SAR studies on a novel series of human cytomegalovirus primase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2007.01.109

A novel series of imidazolylpyrimidines were found to possess inhibitory activity against the human CMV UL70 primase. Extensive SAR studies on an HTS lead led to potent, orally bioavailable compounds with anti-CMV IC₅₀ values of 150 nM in both

Full text of DOI:10.1016/j.bmcl.2007.01.109

Bioorganic & Medicinal Chemistry Letters 17 (2007) 2188–2192
SAR studies on a novel series of human cytomegalovirus
primase inhibitors
X. Chen,* J. Adrian, T. Cushing, H. DiMaio, L. Liang, V. Mayorga, S. Miao,
M. G. Peterson, J. P. Powers, F. Spector, C. Stein, M. Wright, D. Xu,
Q. Ye and J. Jaen
Amgen Inc., 1120 Veterans Boulevard ASF2-3, South San Francisco, CA 94080, USA
Received 29 November 2006; accepted 24 January 2007
Available online 8 February 2007
Abstract—A novel series of imidazolylpyrimidines were found to possess inhibitory activity against the human CMV UL70 primase.
Extensive SAR studies on an HTS lead led to potent, orally bioavailable compounds with anti-CMV IC50 values of 150 nM in both
viral yield and viral DNA replication assays and with a much reduced cytotoxicity compared to marketed treatments ganciclovir and
cidofovir.
Ó 2007 Elsevier Ltd. All rights reserved.
Cytomegalovirus (CMV) is a subgroup of herpesviruses
that is widely distributed in nature. Human cytomegalo-
virus (hCMV) infects, by age 40, about 50% of the pop-
ulation outside of urban centers and up to 80% of the
population within the cities in North America.1 CMV
infection in normal hosts results in a sub-clinical
response that is usually not diagnosed or treated. How-
ever, infection of allograft recipients (kidney, liver,
Figure 1. Proposed mechanism of action for compound 1 and analogs.
heart, and bone marrow) is a major cause of graft rejec-
tion.2–4 Human CMV also causes retinitis in AIDS
patients with low CD4 counts, a situation expected to
N
N
continue unless an appropriate treatment against hCMV
is implemented. The existing treatments for hCMV suf-
fer from poor oral bioavailability, sub-optimal antiviral
activity, as well as propensity for induction of severe
neutropenia.5 Additionally, the emergence of ganciclo-
vir- and cidofovir-resistant virus has resulted in an in-
crease in the rate of disease relapse in the high-risk
patient population.6 High-throughput screening using
a cell-based hCMV-luciferase replication assay discov-
ered compound 1.7 This and related compounds were
found to be inhibitors of the hCMV UL70 primase.
N
N
NO2
NO2
N
N
Ph
Bn
N
N
N
N
1, IC50 = 0.3μM
2, IC50 =0.06μM
N
N
O
O N
O
N
N
NO2
NO2
NO2
N
N
In an earlier publication,8 we disclosed the initial SAR
results and radiolabeling experiments that led to the
Bn
Bn
N
Bn
N
N
N
N
N
5, IC50 = 25μM
3, IC50 =30μM
4, IC50>30μM
Figure 2. Attempts to prepare a non-covalent CMV primase inhibitor.
(IC50 values shown refer to inhibition of viral DNA replication
determined by Dot blot assay9).
Keywords: Anti-viral; hCMV; Primase; Cytomegalovirus; Inhibitor.
