Antagonists of human CCR5 receptor containing 4-(pyrazolyl)piperidine side chains. Part 3: SAR studies on the benzylpyrazole segment

Article abstract of DOI:10.1016/j.bmcl.2003.12.006

Extensive SAR studies in our benzylpyrazole series of CCR5 antagonists have shown that both lipophilic and hydrophilic substituents on the phenyl of the benzyl group increase antiviral potency. However, improvements in pharmacokinetic profiles were generally only observed with more lipophilic substitutions. 4-Biphenyl (51) performed the best in this regard. Highly lipophilic substituents impart undesirable ion channel activity to these CCR5 antagonists. Alkoxy substituents provide a good balance of antiviral activity, pharmacokinetic parameters, and selectivity. Compounds 42b and 42d, containing a 3,4-dimethoxy substituent, are considered the most promising improvements over parent compounds 9. They demonstrate improved antiviral activity while retaining good pharmacokinetic profile and selectivity.

Full text of DOI:10.1016/j.bmcl.2003.12.006

Bioorganic & Medicinal Chemistry Letters 14 (2004) 947–952
Antagonists of human CCR5 receptor containing
4-(pyrazolyl)piperidine side chains. Part 3: SAR studies on
the benzylpyrazole segment
Min Shu,a,* Jennifer L. Loebach,a Kerry A. Parker,a Sander G. Mills,a Kevin T. Chapman,a
Dong-Ming Shen,a,* Lorraine Malkowitz,b Martin S. Springer,b Sandra L. Gould,b
Julie A. DeMartino,b Salvatore J. Siciliano,b Jerry Di Salvo,b Kathy Lyons,a
James V. Pivnichny,a Gloria Y. Kwei,c Anthony Carella,d Gwen Carver,d Karen Holmes,d
WilliamA. Schleif, d Renee Danzeisen,d Daria Hazuda,d Joseph Kessler,d
Janet Lineberger,d Michael D. Millerd and Emilio A. Eminid
aDepartment of Medicinal Chemistry, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
bDepartment of Immunology, Merck Rearch Laboratories, PO Box 2000, Rahway, NJ 07065, USA
cDepartment of Drug Metabolism, Merck Rearch Laboratories, PO Box 2000, Rahway, NJ 07065, USA
dDepartment of Antiviral Research, Merck Research Laboratories, PO Box 4, West Point, PA 19486, USA
Received 2 September 2003; revised 24 November 2003; accepted 2 December 2003
Abstract—Extensive SAR studies in our benzylpyrazole series of CCR5 antagonists have shown that both lipophilic and hydro-
philic substituents on the phenyl of the benzyl group increase antiviral potency. However, improvements in pharmacokinetic
profiles were generally only observed with more lipophilic substitutions. 4-Biphenyl (51) performed the best in this regard. Highly
lipophilic substituents impart undesirable ion channel activity to these CCR5 antagonists. Alkoxy substituents provide a good
balance of antiviral activity, pharmacokinetic parameters, and selectivity. Compounds 42b and 42d, containing a 3,4-dimethoxy
substituent, are considered the most promising improvements over parent compounds 9. They demonstrate improved antiviral
activity while retaining good pharmacokinetic profile and selectivity.
# 2004 Elsevier Ltd. All rights reserved.
Since the discovery of the chemokine receptor CCR5 as
a co-receptor with CD4 for HIV-1 cell entry, there has
been an intense interest to discover small molecule CCR5
antagonists as potential agents for the treatment of
HIV-1 infection.1ꢀ12 Our colleagues fromthese laboratories
have reported on 1-amino-2-aryl-4-(piperidin-1-yl)-
butanes,1,2,12 trans-1,3,4-trisubstituted pyrrolidines,2ꢀ7
and related compounds11 as CCR5 antagonists. Excel-
lent in vitro antiviral activities against HIV-1 and
good pharmacokinetic profiles in animals have been
demonstrated for some of these compounds.
In the preceding two papers, we reported that replacing
a phenylpropyl side chain with a heterocycle such as a
benzylpyrazole or a benzylisoxazole provided com-
pounds such as 1 with potent CCR5 binding and in
vitro antiviral activities.13 Although compounds in this
Keywords: CCR5 antagonist; HIV-1; Antiviral; Pyrazole.
* Corresponding authors. Fax: +1-732-594-8080; e-mail: min_shu@
merck.com; dongming_shen@merck.com
0960-894X/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.12.006

948
M. Shu et al. / Bioorg. Med. Chem. Lett. 14 (2004) 947–952
cyclohexyl series do not have suitable off-target selectivity
or pharmacokinetic profiles, further investigation led to
the isopropyl compound 9a having excellent selectivity
and enhanced pharmacokinetics.14 However, 9a showed
considerably less antiviral activity in comparison to 1.
