Design, synthesis, and antiviral activity of adenosine 5′-phosphonate analogues as chain terminators against hepatitis C virus

Article abstract of DOI:10.1021/jm049029u

A series of adenosine 5′-phosphonate analogues were designed to mimic naturally occurring adenosine monophosphate. These compounds (1-5) were synthesized and evaluated in a cellular hepatitis C virus (HCV) replication assay. To improve cellular permeabili

Full text of DOI:10.1021/jm049029u

J. Med. Chem. 2005, 48, 2867-2875
2867
Design, Synthesis, and Antiviral Activity of Adenosine 5′-Phosphonate
Analogues as Chain Terminators against Hepatitis C Virus
Yung-hyo Koh,† Jae Hoon Shim,† Jim Zhen Wu, Weidong Zhong, Zhi Hong, and Jean-Luc Girardet*
Drug Discovery, R&D, Valeant Pharmaceuticals International, 3300 Hyland Avenue, Costa Mesa, California 92626
Received November 30, 2004
A series of adenosine 5′-phosphonate analogues were designed to mimic naturally occurring
adenosine monophosphate. These compounds (1-5) were synthesized and evaluated in a cellular
hepatitis C virus (HCV) replication assay. To improve cellular permeability and enhance the
anti-HCV activity of these phosphonates, a bis(S-acyl-2-thioethyl) prodrug for compound 5 was
prepared, and its cellular activity was determined. To elucidate the mechanism of action of
these novel adenosine phosphonates, their diphosphate derivatives (1a-5a) were synthesized.
Further nucleotide incorporation assays by HCV NS5B RNA-dependent RNA polymerase
revealed that 2a and 3a can serve as chain terminators, whereas compounds 1a, 4a, and 5a
are competitive inhibitors with ATP. Additional steady-state kinetic analysis determined the
incorporation efficiency of 2a and 3a as well as the inhibition constants for 1a, 4a, and 5a.
The structure-activity relationships among these compounds were analyzed, and the implica-
tion for nucleoside phosphonate drug design was discussed.
Introduction
fovir, adefovir, and cidofovir are acyclic nucleoside
phosphonates.5 Generally, nucleosides are prodrugs, and
their biological activity is exerted by their 5′-triphos-
phate derivative. Intracellular phosphorylation is re-
quired to convert a nucleoside drug to its active 5′-
triphosphate in order to be incorporated into a viral
DNA. The sequential phosphorylation is normally ex-
ecuted by three viral or cellular kinases, including
nucleoside kinase, nucleoside monophosphate kinase,
and nucleoside diphosphate kinase.6 Once incorporated
into an elongating viral DNA chain, the nucleotide
inhibitor functions as a chain terminator to abort viral
DNA synthesis.
HCV is the major causative agent for non-A, non-B
virally induced hepatitis.1 As an insidious and deadly
disease, HCV infection is responsible for an emerging
pandemic of chronic liver diseases. There are 170 million
infected individuals worldwide and approximately 4
million HCV carriers in the United States alone. Un-
resolved acute HCV infections progress to a chronic
disease that persists for decades. As much as 20% of
the infected individuals eventually develop liver cir-
rhosis with 1-5% subsequently leading to hepatocellu-
lar carcinoma.2 This type of cancer accounts for nearly
10 000 deaths annually in the US. The current standard
of care treatment is a combination of subcutaneous
pegylated interferon-R with oral nucleoside drug rib-
avirin.3 The sustained viral response, defined as unde-
tectable viral load for 6 months following cessation of
therapy, is around 54-56% for this combination therapy.
Therefore, there is still room for the treatment to be
improved in term of efficacy as well as safety. One
approach is to develop direct antiviral chemotherapy
against a viral target. HCV NS5B is the viral RNA-
dependent RNA polymerase (RdRp) that is responsible
for viral genome replication. It constitutes a valid target
for drug discovery.4
Nucleoside chain terminators comprise a major class
of drugs that targets a viral polymerase. The chemically
unique chain terminators that have been proved clini-
cally useful include acyclovir (ACV) and gancyclovir
(GCV) for herpes virus infection, zidovudine (AZT),
didanosine (ddI), zalcitabine (ddC), stavudine (d4T),
abacavir (ABC), and tenofovir (TDF) for HIV/AIDS,
lamivudine (3TC) and adefovir (ADV) for HBV, and
cidofovir for cytomegalovirus retinitis in AIDS patients.
Although most of them are nucleoside analogues, teno-
A nucleoside 5′-phosphonate is essentially a nucleo-
side monophosphate analogue. However, a phosphonate
has the advantage over its phosphate counterpart of
being metabolically stable, as its phosphorus-carbon
bond is not susceptible to phosphatase hydrolysis. More
importantly, the presence of a 5′-phosphonate allows the
first phosphorylation step required for nucleoside acti-
vation to be skipped, therefore bypassing this inefficient
and often rate-limiting step in the conversion to 5′-
triphosphate. Like a nucleoside monophosphate, a nu-
cleoside phosphonate can be further phosphorylated by
cellular nucleotide kinases. The concept of nucleoside
phosphonate has been applied to design chain termina-
tors for anti-HIV chemotherapy and proved to be valid.
9-(2-Phosphonylmethoxypropyl)adenine (PMPA) and
9-(2-phosphonylmethoxyethyl)adenine (PMEA) are two
effective and potent nucleoside phosphonate chain
terminators for HIV reverse transcriptase (RT).7-9
Mechanism of action studies show that incorporation
of PMPA diphosphate (PMPApp) by HIV-RT is almost
as efficient as that of dATP or ddATP.10 Despite this
success, the concept of nucleoside phosphonate has not
been fully explored for RNA viruses.5 In this study, we
designed and synthesized a series of 5′-deoxyadenosine
* To whom correspondence should be addressed. Telephone: 714-
545-0100, ext. 4204. Fax: 714-641-7222. E-mail: jlgirardet@valeant.com.
† These investigators contributed equally to this work.
10.1021/jm049029u CCC: $30.25 © 2005 American Chemical Society
Published on Web 03/15/2005

2868 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8
Koh et al.
Scheme 1a
a Reagents and conditions: (a) Bu3N, 1,1′-carbonyldiimida-
zole, DMF; (HNBu3)2H2P2O7; TEAB buffer.
Results
Synthesis of Adenosine Phosphonate Analogues
and Their Diphosphate Derivatives. Adenosine
phosphonate 1 was synthesized following the literature
procedure of Raju et al.11 To prepare its diphosphate
derivative 1a, 1 was reacted with 1,1′-carbonyldiimida-
zole (CDI) in DMF followed by bis(tri-n-butylammo-
nium) pyrophosphate. The diphosphate obtained from
the reaction contained a 2′,3′-cyclic carbonate moiety
because of the coupling between 2′- and 3′-hydroxyl
groups of 1 with CDI. The cyclic carbonate was hydro-
lyzed by treatment with triethylammonium bicarbonate
(TEAB) buffer to give the phosphono diphosphate 1a
(Scheme 1).
For the synthesis of nucleoside phosphonate 2, the
previously reported synthetic routes18,19 were adopted
with some modifications (Scheme 2). Treatment of
N6,2′,3′-tribenzoyladenosine (6) with dicyclohexylcarbo-
diimide11 and DMSO in the presence of pyridinium
trifluoroacetate produced a 5′-aldehyde, which was later
reacted with diphenyl triphenylphosphoranylidenem-
ethylphosphonate20 to yield the corresponding 5′,6′-vinyl
phosphonate. Catalytic hydrogenation of the vinyl phos-
phonate in the presence of palladium on carbon gave a
saturated phosphonate ester 7. Transesterification of 7
with sodium benzoxide resulted in 8 in a good yield.
Deprotection of benzoyl groups of 8 followed by pal-
ladium-catalyzed hydrogenolysis of benzyl ester pro-
vided the nucleoside phosphonate 2. Its diphosphate
derivative 2a was prepared using the same reaction
conditions as described in Scheme 1.
Figure 1. Examples of nucleoside phosphonates designed for
DNA virus and retrovirus.
phosphonates and evaluated their antiviral activity
against HCV, an RNA virus.
