Structure - activity relationship of heterobase-modified 2′-C-methyl ribonucleosides as inhibitors of hepatitis C virus RNA replication

Article abstract of DOI:10.1021/jm040068f

Hepatitis C virus infection constitutes a significant health problem in need of more effective therapies. We have recently identified 2′-C-methyladenosine and 2′-C-methylguanosine as potent nucleoside inhibitors of HCV RNA replication in vitro. However, both of these compounds suffered from significant limitations. 2′-C-Methyladenosine was found to be susceptible to enzymatic conversions by adenosine deaminase and purine nucleoside phosphorylase, and it displayed limited oral bioavailability in the rat. 2′-C-Methylguanosine, on the other hand, was neither efficiently taken up in cells nor phosphorylated well. As part of an attempt to address these limitations, we now report upon the synthesis and evaluation of a series of heterobase-modified 2′-C-methyl ribonucleosides. The structure-activity relationship within this series of nucleosides reveals 4-amino-7-(2-C-methyl- β-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine and 4-amino-5-fluoro-7-(2-C- methyl-β-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine as potent and noncytotoxic inhibitors of HCV RNA replication. Both 4-amino-7-(2-C-methyl- β-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine and 4-amino-5-fluoro-7-(2-C- methyl-β-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine display improved enzymatic stability profiles as compared to that of 2′-C-methyladenosine. Consistent with these observations, the most potent compound, 4-amino-5-fluoro-7H-pyrrolo[2,3-d]pyrimidine ribonucleoside, is orally bioavailable in the rat. Together, the potency of the 2′-C-methyl-4-amino- pyrrolo[2,3-d]pyrimidine ribonucleosides and their improved pharmacokinetic properties relative to that of 2′-C-methyladenosine suggests that this class of compounds may have clinical utility.

Full text of DOI:10.1021/jm040068f

5284
J . Med. Chem. 2004, 47, 5284-5297
Str u ctu r e-Activity Rela tion sh ip of Heter oba se-Mod ified 2′-C-Meth yl
Ribon u cleosid es a s In h ibitor s of Hep a titis C Vir u s RNA Rep lica tion
Anne B. Eldrup,*,† Marija Prhavc,† J ennifer Brooks,† Balkrishen Bhat,† Thazha P. Prakash,† Quanlai Song,†
Sanjib Bera,† Neelima Bhat,† Prasad Dande,† P. Dan Cook,† C. Frank Bennett,† Steven S. Carroll,‡
Richard G. Ball,§ Michele Bosserman,‡ Christine Burlein,‡ Lawrence F. Colwell,§ J ohn F. Fay,‡
Osvaldo A. Flores,‡ Krista Getty,‡ Robert L. LaFemina,‡ J oseph Leone,§ Malcolm MacCoss,§
Daniel R. McMasters,§ J oanne E. Tomassini,‡ Derek Von Langen,§ Bohdan Wolanski,‡ and David B. Olsen‡
Department of Medicinal Chemistry, Isis Pharmaceuticals, 2280 Faraday Avenue, Carlsbad, California 92008, Department of
Biological Chemistry, Merck Research Laboratories, P.O Box 4, West Point, Pennslyvania 19486, and Department of Medicinal
Chemistry, Merck Research Laboratories, P.O. Box 2000, Rahway, New J ersey 07065
Received March 19, 2004
Hepatitis C virus infection constitutes a significant health problem in need of more effective
therapies. We have recently identified 2′-C-methyladenosine and 2′-C-methylguanosine as
potent nucleoside inhibitors of HCV RNA replication in vitro. However, both of these compounds
suffered from significant limitations. 2′-C-Methyladenosine was found to be susceptible to
enzymatic conversions by adenosine deaminase and purine nucleoside phosphorylase, and it
displayed limited oral bioavailability in the rat. 2′-C-Methylguanosine, on the other hand, was
neither efficiently taken up in cells nor phosphorylated well. As part of an attempt to address
these limitations, we now report upon the synthesis and evaluation of a series of heterobase-
modified 2′-C-methyl ribonucleosides. The structure-activity relationship within this series
of nucleosides reveals 4-amino-7-(2-C-methyl-â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine
and 4-amino-5-fluoro-7-(2-C-methyl-â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine as potent
and noncytotoxic inhibitors of HCV RNA replication. Both 4-amino-7-(2-C-methyl-â-D-ribo-
furanosyl)-7H-pyrrolo[2,3-d]pyrimidine and 4-amino-5-fluoro-7-(2-C-methyl-â-D-ribofuranosyl)-
7H-pyrrolo[2,3-d]pyrimidine display improved enzymatic stability profiles as compared to that
of 2′-C-methyladenosine. Consistent with these observations, the most potent compound,
4-amino-5-fluoro-7H-pyrrolo[2,3-d]pyrimidine ribonucleoside, is orally bioavailable in the rat.
Together, the potency of the 2′-C-methyl-4-amino-pyrrolo[2,3-d]pyrimidine ribonucleosides and
their improved pharmacokinetic properties relative to that of 2′-C-methyladenosine suggests
that this class of compounds may have clinical utility.
In tr od u ction
specific inhibitors. Recently, a number of nucleoside
inhibitors (NIs)5-9 and non-nucleoside inhibitors
(NNIs)10-13 of the NS5B polymerase have been de-
scribed. Although NIs are dependent on the metabolism
of the nucleoside to the corresponding triphosphates for
their activity, they possess advantages over the NNIs
in that the mode of action (chain termination) is well
defined and can be easily verified. Moreover, NIs acting
by chain termination must, by mechanistic necessity,
bind in the highly conserved active-site region of HCV
NS5B and thus have a high likelihood of affecting RNA
replication for all HCV genotypes.
Hepatitis C virus (HCV) infection constitutes a serious
health problem in need of more efficacious therapies.
Current estimates suggest that about 170 million people
suffer from HCV infection worldwide. Although HCV
infection is often asymptomatic, it can progress to
chronic hepatitis, leading to liver cirrhosis and hepa-
tocellular carcinoma in a significant percentage of
patients. Combination therapy using pegylated inter-
feron-R and ribavirin have significantly improved the
sustained response rates for patients infected with HCV
genotypes 2 or 3, but the response rate for patients
infected with genotype 1 remains at a disappointing 42-
46%.1
HCV encodes a series of viral proteins including the
NS2/3 autoprotease, the NS3 serine protease/NTPase/
helicase, and the NS5B RNA-dependent RNA poly-
merase (RdRp).2-4 The RdRp, encoded within the 3′-
terminal portion of the HCV genome, is essential for
viral replication and therefore represents a valid target
for therapeutic intervention through the design of
We recently reported the synthesis and structure-
activity relationship (SAR) of purine ribonucleosides as
inhibitors of HCV RNA replication.8 In this study, 2′-
C-methyladenosine and 2′-C-methylguanosine (Figure
1) were identified as potent, specific inhibitors of HCV
RNA replication in cell culture, and their corresponding
triphosphates were shown to be inhibitors of HCV
NS5B-mediated RNA synthesis. In line with these
findings, 2′-C-methyladenosine and 2′-C-methylcytidine
were described as specific attenuators of HCV RNA
levels by Stuyver et al.9 We found the SAR of the 2′-
and 3′-ribonucleoside positions to be quite stringent;
however, HCV NS5B appeared to be somewhat amen-
able to the modification of the purine heterobase.8
* Author to whom correspondence should be addressed. Phone: 760-
603-3852. Fax: 760-603-4654. E-mail aeldrup@isisph.com.
† Department of Medicinal Chemistry, Isis Pharmaceuticals.
‡ Department of Biological Chemistry, Merck Research Laboratories.
§ Department of Medicinal Chemistry, Merck Research Laboratories.
10.1021/jm040068f CCC: $27.50 © 2004 American Chemical Society
Published on Web 09/15/2004

Heterobase-Modified 2′-C-Methyl Ribonucleosides
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 21 5285
F igu r e 1. Structure of nucleosides 1 and 2, inhibitory potency of the corresponding nucleoside 5′-triphosphates on HCV NS5B-
mediated RNA synthesis (HCV NS5B, IC50/µM), inhibitory potency of nucleosides on HCV RNA replication in a subgenomic replicon
assay harbored in HB-1 cells (HCV replicon, EC50/µM), cellular uptake and metabolism to triphosphate in Huh-7 cells, and oral
bioavailability (%F).
Several NTPs containing heterobase modifications were
found to be inhibitors of HCV-mediated RNA synthesis,
whereas the corresponding nucleosides generally failed
to inhibit HCV RNA replication in the cell-based,
subgenomic replicon assay.8 These data suggest that
cellular uptake and/or kinase mechanisms are highly
sensitive to modifications in the heterobase of ribo-
nucleosides. A closer look at the two most potent NIs
indicated that, for 2′-C-methylguanosine, cellular up-
take and intracellular metabolism to the corresponding
triphosphate were inefficient, suggesting that the po-
tency of this NI may be limited in vivo.7 Furthermore,
2′-C-methyladenosine was found to be a substrate for
adenosine deaminase (ADA)-mediated and purine nu-
cleoside phosphorylase (PNP)-mediated conversions,
suggesting that the high clearance rate and lack of oral
bioavailability observed for this NI7 may be due in part
to metabolic conversion.8
Ch em istr y
The effect of the modification of the heterobase of 2′-
C-methyladenosine was investigated by first synthesiz-
ing the corresponding 7-deaza derivative, 4-amino-
7-(2-C-methyl-â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimi-
dine (9), according to the route depicted in Scheme 1.
Starting with 3,5-bis-O-(2,4-dichlorophenylmethyl)-1-O-
methyl-R-D-ribofuranose (3), oxidation of the 2-hydroxyl
group was performed using Dess-Martin periodinane
in dichloromethane to give 4. Subsequent stereospecific
addition of a methyl group on the â face of the ribofura-
nose to give 3,5-bis-O-(2,4-dichlorophenylmethyl)-2-C-
methyl-1-O-methyl-R-D-ribofuranose (5) was accom-
plished by reaction with methylmagnesium bromide in
diethyl ether. The ribofuranose 5 was next converted
to the corresponding 1-bromo compound (with hydrogen
bromide/acetic acid in dichloromethane) and then re-
acted with the sodium salt of 4-chloro-1H-pyrrolo[2,3-
d]pyrimidine (6)14 (formed with sodium hydride) in
acetonitrile to yield exclusively the â anomer, 7. The
removal of the dichlorophenylmethyl protection groups
was performed using boron trichloride in dichloro-
methane to give 4-chloro ribonucleoside 8. This com-
pound was converted to the desired 4-amino derivative,
9, by ammonolysis at elevated temperature. The 8-aza-
7-deaza derivative of 2′-C-methyladenosine, 4-amino-
1-(2-C-methyl-â-D-ribofuranosyl)-1H-pyrazolo[3,4-d]py-
rimidine (12), was also synthesized from ribofuranose
5, again through the formation of the 1-bromo derivative
and reaction with the sodium salt of 4-amino-1H-
pyrazolo[3,4-d]pyrimidine (10) (formed by sodium hy-
dride) in 1-methyl-2-pyrrolidinone to give 11 (Scheme
2). The deprotection of 11 was carried out with boron
trichloride in dichloromethane to give the desired ribo-
nucleoside, 12. The synthesis of the 3,7-dideaza deriva-
tive of 2′-C-methyladenosine, 4-amino-1-(2-C-methyl-â-
As part of an effort to address the limitations of 2′-
C-methyladenosine (1) (Figure 1) with respect to its
enzymatic stability and bioavailability, we report here
the synthesis of a series of heterobase-modified 2′-C-
methylribonucleosides and evaluate their potency as
inhibitors of HCV RNA replication. Our analysis of
these heterobase-modified 2′-C-methylnucleosides leads
to the identification of 4-amino-7-(2-C-methyl-â-D-ribo-
furanosyl)-7H-pyrrolo[2,3-d]pyrimidine and 4-amino-2-
fluoro-7-(2-C-methyl-â-D-ribofuranosyl)-7H-pyrrolo[2,3-
d]pyrimidine as potent, noncytotoxic inhibitors of HCV
RNA replication. Furthermore, these nucleosides are
found to possess attractive profiles in terms of enzymatic
stability when compared to that of 2′-C-methyladenos-
ine. Most importantly, as exemplified by the more potent
fluoro-substituted derivative, this class of inhibitors
possesses a pharmacokinetic profile that may allow oral
dosing of this compound.

