The application of the phosphoramidate ProTide approach confers micromolar potency against Hepatitis C virus on inactive agent 4′-azidoinosine: Kinase bypass on a dual base/sugar modified nucleoside

Article abstract of DOI:10.1016/j.bmcl.2009.03.138

Novel phosphoramidate ProTides derived from 4′-azidoinosine have been prepared and evaluated in the replicon assay against hepatitis C Virus (HCV). The parent nucleoside analogue is inactive in this assay, while the ProTides are active at low μM levels in

Full text of DOI:10.1016/j.bmcl.2009.03.138

Bioorganic & Medicinal Chemistry Letters 19 (2009) 3122–3124
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The application of the phosphoramidate ProTide approach confers
micromolar potency against Hepatitis C virus on inactive agent 40-azidoinosine:
Kinase bypass on a dual base/sugar modified nucleoside
a
b
b
Christopher McGuigan a, , Felice Daverio , Isabel Nájera , Joseph A. Martin ,
*
Klaus Klumpp b, David B. Smith b
a Welsh School of Pharmacy, Cardiff University, King Edward VII Avenue, Cardiff CF10 3NB, UK
b Roche Palo Alto LLC, 3431 Hillview Avenue, Palo Alto, CA 94304, USA
a r t i c l e i n f o
a b s t r a c t
Novel phosphoramidate ProTides derived from 40-azidoinosine have been prepared and evaluated in the
replicon assay against hepatitis C Virus (HCV). The parent nucleoside analogue is inactive in this assay,
while the ProTides are active at low lM levels in some cases. This is a rare example of an inosine nucle-
oside analogue with potent antiviral activity and further supports the notion of ProTides as a drug discov-
ery motif.
Article history:
Received 23 February 2009
Revised 24 March 2009
Accepted 25 March 2009
Available online 29 March 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
HCV
Protides
Nucleotides
Hepatitis C virus
Prodrugs
Antiviral
According to the World Health Organization (WHO) more than
170 million people are infected with the hepatitis C virus (HCV),
representing 3% of the world population.1 Current therapy consists
of ribavirin and pegylated interferon. Both agents are broad spec-
trum and more specific therapies for HCV are under development
in many laboratories.2
rate-limiting, and that the free parent nucleoside monophosphates
are of limited direct use in therapy due to poor membrane perme-
ation and instability to dephosphorylation.7 Thus, we8 and others9–
12
have developed approaches to the intracellular delivery of pre-
formed nucleotides via pro-drug (‘ProTide’) approaches. Indeed,
application of our phosphoramidate ProTide approach to 40-azido-
One of our laboratories has reported extensively on various
40-substituted ribonucleoside analogues as specific inhibitors of
HCV, with 40-azidocytidine (1) emerging as an important lead.3
The 20,30,50-triester pro-drug of this agent has demonstrated effi-
cacy in phase 2a studies in HCV infected patients. 4–5 Further stud-
ies have focused on new nucleoside analogs with improved potency
by affecting either intrinsic potency or phosphorylation efficiency.
The limitations imposed by host nucleoside kinases can be severe
and many nucleoside analogues are inactive due to poor phosphor-
ylation. A good example in the HCV space is 40-azidouridine (2).
uridine (2) notably conferred sub-lM HCV potency on the inactive
parent,6 this being taken as proof of successful monophosphate
delivery in vitro by the ProTide and further supporting the notion
that (2) (and perhaps other analogues) fail as antivirals due to a
poor initial phosphorylation step.
The combination of nucleosides and nucleotides with different
bases may increase the overall efficiency of nucleoside analogs
when used in combination therapy. In an attempt to identify
new series of nucleotides with antiviral potency against HCV, we
herein report the application of this technology for the first time
to an inosine nucleoside, 40-azidoinosine (3) with positive results
emerging.