In the literature, several 5′-phosphate isosteres have
been adopted to prepare nucleoside phosphonates. As
shown in Figure 1, compounds I11 and III12,13 are simple
5′-deoxy nucleoside phosphonates, in which the 5′-
oxygen of a nucleoside phosphate is either removed or
replaced by a methylene group. Alternatively, in com-
pounds II14 and IV,15 the 5′-oxygen and -carbon are
simply swapped to yield a phosphonate. All of the
resultant phosphonates mimic the overall shape and
geometry of a nucleoside monophosphate. We applied
the same strategy to an adenosine ribonucleotide scaf-
fold and designed a series of 5′-deoxyadenosine phos-
phonate analogues (Figure 2, compounds 1-4). It has
been reported that addition of a 2′-methyl group onto
ATP yields a potent chain terminator for HCV NS5B
RdRp.16 Hence, an adenosine phosphonate with the 2′-
methyl addition was also conceived (Figure 2, compound
5). Since the ionic character of a phosphonic acid would
present an obstacle for cellular permeability, an S-acyl-
2-thioethyl (SATE) prodrug (5b) for compound 5 was
also prepared. Esterification of a phosphonic acid with
two SATE groups has been proven to be a feasible
strategy to deliver a phosphate or phosphonate drug into
cells.17 We report here the synthesis of adenosine
phosphonate analogues 1-5, their corresponding diphos-
phate derivatives 1a-5a, and a bis-SATE prodrug 5b.
The cellular activity of 1-5 and 5b were evaluated in a
HCV replication assay. Moreover, kinetic analysis was
performed for phosphono diphosphate derivatives 1a-
5a as an ATP analogue for NS5B RdRp. The implication
for future nucleoside phosphonate chain terminator
design is discussed.
Nucleoside phosphonate 3 was synthesized by follow-
ing a literature procedure14 in nine steps starting from
adenosine. The diphosphate derivative 3a was prepared
by activation of 3 with CDI followed by reaction with
pyrophosphate as described in Scheme 1 (Scheme 3).
Merlo et al. had previously reported the synthesis of
the carbocyclic nucleoside phosphonate 4 by employing
enzyme-catalyzed resolution method.21 Our chemical
Figure 2. Adenosine phosphonate analogues designed for HCV NS5B RdRp.

Chain Terminators against Hepatitis C Virus
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8 2869
Scheme 2a
a Reagents and conditions: (a) DCC, DMSO, TFA, pyridine; (PhO)2P(dO)CHPPh3; (b) H2, Pd-C, CH3OH; (c) NaH, BnOH, DMSO; (d)
NH3, CH3OH; (e) H2, Pd-C, CH3OH; (f) Bu3N, 1,1′-carbonyldiimidazole, DMF; (HNBu3)2H2P2O7; TEAB buffer.
Scheme 3a
pound 14 was obtained via a silylation, benzoylation,
and desilylation sequence. The same chemistry de-
scribed in Scheme 2 was applied to 14 to make phos-
phonate 5 and its diphosphate derivative 5a.
The bis-SATE prodrug of 5, 5b, was synthesized
following a published procedure as illustrated in Scheme
6.17 Briefly, phosphonate 5 was reacted with thioester
17 in the presence of 1-mestylene-2-sulfonyl-3-nitro-
1,2,4-triazole (MSNT). However, the reaction yielded a
3′,5′-cyclic phosphodiester 18 instead of 5b. To avoid the
unwanted cyclization, the 3′-hydroxyl group of 16 was
protected with a tert-butyldimethylsilyl group. Hydroly-
sis of the benzoyl amide followed by hydrogenolysis of
the benzyl ester provided the protected phosphonate 19,
which was then treated with thioester 17 to afford a
protected tert-butyl SATE phosphodiester. Finally, desi-
lylation of the phosphodiester with tetrabutylammo-
nium fluoride in the presence of acetic acid resulted in
5b with a good yield.
Cellular Activity of Adenosine Phosphonates. A
previously reported HCV subgenomic replicon system25
was used to evaluate the antiviral activity of adenosine
phosphonate analogues 1-5 and bis-SATE prodrug 5b.
In addition, the cytotoxicity of these compounds was
assessed in Huh7 cells. The results are summarized in
Table 1. Due to its poor solubility in DMSO, no repro-
ducible results were obtained for compound 1. For the
remaining adenosine phosphonates, compounds 4 and
5 are relatively potent with an EC50 of around 30 µM.
By comparison, compounds 2 and 3 are weaker with an
EC50 of 200 and 65 µM, respectively. None of them
a Reagents and conditions: (a) Bu3N, 1,1′-carbonyldiimidazole,
DMF; (HNBu3)2H2P2O7; TEAB buffer.
synthesis of 4 began with (1R,4S)-9-(4-hydroxycyclo-
pent-2-en-1-yl)adenine 9, which was prepared by
Pd(0)-catalyzed coupling of 6-chloropurine with chiral
cyclopentene-1,3-diol 1-acetate (Scheme 4).22,23 Initial
attempts to alkylate the oxygen at C4 of 9 with diethyl
p-tolylsulfonyloxymethyl phosphonate 11 in the pres-
ence of a catalytic base resulted in a mixture of several
inseparable products. Therefore, we decided to first
convert 9 to its N,N-dimethylformamidine derivative 10.
Further treatment of the protected nucleoside 10 with
phosphonate ester 11 led to clean O-alkylation. Depro-
tection of the amidine group with methanolic ammonia,
followed by bis-hydroxylation with osmium tetraoxide,
gave phosphonate ester 12. Hydrolysis of 12 by treat-
ment with bromotrimethylsilane afforded the carbocylic
phosphonate 4 in good yield. A diphosphate derivative
4a was prepared from 4 using the similar reaction
conditions as described in Scheme 1.
Synthesis of 2′-C-methyl phosphonate 5 was ac-
complished by employing a similar synthetic approach
developed for the synthesis of phosphonate 2 (Scheme
5). Starting from 2′-C-methyl adenosine (13),24 com-
Scheme 4a
a Reagents and conditions: (a) (EtO)2CHN(CH3)2, DMF; (b) NaH, DMF, (EtO)2POCH2OTs (11); (c) NH3, MeOH; (d) OsO4, 4-methyl-
morphorine N-oxide, acetone-t-BuOH-H2O; (e) TMSBr, DMF; (f) Bu3N, 1,1′-carbonyldiimidazole, DMF; (HNBu3)2H2P2O7; TEAB buffer.

2870 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8
Koh et al.
Scheme 5a
a Reagents and conditions: (a) TBSCl, pyridine; (b) Bz2O, pyridine; (c) TBAF, AcOH, THF; (d) DCC, DMSO, TFA, pyridine;
(PhO)2P(dO)CHPPh3; (e) H2, Pd-C, CH3OH; (f) NaH, BnOH, DMSO; (g) NH3, CH3OH; (h) H2, Pd-C, CH3OH; (i) Bu3N, 1,1′-
carbonyldiimidazole, DMF; (HNBu3)2H2P2O7; TEAB buffer.
Scheme 6a
a Reagents and conditions: (a) TBSCl, imidazole, DMF; (b) NH3, MeOH; (c) H2, Pd-C, MeOH; (d) 17, 1-mestylene-2-sulfonyl-3-nitro-
1,2,4-triazole, pyridine; (e) TBAF, AcOH, THF.