Herein we report the results of our structure–activity
relationship studies on the benzylpyrazole region of
these compounds. The aim was to discover compounds
having the antiviral activity of 1 and the selectivity and
pharmacokinetic profile of 9a. Three isomeric butyl
analogues of 9a14 were also pursued during this study.
the isopropyl and (R)-s-Bu compounds were usually the
least active analogues.
Replacing the benzyl in 9 with cyclohexylmethyl (10)
resulted in about a 10-fold loss of intrinsic antiviral
activity. The losses after accounting for serumprotein
binding are about 3-fold.
A series of halogen
substitutions (11–17) showed little impact on antiviral
activity compared to the parent compound 9a in the
isopropyl series, but in the cyclohexyl series, there was
some improvement in antiviral activity. Adding one
trifluoromethyl group (18) enhanced antiviral activity
slightly. However, the 3,5-bis(trifluoromethyl)phenyl
compounds (19) were much less active than their 4-tri-
fluoromethylphenyl analogues 18 in the antiviral assay.
As was observed previously in the 4-(3-phenylpropyl)-
piperidine series,7 appending electron-withdrawing
groups, such as cyano, to the benzyl ring enhanced
antiviral activity in these pyrazole compounds (20 and
21), especially for 20b and 20d after consideration of
their serumprotein binding. However, this substitution
sometimes had a negative effect on the selectivity of
these compounds. For example, compound 20d had a Ki
of 0.99 mM for the hERG K+ channel versus 14.3 mM
for 9d.20 Inhibition of the hERG K+ channel has
caused several product withdrawals in the industry and
is currently a widespread concern in drug discovery.21,22
Therefore, structural changes that increase the hERG
K+ channel inhibition activity are to be avoided. How-
ever, the hERG K+ channel Ki’s for 9c and 20c were
similar at 29 and >10 mM. The isomeric 3-cyano com-
pounds 21 offered no advantage in activity over 20. On
the other hand, the presence of a polar 4-methylsulfonyl
group as in 22 gave compounds which were inactive at
the hERG K+ channel while retaining the CCR5 bind-
ing and antiviral activity. Increasing the size of the
lipophilic group on the sulfone (23) afforded com-
pounds with similar antiviral activity. However, their
selectivity was diminished. For example, 23d had a Ki of
3.2 mM at the hERG K+ channel compared to >10 mM
for 22d. Increasingly basic functional groups showed
larger decreases in intrinsic antiviral activity in 24 and
25. However, these compounds exhibited substantially
less serumprotein binding than the parent compounds.
Therefore, in the presence of human serum, they gen-
erally showed better antiviral activity than the parent
compounds 9.
Our initial route for the syntheses of the 4-(py-
razolyl)piperidines gave low selectivity for the desired
pyrazole isomer 7a over 7b.13 In order to improve the
selectivity and functional group compatibility, a new
route to pyrazole 7a was developed (Scheme 1). The com-
mercially available acid 2 was converted to the dibromo
olefin 4 via the aldehyde 3. Converting 4 to the terminal
acetylene anion15 and quenching with chlorosilane
followed by Si–Sn exchange16 gave the acetylenic tin
compound 5. We also found that the above two steps
could be combined using tri-n-butyltin chloride during
quenching of the lithiumacetylide. Pd-catalyzed
coupling of 5 with acid chlorides provided acetylenic
ketone intermediates 6 in good yields.17 Cyclization of
6 with hydrazines gave 7a/7b isomeric ratios as high as
20 when the hydrazine was in its free-base form
(R0¼Me). When ethylhydrazine oxalate and DIEA were
used (R0¼Et), the ratio was about 8 compared to 2–3
13
fromthe diketone route described previously.
Both
routes were used to prepare 7a for this study.
Reductive amination of 7a with aldehyde 8 or its
p-methoxybenzyl analogues18 followed by deprotection
afforded the desired compounds for testing (Table 1).
The parent benzyl compounds 9a–e14 are listed for
comparison.
In general, all compounds in Table 1 showed very good
CCR5 binding activities with IC50’s in the low nM range
or below. Antiviral activities in a cell culture HIV-1
infectivity inhibition assay using HeLa cells (hereafter
referred to as HeLa assay), especially in the presence of
50% normal human serum, were considered a more
critical measure of activity for these compounds. The
antiviral activities of compounds with various R-groups
in Table 1 parallel their parent compounds 9a–e. Thus,
ꢁ
.
Scheme 1. Reagents: (a) BH3 THF, THF, 100%; (b) oxalyl chloride, dimethyl sulfoxide, Et3N, dichloromethane, ꢀ78 C–rt, 86%; (c) CBr4, Ph3P,
DCM, 80–90%; (d) (1) n-BuLi, THF; (2) Me3SiCl, 89%; (e) Bu4NF, (Bu3Sn)2O, 60%; (f) (1) n-BuLi, THF; (2) Bu3SnCl, 80%; (g) 2% Pd(PPh3)4 or
Pd(PPh3)2Cl2, 1,2-dichloroethane, reflux, N2, 50–93%; (h) 60–80 ꢁC, ethanol; (i) HCl in MeOH or TFA, anisole, dichloromethane, rt, 46–85% (7a),
4–11% (7b); (j) NaBH(OAc)3, DIEA, 1,2-dichloroethane; (k) Pd/C, H2, MeOH or alternatively in the case of PMB esters, 96% formic acid, rt, 60–
80% (two steps).

