmol) was added (TLC: CH2Cl2/MeOH, 9.8:0.2). After
48 h at room temperature, the solvent was evaporated and
the crude residue was purified by DEAE Sephadex A-25 (2 ×
30 cm) column with TEAB (linear gradient 0.01-1 M), pH 7.5,
at a flow rate of 50 mL/h over 10 h. On freeze-drying of the
appropriate fraction a white mass was obtained. The tetra-n-
butylammonium cation was then exchanged for sodium by
Dowex AG 50W-X2 column chromatography (50-100 mesh,
Na+ form) and eluting with 2 column volumes of deionized
water. The eluent was freeze-dried, and the solid obtained was
further purified by HPLC with CH3CN/H2O (linear gradient
from 1:9 to 6:4) to give, on freeze-drying, 50 mg of compound
17a : retention time ) 3.08; yield 19%; white solid; mp 120-
123 °C (MeOH/Et2O); UV (H2O) λmax 263 nm (ꢀ 11700), λmin
225 nm (ꢀ 3500); 1H NMR (D2O) δ 7.87 (d, J ) 8 Hz, 1H, H6);
6.37 (d, J ) 5.6 Hz, 1H, H1′); 5.94 (d, J ) 8 Hz, 1H, H5); 5.90-
5.75 (m, 1H, Hb-allyl); 5.29-5.16 (m, 2H, Hc-allyl); 4.50-4.40
(m, 2H, H2′ and H3′); 4.30-4.10 (m, 3H, H4′, H5′ and H5”);
4.10-3.90 (m, 2H, Ha-allyl); 31P NMR (D2O) δ -9.4 (d, J p,p
) 19.6 Hz, 1P, PR); -10.3 (d, J p,p ) 19.18 Hz, 1P, Pâ); MALDI
MS (M + Na)+ 468.4 Da.
2′-O-Allyl-3′-O-ter t-bu tyld im eth ylsilyl-a r a A (14b). Com-
pound 9 (129 mg 0.24 mmol) was deprotected with 80% acetic
acid (2.5 mL) as described for compound 13a . The resulting
oil was purified by silica gel column chromatography (EtOAc/
hexane, 8:2) to give 68 mg of 14b: yield 67%; colorless syrup;
1H NMR (CDCl3) δ 8.28 (s, 1H, H2); 8.04 (s, 1H, H8); 6.43 (d,
J ) 5.6 Hz, 1H, H1′); 5.86 (sbr, 2H, NH2); 5.62-5.40 (m, 1H,
Hb-allyl); 5.05-4.95 (m, 3H, Hc-allyl and OH); 4.70-4.50 (m,
1H, H3′); 4.19-3.64 (m, 6H, Ha-allyl, H2′, H4′, H5′, H5”); 0.90
(s, 9H, 1 × tBut-Si); 0.16 (s, 6H, 2 × Me2-Si); MALDI MS (M)+
422.9 Da.
2′-O-Allyl-5′-(p h en yleth oxy-L-a la n yl p h osp h a te)-a r a U
(18). Compound 4 (155 mg, 0.4 mmol) was dissolved in dry
THF (10 mL). Ethoxy-L-alanyl monochlorophosphate (1.2 mL,
1.2 mmol), prepared according to McGuigan et al.,14 and
1-methylimidazole (192 µL, 2.4 mmol) were added dropwise
to the solution at room temperature under vigorous stirring
and positive argon pressure; a green gum precipitated from
the pale-yellow solution. After 5 h, ethoxy-L-alanyl monochlo-
rophosphate (0.5 mL, 0.5 mmol) and 1-methylimidazole (65
µL, 0.8 mmol) were further added; after 72 h, the solvent was
evaporated and the residue was dissolved in CH2Cl2 (15 mL)
and washed with 1 M HCl (1 × 10 mL), aqueous saturated
NaHCO3 solution (1 × 10 mL), and brine (1 × 10 mL). The
organic layer was dried and evaporated, and the crude solid
was purified by silica gel column chromatography (CH2Cl2/
MeOH, linear gradient from 9.8:0.2 to 9:1) and also by
preparative HPLC (linear gradient from H2O to CH3CN) to
give, on freeze-drying, 43 mg of compound 18: yield 20%; white
foam; 1H NMR (CDCl3) δ 11.0-10.50 (br, 1H, NH); 8.70-8.55
(br, 1H, NH); 7.40-7.10 (m, 6H, H-Ar and H6); 6.21 and 6.25
(d × 2, J ) 5.4 Hz, 1H, H1′); 5.75-5.45 (d × 2, J ) 8 Hz and
m, 2H, H5 and Hb-allyl); 5.45-5.30 (br, 1H, OH); 5.30-5.00
(m, 2H, Hc-allyl); 4.29-3.62 (m, 10H, Ha-allyl, H2′, H3′, H4′,
H5′, H5”, CHNH, CH2OC); 1.22-1.13 (m, 6H, MeCHNH,
MeCH2OC); 31P NMR (CDCl3) δ 3.97, 3.65 (1:1); MALDI MS
(M + Na)+ 562.8 Da. Anal. (C23H30N3O10P) C, H, N.
