Phosphorodiamidates as a promising new phosphate prodrug motif for antiviral drug discovery: Application to anti-HCV agents

Article abstract of DOI:10.1021/jm2011673

We herein report phosphorodiamidates as a significant new phosphate prodrug motif. Sixty-seven phosphorodiamidates are reported of two 6-O-alkyl 2′-C-methyl guanosines, with significant variation in the diamidate structure. Both symmetrical and asymmetric phosphorodiamidates are reported, derived from various esterified amino acids, both D and L, and also from various simple amines. All of the compounds were evaluated versus hepatitis C virus in replicon assay, and nanomolar activity levels were observed. Many compounds were noncytotoxic at 100 μM, leading to high antiviral selectivities. The agents are stable in acidic, neutral, and moderately basic media and in selected biological media but show efficient processing by carboxypeptidases and efficiently yield the free nucleoside monophosphate in cells. On the basis of in vitro data, eight leads were selected for additional in vivo evaluation, with the intent of selecting one candidate for progression toward clinical studies. This phosphorodiamidate prodrug method may have broad application outside of HCV and antivirals as it offers many of the advantages of phosphoramidate ProTides but without the chirality issues present in most cases.

Full text of DOI:10.1021/jm2011673

Article
pubs.acs.org/jmc
Phosphorodiamidates as a Promising New Phosphate Prodrug Motif
for Antiviral Drug Discovery: Application to Anti-HCV Agents
Christopher McGuigan,*,† Karolina Madela,† Mohamed Aljarah,† Claire Bourdin,† Maria Arrica,†
Emma Barrett,† Sarah Jones,† Alexander Kolykhalov,‡ Blair Bleiman,‡ K. Dawn Bryant,‡ Babita Ganguly,‡
Elena Gorovits,‡ Geoffrey Henson,‡ Damound Hunley,‡ Jeff Hutchins,‡ Jerry Muhammad,‡
Aleksandr Obikhod,‡ Joseph Patti,‡ C. Robin Walters,‡ Jin Wang,‡ John Vernachio,‡
Changalvala V. S. Ramamurty,§ Srinivas K. Battina,§ and Stanley Chamberlain‡
†Welsh School of Pharmacy, Cardiff University, King Edward VII Avenue, Cardiff CF10 3NB, U.K.
‡Inhibitex Inc., 9005 Westside Parkway, Alpharetta, Georgia 30004, United States
§CiVentiChem, 1001 Sheldon Drive, Cary, North Carolina 27513, United States
S
* Supporting Information
ABSTRACT: We herein report phosphorodiamidates as a significant new phosphate prodrug motif. Sixty-seven
phosphorodiamidates are reported of two 6-O-alkyl 2′-C-methyl guanosines, with significant variation in the diamidate structure.
Both symmetrical and asymmetric phosphorodiamidates are reported, derived from various esterified amino acids, both D and L,
and also from various simple amines. All of the compounds were evaluated versus hepatitis C virus in replicon assay, and
nanomolar activity levels were observed. Many compounds were noncytotoxic at 100 μM, leading to high antiviral selectivities.
The agents are stable in acidic, neutral, and moderately basic media and in selected biological media but show efficient processing
by carboxypeptidases and efficiently yield the free nucleoside monophosphate in cells. On the basis of in vitro data, eight leads
were selected for additional in vivo evaluation, with the intent of selecting one candidate for progression toward clinical studies.
This phosphorodiamidate prodrug method may have broad application outside of HCV and antivirals as it offers many of the
advantages of phosphoramidate ProTides but without the chirality issues present in most cases.
cell uptake, a number of phosphate (nucleotide) prodrug motifs
have emerged.4 These have included our aryl phosphoramidate
(“ProTide”) approach,5 the amidate diester method of Wagner6
and McKenna,7 the lipid diester approach of Hosteller,8
thioester approaches of Gosselin,9 cytochrome based meth-
ods,10 and the chemical driven cycloSAL method of Meier.11
Each of these methods has its strengths and weaknesses. In
general, fully blocked prodrugs (ProTides, CycloSAL, etc.) may
give better delivery but do often generate a chiral phosphate
center leading to isomer issues. Phosphate diester methods
avoid the chirality issue but may have delivery challenges.4
Despite this, several phosphate prodrugs of antiviral nucleosides
have progressed into clinical trials. This includes Inhibitex’s
INX-18912 (5, Figure 2) and Pharmasset’s PSI-797713 (6), both
based directly on phosphoramidate ProTides, and Idenix’s
INTRODUCTION
■
There is an urgent ongoing need for improved therapeutic
agents for hepatitis C Virus (HCV) infection, with an
increasing emphasis on direct acting antivirals (DAAs), and in
particular, inhibitors of the viral NS5B RNA polymerase.1
Nucleoside-based inhibitors of the polymerase are considered
particularly valuable on the basis of the high genetic barrier to
resistance.2 Thus, a number of nucleoside modifications (1−4,
Figure 1) have emerged with anti-HCV activity in vitro, and
several have progressed to clinical evaluation.
One issue common to all nucleoside analogues, either
antiviral or anticancer, is an absolute need for nucleoside
kinase-mediated activation to their 5′-monophosphate forms. In
some cases, as in the present anti-HCV field, further
phosphorylation to the 5′-triphosphate is also required, and in
general the first phosphorylation step is considered rate-
limiting.3 With this in mind, and given that the free phosphate
derivatives are not considered useful drug entities due to poor
Received: September 2, 2011
Published: October 31, 2011
© 2011 American Chemical Society
8632
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645

Journal of Medicinal Chemistry
Article
Figure 1. Some anti-HCV nucleoside analogues.
Figure 2. Some anti-HCV phosphate prodrug compounds in clinical trials.
a
Scheme 1. Synthetic Route to Phosphorodiamidates
a
Reagents and conditions: (a) Et3N (1 equiv), POCl3 (1 equiv), dry THF; (b) for symmetrical diamidates, amine (5 equiv), Et3N (10 equiv), dry
DCM; for asymmetric diamidates, amine 1 (1 equiv), Et3N (2 equiv) then amine 2 (5 equiv), Et3N (10 equiv), dry DCM.
IDX18414 (7), which is a hybrid phosphoramidate/SATE
prodrug. Each of these is now in human trials for efficacy versus
HCV.
Most chiral nucleotide prodrugs that have entered into the
CHEMISTRY
■
Following the success of 5 in early clinical trials for HCV23 and
our recent disclosure that alternative 6-alkoxy (and other)
groups may be equally effective,24 we used the 6-alkoxy
derivatives of 2′-C-methylguanosine as our starting point. These
were prepared by methods we have described.12 Following our
much earlier reports on phosphoramidates of AZT,17 we used a
single synthetic route based on treating the unprotected
nucleoside with phosphoryl chloride to generate the
intermediate dichloridate which was not isolated (Scheme 1).
In the first instance, we used the 6-methoxy analogue 8, the
base nucleoside behind 5. Reaction with POCl3 in THF in the
presence of Et3N for 45 min (when 31P NMR showed no
presence of POCl3) followed by addition of an excess of various
amino acid esters (or other amines), lead to compounds 10−76
as shown in Table 1.
As L-alanine is often a preferred motif in phosphoramidate
ProTides and the final putative activating step for these
diamidates would be the same loss of amino acid, we thought
that L-alanine may be beneficial here, so our first SAR family
was symmetrical substituted L-alanines with linear esters
methyl-(10), ethyl-(11), n-propyl-(12), n-butyl-(13), and n-
pentyl-(14); branched esters isopropyl-(15) and (R,S)-2-butyl-
(16), and 3,3-dimethylbutyl-(17); followed by cyclic esters
cyclobutyl-(18), cyclopentyl-(19) cyclohexyl-(20), and 2-
tetrahydropyranyl-(24). We have often reported that benzyl
esters can be rather effective in ProTides, so we prepared the
benzyl-(21), S-phenethyl-(22), 2,4-difluorobenzyl-(23), and 2-
indanyl-(25) L-alanine esters. Given the extremely high potency
clinic to date have been progressed as diastereoisomeric
mixtures at the phosphate center. NewBiotics’ anticancer
agent NB1011 is a further example in this category.15 In
general, when tested, both stereoisomers often display similar
biological activity in vitro and always release the same
pharmacophore after initial metabolism.16 One exception to
this in the HCV field to date is Pharmasset, who has developed
a large scale separation technique for their mixed compound
PSI-7851, now pursued as the single stereoisomer 6 as above.13
With this in mind, we sought to revisit the notion of an
achiral phosphate prodrug motif with a phosphoramidate core.
In addition, we particularly wanted to formulate a prodrug
whose promoieties were nontoxic and preferably natural. Thus
our attention turned to phosphorodiamidates. Indeed, over 20
years ago, our group was among the first to report
phosphorodiamidate-based nucleotide prodrugs. Thus, seeking
to deliver the 5′-monophosphate of the anti-HIV agent AZT
into cells, we prepared a series of L-amino acid phosphor-
odiamidates with a variety of esters.17 There have since been
reports of this motif on phosphonates18 and non-nucleosides19
but little further work on nucleosides.20−22
In this paper, we report the very promising profile of a family
of such phosphorodiamidates of 2′-C-methylguanosine and its
derivatives, leading toward selection of a clinical candidate.
8633
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645