M. Shu et al. / Bioorg. Med. Chem. Lett. 14 (2004) 947–952
949
Surprisingly, adding electron-donating groups such as
alkoxyls also improved antiviral activity over the parent
compounds 9 in most cases. For example, the 4- and 3-
methoxy and 4-ethoxy derivatives (26–28) were more
active than 9. However, the 3-ethoxy derivatives (29)
showed no such improvement. As the normal alkyl
groups become longer, the 4-n-propoxy analogues (30)
are more active than parent compounds 9 while n-
butoxy derivatives (31) are less active. Branched and
cyclic alkoxy compounds (32–35) also showed better
antiviral activity with the t-butoxy compounds (35)
being the most active. 4-Benzyloxy and 4-phenoxy
compounds (36–37) were all less active. Substitution
with 4-trifluoromethoxy and 4-(2,2,2-trifluoroethoxy)
(38 and 39) gave similar antiviral activity as 9 in the
presence of human serum. More polar compounds (40)
showed better antiviral activity, even in the presence of
serum. Yet another example of polar substituent
improving activity is the phenol (41).
Next, four pairs of substituents that enhanced antiviral
activity when present individually were combined (42–
Table 1. CCR5 binding affinity and antiviral activity of substituted benzylpyrazole compoundsa
Compd, Ar=
a R=i-Pr
b R=(S)-s-Bub
c R=(R)-s-Bub
d R=t-Bu
e R=c-Hex
9, Phenyl
10, c-Hexyl
11, 4-Fluorophenyl
12, 3-Fluorophenyl
13, 2-Fluorophenyl
14, 3,4-Difluorophenyl
15, 3,5-Difluorophenyl
16, 2,4-Difluorophenyl
17, 4-Chlorophenyl
18, 4-Trifluoromethylphenyl
19, 3,5-Bis(trifluoromethyl)phenyl
20, 4-Cyanophenyl
1.5 (3.7/300)
2.3 (33/>300)
1.1 (3.7, 11/>300)
1.3 (11/300)
2.2 (11/>300)
1.1 (3.7, 11/300)
2.6 (11/300)
5.0 (100/>300)
1.3 (3.7/300)
2.0 (3.7/100)
3.2 (100/>300)
3.1 (1.2/100)
1.7 (3.7/100)
3.5 (3.7/33)
0.8 (<0.14/100)
1.2 (1.2/300)
0.9 (0.4/300)
0.8 (<0.14/100)
1.1 (3.7/300)
1.6 (0.4/100)
1.1 (33, ND)
1.0 (<0.14/33)
0.5 (<0.14/33)
0.9 (<0.14/11)
1.5 (0.4/33)
0.9 (0.4/100)
1.5 (0.4/33)
0.5 (<0.14/33)
2.1 (11/>300)
1.3 (<0.14/3.7)
0.9 (<0.14/11)
0.5 (0.4/100)
1.6 (33/>300)
1.3 (0.4/33)
0.3 (0.4/33)
2.3 (11/>300)
1.3 (<0.14/3.7)
1.0 (<0.14/11)
1.6 (<0.14/11)
1.1 (<0.14/3.7)
1.7 (0.4/11)
21, 3-Cyanophenyl
22, 4-Methylsulfonylphenyl
23, 4-Phenylsulfonylphenyl
24, 3-Pyridinyl
25, 3-(Aminomethyl)phenyl
26, 4-Methoxyphenyl
27, 3-Methoxyphenyl
28, 4-Ethoxyphenyl
2.3 (0.4/11)
0.8 (0.4/33)
3.9 (0.4/33)
1.3 (0.4/33)
13.1 (11/300)
0.6 (<0.14/11)
1.1 (0.4/33)
2.7 (3.7/33)
0.5 (1.2/11)
1.3 (0.4/100)
1.0 (0.4/300)
1.5 (<0.14/33)
0.7 (<0.14/33)
0.9 (0.4/33, 100)
0.4 (<0.14/100)
0.8 (0.4/300)
0.8 (1.2/33)
0.4 (<0.14/33)
0.7 (0.4/33)
0.7 (<0.14/11)
2.0 (11/300)
1.1 (3.7/300)
0.4 (3.7/300)
0.6 (1.2/100)
0.6 (1.2/100)
0.5 (<0.14/11)
1.1 (0.4/11)
1.2 (<0.14/11)
0.7 (<0.14/33)
0.9 (<0.14/33)
0.7 (<0.14/11)
0.2 (<0.14/11)
0.2 (<0.14/100)
0.6 (<0.14/33)
0.6 (0.4/33)
1.1 (<0.14/33)
1.0 (<0.14/33)
1.2 (<0.14/100)
0.6 (<0.14/33)
0.9 (<0.14/100)
1.1 (<0.14/100)
1.0 (0.4/11)
0.3 (<0.14/11)
1.3 (<0.14/11)
0.8 (<0.14/33)
1.3 (<0.14/11)
0.7 (3.7/300)
0.4 (1.2/100)
1.1 (0.4/33)
0.6 (0.4/33)
0.3 (<0.14/11)