2′-O-Allyl-5′-(p h en yleth oxy-L-a la n yl p h osp h a te)-a r a C
(19). Compound 7 (88 mg, 0.31 mmol) was dissolved in dry
THF (10 mL) and reacted with ethoxy-L-alanyl monochloro-
phosphate (0.9 mL, 0.9 mmol)14 and 1-methylimidazole (150
µL, 1.86 mmol) as described for compound 4. The crude residue
was purified by silica gel column chromatography (linear
gradient from CH2Cl2 to CH2Cl2/MeOH 9:1) and preparative
HPLC with CH3CN/H2O (linear gradient from 1:9 to 6:4) to
give, on freeze-drying, 30 mg of compound 19: yield 18%; pale-
yellow foam; 1H NMR (CDCl3) δ 10.0-9.85 (sbr, 1H, NH); 7.41
and 7.38 (d × 2, J ) 8.2 Hz, 1H, H6); 7.30-7.16 (m, 5H, H-Ar);
7.15 (m, 2H, NH2); 6.20 and 6.23 (d × 2, J ) 5.4 Hz, 1H, H1′);
5.75-5.49 (m, 2H, Hb-allyl and OH); 5.45-5.30 (d × 2, J ) 8
Hz, 1H, H5); 5.13-5.03 (m, 2H, Hc-allyl); 4.29-3.60 (m, 10H,
Ha-allyl, H2′, H3′, H4′, H5′, H5”, CHNH, CH2OC); 1.22-1.13
(m, 6H, MeCH, MeCH2OC); 31P NMR (CDCl3) δ 5.26, 5.04 (1:
1); MALDI MS (M + Na)+ 561.2 Da. Anal. (C23H31N4O9P) C,
H, N.
2′-O -Ally l-3′-O -t er t -b u t y ld im e t h y ls ily l-5′-O -t o s y l-
a r a A (15b). Compound 14b (250 mg, 0.59 mmol) was reacted
with tosyl chloride (124 mg, 0.65 mmol) as described for
compound 14a . The residue was purified by silica gel column
chromatography (EtOAc/hexane, 1:1); 272 mg of 15b was
1
obtained: yield 80%; white foam; H NMR (CDCl3) δ 8.31 (s,
1H, H2); 7.96 (s, 1H, H8); 8.27 (d, J ) 10.8 Hz, 2H, Ph); 7.29
(d, J ) 10.8. Hz, 2H, Ph); 6.46 (d, J ) 4.4 Hz, 1H, H1′); 5.75
(sbr, 2H, NH2); 5.52-5.44 (m, 1H, Hb-allyl); 5.04-4.95 (m, 2H,
Hc-allyl); 4.40-3.98 (m, 1H, H3′); 4.30-4.20 (m, 2H, H2′ and
H4′); 4.15-3.90 (m, 2H, H5′ and H5”); 3.85-3.60 (m, 2H, Ha-
allyl); 2.45 (s, 3H, Me-Ph); 0.89 (s, 9H, 1 × tBut-Si); 0.12 (s,
6H, 2 × Me2-Si).
Biology. Ma ter ia ls a n d m eth od s: Enzymatic assays on
RNR (R1 and R2 subunits) were performed according to the
procedures described by Engstro¨m18 and Shewach.19
2′-O-Allyl-5′-O-tosyl-a r a A (16b). Compound 15b (120 mg,
0.21 mmol) was dissolved in CH3CN (5 mL) and deprotected
with 2% HF/H2O (3 × 5 mL) as described for compound 15a .
The residue was purified by silica gel column chromatography
(CH2Cl2/MeOH, linear gradient from 9.5:0.5 to 9:1) to give 57
mg of compound 16b: yield 59%; white foam; 1H NMR (CDCl3)
δ 8.30 (s, 1H, H2); 8.06 (s, 1H, H8); 7.72 (d, J ) 8.2 Hz, 2H,
OTs); 7.25 (d, J ) 8.2 Hz, 2H, OTs); 6.52 (d, J ) 5 Hz, 1H,
H1′); 6.15 (sbr, 2H, NH2); 5.50-5.41 (m, 2H, Hb-allyl and OH);
5.01-4.91 (m, 2H, Hc-allyl); 4.70-4.57 (m, 1H, H3′); 4.40-
4.10 (m, 4H, H2′, H4′, H5′ and H5”); 4.00-3.60 (m, 2H, Ha-
allyl); 2.39 (s, 3H, Me-Ph).