Journal of Medicinal Chemistry
Article
Table 1. Replicon Data and Cytotoxicity of Nucleoside Analogues and Phosphorodiamidates
EC50 (μM)
EC90 (μM)
av SD
8.3
compd
AA ester/amine
AA ester/amine
nucleoside
2′CMeG
av
2.2
SD
CC50 (μM)
4
1.4
7.5
>100
>100
>100
7.01
>100
>100
>100
>100
65
8
6OMe2′CMeG
6OEt2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
4.4
9.5
2.2
19.4
24.5
10.2
9
1.84
0.01
0.76
0.08
0.05
0.01
0.01
0.18
0.00
0.01
0.05
0.01
0.02
1.78
5
L-Ala OCH2tBu
NaphO
0.01
5.90
1.20
0.28
0.07
0.03
0.49
0.15
0.02
0.32
0.06
0.05
0.49
0.49
0.26
13.33
0.58
0.06
0.07
0.21
0.11
10.18
0.61
0.60
0.13
0.45
0.38
0.47
5.17
4.00
0.40
2.90
0.60
0.25
2.22
0.50
0.32
0.05
0.52
0.81
2.50
0.12
0.72
0.11
0.32
0.27
0.54
3.80
37.80
0.04
>10
0.02
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
L-Ala OMe
L-Ala OMe
L-Ala OEt
L-Ala OEt
4.10
1.20
0.23
0.11
2.10
0.53
0.09
0.80
0.22
0.19
1.60
1.70
1.07
29.30
1.30
0.20
0.26
0.66
0.50
0.09
0.53
0.02
0.02
0.09
0.10
0.07
L-Ala OnPr
L-Ala OnPr
L-Ala OButyl
L-Ala OPentyl
L-Ala OiPr
L-Ala OButyl
L-Ala OPentyl
L-Ala OiPr
>100
>100
71
L-Ala O-(R,S)-2Bu
L-Ala O-3,3-dimethylbutyl
L-Ala OcBu
L-Ala O-(R,S)-2Bu
L-Ala O-3,3-dimethylbutyl
L-Ala OcBu
>100
>100
>100
>100
88
L-Ala OcPentyl
L-Ala OcHx
L-Ala OcPentyl
L-Ala OcHx
0.03
0.08
L-Ala OBn
L-Ala OBn
L-Ala OSPhEt
L-Ala O-2,4-diFBn
L-Ala OTHP
L-Ala OIndanol
L-Ala OCH2tBu
L-Ala OCH2iPr
L-Ala OCH2cPropyl
D-Ala OCH2tBu
L-Asp OMe
L-Ala OSPhEt
L-Ala O-2,4-diFBn
L-Ala OTHP
L-Ala OIndanol
L-Ala OCH2tBu
L-Ala OCH2iPr
L-Ala OCH2cPropyl
D-Ala OCH2tBu
L-Asp OMe
0.21
13.02
0.39
0.04
0.01
0.00
0.03
3.17
0.08
0.13
0.02
0.52
0.02
0.15
1.94
0.47
0.11
0.33
0.17
0.04
0.92
15.14
0.10
54
>100
70
0.12
>100
>100
>100
>100
>100
>100
>100
>100
20
0.01
0.001
0.19
>40
L-Asp OBn
L-Asp OBn
1.90
1.80
0.60
2.00
1.31
2.40
0.14
0.09
0.20
2.15
0.21
0.71
2.66
1.16
0.10
1.36
0.11
0.24
L-Gly OBn
L-Gly OBn
L-Gly OCH2tBu
L-Leu OcHx
L-Gly OCH2tBu
L-Leu OcHx
L-Leu OBn
L-Leu OBn
39
L-Leu OCH2tBu
L-Ile OMe
L-Leu OCH2tBu
L-Ile OMe
27
29.5
>100
14
L-Ile OcHx
L-Ile OcHx
7.10
1.56
6.30
2.60
0.77
5.00
2.00
1.35
0.21
1.80
2.00
10.00
0.66
2.50
0.50
1.80
1.50
2.40
L-Ile OBn
L-Ile OBn
24
L-Ile OCH2tBu
L-Met OcHx
L-Ile OCH2tBu
L-Met OcHx
L-Met OBn
15
51
L-Met OBn
>100
>100
25
L-Met OCH2tBu
L-Phe OcHx
L-Met OCH2tBu
L-Phe OcHx
L-Phe OBn
0.59
0.14
0.02
0.21
0.05
0.35
0.04
2.20
1.07
0.12
0.23
0.21
0.72
0.35
L-Phe OBn
67
L-Phe OCH2tBu
L-Pro OBn
L-Phe OCH2tBu
L-Pro OBn
24
>100
56
L-Pro OCH2tBu
L-Val OcHx
L-Pro OCH2tBu
L-Val OcHx
20
L-Val OBn
L-Val OBn
49
L-Val OCH2tBu
L-Tyr (tBu) OMe
L-PhG OcHx
L-PhG OCH2tBu
L-Val-L-Ala OCH2tBu
β-Ala OBn
L-Val OCH2tBu
L-Tyr (tBu) OMe
L-PhG OcHx
L-PhG OCH2tBu
L-Val-L-Ala OCH2tBu
β-Ala OBn
32
0.04
0.00
0.03
0.26
2.03
3.80
0.10
0.04
0.09
0.46
89
18
24
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>10
butylamine
butylamine
>40
morpholine
morpholine
>100
>100
L-Ala OcPentyl
L-Ala OCH2tBu
L-Ala OCH2tBu
L-Ala OCH2tBu
L-Ala OCH2tBu
L-Ala OCH2tBu
L-Ala OcHx
0.04
0.15
0.15
0.04
0.49
0.17
0.01
0.04
0.01
0.01
0.16
0.03
0.07
0.08
0.02
L-Ala OBn
0.40
0.57
0.15
2.20
0.69
L-Ala OtBu
L-Ala OcHx
L-Pro OMe
L-Val OMe
0.06
0.15
8634
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645

Journal of Medicinal Chemistry
Article
Table 1. continued
EC50 (μM)
av
EC90 (μM)
av SD
1.68
compd
AA ester/amine
L-Ala OBn
AA ester/amine
butylamine
nucleoside
SD
CC50 (μM)
65
66
67
68
69
70
71
72
73
74
75
76
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OMe2′CMeG
6OEt2′CMeG
6OEt2′CMeG
6OEt2′CMeG
0.44
0.41
0.11
0.87
0.43
0.27
0.13
3.70
0.87
0.05
0.29
0.04
0.28
0.28
0.02
0.11
0.02
0.01
0.05
0.94
0.75
0.05
0.45
0.11
0.00
0.18
>100
>100
>100
>100
>100
>100
>100
>100
>100
69
L-Ala OCH2tBu
L-Ala OCH2tBu
L-Ala OCH2tBu
L-Ala OCH2tBu
L-Ala OCH2tBu
L-Ala OCH2tBu
L-Ala OCH2tBu
L-Ala OCH2tBu
L-Ala OcHx
butylamine
pentylamine
cyclopropylamine
BnNH
1.50
0.52
4.70
1.90
0.82
0.56
PhNH
NaphNH
diethylamine
pyrrolidine
L-Ala OcHx
L-Ala OBn
L-Ala OCH2tBu
>10
0.02
0.00
0.04
0.01
4.00
0.15
1.00
0.11
0.16
0.02
0.29
0.04
L-Ala OBn
62
L-Ala OCH2tBu
72
of the neopentyl alanine compound as a naphthyl ProTide
(5),12 we were keen to explore the methylene bridged family
starting with the neopentyl-(26) and 2-methylpropyl-(27)
compounds and also the methylene cyclopropyl-(28).
On the basis of our recent report that the 6-ethoxy group can
well substitute for the purine 6-methoxy as a core nucleoside
for anti-HCV ProTides,21 we applied the diamidate method to
the O6-ethyl-2′-C-methylguanosine nucleoside, preparing the
symmetrical cyclohexyl-(74), benzyl-(75), and neopentyl-(76)
L-alanine analogues.
In aryl phosphoramidate ProTides, we have reported on
several occasions the sometimes strong preference, for L-amino
acids over D-analogues,25 and all of the first examples reported
here are L-amino acid derived. However, we did synthesize the
neopentyl-D-alanine analogue 29 of 8.
Although L-alanine is generally strongly preferred, other
natural and unnatural amino acids are also effective as
ProTides,26 and so we did vary the amino acid motif itself.
An attempt was made to consistently select from the same three
ester groups, neopentyl, cyclohexyl, and benzyl, for each new
amino acid to facilitate comparison. Thus, symmetrical
diamidates were prepared from dimethyl-(30) and dibenzyl-
(31) L-aspartic acid, benzyl-(32) and neopentyl-(33) glycine,
cyclohexyl-(34), benzyl-(35), and neopentyl-(36) L-leucine,
methyl-(37), cyclohexyl-(38), benzyl-(39), and neopentyl-(40)
L-isoleucine, cyclohexyl-(41), benzyl-(42), and neopentyl-(43)
L-methionine, cyclohexyl-(44), benzyl-(45), and neopentyl-
(46) L-phenylalanine, benzyl-(47) and neopentyl-(48) L-
proline, and cyclohexyl-(49), benzyl-(50), and neopentyl-(51)
L-valine. The tyrosine methyl ester diamidate was also prepared
as its para-O-tert-butyl ether-(52). Diamidates of the unnatural
amino acid L-phenylglycine were prepared as its cyclohexyl-
(53) and neopentyl-(54) esters. In a prior program on
ProTides of d4T for HIV, we reported a complete loss of
activity on extending from alanine to β-alanine and beyond,27
but it was unclear whether similar restrictions would apply here,
hence we prepared the symmetrical benzyl-β-alanine diamidate-
(56).
unknown, and so we prepared the symmetrical butylamine-(57)
and morpholinyl-(58) diamidates. Notably, from a synthetic
perspective, none of the above amino acid and amine variations
presented particular challenges although the yields from these
reactions were not high and remain unoptimized.
Besides seeking to probe the SAR and potential advantages
of symmetrical diamidates, as above, we also wondered if the
synthetic method would allow access to asymmetrically mixed
diamidates. To do this, we slightly adapted the synthetic route
(Scheme 1) to allow the stepwise introduction of two separate
amino acids or one amino acid and one amine. To limit the
large number of possible combinations, and to facilitate
interpretation of the data, one of the amines was generally
kept constant as neopentyl L-alanine and variations were made
in the second amine. Initially, different L-alanine esters, benzyl-
(60), tert-butyl-(61), and cyclohexyl-(62), were combined with
neopentyl L-alanine, followed by different amino acids methyl L-
proline-(63) and methyl L-valine-(64), and then a number of
different simple amines (65−73). Synthetically, it did not
matter much which amine was introduced first. Each of these
asymmetric diamidates was isolated as a roughly 1:1 mixture of
phosphate diastereoisomers as revealed by 31P NMR and
HPLC. No attempt was made to separate the diastereoisomers,
and they were tested as mixtures.
In this way, a substantial set of symmetrical and asymmetric
phosphorodiamidates of 6-O-methyl-2′-C-methylguanosine and
three derivatives of the 6-O-ethyl analogue were prepared and
fully characterized. As expected, all of the symmetrical
diamidates reported above were observed as one peak by 31P
NMR and one signal on analytical HPLC. The 31P NMR shifts
of the symmetrical amino aryl diamidates were ca. 13 ppm,
being rather downfield of our usual aryloxy phosphoramidate
ProTides.28 The asymmetric diamidates gave two peaks by 31P
NMR and for HPLC, in roughly 1:1 ratios, as typically observed
for phosphoramidate ProTides. Other spectroscopic and
analytical data fully confirmed the structure and purity of the
diamidates herein reported.
Finally, in this series, we wondered if the chemistry
methodology would extend to a dipeptide and whether such
an adduct might be active and so we successfully incorporated
an L-valyl-L-alanine neopentyl ester (55).
Simple amines have not been found to be useful as the amino
component of aryl phosphoramidates, although Idenix has
successfully incorporated benzylamine into their clinical
analogue IDX184.14 However, for diamidates, the SAR was
BIOLOGICAL ACTIVITY IN REPLICON
■
Each of the diamidates described above were tested for HCV
inhibition in a replicon assay, with the clinical anti-HCV
ProTide (5) as positive control (Table 1). Both EC50 and EC90
values are reported, along with standard deviations, in a 72 h
HCV replicon assay. In general EC90 values were 2−5-fold
higher than EC50 values, and the discussion below will focus on
8635
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645