Thus, while the 50-triphosphate of (2) is a potent, sub-
lM, inhibitor
of HCV RNA polymerase,6 the nucleoside itself (2) is inactive as an
inhibitor of HCV in the cell based replicon assay, presumably due
to poor phosphorylation in vitro.6
O
NH2
N
O
N
N
NH
NH
It is considered that the first phosphorylation step of parent
(1)
(2)
(3)
nucleoside analogues to their 50-monophosphates is often
N
O
N
O
N
HO
N3
HO
N3
HO
N3
O
O
O
* Corresponding author. Tel./fax: +44 29 20874537.
E-mail address: mcguigan@cardiff.ac.uk (C. McGuigan).
OH OH
OH OH
OH OH
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.03.138

C. McGuigan et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3122–3124
3123
3.3–3.7. Proton NMR signals for the base moiety (H2, H8) also
showed splitting due to the diastereomers. Each of the ProTides
5a–h were evaluated for their ability to inhibit the proliferation
of HCV in the replicon assay as described.3
Data are presented in Table 1 as EC50 values in
the concentration of compound reducing HCV replication by 50%.
All of the compounds showed cytotoxicity (CC50) values >100 M.
The parent compound (3) did not inhibit HCV replication (EC50
>100 M).
In contrast several of the ProTides did inhibit HCV replication at
low M levels. These data support the notion that (3) is inactive due
lM, representing
l
l
l
to poor phosphorylation in vitro, that it is the first phosphorylation
step that is rate limiting, and that some of the ProTides can success-
fully deliver the 50-monophosphate of (3) intracellularly.
Scheme 1.
Considering previous SARs reported by us for ProTides in gen-
eral8 and for HCV in particular6,13 we chose to concentrate this ini-
tial study on alanine-based ProTides, we have previously found
benzyl esters of alanine to be a particularly effective ProTide motif
in vitro, and ethyl and isopropyl esters to substitute for benzyl in
many cases.8 t-Butyl esters have in general been found to be poorly
effective, and this has been attributed to poor cleavage of these es-
ters in vitro, this cleavage being an essential first step in ProTide
activation.16 Thus, it was not surprising that the phenyl t-butylala-
Thus 40-azidoadenose was prepared from inosine in a 5-step
route similar to that we have published for the corresponding aden-
osine analogue.13 Briefly, inosine was 50-iodinated using triphenyl
phosphine and iodine, the product protected as its 20,30-bis
(TBDMS) ether and the 40,50-olefin generated under basic conditions
(Scheme 1).
nine compound 5c was poorly active (EC50>100 lM). However, it
Stereo- and regio-specific reaction of the olefin with DMDO affor-
ded the key epoxide intermediate13,14 which was ring opened with
TMS azide and tin(IV) chloride to give the protected 40-azido nucle-
oside. Desilylation with TBAF give 40-azidoinosine (3). As we have
previously noted, protection of the 20,30-diol with a cyclopentylid-
ene group enhances the yield and regio selectively of 50-ProTide
formation,6,13 and so (3) was thus protected. Appropriate phosphor-
ylating agents were prepared by methods we have extensively
reported.6,13 Thus, phenyl or 1-naphthyl phosphorodichloridates
were allowed to react with various alanine ester hydrochlorides to
give the target phosphorochloridates, which were reacted with
20,30-O-cyclopentylidene-40-azidoinosine in THF in the presence of
t-butyl magnesium chloride, to yield protected ProTides 4a–h
(Scheme 2).
In both phenyl and 1-naphthl series the esters prepared were
ethyl, i-propyl, t-butyl and benzyl. Following coupling, the pro-
tected ProTides were deprotected to yield target compounds 5a-
h.15 Coupling yields ranged from 48% to 100% while deprotection
yields were in general ca. 70–80% except for the t-butyl esters 5c
and 5g which were 20–30%. The reason for the reduced yield on
the deprotection step in these cases is unclear. In each case the Pro-
Tides were isolated as roughly 1:1 mixtures of phosphate diastere-
oisomers as evidenced by two closely spaced 31P NMR signals as ca.
was a surprise that the corresponding ethyl (5a), isopropyl (5b)
and benzyl (5d) esters were also inactive. One possible reason for
the poor activity of this phenyl family could be their relatively
low lipophilicity (Table 1) with ClogP values of ca. ꢀ1 to ꢀ2. While
they are still ca. 2ꢀ logs more lipophilic than (3), their lipophilicity
values may predict poor passive diffusion into cells.