1.4 (0.4/11)
1.5 (0.4/33)
0.5 (<0.14/100)
0.5 (<0.14/33)
0.7 (<0.14/11)
0.4 (<0.14/11)
0.3 (<0.14/300)
0.3 (<0.14/11)
0.3 (<0.14/11)
0.3 (0.4/11)
29, 3-Ethoxyphenyl
0.8 (<0.14/>300)
1.2 (1.2/300)
0.8 (3.7/100)
0.3 (<0.14/33)
0.4 (<0.14/33)
30, 4-n-Propoxyphenyl
31, 4-n-Butoxyphenyl
32, 4-Cyclopropoxyphenyl
33, 4-Cyclobutoxyphenyl
34, 4-i-Propoxyphenyl
35, 4-t-Butoxyphenyl
36, 4-Benzyloxyphenyl
37, 4-Phenoxyphenyl
38, 4-Trifluoromethoxyphenyl
39, 4-(2,2,2-Trifluoroethoxy)phenyl
40, 4-Difluoromethoxyphenyl
41, 4-Hydroxyphenyl
42, 3,4-Dimethoxyphenyl
43, 1,3-Benzodioxol-5-yl
44, 4-Methoxy-3-cyanophenyl
45, 4-Ethoxy-3-fluorophenyl
46, 1-Naphthyl
0.8 (3.7/>300)
0.5 (11/300)
1.0 (1.2/300)
0.6 (1.2/100)
0.7 (1.2/100)
0.8 (0.4/33)
0.2 (<0.14/33)
0.8 (<0.14/100)
1.6 (<0.14/3.7)
0.5 (3.7/100)
0.5 (3.7/300)
0.6 (1.2/100)
0.5 (3.7/100)
0.6 (<0.14/11)
0.6 (<0.14/11)
2.6 (0.4/33)
0.5 (<0.14/11)
0.6 (11/>300)
0.8 (3.7/300)
0.5 (1.2/100)
0.5 (<0.14/100)
0.6 (1.2/33)
0.5 (11/33)
0.8 (1.2/300)
0.4 (<0.14/100)
0.4 (<0.14/33)
0.2 (<0.14/33)
0.2 (<0.14/300)
47, 2-Naphthyl
48, 4-Tolyl
49, 4-i-Propylphenyl
50, 4-t-Butylphenyl
51, 4-Biphenyl
0.8 (1.2/300)
0.5 (1.2/300)
0.4 (0.4/33)
0.6 (1.2/100)
0.3 (<0.14/100)
0.3 (0.4/33)
0.3 (0.4/11)
a Data shown as CCR5 IC50, nM (HeLa IC90, nM/HeLa IC90, nM in the presence of 50% normal human serum). CCR5 IC50’s reported are averages
of triplicate measurements whose standard errors were normally <15% in a given assay. Assay to assay variability was within ꢂ2-fold based on a
standard compound. CCR5 IC50’s were measured as displacement of [125I]-labeled MIP-1a fromthe CCR5 receptor expressed on CHO cell
membranes. See ref 3 note 20 for assay protocol. See ref 19 for HeLa assay conditions.
b The structures of 9b and 9c are shown in ref 23.

950
M. Shu et al. / Bioorg. Med. Chem. Lett. 14 (2004) 947–952
45). These compounds did not show any additive
improvements in antiviral activity (compare 44 with 21
and 26, for example).
could not be considered predictive of the human profiles
for those compounds. Thus, compounds having very
low volume were often not evaluated further in our
series because of difficulties in measuring plasma
binding for a large number of compounds. It should be
noted, however, 9a had similar binding to rat and
human plasmas.14
Replacing the phenyl group in 9 with a 1-naphthyl
group (46) improved antiviral activity. However, 2-
naphthyl (47) did not show an improvement in the pre-
sence of human serum. Adding alkyl groups to the
4-position of the benzyl group (48–50) enhanced the
antiviral activity. The size of the alkyl groups did not
appear to be important since methyl, i-propyl and
t-butyl all gave the same activity. The 4-biphenyl com-
pounds (51) showed some enhancement in antiviral
activity, especially for 51e.
Addition of 4-fluoro in 3-phenylpropyl series7 reduced
clearance. However, the trend was the opposite for
4-fluoro analogue 11a. The 3-fluoro compound 12a had
similar or only slightly better exposure than 9a con-
sidering the data were obtained using mixture dosing.