Journal of Medicinal Chemistry
Article
Table 2. Replicon Activity for Symmetrical Phosphorodiamidates
replicon EC50 (μM)
(R2)
amino acid
L-alanine
compd
cyclohexyl
0.05
benzyl
neopentyl
CC50 in Huh7 cells (μM)
20
21
19
>100
>100
>100
0.49
0.06
0.11
D-alanine
glycine
29
>100
32
33
0.60
0.38
0.40
>100
>100
0.13
0.47
2.9
a
L-leucine
L-isoleucine
L-valine
34
35
36
0.45
20
39
27
38
39
40
4.0
2.5
14
24
15
49
50
51
20
49
32
0.12
0.52
b
0.72
L-proline
47
48
>100
56
0.81
L-methionine
41
42
43
0.60
0.32
51
0.25
>100
>100
b
2.22
L-phenylglycine
L-phenylalanine
53
54
18
24
0.27
0.05
a
44
45
46
0.50
25
67
24
0.32
0.61
L-aspartic (OBn)
31
>100
a
b
Large standard deviation. Single assay result.
the EC50 numbers. In addition, the cell cytotoxicities (CC50) in
the replicon cell line (Huh7) are reported.
increase in potency for the n-alkyl esters as they extended from
methyl (10 EC50 = 6 μM) to n-pentyl (14 EC50 = 0.03 μM).
The n-pentyl L-alanine 14 is thus about 3 times less active than
5, but notably it is also about 10 times less cytotoxic. It is
interesting to note that the calculated lipophilicity (ClogP)
values for this series range from 0.5 (methyl) to 4.7 (n-pentyl).
It maybe that lipophilicity and potency correlate, however,
there are exceptions. Clearly, the n-pentyl 14 is an interesting
compound with a very attractive SI.
Branching the amino acid ester at the α (Table 1, 15, 16), β
(26, 27), or γ (17) position does not significantly change
activity or toxicity relative to the straight chain compounds.
Thus, the isopropyl ester 15 and the n-propyl ester 12 have
As noted in Table 1, the three parent nucleosides, 2′-C-
methyl guanosine 4 and the 6-methoxy 8 and 6-ethoxyl 9
analogues, display only modest anti-HCV activity, with EC50
values in the 2−10 μM range. By comparison, 5 is active at
nanomolar levels, with an EC50 of 10 nM and an EC90 of 40
nM, representing a ca. 400-fold potency boost over the
respective nucleoside.12 Compound 5 does show some
cytotoxicity to Huh7 cells in this assay at 7 μM, but its high
potency still leads to a significant SI of ca. 700.
Examining the potency of the first family of symmetrical L-
alanine phosphorodiamidates, we note a consistent and clear
8636
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645

Journal of Medicinal Chemistry
Article
similar activities, and the β branched neopentyl ester 26 is
similar to the n-pentyl ester 14. Indeed, several of these
branched esters such as compounds 17 and 27 are very potent,
with excellent SI values. We were particularly interested in the
neopentyl L-alanine analogue 26 given its similarity to our
clinical agent 5. This compound reveals an EC50 of 0.06 μM
and EC90 of 0.2 μM and is only 5-fold less active in this assay
than 5. However, in common with many of the phosphor-
odiamidates, this compound is significantly less cytotoxic, with
CC50 >100 μM. Thus, the SI for 26 is >1600, which exceeds
that of 5.
Scheme 2. Putative Initial Activation Route of
Phosphorodiamidates
Cyclic esters of L-alanine such as cyclopentyl-(19) and
cyclohexyl-(20) are also very potent in the replicon assay and
show no Huh7 cell toxicity at 100 μM. However, the cyclobutyl
derivative 18 is about 5-fold less active with an EC50 of 0.32
μM, and the tetrahydropyranyl ester 24 is more than 200-fold
less active compared to 20.
Interestingly, the benzyl ester derivatives of L-alanine (Table
1, 21−23) are somewhat less active than many of the alkyl ester
analogues, being 5−10-times less active than the cyclohexyl
compound for example. This is in marked contrast to the
phosphoramidate ProTides SAR and much of our prior
experience.28
An observation from our phosphoramidate SAR is that the
purine C-6-substituent can be varied considerably and that the
6-ethoxy may be particularly effective.24 As noted in Table 1,
the same applies to this series, with the cyclohexyl-(74), benzyl-
(75), and neopentyl-(76) purine C-6-ethoxy derivatives all
being equipotent with their C-6-methoxy analogues. Indeed,
with an EC90 of 110 nM, the neopentyl L-alanine analogue of
the 6-ethoxy nucleoside 76 emerged as one of the most potent
compounds in the present study.
The SAR next turned to symmetrical phosphorodiamidates
with amino acids other than L-alanine. The subtlest change was
to make the D-alanine analogue as its neopentyl ester 29. In
contrast to our earlier work on D-amino acids in aryl
phosphoramidate ProTides,25 here we see only a slight (ca.
2-fold) loss of activity. This again points to quite a new and
separate SAR for these diamidates as compared to aryloxy
phosphoramidates.
In generating the amino acid SAR in this series of
symmetrically substituted phosphorodiamidates, effort was
made to use three similar esters for each different amino acid
derivative to facilitate comparison. The three ester groups
selected were the cyclohexyl, benzyl, and neopentyl. Table 2
shows the compiled data.
A number of observations can be made from the data in
Table 2. First, the benzyl esters do not greatly distinguish the
different amino acids. The replicon activities for nine amino
acids with benzyl ester groups range from 0.25 to 0.61 μM, only
2- to 3-fold, which is similar to the variability in the replicon
assay.
The neopentyl and cyclohexyl ester groups distinguish the
different amino acids and give a similar rank order. Thus, the
best amino acid for both the cyclohexyl and the neopentyl
esters is the L-alanine, whereas the worst is L-isoleucine.
Overall evaluation of the amino acid SAR leads to conclusion
that small amino acid substituents like glycine or alanine are
preferred and that increasing the size of the amino acid R1
group as for L-leucine and L-methionine is detrimental to
replicon activity, as is branching of the amino acid as for L-
valine, L-isoleucine, and L-phenylglycine.
The exception to this rule is that the neopentyl L-
phenylalanine derivative 46 is very potent (EC50 = 0.05 μM)
in the replicon assay. The cyclohexyl L-phenylalanine derivative
44 was tested twice, and one value was 10-fold higher than the
other (EC50 = 0.083 and 0.92). It is possible that further testing
would refine these data. Further supporting this surprising L-
phenylalanine SAR, the L-tyrosine derivative 52, as the para-O-
tert-butyl ether, also had good activity and limited toxicity in the
replicon assay (EC50 = 0.11 μM, CC50 = 89 μM).
A further observation in this series of symmetrical
phosphorodiamidates is that some cytotoxicity is observed in
Huh7 cells for the amino acids other than alanine and glycine.
In fact, the cyclohexyl L-isoleucine 38 demonstrates that not all
phosphorodiamidates are equivalent. Compound 38 has an
EC50 of 4 μM and a CC50 of 14 μM, for an SI of only ca. 4, and
being the most cytotoxic diamidate studied.
As an extension of the growing symmetrical phosphorodia-
midate SAR, a more complicated, dipeptide, L-Val-L-Ala
neopentyl ester 55, in which the L-valine nitrogen forms the
phosphoramidate linkage, was prepared. Testing in the replicon
assay shows that 55 has similar replicon activity (Table 1, EC50
= 0.54 μM) as the neopentyl L-valine derivative 51. It is unclear
how this dipeptide would fit in with the putative metabolic
process (see Scheme 2) and may suggest the existence of
another metabolic pathway leading to the triphosphate
formation for this compound.
Working the SAR in the opposite direction, simple amine
substituted phosphorodiamidates were tested in the replicon
assay. The n-butyl amine analogue 57, which may be
8637
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645

