H. Matter et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1817–1822
1821
a CF3-group would improve binding affinity. While derivatives 8f
and 8g exhibit slightly increased activity with 1.30 and
2.21
pounds (column MLab Keto in Table 4). In contrast CYP2D6 appar-
ently is much less likely to be involved in this metabolization
process, as the addition of the CYP2D6 inhibitor quinidine (column
MLab Quin in Table 4) does not significantly alter the experimen-
tally observed metabolization rates. From the inspection of Table
4, compound 9m exhibited interesting in vitro ADMET and physi-
cochemical properties with moderate human metabolic lability
and still acceptable logD7.4. This compound is the most potent
member of this series in HepG2 and H4IIE cell-based SCD1 assays
a
l
M compared to the original hit, this option was not explored
further with a larger range of R2 substituents, since alternative sub-
stituents combined with the carboxamide linkers lead to signifi-
cantly improved activity (cf. Table 3).
Table 3 summarizes our efforts to investigate the ortho-position
on the distal phenoxy-moiety of the original hit 2a. The ROCS-de-
rived alignment of 2a with the virtual screening query 1 suggests
that inserting the lipophilic substituent of the query 1 at this posi-
tion in our series should be tolerated by the enzyme. We explored
this option combined with different carboxamide substituents (R1
in Table 3). Introducing an ortho-chlorine substituent on the distal
ring with a thiazolyl-substituent attached to the carboxamide
with IC50 values of 0.157 and 0.182 lM (see Table 4), respectively.
It was also potent with respect to ACC phosphorylation. Therefore,
this compound was selected for further investigation and renamed
SAR224.
An in-depth pharmacokinetic and pharmacological character-
ization of SAR224 was performed. All experimental procedures
were conducted in accordance to the German Animal Protection
Law and international animal welfare legislation and rules and per-
formed as previously described.8,9,19 After intravenous administra-
tion of 5 mg/kg SAR224 in solution to male ZDF rats, a low mean
plasma clearance (0.25 L/h/kg), a large volume of distribution
(Vss = 1.7 L/h/kg) and a long half-life (6.5 h) were observed. After
a single oral dose of 30 mg/kg SAR224 in suspension to male ZDF
rats, the mean plasma concentration profile showed a slow
increase from 0.25 to 6 h after administration and a mean Cmax va-
lue of 1950 ng/mL after 4 h. SAR224 showed a long mean apparent
plasma half-life (8.4 h) and a relative bioavailability of 25% (Fig. 3,
Upper). To evaluate the in vivo pharmacological activity of SAR224,
the compound was administered once orally at 30 mg/kg to
8 week-old male ZDF rats, a well-known animal model of obesity
and diabetes. After 6 h the compound significantly decreased the
serum fatty acid desaturation index (À84.4 15.9%) compared to
(R1 = A) results in potent derivatives, for example, 9a (IC50
:
0.115 lM).
The nature of the amine substituent of R1 significantly influ-
ences SCD1 activity. Replacing the thiazole-ring by a chloro-thio-
phene (R1 = B) results in a lower activity (9b: 0.832
lM), while
aliphatic substituents do not improve activity, as demonstrated
by 9c and 9d with IC50 values of 0.966 and 2.72 lM, respectively.
Interestingly this ranking of R1-substituents is not respected if
the ortho-chlorine of substituent R2 is replaced by an ortho-trifluo-
romethyl group. Here the most active compound is obtained with
the chloro-thiophene moiety (R1 = B), 9f (IC50: 0.142
lM) rather
than with the thiazolyl-moiety, 9e (IC50: 0.212 lM). Combining ali-
phatic R1-substituents with ortho-CF3-substituted biarylethers
does not improve activity, as seen by compounds 9g and 9h with
IC50 values of 0.539 and 0.434
carboxamide substituent by a methyl-group significantly affects
activity (9i: 27.2 M), and even more so with a phenylpropyl-
substituent (9k: >100 M). similar decrease in activity is
observed for the corresponding ethyl-ester substitution at position
R1 (9j: 38.3
M).
l
M, respectively. Replacing the R1-
l
l
A
100000
10000
1000
l
Finally we replaced the ortho-substituted biphenylether by
substituted benzophenone-derivatives (9l–9q). Our intention was
to explore the effect of altered hydrogen-bond accepting properties
on SCD1 inhibition by comparing the carbonyl oxygen to the aro-
matic ether-oxygen which cannot play the role of H-bond acceptor.
Once again the most significant influence on activity is observed
with a thiazole-substituent (R1 = A; 9l: 0.078
thiophene (R1 = B; 9m: 0.136
M), while aliphatic substituents at
R1 once more result in decreased SCD1 inhibition (9n: 3.86
M).
lM) and chloro-
l
l
100
i.v.
These favorable motifs in 9l and 9m were further explored by
changing the distance between the aromatic thiazole ring and
the carboxamide linker.
p.o.
10
Removing the methylene-linker from R1 = A reduces activity, as
0
6
12
18
24
TIME (hours)
can be seen by comparing 9l (IC50: 0.078
4.52 M. Removal of the chlorine substituent in 9m has only a
slightly favorable effect on SCD1 inhibition, leading to 9p with an
IC50 value of 0.099 M. Our final investigation in this series was
to reverse the order of the carboxamide linker, resulting in the
much less active compound 9q (3.20 M).
lM) and 9o (IC50:
l
1,4
1,2
1
l
l
*
This systematic SAR investigation had lead to several molecules
that were interesting as potent SCD1 inhibitors, and which were
then profiled in further assays. A summary of additional in vitro
data, giving cellular SCD1 activities in rat H4IIE and human HepG2
liver cell lines, effect on ACC phosphorylation in human HepG2
0,8
0,6
0,4
0,2
0
cells (at 10 lM concentration with in-cell Western immunolabel-
ing),8 ADME and physicochemical properties is provided in Table
4. In general, the entire chemical series is characterized by high
Caco-2 permeability and acceptable to moderate metabolic lability
rates (e.g., column MLab Human reporting the percentage of a com-
pound metabolized). Addition of the CYP3A4/5 inhibitor ketocona-
zole to the metabolic lability assay reveals an important influence
of CYP3A4/5 on the observed metabolic degradation of the com-
lean
obese
SAR224
Figure 3. Upper: Pharmacokinetic parameters of SAR224 following an intravenous
bolus administration of 5 mg/kg SAR224 and following a single oral dose of 30 mg/
kg SAR224 to male obese ZDF rats. Results are means SD (n = 3). Lower: Decreased
serum fatty acid desaturation indices (À84.4 15.9%) by treatment of male obese
ZDF rats with 30 mg/kg SAR224. Results are means SEM (n = 8) and ⁄p <0.05
versus the vehicle-treated, obese controls.