Cytosta tic a ctivity of test com p ou n d s: All assays were
performed in 96-well microtiter plates. To each well were
added 5-7.5 × 104 cells and a given amount of the test
compound. The cells were allowed to proliferate for 48 h
(murine leukemia L1210 and murine mammary carcinoma
FM3A) or 72 h (human lymphocyte CEM and Molt) at 37 °C
in a humidified CO2-controlled atmosphere. At the end of the
incubation period, the cells were counted in a Coulter counter.
The IC50 (50% effective inhibitory concentration) was defined
as the concentration of compound that reduced the number of
viable cells by 50%.
Nu cleosid e k in a se a ssa y: Purified cytosolic human thy-
midine kinase and human deoxycytidine kinase activity was
assayed as follows: the reaction mixture contained 50 mM Tris
HCl, pH 8.0, 2.5 mM MgCl2, 10 mM dithiothreitol, 2.5 mM
ATP, 1 mg/mL bovine serum albumin, 10 mM NaF, and
[methyl-3H]thymidine or [5-3H]deoxycytidine at 1 µM (≈0.1-
0.4 µCi/assay) in a total volume of 50 µL. Assays were perfomed
at 37 °C during a 30-min incubation period. Aliquots of 45 µL
2′-O-Allyl-a r a A 5′-d ip h osp h a te (17b). Compound 16b
(155 mg, 0.33 mmol) was dissolved in anhydrous CH3CN (200
µL) and converted into the corresponding diphosphate with
tris(tetra-n-butylammonium)hydrogen pyrophosphate7 (451
mg, 0.38 mmol) as described for compound 16a . The appropri-
ate fractions, deriving from the Dowex AG 50W-X2 (50-100
mesh, Na+ form) column chromatography, were freeze-dried,
and the solid obtained was further purified by HPLC (linear
gradient from H2O to CH3CN/H2O, 6:4) to give, on freeze-
drying, 69 mg of compound 17b: retention time ) 3.58; yield

3250 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17
Manfredini et al.
(9) Divakar, K. J .; Reese, C. B. 4-(1,2,4-Triazol-1-yl)- and 4-(3-nitro-
1,2,4-triazol-1-yl)-1-(â-D-2,3,5-tri-O-acetylarabinofuranosyl)py-
rimidin-2(1H)-ones. Valuable intermediates in the synthesis of
derivatives of 1-(â-D-arabinofuranosyl)cytosine (Ara-C). J . Chem.
Soc., Perkin Trans. 1 1982, 1171-1176.
(10) Baker, D. C.; Kumar, S. D.; Waites, W. J .; Harnett, G.; Shannon,
W. M.; Higuchi, W. I.; Lambert, W. J . J . Med. Chem. 1984, 27,
270-274.
(11) Zhang, W.; Robins, M. J . Removal of silyl protecting groups from
hydroxyl functions with ammonium fluoride in methanol. Tet-
rahedron Lett. 1992, 33, 1177-1180.
(12) Kova´cs, T.; O¨ tvo¨s, L. Simple synthesis of 5-vinyl and 5-ethynyl-
2′-deoxyuridine 5′-triphosphates. Tetrahedron Lett. 1988, 29,
4525-4528.
(13) Davisson, V. J .; Davis, D. R.; Dixit, V. M.; Poulter, D. C.
Synthesis of nucleotide 5′-diphosphates from 5′-O-tosyl nucleo-
sides. J . Org. Chem. 1987, 52, 1794-1801.
(14) McGuigan, C.; Pathirana, R.; Mahmood, N.; Devine, K.; Hay,
A. Aryl phosphate derivatives of AZT retain activity against
HIV1 in cell lines which are resistant to the action of AZT.
Antiviral Res. 1992, 17, 311-321.
(15) Eriksson, M.; Uhlin, U.; Ramaswamy, S.; Ekberg, M.; Regnstro¨m,
K.; Sjo¨berg, B. M.; Eklund, H. Binding of allosteric effectors to
ribonucleotide reductase protein R1: reduction of active-site
cysteines promotes substrate binding. Structure 1997, 5, 1077-
1092.
of the reaction mixtures were spotted onto Whatman DE-81
filter paper disks. The filters were subsequently washed three
times for 5 min in 1 mM ammonium formate, one time for 1
min in H2O, and one time for 5 min in ethanol. The radioactiv-
ity was determined by scintillation counting.
Molecu la r m od elin g: Computational studies were con-
ducted on a Silicon Graphics Indigo 2 and O2 workstations
using MacroModel20 (5.5) software and the Amber21 force field.
