Dual pro-drugs of 2′-C-methyl guanosine monophosphate as potent and selective inhibitors of hepatitis C virus

Article abstract of DOI:10.1016/j.bmcl.2011.06.013

We have previously reported the power of combining a 5′- phosphoramidate ProTide, phosphate pro-drug, motif with a 6-methoxy purine pro-drug entity to generate highly potent anti-HCV agents, leading to agents in clinical trial. We herein extend this work with the disclosure that a variety of alternative 6-substituents are tolerated. Several compounds exceed the potency of the prior 6-methoxy leads, and in almost every case the ProTide is several orders of magnitude more potent than the parent nucleoside. We also demonstrate that these agents act as pro-drugs of 2′-C-methyl guanosine monophosphate. We have also reported the novel use of hepatocyte cell lysate as an ex vivo model for ProTide metabolism.

Full text of DOI:10.1016/j.bmcl.2011.06.013

Bioorganic & Medicinal Chemistry Letters 21 (2011) 6007–6012
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Dual pro-drugs of 20-C-methyl guanosine monophosphate as potent and
selective inhibitors of hepatitis C virus
Christopher McGuigan a, , Karolina Madela a, Mohamed Aljarah a, Arnaud Gilles a, Srinivas K. Battina c,
⇑
Changalvala V. S. Ramamurty c, C. Srinivas Rao c, John Vernachio b, Jeff Hutchins b, Andrea Hall b,
Alexander Kolykhalov b, Geoffrey Henson b, Stanley Chamberlain b
a Welsh School of Pharmacy, Cardiff University, King Edward VII Avenue, Cardiff CF10 3NB, UK
b Inhibitex Inc., 9005 Westside Parkway, Alpharetta, GA 30009, USA
c CiVentiChem, 1001 Sheldon Drive, Cary, NC 27513, USA
a r t i c l e i n f o
a b s t r a c t
We have previously reported the power of combining a 50-phosphoramidate ProTide, phosphate pro-
drug, motif with a 6-methoxy purine pro-drug entity to generate highly potent anti-HCV agents, leading
to agents in clinical trial. We herein extend this work with the disclosure that a variety of alternative 6-
substituents are tolerated. Several compounds exceed the potency of the prior 6-methoxy leads, and in
almost every case the ProTide is several orders of magnitude more potent than the parent nucleoside.
We also demonstrate that these agents act as pro-drugs of 20-C-methyl guanosine monophosphate. We
have also reported the novel use of hepatocyte cell lysate as an ex vivo model for ProTide metabolism.
Ó 2011 Elsevier Ltd. All rights reserved.
Article history:
Received 10 May 2011
Revised 3 June 2011
Accepted 4 June 2011
Available online 17 August 2011
Keywords:
HCV
ProTide
Phosphoramidate
Nucleoside
Nucleotide
Polymerase
NS5B
Nucleoside analogues continue to play a vital role in the search
for improved therapies for hepatitis C Virus (HCV) infection.1 Sev-
eral families of modified nucleosides have been reported as inhib-
itors of the HCV NS5B RNA polymerase, including 40-modified2 and
20-modified3 nucleosides. All of these nucleosides act following
sequential phosphorylation to their bioactive 50-triphosphate
forms. As with many nucleoside analogues this phosphorylation,
and particularly the first, nucleoside kinase-mediated, step may
be limiting to their bio-activity.3 In these circumstances one of a
number of phosphate pro-drug methods may be used.4 We have
developed a phosphoramidate ProTide approach5 and applied it
extensively to a variety of bioactive nucleosides. Recently we re-
ported its effectiveness when applied to the anti-HCV agent 20-C-
⇑
Corresponding author. Tel./fax: +44 2920874537.