Compounds 22a, 23a, and 41a were the most active
antivirals in the isopropyl series. The PK properties of
22a and 23a were similar and poor in the rat. Com-
pound 41a was not dosed because 41b–e all showed
poor pharmacokinetics. Mixture dosing of 14a, 17a,
18a, 20a, 26a, 38a, and 40a gave similar pharmaco-
kinetics as 9a. The only compound showing a clearly
better PK profile than 9a in this series was 51a, albeit
with a very low Vdss. It appeared difficult to improve
the antiviral activity of 9a in the isopropyl series while
preserving the good PK and selectivity.24
The pharmacokinetic (PK) profiles of selected com-
pounds from Table 1 were first evaluated in the rat
(Table 2). In a related series of compounds, some com-
pounds were extremely tightly bound to rat plasma
(>99.97%) but only moderately bound to human
plasma (ꢃ97%). In the rat, this correlated with very
low volumes of distribution, low clearances and long
half-lives. However, due to the discordance in protein
binding, it was recognized that good rat PK results
The PK parameters and selectivity profile of the (S)-s-
Bu series is typical among the related series. The PK of
9b was characterized by moderate clearance, short half-
life, and low oral AUCN. Most of the compounds in
this series displayed similar or only marginally
improved PK’s over 9b. Compounds demonstrating
better PK parameters than the parent 9b included the
alkoxy compounds 28b, 31b–33b, 35b, 42b, fluoroalkoxy
analogue 38b, alkyl derivatives 49b, 50b, and biphenyl
51b. However, their ion channel activities increase with
increasing lipophilicity and decreasing polarity of the
substituents. The best combination of antiviral activity,
pharmacokinetics, and selectivity was found in the 3,4-
dimethoxy compound 42b. The improved selectivity of
dialkoxy compounds appears general. Another example
was 43b with a Ki of 27.5 mM at the hERG K+ channel.
One more interesting compound was 38b which showed
a reasonable iv profile in the dog (Table 3), in addition
to its good antiviral activity.
Table 2. Pharmacokinetic parameters of CCR5 antagonists in the rat
and activity at the hERG K+ ion channela
Compd
Clp
AUCNpo
.
Vdss t1/2
F
hERG K+
(mL/min/kg) (mM h/dose) (L/kg) (h) (%) Ki (mM) (n)
9a
11a
12ab
22ab
51ab
9b
3.2
7.6
3.9
11
0.32
25
1.1
0.33
3.3
0.17
22
0.16
0.38
1.2
5.7
7.5
5.2
1.3
2.6
0.28 1.4 11
0.35 0.91 7.8 ND
0.16 0.80 40 ND
0.25 0.38 5.4 ND
0.12 5.7 26 ND
0.64 0.54 13 30 (2)
0.15 0.98 5.9 16
>10
28b
4.4
4.8
1.9
2.7
4.0
8.8
3.5
7.4
3.3
1.6
0.52
9.8
0.75
0.97
1.8
5.7
2.4
1.7
0.62
0.98
3.9
2.1
1.1
0.23
31bb
32bb
33bb
35bb
38bb
39bb
42bb
49bb
50bb
51bb
20c
0.40 1.2 22
0.07 1.0 36
0.21 1.2 68
0.31 1.1 76
0.73 2.3 41
0.21 1.1 36
0.53 3.3 17
0.18 1.18 19
ND
ND
5.7 (2)
ND
ND
3.8 (2)
22 (3)
8.1 (1)
0.66
1.7
1.7
0.15 1.4 9.2 ND
0.15 4.1 36 3.0 (2)
0.33 0.74 3.1 >10 (1)
0.10 2.1 26 3.4 (2)
0.17 2.7 23 5.6
0.05 0.73 8.3 37 (2)
20
0.09
9.9
6.6
1.3
3.7
5.2
12
28
14
2.7
4.2
19
39
0.01
The less active (R)-s-Bu series generally showed better
PK profiles. This appears to hold with less lipophilic
substituents such as 33c, 35c, 40c, 45c, and 46c. For
33cb
35cb
42cb
9db
20db
28db
33db
34d
0.24 0.79 71
0.13 1.2 43
0.20 1.4 68
0.18 3.5 65
0.15 1.4 49
0.27 1.1 39
0.13 1.1 33
0.22 2.5 77
0.14 8.9 33
14
1.0 (2)
ND
3.5 (2)
9.6
6.0 (2)
11 (3)
2.1 (2)
2.3 (2)
ND
Table 3. Pharmacokinetic parameters of CCR5 antagonists in the
doga
Compd
Clp
(mL/min/kg)
AUCNiv,
.