5286 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 21
Eldrup et al.
Sch em e 1a
a
Reagents and conditions: (i) Dess-Martin periodinane in dichloromethane 3 days at 0 °C to room temperature; (ii) MeMgBr in diethyl
ether 7 h at -55 to -15 °C; (iii) HBr/acetic acid in dichloromethane, 4 h at 0 °C to room temperature; then sodium salt of 6 in acetonitrile,
24 h at room temperature (for 7) or sodium salt of 17, KOH, tris[2-(2-methoxyethoxy)ethyl]amine in acetonitrile, 1 h at room temperature
(for 18) or sodium salt of 28 in acetonitrile, 24 h at room temperature (for 29); (iv) boron trichloride in dichloromethane 5.5 h at -78 to
-20 °C; (v) methanolic ammonia, 14 h at 85 °C (for 9) or 1 N NaOH, 5 h at reflux (for 20) or methanolic ammonia, 14 h at 100 °C (for 31).
Sch em e 2a
Sch em e 3a
a
Reagents and conditions: (i) hydrogen bromide in dichlo-
romethane, 10 min at room temperature; (ii) 4-amino-1H-pyra-
zolo[3,4-d]pyrimidine and NaH in 1-methyl-2-pyrrolidinone, 30
min at room temperature; (iii) boron trichloride in dichloromethane
2 h at -78 °C to -30 °C.
a
Reagents and conditions: (i) hydrogen bromide in dichlo-
°C; (ii) 4,6-dichloro-1H-pyrrolo[2,3-
romethane, 30 min at
0
c]pyridine and NaH in 1-methyl-2-pyrrolidinone, 60 min at room
temperature; (iii) boron trichloride in dichloromethane 2 h at -78
to -30 °C; (iv) liquid ammonia and CuI, 24 h at 130 °C; (v) Pd/C
(10%) and concentrated aqueous ammonia in methanol under
hydrogen (1 atm).
D-ribofuranosyl)-1H-pyrrolo[3,2-c]pyridine (16), was
carried out using a similar strategy (Scheme 3). Briefly,
the bromo derivative of 5 was coupled with the anion
of dichloro-1H-pyrrolo[3,2-c]pyridine (13)15 to yield ri-
bonucleoside 14. Removal of the dichlorophenylmethyl
protection groups and ammonolysis were performed
with boron trichloride in dichloromethane and liquid
ammonia/CuI,16 respectively, to give 4-amino-6-chloro
derivative 15. Catalytic hydrogenation with Pd/C in the
presence of aqueous ammonia gave the desired com-
pound, 16. The analogous guanosine derivative of 9,
2-amino-7-(2-C-methyl-â-D-ribofuranosyl)-7H-pyrrolo-
[2,3-d]pyrimidin-4(3H)-one (20), was synthesized ac-
cording to the procedure in Scheme 1 using 2-amino-4-
chloro-7H-pyrrolo[2,3-d]pyrimidine (17) as the heterobase.

Heterobase-Modified 2′-C-Methyl Ribonucleosides
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 21 5287
Sch em e 4a
Sch em e 5a
a
Reagents and conditions: (i) N-chlorosuccinimide in DMF, 1
h at room temperature (for 25) or N-bromosuccinimide in DMF,
20 min at 0 °C, and then at room temperature for 10 min (for 26).
Sch em e 6a
a
Reagents and conditions: (i) n-BuLi in tetrahydrofuran, 30
min at -78 °C; (ii) Me3SnCl in tetrahydrofuran, 24 h at room
temperature; (iii) SELECTFLUOR in acetonitrile, 7 h at room
temperature; (iv) 23, triphenylphosphine, and DEAD in tetrahy-
drofuran, 24 h at room temperature; (v); liquid ammonia and
dioxane, 24 h at 85 °C.
Treatment of protected ribonucleoside 18 with boron
trichloride gave 19, which was converted to the desired
compound, 20, through basic hydrolysis of the 4-chloro
functionality.
a
Reagents and conditions: (i) aqueous ammonia, methanolic
ammonia and hydrogen peroxide (30%), at room temperature for
18 h.
In order to examine the effect of substitution at the 5
position of the pyrrolo[2,3-d]pyrimidine heterobase, a
series of derivatives (25-27 and 31-34) were prepared.
The synthesis of the 5-fluoro derivative (25) was ac-
complished through the synthesis of the fluoro-substi-
tuted heterobase, 4-chloro-5-fluoro-1H-pyrrolo[2,3-d]py-
rimidine (23, Scheme 4). Compound 23 could subse-
quently be coupled to the 2-C-methylribose. Lithiation
of 5-bromo-4-chloro-1H-pyrrolo[2,3-d]pyrimidine17 and
subsequent quenching of the resulting dianion with
trimethylstannane chloride gave the desired 5-trimeth-
ylstannane, 22. This was then reacted in situ with
1-(chloromethyl)-4-fluoro-1,4-diazoniabicyclo[2.2.2]-
octane bis(tetrafluoroborate) (SELECTFLUOR) to give
the desired 5-fluoro derivative, 23. A variety of condi-
tions for glycosylation of this heterobase were examined,
including coupling via the herein described bromo
derivative of 5. None of these conditions gave the desired
compound in appreciable yield. However, the coupling
of 23 with previously described 2,3,5-tri-O-benzoyl-2-
C-methylD-ribofuranose18 proceeded under Mitsunobu
conditions (triphenylphosphine and diethyl azodicar-
boxylate in tetrahydrofuran) to give the desired N-7
glycosylated compound as a mixture of R and â anomers
as well as the N-1 isomeric derivatives (Scheme 4). This
mixture was deprotected directly using liquid ammonia
at elevated temperature and subsequently separated by
preparative HPLC (see general methods) to give the
desired N-7 glycosylated â isomer 25. Electrophilic
substitutions on 4-amino-7-(â-D-ribofuranosyl)-7H-pyr-
rolo[2,3-d]pyrimidine are known to yield the corre-
sponding 5-substituted derivatives (corresponding to the
purine N-7 position).19 Hence, for the synthesis of
4-amino-5-chloro-7-(2-C-methyl-â-D-ribofuranosyl)-7H-
pyrrolo[2,3-d]pyrimidine (26), compound 9 was reacted
with 1 equiv of N-chlorosuccinimide in dimethylforma-
mide (Scheme 5). The bromo derivative 27 was prepared
from 9 by a reaction with N-bromosuccinimide under
similar conditions. 4-Amino-5-methyl-7-(2-C-methyl-â-
D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine (31) was
synthesized by a similar procedure to that used for 9
and 20 (Scheme 1). Briefly, 4-chloro-5-methyl-1H-pyr-
rolo[2,3-d]pyrimidine (28) was prepared according to
known procedures.19 The sodium salt of this compound
(formed with sodium hydride) was reacted with the
1-bromo compound, prepared from ribofuranose 5, in
acetonitrile to give exclusively the â anomer. The
intermediate 29 was subsequently deprotected with
boron trichloride in dichloromethane to give chloro
derivative 30. Finally, ammonolysis was carried out
using methanolic ammonia at elevated temperature to
furnish the desired 2′-C-methylribonucleoside 31. Two
additional derivatives, 4-amino-7-(2-C-methyl-â-D-ribo-
furanosyl)-7H-pyrrolo[2,3-d]pyrimidine-5-carbonitrile
(32) and 4-amino-7-(2-C-methyl-â-D-ribofuranosyl)-7H-
pyrrolo[2,3-d]-pyrimidine-5-carboxamide (33), were syn-
thesized as previously described.20 The 5-carboxylic acid
derivative 34 was prepared from methyl derivative
31 by oxidation with a mixture of aqueous and metha-
nolic ammonia and hydrogen peroxide as the oxidant
(Scheme 6).
To evaluate the effect of substitution of the 2-position
of the active 4-amino-7-(2-C-methyl-â-D-ribofuranosyl)-
7H-pyrrolo[2,3-d]pyrimidine (9), the corresponding 2-ami-
no and 2-fluoro derivatives were synthesized according
to the route shown in Scheme 7. Thus, 2,4-diamino-7-
(2-C-methyl-â-D-ribofuranosyl)-1H-pyrrolo[2,3-d]pyrimi-
dine (35) was readily obtained from 19 by ammonolysis
at elevated temperature. Subsequent, the selective
transformation of the more activated 2-amino group at
low temperature yielded 4-amino-2-fluoro-7-(2-C-meth-
yl-â-D-ribofuranosyl)-1H-pyrrolo[2,3-d]pyrimidine (36).
To evaluate the effect of changes in the 5 position of
pyrrolo[2,3-d]pyrimidine derivative 20, we synthesized
its 5-chloro derivative through a two-step transforma-
tion of 19, first with N-chlorosuccinimide in dimethyl-
formamide to give the 5-chloro derivative and then by

5288 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 21
Eldrup et al.
Sch em e 7a
a
Reagents and conditions: (i) concentrated aqueous ammonia, overnight at 100 °C; (ii) HF in pyridine and tert-butyl nitrite at -25 °C.
Sch em e 8a
transfer or replacement of one or more of the non-
hydrogen-bond-forming endocyclic nitrogens with car-
bon to give 4-amino-pyrrolo[2,3-d]pyrimidine (9), 4-amino-
pyrazolo[3,4-d]pyrimidine (12), and 4-amino-pyrrolo[3,2-
c]pyridine (16) substituted 2′-C-methylribonucleosides.
One of these nucleosides, 9, in which the N-7 nitrogen
of the purine had been replaced with carbon, displayed
potent inhibition of HCV RNA replication in the cell-
based assay (EC50 ) 0.25 µM). The other two nucleo-
sides, 12, in which N-7 was transferred to the 8 position,
and 16, in which both the purine N-7 and N-3 had been
replaced with carbon, had no effect on intracellular HCV
RNA levels (EC50 > 100 µM). Because 2′-C-methylguan-
osine (2) was also identified as an active inhibitor of
HCV RNA replication in our previous investigation of
purine nucleosides, we decided to evaluate the corre-
sponding guanosine mimic of 9, 2-amino-7-(2-C-methyl-
â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidin-4(3H)-
one (20). This compound tested inactive in the cell-based
assay (EC50 > 100 µM). To gain a better understanding
of the reason for the discrepancy between the inhibitory
activities of 9 and 20, we synthesized and evaluated the
corresponding triphosphates as inhibitors of HCV NS5B-
mediated RNA synthesis.5 The triphosphates of both 9
and 20 were found to be potent inhibitors of RNA
synthesis mediated by HCV NS5B (both yielding IC50
values of 0.12 µM).
On the basis of the finding that the potency of 9 is
largely unchanged relative to its 7-aza derivative 1
(EC50 ) 0.26 and 0.25 µM, respectively), we elected to
further explore the structural space around the purine
7 position (corresponding to the pyrrolo[2,3-d]pyrimidine
5-position). The effect of relatively small, electronegative
substituents such as fluoro, chloro, and bromo was
investigated by evaluation of the 5-fluoro (25), 5-chloro
(26), and 5-bromo (27) derivatives. Potent inhibition of
HCV RNA replication (EC50 ) 0.07, 0.13, and 0.24 µM,
respectively) was observed upon treatment with these
compounds. The potency of the 5-halogen-substituted
NIs declined with increasing atomic radius and decreas-
ing electronegativity of the 5-substituent, whereas a
concomitant increase in cytotoxicity was observed (CC50
(24 h) > 100, ∼50, and 40 µM for 25, 26, and 27,
respectively) with significant increases in toxicity for the
5-chloro and 5-bromo derivatives during extended cul-
ture times. For the most potent derivative (25), we
evaluated the ability of the corresponding triphosphate
to inhibit RNA synthesis mediated by HCV NS5B. The
triphosphate of 25 was found to be about equipotent
with the triphosphate of unsubstituted 9 (IC50 ) 0.19
and 0.12 µM, respectively).
a
Reagents and conditions: (i) N-chlorosuccinimide in DMF, 2
h at room temperature; (ii) 2 N aqueous NaOH, 2 h at reflux.
Ta ble 1. Inhibitory Potency (EC50) of Heterobase Modified
Nucleosides on HCV RNA Replication in a Subgenomic
Replicon Harbored in HB-1 Cells and Cytotoxicity (CC50) as
Measured by MTS
HCV replicon
MTS
compd
X
Y
Z
V
W
H
[EC50] (µM) [CC50] (µM)
1
2
CH
CH
N
N
N
N
N
N
NH2
OH NH2
NH2
NH2
0.26
3.5
>100
>100
>100
>100
>100
>100
>100
∼50
9
CH CH
N CH
CH CH
CH CH
CH CF
CH CCl
CH CBr
CH CMe
CH CCN
CH CCONH2
CH CCOOH
CH CH
H
H
H
0.25
>100
>100
>100
0.07
0.13
0.24
34
3.1
0.08
70
100
100
12
16
20
25
26
27
31
32
33
34
35
36
37
CH NH2
N
N
N
N
N
N
N
N
N
N
N
OH NH2
NH2
NH2
NH2
NH2
NH2
NH2
NH2
NH2 NH2
NH2
OH NH2
H
H
H
H
H
H
H
40
>100
∼50
∼33
>100
>100
>100
>100
CH CH
CH CCl
F
>100
alkaline hydrolysis to give 2-amino-5-chloro-7-(2-C-
methyl-â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidin-
4(3H)-one (37) (Scheme 8).
Resu lts
A series of nucleosides modified in the purine hetero-
base were synthesized and screened in a cell-based,
subgenomic replicon assay5 for their ability to inhibit
HCV RNA replication (Table 1). In addition, cytotoxicity
was evaluated in parallel in an MTS [3-(4,5-dimeth-
ylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophe-
nyl)-2H-tetrazolium]-based assay.5 In an initial set of
compounds, the heterobase of the previously identified
inhibitor, 2′-C-methyladenosine (1), was modified by the
To evaluate the effect of other substituents at this
position, we tested 4-amino-5-methyl-7-(2-C-methyl-â-