E-mail addresses: mcguigan@cf.ac.uk, mcguigan@cardiff.ac.uk (C. McGuigan).
methylguanosine (1).6 We further reported that combining the
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.06.013

6008
C. McGuigan et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6007–6012
Figure 1. Reagents and conditions: (a) DBU, TMSOTf, ACN, 65 °C, 4–6 h, 76%; (b) NH3/MeOH, 12 h, rt, 90% (for 6a); NaOMe, MeOH, rt, 12 h, 78% (for 6b); NaOEt/EtOH, 50 °C,
5 h, 84% (for 6c); MeNH2/Et3N/EtOH, 85 °C, sealed tube, 92% (for 6d); NaSMe/EtOH, rt, 18 h, 87% (for 6e); NH3/MeOH, rt, 12 h, 90%, NaH, 3-methoxy-1-propanol, 50% (for 6f);
BnCH2NH2, EtOH, reflux, 16 h, 55%, NH3/MeOH, rt, 94% (for 6g); BnNH2, EtOH, reflux, 16 h, 98%, NH3/MeOH, rt, 89% (for 6h); NaOCH(Me)2, rt, 18 h then NaOMe, rt, 18 h, 70%
(for 6i).
Figure 2. Reagents and conditions: (a) N-methylimidazole, THF, 16 h, rt, 11–28%; (b) tBuMgCl, THF, 22-25%.
50-phosphate ProTide moiety with a C6 methoxy pro-drug entity
further significantly boosted the potency of the compound to give
the clinical candidate INX-189 (2) with nanomolar activity against
HCV.7 This agent has recently successfully completed phase 1b
clinical trial in HCV infected patients.
Several reasons were suggested for the potency boost presented
on C6 modification. Amongst them was the ca 2-log increase in
lipophilicity upon this modification, and potential consequential
enhancements in passive cell uptake.7
Since the ProTide motif is thought to function by intracellular
50-monophosphate release it was considered that the 50-mono-
phosphate from (2) was a substrate for adenylate deaminase to lib-
erate 20-methylguanosine monophosphate as an essential
precursor to the bioactive 50-triphosphate. Given the known wide
substrate specificity of adenylate deaminase8 we wondered if
alternative 6-substituents might substitute for the 6-methoxy
and might further boost lipophilicity and enhance potency. We
herein report the notable success of this approach.
(4) in the presence of TMS triflate to give the protected beta nucle-
oside (5) in 76% yield. This was deprotected with methanolic
ammonia to give (6a) in 78% yield. Use of alternatives deprotec-
tion/substitution conditions gave the various 6-substituted
analogues (6b–i) as shown in Figure 1. Thus, sodium methoxide
and ethoxide respectively gave (6b–c), methylamine (6d), and
thiomethoxide gave (6e). Other analogues were similarly
prepared.
The anti-HCV activity of (6a–i) was studied in sub-genomic rep-
licon assay and the data are reported in Table 1. In general the
compounds are active at
C6 substituents.
lM levels; activity decreasing for larger
Following the discovery of the power of the naphthyl neopen-
tylalanine ProTide motif in (2) we now applied this moiety to
(6a–i) as shown in Figure 2. Thus, 1-naphthyl neopentylalaninyl
phosphorochloridate was allowed to react with (6a–i) to generate
the corresponding ProTides (7a–i) in moderate yield.
All compounds were isolated as roughly equimolar mixtures of
phosphorus diastereoisomers as evidenced by 31P NMR and HPLC.9
The ProTides were tested as inhibitors of HCV as noted above,
with data being shown in Table 2.
As shown in Figure 1 we used the 6-chloro nucleoside as a route
into various 6-substituted analogues. Thus, tetrabenzoyl 20-C-
methyl-D-ribose (3) was condensed with 2-amino-6-chloropurine

C. McGuigan et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6007–6012
6009
Figure 3. Putative mechanism of ProTide activation.
Figure 4. Carboxypeptidase Y mediated hydrolysis of (8a) followed by 31P NMR.