*
Corresponding author. Tel.: +1 650 244 2596; fax: +1 650 837
9369; e-mail: xiaoqic@amgen.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.01.109

X. Chen et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2188–2192
2189
Table 2. Inhibition of hCMV replication (Dot blot assay9) by
2-alkylamine analogs
identification of UL70 as the viral target. Resistant virus
cultivated in the presence of 1 carried a single point
mutation of P571 – S571. It was believed that the thiol
group of cysteine 570 of the UL70 primase acts as a
nucleophile that attacks compound 1 with displacement
of the imidazole moiety. This results in the formation of
a covalent link between the target enzyme and the drug,
rendering the primase inactive (Fig. 1). Not surprisingly,
early attempts to find a bioisosteric replacement of the
imidazole, such as a C-linked 5-oxazole, 5-oxadiazole
or 4-isoxazole, met with failure (Fig. 2). Removal of
the imidazole or methyl imidazole resulted in a tremen-
dous reduction in antiviral activity. Previous SAR
N
N
NO2
N
R
N
H
N
Compound
R
H
IC50 (lM)
26
2.0
27
28
1.0
0.5
Table 1. Inhibition of hCMV replication (Dot blot assay9) by
2-substituted benzylamine analogs
29
30
31
32
33
34
35
36
0.3
0.13
2.3
0.35
2.3
0.2
1.3
0.9
N
N
NO2
N
X Ph
N
H
N
Ph
Compound
X
IC50 (lM)
O
6
7
2-CH3
3-CH3
4-CH3
4-Cl
0.2
0.5
1.6
0.3
0.15
0.2
0.25
0.3
15
S
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
4-F
O2
S
4-CF3
3-F
3-OCF3
O
3,5-di-CF3
Ph@3-Py
2-OCF3
0.8
0.2
0.3
0.3
0.4
0.25
0.2
0.15
0.2
0.2
0.5
37
38
0.6
0.2
3-OCF3
3-OCH3
3,4-di-F
2-Cl, 4-F
2-F, 4-CF3
2,5-di-F
3,4-di-Cl
2,3-di-Cl
Ph@2-Py
39
40
41
0.12
0.2
OH
Cl
OH
0.15
a
b
NO2
NO2
NO2
N
N
N
HO
N
Cl
N
Cl
N
III
42
43
0.15
0.08
I
II
N
OH
Cl
N
c
d
e
NO2
NO2
N
N
NO2
N
RHN
N
RHN
N
RHN
N
VI
V
IV
44
45
0.1
0.5
Scheme 1. Reaction conditions for analog synthesis: (a) Et4NCl, N,
N-dimethylaniline, POCl3, CH3CN, 87%; (b) NaOAc, AcOH, EtOH,
H2O, 79%; (c) RNH2, chloroform, reflux, 35–80%; (d) POCl3, 100 °C,
70–85%; (e) 2-methyl imidazole, CH3CN, 80 °C, 24 h, 95%.

2190
X. Chen et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2188–2192
studies also indicated that replacement of the nitro
group with other electron-withdrawing groups also
significantly diminished the antiviral activity. Analogs
with 2-methyl imidazole had similar antiviral activity
as analogs containing unsubstituted imidazole.8
group in the benzylamino series. Hindered secondary
amino groups were expected also to be less metabolically
labile than the benzylamino group toward oxidative
dealkylation. The polarity of the alkylamino group
Table 4. Inhibition of hCMV replication (Dot blot assay9) by
2-heterocyclo-amine analogs
Subsequent SAR studies were focused on close analogs
of 1 by investigating the effect of substitutions on the
phenyl ring while maintaining the methyl imidazole moi-
ety intact. SAR results are summarized in Table 1. The
antiviral activity (IC50) was measured by using a Dot
blot CMV DNA replication assay.9 In general, the
SAR trend is rather flat; there are no large increases in
activity due to change of substitution patterns in this
series. Most active inhibitors (IC50 6 200 nM) contain
a small electron-withdrawing group on the phenyl ring.
There is no additive SAR effect from multiple substitu-
tions. Replacement of the electronically deficient phenyl
ring in 10 by a pyridine group (compound 10 vs 15 and
25) resulted in a slight reduction of antiviral activity.
When compound 10 was dosed in rats, it showed rela-
tively high clearance and poor oral bioavailability.
Additionally, the majority of the compounds in Table
1 had poor aqueous solubility that prevented them from
further in vivo evaluation.
