Design, synthesis, and in vivo activity of novel inhibitors of delta-5 desaturase for the treatment of metabolic syndrome

Article abstract of DOI:10.1016/j.bmcl.2015.07.066

The synthesis, SAR, and in vivo activity of inhibitors of delta-5 desaturase are described. Ring-constraint of the initial series provided access to a variety of in vitro active chemotypes, from which the indazole was selected. Examples from the indazole

Full text of DOI:10.1016/j.bmcl.2015.07.066

Bioorganic & Medicinal Chemistry Letters 25 (2015) 3836–3839
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis, and in vivo activity of novel inhibitors of delta-5
desaturase for the treatment of metabolic syndrome
a,
a
b
b
b
Simon D. P. Baugh a, , Praveen K. Pabba , Joseph Barbosa , Eric Coulter , Urvi Desai , Jason P. Gay ,
Suma Gopinathan b, Qiang Han a, Rajee Hari b, S. David Kimball a, Huy V. Nguyen a, Chi-You Ni a,
David R. Powell b, Arian Smith b, Kristen M. Terranova a, Alan Wilson b, Xuan-Chuan Yu b,
Victoria K. Lombardo a
⇑
⇑
a Lexicon Pharmaceuticals, 350 Carter Road, Princeton, NJ 08540, United States
b Lexicon Pharmaceuticals, 8800 Technology Forest Place, The Woodlands, TX 7738, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
The synthesis, SAR, and in vivo activity of inhibitors of delta-5 desaturase are described. Ring-constraint
of the initial series provided access to a variety of in vitro active chemotypes, from which the indazole
was selected. Examples from the indazole series displayed in vivo activity in reducing the enzymatic
activity of liver delta-5 desaturase.
Received 23 June 2015
Revised 17 July 2015
Accepted 21 July 2015
Available online 26 July 2015
Ó 2015 Elsevier Ltd. All rights reserved.
Keywords:
Delta-5 desaturase
Metabolic syndrome
Indazole
Ring constraint
In vivo efficacy
Metabolic syndrome is associated with cardiovascular disease
as well as type 2 diabetes—in particular insulin resistance, high
blood pressure, central obesity, decreased high-density lipoprotein
(HDL) cholesterol, and elevated triglycerides.1–3 A correlation has
been observed between the level of activity of desaturases and var-
ious disease states, including obesity.4,5 For instance, regulation of
both delta-5 desaturase [also known as fatty acid desaturase
(FADS1)] and delta-6 desaturase activity has been associated with
changes in body weight in mice, humans and other animals.
Desaturases are a family of enzymes that are involved in the pro-
duction of poly-unsaturated fatty acids, beginning with the desat-
uration of palmitic acid, and are named according to the type of
fatty acid they generate (delta-5, delta-6, and delta-9 desaturase).6
Our interest in the desaturase enzymes came about through the
production of almost 5000 lines of knock out (KO) mice.7 In our
hands, delta-5 desaturase KO mice had decreased body fat,
improved glucose tolerance, and lower fasting serum levels of
insulin, cholesterol and triglycerides when compared to wild type
littermate control mice.8 With our knowledge of the high degree of
homology between murine and human delta-5 desaturase, and as
delta-5 desaturase is one of the four desaturases present in
humans, we initiated an effort to discover inhibitors of this target
for the treatment of body composition disorders, such as obesity
and diabetes.
To date, there have been a structurally diverse array of small
molecule inhibitors of delta-5 desaturase previously reported in
the literature, including compounds such as Sesamin 1,9 para-
isopentoxyaniline 2,10 fused-pyrimidones 3,11 dibenzoazepine
4,12 and amide 5,12 Figure 1.
We identified compound 5 as an interesting small molecule
lead. It has previously been disclosed as a delta-5 desaturase inhi-
bitor that was selective over delta-6 and delta-9 desaturase.12
Obukowicz et al.12 investigated compound 5 with a view to its
potential anti-inflammatory effects. From this starting point, we
first investigated variations in the linker between the two aromatic
rings, and the amide linkage proved to be the optimal one. It was
found that the reverse amide, reduction of the amide to an amine,
and methylation of the amide nitrogen all removed inhibitory
activity.
The synthesis of our derivatives was carried out following the
synthetic methods shown in Scheme 1. The appropriate aniline
was reacted with a functionalized isatoic anhydride or the neces-
sary benzoic acid to yield amides 5–12.
We initially set out to survey changes upon the right-hand aro-
⇑
Corresponding authors. Tel.: +1 215 589 6408.
E-mail address: simondpbaugh@gmail.com (S.D.P. Baugh).
matic ring. Under our assay conditions,8 compound 5 showed an
http://dx.doi.org/10.1016/j.bmcl.2015.07.066
0960-894X/Ó 2015 Elsevier Ltd. All rights reserved.