It is striking that while all of the parent nucleosides (6a–i) are
poorly active, with EC50 values of ca 5–30 M, all of the ProTides
are active sub- M. Particularly active are the 6-ethoxy, 6-methoxy
and 6-chloro compounds with EC90values of <100 nM. The largest
C6 substituted compounds such as (7h) seem to lose some activity,
implying some size restrictions at C6, but even here the ProTide re-
mains 30-fold more active than the parent (6h). The ProTides are
To confirm the possible mode of action of these compounds we
conducted a number of assays. In the first instance we performed a
Carboxypeptidase Y assay, an in vitro probe of ProTides activation
pathway (Fig. 3).
l
l
The data shown in Figures 4 and 5 represent 31P NMR spectra
recorded every 7 min during 14 h incubation period of (8a) and
(8e) with Carboxypeptidase Y in acetone-d6 and Trizma buffer
(pH 7.6).
cytotoxic at
lM levels but their high potency still leads to thera-
peutic index values in the 1000 range. Given the extremely high
potency of the 6-O-ethoxy analogue we decided to vary the ProTide
motif on (7c). Following the same procedures as in Figure 2 but
varying the phosphorochloridate we converted (6c) to (8a–f)
(Table 3).
In case of the naphthyl benzylalanine compound (8a) (Fig. 4),
the experiment showed fast hydrolysis of starting material
(dP = 3.62, 4.11 ppm) to the intermediate lacking the ester group
(dP = 4.69, 4.83 ppm). Both diastereoisomers of (8a) appear to be
processed with roughly similar efficacy as far as the data
allow us to discern. The single peak at 6.95 ppm corresponds to
the final product of the hydrolysis—the achiral aminoacyl phos-
phate. The estimated half-life of (8a) was less than 5 min. In case
All of these compounds show sub-lM activity in replicon but
none are as active as (7c). In general, the alanine compounds are
more active than the valine ones as we have previously noted.6

6010
C. McGuigan et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6007–6012
Figure 5. Carboxypeptidase Y mediated hydrolysis of (8e) followed by 31P NMR.
Figure 6. Metabolism of (8c) in HuH7 cell lysate followed by 31P NMR.
of the isopropyl analogue (8e) (Fig. 5) conversion of the starting
material was slower, with a half-life of 4 h and in this case one
of the diastereoisomers of (8e) seems to be processed rather more
rapidly.
These results are consistent with the replicon data, showing
that (8a), which is processed faster, is also found to be more active
ence of newly formed peak at 0.98 ppm was observed suggesting
successful liberation of the monophosphate species (Fig. 6). This
signal had the same chemical shift as an authentic monophosphate
under these conditions. Given that its release is considered to be a
pre-requisite for antiviral action, via the triphosphate, we view
these data as encouraging and consistent with the replicon data
in Tables 1–3 above.
(EC50 = 0.04 lM vs 0.14 lM for (8e)). The carboxypeptidase assay is
thus a potentially useful predictive model for the in vitro bio-acti-
vation of these agents.
As far as we are aware, this is the first time that P-31 NMR of
hepatocyte lysate has been used as an ex vivo assay of anti-HCV
ProTide metabolism.
Since we regard each of these 6-modified nucleotides as pro-
drugs of the guanine nucleotides, we also conducted an adenosine
deaminase (ADA) assay on (6b), with spectra recorded each min-
ute.10 The results are shown in Fig 7.
HuH7 cell lysates have been prepared to examine monophos-
phate formation during the ProTide hydrolysis. An NMR based as-
say was performed on (8c) using HuH7 cell lysate (107 cells) in
acetone-d6 and Trizma buffer (pH 7.6) at 37 °C. 31P NMR spectra
were recorded every 1 h for 11 h. After 1 h of incubation, the pres-

C. McGuigan et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6007–6012
6011
Table 1
HCV Replicon activity and cytotoxicity of 6-substituted 20-C-methylguanosine
nucleosides
Compound
R1
EC50
/
l
Ma
CC50/ l
Mb
6a
6b
6c
6d
6e
6f
6g
6h
6i
Cl
OMe
OEt
NHMe
SMe
O(CH2)3 OMe
NHCH2 Bn
NHBn
8.5
4.6
8.8
13
11
12
24
30
>100
>100
>100
>100
>100
>100
>100
ND
OCH(CH3)2
9.3
>100
For structures see Figure 1.
a
50% effective concentration. Replicon data for genotype 1b in HUH7 cells with
48 h exposure.
