E. Gunic et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2456–2458
Table 1. Anti-HCV activity of compounds 5a–42b
2457
phosphoramidates 25–33. These amidates were con-
densed with 5a and 5b to afford prodrugs 34a,b–42a,b
in decent yields Scheme 3. Each compound was a diaste-
reoisomeric mixture that we did not attempt to separate.
These phosphoramidate prodrugs were tested for their
anti-HCV activity, and most of them showed submi-
cromolar activity without any cytotoxicity. Several
trends were noticeable: the first one was that, as previ-
ously noticed with the SATE prodrugs, the 2-amino
nucleosides were more active than their 2-hydrogen ana-
logues. This result was interesting because our nucleo-
sides are bearing a heterocyclic base that is a hybrid
between a guanine and an adenine. The other observa-
tion was that the prodrugs with a methyl group at the
R2 position were generally a lot more potent than those
with a hydrogen at the same position, as seen with the
pairs 34a–36a, 34b–36b and 35b–37b. Adding a second
methyl at R3 neither improved nor diminished the activ-
ity further, as shown with compounds 35a–41a and 35b–
41b. The most active compounds (39b and 42b) were
4- to 6-fold less potent than their SATE prodrug coun-
terparts (7b), but at this point, we were more interested
in pharmacokinetic properties than optimized activities.
It is interesting to note that the cyclic monophosphate
SATE prodrugs of the same nucleosides exhibited very
similar antiviral activities.5
Compound
R1
R2
R3
R4
R5
EC50, lMa
5a
5b
H
na
na
na
na
na
na
H
na
na
na
na
na
na
H
300
92
NH2
H
7a
7b
na
na
na
na
0.060
0.024
26
NH2
H
34a
34b
35a
35b
36a
36b
37a
37b
38a
38b
39a
39b
40a
40b
41a
41b
42b
Me
Me
Me
Me
H
Me
Me
Me
Me
Me
Me
Me
Me
Bn
NH2
H
H
H
1.2
2.6
H
Cl
Cl
H
NH2
H
H
0.80
220
22
H
NH2
H
H
H
H
H
H
Cl
Cl
H
39
NH2
H
H
H
8.9
2.9
Me
Me
Me
Me
Me
Me
Me
Me
Me
H
NH2
H
H
Bn
Bn
H
2.2
H
Cl
Cl
Cl
Cl
Cl
Cl
Cl
0.68
0.11
2.5
NH2
H
H
Bn
H
iPr
iPr
Me
Me
cPent
NH2
H
H
0.27
1.4
Me
Me
H
NH2
NH2
0.26
0.15
a Values are means of multiple experiments. na, not applicable.
In order to bypass this first enzymatic phosphorylation
step, we elected to synthesize the bis(tBuSATE) mono-
phosphate prodrugs of these nucleosides following a
well-established literature procedure.4 The phospho-
ramidite 6 was synthesized and condensed with 5a and
5b in the presence of tetrazole in dimethylformamide,
followed by oxidation of the resulting phosphite triesters
to afford the phosphotriesters 7a and 7b (Scheme 2).
These two prodrugs were tested for their antiviral activ-
ity in the replicon assay (Table 1). Their EC50s were 60
and 24 nM, respectively, which represents a 3log
improvement in antiviral efficacy when compared to
their parent nucleosides 5a and 5b. However, these
SATE prodrugs showed poor stability in human plasma
with half-lives of just a few minutes. This characteristic
was also observed with similar prodrug moieties at-
tached to cyclic monophosphates.5 This limitation led
us to turn our focus toward a different prodrug ap-
proach involving phosphoramidate esters. These pro-
drugs are also well documented in the literature6 and
have demonstrated some good plasma stability in vari-
ous species.
Selected compounds were tested for their in vitro stabil-
ity in multiple assays, including human plasma, human
simulated gastric fluid (SGF), and human simulated
intestinal fluid (SIF). All were stable in human plasma
and SGF for up to 1 h, but the SIF stability proved to
be more problematic. As described in the literature,7
the degradation products observed resulted from the
hydrolysis of the amino acid ester followed by a hydro-
lysis of the phenol group. As seen in Table 2, com-
pounds with R2 and R3 as hydrogens performed
poorly. This was not an issue for us because they were
also poorly active (36a,b–37a,b). However, the most ac-
tive compounds bearing a methyl as R2 and a hydrogen
as R3 also showed poor stability, especially when R4 was
a benzyl (38a,b–39a,b). Interestingly, replacing the ben-
zyl by a more bulky cyclopentyl (42b) did not help the
situation. However, replacing the R4 benzyl with an iso-
R2
O R2
R2
O
R4
O
R4
R4-OH
a)
HO
NH
Boc
O
NH2
O
NH
Boc
R3
R3
b)
R3
The aminoacids 8–10 were esterified, then their Boc pro-
tecting group was removed before condensation with
phosphodichlorides 23 and 24 which yielded the chloro-
HCl
17-22
8-10
11-16
R5
O
Cl
c)
P
Cl
O
23,24
O
S
O
H
N
N
R2
R4
O
P
O
O
O
H
H
N
N
N
R2
N
N
N
S
O
O
O
R4
O
Cl
O
P
O
N
N
O
O
O
O
H
N
R3
S
S
S
S
R1
O
N
N
O
N
O
O
O
O
O
O
R3
5a,b
d)
P
N
P
R1
O
a)
b)
HO
OH
R5
25-33
34a-41a R1=H
HO
OH
34b-42b R1=NH2
R5
6
7a R1=H
7b R1=NH2
Scheme 3. Reagents and conditions: (a) TEA, DMAP, isopropenyl
chloroformate, DCM, 0 °C to rt, 1–20 h; (b) TFA, DCM, rt, 2 h; (c)
TEA, DCM, ꢀ10 °C, 2 h; (d) 1-methylimidazole, DCM, rt, 20 h.
Scheme 2. Reagents and conditions: (a) 5a or 5b, tetrazole, DMF, rt,
2 h; (b) tBuOOH, DMF, rt, 3 h.