Heterobase-Modified 2′-C-Methyl Ribonucleosides
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 21 5289
D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine (31). This
methyl derivative displayed a weak inhibitory effect on
HCV RNA replication (EC50 ) 34 µM) and was not
cytotoxic at up to 100 µM, as measured in the MTS
assay. In light of the high potency of the other 5-sub-
stituted derivatives, we decided to evaluate the corre-
sponding triphosphate as an inhibitor of HCV NS5B-
mediated RNA synthesis. The triphosphate of 31 was
found to be a potent inhibitor of RNA synthesis (IC50
)
0.57 µM). The effect of larger, electronegative substit-
uents in the 5 position was further probed by the
evaluation of the carbonitrile (32), the carboxamide (33),
and the carboxylic acid (34). Compounds 32 and 33 gave
rise to moderate to potent inhibition of HCV RNA
replication (EC50 ) 3.1 and 0.08 µM, respectively),
whereas carboxylic acid 34 was significantly less potent
in the cell-based assay (EC50 ) 70 µM). Both 32 and
33, however, displayed cytotoxicity (CC50 ≈ 50 and 33
µM, respectively) in the MTS assay.
The effect on activity of substituents at the pyrrolo-
[2,3-d]pyrimidine 2-position was investigated by the
evaluation of 2,4-diaminopyrrolo[2,3-d]pyrimidine-2′-C-
methylribonucleoside (35) and 4-amino-2-fluoro-pyrrolo-
[2,3-d]pyrimidine-2′-C-methylribonucleoside (36). Both
compounds had minimal effect on reducing HCV RNA
levels in the cell-based assay (EC50 > 100 µM).
F igu r e 2. ORTEP perspective view of 9 showing the crystal-
lographic numbering scheme. The thermal ellipsoids for the
non-H atoms are drawn at the 50% level, and the H atoms
are an arbitrary size.
The high potency of the 5-substituted 4-amino-pyr-
rolo[2,3-d]pyrimidines 25-27 and 32-33 led us to
examine whether a similar type of substitution on
guanosine analogue 20 would confer activity in the cell-
based assay for HCV replication. Unfortunately, the
evaluation of the 5-chloro derivative 37 revealed no
significant effect on intracellular HCV RNA levels (EC50
> 100 µM).
Ca lcu la t ion of Con for m a t ion a l P r efer en ce for
Ribon u cleosid es 9, 12, a n d 16. The conformational
preferences of 9, 12, and 16 were calculated by molec-
ular mechanics as described in the Experimental Sec-
tion. The conformational analysis revealed no remark-
able difference among the three compounds that might
help explain the lack of potency of 12 and 16. All were
calculated to prefer a Northern, 3′-endo ribose confor-
mation by 2.5-3.0 kcal/mol over the Southern, 2′-endo
conformer as a result of the presence of the 2′-C-methyl
group and to prefer an anti conformation of the base by
a margin of 0.4-2.3 kcal/mol. Minor differences in the
shape of the heterobases were observed as a result of
the differences in the positioning of the endocyclic
nitrogens, but the differences were too subtle to explain
the dramatic differences in the potency of these ribo-
nucleosides.
F igu r e 3. Stereoview of the crystallographically determined
structure of 9 (green) and its global minimum conformation
by molecular mechanics (orange).
side chain has relatively high B factors and that its
conformation differs in the various structures, suggest-
ing a flexibility that might facilitate interaction with
appropriate inhibitor functionality. Increased potency
upon substitution with the electronegative fluorine (25)
and with a hydrogen-bonding cyano (32) or carboxamide
group (33) is consistent with direct interaction with
Lys141. The carboxylic acid derivative 34, which we
hoped could form a salt bridge with the Lys141 basic
side chain, unfortunately proved inactive. This inactiv-
ity, however, may be speculated to be the result of
decreased cellular permeability due to the molecule’s net
charge under assay conditions.
Ca lcu la tion of Con for m a tion a l P r efer en ce a n d
Solid -Sta te Str u ctu r e of 9. The solid-state structure
of 9 (Figure 2) was determined by single-crystal X-ray
crystallography. The conformational space of 9 was
evaluated using molecular mechanics with the MMFFs
force field and a dielectric constant of 50. A large
number of conformations were generated using the J G
distance geometry program and minimized with Batch-
Min. The lowest-energy conformation found was over-
laid with the crystallographically determined structure
with respect to the furanose ring atoms (Figure 3). The
RMSD for the ring atom positions was 0.017 Å, and the
Eva lu a tion of P ossible Con ta cts betw een Su b-
stitu en ts in th e P yr r olo[2,3-d ]p yr im id in e 5 P osi-
tion a n d HCV NS5B. A model of nucleoside inhibitors
in the HCV NS5B active site was created7 from the HCV
NS5B crystal structure21 and the crystal structure of
the æ6 RdRp with an initiation complex bound.22 The
steric and electrostatic environment of the inhibitor 5
position in the model was examined to help rationalize
the SAR at this position. The closest polar functionality
in the enzyme accessible from the inhibitor 5 position
is the side-chain amino group of Lys141, at a distance
of 4 Å. The examination of several overlaid HCV NS5B
crystal structures from the PDB reveals that the Lys141

5290 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 21
Eldrup et al.
F igu r e 4. Plot of pseudorotational parameter P versus relative energy in kcal mol-1 for conformations of 9 calculated by full
minimization (9) or minimization with the furanose ring conformation constrained (+). Heterobase orientation is indicated by
color (blue ) anti; red ) syn).
Ta ble 2. Phosphorolysis of the Glycosidic Bond of Adenosine,
2′-C-Methyladenosine (1), 4-Amino-7-(2-C-methyl-â-D-ribofur-
anosyl)-7H-pyrrolo[2,3-d]pyrimidine (9), and 4-Amino-5-fluoro-
7-(2-C-methyl-â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine
(25) Catalyzed by Purine Nucleoside Phosphorylase (PNP)a
Ta ble 3. Pharmacokinetic Parameters in Rat of Compound 25
Were Compared to That of 2′-C-Methyladenosine (1)
1
25
Clp (mL/min/kg)
1/2 (h)
Cmax (µM)
Tmax (h)
F (%)
>200
0.3
nd
44
T
5.1
0.6
0.5
51
% phosphorolysis by PNP
% conversion by ADA
nd
1.8 U/mL
18 U/mL 0.00625 U/mL 1.25 U/mL
0
adenosine
1
9
100
19
0
100
100
0
96
0
0
100
100
0
a
Male CRL rats were dosed iv (n ) 2) at 1.0 mg/kg and po at
2 mg/mL (N ) 3).
25
0
0
0
0
the Northern conformation of 9 that was observed
crystallographically is calculated to lie 2.5 kcal/mol
below the lowest-energy Southern conformation; the
anti base conformation is preferred to syn by 0.6 kcal/
mol.
a
The results indicate that under the experimental conditions
adenosine and 1 are substrates for PNP-mediated phosphorolysis,
whereas no conversion of pyrrolo[2,3-d]pyrimidine derivatives 9
and 25 was detected. Adenosine deaminase-catalyzed conversion
of adenosine, 2′-C-methyladenosine (1), 4-amino-7-(2-C-methyl-â-
D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine (9), and 4-amino-5-
fluoro-7-(2-C-methyl-â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimi-
dine (25). Under the reaction conditions, adenosine and 1 are
converted to inosine and 2′-C-methylinosine, respectively, whereas
no conversion of pyrrolo[2,3-d]pyrimidine derivatives 9 and 25 was
detected.
En zym a tic Con ver sion s Med ia ted by P u r in e
Nu cleosid e P h osp h or yla se a n d Ad en osin e Dea m i-
n a se. The ability of the two most potent NIs 4-amino-
7-(2-C-methyl-â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyri-
midine (9) and 4-amino-5-fluoro-7-(2-C-methyl-â-D-
ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine (25) to func-
tion as substrates for PNP and ADA were examined
(Table 2). The susceptibility of adenosine, and analogues
1, 9, and 25, to cleavage by PNP, was determined at
1.8 and 18 U/mL and reaction times from 170 h (at 4-10
°C). The positive control for the assay, adenosine, was
quantitatively cleaved while the otherwise unmodified
2′-C-methyladenosine (1) was only slowly converted. No
appreciable conversion of the pyrrolo[2,3-d]pyrimidine
derivatives 9 and 25 was observed (Table 2). Similarly,
pseudorotational parameters P and τ were in close
agreement (crystal: P ) 15.4°, τ ) 39.0°; MMFFs: P )
16.4°, τ ) 39.0°). The orientation of hydroxyl groups and
the 4-amino-7H-pyrrolo[2,3-d]pyrimidine base was also
in agreement between the modeled and experimentally
determined structures. Figure 4 depicts a plot of relative
energy versus pseudorotational angle for the conforma-
tions determined by full minimization (9) and con-
strained minimization (+). On the basis of this analysis,

Heterobase-Modified 2′-C-Methyl Ribonucleosides
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 21 5291
we evaluated the ability of these compounds to function
as substrates in the ADA-mediated conversion to the
corresponding inosine derivatives at enzyme concentra-
tions of 0.00625 and 1.25 U/mL. At the lower enzyme
concentration, adenosine was converted to inosine,
whereas 2′-C-methyladenosine (1) was not measurably
affected. At higher enzyme concentrations, however,
both adenosine and 1 were quantitatively converted to
their corresponding inosine derivatives, whereas the
pyrrolo[2,3-d]pyrimidine derivatives 9 and 25 were not
substrates for this enzyme (Table 2).
P h a r m a cok in etic Stu d ies. To evaluate the poten-
tial clinical utility of the 4-amino-7H-pyrrolo[2,3-d]-
pyrimidine ribonucleosides, we compared the pharma-
cokinetic profile in the rat of 25 to that of 2′-C-
methyladenosine 1 (Table 2). The oral bioavailablity of
25 was 51%, with a plasma clearance of 44 mL min-1
kg-1 and a half-life of 5.1 h. In comparison, plasma
levels were below the level of detection post per oral
dosing of 1 with clearance in excess of hepatic bloodflow
(>200 mL min-1 kg-1).5
The potency of 4-amino-7-(2-C-methyl-â-D-ribofura-
nosyl)-7H-pyrrolo[2,3-d]pyrimidine (9) in the cell-based
assay was not mirrored by the corresponding guanosine
mimic, 2-amino-7-(2-C-methyl-â-D-ribofuranosyl)-7H-
pyrrolo[2,3-d]pyrimidin-4(3H)-one (20), in the cell-based
assay. Despite this dramatic difference, the triphos-
phates of both 9 and 20 were equally potent inhibitors
of HCV NS5B-mediated RNA synthesis, suggesting that
the lack of potency of 20 in the cell-based assay is not
due to lack of recognition of the nucleoside triphosphate
or lack of ability of the triphosphate to function as a
“chain terminator” in the HCV NS5B-catalyzed polym-
erization reaction. In contrast, these data suggest that
the inactivity of 20 is likely due to lack of cellular uptake
and/or inefficient intracellular metabolism to the tri-
phosphate. This conclusion is supported by experiments
in which radiolabled compound was used to demonstrate
that uptake and metabolism to the triphosphate in
replicon-containing cells are inefficient for compound 20
(data not shown).
The similar potencies of 2′-C-methyladenosine (1) and
its pyrrolo[2,3-d]pyrimidine analogue (9) indicated some
tolerance for changes in the stereoelectronic features of
the pyrrolo[2,3-d]pyrimidine 5 position (corresponding
to the purine 7 position). The extent of this tolerance
was further defined by the activities of the ribonucleo-
side derivatives substituted with fluoro (25), chloro (26),
bromo (27), nitrile (32), and carbamoyl (33) in the
pyrrolo[2,3-d]pyrimidine 5 position. The potency of the
5-halogen substituted 2′-C-methyl ribonucleosides de-
clined with increasing atomic radius of the substituent,
whereas the cytotoxicity was found to increase, leaving
a narrow 100-fold window between activity and cyto-
toxicity for 5-bromo derivative (27). The high potency
of the 5-fluoro derivative in the cell-based assay repre-
sents an approximate 4-fold improvement relative to the
nonsubstituted derivative (9). In contrast, the intrinsic
potency of 25 was slightly less than that of 9. This
apparent discrepancy was addressed by experiments in
which a radiolabled compound was used to demonstrate
a more efficient uptake and metabolism to the triphos-
phate in the replicon-containing cells for compound 25
as compared to those of 9 (data not shown). Despite this
difference, cellular uptake and metabolism to their
respective triphosphates were generally efficient for
these adenosine analogues and, hence, in line with their
relatively high potencies in the cell-based assay as
compared to their intrinsic potencies. The addition of a
carbamoyl substituent resulted in a compound that was
about equipotent to the fluoro derivative but that
displayed significant cytotoxicity. A similar increase in
cytotoxicity was seen for the nitrile derivative. Notably,
the 5-methyl substituted derivative (31) was weakly
inhibitory on HCV RNA replication in the cell-based
assay, whereas the corresponding triphosphate was a
potent inhibitor of RNA synthesis by HCV NS5B. This
pattern of activity is remarkably similar to that seen
with guanosine mimic 20. Thus, the presence of the
nonpolar methyl substituent in the pyrrolo[2,3-d]pyri-
midine 5 position appears to interfere with cellular
uptake mechanisms and/or with the efficiency of intra-
cellular conversion to the triphosphate. Likewise, the
negative charge of the carboxylic acid derivative (34)
may impede transport across the cell membrane and/
Discu ssion
2′-C-Methyladenosine is a potent inhibitor of HCV
RNA replication in vitro but is not orally bioavailable
in rats.5 We have suggested that the lack of oral
bioavailability may be due in large part to the ability
of 2′-C-methyladenosine to function as a substrate for
PNP and ADA.8 In light of these observations, we
decided to examine 2′-C-methylribonucleosides contain-
ing heterobases with altered placement and number of
non-hydrogen-bond-forming, endocyclic nitrogens for
their ability to inhibit HCV RNA replication. This
strategy for improving the activity and bioavailability
was supported by previous studies, in which it was
shown that nucleoside analogues containing 4-amino-
pyrrolo[2,3-d]pyrimidine heterobases were inert to ADA-
mediated hydrolysis.23 We suspected that other meta-
bolic enzymes that act on purine ribonucleosides might
display similar sensitivity to subtle changes in the
purine heterobase.
The replacement of the 6-aminopurine heterobase of
the potent nucleoside inhibitor 1 with 4-amino-pyrrolo-
[2,3-d]pyrimidine generated an equipotent inhibitor (9)
of HCV RNA replication. However, two closely related
ribonucleosides, in which 6-aminopurine was substi-
tuted with either 4-amino-pyrazolo[3,4-d]pyrimidine
(12) or 4-amino-pyrrolo[3,2-c]pyridine (16), lacked any
ability to attenuate HCV RNA levels. In an effort to
better understand the inactivities of ribonucleosides 12
and 16, we calculated the conformational preferences
for both compounds, as well as for 9. The calculated
conformational preference of 9 was in agreement with
the solid-state structure determined by single-crystal
X-ray crystallography and is similar to the previously
described conformational preference for the correspond-
ing purine derivatives. The computations of conforma-
tional preference for 12 and 16 revealed very limited
differences in the shape of the heterobases relative to
9. These differences did not significantly affect the syn/
anti position of the heterobase nor the north/south
preference of the 2′-C-methylribose. Hence, our compu-
tations do not explain the observed inactivity of 12 and
16.