Figure 7. Adenosine deaminase mediated deamination of (6b) followed by UV
spectroscopy.
b
50% cytotoxic concentration in HUH7 cells.
has an identical UV spectrum to (1) as expected. Compound (6c)
was found to be suitable substrate for the enzyme, however the
conversion rate to (1) was much slower than for (6b). We expect
that the monophosphates of (6a–c) released from the ProTide
in vitro, may similarly be substrates for adenylate deaminase, giv-
ing the 20-C-methylguanosine monophosphate in each case. ADA
assay performed on (6d) and (6e) showed no transformation with-
in the reaction time of 24 h which may suggest that broader sub-
strate specificity might operate at the adenylate deaminase level
or that the deamination of these nucleoside analogues and their
corresponding monophosphates is catalysed by a different type
of deaminase for example, human abacavir monophosphate deam-
inase.11 To further pursue this notion we conducted the replicon
anti-viral assay in the presence and absence of pentostatin, a
known inhibitor of both adenosine and adenylate deaminase (data
not shown) and noted that each of the ProTides herein described
lost entirely their activity in the presence of pentostatin, clearly
confirming an absolute need for ‘deamination’ to be active.
Table 2
HCV Replicon activity and cytotoxicity of 1-naphthyl neopentyl alanine ProTides of 6-
substituted 20-C-methylguanosine nucleosides
Compound
R1
EC50
/
a l
M
EC90
/
b l
M
CC50 / M
c l
7a
7b
7c
7d
7e
7f
7g
7h
7i
Cl
OMe
OEt
NHMe
SMe
O(CH2)3OMe
NHCH2Bn
NHBn
0.012
0.010
0.008
0.035
0.037
0.043
0.270
0.79
0.028
0.038
0.032
0.103
0.16
0.12
1.30
2.7
0.065
16
7
10
38
41
31
42
15
11
OCH(CH3)2
0.023
For structures see Figure 1.
a
50% effective concentration. Replicon data for genotype 1b in HUH7 cells with
48 h exposure.
b
90% effective concentration. Replicon data for genotype 1b in HUH7 cells with
48 h exposure.
c
50% cytotoxic concentration in HUH7 cells.
In conclusion, we herein report that a wide range of 6-substitu-
ents is acceptable at the ProTide level of 20-C-methyl guanosine
based anti-HCV agents. In every case, the parent nucleosides are
Table 3
HCV Replicon activity and cytotoxicity of various ProTides of 6-ethoxy-20-C-
methylguanosine
poorly active, being active at only high
lM levels. In every case
their ProTides are active sub- M; in most cases the potency of
l
the nucleoside is enhanced >100 fold. Alanine emerges again as
the preferred amino acid, and several ester variants lead to very
potent compounds. In some cases the potency and selectivity
in vitro exceeds that of the clinical candidate (2), INX-08189. In
every case ‘deamination’ at the C6-position, presumably at the
monophosphate level, is implicated as essential for activity, and
thus these compounds are dual pro-drugs. We have also reported
the novel use of hepatocyte cell lysate as an ex vivo model for
ProTide metabolism.
Acknowledgements
Compound
R1
R2
R3
EC50
/
l
Ma
CC50/l
Mb
8a
8b
8c
8d
8e
8f
OEt
OEt
OEt
OEt
OEt
OEt
Bn
0.04
0.17
11
24
12
47
40
17
L
L
L
L
L
L
-Ala
This study was supported by a grant from Inhibitex, Inc. to C.M.