Journal of Medicinal Chemistry
Article
Figure 3. 31P NMR kinetic study of 26 in the presence of carboxypeptidase Y. Conditions: 5.0 mg of 26 in 200 μL of acetone-d6 + 400 μL of Trizma
buffer (pH 7.4), 0.3−0.5 mg of carboxypeptidase Y in 200 μL of Trizma buffer (pH 7.4).
considered similar to the L-leucine derivatives, but without the
α carboxylate, has a replicon EC50 of 38 μM, which is about
100-fold less active than the L-leucine derivatives. This is
consistent with the proposed metabolic route (Scheme 2)
where the carboxylate anion plays a key role. The positioning of
the carboxylate β to the amine, as for the β-alanine analogue 56,
results in moderate activity in the replicon assay (Table 1, EC50
= 3.8 μM). The morpholino phosphorodiamidate 58, an
example of a secondary amine, is inactive even at 100 μM in the
replicon assay. Both of these simple amine phosphorodiami-
dates 57 and 58 are much less active than the parent nucleoside
8 in the replicon assay, supporting our understanding that there
is a limited direct cleavage of the P−O bond, leading to free
nucleoside, in the replicon assay.
Although one of the key motivations for this new ProTide
motif was to generate an achiral phosphate center, we were
interested in understanding some aspects of the asymmetric
diamidate SAR. Our synthetic route (Scheme 1) did allow the
stepwise addition of two different amines. Each of these
asymmetric diamidates were isolated as a roughly 1:1 mixture of
phosphorus diastereoisomers, and no attempt was made to
separate them and they were tested as mixtures. To facilitate
comparison, generally, one of the two amines was maintained as
the L-alanine neopentyl ester (Table 1).
From this basis, our SAR study was focused in three areas.
First, we explored compounds 60−62 in which the second
amine was also L-alanine but with different ester groups. The
replicon results indicated that the presence of the L-alanine
neopentyl ester provides good antiviral potency even in the
presence of difficult to cleave29 tert-butyl L-alanine ester 61.
The presence of two easily cleavable esters such as for
derivatives 59 and 62 slightly increased the replicon activity
comparing to 61 but one “cleavable” group was sufficient.
Next, the SAR of asymmetrical phosphorodiamidates, where
the neopentyl L-alanine is combined with different amino acids
was briefly explored. The methyl L-proline 63 and the methyl L-
valine 64 amino acids had moderate replicon activity (Table 1),
more similar to the symmetrical L-prolines and L-valines than
the very potent symmetrical neopentyl L-alanine 26. It might be
that compounds 63 and 64 have different metabolic
intermediates (78, Scheme 2) than does 26.
Third, the phosphorodiamidate SAR of simple amines
combined with neopentyl L-alanine was extensively studied
(66−73, Table 1). In general, these compounds had moderate
to good activity, with EC50 values ranging from 0.1 to 0.9 μM
for monosubstituted amines, more active than the symmetrical
simple amines, but less active than the symmetrical neopentyl L-
alanine derivative (26). Presumably, conversion of compounds
66−73 involves cleavage of the L-alanine neopentyl ester
followed by the elimination of the simple amine to form the key
metabolic intermediate (78, Scheme 2). Thus, it might be
expected that the amines that make the best leaving groups
such as phenylamine 70 and naphthylamine 71, would be the
most active (EC50 = 0.27 and 0.13 μM) and that a poor leaving
group such as diethyl amine, in derivative 72, would be the least
active (EC50 = 3.7 μM, Table 1).
In conclusion, all but one of the 67 phosphorodiamidates
tested in the replicon assay were active at or below micromolar
concentrations, with 55 of them active below 1 μM and 12
active below 100 nM. Many are noncytotoxic at high μM levels,
some at 100 μM, giving attractive SI values for many
compounds. Clearly several of these compounds were worthy
of advancing to in vivo studies.
STABILITY ASSAYS
■
Given the new structural motif we are reporting and the very
promising replicon data (as described above), we sought to
establish some outline stability data under a variety of
conditions. To begin with, the lead symmetrical neopentylala-
nine compound 26 was dissolved in pH 7 phosphate buffer at
8638
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645

Journal of Medicinal Chemistry
Article
Figure 4. 31P NMR kinetic study of 21 in the presence of carboxypetidase Y. Conditions: 5.1 mg of 21 in 200 μL of acetone-d6 + 400 μL of Trizma
buffer (pH 7.4), 0.3−0.5 mg of carboxypeptidase Y in 200 μL of Trizma buffer (pH 7.4).
25 °C and monitored by 31P NMR over 14 h. Compound 26
was observed at 13.2 ppm, and no additional signal(s) appeared
over the course of the experiment (see Supporting Information
for stability spectra). HPLC also indicated no detectable
decomposition. We have previously reported acid stability
data27 for several acyclovir aryloxy phosphoramidate ProTides.
As a representative example, the phosphorodiamidate (26) was
dissolved in citric acid−HCl buffer at pH 2 and maintained at
either 37 or 47 °C while being monitored by 31P NMR over 14
h. Despite the acidic pH and elevated temperatures, the
samples remained entirely stable over the course of the assay. A
base stability study was conducted on compound 26 at pH 8.5/
37 °C and pH 11/37 °C, and these studies revealed slow
decomposition. After 17 h at pH 11, the majority species was
still unchanged (26), with only minor peaks at 6.9 ppm and
−4.3 ppm. The peak at 6.9 ppm corresponds to an amino acid
phosphoramidate derivative similar to compound 77 (see
Scheme 2 below), suggesting one phosphorodiamidate P−N
bond on compound 26 had been hydrolyzed under the aqueous
basic conditions. However, the decomposition is rather slow,
with a half-life exceeding 100 h (if first order). From this initial
stability data, it seems possible that the phosphorodiamidates
have a pH stability profile that is consistent with oral dosing.
Next, we studied the stability of compound 26 with human
serum at 37 °C. As before, the solution was monitored at 1 h
intervals over by 31P NMR over 12 h. The phosphorodiamidate
26 gave a peak at 14.5 ppm under these conditions, and no sign
of degradation was observed by 31P NMR. Thus, as noted for
aryloxy phosphoramidates,30 the present phosphorodiamidates
seem essentially stable in human serum, certainly for periods of
hours appropriate for human dosing.
enzyme carboxypeptidase Y as an in vitro model for studying
the initial steps of ProTide30 activation. We sought to apply this
method to our new family of anti-HCV phosphorodiamidates.
Thus, compounds of interest were dissolved in acetone-d6 and
TRIZMA buffer at pH 7.4, and the 31P NMR spectrum was
recorded as the baseline. Then carboxypeptidase Y (cathepsin)
was added and spectra recorded at intervals up to 13 h (Figure
3). The data from these experiments were used to map a
possible phosphorodiamidate metabolic pathway (Scheme 2).
The first compound studied was the symmetrical neopentyl
ester of L-alanine 26, which has a 31P NMR shift of 14.2 ppm at
baseline (Figure 3). After 10 min, a small downfield metabolite
peak is observed at 14.9 ppm, which is consistent with cleavage
of the neopentyl ester. The peak at 14.9 ppm builds up for
approximately one hour, then diminishes, with a new peak
emerging at 6.95 ppm. This is first observed at ca. 40 min and
continues to grow through the course of the experiment. The
peak at 6.95 corresponds to the key amino acid phosphate
derivative 77 (R = CH3) (Scheme 2). This was determined by
synthesizing compound 77 (R = CH3) and using it as an
analytical reference standard (data not shown). Intermediate 77
could form by intramolecular attack of the amino acid
carboxylate anion onto the phosphorus, with elimination of
the second amino acid, followed by spontaneous hydrolysis of
the cyclic, mixed anhydride intermediate 78a,b. No direct 31P
NMR evidence (Figure 3) was obtained for this cyclic
compound, consistent with it being a short-lived metabolic
intermediate. As discussed in our publications on aryloxy amino
acid phosphoramidates, final conversion to the nucleoside
monophosphate is generally thought to be effected by enzymes
of the histidine triad nucleotide binding (Hint) family (Scheme
2).30,33
As we have reported previously,12,30 the initial step of the
conversion of aryloxy amino acid ester phosphoramidates
(ProTides) to phosphates is thought to involve an enzyme-
mediated cleavage of the amino acid ester group. We have also
reported on the use of 31P NMR and a buffered solution of the
This initial evidence indicates that nucleoside phosphor-
odiamidates are converted to nucleoside monophosphates in a
similar fashion as aryl phosphoramidates, that is, via metabolite
77. To further support this hypothesis, a second phosphor-
8639
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645