Ack n ow led gm en t. This work was supported by the
University of Ferrara (research grants for 1997-1998)
and the Belgian Fonds voor Geneeskundig Wetenschap-
pelijk Onderzoek (Project Number 3.0180.95). Valentina
Buzzoni was a Research Fellow, from the University of
Ferrara, at the Umeå University, Sweden (year 1998).
We thank Lizette van Berckelaer for excellent technical
help.
Refer en ces
(1) (a) Stubbe, J . Ribonucleotide reductases in the twenty-first
century. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 2723-2724. (b)
Reichard, P. The evolution of ribonucleotide reduction. TIBS
1997, 22, 81-85.
(16) Mann, G. J .; Gra¨slund, A.; Ochiai, E. I.; Ingemarson, R.;
Thelander, L. Purification and characterization of recombinant
mouse and Herpes Simplex Virus ribonucleotide reductase R2
subunit. Biochemistry 1991, 30, 1939-1947.
(17) Davis, R.; Thelander, M.; Mann, G. J .; Behravan, G.; Soucy, F.;
Beaulieu, P.; Lavalle´e, P.; Gra¨slund, A.; Thelander, L. Purifica-
tion, characterization and localization of subunit interaction area
of recombinant mouse Ribonucleotide Reductase R1 subunit. J .
Biol. Chem. 1994, 269, 23171-23176.
(2) Zhou, B.; Tsai, P.; Ho, R.; Shih, J .; Yen, Y. Overexpression of
transfected human ribonucleotide reductase M2 subunit in
human cancer cells enhances their invasive potential. Clin. Exp.
Metastasis 1998, 16, 43-49.
(3) van der Donk, W.; Yu, G.; Perez, L.; Sanchez, R. J .; Stubbe, J .;
Samano, V.; Robins, M. J . Detection of a new substrate-derived
radical during inactivation of ribonucleotide reductase from
Escherichia coli by gemcitabine 5′-diphosphate. Biochemistry
1998, 37, 6419-6426.
(4) Manfredini, S.; Baraldi, P. G.; Bazzanini, R.; Marangoni, M.;
Simoni, D.; Balzarini, J .; De Clercq, E. Synthesis and cytotoxic
activity of 6-vinyl- and 6-ethynyluridine and 8-vinyl and 8-ethy-
nyladenosine. J . Med. Chem. 1995, 38, 199-203.
(18) Engstro¨m, Y.; Eriksson, S.; Thelander, L.; A° kerman, M. Ribo-
nucleotide reductase from Calf Thymus. Purification and proper-
ties. Biochemistry 1979, 14, 2941-2948.
(5) Manfredini, S.; Baraldi, P. G.; Bazzanini, R.; Simoni, D.;
Balzarini, J .; De Clercq, E. Synthesis and antiproliferative
activity of 2′-O-allyl-1-beta-D-arabinofuranosyl-uracil, -cytosine
and -adenine. Bioorg. Med. Chem. Lett. 1997, 7, 473-479.
(6) Gopalakrishnan, V.; Kumar, V.; Ganesh, K. N. Regioselective
2′,3′-O-allylation of pyrimidine ribonucleosides using phase
transfer catalysis. Nucleosides Nucleotides 1992, 11, 1263-1273
and references therein quoted.
(7) Sproat, B. S.; Iribarren, A. M.; Garcia, R. G.; Beijer, B. New
synthetic routes to synthons suitable for 2′-O-allyl-oligoribo-
nucleotide assembly. Nucleic Acids Res. 1991, 19, 733-738 and
references therein quoted.
(19) Shewach, D. S. Quantitation of deoxyribonucleoside 5′-triphos-
phates by a sequential boronate and anion-exchange high-
pressure liquid chromatographic procedure. Anal. Biochem.
1992, 206, 178-182.
(20) Mohamadi, F.; Richards, N. G. J .; Guida, W. C.; Liskamp, R. M.
J .; Lipton, M. A.; Caulfield, C. E.; Chang, G.; Hendrickson, T.
F.; Still, W. C. MacroModel - an Integrated Software System
for Modeling Organic and Bioorganic Molecules Using Molecular
Mechanics. J . Comput. Chem. 1990, 11, 440-467.
(21) Weiner, S. J .; Kollman, P. A.; Case, D. A.; Singh, U. C.; Ghio,
C.; Alagona, G.; Profeta, S., J r.; Weiner, P. A New Force Field
for Molecular Mechanical Simulation of Nucleic Acids and
Proteins. J . Am. Chem. Soc. 1984, 106, 765-784.
(8) Hampton, A.; Nichol, A. W. Nucleotides. V. Purine ribonucleoside
2′,3′-cyclic carbonates. Preparation and use for the synthesis of
5′-monosubstituted nucleosides. Biochemistry 1966, 5, 2076-
2082.
J M9807095