Table 1. Antiviral Activity of Adenosine Phosphonate 1-5 and
a Bis-SATE Prodrug 5b in a HCV Subgenomic Replicon Assay
and Their Cytotoxicity in Huh7 Cellsa
Incorporation of Nucleoside Phosphono Diphos-
phates by NS5B RdRp. All diphosphate derivatives
1a-5a of adenosine phosphonates 1-5 resemble the
conformation of ATP. They can be potentially recognized
by NS5B RdRp, be incorporated into elongating viral
RNAs, and act as a chain terminator. Alternatively,
even if they cannot be utilized as a substrate, they could
serve as a competitive inhibitor against ATP. We
investigated the mode of action of these adenosine
phosphono diphosphates in a nucleotide incorporation
assay as previously described.26 In the experiments, an
oligo template 5′-AAAAAAAUGC-3′ and a dinucleotide
primer 33pGpC were used. An adenosine nucleotide
substrate would form a Watson-Crick base pair with
U at the third base position in the template. A chain
terminator would be incorporated and generate a tri-
nucleotide product. The product is visualized after gel
separation followed by Phosphor Imager analysis. As
shown in Figure 3, among the five nucleoside phosphono
diphosphates tested, only 2a and 3a were incorporated
by NS5B RdRp and gave a detectable trinucleotide
product. Judging from the amount of products formed
on the gel, these two compounds are less active than
compd EC50 (µM) CC50 (µM) compd EC50 (µM) CC50 (µM)
1b
2
3
ND
200
65
ND
>300
>300
4
30
35
25
300
>300
30
5
5b
a The reported data are an average of at least two sets of results.
b The values of EC50 and CC50 were not determined (ND) due to
the poor solubility of the compound.
shows any significant cytotoxicity up to 300 µM. Addi-
tion of the 2′-methyl group reduces the EC50 of 2 by
6-fold. This result is consistent with the report in which
addition of a 2′-methyl group to ATP makes the result-
ing triphosphate a potent HCV inhibitor.16 Interestingly,
the antiviral activity of compounds 3 and 4 does not
differ much despite their different sugar moieties. This
conclusion indicates that the ribose bridge oxygen is not
critical for cellular activity. Finally, the bis-SATE pro-
drug 5b was found to be slightly more potent than its
parental compound 5, but its cytotoxicity was somehow
enhanced with the CC50 decreased from more than 300
µM for 5 to about 30 µM (Table 1).

Chain Terminators against Hepatitis C Virus
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8 2871
the next elongating nucleotide UTP in the incorporation
experiments described above. We were unable to detect
any extended products in these experiments (data not
shown). This could attribute to the fact that these two
nucleoside phosphonates are functioning as a chain
terminator. Although they possess the 3′-hydroxyl group
for elongation, the conformation distortion relayed from
the phosphonate bond could prevent them from further
chain elongation. Nucleoside analogues that retain a 3′-
hydroxyl have been reported as a chain terminator, such
as 2′-methyl ATP16 for HCV NS5B RdRp, and arabino-
syladenine triphosphate27 and arabinofuranosylcytidine
triphosphate28 for DNA polymerase â. Nevertheless,
there is a slight possibility that a lower concentration
of trinucleotides formed in the reaction makes any chain
elongation difficult to detect under the described assay
conditions. Further in vitro experiments using the
chemical synthesized primers 3′-ended with 2 or 3 are
needed to confirm the conclusion.
Figure 3. Incorporation of each of adenosine phosphono
diphosphates 1a-5a into a dinucleotide pGpC primer by HCV
NS5B RdRp. An incorporation reaction contained 2.5 µM of
NS5B RdRp, 20 µM of [5-33P]GpC primer, 20 µM of 5′-
AAAAAAAUGC-3′ template, and 1 mM of the following nucle-
otide. Lane 1, none; lane 2, ATP; lane 3, 1a; lane 4, 2a; lane
5, 3a; lane 6, 4a; lane 7, 5a. The reaction products were
resolved on a 25% polyacrylamide-7 M urea-TBE gel and
were analyzed using a Phosphor Imager. The mobility of the
radiolabeled dinucleotide primer and trinucleotide product was
marked.
Inhibition experiments for compounds 1a, 4a, and 5a
were performed to determine whether they can compete
with ATP to inhibit NS5B RdRp activity. All of them
were found to behave as a competitive inhibitor against
ATP (data not shown). Further kinetic analysis deter-
mined their apparent Ki values from Dixon plots at the
fixed concentration of ATP (1 mM) (Table 2). The Ki
values range from 200 to 300 µM, which are about 1.5-
2.5-fold higher than Km for ATP. These results suggest
that all three compounds are capable of interacting with
NS5B RdRp directly. However, their apparent binding
affinity is weaker than that of endogenous ATP.
Table 2. Steady-State Kinetic Parameters for ATP and
Adenosine Phosphono Diphosphates 1a-5a in an NS5B RdRp
Catalyzed Nucleotide Incorporation Assaya
compd Ki (µM)
Km (µM)
133 ( 12
Vmax (µM/min) Vmax/Km (×103)
ATP
0.23 ( 0.07
1.7 ( 0.5
1a
2a
3a
4a
5a
201 ( 47
143 ( 24
0.028 ( 0.002
0.19 ( 0.03
241 ( 21 0.0029 ( 0.0001 0.012 ( 0.001
229 ( 28
327 ( 64
a The reported data are an average of at least two sets of results.
ATP for NS5B RdRp. To determine their incorporation
efficiency (Vmax/Km), steady-state kinetic analysis was
performed. The apparent Km and Vmax for 2a, 3a, and
ATP were obtained at the fixed concentration of the
template and primer (20 µM), and their incorporation
efficiency was calculated. Conversely, for compounds 1a,
4a, and 5a, which cannot be utilized as a substrate for
NS5B RdRp, we determined their inhibition constant
(Ki) against ATP through a Dixon plot. All the obtained
kinetic parameters are summarized in Table 2. For
compound 2a, its Km value is very similar to that of
ATP, but its Vmax is 10-fold lower, resulting in an
incorporation efficiency of 1.9 × 10-4 min-1. These
kinetic results suggest that the presence of the 5′-oxygen
does not contribute much to the binding to NS5B RdRp
(Km), but it is critical for catalysis (Vmax). Compound 3a
has a lower incorporation efficiency than that of 2a with
a Vmax of 0.0029 µM/min and Km around 241 µM.
Compared to ATP, its Vmax value is reduced by almost
100-fold. Its incorporation efficiency of 1.2 × 10-5 min-1
is 140-fold lower than that of ATP (Table 2). Apparently,
the exchange of 5′-oxygen and -carbon has a relatively
small effect on substrate binding but a severe conse-
quence on catalysis. Intriguingly, although compounds
1a, 4a, and 5a are structurally similar to ATP, they
cannot be utilized as a substrate for NS5B RdRp.
Therefore, the structural alternations of removal of the
5′-oxygen (1a) or ribose bridge oxygen (4a) and addition
of a 2′-methyl group (5a) are detrimental to the activity
of an adenosine phosphonate.
Discussion
The concept of nucleoside phosphonate chain termi-
nator has been successfully utilized in designing novel
and potent drugs against retroviruses. Notably, teno-
fovir disoproxil fumarate, a prodrug of PMPA; adefovir
dipivoxil, a prodrug of PMEA; and cidofovir have been
approved by the FDA for the treatment of HIV/AIDS,
HBV, and CMV infections, respectively. In this study,
we extended the idea to HCV, an RNA virus. In this
regard, a series of five adenosine phosphonate analogues
were conceived and synthesized (Figure 2). They exhib-
ited varied antiviral activity in an HCV replicon assay
with an EC50 ranging from 35 to 200 µM. To eliminate
the potential cell penetration issue associated with the
anionic phosphonate moiety, we further modified com-
pound 5 by masking its anionic charges with two SATE
groups and prepared a bis-SATE prodrug 5b. Compared
to 5, the anti-HCV activity of 5b is moderately im-
proved. But its cytotoxicity is drastically increased. The
result is not totally unexpected. It implicates that the
nucleoside phosphonate 5 may possess a broad activity
against host polymerases. The cytotoxicity is exposed
once it gets efficiently delivered into cells.
To understand the mechanism of action of these
adenosine phosphonates, their diphosphate derivatives
were synthesized so that they could be tested in an HCV
NS5B RdRp catalyzed nucleotide incorporation assay.
Among the five adenosine phosphono diphosphates, only
2a and 3a were incorporated by NS5B RdRp, implying
that they can potentially serve as chain terminators.
Their incorporation efficiency varies, providing an in-
teresting structure-activity relationship (SAR). The
We further evaluated 2a and 3a as a potential chain
terminator and investigated whether a trinucleotide
ended by 2 or 3 can be further elongated by supplying

2872 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8
Koh et al.
Vmax/Km for 2a and 3a is 10-100-fold lower than that
of ATP, respectively. Therefore, the substitution of 5′-
oxygen by a methylene or exchange of the 5′-carbon and
-oxygen position of ATP not only does not potentiate but
actually reduces a nucleotide’s capability as a chain
terminator. Similar results have been observed for the
phosphonate analogues of AZTTP in the HIV-RT cata-
lyzed reaction.29 Isosteric substitution of the 5′-oxygen
by methylene for AZTTP reduces the catalytic efficiency
of HIV-RT by 1800-fold. The same substitution for
ddUTP makes it a very poor substrate of HIV-RT.30
in this study are not adequate to improve a chain
terminator’s catalytic efficiency nor the binding affinity
to NS5B RdRp. Further medicinal chemistry efforts are
needed to come up with novel and potent nucleoside
phosphonate chain terminators for HCV.