5292 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 21
Eldrup et al.
or interfere with enzymes involved in the requisite
conversion to the triphosphate, giving rise to its poor
activity in the HCV replicon assay.
as illustrated by the oral bioavailability of the fluori-
nated derivative 25 in the rat. Hence, in contrast to 2′-
C-methyladenosine, the 4-amino-7-(2-C-methyl-â-D-
ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine inhibitor class
displays pharmacokinetic profiles consistent with clini-
cal utility.
The examination of the 5-substituted pyrrolo[2,3-d]-
pyrimidine NIs in the model created from the crystal
structure of HCV NS5B and a crystal structure of the
æ6 RdRp bound to the æ6 initiation complex supports
the observed activity of the fluorine-, nitrile-, and
carbamoyl-substituted derivatives on the basis of a
suggested interaction with Lys141. The model does not
indicate a very tight environment around the pyrrolo-
[2,3-d]pyrimidine 5 position and, hence, does not con-
tradict the observed intrinsic potency of the triphos-
phate of 5-methyl derivative 31.
Exp er im en ta l Section
Gen er a l P r oced u r es. 4-Chloro-1H-pyrrolo[2,3-d]pyrimi-
dine,14 2-amino-4-chloro-1H-pyrrolo[2,3-d]pyrimidine,24 2,4-
dichloro-1H-pyrrolo[3,2-c]pyridine,15 and 4-chloro-5-methyl-1H-
pyrrolo[2,3-d]pyrimidine19 were synthesized according to existing
literature procedures. 4-Amino-1H-pyrazolo[3,4-d]pyrimidine
(purchased from Aldrich A7,780-6) and 3,5-bis-O-(2,4-dichlo-
rophenylmethyl)-1-O-methyl-R-D-ribofuranose (Lipomed NUC-
55) were used as received. TLC was performed on silica 60
(Merck 5554 aluminum sheet), column chromatograpy on silica
60 (230-400-mesh ASTM) (Merck 9385). Molecular sieves (4
Å, Mallinckrodt 4494) were activated prior to use (120 °C in
vacuo). 1H and 13C NMR spectra were obtained at 200 MHz
(Varian Mercury VX) in 5-mm tubes unless otherwise indi-
cated; chemical shifts are positive in the low-field direction.
FAB mass spectra were recorded on a J EOL Hx110/110 mass
spectrometer. High-resolution FAB spectra were obtained from
UC Berkeley. All nucleosides screened in the cell-based assay
were purified by preparative RP HPLC using a reverse-phase
column (Phenomenex, Luna 10 µm, 21.2 mm × 250 mm, A )
water, B ) acetonitrile, 2 to 95% B in 60 min, flow 3 mL
min-1). Nucleosides were analyzed for purity by analytical RP
HPLC (Phenominex C18 aqua 5 µm, 150 mm × 4.6-mm column
A ) 100 mM triethylammoniumacetate, pH 7, B ) 10% 100
mM triethylammoniumacetate in acetonitrile, 100% B in 25
min., flow 1 mL min-1). The purity of the nucleosides screened
was generally greater than 98%.
Assa y for In h ibition of HCV RNA Rep lica tion in Cells.
Inhibition of HCV RNA replication was assayed in a subge-
nomic bicistronic replicon assay in HB-1 cells by in situ RPA
as previously described.5 The replicon EC50 values are the
average of at least two separate potency determinations at 24
h (Table 1).
Assa y for In h ibition of NS5B-Med ia ted RNA Syn th esis
in Vitr o. Inhibition of the enzymatic activity of HCV NS5B∆21
by the corresponding nucleoside triphosphates was determined
as previously described.5 The IC50 values for all active com-
pounds are the averages of at least three separate determina-
tions.
Assa y for P NP -Med ia ted P h osp h or olysis. Assay reac-
tions (100 mM HEPES pH 7.0, 50 mM sodium phosphate pH
7.0, 200 µM nucleoside) were initiated with addition of 1.8 or
18 U/mL PNP (Sigma N-3514, isolated from human blood).
The reactions were incubated at 4 °C for 170 h. The RP-HPLC
analysis utilized a C18 column (Vydac C218TP, 150 mm × 4.6
mm, 5 µM) and guard column (Perkin-Elmer 0711-0092, 15
mm × 3.2 mm, 7 µM) with 10 mM potassium phosphate, 2
mM TBA (buffer A) and 50% methanol, 10 mM potassium
phosphate, 2 mM TBA (buffer B), 1 mL/min flow rate. A 50-
µL reaction volume injection and elution over the gradient
yielded quantifiable nucleoside peaks (1-5% B over 15 min;
5-75% B over 2.5 min; hold 75% for 1 min), monitoring at
260 nm.
Assa y for ADA-Med ia ted Con ver sion . Assay reactions
included 5 mM Tris-HCl pH 7.4, 5 mM potassium phosphate,
and 0.00625 or 1.25 U/mL adenosine deaminase (ADA, Sigma
A-5168, type IX from calf spleen). The reactions were initiated
with addition of 25 µM nucleoside. The reactions were incu-
bated at room temperature for 30 min (0.00625 U/mL ADA)
or 24 h (1.25 U/mL ADA) before the enzyme was heat
inactivated (75 °C for 15 min). The RP-HPLC analysis utilized
a SUPELCOSIL LC-18-S column (Supelco 58931, 150 mm ×
4.6 mm, 5 µM) with 97.5%/2.5% 50 mM potassium phosphate
pH 4.4/methanol (buffer A) and 80%/20% 50 mM potassium
phosphate pH 4.4/methanol (buffer B), 1 mL/min flow rate. A
250-µL reaction volume injection and elution over the gradient
The purine 2 position of adenosine is not utilized for
Watson-Crick hydrogen-bond formation, and hence, we
considered it possible that modification at this position
may yield compounds with significant antiviral potency.
However, even a very conservative change from hydro-
gen to fluoro at C2 on active nucleoside 9 rendered
halogenated derivative 36 inactive in the replicon assay,
suggesting that this position is not available for modi-
fication.
The findings presented here demonstrate that the
4-amino-pyrrolo[2,3-d]pyrimidine heterocycle is a good
bioisosteric replacement for the endogenous adenine
with respect to cell penetration, phosphorylation, and
subsequent chain termination catalyzed by HCV NS5B
polymerase. In contrast, the 4-amino-pyrazolo[3,4-d]-
pyrimidine and 4-amino-pyrrolo[3,2-c]pyridine hetero-
cycles are inefficient bioisosters for adenine. In addition,
our observations reveal a remarkable intolerance of
HCV NS5B to changes in the non-Watson-Crick hy-
drogen bond participating 2 and 6 positions of the
4-amino-pyrrolo[2,3-d]pyrimidine heterocycle. These re-
sults are in agreement with the results previously
obtained in our laboratories, suggesting that changes
in the 2 and 8 positions of 2′-C-methyladenosine did not
generally lead to compounds that displayed significant
inhibitory potency.8 The substitution of the 4-amino-
pyrrolo[2,3-d]pyrimidine 5 position with polar substit-
uents is tolerated and leads to increased potency when
the substituents are small and polar/electronegative.
Furthermore, the inactivity of the 2-amino-5-chloro-7-
(2-C-methyl-â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimi-
din-4(3H)-one derivative (37) shows that the inhibitory
activities of the 5-substituted 4-amino-pyrrolo[2,3-d]-
pyrimidine derivatives on HCV RNA replication do not
generally extend to the 2-amino-pyrrolo[2,3-d]pyrimi-
din-4(3H)-one series. With respect to the potency of the
4-amino-pyrrolo[2,3-d]pyrimidine derivatives, 4-amino-
5-fluoro-pyrrolo[2,3-d]pyrimidine appears to be an es-
pecially effective bioisosteric replacement for adenine.
The selectivity indexes for 4-amino-7-(2-C-methyl-â-D-
ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine (9) and 4-
amino-5-fluoro-7-(2-C-methyl-â-D-ribofuranosyl)-7H-pyr-
rolo[2,3-d]pyrimidine (25) are both >800-fold. Most
importantly, bioisosteric replacement of adenine with
4-amino-pyrrolo[2,3-d]pyrimidine or 4-amino-5-fluoro-
pyrrolo[2,3-d]pyrimidine abolishes the ability of the
resulting 2′-C-methyl ribonucleosides to function as
substrates for metabolic conversions mediated by ADA
and PNP. The differences in susceptibility to enzymatic
conversions lead to improved pharmacokinetic profiles,