N
R1
N
NO2
N
R
N
H
N
Compound
R
IC50 (lM)
53
3
O
54
55
56
0.4
>5
4
S
O2S
O
57
0.4
0.5
0.15
0.15
0.7
0.25
2
O
We speculated that a sterically bulky alkylamino group
could mimic the hydrophobic interactions of the benzyl
58a
59c
60c
61
O
Table 3. Inhibition of hCMV replication (Dot blot assay9) by
2-cycloalkylamine analogs
O
O
N
N
NO2
N
R
N
H
N
O
Compound
R
IC50 (lM)
S
46
0.1
62
47
48
39
49
50
51
52
0.7
63
O2S
O
0.09
0.12
0.3
64
1
O
65
0.3
0.6
0.2
0.13
66b
67c
68
O
O
1.5
O
0.1
a For the synthesis of alkyl amine for 58 see Ref. 12.
b Synthesis of alkyl amine for 66 see Ref. 13.
c Synthesis of alkyl amines for 59, 60, and 67, see Ref. 14.
0.15

X. Chen et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2188–2192
2191
Table 5. Antiviral activity, cellular toxicity, and bone marrow cellular toxicity comparison of 60, GCV (ganciclovir), and CDV (cidofovir)
Compound
CMV replication9
CMV yield17
HFF Tox.15
Jurkat Tox.15
CFU-GM Tox.16
IC50 (lM)
EC99.9 (lM)
CC50 (lM)
CC50 (lM)
CC50 (lM)
60
GCV
CDV
0.13
1
0.4
5
100
>100
100
50
>100
100
50
30
6
0.35
0.9
could then be subsequently adjusted to increase aqueous
solubility.
rate of oxidative elimination in vivo. When compound
46 (Table 3) was incubated with rat liver S9 microsomes,
the hydroxylation of the cyclohexyl group also appeared
to be a major route of metabolism. Hence, it was neces-
sary to design newer analogs with improved aqueous
solubility and oxidative stability in order to obtain good
oral CMV inhibitors.
The general procedure for the preparation of this series
of compounds is depicted in Scheme 1. Dichlorination
of dihydroxypyrimidine I with phosphorooxytrichloride
in refluxing acetonitrile produced 2,4-dichloropyrimi-
dine II in 87% yield.10 The 4-chloro atom in II was selec-
tively removed under careful basic hydrolysis with
NaOAc and HOAc to give III.11 Amino substitution
of 2-chloropyrimidine III in refluxing chloroform fur-
nished compounds of general formula IV in 35–80%
yields. The 4-chloro group of V was then reintroduced
by reacting IV with phosphorooxytrichloride. The imid-
azole group was introduced by reaction with 2-methyl
imidazole in refluxing acetonitrile to yield the final com-
pounds VI (Scheme 1). All final compounds showed sat-
To address this issue, we explored heterocyclic alkyl
analogs (Table 4). Simple tetrahydropyran replacement
of the cyclohexyl group resulted in a large decrease in
antiviral activity (53 vs 38), while the more polar thiacy-
clohexane or thiacyclopentyl groups further decreased
the antiviral activity (55 and 63). This observation is
consistent with the earlier SAR trend that polar alkyl
groups containing sulfone and ethers are not very well
tolerated in the side chain. Fortunately, analogs with
20-methyl substitution of the tetrahydropyran or the
tetrahydrofuran group achieved similar potency as the
20-methyl cyclohexyl analogs. Tetrahydrofuran or tetra-
hydropyran analogs had a good balance of potent anti-
viral activity, good aqueous solubility, and in vitro
metabolic stability. Compounds 60 and 68 showed good
pharmacokinetic profiles in rats with clearance values of
1.4 and 2.2 L/h/kg, and oral bioavailabilities of 21% and
57%, respectively. Additionally, compound 60 showed a
much larger window between antiviral activity and cellu-
lar toxicity in the HFF, Jurkat,15 and human bone mar-
row cells, when compared with ganciclovir and cidofovir
(Table 5).16
1
isfactory H NMR and LC mass spectral analysis. All
the compounds are racemic except as noted.