S. D. P. Baugh et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3836–3839
3837
Table 1
O
F
SAR of amide analogs as inhibitors of delta-5 desaturase
F
F
O
O
F
O
N
F
O
X
O
H
H
N
O
N
H
Z
O
O
N
H
O
F
F
F
NH2
Y
3
2
O
Compound
X
Y
Z
IC50 (nM)
O
1
5
6
7
8
9
10
11
12
NH2
NH2
NH2
OH
NH2
NH2
OH
H
H
H
H
F
Cl
F
58
516
843
174
67
34
15
925
H
N
NH2
O
i-Pr
Cl
Cl
Cl
Cl
Cl
Cl
N
H
Cl
OCH3
OCH2CH3
5
4
OH
Figure 1. Representative literature described delta-five desaturase inhibitors.
IC50 of 58 nM (see Table 1). Moving the chlorine atom of compound
5 from the meta position to either the ortho- or para-position gave
compounds with markedly lower potency (39% and 36% enzyme
quinazoline, and naphthalene derivatives 12–14, all showed good
potency, whereas the furo[3,2-c]pyridyl derivative 15 was much
less active. Thieno[3,2-d]pyrimidine 16 showed good activity,
whereas the regioisomeric thieno[2,3-d]pyrimidine 17 was sub-
stantially less potent. The indane, compound 18, was reasonably
potent as the racemate, and the indazole, compound 19, also
showed good potency. From these, as well as other ring-con-
strained analogs not shown, the indazole derivative 19 was
selected for further investigation. With the previous amide and
phenol series’ showing improved activity upon substitution upon
the left-hand aromatic ring, we next set about producing substi-
tuted analogs of 19.
inhibition, respectively, at 1 lM compound concentration), thus
our work focused primarily on changes at the meta position.
Examples of the functional group variations that we employed
included changing the chlorine atom to a fluorine or isopropyl
group, Table 1, compounds 6 and 7, both of which resulted in sig-
nificantly reduced activity. Di-substituted right-hand side deriva-
tives of compound 5 were also pursued, however they too
resulted in reduced activity—for example, 3,5-dichloro (245 nM),
and 3-chloro-4-fluoro (155 nM).
Maintaining the preferred amide linker and right-hand side aro-
matic ring, we next made changes to the left-hand ring. Moving
from X = amino to X = hydroxyl (5 vs 8) showed a slight drop in
activity. Attempts to replace the amino or hydroxyl with other
functionalities such as methyl, cyano, and chloro all provided
derivatives with significantly lower activity, and moving the amino
group around the left-hand ring demonstrated that the optimal
position for the amino group was at the 2-position (for instance,
the 3-amino derivative showed only 54% enzyme inhibition at
The substituted indazole derivatives were synthesized either
via displacement of the 3-halo indazole with the appropriate ani-
line, coupling of the requisite 3-amino indazole with the appropri-
ate aryl boronic acid in the presence of copper(II) acetate, or by
palladium catalyzed aryl amination, Scheme 3.8
In some cases, replacement of the Z chloro substituent with flu-
orine did still produce active molecules, such as 20, but this was
not general, and thus chloro was the Z group of choice, Table 3.
As expected from our previous work, substitutions at the 4-, 6-,
and 7-position of the indazole ring with groups larger than fluorine
were almost exclusively detrimental to the activity against the tar-
get, with a notable exception being the 7-chloro derivative of com-
pound 19 which showed an IC50 of 33 nM. Hence, we focused on
variations at the 5-position. Introduction of fluorine at the 5-posi-
tion, compound 21, maintained potency, whereas dichloro deriva-
1 lM compound concentration). Addition of the Y-substituent at
the 5-position resulted in increased potency (10 and 11). Steric
effects were also found, with a drastic loss in potency seen when
changing from 5-methoxy to 5-ethoxy (11 vs 12).
After exploring the SAR with amide based derivatives, we found
that they possessed poor metabolic stability and pharmacokinetics.
Since we believed that the activity of the compounds was related
to a six-membered hydrogen bond between the amide carbonyl
and the anilinic/phenolic functionality.13,14 we next focused on
making cyclized versions of these derivatives.
The syntheses of the initial ring-constrained analogs were car-
ried out by following one of the four methods shown in
Scheme 2. The compounds were either synthesized by SNAr dis-
placement of an aryl halide, Buchwald–Hartwig aryl amination,
copper(II) acetate mediated coupling of an aryl amine (Chan–Lam
coupling), or by reaction neat of the chloro-indazole with the ani-
line hydrochloride.8
tive 22 showed
a substantial improvement in activity, and
brominated compound 23 resulted in an appreciable increase in
activity compared to compound 19. Methyl derivative 24 was
highly potent, whilst the trifluoromethyl variation 25 resulted in
diminished activity. Methoxy substituted 26 was also very active,
whilst trifluoromethoxy-containing 27 was not.
From these indazole derivatives, four compounds, 19, 20, 21
and 26 were selected to be tested in an in vivo liver enzyme inhi-
bition experiment—delta-5 desaturase is highly expressed in the
mammalian liver.6 Each compound was given orally along with
vehicle control, using a PEG/Solutol (20:80) dosing vehicle, to mice
(n = 4) at a dose of 30 or 60 mg/kg, then 30 min later 14C-di-homo-
gamma-linolenic acid (DGLA) was given using an intraperitoneal
As can be seen from Table 2, a variety of different bicyclic sys-
tems as ring-constraints were well tolerated. Quinoline,
Scheme 1. Reagents and conditions: (a) aniline, EtOH, 100 °C; (b) aniline, PCI3, PhMe, 120 °C. For lists of Y and Z see Table 1.