Experimental Section
General Chemical Methods. 1H NMR spectra were
recorded on a Varian Mercury 300 spectrometer with TMS as
the internal standard. Mass spectral data were obtained on
an API 150EX electrospray mass spectrometer. Anhydrous
solvents were purchased from Acros or Fluka and were used
without further treatment unless noted. TLC plates and silica
gel for flash chromatography were supplied by ICN Biomedi-
cals or Fisher Scientific. Solvent ratios were based on volume
in case that solvent mixture was used.
The failure of three other phosphono diphosphates,
1a, 4a, and 5a, to be accepted as a substrate by NS5B
RdRp highlights the other key structural features
required by a nucleotide chain terminator. A structural
comparison of 1a to 2a, 3a to 4a, and 2a to 5a
respectively suggests that shortening spacing between
the 5′-carbon to phosphorus, replacement of the ribose
bridge oxygen by a carbon, or addition of a 2′-methyl
group is detrimental for a phosphonate to function as a
chain terminator. The incompetency of these nucleoside
phosphonates can be further comprehended in the light
of steady-state kinetics. As determined in this study,
the Km or Ki values are relatively unchanged for all five
phosphono diphosphates compared to ATP’s (Table 2).
But their Vmaxs are greatly diminished. As a result, the
presence of a phosphonate linkage has a negligible effect
on binding affinity but a severe consequence on catalytic
efficiency. Such a reduction in the catalytic rate by the
employment of a phosphonate analogue has been re-
ported for AZTTP on HIV-RT.29 Conversely, these
results are quite different from the published kinetic
analysis of PMPA.10 PMPA is an acyclic nucleoside
analogue that structurally differs from canonical nucle-
otides. Interestingly, its diphosphate derivative is com-
parable to ATP as a substrate for HIV-RT. It is likely
that the acyclic sugar moiety in PMPA is flexible enough
for the polymerase complex to bring the R-phosphorus
of PMPA to the 3′-OH of the primer in a productive
orientation. In contrast, the relatively rigid ribose ring
in our compounds or in the phosphonate analogue of
AZTTP and ddUTP might render a higher energy
barrier for formation of a productive complex between
polymerase and a nucleotide. Consistent with this
argument is the observation that shortening the length
between the R-phosphorus and ribose or employing a
carbocyclic ring was more detrimental to polymerase
function than isosteric exchanges. It will be a challenge
to design a PMEA- or PMPA-like acyclic nucleoside
phosphonate analogue for an RNA virus, since an RNA
polymerase requires the 2′-OH of the ribose ring for
proper substrate recognition and there is no simple
solution to retain the 2′-OH moiety in an acyclic nucleo-
side.
9-[5′-Deoxy-5′-[(hydroxypyrophosphoroxy)phosphinyl]-
â-D-ribo-furanosyl]adenine (1a). A solution of 9-[5′-deoxy-
5′-(dihydroxyphosphinyl)-â-D-ribofuranosyl]adenine (1a) (19
mg, 0.059 mmol) and tributylamine (100 µL, 0.41 mmol) in
water (0.5 mL) was evaporated under reduced pressure. The
residue was repeatedly coevaporated with anhydrous ethanol
and toluene. To the resulting powder were added DMF (2 mL)
and 1,1′-carbonyldiimidazole (64 mg, 0.39 mmol). After stirring
at room temperature for 3.5 h, the reaction was quenched by
addition of methanol (20 µL). Bis(tri-n-butylammonium)-
pyrophosphate (153 mg, 0.34 mmol) was added, and the
stirring was continued for 16 h. White precipitate was removed
by centrifugation. The supernatant was collected and concen-
trated under reduced pressure. The resulting residue was
dissolved in 1 M solution of triethylammonium bicarbonate
buffer (3 mL, pH ) 8.5) and stirred at room temperature for
2 h. The mixture was concentrated under reduced pressure.
It was purified by RP-HPLC (250 × 10 mm, YMC-Pack ODS-
AQ, Waters) with a linear gradient of acetonitrile from 0 to
20% (for 60 min) in 50 mM tetraethylammonium acetate buffer
(TEAA, pH 7.0, tR ) 22 min). The yield was 7.2 mg (25%) of
1
1a. H NMR (D2O): δ 8.25 (s, 1H), 8.08 (s, 1H), 5.89 (d, J )
6.0 Hz, 1H), 4.65 (m, 1H), 4.27 (m, 2H), 3.02 (q, J ) 7.5 Hz,
36H), 2.20 (m, 2H), 1.10 (t, J ) 7.5 Hz, 48H). 31P NMR (D2O):
δ 13.2, -9.79, -22.2. MS (ES): m/e 489.6 (M - H).
9-[5′,6′-Dideoxy-6′-(hydroxyphosphinyl)-â-D-ribo-hexo-
furanosyl]adenine (2). A solution of 8 (190 mg, 0.30 mmol)
in methanolic ammonia (40 mL, saturated at 0 °C) was stirred
at room temperature in a sealed bomb for 24 h. The bomb was
cooled to 0 °C before opening. The reaction mixture was stirred
at room temperature for 1 h and then concentrated to dryness.
Silica gel chromatography (CH2Cl2:MeOH ) 90:10) yielded 146
mg (92%) of 9-[5′,6′-dideoxy-6′-(dibenzyloxyphosphinyl)-â-D-
ribo-hexofuranosyl] adenine as a pale foam. To a solution of
the compound obtained above (146 mg, 0.28 mmol) in MeOH
(20 mL) was added Pd-C (10%, 150 mg). The reaction mixture
was stirred under atmospheric pressure of hydrogen. After 16
h, the mixture was filtered through Celite to remove the
catalyst. The filter cake was washed with H2O (20 mL) and
the combined filtrate was evaporated to dryness. The resulting
solid obtained was washed with CH2Cl2 (1 mL) to yield 53 mg
(55%) of 2 as a white solid. 1H NMR (DMSO-d6): δ 8.41 (s,
1H), 8.25 (s, 1H), 7.38 (br s, 2H), 5.94 (d, J ) 4.5 Hz, 1H), 4.69
(m, 1H), 4.16 (m, 1H), 3.98 (m, 1H), 1.94 (m, 2H), 1.59 (m,
2H). HRMS (M + H)+ for C11H16N5O6P: calcd m/z 346.0916,
obs 346.0905.
In summary, several adenosine phosphonate ana-
logues were successfully synthesized. These AMP mim-
ics demonstrated some promising but relatively weak
anti-HCV activity. Further mechanistic studies indicate
that compounds 2a and 3a can potentially function as
a chain terminator, thereby proving the principle that
the nucleoside phosphonate concept is applicable to an
RNA virus. But challenges remain as neither 2a nor 3a
are potent enough to compete with ATP. The structural
modifications on the ribose and around the 5′-moiety
9-[5′,6′-Dideoxy-6′-[(hydroxypyrophosphoroxy)phos-
phinyl]-â-D-ribo-hexofuranosyl]adenine (2a). Diphosphate
2a (8.0 mg, 28%, tR ) 22 min) was prepared from 2 (19 mg,
0.056 mmol) by a similar procedure as that described for the
1
synthesis of 1a. H NMR (D2O): δ 8.23 (s, 1H), 8.09 (s, 1H),
5.89 (d, J ) 6.3 Hz, 1H), 4.60 (m, 1H), 4.14 (dd, J ) 3.6, 5.4
Hz, 1H), 4.04 (m, 1H), 3.02 (q, J ) 7.5 Hz, 36H), 1.88 (m, 2H),
1.73 (m, 2H), 1.10 (t, J ) 7.5 Hz, 48H). 31P NMR (D2O): δ
19.6, -9.85, -22.3. MS (ES): m/e 503.6 (M - H).