C.M. is a board member and shareholder of Inhibitex, Inc. All
authors from Inhibitex, Inc. own options and are shareholders.
We would like to thank Helen Murphy for secretarial assistance.
NeoPnt
S-PhEt
THP
-Val
-Ala
-Ala
-Ala
-Ala
0.085
0.08
iPr
0.14
2,4F2Bn
0.034
References and notes
a
50% effective concentration. Replicon data for genotype 1b in HUH7 cells with
48 h exposure.
1. Brown, N. A. Expert Opin. Invest Drugs 2009, 18, 709.
2. Klumpp, K.; Leveque, V.; Le Pogam, S.; Ma, H.; Jiang, W.-R.; Kang, H.;
Granycome, C.; Singer, M.; Laxton, C.; Hang, J. Q.; Sarma, K.; Smith, D. B.;
Heindl, D.; Hobbs, C. J.; Merrett, J. H.; Symons, J.; Cammack, N.; Martin, J. A.;
Devos, R.; Najera, I. J. Biol. Chem. 2006, 281, 3793.
3. Eldrup, A. B.; Prhavc, M. I.; Brooks, J.; Bhat, B.; Prakash, T. P.; Song, Q.; Bera, S.;
Bhat, N.; Dande, P.; Cook, P. D.; Bennett, C. F.; Caroll, S. S.; Ball, R. G.; Bosserman,
b
50% cytotoxic concentration in HUH7 cells.
These data demonstrate that (6b) is an excellent substrate for
adenosine deaminase-mediated deamination. The final product

6012
C. McGuigan et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6007–6012
M.; Burlein, C.; Colwell, L. F.; Fay, J. F.; Flores, O. A.; Gerry, K.; LaFemina, R. L.;
Leone, J.; MacCoss, M.; McMasters, D. R.; Tomassini, J. E.; Langen, D. V.;
Wolanski, B.; Olsen, D. B. J. Med. Chem. 2004, 47, 5284.
1H NMR (500 MHz, MeOD-d4) d 8.21–8.14 (m, 1H, H-8 napht), 7.98 and 7.96
(2s, 1H, H-8), 7.86–7.79 (m, 1H, H-5 napht), 7.66 and 7.63 (2d, J = 8.3 Hz, 1H, H-
4 napht), 7.56–7.43 (m, 3H, H-7, H-6, H-3 napht), 7.41–7.34 (m, 1H, H-2 napht),
6.03 and 6.02 (2s, 1H, H-10), 4.67–4.71 (m, 2H, H-50, H-500), 4.54–4.47 (2q,
J = 7.0 Hz, 2H, OCH2CH3), 4.37-4.25 (m, 2H, H-30, H-40), 4.12–4.04 (m, 1H, CH
Ala, 3.69 and 3.65 (2 AB systems, J = 10.5 Hz, 2H, CH2 ester), 1.41 (t, J = 7.1 Hz,
3H, OCH2CH3), 1.33 (d, J = 7.1 Hz, 3H, CH3 Ala), 0.98 and 0.96 (2s, 3H, CH3), 0.84
and 0.82 (2s, 9H, 3ꢀ CH3 ester)
4. Gundmundsson, K. S.; Wang, Z. C.; Daluge, S. M.; Johnson, L. C.; Hazen, R.;
Condreay, L. D.; McGuigan, C. Nucleos Nucleotide Nucl. 2004, 23, 1929.
5. McGuigan, C.; Cahard, D.; Sheeka, H. M.; DeClercq, E.; Balzarini, J. J. Med. Chem.
1996, 39, 1748.