Vdss
(L/kg)
t1/2
(h)
F
(%)
38d
42d
(mM h/dose)
50db
51db
9e
38bb
9d
3.8
13
25
5.5
22
108
13
6.8
2.3
1.2
4.9
1.3
0.24
2.3
1.9
0.87
0.50
2.1
0.86
0.99
4.6
3.1
ND
44
ND
ND
7.9
ND
ND
ND
0.64
0.75
2.8
2.7
0.55
1.8
9db
84
27
5.3
3.9
1.6
0.83
1.0 19
3
35eb
51eb
0.18
0.88
ND
0.4 (2)
18db
20d
22db
28db
42db
0.53 1.3 18
a All compounds were dosed at 0.5 mpk iv and 2.0 mpk orally.
b Dosed as a mixture of five compounds. In cases where comparisons
were available, such as 9a, 11a, and 42d, mixture dosing increased
AUCNpo and %F by about 2- to 3-fold while other parameters were
similar.
1.8
1.8
14
2.1
a All compounds were dosed at 0.5 mpk iv and 2.0 mpk orally.
b Dosed as a mixture of five compounds at 0.5 mpk each via iv only.

M. Shu et al. / Bioorg. Med. Chem. Lett. 14 (2004) 947–952
951
Table 4. Antiviral activity in a PBMC cell culture viral spread assay25 and selectivity profile of select CCR5 antagonists
Compd
PBMC viral
spread inhibition, IC95
IC50 (mM) or % inhibition at 10 mM on selected counter-screen targets
nM (n)
50% NHS,
nM (n)
Human
adrenergic NE
transporter (%)
Rat L-type Ca2+
benzothiazepine
(%)
Human
dopamine
D3 (%)
Human dopamine
transporter (%)
Rat Na+ channel,
site 2 (%)
9a
22ꢂ16 (26)
2.5ꢂ1.5 (11)
5.5ꢂ4.1 (11)
1.9ꢂ1.4 (13)
6.2ꢂ2.1 (6)
1.6ꢂ0.8 (6)
350ꢂ320 (14)
20ꢂ19 (8)
21, 27
42
15
ꢀ3, 13
46
36
45, 36
2.7 mM
35
4 mM
39
44
15, 48
3.6 mM
ꢀ4
32, 18
1.2 mM
7.2 mM
44
38b
42b
20d
22d
42d
28ꢂ16 (11)
9.3ꢂ8.5 (6)
8.1ꢂ2.7 (6)
31ꢂ35 (6)
23
13
3
1
15
ꢀ8
17
36
9
7
3.2 mM
more polar groups, PK parameters are either similar to
those of the (S)-s-Bu series (20c, 23c, and 41c) or worse
as in the case of 42c (Table 2).
bioavailable, even in the cyclohexyl series (51e). Highly
lipophilic substituents imparted undesirable ion channel
activity. Alkoxy substituents generally provided a good
balance of antiviral activity, pharmacokinetic para-
meters, and selectivity. Among these, the 3,4-dimethoxy
group was the most outstanding. For example, com-
pounds 42b and 42d represent the most promising
improvements over their parent compounds 9. They
showed improved antiviral activity while retaining good
pharmacokinetic profile and selectivity.
Although the parent compound in the t-Bu series, 9d,
had good PK, most compounds having more polar
substituents in this series had less favorable pharmaco-
kinetics. Compounds with mostly lipophilic substituents
showed better PK profiles than 9d. In both the alkoxy
(26d–35d) and alkyl/aryl (48d–51d) subclasses, increas-
ing the size of the substituents is consistently associated
with higher oral AUCN, longer half-life, and lower
clearance. However, increasing lipophilicity was again
associated with higher activity at the hERG K+ chan-
nel. Because the parent compound 9d did not show a
good half-life in the dog, several compounds from this
series were screened (Table 3). The results showed that
18d, 28d, and 42d showed improvements over 9d. Over-
all, 42d showed the most promising combination of
antiviral activity, PK profiles, and selectivity among the
t-Bu series.
Acknowledgements
The authors would like to thank Dr. G. R. Kieczy-
kowski, Mr. J. F. Leone and Mr. F. Wong for synthetic
assistance and Drs. C. G. Caldwell and P. E. Finke for
reviewing the manuscript.
References and notes
Most compounds in the cyclohexyl series had similarly
inferior rat PK profiles as the parent 9e. Only the
t-butoxy compound 35e and 4-biphenyl 51e showed a
significant improvement in PK parameters (Table 2).
The latter displayed an acceptable PK profile. However,
its selectivity was inadequate.
1. Finke, P. E.; Oates, B.; Mills, S. G.; MacCoss, M.;
Malkowitz, L.; Springer, M. S.; Gould, S. L.; DeMartino,
J. A.; Carella, A.; Carver, G.; Holmes, K.; Danzeisen, R.;
Hazuda, D.; Kessler, J.; Lineberger, J.; Miller, M.;
Schlief, W. A.; Emini, E. A. Bioorg. Med. Chem. Lett.
2001, 11, 2475 and references cited therein.