Analogs with the proper alkylamine substituents display
potent anti-CMV activity (Table 2). The antiviral activ-
ity increases as the size of the alkyl side chain increases
(from 26 to 30). However, if the substitution is too large
like 31 or 32, there is not further increase of antiviral
activity. Polar groups, such as sulfone and ethers, are
not very well-tolerated in the side chain (33, 35, and
36). The cyclohexyl analog (38) has good antiviral activ-
ity compared with the cyclopentyl analog (37). Addi-
tional substitution around the cyclohexyl ring, in
particular, 20-methyl substitution, provides the best anti-
viral activity (43). Only small alkyl-substituents on the
20-position of the cyclohexyl ring are tolerated in this
series.
Summary. A novel class of non-nucleoside CMV inhib-
itors has been identified. The compounds selectively and
irreversibly bind to CMV UL70 primase, resulting in the
blockade of CMV replication and infection. Detailed
SAR studies on this series of CMV inhibitors yielded
several potent and orally bioavailable drug candidates
that have the potential to become a novel therapeutic
approach to CMV infection.
Inhibitors with a methyl substitution at the 30 or 40 posi-
tion of the cyclohexyl group are not as potent as the 20
substituted analog. 20-Methylcyclohexyl appears to be
optimal in this series. Substituents larger than ethyl
group (data not shown) compromise antiviral activity.
For substituted cyclohexyl side chains, trans-substitu-
tion leads to better antiviral activity than cis-substitu-
tion. For example, trans compound 46 is sevenfold
more potent than its cis isomer 47. Most compounds
containing cycloamino substituents at the 2-position of
the pyrimidine ring still displayed poor (<5%) oral bio-
availability in rats. In contrast to the in vitro glutathione
reactivity of these compounds, glutathione replacement
of the methyl imidazole or imidazole does not appear
to be a major metabolic pathway in vivo. For example,
compound 26 showed bioavailability of 59% after oral
dosing of 2 mg/kg to rats. The poor oral bioavailability
of the more potent analogs is likely due to their limited
aqueous solubility, hence the poor absorption, and high
Acknowledgment
We acknowledge the late Professor William H. Burns,
Department of Medicine, Medical College of Wisconsin,
for performing the bone marrow toxicity studies.
References and notes
1. <http://www.cdc.gov/cmv/facts.htm.>
2. Ho, M. Cytomegalovirus: Biology and Infection. In
Greenough, W. B., Merigan, T. C., Eds.; Plenum Publishing:
New York, 1982; pp 171–204.

2192
X. Chen et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2188–2192
3. Meyers, J. D.; Spencer, H. C., Jr.; Watts, J. C.; Gregg, M.
B.; Stewart, J. A.; Troupin, R. H.; Thomas, E. D. Ann.
Intern. Med. 1975, 82, 181.
described in reference (Amouroux, R.; Slassi, A.; Saluzzo,
C. Heterocycles, 1993, 36(9) 196), followed by
hydrogenation.
4. Neiman, P.; Wasserman, P. B.; Wentworth, B. B.; Kao, G.
F.; Lerner, K. G.; Strob, R.; Buckner, C. D.; Clit, R. A.;
Fefer, A.; Fass, L.; Glucksberg, H.; Thomas, E. O.
Transplantation 1973, 15, 478.
5. Goodrich, J. M.; Bowden, R. A.; Fisher, L.; Keller, C.;
Schoch, G.; Meyers, J. D. Ann. Intern. Med. 1993, 118(3),
173.
6. Bhorade, S. M.; Lurain, N. S.; Jordan, A.; Leischner, J.;
Villanueva, J.; Durazo, R.; Creech, S.; Vigneswaran, W.
T.; Garrity, E. R. J. Heart Lung Transplant. 2002, 21(12),
1274.
7. The luciferase screen involved infection of human foreskin
fibroblasts (HFF) with a recombinant virus (rCMVLUC)
containing a luciferase reporter gene under the control of
HCMV UL99 promoter, inserted in the CMV genome
between US9 and US10. Luciferase activity was measured
following incubation with various concentrations of the
drugs.