Heterobase-Modified 2′-C-Methyl Ribonucleosides
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 21 5293
yielded quantifiable nucleoside peaks (0% B hold for 3 min,
0-100% B over 9 min, 100% hold for 3 min, 100% to 60% over
4.5 min) monitoring at 254 and 280 nm. Injections of adenosine
and inosine yielded linear standard curves allowing for the
quantitation of substrate utilization and product formation.
P h a r m a cok in etic Stu d ies. To evaluate the potential
clinical utility of the 4-amino-7H-pyrrolo[2,3-d]pyrimidine
ribonucleosides, the pharmacokinetic profile in the rat of the
most potent compound, 25, was compared to that of 2′-C-
methyladenosine 1 (Table 2). Male CRL rats were dosed iv (N
) 2) at 1.0 mg/kg and po at 2 mg/mL (N ) 3). Plasma samples
were obtained and analyzed as previously described.7
and brine (600 mL), dried over MgSO4, filtered, and evaporated
in vacuo to give the title compound (34.2 g, 95%) as a colorless
oil, which was used without further purification in the next
1
step. H NMR (CDCl3): δ 3.50 (s, 3H), 3.79 (dd, J ) 11.3 Hz,
3.5 Hz, 1H), 3.94 (dd, J ) 11.3 Hz, 2.3 Hz, 1H), 4.20 (dd, J )
1.3 Hz, 8.4 Hz, 1H), 4.37 (ddd, J ) 8.4 Hz, 3.5 Hz, 2.3 Hz,
1H), 4.58, 4.69 (2d, J ) 13.0 Hz, 2H), 4.87 (d, J ) 1.3 Hz, 1H),
4.78, 5.03 (2d, J ) 12.5 Hz, 2H), 7.19-7.26, 7.31-7.42 (2m,
6H).
3,5-Bis-O-(2,4-d ich lor op h en ylm eth yl)-2-C-m eth yl-1-O-
m eth yl-r-D-r ibofu r a n ose (5). To a solution of methylmag-
nesium bromide in diethyl ether (0.48 M, 300 mL) at -55 °C
was added dropwise a solution of 3,5-bis-O-(2,4-dichloro-
phenylmethyl)-1-O-methyl-R-D-erythro-pentofuranos-2-ulose (4)
(17.40 g, 36.2 mmol) in diethyl ether (125 mL). The reaction
mixture was allowed to warm to -30 °C and was stirred for 7
h at -30 to -15 °C, then poured into ice-cold water (500 mL),
and the mixture vigorously stirred at room temperature for
0.5 h. The mixture was filtered through a Celite pad which
was thoroughly washed with diethyl ether. The organic layer
was dried (MgSO4), filtered, and evaporated in vacuo. The
residue was purified on silica gel using hexanes and hexanes/
ethyl acetate (9:1) as the eluent to give the title compound (16.7
Ca lcu la tion s. Nucleoside conformational preferences were
calculated by molecular mechanics using the MMFFs force
field and a dielectric constant of 50. For each nucleoside, 1000
conformers were generated using the J G distance geometry
program.25 and were minimized to low gradient using Batch-
Min.26 To estimate the barrier to interconversion between the
Northern and Southern conformers and the steepness of the
minima, the conformers generated by distance geometry were
also subjected to constrained minimization, holding one of the
ribose ring dihedral angles at a random value between -42°
and +42° by means of a harmonic force constant of 1000 kJ
mol-1 rad-2
.
1
g, 93%) as a colorless oil. H NMR (CDCl3): δ 1.36 (d, J ) 0.9
Hz, 3H), 3.33 (q, J ) 0.9 Hz, 1H), 3.41 (d,, J ) 3.3 Hz, 1H),
3.46 (s, 3H), 3.66 (d, J ) 3.7 Hz, 2H), 4.18 (apparent q, 1H),
4.52 (s, 1H), 4.60 (s, 2H), 4.63, 4.81 (2d, J ) 13.2 Hz, 2H),
Solid -Sta te Str u ctu r e of 9. The structure of compound 9,
12H16N4O4, has been determined by single-crystal X-ray
C
crystallography. Diffraction-quality crystals were grown by the
slow evaporation of a CH3NO2/CH3CH2OH solution of the
compound, and the crystals that were obtained are orthor-
hombic with space group P21212 and cell constants of a )
16.330(6) Å, b ) 8.353(6) Å, c ) 8.78(1) Å, V ) 1198(2) Å3,
7.19-7.26, 7.34-7.43 (2m, 6H). HRMS FAB: calcd for C21H21
-
Cl4O4 + Li+ 501.0381, found 501.0378.
4-Ch lor o-7-[3,5-bis-O-(2,4-d ich lor op h en ylm eth yl)-2-C-
m e t h yl-â-D -r ib o fu r a n o syl]-7H -p y r r olo[2,3-d ]p y r im i-
d in e (7). To a solution of 3,5-bis-O-(2,4-dichlorophenylmethyl)-
2-C-methyl-1-O-methyl-R-D-ribofuranose (5) (9.42 g, 19 mmol)
in anhydrous dichloromethane (285 mL) at 0 °C was added
HBr (5.7 M in acetic acid, 20 mL, 114 mmol) dropwise. The
resulting solution was stirred at 0 °C for 1 h and then at room
temperature for 3 h, evaporated in vacuo, and coevaporated
with anhydrous toluene (3 × 40 mL). The oily residue was
dissolved in anhydrous acetonitrile (50 mL) and added to a
solution of the sodium salt of 4-chloro-1H-pyrrolo[2,3-d]pyri-
midine (6) in acetonitrile (generated in situ from 4-chloro-1H-
pyrrolo[2,3-d]pyrimidine (8.76 g, 57 mmol) in anhydrous
acetonitrile (1000 mL), and NaH (60% in mineral oil, 2.28 g,
57 mmol), after 4 h of vigorous stirring at room temperature).
The combined mixture was stirred at room temperature for 1
day and then evaporated to dryness. The residue was sus-
pended in water (250 mL) and extracted with ethyl acetate (2
× 500 mL). The combined extracts were washed with brine
(300 mL), dried over Na2SO4, filtered, and evaporated. The
crude product was purified on silica gel using ethyl acetate/
hexane (1:3 and 1:2) as the eluent. Fractions containing the
product were combined and evaporated in vacuo to give the
desired product (5.05 g, 43%) as colorless foam. 1H NMR
(CDCl3): δ 0.93 (s, 3H), 3.09 (s, 1H), 3.78 (dd, J ) 10.9 Hz, 2.5
Hz, 1H), 3.99 (dd, J ) 10.9 Hz, 2.2 Hz, 1H), 4.23-4.34 (m,
2H), 4.63, 4.70 (2d, J ) 12.7 Hz, 2H), 4.71, 4.80 (2d, J ) 12.1
Hz, 2H), 6.41 (s, 1H), 6.54 (d, J ) 3.8 Hz, 1H), 7.23-7.44 (m,
6H), 7.69 (d, J ) 3.8 Hz, 1H), 8.64 (s, 1H). HRMS FAB: calcd
for C26H22Cl5N3O4 + H+ 616.0131, found 616.0118.
4-Ch lor o-7-(2-C-m eth yl-â-D-r ibofu r a n osyl)-7H-p yr r olo-
[2,3-d ]p yr im id in e (8). To a solution of 4-chloro-7-[3,5-bis-O-
(2,4-dichlorophenylmethyl)-2-C-methyl-â-D-ribofuranosyl]-7H-
pyrrolo[2,3-d]pyrimidine (7) (5.42 g, 8.8 mmol) in dichlorometh-
ane (175 mL) at -78 °C was added boron trichloride (1 M in
dichloromethane, 88 mL, 88 mmol) dropwise. The mixture was
stirred at -78 °C for 2.5 h, and then at -30 to -20 °C for 3 h.
The reaction was quenched by addition of methanol/dichlo-
romethane (1:1) (90 mL), and the resulting mixture was stirred
at -15 °C for 30 min, then neutralized with aqueous ammonia
at 0 °C, and stirred at room temperature for 15 min. The solid
was filtered and washed with dichloromethane/methanol
(1:1, 250 mL). The combined filtrate was evaporated, and the
residue was purified by flash chromatography on silica gel
and Z ) 4. The calculated density is 1.555 g cm-3
.
All diffraction measurements were made using monochro-
matized Mo KR radiation (λ ) 0.71073 Å) on a CCD area-
detector equipped diffractometer, at T ) 100 K, to a θ limit of
26.48°. There are 2472 unique reflections out of 12 981
measured, with 2127 observed at the I g 2σ(I) level. The
structure was solved by direct methods and was refined using
full-matrix least-squares on F2 using 185 parameters and all
unique reflections. The refinement converged with agreement
statistics of R ) 0.035, wR ) 0.080, S ) 1.01, (∆/σ)max e 0.01.
A computer-generated perspective view of the molecule is
shown in Figure 2. Tables of the atomic parameters for the
model along with interatomic distances and angles are avail-
able in the Supporting Information. The atomic coordinates
for this structure have been deposited with the Cambridge
Crystallographic Data Centre (deposition no. 245444) and can
be obtained, on request, from data_request@ccdc.cam.ac.uk or
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, U.K.
Syn t h esis, P u r ifica t ion , a n d QC of Nu cleosid e Tr i-
p h osp h a tes. Nucleoside triphosphates were synthesized as
previously described.8 Purification was performed by anion
exchange chromatography, and appropriate fractions were
subsequently desalted by reverse-phase HPLC as described.8
Purity was determined by analytical reverse-phase HPLC and
anion exhange HPLC. The identities of the nucleoside triph-
osphates were confirmed by mass spectrometry as described.8
The yield of purified nucleoside triphosphate following this
procedure ranged from 5 to 15 mg. Purity of >90% was
generally attained.
3,5-Bis-O-(2,4-d ich lor op h en ylm eth yl)-1-O-m eth yl-r-D-
er yth r o-p en tofu r a n os-2-u lose (4). To an precooled (0 °C)
suspension of Dess-Martin periodinane (50.0 g, 118 mmol)
in dichloromethane (350 mL) was added a solution of 3,5-bis-
O-(2,4-dichlorophenylmethyl)-1-O-methyl-R-D-ribofuranose (3)
(36.2 g, 75 mmol) in dichloromethane (200 mL) dropwise over
0.5 h. The reaction mixture was stirred at 0 °C for 0.5 h and
then at room temperature for 3 days. The mixture was diluted
with diethyl ether (600 mL) and poured into a precooled (0
°C) mixture of Na2S2O3‚5H2O (180 g) in saturated aqueous
sodium hydrogencarbonate (1400 mL). The layers were sepa-
rated, and the organic layer was washed with saturated
aqueous sodium hydrogencarbonate (600 mL), water (800 mL),
using dichloromethane and
a dichloromethane/methanol