2. Willoughby, C. W.; Berk, S. C.; Rosauer, K. G.;
Delgrado, S.; Chapman, K. T.; Gould, S. L.; Springer,
M. S.; Malkowitz, L.; Schlief, W. A.; Hazuda, D.; Miller,
M.; Kessler, J.; Danzeisen, R.; Holmes, K.; Lineberger, J.;
Carella, A.; Carver, G.; Emini, E. A. Bioorg. Med. Chem.
Lett. 2001, 11, 3137.
3. Hale, J. J.; Budhu, R. J.; Mills, S. G.; MacCoss, M.;
Malkowitz, L.; Siciliano, S.; Gould, S. L.; DeMartino, J. A.;
Springer, M. S. Bioorg. Med. Chem. Lett. 2001, 11, 1437.
4. Hale, J. J.; Budhu, R. J.; Holson, E. B.; Finke, P. E.;
Oates, B.; Mills, S. G.; MacCoss, M.; Gould, S. L.;
DeMartino, J. A.; Springer, M. S.; Siciliano, S.;
Malkowitz, L.; Schlief, W. A.; Hazuda, D.; Miller, M.;
Kessler, J.; Danzeisen, R.; Holmes, K.; Lineberger, J.;
Carella, A.; Carver, G.; Emini, E. A. Bioorg. Med. Chem.
Lett. 2001, 11, 2741.
A selected group of compounds was further evaluated in
an in vitro peripheral blood mononuclear cell (PBMC)
HIV-1 spread assay25 and selectivity in five counter-
screens (Table 4). They all showed significantly
improved antiviral activity over our lead 9a in the
PMBC assay. However, poor selectivity at the hERG
K+ channel for 20d, multiple low mM activities in the
counter-screens for 38b, and a poor dog PK profile for
22d precluded their further study. On balance, the 3,4-
dimethoxy derivatives 42b and 42d showed the best com-
bination of antiviral activity, PK profile, and selectivity.
Extensive SAR studies in five series of compounds have
shown that both lipophilic and hydrophilic substituents
on the phenyl of the benzyl group in compounds 9a–e
increase their antiviral potency. However, improve-
ments in pharmacokinetic profiles in the rat and/or dog
were generally only observed with more lipophilic sub-
stitutions. For example, 4-biphenyl compounds 51 are
5. Lynch, C. L.; Hale, J. J.; Budhu, R. J.; Gentry, A. L.;
Mills, S. G.; Chapman, K. T.; MacCoss, M.; Malkowitz,
L.; Springer, M. S.; Gould, S. L.; DeMartino, J. A.;
Siciliano, S. J.; Cascieri, M. A.; Carella, A.; Carver, G.;
Holmes, K.; Schlief, W. A.; Danzeisen, R.; Hazuda, D.;
Kessler, J.; Lineberger, J.; Miller, M.; Emini, E. A.

952
M. Shu et al. / Bioorg. Med. Chem. Lett. 14 (2004) 947–952
Bioorg. Med. Chem. Lett. 2002, 12, 3001 and references
cited therein.
14. Shen, D.-M.; Shu, M.; Willoughby, C. W.; Shah, S.;
Lynch, C. L.; Hale, J. J.; Mills, S. G.; Chapman, K. T.;
Malkowitz, L.; Springer, M. S.; Gould, S. L.; DeMartino,
J. A.; Siciliano, S. J.; Lyons, K.; Pivnichny, J. V.; Kwei,
G. Y.; Carella, A.; Carver, G.; Holmes, K.; Schleif, W. A.;
Danzeisen, R.; Hazuda, D.; Kessler, J.; Lineberger, J.;
Miller, M. D.; Emini, E. A. Bioorg. Med. Chem. Lett.
2004, 14, preceding paper in this issue. doi:10.1016/
j.bmcl.2003.12.005.
6. Willoughby, C. W.; Rosauer, K. G.; Hale, J. J.; Budhu,
R. J.; Mills, S. G.; Chapman, K. T.; MacCoss, M.;
Malkowitz, L.; Springer, M. S.; Gould, S. L.; DeMartino,
J. A.; Siciliano, S. J.; Cascieri, M. A.; Carella, A.; Carver,
G.; Holmes, K.; Schlief, W. A.; Danzeisen, R.; Hazuda,
D.; Kessler, J.; Lineberger, J.; Miller, M.; Emini, E. A.
Bioorg. Med. Chem. Lett. 2003, 13, 427.
7. Lynch, C. L.; Willoughby, C. W.; Hale, J. J.; Holson,
E. J.; Budhu, R. J.; Gentry, A. L.; Rosauer, K. G.;
Caldwell, C. G.; Chen, P.; Mills, S. G.; MacCoss, M.;
Berk, S.; Chen, L.; Chapman, K. T.; Malkowitz, L.;
Springer, M. S.; Gould, S. L.; DeMartino, J. A.; Siciliano,
S. J.; Cascieri, M. A.; Carella, A.; Carver, G.; Holmes, K.;
Schlief, W. A.; Danzeisen, R.; Hazuda, D.; Kessler, J.;
Lineberger, J.; Miller, M.; Emini, E. A. Bioorg. Med.