6. McGuigan, C.; Gilles, A.; Madela, K.; Aljarah, M.; Holl, S.; Jones, S.; Vernachio, J.;
Hutchins, J.; Ames, B.; Bryant, D.; Gorovits, E.; Ganguly, B.; Hunley, D.; Hall, A.;
Kolykhalov, A.; Liu, Y.; Muhammad, J.; Raja, N.; Walters, R.; Wang, J.;
Chamberlain, S.; Henson, G. J. Med. Chem. 2010, 53, 4949.
7. McGuigan, C.; Madela, K.; Aljarah, M.; Gilles, A.; Brancale, A.; Zonta, N.;
Chamberlain, S.; Vernachio, J.; Hutchins, J.; Hall, A.; Ames, B.; Gorovits, E.;
Ganguly, B.; Kolykhalov, A.; Wang, J.; Muhammad, J.; Patti, J. M.; Henson, G.
BMCL 2010, 20, 4850.
8. Simon, L. N.; Bauer, R. J.; Tolman, R. L.; Robins, R. K. Biochemistry 1970, 9, 573.
9. Procedure for 7c. To nucleoside (1 mol equiv) in anhydrous THF,
phosphorochloridate (3 mol equiv) in THF was added, followed by addition of
N-methyl-imidazole (5 mol equiv). The mixture was stirred overnight under
argon atmosphere. Solvent was removed under reduce pressure. To remove the
N-methyl-imidazole, the phosphoramidate was dissolved in chloroform and
washed 3 times with hydrochloric acid (HCl 0.5 N). The organic layer was then
dried over sodium sulfate and evaporated under reduce pressure. The residue
was then purified on silica gel using CHCl3 to CHCl3/MeOH 95:5 as an eluent, to
give the pure phosphoramidate as a white solid. Starting from 500 mg of
nucleoside. Yield 28% (285 mg).
3
13C NMR (126 MHz, MeOD-d4) d 175.06, 174.80 (2d, JC-C-N-P = 5.0 Hz, C@O),
162.39 (C-6), 161.87, 161.86 (C-2), 154.57, 154.53 (C-4), 148.01, 147.96 (2d, 2JC-
O-P = 3.6 Hz, C-1 napht), 139.32, 139.04 (C-8), 136.29, 136.26 (C-10 napht),
128.84, 128.79 (C-5 napht, 127.91, 127.86 (C-6 napht), 127.75, 127.72 (C-7
napht), 126.53, 126.40 (C-9 napht), 125.96, 125.93 (C-3 napht), 122.81, 122.76
(C-4 napht), 116.23, 116.18 (2d, 3JC-C-O-P = 2.9 Hz, C-2 napht), 115.63, 115.62 (C-
3
5), 93.34, 93.20 (C-10), 82.33, 82.16 (2d, JC-C-O-P = 8.0 Hz, C-40), 79.98, 79.94
(CH2 ester), 75.37 (C-20), 74.96, 74.72 (C-30), 68.14, 67.67 (2d, 2JC-O-P = 5.3 Hz, C-
50), 63.59, 63.57 (OCH2CH3), 51.79, 51.72 (CH Ala), 32.26, 32.23 (C(CH3)3 ester),
26.72, 26.69 (C(CH3)3 ester), 20.84, 20.64 (2d, 3JC-C-N-P = 6.6 Hz, CH3 Ala), 20.34,
20.31 (20-CH3), 14.89 (OCH2CH3) 31P NMR (202 MHz, MeOD-d4) d 4.28 and 4.21.
HPLC tR = 22.92, 23.16 min (Varian Polaris C18-A (10 lM) analytic column;
elution was performed using a mobile phase consisting of water/acetonitrile in
gradient 90/10 to 0/100 v/v in 30 min).
10. 1 ml of 44
lM solution of nucleoside in phosphate buffer (pH7.4), 30 lL of ADA
solution. Spectra recorded in 1 min intervals for 30 min.
11. Schinkmanova, M.; Votruba, I.; Shibata, R.; Han, B.; Liu, X.; Cihlar, T.; Holy, A.
Collect. Czech. Chem. Commun. 2008, 2, 275–291.