Journal of Medicinal Chemistry
Article
Figure 5. 31P NMR kinetic study of 65 in the presence of carboxypetidase Y. Conditions: 4.9 mg of 65 in 200 μL of acetone-d6 + 400 μL of Trizma
buffer (pH 7.4), 0.3−0.5 mg of carboxypeptidase Y in 200 μL of Trizma buffer (pH 7.4). * Phosphate impurity in Trizma buffer; in blank and
remains unchanged throughout.
Figure 6. 31P NMR kinetic study of 73 in the presence of carboxypetidase Y. Conditions: 5.3 mg of 73 in 200 μL of acetone-d6 + 400 μL of Trizma
buffer (pH 7.4), 0.3−0.5 mg of carboxypeptidase Y in 200 μL of Trizma buffer (pH 7.4).
odiamidate, the symmetrical benzyl L-alanine derivative 21, was
studied in the same system (Figure 4). In this case, a similar 31
79a and 79b are observed, but cleavage of both benzyl esters
does not occur; (c) the peak at 14.46 ppm represents mono
ester cleavage, where the 31P NMR signals of 79a and 79b
overlap, and the peak at 14.78 represents diester cleavage to
give 80.
Regardless of how the amino acid ester groups are cleaved by
this particular enzyme, the important information from the 31P
NMR experiments is that compound 21 goes through the same
common intermediate 77 (R = CH3) as does compound 26.
In Supporting Information, we present further detail of the
kinetics of appearance of each species in the carboxypeptidase Y
mediated cleavage of 21.
P
NMR pattern, with the formation of the same key metabolite at
6.94 ppm (77, R = CH3) is observed, but there are subtle
differences. Parent 21 shows one phosphorus 31P NMR peak at
14.08 ppm at baseline, which disappears within the first 30 min
upon incubation with enzyme. However, in this case, two
downfield singlet peaks are observed, one small and transient at
14.46 ppm and the other larger and longer lived at 14.78 ppm.
Three possibilities exist to explain these two new downfield
peaks: (a) one specific benzyl ester is cleaved by carbox-
ypeptidase Y to give either 79a or 79b, but not both, followed
by a second ester cleavage to give 80; (b) both pro-pR and pro-
pS benzyl esters are cleaved by carboxypeptidase Y and both
It should be noted that 26 is nearly 10-fold more active than
21 in the HCV replicon assay (Table 1), however comparison
8640
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645

Journal of Medicinal Chemistry
Article
Figure 7. (a) Docking of compound 65−Rp isomer within the catalytic site of carboxypeptidase Y. (b) Docking of compound 65−Sp isomer within
the catalytic site of carboxypeptidase Y.
Table 3. Rat Liver Triphosphate Levels from PK Analysis of Phosphorodiamidates
a
b
c
d
compd
AA ester
replicon EC50 (μM)
Cmax (ng/g)
Clast (ng/g)
Tmax (h)
AUC0−t (ng·h/g)
e
5
L-Ala OCH2tBu
L-Ala OCH2tBu
L-Ala O-n-butyl
L-Ala O-n-pentyl
L-Ala O-(R,S)-2-butyl
L-Ala O‑cyclopentyl
L-Ala O-2‑indanol
D-Ala OCH2tBu
L-Val OBn
0.01
0.06
0.07
0.03
0.15
0.06
0.58
0.11
0.12
1446.7
1580.0
1553.3
1042.3
907.7
983.3
800.0
719.7
506.0
538.7
610.0
632.0
441.7
8.0
8.0
8.0
8.0
8.0
8.0
24.0
8.0
4.0
24557
25147
27671
18600
16252
24546
10642
11608
4478
26
13
14
16
19
25
29
50
1523.3
632.0
598.3
359.5
73.6
a
b
c
Cmax: Maximum observed concentration. Clast: Concentration at the last measurable time-point. Tmax: Time at which maximum concentration was
d
e
observed. AUC0−t: area under the concentration−time curve from time 0 to the last measurable concentration. Compound 5 is a the
phosphoramidate INX-189 (see Figure 2).
of Figures 3 and 4 shows that the speed of carboxypeptidase Y
processing is very similar for the two prodrugs, and thus the
replicon and enzyme assays do not correlated directly, but we
do regard enzyme mediated ester cleavage as a useful tool to
study our prodrugs in vitro.
If a carboxylate anion intermediate such as 79a, 79b, or 80
can eliminate an amino acid moeity to give intermediate 77, we
wondered if simple primary or secondary amines could likewise
be eliminated from asymmetric phosphorodiamidates such as
65 and 73. Figure 5 shows the 31P NMR traces for an
experiment with the benzyl L-alanine, n-butyl amine derivative
65, and carboxypeptidase Y.
Because 65 is asymmetrical, two peaks are seen in the
baseline 31P NMR (16.59 and 16.75 ppm). Both peaks
disappear within 120 min, but the diastereomer at 16.75 is
cleaved faster. Once again. a peak at 6.94, identified as
compound (77, R = CH3), is observed growing in magnitude
during the course of the experiment. A very similar pattern is
observed in Figure 6, when the neopentyl L-alanine, pyrrolidine
asymmetrical phosphoramidate 73 was incubated with
carboxypeptidase Y. Again compound 77 (R = CH3) is
observed at 6.94 ppm and grows in over 14 h. Interestingly, in
this case, only one of the two diastereomers of compound 73
(31P NMR δ = 15.54 ppm) is cleaved by carboxypeptidase Y
over the course of the experiment.
asymmetrical phosphoramidates containing an amino acid
carboxylate by showing that a second amino acid or a primary
amine or a secondary amine can also be eliminated to give the
key intermediate 77. The data on 21, 26, 65, and 73 supports
the notion that only one ester cleavage is necessary and that the
second amine loss is rapid following the first ester cleavage.
Docking Studies. To further support the above, we
conducted some docking studies on several asymmetric
diamidates, using published31 crystal structure of carboxypepti-
dase Y. Thus, as shown in Figure 7a,b the Sp diastereomer of
65 binds significantly better than the Rp diastereomer.
In the case of the Sp isomer, the stabilization by H-binding of
two glycine residues (Gly52, 53) is notable and the nucleophilic
active site of Ser146 is also well positioned. These docking data
would then suggest that one diastereoisomer of 65 might be
processed better than the other. On the basis of the clear
kinetic difference noted above (Figure 5), perhaps the more
downfield species in 31P NMR of 65 is the Sp isomer. It is
interesting to wonder if the differing kinetics of metabolism
here may lead to a difference in biological potency as we have
noted in some cases for aryl phosphoramidates.5 However, we
were unable to separate the compound 65 diastereomers to test
this hypothesis.
Pharmacokinetics. The HCV replicon, stability, and
carboxypeptidase Y data suggests that our new phosphor-
odiamidate prodrug strategy may be a promising means of
delivering 2′-C-methylguanosine triphosphate into cells. The
Summarizing this portion or our work, we have built on our
previous understanding that aryloxy groups are eliminated from
8641
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645

Journal of Medicinal Chemistry
Article
next hurdle was to determine if this prodrug strategy works in
vivo in a rodent. Because of our growing experience with
phosphoramidates such as 5 in rats, we decided to continue
with the rat as our initial PK model for the phosphorodiami-
dates. Eight symmetrical derivatives were selected for rat PK
studies (13, 14, 16, 19, 25, 26, 29, and 50, Table 3) based on
activity in the replicon assay and structural considerations. We
focused on symmetrical diamidates because they are
represented by single diastereoisomers, which would simplify
further development. All but two were L-alanine derivatives
because we wanted to fully explore the ability of these relatively
simple derivatives to provide sufficient liver triphosphate levels.
Compounds were formulated in 95% Capmul MCM/5%
Tween 80, and doses of 10 mg/kg were administered by oral
gavage to male Sprague−Dawley rats. Liver samples were
collected up to 24 h postadministration and were snap-frozen in
liquid nitrogen. Liver concentrations of 2′-C-methylguanosine
triphosphate were determined by LC-MS/MS. Results for these
eight phosphorodiamidates are compared to 5 in Table 3.
The main observation is that all compounds tested produced
significant levels of triphosphate in rat livers from a 10 mg/kg
dose, further validating phosphorodiamidates as effective
phosphate prodrugs. In addition, several of the phosphor-
odiamidates provide similar liver triphosphate exposures as the
clinical compound 5, which has shown efficacy against HCV in
phase Ib clinical trials.23 The amount of 2′-C-methylguanosine
triphosphate necessary to achieve an EC90 in the replicon assay
can be determined by measuring triphosphate levels in the
replicon cells upon incubation with a 2′-C-methylguanosine
based inhibitor such as 5. This EC90 triphosphate level in cells
can be extrapolated to EC90 triphosphate levels in the liver, as
measured in ng of triphosphate/gram of liver tissue. The value
we have calculated is 243 pmol of triphosphate per gram of liver
(equal to 131 ng/g).32 Thus, for all phosphorodiamidates
tested, except for 50, the level of triphosphate 24 h post dose, is
several fold above 131 ng/g level necessary to achieve 90%
inhibition of HCV replication (Table 3).
An important part of our evaluation of any new prodrug
approach for delivering 2′-C-methylguanosine triphosphate,
including these phosphorodiamidates, is measurement of
systemic nucleoside (2′-C-methylguanosine, 4) levels after
oral dosing of the prodrug. Our desire is to limit the systemic
exposure of this nucleoside. The plasma 2′-C-methylguanosine
levels for these eight phosphorodiamidates were measured, and
all but compound 13 had lower nucleoside Cmax values than 5
(<100 nM), and it is only slightly higher (data not shown).
The neopentyl L-alanine diamidate 26 has both the highest
Cmax and Clast of any diamidate tested. Figure 8 shows rat liver
triphosphate levels after a 10 mg/kg dose for the clinical
compound 5 and for 26. Triphosphate levels were measured at
eight time points over 24 h. It is clear that both prodrug
strategies produce similarly high levels of triphosphate. The n-
butyl L-alanine ester 13 has the highest overall triphosphate
AUC (Table 3), and the cyclopentyl ester 19 also has excellent
AUC and Clast values. These compounds along with the n-
pentyl ester 14 were advanced to monkey PK studies. A
combination of rat PK, monkey PK, and preliminary rodent
toxicology studies will be used to help select a clinical
candidate. These additional studies will be reported elsewhere
upon completion.
Figure 8. Rat liver 2′-C′-methyl guanosine triphosphate levels from
compounds 5 and 26.
CONCLUSIONS
■
In conclusion, we report on a new family of phosphate
prodrugs based on phosphorodiamidates. This type of 5′-
monophosphate prodrug has the advantage that it can be
designed to be achiral at the phosphate center if desired. We
report on a range of novel prodrugs derived from amino acids
and simple amines and build a substantial HCV replicon SAR
for both symmetrical and asymmetrical phosphorodiamidates of
2′-C-methyl-6-O-methylguanosine. The replicon data suggests
that one aminoacyl ester is essential for potent activity versus
HCV. The phosphorodiamidates are stable in acid and mild
base and also in human serum. Carboxypeptidase Y is able to
activate these compounds to a nucleoside aminoacid phosphate
key intermediate, which is also the essential metabolic
intermediate for our earlier aryloxy ProTides. Many of the
novel compounds in this study show low nanomolar activity
versus HCV in replicon coupled with low cytotoxicity in the
Huh7 replicon cell line. Eight potent HCV inhibitors were
advanced to PK studies in Sprague−Dawley rats, and it was
demonstrated that they all provided substantial 2′-C-methyl-
guanosine triphosphate levels, in rat livers, that were
maintained over a period of 24 h. This body of work has
validated phosphorodiamidates as prodrugs for 2′-C-methyl-
guanosine both at the in vitro and the in vivo levels. Further in
vivo studies are underway that are intended to lead toward
selection of a phosphorodiamidate prodrug for HCV clinical
studies.
EXPERIMENTAL SECTION
■
General. Anhydrous solvents were purchased from Aldrich and
used without further purification. All reactions were carried out under
an argon atmosphere. Reactions were monitored with analytical TLC
on silica gel 60-F254 precoated aluminum plates and visualized under
UV (254 nm) and/or with 31P NMR spectra. Column chromatography
was performed on silica gel (35−70 μM). Proton (1H), carbon (13C),
and phosphorus (31P) NMR spectra were recorded on a Bruker
Avance 500 spectrometer at 25 °C. Spectra were autocalibrated to the
deuterated solvent peak, and all 13C NMR and 31P NMR were proton-
decoupled. Analytical and semipreparative HPLC were conducted by
Varian Prostar (LC Workstation-Varian prostar 335 LC detector)
using Varian Polaris C18-A (10 μM) as an analytic column and Varian
Polaris C18-A (10 μM) as a semipreparative column; elution was
performed using a mobile phase consisting of water/acetonitrile in
gradient (system1, 90/10 to 0/100 v/v in 30 min) or water/methanol
(system 2, 90/10 to 0/100 v/v in 30 min). High-resolution mass
spectra (HRMS) was performed as a service by Cardiff University,
using electrospray (ES). Compound purity was assured by a
8642
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645