8. Cushing, T. D.; Adrian, J.; Chen, X.; DiMaio, H.;
Doughan, B.; Flygare, J.; Liang, L.; Mayorga, V.; Miao,
S.; Mellon, H.; Peterson, M. G.; Powers, J. P.; Spector, F.;
Stein, C.; Wright, M.; Xu, D.; Ye, Q.; Jaen, J. Bioorg.
Med. Chem. Lett. 2006, 16(18), 4879.
9. The Dot blot assay was used to determine the drug
susceptibility against CMV. The foreskin fibroblast (HFF)
cells were infected with cytomegalovirus (rCMVLUC).
The infected HFF cells were incubated with various
concentrations of the drug for 5 days. The rCMVLUCD-
NA was then harvested and immobilized on a membrane.
The viral DNA was quantified using a radio-labeled probe
containing CMV-specific sequences hybridized with the
membrane. The IC50 values represent the drug concentra-
tion required to inhibit 50% of the amount of radioactivity
captured by the membrane when the infected cells were
treated only with drug vehicle.
10. (a) Clark, J.; Ramsden, T. J. Chem. Soc., C 1971, 675; (b)
Hanson, J.C. PCT Int. Appl. WO 91/01310. see also
Hodgetts, K.J.; Yoon, T.; Huang, J.; Gulianello, M.;
Kieltyka, A.; Primus, R.; Brodbeck, R.; De Lombaert, S.;
Doller, D. Bioorg. Med. Chem. Lett. 2003, 13(15), 2497.
11. Clark, J.; Ramsden, T. J. Chem. Soc. C 1971, 679.
12. The starting amine was synthesized by azide replacement
of the corresponding alcohol synthesized by the method
13. The starting amine was prepared similarly to examples
in Ref. 12 starting with the same alcohol followed
by Mitsunobu reaction with p-chlorobenzoic acid
and followed by hydrolysis, azide replacement, and
reduction.
14. The starting amine was prepared by methyl Grignard
reagent opening of epoxide catalyzed by CuI, azide
replacement of the alcohol, and azide reduction. The
enantiomerically pure amine was prepared by chiral reso-
lution with D- or L-malic acid recrystallized in ether. The
absolute stereo chemistry was determined by X-ray
crystallography of the crystal of 4-iodo-N-((3R,4S)-4-
methyl-tetrahydrofuran-3-yl)benzeneulfonamide and 4-iodo-
N-((3S,4R)-4-methyl-tetrahydrofuran-3-yl)benzeneulfonamide.
15. The cytotoxicity assay in HFF or Jurkat cells was
performed by Alamar blue assay. HFF or Jurkat cells
were seeded into 96-well plates and then incubated with
various concentrations of compound. This was followed
by Alamar blue (10 lL) staining at various time points
(0, 24, 48, and 72 h). After incubating an additional 3 h,
the fluorescence was determined (PerSeptive Biosystems
CytoFluor II) and quantified (MS Excel). The IC50 value
represents the concentration of compound that inhibits
cell growth by 50%.
16. The cytotoxicity assay in bone marrow cells was
performed in the following way. Human bone marrow
cells were harvested, washed with Iscoves modified
Dulbecco’s medium (IMDM)–2% FBS, and the viable
nucleated or mononuclear cell number was determined.
The cells were diluted with IMDM–2% FBS to 106 cells/
mL, at which point the drug was added. The cells were
diluted 1:10 methocult (stem cell technologies), vortexed,
and plated onto tissue culture dishes. After incubating at
37 °C in a humid 5% CO2 atmosphere for 15 days, the
CFU-GM colonies were quantified by microscopic
inspection (inverted microscope). The concentration of
drug that inhibited colony formation by 50% (IC50) was
calculated.
17. The viral yield from the rCMVLUC-infected HFF
cells was determined by the immunofluorescence assay
using human antiserum to HCMV after 5–7 days
incubation in the presence of various concentrations
of the drug.