Table 1
a
Compd
Aryl
Ester
%
EC50
lM
ClogPb
3
—
Ph
Ph
Ph
—
Et
—
>100
>100
>100
>100
>100
16.4
6.1
ꢀ3.6
ꢀ2.1
ꢀ1.8
ꢀ1.4
ꢀ1.0
ꢀ1.0
ꢀ0.7
ꢀ0.3
0.2
5a
5b
5c
5d
5e
5f
68, 76
87, 73
71, 27
57, 70
48, 62
70, 80
100, 18
59, 66
iPr
tBu
Bn
Et
iPr
tBu
Bn
Ph
Nap
Nap
Nap
Nap
5g
5h
>100
4.3
a
Isolated yield of 5a–h over each of the 2 steps from protected (3).
Calculated logP (ClogP) based on ChemDraw Ultra 11.0.1.
b
O
N
N
N
O
NH
N
NH
O
HO
N
O
N
O
HO
N3
Me
O
O
P
H
N3
O
PTSA, 100 oC, 2h
70%
N
C
H
Cl
O
O
R
O
O
OH OH
Ar
tBuMgBr
THF, RT, 17h
O
N
O
N
N
NH
N
NH
O
Me
O
P
H
N
Me
O
O
P
C
H
O
N
N
H
80% aq HCOOH
R
O
C
H
O
N
O
O
R
O
Ar
O
O
N3
Ar
RT, 5h
(5a-h)
N3
O
O
(4a-h)
For Ar, R, and
% see Table
OH OH
Scheme 2.

3124
C. McGuigan et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3122–3124
6. Perrone, P.; Luoni, G.; Kelleher, M. R.; Daverio, F.; Angell, A.; Mulready, S.;
With this in mind and also noting our recent observations of the
Congiatu, C.; Rajyaguru, S.; Martin, J.; Levêque, V.; Le Pogam, S.; Najera, I.;
Klumpp, K.; Smith, D. B.; McGuigan, C. J. Med. Chem. 2007, 1840.
7. Hersh, M. R.; Kuhn, J. G.; Philips, J. L. Cancer Chemother. Pharmacol. 1986, 17,
277.
8. Cahard, D.; McGuigan, C.; Balzarini, J. Mini-Rev. Med. Chem. 2004, 4, 371.
9. Poijärvi-Virta, P. Curr. Med. Chem. 2006, 13, 3441.
particular efficacy of 1-naphthyl ProTides,6,13,17 we prepared a par-
allel series of naphthyl ProTides of alanine esters.
Although the ethyl compound (5e) was only active above 10
the isopropyl (5f) and benzyl (5h) esters were active at low
lM,
lM
levels. Notably, the t-butyl ester (5g) remained inactive in this sys-
tem. In each case, the naphthyl phosphates were calculated to be ca
10-times more lipophilic than their phenyl analogues. To some
extent there was a tendency towards potency increasing with lipo-
philicity, which may support the notion that cell entry was limiting,
but the benzyl ester in the phenyl series (5d) and the ethyl ester in
the naphthyl series (5e) happened to share a common logP and yet
only the naphthyl system was active, indicating lipophilicity to be
at best only part of the reason for the different profiles. The altered
pKa and leaving group ability of the 1-naphthyl group may be an
additional parameter, as loss of the aryl moiety is considered to
be the essential second step in ProTide activation. This is the first
example that we are aware of where a family of ProTides depends
for its activity on the presence of a naphthyl moiety rather than a
phenyl group.