Journal of Medicinal Chemistry
Article
combination of high field multinuclear NMR (H, C, P) and HPLC.
Purity by the latter was always >95% with no detectable parent
nucleoside for all final products.
4.21−4.16 (m, 1H, H4′), 4.07 (s, 3H, OCH3), 3.98−3.89 (m, 2H, 2×
CHα Ala), 3.70, 3.69 (2s, 6H, 2× OCH3 ester), 1.33 and 1.32 (2d, 6H,
J = 7.1 Hz, CH3 Ala), 1.00 (s, 3H, CH3).
13C NMR (126 MHz, MeOD-d4) δ 176.10 (2d, 3JC−C−N−P = 4.8 Hz,
2× CO Ala), 162.75 (C6), 161.95 (C2), 154.21 (C4), 139.33 (C8),
Standard Procedure A: Synthesis Of Symmetrical Diami-
dates. To a suspension of the nucleoside (1.0 mol equiv) in
anhydrous tetrahydrofuran, triethylamine (1.0−1.2 mol equiv) was
added. After stirring for 30 min at room temperature, phosphoryl
chloride (1.0−1.2 mol equiv) was added dropwise at −78 °C. The
reaction mixture was stirred for 30 min at −78 °C and then allowed to
warm to room temperature. Anhydrous dichloromethane was added,
followed by amino acid ester (5.0 mol equiv) and triethylamine (10.0
mol equiv) at −78 °C. After stirring at room temperature for 20 h,
water was added and the layers are separated. The aqueous phase was
extracted with dichloromethane and the organic phase washed with
brine. The combined organic layers were dried over anhydrous sodium
sulfate, filtered, and evaporated to dryness. The resulting residue was
purified by silica gel column chromatography using as an eluent a
gradient of methanol in dichloromethane or chloroform. In some
cases, a subsequent repurification was necessary either by preparative
HPLC (gradient of methanol in water) or preparative TLC.
Standard Procedure B: Synthesis of Symmetrical Diami-
dates. To a solution of the nucleoside (1.0 mol equiv) in anhydrous
triethylphosphate was added phosphoryl chloride (2.0 mol equiv) at 0
°C. The reaction mixture was stirred for 24 h at 5 °C. Anhydrous
dichloromethane was added to the reaction mixture followed by amino
acid ester (5.0 mol equiv) and diisopropylethylamine (10.0 mol equiv)
at 0 °C. After stirring at 5 °C for 5 days, water was added and the
layers were separated. The aqueous phase was extracted with
dichloromethane and the organic phase washed with brine. The
combined organic layers were dried over anhydrous sodium sulfate,
filtered, and evaporated to dryness. The resulting residue was purified
by silica gel column chromatography using as an eluent a gradient of
methanol in dichloromethane. A subsequent repurification, if
necessary, was accomplished either by preparative HPLC (gradient
of methanol in water) or preparative TLC.
Standard Procedure C: Synthesis of Asymmetrical Diami-
dates. To a suspension of the nucleoside (1.0 mol equiv) in
anhydrous tetrahydrofuran, triethylamine (1.0 mol equiv) was added.
After stirring for 30 min at room temperature, phosphoryl chloride
(1.0 mol equiv) was added dropwise at −78 °C. The reaction mixture
was stirred for 30 min at −78 °C and then allowed to warm to room
temperature. Anhydrous dichloromethane was added, followed by the
addition of amino acid ester or amine (1 mol equiv) and anhydrous
triethylamine (2 or 1 mol equiv, respectively) at −78 °C. Reaction was
warmed to room temperature and monitored by 31P NMR. When
NMR indicated completion of the reaction (no starting material,
presence of monosubstituted product), a second amino acid ester or
amine (5 mol equiv) was added followed by the addition of
triethylamine (10 or 5 mol equiv, respectively) at −78 °C. After
stirring at room temperature for 16−20 h, water was added and the
layers were separated. The aqueous phase was extracted with
dichloromethane. The combined organic layers were dried over
anhydrous sodium sulfate, filtered, and evaporated to dryness. The
resulting residue is purified by silica gel column chromatography using
as an eluent a gradient of methanol in chloroform.
3
115.54 (C5), 93.19 (C1′), 82.33 (d, JC−C−O−P = 7.6 Hz, C4′), 80.02
2
(C2′), 74.73 (C3′), 66.13 (d, JC−O−P = 4.8 Hz, C-5′), 54.21 (OCH3),
52.72 (2× CH3 ester), 51.01, 50.93 (2d, 2JC−N−P = 2.3 Hz, 2× CH Ala),
3
20.86, 20.67 (2d, JC−C−N−P = 6.2 Hz, 2× CH3 Ala), 20.27 (2′-CH3).
31P NMR (202 MHz, MeOD-d4) δ 14.00. HPLC tR = 8.86 min
(system 1).
Example: Synthesis of 2-Amino-6-methoxy-9-(2′-C-methyl-
β-D-ribofuranosyl) Purine 5′-O-Bis(benzoxy-L-alaninyl) Phos-
phate (21). The phosphorodiamidate 21 was prepared according to
the standard procedure B.
In the first step, a solution of 6-O-methyl-2′-C-methylguanosine
(250 mg, 0.803 mmol) in anhydrous triethylphosphate (1 mL) was
reacted with phosphorus oxychloride (148 μL, 1.61 mmol). In the
second step, anhydrous dichloromethane (4 mL), the tosylate salt of
benzoxy-L-alanine (1.41 g, 4.02 mmol), and diisopropylethylamine
(1.40 mL, 8.03 mmol) were added to the previous mixture. After
workup, silica gel column chromatography and preparative HPLC,
1
50.1 mg of 21 was obtained in 8.7% yield as an off white solid. H
NMR (500 MHz, MeOD-d4) δ 7.96 (s, 1H, H-8), 7.34−7.25 (m, 10H,
2× Ph), 5.99 (s, 1H, H1′), 5.16−5.02 (m, 4H, 2× CH2 ester), 4.41−
4.31 (m, 2H, H-5′), 4.29 (d, 1H, J = 9.0 Hz, H3′), 4.21−4.15 (m, 1H,
H4′), 4.04 (s, 3H, OCH3), 4.02−3.94 (m, 2H, 2× CH Ala), 1.33 (d,
6H, J = 7.1 Hz, 2× CH3 Ala), 0.99 (s, 3H, CH3). 13C NMR (126 MHz,
3
MeOD-d4) δ 175.42, 175.36 (2d, 2× CO, JC−C−N−P = 6.3 Hz,
ester), 162.72 (C6), 161.91 (C2), 154.56 (C4), 139.31 (C8), 137.32,
137.29 (d, 2× C ipso OCH2Ph), 129.55, 129.5, 129.24, 129.21
(OCH2Ph), 115.55 (C5), 93.17 (C1′), 82.37 (C4′), 80.01 (C2′), 74.81
(C3′), 67.89,67.87 (2× OCH2Ph), 66.26 (C5′), 54.19 (OCH3), 51.13,
51.08 (2d, 2× Cα Ala), 20.79−20.58 (2d, 3JC−C−N−P = 6.3 Hz, 2× CH3
Ala), 20.26 (2′CCH3). 31P NMR (202 MHz, MeOD-d4) δ 13.93.
HPLC tR = 13.16 min (system 1).
Example: Synthesis of 2-Amino-6-methoxy-9-(2′-C-methyl-
β-D-ribofuranosyl) Purine 5′-O-[(Benzoxy-L-alaninyl)-(2,2-dime-
thylpropoxy-L-alaninyl)] Phosphate (60). The phosphorodiami-
date 60 was prepared according to the standard procedure C.
In the first step, a solution of 6-O-methyl-2′-C-methylguanosine
(250 mg, 0.803 mmol) in anhydrous tetrahydrofuran (5 mL) was
allowed to react with triethylamine (110 μL, 0.803 mmol) and
phosphorus oxychloride (70 μL, 0.803 mmol). The tosylate salt of
benzoxy-L-alanine (282 mg, 0.803 mmol) and triethylamine (110 μL,
0.803 mmol) were added. Anhydrous dichloromethane (4 mL) and
the tosylate salt of neopentyloxy-L-alanine (1.33 g, 4.02 mmol) and
triethylamine (1.12 mL, 8.03 mmol) were added as described in
method C. After workup and silica gel column chromatography, 25 mg
of the prodrug was obtained in 4% yield as an off-white solid. 1H NMR
(500 MHz, MeOH-d4) 7.97, 7.96 (2s, 1H, H8), 7.36−7.30 (m, 5H,
OCH2Ph), 5.98, 5.97 (2s, 1H, H1′), 5.18- 5.09 (m, 2H, OCH2Ph),
4.39−4.33 (m, 2H, H5′), 4.28 (2d, J= 8.00 Hz, 1H, H3′), 4.20−4.16
(m, 1H, H4′), 4.06, 4.05 (2s, 3H, 6OCH3), 4.02−3.94 (m, 2H, 2×
CHα Ala), 3.84, 3.82, 3.72, 3.67 (2AB, JAB= 10.50 Hz, 2H,
CH2C(CH3)3), 1.39−1.32 (m, 6H, 2× CH3 Ala), 0.97 (s, 3H,
2′CCH3), 0.93, 0.91 (2s, 9H, CH2C(CH3)3). 13C NMR (126 MHz,
MeOH-d4) 175.54, 175.43, 175.39 (CO ester), 162.73, 162.71
(C6), 161.93, 161.89 (C2), 154.57, 154.55 (C4), 139.32, 139.08 (C8),
137.39 (ipso OCH2Ph), 129.55, 129.35, 129.25, 129.23, 129.20,
129.16, 128.27, 128.00 (OCH2Ph), 116.19, 115.54 (C5), 93.34, 93.18
(C1′), 82.39, 82.33 (C4′), 80.01, 79.99 (C2′), 75.34, 75.04 (CH2C-
(CH3)3), 74.84, 74.82 (C3′), 67.88, 67.85 (OCH2Ph), 67.86 (d,
Example: Synthesis of 2-Amino-6-methoxy-9-(2′-C-methyl-
β-D-ribofuranosyl) Purine 5′-O-Bis(methoxy-L-alaninyl) Phos-
phate (10). The phosphorodiamidate 10 was prepared according to
the standard procedure B.
In a first step, a suspension of 6-O-methyl-2′-C-methylguanosine
(250 mg, 0.803 mmol) in anhydrous tetrahydrofuran (4 mL) was
reacted with triethylamine (135 μL, 0.964 mmol) and phosphorus
oxychloride (89 μL, 0.964 mmol). In a second step, anhydrous
dichloromethane (4 mL), L-alanine methyl ester hydrochloride salt
(561.1 mg, 4.02 mmol), and triethylamine (1.12 mL, 8.03 mmol) were
added.
2
2JC−O−P = 3.75 Hz, C5′), 66.36 (d, JC−O−P = 5.50 Hz, C5′), 54.18,
54.01 (6OCH3), 49.69, 49.64, 49.52, 49.46 (2× Cα Ala), 32.28, 32.25
(CH2C(CH3)3), 26.74, 26.71 (CH2C(CH3)3), 21.07, 20.90, 20.79,
20.66 (4d, 3JC−C−N−P = 6.25 Hz, 2× CH3 Ala), 20.39, 20.25 (2′CCH3).
31P NMR (202 MHz, MeOH-d4) 13.98, 13.94. HPLC tR = 16.11, 16.80
min (system 1). MS (TOF ES+) m/z: 716.28 (M + Na+, 100%).
HRMS C30H44N7O10P1 calculated, 694.2966; found, 694.2956
After workup, silica gel column chromatography and preparative
HPLC, 19.8 mg of (10) was obtained in 4.4% yield as an off-white
1
solid. H NMR (500 MHz, MeOD-d4) δ 7.99 (s, 1H, H-8), 5.99 (s,
1H, H1′), 4.44−4.34 (m, 2H, H5′), 4.31 (d, 1H, J = 8.9 Hz, H3′),
8643
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645