5294 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 21
Eldrup et al.
(99:1, 98:2, 95:5, and 90:10) gradient as the eluent to furnish
the desired compound (1.73 g, 66%) as colorless foam, which
turned into an amorphous solid after treatment with aceto-
nitrile. 1H NMR (DMSO-d6): δ 0.64 (s, 3H), 3.61-3.71 (m, 1H),
3.79-3.88 (m, 1H), 3.89-4.01 (m, 2H), 5.15-5.23 (m, 3H), 6.24
(s, 1H), 6.72 (d, J ) 3.8 Hz, 1H), 8.13 (d, J ) 3.8 Hz, 1H), 8.65
(s, 1H). HRMS FAB: calcd for C13H18N4O4 + H+ 300.0751,
found 300.0736.
the solution was saturated (5 min). The resulting solution was
stirred at 0 °C for 30 min, evaporated in vacuo, and coevapo-
rated with anhydrous toluene (20 mL). To this compound was
added the sodium salt of 4,6-dichloro-1H-pyrrolo[3,2-c]pyridine
(13) in 1-methyl-2-pyrrolidinone (generated in situ from 4,6-
dichloro-1H-pyrrolo[3,2-c]pyridine (1.13 g, 6.04 mmol) in an-
hydrous 1-methyl-2-pyrrolidinone (50 mL) and NaH (60%, 0.26
g, 6.50 mmol) after 1 h of vigorous stirring at room tempera-
ture). The resulting solution was stirred overnight at room
temperature. The mixture was diluted with toluene (200 mL)
and filtered to remove any solid. The filtrate was washed with
brine (3 × 100 mL) and was then concentrated under reduced
pressure. The residue was purified on silica gel using ethyl
acetate/hexane (20:80) to give the desired compound (0.98 g,
37%) as a colorless solid. 1H NMR (CDCl3): δ 0.85 (s, 3H), 3.14
(s, 1H), 3.76 (m, 1H), 3.95 (m, 1H), 4.22 (m, 1H), 4.60-4.76
(m, 5H), 5.85 (s, 1H), 6.58 (d, J ) 3.4 Hz, 1H), 7.22-7.45 (m,
7H), 7.54 (d, J ) 3.5 Hz, 1H). HRMS: calcd for C27H22Cl6N2O4
+ H 648.9787, found 648.9783.
4-Am in o-7-(2-C-m eth yl-â-D-r ibofu r a n osyl)-7H-p yr r olo-
[2,3-d ]p yr im id in e (9). To 4-chloro-7-(2-C-methyl-â-D-ribo-
furanosyl)-7H-pyrrolo[2,3-d]pyrimidine (8) (1.54 g, 5.1 mmol)
was added methanolic ammonia (saturated at 0 °C) (150 mL).
The mixture was heated in a stainless steel autoclave at 85
°C for 14 h and then cooled and evaporated in vacuo. The crude
mixture was purified on silica gel using dichloromethane/
methanol (9:1) as the eluent to give the title compound as a
colorless foam (0.80 g, 56%), which separated as an amorphous
solid after treatment with acetonitrile. The amorphous solid
1
was recrystallized from methanol/acetonitrile; mp 222 °C. H
NMR (DMSO-d6): δ 0.62 (s, 3H), 3.57-3.67 (m, 1H), 3.75-
3.97 (m, 3H), 5.00 (s, 1H), 5.04 (d, J ) 6.8 Hz, 1H), 5.06 (t, J
) 5.1 Hz, 1H), 6.11 (s, 1H), 6.54 (d, J ) 3.6 Hz, 1H), 6.97 (br
s, 2H), 7.44 (d, J ) 3.6 Hz, 1H), 8.02 (s, 1H). HRMS FAB: calcd
for C12H16N4O4 + H+ 281.1250, found 281.1243.
4-Am in o-6-ch lor o-1-(2-C-m eth yl-â-D-r ibofu r a n osyl)-1H-
p yr r olo[3,2-c]p yr id in e (15). To 4,6-dichloro-1-[3,5-bis-O-(2,4-
dichlorophenylmethyl)-2-C-methyl-â-D-ribofuranosyl]-1H-pyr-
rolo[3,2-c]pyridine (14) (0.65 g, 1.0 mmol) in dichloromethane
(10 mL) at -78 °C was added boron trichloride (1 M in
dichloromethane, 10 mL, 10 mmol) dropwise. The mixture was
stirred at -78 °C for 0.5 h and then at -45 to -30 °C for 2 h.
The reaction was quenched by addition of sodium acetate (3.0
g) and methanol (20 mL). The solution was evaporated, and
the residue was purified by flash chromatography over silica
gel using dichloromethane and then methanol/dichloromethane
(5:95) as the eluents to furnish the desired compound (200 mg,
64%) as slightly tan solid. The above solid was heated with
liquid ammonia (ca. 10 mL) and CuI (0.1 g) at 130 °C for 24 h.
The mixture was evaporated, and the residue was purified by
column chromatography on silica gel. The desired product was
eluted with methanol/dichloromethane (5:95 to 10:90) to afford
a colorless solid (120 mg, 38% for the two steps). 1H NMR
(DMSO-d6): δ 0.60 (s, 3H), 3.65 (m, 1H), 3.78 (m, 3H), 5.00
(d, J ) 6.1 Hz, 1H), 5.08 (t, J ) 4.9 Hz, 1H), 5.19 (s, 1H), 5.73
(s, 1H), 6.55 (s, 2H), 6.63 (d, J ) 3.0 Hz, 1H), 6.82 (s, 1H),
4-Am in o-1-[3,5-bis-O-(2,4-d ich lor op h en ylm eth yl)-2-C-
m e t h yl-â-D-r ib ofu r a n osyl]-1H -p yr a zolo[3,4-d ]p yr im i-
d in e (11). 3,5-Bis-O-(2,4-dichlorophenylmethyl)-2-C-methyl-
1-O-methyl-R-D-ribofuranose (5) (1.00 g, 2.02 mmol) was
dissolved in dichloromethane (20 mL), and HBr gas was
bubbled in the solution for 5 min until it was saturated. The
resulting solution was stirred at room temperature for 10 min,
evaporated in vacuo, and coevaporated with anhydrous toluene
(10 mL). To this compound was added the sodium salt of
4-amino-1H-pyrazolo[3,4-d]pyrimidine (10) in 1-methyl-2-pyr-
rolidinone (generated in situ from 4-amino-1H-pyrazolo[3,4-
d]pyrimidine (0.43 g, 3.18 mmol) in anhydrous 1-methyl-2-
pyrrolidinone (10 mL) and NaH (60% in mineral oil, 150 mg,
3.8 mmol) after 30 min of vigorous stirring at room temper-
ature). The resulting solution was stirred overnight at room
temperature. The mixture was diluted with toluene (50 mL),
washed with brine (3 × 50 mL), and concentrated in vacuo.
The residue was purified on silica gel using ethyl acetate as
the eluent. Fractions containing the product were pooled and
evaporated in vacuo to give the desired compound (400 mg,
7.49 (d, J ) 3.4 1H). HRMS FAB: calcd for C13H16ClN3O4
+
H 314.0908, found 314.0902.
4-Am in o-1-(2-C-m eth yl-â-D-r ibofu r a n osyl)-1H-p yr r olo-
[3,2-c]p yr id in e (16). 4-Amino-6-chloro-1-(2-C-methyl-â-D-ri-
bofuranosyl)-1H-pyrrolo[3,2-c]pyridine (15) (40 mg, 0.13 mmol),
concentrated aqueous ammonia (0.5 mL), and Pd/C (10%, 20
mg) in methanol (5 mL) were stirred under a normal atmo-
sphere of hydrogen overnight. The mixture was filtered and
evaporated. The residue was purified by preparative reverse-
phase HPLC using a linear gradient of acetonitrile-water
(1:99-15:85 in 15 min, C18 column, 9 mL/min) to give a
1
33%) as a colorless solid. H NMR (DMSO-d6): δ 0.97 (s, 3H),
3.77 (m, 2H), 4.23 (m, 1H), 4.42 (m, 1H), 4.49 (s, 2H), 4.65 (m,
2H), 5.48 (s, 1H), 6.14 (s, 1H), 7.13-7.63 (m, 6H), 7.78 (br s,
2H), 8.19 (s, 1H), 8.20 (s, 1H). HRMS FAB: calcd. for C25H23
-
Cl4N5O4 + H 598.0582, found 598.0581.
4-Am in o-7-(2-C-m eth yl-â-D-r ibofu r an osyl)-1H-pyr azolo-
[3,4-d ]p yr im id in e (12). To a solution of 4-amino-1-[3,5-bis-
O-(2,4-dichlorophenylmethyl)-2-C-methyl-â-D-ribofuranosyl]-
1H-pyrazolo[3,4-d]pyrimidine (11) (0.20 g, 0.33 mmol) in
dichloromethane (10 mL) at -78 °C was added boron trichlo-
ride (1 M in dichloromethane) (3.0 mL, 3.0 mmol) dropwise.
The mixture was stirred at -78 °C for 0.5 h and then at -45
to -30 °C for 2 h. The reaction was quenched by addition of
sodium acetate (1.0 g) and methanol (10 mL). The solution was
evaporated in vacuo, and the residue was purified by flash
chromatography over silica gel using dichloromethane and
dichloromethane/methanol (95:5 to 90:10) as the eluents to
furnish the desired compound (60 mg, 64%) as slightly tan
solid. This compound was recrystallized from methanol and
acetonitrile to give the desired compound (40 mg, 41%) as an
1
colorless solid (20 mg, 56%). H NMR (DMSO-d6): δ 0.62 (s,
3H), 3.67 (m, 1H), 3.88 (m, 3H), 5.16 (m, 2H), 5.34 (s, 1H),
5.95 (s, 1H), 7.06 (d, J ) 3.2 Hz, 1H), 7.25 (d, J ) 7.1 Hz, 1H),
7.60 (d, J ) 7.1 Hz, 1H), 7.85 (d, J ) 3.4 Hz, 1H). 8.15 (s, 2H).
HRMS FAB: calcd for C13H17N3O4 + H 280.1297, found
280.1290.
2-Am in o-4-ch lor o-7-[3,5-b is-O-(2,4-d ich lor op h e n yl-
m eth yl)-2-C-m eth yl-â-D-r ibofu r a n osyl)-7H-p yr r olo[2,3-d ]-
p yr im id in e (18). To a precooled (0 °C) 3,5-bis-O-(2,4-dichlo-
rophenylmethyl)-2-C-methyl-1-O-methyl-R-D-ribofuranose (5)
(1.27 g, 2.57 mmol) in dichloromethane (30 mL) was added
HBr (5.7 M in acetic acid; 3 mL) dropwise. The reaction
mixture was stirred at room temperature for 2 h, concentrated
in vacuo, and coevaporated with toluene (2 × 15 mL). The
resulting oil was dissolved in acetonitrile (15 mL) and added
dropwise into a well-stirred mixture of 2-amino-4-chloro-7H-
pyrrolo[2,3-d]pyrimidine (433 mg, 2.57 mmol), KOH (85%,
powdered) (0.51 g, 7.7 mmol), and tris[2-(2-methoxyethoxy)-
ethyl]amine (165 µL, 0.51 mmol) in acetonitrile (30 mL). The
resulting mixture was stirred at room temperature for 1 h,
filtered, and evaporated in vacuo. The residue was purified
on a silica gel column using hexanes/ethyl acetate (5:1, 3:1,
and 2:1) as the eluent to give the title compound as a colorless
1
off-white solid. H NMR (DMSO-d6): δ 0.75 (s, 3H), 3.59 (m,
1H), 3.69 (m, 1H), 3.91 (m, 1H), 4.12 (m, 1H), 4.69 (t, J ) 5.1
Hz, 1H), 5.15 (m, 2H), 6.13 (s, 1H), 7.68-7.96 (br, 2H,), 8.18
(s, 1H), 8.21 (s, 1H). HRMS FAB: calcd for C11H15N5O4 + H
282.1202, found 282.1203.
4,6-Dich lor o-1-[3,5-bis-O-(2,4-d ich lor op h en ylm eth yl)-
2-C-m e t h yl-â-D-r ib ofu r a n osyl]-1H -p yr r olo[3,2-c]p yr i-
d in e (14). Through a solution of 3,5-bis-O-(2,4-dichlorophen-
ylmethyl)-2-C-methyl-1-O-methyl-R-D-ribofuranose (5) (2.00 g,
4.04 mmol) in dichloromethane (50 mL) was bubbled HBr until