10. Hecker, S. J.; Erion, M. D. J. Med. Chem. 2008, 51, 2328.
11. Wagner, C. R.; Iyer, V. V.; McIntee, E. J. Med. Res. Rev. 2000, 20, 417.
12. Gisch, N.; Balzarini, J.; Meier, C. J. Med. Chem. 2007, 50, 1658.
13. Perrone, P.; Daverio, F.; Valente, R.; Rajyaguru, S.; Martin, J.; Le´vêque, V.; Le
Pogam, S.; Najera, I.; Klumpp, K.; Smith, D. B.; McGuigan, C. J. Med. Chem. 2007,
50, 5463.
14. Haraguchi, K.; Takeda, S.; Tanaka, H. Org. Lett. 2003, 5, 1399.
15. Selected synthetic procedures and spectroscopic data: 20,30-O-cyclopentan-
dienide-40-azidoinosine-50-phenyl(ethoxy-
L
-alaninyl)phosphate (4a) To
a
solution of 20,30-O-cyclopentadienide-40-azidoinosine (55 mg, 0.147 mmol), in
THF a 1 M solution of tert-butylmagnesium chloride in THF (1 M soln in THF,
366
15 min. A 1 M solution of phenyl ethylalaninyl phosphorochloridate in THF
(0.366 mmol, 366 L) was added dropwise. The mixture was stirred at rt for
lL) was added dropwise and the mixture left to equilibrate at rt for
l
14 h. and the solvent removed and the crude product purified by silica column
chromatography using CHCl3/MeOH (from 95/5) as eluent. The appropriate
fractions were collected and the solvent removed under reduced pressure to
afford a white solid (63 mg, 68%). 31P NMR (MeOD; 202.5 MHz): d 3.54, 3.17. 1
H
NMR (MeOD; 500 MHz): d 8.26, 8.25 (1H, 2s, H-8), 8.05 (1H, s, H-2), 7.91–7.12
(5H, m, PhO), 6.51–6.50 (1H, m, H-10), 5.41–5.36 (1H, m, H-20), 5.27–5.25 (1H,
m, H-30), 4.32–4.26 (2H, m, H-50), 4.23–4.10 (2H, m, CH3CH2), 3.92–3.87 (1H, m,
CH3CH), 2.27–2.13 (2H, m, cyclopentylidene), 1.86–1.71 (6H, m, cyclopen-
tylidene), 1.32, 1.31 (3H, 2d, 3J = 7.00 Hz, CH3CH), 1.24 (3H, t, 3J = 7.15 Hz,
CH3CH2).
In conclusion, we have developed a synthesis of 40-azidoinosine
from inosine and reported the preparation of a family of 8 ProTides
thereof.
While the parent nucleoside analogue is inactive against HCV in
replicon, some of the ProTides are active at low lM levels. Notably,
activity depends on the presence of a naphthyl phosphate, and is
lost for t-butyl esters. This study represents one of very few high-
lighting activity from inosine based systems and suggests that the
potential of nucleosides based on this base motif may be unleashed
by ProTide methods.