Journal of Medicinal Chemistry
Article
Biological Methods. Replicon Assays. The HCV inhibitory
activity of compounds was evaluated in an Huh7 cell line expressing a
stable, bicistronic subgenomic HCV genotype 1b (Con1) replicon
encoding the Renilla luciferase reporter gene (Apath, LLC, Brooklyn,
NY) as previously described.32 Cellular cytotoxicity was evaluated
using the CellTiter-Glo Luciferase assay (Promega, Madison WI). A
day before testing, 2 × 104 Huh7 cells were seeded in 96-well flat
bottom white plates (Nunc, Roskilde, Denmark). Four-fold serial drug
dilutions were made in growth medium and added to the cells. No
drug controls were included in each plate. The plates were incubated
in the presence of test compound for 3 days at 37 °C with 5% CO2.
Luciferase reagent was added to cells and plates were incubated for 20
min before measuring relative luminescent units (RLU) in a
luminometer (Veritas, Turner Biosystems, Sunnyvale, CA).
Pharmacokinetic Studies in Rats. Rat studies were conducted at
Inhibitex, Inc., in accordance with NIH Guidelines and following
protocols approved by the Institutional Animal Care and Use
Committee (IACUC) of Inhibitex, Inc. Studies were carried out as
previously described.32 Test compounds were formulated in 95%
Capmul MCM (ABITEC Corp., Janesville, WI)/5% Tween 80 (Sigma,
St. Louis, MO), and doses of 10 mg/kg were administered by oral
gavage to male Sprague−Dawley rats (Taconic Farms, Germantown,
NY). Liver samples were collected as a terminal procedure up to 24 h
postadministration and were snap-frozen immediately upon collection
in liquid nitrogen. Liver samples were stored frozen at ≤−80 °C prior
to analysis.
retrovirus specificity and intracellular metabolism of 2′,3′-didehydro-
2′,3′-dideoxythymidine (d4T, stavudine) and its d4T 5′-
monophosphate triester prodrug. Mol. Pharmacol. 1996, 50, 1207−
1213.
(4) Hecker, S. J.; Erion, M. D. Prodrugs of Phosphates and
Phosphonates. J. Med. Chem. 2008, 51, 2328−2345.
(5) McGuigan, C.; Mehellou, Y.; Balzarini, J. Aryloxy
Phosphoramidate Triesters: A Technology for Delivering
Monophosphorylated Nucleosides and Sugars into Cells.
ChemMedChem 2009, 4, 1779−1791.
(6) Drontle, D. P.; Wagner, C. R. Designing a pronucleotide
stratagenic lessons from amino acid phosphoramidates of anticancer
and antiviral pyrimidines. Mini-Rev. Med Chem. 2004, 4, 409−419.
(7) McKenna, C. E.; Kashemirov, B. A.; Eriksson, U.; Amidon, G. L.;
Kish, P. E.; Mitchell, S.; Kim, J.-S.; Hilfinger, J. M. Cidofovir peptide
conjugates as prodrugs. J. Organomet. Chem. 2005, 2673−2678.
(8) Hosteller, K. Y.; Stuhmiller, L. M.; Lenting, H. B. M.; Van den
Bosch, H.; Richman, D. D. Synthesis and antiretroviral activity of
phospholipid analogs of azidothymidine and other antiviral
nucleosides. J. Biol. Chem. 1990, 265, 6112−6117.
(9) Perigaud, C.; Aubertin, A.-M.; Benzaria, S.; Pelicano, H.; Giradet,
J.-L.; Maury, G.; Gosselin, G.; Kirn, A.; Imbach, I.-L. Equal inhibition
of the replication of human immunodeficiency virus in human T-cell
culture by ddA bis(SATE) phosphotriester and 3′-azido-2′,3′-
dideoxythymidine. Biochem. Pharmacol. 1994, 48, 11−14.
(10) Erion, M. D.; Reddy, K. R.; Boyer, S. H.; Matelich, M. C.;
Gomez-Galeno, J.; Lemus, R. H.; Ugarkar, B. G.; Colby, T. J.;
Schanzer, J.; Van Poelje, P. D. Design synthesis and characterization of
a series of cytochrome P(450) 3A-activated prodrugs (HepDirect
prodrugs) useful for targeting phosph(on)ate-based drugs to the liver.
J. Am. Chem. Soc. 2004, 126, 5154−5163.
(11) Meier, C. 2-Nucleos-5′-O-yl-4H-1,2,3-benzodioxaphosphinin-2-
oxides a new concept for lipophilic, potential prodrugs of biologically
active nucleoside monophosphates. Angew. Chem., Int., Ed. Engl. 1996,
35, 70−72.
(12) 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. Design, synthesis and evaluation of a novel
double pro-drug: INX-08189. A new clinical candidate for hepatitis C
virus. Bioorg. Med. Chem. Lett. 2010, 20, 4850−4854.
Bioanalysis of Pharmacokinetic Samples. The concentration of
2′-C-MeGTP in liver samples from rats was measured by LC-MS/MS
as described previously.28 The assay was linear (r2 ≥ 0.99) in the
concentration range of 100−4000 ng per gram of tissue with ≥85%
accuracy and ≤2% CV. Noncompartmental pharmacokinetic analyses
were performed on the liver concentration data using WinNonlin v5.2
software (Pharsight, St. Louis, MO) as described previously.32
ASSOCIATED CONTENT
* Supporting Information
■
S
Preparative methods, spectroscopic and analytical data on
target compounds plus 31P NMR stability assays and metabolic
activation data. This material is available free of charge via the
Internet at http://pubs.acs.org.
(13) Furman, P. A.; Wang, P. Y.; Niu, C. R.; Bao, D. H. Symonds, W.;
Nagarathnam, D.; Steuer, H. M.; Rachakonda, S.; Ross, B. S.; Otto, M.
J.; Sofia, M. PSI-7851: A novel liver-targeting nucleotide prodrug for
the treatment of Hepatitis C. Hepatology 2009, 48 (Suppl), 59th
Annual Meeting of the American Association for the Study of Liver
Diseases, 2008, San Francisco, CA, poster 1901.
AUTHOR INFORMATION
Corresponding Author
*Phone/Fax: +44 29 20874537. E-mail: mcguigan@cardiff.ac.
■
uk.
(14) Zhou, X. J.; Pietropaolo, K.; Chen, J.; Khan, S.; Sullivan-Bolyai.,
J.; Mayers, D. Safety and pharmacokinetics of IDX184, a liver-targeted
nucleotide polymerase inhibitor of hepatitis C virus, in healthy
subjects. Antimicrob. Agents Chemother. 2011, 1, 76−81.
(15) Pengram, M.; Ku, N.; Shepard, M.; Speid, L.; Lenz, H.-J.
Enzyme catalyzed therapeutic activation (ECTA) NB1011
(thymectacin) selectively targets thymidylate synthase (TS)-
overexpressing tumor cells: Preclinical and phase 1 clinical results.
Eur. J. Cancer 2002, 38 (suppl, 7), 99.
(16) Congiatu, C.; Brancale, A.; Mason, M. D.; Jiang, W. G.;
McGuigan, C. Novel Potential Anticancer Naphthyl
Phosphoramidates of BVdU: Separation of Diastereoisomers and
Assignment of the Absolute Configuration of the Phosphorus Center.
J. Med. Chem. 2006, 49, 452−455.
(17) Jones, B. C. N. M.; McGuigan, C.; O’Connor, T. J.; Jeffries, D.
J.; Kinchington, D. Synthesis and anti-HIV activity of some novel
phosphorodiamidate derivatives of 3′-azido-3′-deoxythymidine (AZT).
Antivir. Chem. Chemother. 1991, 2, 35−39.
ACKNOWLEDGMENTS
We thank Helen Murphy for secretarial assistance.
■
ABBREVIATIONS USED
■
DAAs, direct acting antivirals; HCV, hepatitis C virus; AZT,
3′azidothymidine; SAR, structure−activity relationships; TLC,
thin layer chromatography; HPLC, high performance/liquid
chromatography; ClogP, calculated logarithm of the octanol/
water partition coefficient; PK, pharmacokinetics
REFERENCES
■
(1) Soriano, V.; Madejon, A.; Vispo, E.; Labarga, P.; Garcia-
Samaniego, J.; Martin-Carbonero, L.; Sheldon, J.; Bottecchia, M.;
Tuma, P.; Barreiro, P. Emerging drugs for hepatitis C. Expert Opin.
Emerging Drugs 2008, 13, 1−19.
(2) Brown, N. A. Progress towards improving antiviral therapy for
hepatitis C with hepatitis C virus polymerase inhibitors. Part 1:
Nucleoside analogues. Expert Opin. Invest. Drugs 2009, 18, 709−725.
(3) Balzarini, J.; Egberink, H.; Hartmann, K.; Cahard, D.;
Vahlenkamp, T.; Thormar, H.; De Clercq, E.; McGuigan, C. Anti-
(18) Serafinowska, H. T.; Ashton, R. J.; Bailey, S.; Harnden, M. R.;
Jackson, S. M.; Sutton, D. Synthesis and in vivo evaluation of prodrugs
of 9-[2-(phosphonomethoxy)ethoxy]adenine. J. Med. Chem. 1995, 38,
1372−1379.
8644
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645