(2R,3R,4S,5R)-9-[Tetrahydro-3,4-dihydroxy-5-[(hydrox-
ylpyrophosphoroxy)phosphonomethoxy]-2-furanyl]ad-

Chain Terminators against Hepatitis C Virus
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8 2873
enine (3a). Diphosphate 3a (16 mg, 35%, tR ) 20 min) was
prepared from (2R,3R,4S,5R)-9-[tetrahydro-3,4-dihydroxy-5-
(phosphonomethoxy)-2-furanyl]adenine (3) (50 mg, 0.091 mmol)
by a similar procedure as that described for the synthesis of
36H), 2.00 (m, 4H), 1.11 (t, J ) 7.5 Hz, 48H), 0.79 (s, 3H). 31
P
NMR (D2O): δ 19.2, -10.3, -23.2. HRMS (M + H)+ for
C12H18N5O6P: calcd m/z 360.1073, obs 360.1085.
N6-Benzoyl-9-[2′,3′-di-O-benzoyl-5′,6′-dideoxy-6′-(diphen-
oxyphosphinyl)-â-D-ribohexofuranosyl]adenine (7). Tri-
fluoroacetic acid (0.15 mL) was added to a solution of N6-
benzoyl-9-(2′,3′-di-O-benzoyl-â-D-ribofuranosyl)adenine (6) (1.65
g, 2.8 mmol), 1,3-dicyclohexylcarbodiimide (2.34 g, 11.4 mmol),
and pyridine (0.32 mL) in anhydrous DMSO (15 mL). The
mixture was stirred at room temperature for 22 h. To the
mixture was added diphenyl(triphosphoranylidene)methyl-
phosphonate (2.89 g, 5.7 mmol) and stirring was continued for
24 h. Excess carbodiimide was hydrolyzed by a careful addition
of a solution of oxalic acid dihydrate (771 mg, 8.6 mmol) in
MeOH (3 mL). After stirring for 10 min, the volatiles were
removed under reduced pressure, and the formed urea was
removed by vacuum filtration. The filtrate was extracted with
EtOAc (300 mL). The organic solution was washed with brine
(2 × 100 mL), dried with Na2SO4, and concentrated to dryness.
Silica gel chromatography (EtOAc) yielded 1.43 g (62%) of N6-
benzoyl-9-[2′,3′-di-O-benzoyl-5′,6′-dideoxy-6′-(diphenoxyphos-
phinyl)-â-D-ribo-hex-5-enofuranosyl]adenine. To a solution of
the vinyl phosphonate (654 mg, 0.81 mmol) in MeOH (20 mL)
was added Pd-C (10%, 200 mg). The reaction mixture was
stirred under 30 psi of hydrogen. After 20 h, the mixture was
filtered through Celite to remove the catalyst, and the filtrate
was evaporated to dryness. Silica gel chromatography (EtOAc:
hexanes ) 75:25) yielded 505 mg (77%) of 7. 1H NMR (DMSO-
d6): δ 11.3 (br s, 1H), 8.76 (s, 1H), 8.74 (s, 1H), 8.06-7.13 (m,
25H), 6.61 (d, J ) 5.1 Hz, 1H), 6.46 (dd, J ) 5.1, 5.4 Hz, 1H),
5.99 (t, J ) 5.7 Hz, 1H), 4.61 (d, J ) 5.4 Hz, 1H), 2.37 (m,
4H).
N6-Benzoyl-9-[5′,6′-dideoxy-6′-(dibenzyloxyphosphinyl)-
â-D-ribo-hexofuranosyl]adenine (8). Sodium hydride (200
mg, 5.0 mmol) was added to 4 mL of benzyl alcohol. The
resulting suspension was stirred for 30 min. To the mixture
was added a solution of 7 (720 mg, 0.88 mmol) in anhydrous
DMSO (4 mL). The reaction was stirred at room temperature
for 4 h and quenched with saturated NH4Cl solution (20 mL).
The mixture was extracted with EtOAc (2 × 60 mL). The
combined organic solution was washed with brine (60 mL),
dried with Na2SO4, and concentrated to dryness. Silica gel
chromatography (EtOAc:MeOH ) 90:10) yielded 420 mg (75%)
of 8. 1H NMR (DMSO-d6): δ 11.2 (br s, 1H), 8.67 (s, 1H), 8.66
(s, 1H), 8.03 (d, J ) 7.5 Hz, 2H), 7.60 (m, 3H), 7.33 (m, 10H),
5.96 (d, J ) 5.4 Hz, 1H), 5.54 (d, J ) 5.7 Hz, 1H), 5.27 (d, J )
5.4 Hz, 1H), 4.96 (m, 4H), 4.73 (q, J ) 5.4 Hz, 1H), 4.10 (m,
1H), 3.91 (m, 1H), 1.91 (m, 4H).
(1R,4S)-N-[9-(4-Hydroxycyclopent-2-en-1-yl)-9H-purin-
6-yl]-N,N-dimethylformamidine (10). To a solution of a
(1R,4S)-9-(4-hydroxycyclopent-2-en-1-yl)adenine (9) (400 mg,
1.8 mmol) in DMF (10 mL) was added N,N-diethylformamide
diethyl acetal (0.47 mL, 2.8 mmol). After stirring at room
temperature for 5 h, the mixture was evaporated under
reduced pressure. Silica gel chromatography (CH2Cl2:MeOH
) 90:10) yielded 491 mg (98%) of 10. 1H NMR (CD3OD): δ
8.91 (s, 1H), 8.43 (s, 1 H), 8.24 (s, 1H), 6.29 (m, 1H), 6.05 (m,
1H), 5.60 (m, 1H), 4.85 (m, 1H), 3.30 (s, 6H), 3.06 (td, J ) 7.8,
15 Hz, 1H), 1.90 (td, J ) 3.6, 15 Hz, 1H).
(1R,4S)-9-[4-(Diethylphosphonomethoxy)-2,3-di-
hydroxycyclopent-2-en-1-yl]adenine (12). To a solution of
10 (271 mg, 1.0 mmol) and NaH (76 mg, 1.9 mmol) in DMF
(10 mL) at 0 °C was added diethyl p-toluenesulfonoxymethyl
phosphanate (11) (610 mg, 1.9 mmol). The reaction was slowly
warmed to room temperature and stirred for 16 h. The mixture
was neutralized with acetic acid, and the volatiles were
concentrated to dryness. The crude mixture was dissolved in
methanolic ammonia (40 mL, saturated at 0 °C) and stirred
at room temperature in a sealed bomb for 24 h. The bomb was
cooled to 0 °C before opening, and the reaction mixture was
concentrated to dryness. Silica gel chromatography (CH2Cl2:
MeOH ) 95:5) yielded 147 mg of a cyclopentene intermediate.
The material was dissolved in a mixture of acetone (6 mL),
t-BuOH (1 mL), and H2O (1 mL) along with 4-methylmorpho-
1
1a. H NMR (D2O): δ 8.35 (s, 1H), 8.13 (s, 1H), 6.08 (d, J )
6.6 Hz, 1H), 5.12 (s, 1H), 4.89 (dd, J ) 4.5, 6.6 Hz, 1H), 4.27
(d, J ) 4.5 Hz, 1H), 3.79 (dd, J ) 9.0, 13.2 Hz, 1H), 3.67 (dd,
J ) 10.2, 13.2 Hz, 1H), 3.02 (q, J ) 7.5 Hz, 36H), 1.11 (t, J )
7.5 Hz, 48H). 31P NMR (D2O): δ 8.5, -9.87, -22.4. MS (ES):
m/e 505.6 (M - H).
(1R,2R,3S,4S)-9-[2,3-Dihydroxy-4-(phosphonometh-
oxy)cyclopent-1-yl]adenine (4). To a solution of 12 (40 mg,
0.10 mmol) in DMF (4 mL) was added bromotrimethylsilane
(0.5 mL) at 0 °C. The reaction mixture was slowly warmed to
room temperature and stirred at the temperature for 6 h. The
volatiles were concentrated to dryness. The oil was diluted
with concentrated NH4OH (1.5 mL), which was subsequently
evaporated to yield a white solid. It was purified by a C18
reverse-phase column chromatography with a gradient from
100% triethylammonium acetate buffer (TEAA buffer, 50 mM)
to a mixture of 80% TEAA buffer and 20% CH3CN. The yield
was 28 mg (80%) of 4 as a white solid. 1H NMR (D2O): δ 8.11
(s, 1H), 7.93 (s, 1H), 4.59 (m, 1H), 4.30 (dd, J ) 5.1, 8.4 Hz,
1H), 4.07 (d, J ) 4.8 Hz, 1H), 3.83 (m, 1H), 3.55 (d, J ) 9.6
Hz, 2H), 3.00 (q, J ) 7.2 Hz, 12H), 2.70 (m, 1H), 1.88 (m, 1H),
1.09 (t, J ) 7.2 Hz, 18H). 31P NMR (D2O): δ 16.6. HRMS (M
+ H)+ for C11H16N5O6P: calcd m/z 346.0916, obs 346.0905.