Chem. Lett. 2003, 13, 119.
8. Baba, M.; Nishimura, O.; Kanzaki, N.; Okamoto, M.;
Sawada, H.; Iizawa, Y.; Shiraishi, M.; Aramaki, Y.;
Okonogi, K.; Ogawa, Y.; Meguro, K.; Fujino, M. Proc.
Natl. Acad. Sci. U.S.A. 1999, 96, 5698.
9. Tagat, J. R.; Steensma, R. W.; McCombie, S. W.; Nazareno,
D. V.; Lin, S.-I.; Neustadt, B. R.; Cox, K.; Xu, S.;
Wojcik, L.; Murray, M. G.; Vantuno, N.; Baroudy, B. M.;
Strizki, J. M. J. Med. Chem. 2001, 44, 3343 and references
cited therein.
10. For a review, see: Kazmierski, W.; Bifulco, N.; Yang, H.;
Boone, L.; DeAnda, F.; Watson, C.; Kenakin, T. Bioorg.
Med. Chem. 2003, 11, 2663.
15. Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 13,
3769.
16. Waner, B. P.; Buchwald, S. L. J. Org. Chem. 1994, 5,
5822.
17. Logue, M. W.; Teng, K. J. Org. Chem. 1982, 47, 2549.
18. For the preparation of 8 and its p-methoxybenzyl analo-
gues, see ref 5 and 14 above.
19. Hazuda, D. J.; Felock, P.; Witmer, M.; Wolfe, A.;
Stillmock, K.; Grobler, J. A.; Espeseth, A.; Gabryelski,
L.; Schleif, W.; Blau, C.; Miller, M. D. Science 2000, 287,
646.
20. Inhibition of radio-labeled MK-499 binding using human
embryonic kidney (HEK) cells expressing hERG. For
procedure, see: Wang, J.; Della Penna, K.; Wang, H.;
Karczewski, J.; Connolly, T. M.; Koblan, K. S.; Bennett,
P. B.; Salata, J. J. Am. J. Physiol.-Heart Circ. Physiol.
2003, 284, H256.
21. Crumb, W.; Cavero, I. Pharm. Sci. Technol. Today 1999,
2, 270.
22. Fermini, B.; Fossa, A. A. Nat. Rev. Drug Discov. 2003, 2,
439.
23. The structures of 9b and 9c are:
11. Lynch, C. L.; Gentry, A. L.; Hale, J. J.; Mills, S. G.;
MacCoss, M.; Malkowitz, L.; Springer, M. S.; Gould,
S. L.; DeMartino, J. A.; Siciliano, S. J.; Cascieri, M. A.;
Doss, G.; Carella, A.; Carver, G.; Holmes, K.; Schlief,
W. A.; Danzeisen, R.; Hazuda, D.; Kessler, J.;
Lineberger, J.; Miller, M.; Emini, E. A. Bioorg. Med.
Chem. Lett. 2002, 12, 677.
12. Kim, D.; Wang, L.; Caldwell, C. G.; Chen, P.; Finke,
P. E.; Oates, B.; MacCoss, M.; Mills, S. G.; Malkowitz,
L.; Gould, S. L.; DeMartino, J. A.; Springer, M. S.;
Hazuda, D.; Miller, M.; Kessler, J.; Danzeisen, R.;
Carver, G.; Carella, A.; Holmes, K.; Lineberger, J.;
Schlief, W. A.; Emini, E. A. Bioorg. Med. Chem. Lett.
2001, 11, 3099 and 3103.
13. Shen, D.-M.; Shu, M.; Mills, S. G.; Chapman, K. T.;
Malkowitz, L.; Springer, M. S.; Gould, S. L.; DeMartino,
J. A.; Siciliano, S. J.; Kwei, G. Y.; Carella, A.; Carver, G.;
Holmes, K.; Schleif, W. A.; Danzeisen, R.; Hazuda, D.;
Kessler, J.; Lineberger, J.; Miller, M. D.; Emini, E. A.
Bioorg. Med. Chem. Lett. 2004, 14, preceding ref 14 in this
issue. doi:10.1016/j.bmcl.2003.12.004.
24. The hERG K+ channel activity of 51a was not measured.
It was noted that both 51b and 51d had considerable
increases in hERG K+ channel activity compared to 9b
and 9d, respectively.
25. Saag, M. S.; Emini, E. A.; Laskin, O. L.; Douglas, J.;
Lapidus, W. I.; Schleif, W. A.; Whitley, R. J.; Hildebrand,
C.; Byrnes, V. W.; Kappes, J. C.; Anderson, K. W.;
Massari, F. E.; Shaw, G. M. New Engl. J. Med. 1993, 329,
1065.