13C NMR (MeOD; 125.7 MHz): d 14.47 (CH3CH2), 20.35, 20.31 (CH3CH), 24.09,
24.81, 36.27, 36.36, 37.31 (CH2, cyclopentylidene), 51.64 (CH3CH), 62.42
(CH3CH2), 69.25, 69.29 (C-50), 83.60, 83.64 (C-30), 85.38 (C-20), 90.29, 90.55 (C-
10), 100.58, 100.66 (C40), 121.28, 121.32, 121.41, 121.45, 126.28, 126.34,
130.75, 130.79 (PhO + C-5), 141.28 (C-8), 147.27 (C-2), 149.36 (C-4), 151.88,
151.93 (‘ipso’, PhO), 158.85 (C6), 174.97 (COOEt). 40-azidoinosine-50-
[phenyl(ethoxy-
L
-alaninyl)] phosphate. (5a) 20,30-O-cyclopentandienide-40-
-alaninyl)]phosphate (4a) (60 mg, 0.095
azidoinosine-50-phenyl (ethoxy-
L
mmol) was dissolved in HCOOH (80% v/v solution in water, 10 mL) and the
mixture stirred at rt for 6 h. The solvent was removed and the crude purified by
silica column chromatography using CHCl3/MeOH (gradient elution from 95/5
to 9/1) as eluent to give 41 mg of a white solid (76%). 31P NMR (MeOD;
202.5 MHz): d 3.50, 3.31. 1H NMR (MeOD; 500 MHz): d 8.28, 8.26 (1H, 2s, H-8),
8.07, 8.03 (1H, s, H-2), 7.36–7.15 (5H, m, PhO), 6.32–6.30 (1H, m, H-10), 4.94–
4.91 (1H, m, H-20), 4.69–4.66 (1H, m, H-30), 4.35–4.20 (2H, m, H-50), 4.16–4.09
(2H, m, CH3CH2), 3.94–3.82 (1H, m, CH3CH), 1.32, 1.31 (3H, 2d, 3J = 7.15 Hz,
CH3CH), 1.23 (3H, t, 3J = 7.10 Hz, CH3CH2). 13C NMR (MeOD; 125.7 MHz): d
14.44, 14.49 (CH3CH2), 20.25, 20.31, 20.45, 20.50 (CH3CH), 51.50, 51.64
(CH3CH), 62.45 (CH3CH2), 68.43, 68.47, 68.70, 68.74 (C-50), 73.99, 74.23,
74.39 (C-30 + C-20), 91.08 (C-10), 99.06, 99.23, 99.31 (C-40), 121.31, 121.33,
121.36, 121.41, 121.15, 126.27 130.77, 130.83 (PhO + C-5), 141.08, 141.15 (C-
8), 147.06 (C-2), 150.04 (C-4), 151.92, 151.97 (‘ipso’, PhO), 158.90 (C-6), 174.02,
175.05 (COOEt).
Acknowledgement
We would like to thank Helen Murphy for excellent secretarial
assistance.
References and notes
1. Boyer, N.; Marcellin, P. J. Hepatol. 2000, 32, 98.
2. Santanonio, T.; Wiegand, J.; Gerlach, J. T. J. Hepatol. 2008, 4, 625.
3. Smith, D. B.; Martin, J. A.; Klumpp, K.; Baker, S. J.; Blomgren, P. A.; Devos, R.;
Granycome, C.; Hang, J.; Hobbs, C. J.; Jiang, W-R.; Laxton, C.; Le Pogam, S.;
Leveque, V.; Ma, H.; Maile, G.; Merrett, J. H.; Pichota, A.; Sarma, K.; Smith, M.;
Swallow, S.; Symons, J.; Vesey, D.; Najera, I.; Cammack, N. Bioorg. Med. Chem.
Lett. 2007, 2570.
4. Pockros, P. J.; Nelson, D.; Godofsky, E. Hepatology 2007, 46, 311.
5. Najera, I.; Le Pogam, S.; Jiang, W. Resistance Profile for 40-Azido-cytidine
(R1479) reveals lack of cross resistance with 20-C-methyl-cytidine, Interferon-
Alfa and Ribavirin. 46th ICAAC San Francisco Abstract, 2006, V1909.
16. McGuigan, C.; Sutton, P. W.; Cahard, D.; Turner, K.; O’Leary, G.; Wang, Y.;
Gumbleton, M.; De Clercq, E.; Balzarini, J. Antivir. Chem. Chemother. 1998, 9,
473.
17. McGuigan, C.; Thiery, J.-C.; Daverio, F.; Jiang, W. G.; Davies, G.; Mason, M.
Bioorg. Med. Chem. 2005, 13, 3219.