(1R,2R,3S,4S)-9-[2,3-Dihydroxy-4-[(hydroxypyrophos-
phoroxy)phosphonomethoxy]cyclopent-1-yl]adenine (4a).
Diphosphate 4a (9.1 mg, 47%, tR ) 27 min) was prepared from
4 (21 mg, 0.038 mmol) by a similar procedure as that described
for the synthesis of 1a. 1H NMR (D2O): δ 8.30 (s, 1H), 8.12 (s,
1H), 4.75 (m, 1H), 4.51 (dd, J ) 4.8, 8.4 Hz, 1H), 4.15 (d, J )
6.0 Hz, 1H), 3.89 (m, 1H), 3.75 (dd, J ) 3.9, 9.6 Hz, 2H), 3.01
(m, 36H), 2.75 (m, 1H), 1.94 (m, 1H), 1.01 (m, 48H). 31P NMR
(D2O): δ 9.58, -9.89, -22.2. MS (ES): m/e 503.6 (M + H).
9-[5′,6′-Dideoxy-6′-(hydroxyphosphinyl)-2′-C-methyl-â-
D-ribo-hexofuranosyl]adenine (5). Compound 5 (31 mg,
60%) was prepared from 15 (123 mg, 0.17 mmol) by a similar
1
procedure as that described for the synthesis of 8 and 2. H
NMR (D2O): δ 8.17 (s, 1H), 8.14 (s, 1H), 5.94 (s, 1H), 3.92 (m,
1H), 3.85 (d, J ) 8.7 Hz, 1H), 1.95 (m, 2H), 1.66 (m, 2H), 0.79
(s, 3H). 31P NMR (D2O): δ 26.1. HRMS (M + H)+ for
C12H18N5O6P: calcd m/z 360.1073, obs 360.1085.
tert-Butyl SATE Phosphodiester of 9-[5′,6′-Dideoxy-6′-
(hydroxylphosphinyl)-2′-C-methyl-â-D-ribo-hexofuran-
osyl]adenine (5b). A solution of 19 (18 mg, 0.032 mmol) and
tributylamine (25 µL, 0.10 mmol) in methanol (0.5 mL) was
evaporated under reduced pressure. The residue was repeat-
edly coevaporated with anhydrous ethanol and toluene. To the
resulting oil were added pyridine (2 mL), thioester 16 (100
mg, 0.61 mmol), and 1-mestylene-2-sulfonyl-3-nitro-1,2,4-
triazole (38 mg, 0.13 mmol). The reaction mixture was stirred
at room temperature for 18 h and quenched by the addition of
tetrabutylammonium bicarbonate buffer (2 mL, 1 M solution,
pH 8.0). The mixture was extracted with CHCl3 (3 × 10 mL).
The organic solution was washed with brine (10 mL), dried
with Na2SO4, and concentrated to dryness to give the corre-
sponding 3′-TBDMS nucleoside phosphonate. To the crude
nucleoside in THF (3 mL) was added a mixture of tetrabut-
ylammonium fluoride (0.76 mL, 1 M in THF, 0.76 mmol) and
acetic acid (0.04 mL). The mixture was stirred at room
temperature for 16 h and evaporated under reduced pressure.
Silica gel chromatography (EtOAc:MeOH ) 95:5) yielded 10
mg (49%) of 5b. 1H NMR (CD3OD): δ 8.21 (s, 1H), 8.16 (s,
1H), 6.02 (s, 1H), 4.11 (m, 6H), 3.15 (t, J ) 6.6 Hz, 4H), 2.14
(m, 4H), 1.22 (s, 18H), 0.96 (s, 3H). HRMS (M + H)+ for
C26H42N5O8PS2: calcd m/z 648.2290, obs 648.2306.
9-[5′,6′-Dideoxy-6′-[(hydroxypyrophosphoroxy)phos-
phinyl]-2′-C-methyl-â-D-ribo-hexofuranosyl]adenine (5a).
Compound 5a (6.8 mg, 21%, tR ) 27 min) was prepared from
5 (22 mg, 0.061 mmol) by a similar procedure as that described
for the synthesis of 1a. 1H NMR (D2O): δ 8.13 (s, 2H), 5.95 (s,
1H), 3.94 (s, 1H), 3.85 (d, J ) 8.7 Hz, 1H), 3.02 (q, J ) 7.5 Hz,

2874 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8
Koh et al.
rine N-oxide (0.10 mL, 0.60 mmol). Subsequently, OsO4 (0.24
mL, 0.020 mmol) was added. The reaction mixture was stirred
at room temperature for 16 h, and the volatiles were concen-
trated to dryness. Silica gel chromatography (CH2Cl2:MeOH
) 95:5) yielded 96 mg (24% from 10) of 12. 1H NMR (CD3OD):
δ 8.22 (s, 1H), 8.17 (s, 1H), 4.86 (m, 1H), 4.59 (dd, J ) 4.8, 8.1
Hz, 1H), 4.20 (m, 5H), 3.99 (d, J ) 9.3 Hz, 2H), 3.95 (m, 1H),
2.82 (m, 1H), 2.17 (m, 1H), 1.37 (t, J ) 6.0 Hz, 6H). 31P NMR
(CD3OD): δ 23.0.
solution was washed with aqueous NaHCO3 solution (30 mL)
and brine (30 mL). It was dried with Na2SO4 and concentrated
to dryness to give N6-benzoyl-9-[5′,6′-dideoxy-6′-(dibenzyloxy-
phosphinyl)-2′-C-methyl-3′-O-tert-butyldimethylsilyl-â-D-ribo-
hexofuranosyl]adenine. The crude nucleoside was subjected to
a similar procedure as that described for the synthesis of 2 to
give 70 mg (95% from 16) of 19. 1H NMR (CD3OD): δ 8.24 (s,
1H), 8.21 (s, 1H), 6.02 (s, 1H), 4.25 (d, J ) 7.5 Hz, 1H), 4.02
(m, 1H), 2.05 (m, 4H), 0.99 (s, 9H), 0.98 (s, 3H), 0.25 (s, 3H).
MS (ES): m/e 474.3 (M + H).
N6-Dibenzoyl-9-(3′-O-benzoyl-2′-C-methyl-â-D-ribo-
furanosyl)adenine (14). To a solution of 13 (1.20 g, 4.2 mmol)
in DMF (10 mL) and pyridine (40 mL) was added tert-
butyldimethylsilyl chloride (633 mg, 4.2 mmol). The reaction
mixture was stirred at room temperature for 24 h and
quenched by addition of methanol (1 mL). The mixture was
evaporated under reduced pressure and diluted with CH2Cl2
(100 mL). The solution was washed with aqueous NaHCO3
solution (50 mL) and brine (50 mL). It was then dried with
Na2SO4 and concentrated to dryness to give the corresponding
5′-TBDMS nucleoside. To the crude nucleoside in pyridine (50
mL) was added benzoic anhydride (3.80 g, 16.8 mmol). The
mixture was heated at 100 °C for 16 h, cooled to room
temperature, and quenched by the addition of methanol (2
mL). The mixture was evaporated under reduced pressure and
diluted with CH2Cl2 (100 mL). The solution was washed with
aqueous NaHCO3 solution (50 mL) and brine (50 mL), dried
over Na2SO4, and concentrated to dryness. Silica gel chroma-
tography (EtOAc:hexanes ) 40:60) yielded 2.05 g (69% from
13) of N6-dibenzoyl-9H-(3′-O-benzoyl-5′-O-tert-butyldimethyl-
silyl-2′-C-methyl-â-D-ribofuranosyl)adenine. To this nucleoside
(448 mg, 0.63 mmol) in THF (10 mL) was added a mixture of
tetrabutylammonium fluoride (0.76 mL, 1 M in THF, 0.76
mmol) and acetic acid (0.04 mL). The mixture was stirred at
room temperature for 16 h and evaporated under reduced
pressure. Silica gel chromatography (EtOAc:hexanes ) 50:50)
HCV Replicon Assay and Cytotoxicity Studies. A
previously reported HCV subgenomic replicon system was used
to test the antiviral effect of adenosine phosphonate analogues
1-5 and a bis-SATE prodrug 5b in Huh7 cells. Conditions for
the assay were similar to those previously described.25 Briefly,
replicon cells were seeded in a 96-well plate at 6000 cells per
well. Compounds were added immediately after seeding. After
48-h incubation, cells were lysed, and luciferase activity in cell
lysate was measured. Reduction of HCV replicon RNA was
calculated on the basis of the reduction of luciferase activity
and expressed as EC50 (concentration to reduce 50% of viral
replication). The cytotoxicity of a compound in Huh7 cell was
determined by a colorimetric assay using 3-(4,5-dimethyl-
thiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) reagent.