Journal of Medicinal Chemistry
Article
(19) Erion, M. D.; Van Poelje, P. D.; Dang, Q.; Kasibhatla, S. R.;
Potter, S. C.; Reddy, M. R.; Reddy, K. R.; Jiang, T.; Lipscomb, W. N.
MB06322 (CS-917): a potent and selective inhibitor of fructose 1,6-
bisphosphatase for controlling gluconeogenesis in type 2 diabetes.
Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 7970−7975; J. Hepatol. 2011,
54 (1), S187.
(33) Chou, T.-F.; Baraniak, J.; Kaczmarek, R.; Zhou, X.; Cheng, J.;
Ghosh, B.; Wagner, C. R. Phosphoramidate Pronucleotides: A
Comparison of the Phosphoramidase Substrate Specificity of Human
and Escherichia coli Histidine Triad Nucleotide Binding Proteins. Mol.
Phamaceutics 2007, 4, 208−217.
(20) Nillroth, U.; Besidsky, Y.; Classon, B.; Chattopadhyaya, J.; Ugi,
I.; Danielson, U. H. Specific interaction between HIV-1 proteinase and
5′-phosphate peptidomimetic derivatives of nucleoside analogs. Drug
Des. Discovery 1995, 13 (1), 43−54.
(21) Nillroth, U.; Vrang, L.; Ahlsen, G.; Besidsky, Y.;
Chattopadhyaya, J.; Ugi, I.; Danielson, U. H. The use of 5′-
phosphate derivatives of nucleoside analogs as inhibitors of HIV-1
replication. Antiviral Chem. Chemother. 1995, 6 (1), 50−64.
(22) Romanowska, K. Biological activity and topological properties of
HIV-1 inhibitors. Pol. J. Chem. 1993, 67 (4), 755−757.
(23) Barry, A.; Patti, J.; Matson, M.; Boehlecke, B.; Wenzel, E.;
Pentikis, H.; Alam, J.; Henson, G. A Study of the Safety and
Pharmacokinetics of Single Ascending Oral Doses of INX-08189
Nucleotide Polymerase Inhibitor, in Healthy Subjects. 46th European
Association for the Study of the Liver, Berlin, Germany, March 30−
April 3, 2011.
(24) McGuigan, C.; Madela, K.; Aljarah, M.; Gilles, A.; Battina, S. K.;
Ramamurty, C. V. S.; Srinivas Rao, C.; Vernachio, J.; Hutchins, J.; Hal,
A.; Kolykhalov, A.; Henson, G.; Chamberlain, S. Dual prodrugs of 2′-
C-methyl guanosine monophosphate as potent and selective inhibitors
of hepatitis C virus. Biomed. Chem. Lett. 2011, 21, 6007−6012.
(25) McGuigan, C.; Salgado, A.; Yarnold, C.; Harries, T. Y.; De
Clercq, E.; Balzarini, J. Novel nucleoside phosphoramidates as
inhibitors of HIV: studies on the stereochemical requirements of the
phosphoramidate amino acid. Antivir. Chem. Chemother. 1996, 7, 184−
188.
(26) McGuigan, C.; Tsang, H.-W.; Cahard, D.; Turner, K.;
Velazquez, S.; Salgado, A.; Bidois, L.; Naesens, L.; De Clercq, E.;
Balzarini, J. Phosphoramidate derivatives of d4T as inhibitors of HIV:
the effect of amino acid variation. Antiviral Res. 1997, 35, 195−204.
(27) McGuigan, C.; Tsang, H.-W.; Sutton, P. W.; De Clercq, E.;
Balzarini, J. Synthesis and anti-HIV activity of some novel chain
extended phosphoramidate derivatives d4T (stavudine) esterase
hydrolysis as a rapid predictive test for antiviral potency. Antivir.
Chem. Chemother. 1998, 9, 109−115.
(28) 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, Walters, N.; Wang, J.; Chamberlain, S.; Henson,
G. Phosphoramidate ProTides of 2′-C-Methylguanosine as Highly
Potent Inhibitors of Hepatitis C virus. Study of Their in Vitro and in
Vivo Properties. J. Med. Chem. 2010, 53, 4949−4957.
(29) McGuigan, C.; Sutton, P. W.; Cahard, D.; Turner, K.; O’Leary,
G.; Wang, Y.; Gumbleton, M.; De Clercq, E.; Balzarini, J. Synthesis,
anti-human immunodeficiency virus activity and esterase lability of
some novel carboxylic ester-modified phosphoramidate derivatives of
stavudine (d4T). Antivir. Chem. Chemother 1998, 9, 473−479.
(30) Derudas, M.; Carta, D.; Brancale, A.; Vanpouille, C.; Lisco, A.;
Margolis, L.; Balzarini, J.; McGuigan, C. The Application of
Phosphoramidate Protide Technology to Acyclovir Confers Anti-
HIV Inhibition. J. Med. Chem. 2009, 52, 5520−5530.
(31) Endrizzi, J. A.; Breddam, K.; Remington, S. J. 2.8-A structure of
yeast serine carboxypeptidase. Biochemistry 1994, 33, 11106−11120.
(32) Vernachio, J. H.; Bleiman, B.; Bryant, K. D.; Chamberlain, S.;
Hunley, D.; Hutchins, J.; Ames, B.; Gorovits, E.; Ganguly, B.; Hall, A.;
Kolykhalov, A.; Liu, Y.; Muhammad, J.; Raja, N.; Walters, R. C.; Wang,
J.; Williams, K.; Patti, J. M.; Henson, G.; Madela, K.; Aljarah, M.;
Gilles, A.; McGuigan, C. INX-08189, a Phosphoramidate Prodrug of 6-
O-Methyl-2′-C-Methyl Guanosine, is a Potent Inhibitor of Hepatitis C
Virus Replication with Excellent Pharmacokinetic and
Pharmacodynamic Properties. Antimicrob. Agents Chemother. 2011,
55, 1843−1851.
8645
dx.doi.org/10.1021/jm2011673 | J. Med. Chem. 2011, 54, 8632−8645