Heterobase-Modified 2′-C-Methyl Ribonucleosides
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 21 5295
foam (0.65 g, 40%). 1H NMR (CDCl3): δ 0.97 (s, 3H 3.11 (br s,
1H), 3.74 (dd, J ) 11.0 Hz, 2.5 Hz, 1H), 3.95 (dd, J ) 11.0 Hz,
1.6 Hz, 1H), 4.25-4.19 (m, 2H), 4.56-4.82 (4d, 4H), 4.98 (br
s, 2H), 6.21 (s, 1H), 6.31 (d, J 3.8 Hz, 1H), 7.22-4.43 (m, 7H).
HRMS: calcd for C26H23Cl5N4O4 + H+ 631.0240, found 631.0231.
silica gel using methanol/dichloromethane (1:9) as the eluent
to give the title compound (0.008 g, 5%). H NMR (methanol-
1
d4): δ 8.07 (s, 1H), 7.41 (d, J ) 2.2 Hz, 1H), 6.25 (d, J ) 1.8
Hz), 4.09-3.95 (m, 3H), 3.82 (dd, J ) 2.7, 12.5 Hz, 1H). 19F
NMR (methanol-d4): δ -170.4. HRMS FAB: calcd for C12H15
FN4O4 + H+ 299.1156, found 299.1150.
-
2-Am in o-4-ch lor o-7-(2-C-m eth yl-â-D-r ibofu r a n osyl)-7H-
p yr r olo[2,3-d ]p yr im id in e (19). To a solution of 2-amino-4-
chloro-7-[3,5-bis-O-(2,4-dichlorophenylmethyl)-2-C-methyl-â-
D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine (18) (630 mg, 1.0
mmol) in dichloromethane (20 mL) at -78 °C was added boron
trichloride (1 M in dichloromethane) (10 mL, 10 mmol). The
mixture was stirred at -78 °C for 2 h and then at -20 °C for
2.5 h. The reaction was quenched with dichloromethane/
methanol (1:1) (10 mL), stirred at -20 °C for 0.5 h, and
neutralized at 0 °C with aqueous ammonia. The solid was
filtered and then washed with dichloromethane/methanol
(1:1), and the combined filtrate was evaporated in vacuo. The
residue was purified on silica gel using dichloromethane/
methanol (50:1 and 20:1) as the eluent to give the title
compound (250 mg, 84%) as a colorless foam. 1H NMR
(methanol-d4) δ: 0.86 (s, 3H), 3.81 (m, 1H), 3.93-4.10 (3m,
3H), 6.16 (s, 1H), 6.38 (d, J ) 3.7 Hz, 1H), 7.43 (d, J ) 3.7,
1H). HRMS: calcd for C12H15ClN4O4 + H+ 315.0860, found
315.0856.
4-Am in o-5-ch lor o-7-(2-C-m eth yl-â-D-r ibofu r a n osyl)-7H-
p yr r olo[2,3-d ]p yr im id in e (26). To a precooled solution (0
°C) of 9 (50 mg, 0.18 mmol) in dimethylformamide (0.5 mL)
was added N-chlorosuccinimide (0.024 g, 0.18 mmol) in di-
methylformamide (0.5 mL) dropwise. The solution was allowed
to warm to room temperature and was stirred for 2 h. The
reaction was quenched with methanol (4 mL) and was evapo-
rated in vacuo. The crude product was purified on silica gel
using methanol/dichloromethane (1:9) as the eluent to give the
desired product (27 mg, 48%) as a colorless solid. 1H NMR
(acetonitrile-d3): δ 0.80 (s, 3H), 3.65-4.14 (overlapping m, 7H),
5.97 (br s, 2H), 6.17 (s, 1H), 7.51 (s, 1H), 8.16 (s, 1H). HRMS:
calcd for C12H15ClN4O4 + H+ 315.0860, found 315.0860.
4-Am in o-5-br om o-7-(2-C-m eth yl-â-D-r ibofu r a n osyl)-7H-
p yr r olo[2,3-d ]p yr im id in e (27). To a precooled solution (0
°C) of 9 (50 mg, 0.18 mmol) in dimethylformamide (0.5 mL)
was added N-bromosuccinimide (0.032 g, 0.18 mmol) in di-
methylformamide (0.5 mL) dropwise. The solution was stirred
at 0 °C for 20 min and then at room temperature for 10 min.
The reaction was quenched by addition of methanol (4 mL)
and was evaporated in vacuo. The crude product was purified
on silica gel using methanol/dichloromethane (1:9) as the
eluent. Fractions containing the product were pooled and
evaporated in vacuo to give the desired product (28 mg, 43%)
2-Am in o-7-(2-C-m eth yl-â-D-r ibofu r a n osyl)-7H-p yr r olo-
[2,3-d ]p yr im id in -4(3H)-on e (20). A mixture of 2-amino-4-
chloro-7-(2-C-methyl-â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]py-
rimidine (19) (90 mg, 0.3 mmol) in aqueous NaOH (2 N, 9 mL)
was heated at reflux temperature for 5 h and then neutralized
at 0 °C with 2 N aqueous HCl and evaporated to dryness.
Purification on silica gel using dichloromethane/methanol
(5:1) as the eluent yielded the title compound (70 mg, 82%) as
1
as a colorless solid. H NMR (acetonitrile-d3): δ 0.69 (s, 3H),
3.46-4.00 (overlapping m, 7H), 5.83 (br s, 2H), 6.06 (s, 1H),
1
a white solid. H NMR (methanol-d4): δ 0.86 (s, 3H), 3.79 (m
7.45 (s, 1H), 8.05 (s, 1H). HRMS: calcd for C12H15BrN4O4
+
1H), 3.90-4.05 (m, 3H), 6.06 (s, 1H), 6.42 (d, J ) 3.7 Hz, 1H),
7.05 (d, J ) 3.7 Hz, 1H). HRMS: calcd for C12H16N4O5 + H+
297.1199, found 297.1200.
H+ 359.0355, found: 359.0350.
4-Ch lor o-7-[3,5-bis-O-(2,4-d ich lor op h en ylm eth yl)-2-C-
m eth yl-â-D-r ibofu r a n osyl]-5-m eth yl-7H-p yr r olo[2,3-d ]p y-
r im idin e (29). To an ice-cold solution of 3,5-bis-O-(2,4-dichloro-
phenylmethyl)-2-C-methyl-1-O-methyl-R-D-ribofuranose (5) (1.06
g, 2.1 mmol) in dichloromethane (30 mL) was added HBr (5.7
M in acetic acid; 2.2 mL) dropwise. The reaction mixture was
stirred at 0 °C for 1 h and then at room temperature for 2 h,
concentrated in vacuo, and coevaporated with toluene (2 × 15
mL). The resulting oil was dissolved in acetonitrile (10 mL)
and was added dropwise into a solution of the sodium salt of
4-chloro-5-methyl-1H-pyrrolo[2,3-d]pyrimidine in acetonitrile
(generated in situ from 4-chloro-5-methyl-1H-pyrrolo[2,3-d]-
pyrimidine) (0.62 g, 3.7 mmol) in anhydrous acetonitrile (70
mL), and NaH (60% in mineral oil, 148 mg, 3.7 mmol), after 2
h of vigorous stirring at room temperature). The combined
mixture was stirred at room temperature for 1 day and then
evaporated to dryness. The residue was suspended in water
(100 mL) and extracted with ethyl acetate (2 × 150 mL). The
combined extracts were washed with brine (50 mL), dried over
Na2SO4, filtered, and evaporated. The crude product was
purified on silica gel using hexane/ethyl acetate (9:1, 5:1, 3:1)
gradient as the eluent to give the desired product (0.87 g, 66%)
4-Ch lor o-5-flu or o-7H-p yr r olo[2,3-d ]p yr im id in e (23). To
a solution of 4-chloro-5-bromo-7H-pyrrolo[2,3-d]pyrimidine (21)
(0.92 g, 4.0 mmol) in tetrahydrofuran (25 mL) was added
n-BuLi (2.5 M solution in hexane, 3.48 mL) dropwise at -78
°C, and the reaction mixture was stirred at -78 °C for an
additional 30 min. To this solution was added trimethylstannyl
chloride (0.88 g, 4.4 mmol) in tetrahydrofuran (8 mL) dropwise
over a period of 10 min. The reaction mixture was allowed to
warm to room temperature and was stirred overnight. Satu-
rated aqueous ammonium chloride (60 mL) was added and
extracted with ethyl acetate (3 × 70 mL). The combined
organic extracts were washed with brine, dried over Na2SO4,
and evaporated to dryness. The residue was purified over silica
gel to give the desired stannane (0.80 g, 63%) as a colorless
solid. To a solution of this compound (1.97 g, 6.20 mmol) in
acetonitrile (60 mL) was added 1-(chloromethyl)-4-fluoro-1,4-
diazoniabicyclo[2.2.2]octane bis(tetrafluoroborate) (2.40 g, 6.5
mmol) in one portion, and the reaction mixture was stirred at
room temperature for 7 h. The precipitate was filtered off, and
the filtrate was evaporated in vacuo. The residue was purified
on silica gel using ethyl acetate/hexane (3:7) as the eluent to
1
as a colorless foam. H NMR (DMSO-d6): δ 0.94 (s, 3H,), 2.31
1
give the title compound (0.22 g, 21%) as a colorless solid. H
(s, 3H), 3.80 (dd, J ) 2.1 Hz, 11.2 Hz, 1H), 4.01 (dd, J ) 1.8
Hz, 11.2 Hz, 1H), 4.26 (m, 2H), 4.62, 4.70 (2d, J ) 12.5 Hz,
2H), 4.72, 4.81 (2d, J ) 12.1 Hz, 2H), 5.30 (s, 1H), 6.39 (s,
NMR (methanol-d4): δ 8.53 (s, 1H), 7.37 (d, J ) 2.8 Hz, 1H).
19F NMR (DMSO-d6): δ -171.5. HRMS FAB: calcd for C6H3-
ClFN3 + H+ 172.0078, found 172.0079.
1H), 7.26-7.45 (m, 7H), 8.56 (s, 1H). HRMS: calcd for C27H24
-
Cl5N3O4 + H+ 630.0288, found 630.0276.
4-Am in o-5-flu or o-7-(2-C-m eth yl-â-D-r ibofu r a n osyl)-7H-
p yr r olo[2,3-d ]p yr im id in e (25). To a solution of 23 (0.075 g,
0.44 mmol), 2,3,5-tri-O-benzoyl-2-C-methyl-D-ribofuranose (24)18
(0.25 g, 0.53 mmol) and triphenylphosphine (0.23 g, 0.88 mmol)
in tetrahydrofuran (15 mL) was added diethyl azodicarboxylate
(DEAD) (0.14 mL, 0.88 mmol). The reaction mixture was
stirred at room temperature overnight. The solution was
directly adsorbed onto silica gel and was purified over silica
gel using ethyl acetate/hexane (1:9) as the eluent. Appropriate
fractions were combined and evaporated. The residue was
dissolved in dioxane (3 mL) and liquid ammonia (4 mL), and
the mixture was heated in a steel bomb at 85 °C for 24 h. The
solvent was evaporated, and the residue was purified over
4-Ch lor o-5-m et h yl-7-(2-C-m et h yl-â-D-r ib ofu r a n osyl)-
7H-p yr r olo[2,3-d ]p yr im id in e (30). To a solution of 4-chloro-
7-[3,5-bis-O-(2,4-dichlorophenylmethyl)-2-C-methyl-â-D-ribo-
furanosyl]-5-methyl-7H-pyrrolo[2,3-d]pyrimidine (29) (0.87 g,
0.9 mmol) in dichloromethane (30 mL) at -78 °C was added
boron trichloride (1 M in dichloromethane, 9.0 mL, 9.0 mmol)
dropwise. The mixture was stirred at -78 °C for 2.5 h and
then at -30 to -20 °C for 3 h. The reaction was quenched by
addition of methanol/dichloromethane (1:1) (9 mL), and the
resulting mixture was stirred at -15 °C for 30 min, and then
neutralized with aqueous ammonia at 0 °C and stirred at room
temperature for 15 min. The solid was filtered and washed