The results were expressed as CC50 (concentration to reduce
50% of cell viability).
Nucleoside Phosphono Diphosphate Incorporation
Assays. HCV NS5B with a C-terminal His-tag and 21 amino
acid truncation was expressed and purified as reported.31 33P-
Labeled pGpC primer was prepared by incubating GpC di-
nucleotide (Sigma) with [γ-33P]ATP (3000 Ci/mmol, Perkin-
Elmer) and T4 polynucleotide kinase (Ambion) as described.25
A typical nucleotide incorporation assay was carried out at 30
°C in 10 µL of 50 mM HEPES (pH 7.3), 10 mM DTT, 5 mM
MgCl2, 20 µM RNA template, 20 µM of [5′-33P]GpC, and
various concentrations of an elongation nucleotide. For the
initial evaluation of nucleotide incorporation by NS5B RdRp,
the reaction was incubated for 1 h with 2.5 µM NS5B RdRp
and 1 mM compound or control ATP. For steady-state kinetic
analysis, a reaction was incubated for 15 min with 1 µM NS5B
RdRp. The assay was quenched by addition of 1 µL of 0.5 M
EDTA. Subsequently, the reaction mixture was treated with
10 µL of loading buffer (90% formamide, 0.025% bromophenol
blue, and 0.025% xylene cyanol) and heated to 70 °C for 2-5
min prior to loading 2 µL onto a denaturing 25% polyacryl-
amide-7 M urea-TBE gel. Electrophoresis was performed in
1× TBE at 70-90 W. Gels were visualized and analyzed using
a Phosphor Imager.
1
yielded 250 mg (67%) of 14. H NMR (DMSO-d6): δ 9.03 (s,
1H), 8.73 (s, 1H), 8.07-7.42 (m, 15H), 6.18 (s, 1H), 6.04 (s,
1H), 5.57 (d, J ) 8.7 Hz, 1H), 5.34 (t, J ) 4.8 Hz, 1H), 4.40
(m, 1H), 3.88-3.67 (m, 2H), 0.90 (s, 3H). MS (ES): m/e 594.2
(M+H).
N6-Benzoyl-9-[3′-O-benzoyl-5′,6′-dideoxy-6′-(diphen-
oxyphosphinyl)-2′-C-methyl-â-D-ribo-hexofuranosyl]adenine
(15). Compound 15 (132 mg, 45%) was prepared from 14 (247
mg, 0.42 mmol) by a similar procedure as that described for
1
the synthesis of 7. H NMR (CDCl3): δ 9.31 (s, 1H), 8.77 (s,
1H), 8.11-7.12 (m, 20H), 6.14 (s, 1H), 5.52 (d, J ) 6.6 Hz,
1H), 4.70 (s, 1H), 4.48 (m, 1H), 2.40 (m, 4H), 1.18 (s, 3H). MS
(ES): m/e 720.3 (M + H).
For steady-state kinetic analysis, apparent Michaelis con-
stant (Km) and the maximal velocity (Vmax) for a phosphono
diphosphate substrate was determined at the fixed concentra-
tion of template/primer (20 µM). The substrate concentration
was varied from Km/2 to 10Km. Initial velocity at each substrate
concentration was determined, and the data were fitted into
a Michaelis-Menten equation for calculation of Km and Vmax
using KaleidaGraph. The inhibitory experiments for a nucleo-
side phosphono diphosphate against ATP incorporation were
performed under the standard assay conditions as described
above. ATP concentration was fixed at 1 mM. Initial velocity
at various concentrations of inhibitor was determined, and the
inhibition constant Ki was calculated by plotting 1/vi vs
inhibitor concentration using a Dixon plot.
Cyclic tert-Butyl SATE Phosphoester of 9-[5′,6′-Dide-
oxy-6′-(hydroxylphosphinyl)-2′-C-methyl-â-D-ribo-hexo-
furanosyl]adenine (18). A solution of 5 (20 mg, 0.056 mmol)
and tributylamine (50 µL, 0.20 mmol) in H2O (0.5 mL) was
evaporated under reduced pressure. The residue was repeat-
edly coevaporated with anhydrous ethanol and toluene. To the
resulting oil were added pyridine (3 mL), thioester 16 (100
mg, 0.61 mmol), and 1-mesitylene-2-sulfonyl-3-nitro-1,2,4-
triazole (83 mg, 0.28 mmol). The reaction mixture was stirred
at room temperature for 18 h and quenched by the addition of
tetrabutylammonium bicarbonate buffer (2 mL, 1 M solution,
pH 8.0). The mixture was extracted with CHCl3 (3 × 10 mL).
The organic solution was washed with brine (10 mL), dried
with Na2SO4, and concentrated to dryness. Silica gel chroma-
tography (CH2Cl2:MeOH ) 95:5) yielded 13 mg (48%) of 18.
1H NMR (CD3OD): δ 8.26 (s, 1H), 8.24 (s, 1H), 6.04 (s, 1H),
4.18 (m, 4H), 3.26 (t, J ) 6.3 Hz, 2H), 2.51-1.95 (m, 4H), 1.22
(s, 9H), 1.08 (s, 3H). MS (ES): m/e 486.3 (M + H).
9-[3′-O-tert-Butyldimethylsilyl-5′,6′-dideoxy-6′-(hy-
droxyphosphinyl)-2′-C-methyl-â-D-ribo-hexofuranosyl]-
adenine (19). To a solution of 16 (114 mg, 0.18 mmol) and
imidazole (55 mg, 0.8 mmol) in DMF (3 mL) was added tert-
butyldimethylsilyl chloride (62 mg, 0.41 mmol). The reaction
mixture was stirred at 45 °C for 24 h and quenched by addition
of methanol (1 mL). The mixture was evaporated under
reduced pressure and diluted with CH2Cl2 (50 mL). The
Acknowledgment. We would like to acknowledge
Dr. Vicky Lai, who performed the HCV replicon assays;
Dr. Stephanie Shaw, for assisting with the manuscript
preparation; and Dr. Teresa Cain, for helping with the
purification of the phosphonates described in this paper.
References
(1) Choo, Q. L.; Kuo, G.; Weiner, A. J.; Overby, L. R.; Bradley, D.
W.; M., H. Isolation of a cDNA clone derived from a blood-borne
non-A, non-B viral hepatitis genome. Science 1989, 244, 359-
362.

Chain Terminators against Hepatitis C Virus
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8 2875
(2) World Health Organization Hepatitis C. Seroprevalence of
hepatitis C virus (HCV) in a population sample. Week. Epi-
demiol. Rec. 1996, 71, 346-349.
(3) Manns, M. P.; McHutchison, J. G.; Gordon, S. C.; Rustgi, V. K.;
Shiffman, M.; Reindollar, R.; Goodman, Z. D.; Koury, K.; Ling,
M.; Albrecht, J. K. Peginterferon alfa-2b plus ribavirin compared
with interferon alfa-2b plus ribavirin for initial treatment of
chronic hepatitis C: A randomised trial. Lancet 2001, 358, 958-
965.
(4) Wu, J. Z.; Hong, Z. Targeting NS5B RNA-dependent RNA
polymerase for anti-HCV chemotherapy. Curr. Drug Targets
Infect. Disord. 2003, 3, 207-219.
(5) Holy, A. Phosphonomethoxyalkyl analogues of nucleotides. Curr.
Pharm. Des. 2003, 9, 2567-2592.