5296 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 21
Eldrup et al.
with dichloromethane/methanol (1:1, 50 mL). The combined
filtrate was evaporated in vacuo, and the residue was purified
on silica gel using dichloromethane and dichloromethane/
methanol (40:1 through 30:1) as the eluent to furnish the
desired compound (0.22 g, 78%) as colorless foam. 1H NMR
(DMSO-d6): δ 0.62 (s, 3H), 2.36 (s, 3H), 3.58-3.94 (m, 4H),
5.04-5.14 (m, 3H), 6.18 (s, 1H, H-1′), 7.77 (s, 1H), 8.54 (s, 1H).
HRMS: calcd for C13H16ClN3O4 + H+ 314.0908, found 314.0905.
4-Am in o-5-m et h yl-7-(2-C-m et h yl-â-D-r ib ofu r a n osyl)-
7H-p yr r olo[2,3-d ]p yr im id in e (31). To 4-chloro-5-methyl-7-
(2-C-methyl-â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine (30)
(0.2 g, 0.64 mmol) was added methanolic ammonia (saturated
at 0 °C) (40 mL). The mixture was heated in a stainless steel
autoclave at 100 °C for 14 h and then cooled and evaporated
in vacuo. The crude mixture was purified on silica gel with
dichloromethane/methanol (30:1, 20:1) gradient as the eluent
to give the title compound (0.13 g, 69%) as a colorless solid.
1H NMR (DMSO-d6): δ 0.60 (s, 3H), 2.26 (s, 3H), 3.52-3.61
(m, 1H), 3.70-3.88 (m, 3H), 5.00 (s, 1H), 4.91-4.99 (m, 3H),
6.04 (s, 1H), 6.48 (br s, 2H), 7.12 (s, 1H), 7.94 (s, 1H). HRMS:
calcd for C13H18N4O4 + H+ 295.1406, found 295.1407.
gel using methanol/dichloromethane (1:9) as the eluent. Frac-
tions containing the product were pooled and evaporated in
vacuo. This product was dissolved in 2 N aqueous NaOH (2
mL) and heated at reflux for 2 h. The suspension was cooled
and neutralized with 2 N aqueous HCl. The suspension was
absorbed on silica gel and was purified on a silica gel column
using methanol/dichloromethane (1:9 through 1:5) as the
eluent. Fractions containing the product were pooled and
evaporated in vacuo to give the desired compound (14 mg, 42%)
as a colorless solid. 1H NMR (methanol-d4): δ 0.85 (s, 3H),
3.77 (m, 1H), 3.90-4.07 (overlapping m, 2H), 6.05 (s, 1H), 7.13
(s, 1H). HRMS calcd for C12H15ClN4O5 + H+ 331.0809, found
331.0815.
Ack n ow led gm en t. We thank Dr. Charles R. Aller-
son of Isis Pharmaceuticals for valuable comments
about the manuscript.
Su p p or tin g In for m a tion Ava ila ble: Tables of experi-
mental details for compound 9, atomic parameters, and
interatomic distances and angles. This material is available
free of charge via the Internet at http://pubs.acs.org.
4-Am in o-7-(2-C-m eth yl-â-D-r ibofu r a n osyl)-7H-p yr r olo-
[2,3-d ]p yr im id in e-5-ca r boxylic a cid (34). 4-Amino-5-meth-
yl-7-(2-C-methyl-â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimi-
dine (34) (0.035 g, 0.11 mmol) was dissolved in a mixture of
aqueous ammonia (4 mL, 30 wt %) and saturated methanolic
ammonia (2 mL), and a solution of H2O2 in water (2 mL, 35
wt %) was added. The reaction mixture was stirred at room
temperature for 18 h. The solvent was removed under reduced
pressure, and the residue was purified by HPLC on a reverse-
phase column (Altech Altima C-18, 10 mm × 299 mm, A )
water, B ) acetonitrile, 10 to 60% B in 50 min, flow 2 mL/
min) to yield the title compound (0.015 g, 41%) as a white solid.
1H NMR (methanol-d4): δ 0.85 (s, 3H), 3.61 (m, 1H), 3.82 (m,
1H) 3.99-4.86 (m, 2H), 6.26 (s, 1H), 8.10 (s, 2H) 8.22 (s, 1H).
HRMS (FAB): calcd for C13H16N4O6 + H+ 325.1148, found
325.1143.
Refer en ces
(1) Seeff, L. B.; Hoofnagle, J . H. National Institutes of Health
Consensus Development Conference Statement: Management
of Hepatitis C: 2002. J une 10-12, 2002. Hepathology 2002, 36,
5, S3-S20.
(2) Hwang, S. B.; Park, K.-J .; Kim, Y.-S.; Sung, Y. C.; Lai, M. M. C.
Hepatitis
C virus NS5B protein is a membrane-associated
phosphoprotein with a predominantly perinuclear localization.
Virology 1997, 227, 439-446.
(3) Behrens, S.-E.; Tomei, L.; De Francesco, R. Identification and
properties of the RNA-dependent RNA polymerase of hepatitis
C virus. EMBO J . 1996, 15, 12-22.
(4) Takamizawa, A.; Mori, C.; Fuke, I.; Manabe, S.; Murakami, S.
Fujita, J .; Onishi, E.; Andoh, T.; Yoshida, I.; Okayama, H.
Structure and organization of the hepatitis C virus genome
isolated from human carriers. J . Virol. 1991, 65, 1105-1113.
(5) Carroll, S. S.; Tomassini, J . E.; Bosserman, M.; Getty, K.;
Stahlhut, M. W.; Eldrup, A. B.; Bhat, B.; Hall, D.; Simcoe, A.
S.; LaFemina, R.; Rutkowski, C. A.; Shim, S.; Wolanski, B.; Yang,
Z.; Migliaccio, G.; De Francesco, R.; Kuo, L. C.; Maccoss, M.;
Olsen, D. B. Inhibition of hepatitis C virus RNA replication by
2′-modified nucleoside analogs. J . Biol. Chem. 2003, 278, 11979-
11984.
2,4-Dia m in o-7-(2-C-m eth yl-â-D-r ibofu r a n osyl)-7H-p yr -
r olo[2,3-d ]p yr im id in e (35). 2-Amino-4-chloro-7-(2-C-methyl-
â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]pyrimidine (19) (21 mg,
0.067 mmol) and concentrated aqueous ammonia (10 mL) were
heated in a stainless steel autoclave overnight at 95 °C and
then cooled and evaporated in vacuo. The crude mixture was
purified on a silica gel using dichloromethane/methanol (10:1
through 5:1) as the eluent to give the title compound (11 mg,
(6) Shim, J .; Larson, G.; Lai, V.; Naim, S.; Wu, J . Z. Canonical 3′-
deoxyribonucleotides as chain terminator for HCV NS5B RNA-
dependent RNA polymerase. Antiviral Res. 1993, 243-251.
(7) Migliaccio, G.; Tomassini, J . E.; Carroll, S. S.; Tomei, L.;
Altamura, S.; Bhat, B.; Bartholomew, L.; Bosserman, M. R.;
Ceccacci, A.; Colwell, L. F.; Cortese, R.; De Francesco, R.; Eldrup,
A. B.; Getty, K. L.; Hou, X. S.; LaFemina, R. L.; Ludmerer, S.
W.; MacCoss, M.; McMasters, D. R.; Stahlhut, M. W.; Olsen, D.
B.; Hazuda, D. J .; Flores, O. A. Characterization of Resistance
to Non-Obligate Chain-Terminating Ribonucleoside Analogs that
Inhibit Hepatitis C Virus Replication in Vitro. J . Biol. Chem.
2003, 278, 49164-49170.
(8) Eldrup, A. B.; Allerson, C. R.; Bennett, C. F.; Bera, S.; Bhat, B.;
Bhat, N.; Bosserman, M. R.; Brooks, J .; Burlein, C.; Carroll, S.
S.; Cook, P. D.; Getty, K. L.; MacCoss, M.; McMasters, D. R.;
Olsen, D. B.; Prakash, T. P.; Prhavc, M.; Song, Q.; Tomassini,
J .; Xia, J . Structure activity relationship of purine ribonucleo-
sides for inhibition of hepatitis C virus RNA-dependent RNA
polymerase. J . Med. Chem. 2004, 47, 2283-2295.
(9) Stuver, L. J .; Mcbrayer, T. R.; Tharnish, P. M.; Hassan, A. E.
A.; Chu, C. K.; Pankiewich, K. W.; Watanabe, K. A.; Schinazi,
R. F.; Otto, M. J . Dynamics of subgenomic hepatitis C virus
replicon levels in Huh-7 cells after exposure to nucleoside
antimetabolites. J . Virol. 2003, 77, 10689-10694.
(10) Chan, L.; Reddy, J .; Proulx, M.; Das, S.; Pereira, O.; Wang, W.;
Siddiqui, A.; Yannopoloulos, C. G.; Poisson, C.; Turcotte, N.;
Drouin, A.; Hicham, Alaoui-Ismaili, M. H.; L’Heureux, L.;
Bilimoria, D.; Nguyen-Ba, N. Identification of N,N-disubstituted
phenylalanines as a novel class of inhibitors of hepatitis C NS5B
polymerase. J . Med. Chem. 2003, 46, 1283-1285.
1
56%) as a colorless solid. H NMR (DMSO-d6): δ 0.68 (s, 3H),
3.48-3.58 (m, 1H), 3.68-3.73 (m, 2H), 3.84 (m, 1H), 4.72 (s,
1H), 4.97-5.03 (m, 2H), 5.45 (s br, 2H), 6.00 (s, 1H), 6.28 (d,
J ) 3.7 Hz, 1H), 6.44 (br s, 2H), 6.92 (d, J ) 3.7 Hz, 1H).
HRMS: calcd for C12H17N5O4 + H+ 296.1359, found 296.1358.
4-Am in o-2-flu or o-7-(2-C-m eth yl-â-D-r ibofu r a n osyl)-7H-
p yr r olo[2,3-d ]p yr im id in e (36). To a solution of HF/pyridine
(70%, 2 mL) and pyridine (0.5 mL) at -25 °C was added 2,4-
diamino-7-(2-C-methyl-â-D-ribofuranosyl)-7H-pyrrolo[2,3-d]py-
rimidine (35) (60 mg, 0.20 mmol) in pyridine (0.5 mL) followed
by tert-butyl nitrite (0.036 mL). Stirring was continued for 5
min at -25 °C, and the solution was then poured into ice water
(5 mL), neutralized with 2 N aqueous NaOH, and evaporated
in vacuo. The residue was purified on silica gel using dichlo-
romethane/methanol (20:1 to 10:1) as the eluent to give the
desired compound (18 mg, 30%) as a colorless powder. 1H NMR
(acetonitrile-d3): δ 0.80 (s, 3H), 3.46 (t, 1H), 3.52 (d, 1H), 3.58
(s, 1H), 3.72-3.82 (m, 1H), 3.85-4.01 (m, 2H), 4.13 (dd, 1H),
6.04 (s, 1H), 6.09 (br s, 2H), 6.55 (d, 1H), 7.37 (d, H). HRMS:
calcd for C12H15FN4O4 + H+ 299.1156, found 299.1157.
2-Am in o-5-ch lor o-7-(2-C-m eth yl-â-D-r ibofu r a n osyl)-7H-
p yr r olo[2,3-d ]p yr im id in -4(3H)-on e (37). To a precooled
solution (0 °C) of 2-amino-4-chloro-7-(2-C-methyl-â-D-ribofura-
nosyl)-7H-pyrrolo[2,3-d]pyrimidine (19) (31 mg, 0.10 mmol) in
dimethylformamide (0.5 mL) was added N-chlorosuccinimide
(14.3 g, 0.11 mmol) in dimethylformamide (0.5 mL) dropwise.
The solution was stirred at room temperature for 2 h, and the
reaction was quenched by addition of methanol (5 mL) and
evaporated in vacuo. The crude product was purified on silica
(11) Wang, W.; Ng, K.; Cherney, M. M.; Chan, L.; Yannopoulos, J .
B.; Morin, N.; Nguyen-Ba, N.; Alaoui-Ismaili, M. H.; Bethell, R.
C.; J ames, M. N. G. Non-nucleoside analogue inhibitors bind to
an allosteric site on HCV NS5B polymerase. J . Biol. Chem. 2003,
278, 9489-9495.
(12) Dhanak, D.; Duffy, K. J .; J ohnston, V. K.; Lin-Goerke, J .; Darcy,
M.; Shaw, A. N.; Gu, B.; Silverman, C.; Gates, A. T.; Nonnema-
cher, M. R.; Earnshaw, D. L.; Casper, D. J .; Kaura, A.; Baker,
A.; Greenwood, C.; Gutshall, L. L.; Maley, D.; DelVecchio, A.;

Heterobase-Modified 2′-C-Methyl Ribonucleosides
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 21 5297
Macarron, R.; Hofmann, G. A.; Alnoah, Z.; Cheng, H. Y.; Chan,
G.; Khandekar, S.; Keenan, R. M.; Sarisky, R. T. Identification
and biological characterization of heterocyclic inhibitors of the
hepatitis C virus RNA-dependent RNA polymerase. J . Biol.
Chem. 2002, 277, 38322-38327.
(20) Murai, Y.; Shiroto, H.; Ishizaki, T.; Iimori, T.; Kodama, Y.;
Ohtsuka, Y.; Oishi, T. A Synthesis and an X-ray analysis of2′-
C-, 3′-C- and 5′-C-methylsangivamycins. Heterocycles 1992, 33,
391-404.
(21) Bressanelli, S.; Tomei, L.; Rey, F. A.; De Francesco, R. Structural
analysis of the hepatitis C virus RNA polymerase in complex
with ribonucleotides. J . Virol. 2002, 76, 3482-3492.
(22) Butcher, S. J .; Grimes, J . M.; Makeyev, E. V.; Bamford, D. H.;
Stuart, D. I. A mechanism for initiating RNA-dependent RNA
polymerization. Nature 2001, 410, 235-240.
(23) Seela, F.; Muth, H. P.; Bindig, U. Synthesis of 6-substituted
7-carbapurine 2′,3′-dideoxynucleosides. solid-liquid phase-
transfer glycosylation of 4-chloropyrrolo[2,3-d]pyrimidine and
deoxygenation of its 2′-deoxyribofuranoside. Synthesis 1988,
670-674.
(24) Seela, F.; Steker, H.; Driller, H. J .; Bindig, U. 2-Amino-2′-
deoxytubercidin and related pyrrolo[2,3-d]pyrimidinyl-2′-deox-
yribofuranosides. Liebigs Ann. Chem. 1987, 15-19.
(25) Crippen, C. M.; Havel, T. F. Distance Geometry and Molecular
Conformation; Bawden, D., Ed.; Wiley: New York, 1998. (Kears-
ley, S. K. Unpublished in-house distance geometry program.)
(26) Mohamadi, F.; Richards, N. G. J .; Guida, W. C.; Liskamp, R.;
Lipton, M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still, W.
C. MacroModel-an integrated software system for modeling
organic and bioorganic molecules using molecular mechanics.
J . Comput. Chem. 1990, 11, 440-467.
(13) Gu, B.; J ohnston, V. K.; Gutshall, L. L.; Nguyen, T. T.; Gontarek,
R. R.; Darcy, M. G.; Tedesco, R.; Dhanak, D.; Duffy, K. J .; Kao,
C. C.; Sarisky, R. T. Arresting initiation of hepatitis C virus RNA
synthesis using heterocyclic derivatives. J . Biol. Chem. 2003,
278, 16602-16607.
(14) Davoll, J . Pyrrolo[2,3-d]pyrimidines. J . Chem. Soc., Abstr. 1960,
131-138.
(15) Schneller, S. W.; Hosmane, R. S. Chlorination of 1H-pyrrolo-
[3,2-c]pyridine-4,6(5H,7H)-dione (3,7-dideazaxanthine) and its
5-methyl derivative. J . Heterocycl. Chem. 1978, 15, 325-326.
(16) Fuhrer, H.; Sutter, P.; Weis, C. D. A new ring closure reaction
of 2-phenoxyphenols and 3-(phenoxy)pyridines. Synthesis of
halogenated 10-methylphenoxazines and 10-methyl[1,4]benzoxa-
zino[3,2-b]pyridines. J . Heterocycl. Chem. 1979, 16, 1121-1134.
(17) Pudlo, J . S.; Saxena, N. K.; Nassiri, M. R.; Turk, S. R.; Drach,
J . C. et al. Synthesis and antiviral activity of certain 4- and 4,5-
disubstituted 7-[(2-hydroxyethoxy)methyl]pyrrolo[2,3-d]pyrim-
idines. J . Med. Chem. 1988, 31, 2086-2092.
(18) Harry-O’kuru, R. E.; Smith, J . M.; Wolfe, M. S. A Short, Flexible
route toward 2′-C-branched ribonucleosides. J . Org. Chem. 1997,
62, 1754-1759.
(19) Pudlo, J . S.; Nassiri, M. R.; Kern, E. R.; Wotring, L. L.; Drach,
J . C.; Townsend, L. B. Synthesis, antiproliferative, and antiviral
activity of certain 4-substituted and 4,5-disubstituted 7-[(1,3-
dihydroxy-2-propoxy)methyl]pyrrolo[2,3-d]pyrimidines. J . Med.
Chem. 1990, 33, 1984-1992.
J M040068F