(6) Johansson, N. G.; Eriksson, S. Structure-activity relationships
for phosphorylation of nucleoside analogues to monophosphates
by nucleoside kinases. Acta Biochim. Pol. 1996, 43, 143-160.
(7) De Clercq, E.; Holy, A.; Rosenberg, I.; Sakuma, T.; Balzarini,
J.; Maudgal, P. C. A novel selective broad-spectrum anti-DNA
virus agent. Nature 1986, 323, 464-467.
(8) Balzarini, J.; Naesens, L.; Herdewijn, P.; Rosenberg, I.; Holy,
A.; Pauwels, R.; Baba, M.; Johns, D. G.; De Clercq, E. Marked
in vivo antiretrovirus activity of 9-(2-phosphonylmethoxyethyl)-
adenine, a selective anti-human immunodeficiency virus agent.
Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 332-336.
(9) Balzarini, J.; Holy, A.; Jindrich, J.; Naesens, L.; Snoeck, R.;
Schols, D.; De Clercq, E. Differential antiherpesvirus and
antiretrovirus effects of the (S) and (R) enantiomers of acyclic
nucleoside phosphonates: Potent and selective in vitro and in
vivo antiretrovirus activities of (R)-9-(2-phosphonomethoxy-
propyl)-2,6-diaminopurine. Antimicrob. Agents Chemother. 1993,
37, 332-338.
(17) Pe´rigaud, C.; Gosselin, G.; Lefebvre, I.; Girardet, J. L.; Benzaria,
S.; Barber, I.; Imbach, J. L. Rational design for cytosolic delivery
of nucleoside monphosphates: “SATE” and “DTE” as enzyme-
labile transient phosphate protecting groups. Bioorg. Med. Chem.
Lett. 1993, 3, 2521-2526.
(18) Jones, G. H.; Moffatt, J. G. The synthesis of 6′-deoxyhomonucleo-
side 6′-phosphonic acids. J. Am. Chem. Soc. 1968, 90, 5337-
5338.
(19) Hollmann, J.; Schlimme, E. Darstellung und Konformation-
szuordnung einiger 5′-homologer adenosinderivate. Liebigs Ann.
Chem. 1984, 98-107.
(20) Jones, G. H.; Hamamura, E. K.; Moffatt, J. G. Stable Wittig
reagent suitable for the synthesis of R,â-unsaturated phospho-
nates. Tetrahedron Lett. 1968, 55, 5731-5734.
(21) Merlo, V.; Reece, F. J.; Roberts, S. M.; Gregson, M.; Storer, R.
Synthesis of optically active 5′-noraristeromycin: Enzyme-
catalysed kinetic resolution of 9-(4-hydroxycyclopent-2-enyl)-
purines. J. Chem. Soc. Perkin Trans. 1 1993, 15, 1717-1718.
(22) Dyatkina, N. B.; Theil, F.; Janta-Lipinski, M. Stereocontrolled
synthesis of the four stereoisomeric diphosphorylphosphonates
of carbocyclic 2′,3′-dideoxy-2′,3′-didehydro-5′-noradenosine. Tetra-
hedron 1995, 51, 761-772.
(23) Dyatkina, N.; Semizarov, D.; Victorova, L.; Krayevsky, A.; Theil,
F.; Janta-Lipinski, M. Synthesis of four stereisomers of carbo-
cyclic 5′-NOR D4A and evaluation of their triphosphates as
substrates for DNA polymerases. Nucleosides, Nucleotides 1995,
14, 723-726.
(24) Harry-O’Kuru, R. E.; Smith, J. M.; Wolfe, M. S. A short, flexible
route toward 2′-branched ribonucleosides. J. Org. Chem. 1997,
62, 1754-1759.
(25) Shim, J.; Larson, G.; Lai, V.; Naim, S.; Wu, J. Z. Canonical 3′-
deoxyribonucleotides as a chain terminator for HCV NS5B RNA-
dependent RNA polymerase. Antivir. Res. 2003, 58, 243-251.
(26) Shim, J. H.; Larson, G.; Wu, J. Z.; Hong, Z. Selection of
3′-template bases and initiating nucleotides by hepatitis C virus
NS5B RNA-dependent RNA polymerase. J. Virol. 2002, 76,
7030-7039.
(27) Ohno, Y.; Spriggs, D.; Matsukage, A.; Ohno, T.; Kufe, D.
Sequence-specific inhibition of DNA strand elongation by incor-
poration of 9-beta-D-arabinofuranosyladenine. Cancer Res. 1989,
49, 2077-2081.
(28) Ohno, Y.; Spriggs, D.; Matsukage, A.; Ohno, T.; Kufe, D. Effects
of 1-beta-D-arabinofuranosylcytosine incorporation on elongation
of specific DNA sequences by DNA polymerase beta. Cancer Res.
1988, 48, 1494-1498.
(29) Freeman, G. A.; Rideout, J. L.; Miller, W. H.; Reardon, J. E. 3′-
Azido-3′, 5′-dideoxythymidine-5′-methylphosphonic acid diphos-
phate: Synthesis and HIV-1 reverse transcriptase inhibition.
J. Med. Chem. 1992, 35, 3192-3196.
(30) Benzaria, S.; Maury, G.; Gosselin, G.; Rittinger, K.; Divita, G.;
Goody, R. S.; Imbach, J.-L. Synthesis of 2′,3′,5′-trideoxyuridine-
5′-methylphosphonic acid and its diphosphate derivative. Study
of the interaction with HIV-1 reverse transcriptase. Antivir.
Chem. Chemother. 1994, 5, 221-228.
(10) Suo, Z.; Johnson, K. A. Selective Inhibition of HIV-1 Reverse
transcriptase by an antiviral inhibitor, (R)-9-(2-phosphonyl-
methoxypropyl)adenine. J. Biol. Chem. 1998, 273, 27250-27258.
(11) Raju, N.; Smee, D. F.; Robins, R. K.; Vaghefi, M. M. Synthesis
and biological properties of purine and pyrimidine 5′-deoxy-5′-
(dihydroxyphosphinyl)-beta-D-ribofuranosyl analogues of AMP,
GMP, IMP, and CMP. J. Med. Chem. 1989, 32, 1307-1313.
(12) Secrist, J. A., 3rd; Riggs, R. M.; Comber, R. N.; Montgomery, J.
A. Synthesis of phosphonate analogues of dideoxyadenosine
(DDA)-, dideoxycytidine (DDC)-, dideoxyinosine (DDI)-, and
deoxythymidine
(DDT)-5′-monophosphates.
Nucleosides,
Nucleotides 1992, 11, 947-956.
(13) Montgomery, J. A.; Thomas, H. J.; Kisliuk, R. L.; Gaumont, Y.
Phosphonate analogue of 2′-deoxy-5-fluorouridylic acid. J. Med.
Chem. 1979, 22, 109-111.
(14) Kim, C. U.; Luh, B. Y.; Martin, J. C. Regiospecific and highly
stereoselective electrophilic addition to furanoid glycals: Syn-
thesis of phosphonate nucleotide analogs with potent activity
against HIV. J. Org. Chem. 1991, 56, 2642-2647.
(15) Khandazhinskaya, A. L.; Shirokova, E. A.; Skoblov, Y. S.;
Victorova, L. S.; Goryunova, L. Y.; Beabealashvilli, R. S.;
Pronyaeva, T. R.; Fedyuk, N. V.; Zolin, V. V.; Pokrovsky, A. G.;
Kukhanova, M. K. Carbocyclic dinucleoside polyphosphonates:
Interaction with HIV reverse transcriptase and antiviral activity.
J. Med. Chem. 2002, 45, 1284-1291.
(31) Zhong, W.; Ferrari, E.; Lesburg, C. A.; Maag, D.; Ghosh, S. K.;
Cameron, C. E.; Lau, J. Y.; Hong, Z. Template/primer require-
ments and single nucleotide incorporation by hepatitis C virus
nonstructural protein 5B polymerase. J. Virol. 2000, 74,
9134-9143.
(16) Carroll, S. S.; Tomassini, J. E.; Bosserman, M.; Getty, K.;
Stahlhut, M. W.; Eldrup, A. B.; Bhat, B.; Hall, D.; Simcoe, A.
L.; LaFemina, R.; Rutkowski, C. A.; 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.
JM049029U