3838
S. D. P. Baugh et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3836–3839
Br
HN
Cl
a
b
R
Cl, Br
HN
R
Cl
14
12, 13, 15-17
Cl
HN
NH2
c
HN
Cl
d
Cl
N
N
19
18
N
N
H
H
Scheme 2. Reagents and conditions: (a) 3-chloroaniline, i-PrOH, HCL (cat.), 60 °C; (b) 3-chloroaniline, Pd2(dba)3, BINAP, NaOtBu, 120 °C; (c) 3-chlorophenyl boronic acid,
Cu(OAc)2, 2,6-lutidine, PhMe, air, 20 °C; (d) 3-chloroaniline. HCI, 180 °C. For lists of R see Table 2.
Table 2
Table 3
SAR of bicyclic analogs as inhibitors of delta-5 desaturase
SAR of indazole analogs as inhibitors of delta-5 desaturase
Compound
R
IC50
Compound
R
IC50
(nM)
(nM)
Compound
Y
Z
IC50 (nM)
19
20
21
22
23
24
25
26
27
H
H
F
Cl
Cl
F
84
59
58
5
21
6
254
8
>3000
12
13
14
15
18
11
16
17
18
19
50
260
173
84
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Br
CH3
CF3
OCH3
OCF3
17
Table 4
Enzyme inhibition of delta-5 desaturase in the mouse liver by selected compounds
371
Compound IC50
Dose
Plasma
Plasma
% Liver
(nM) (mg/
kg)
concentration
0.5 h (nm)
concentration
0.5 h/IC50
delta-5
desaturase
inhibition
relative to
vehicle
injection.8 Plasma was removed from the mice 0.5 h post com-
pound dose, and was bioanalyzed to determine compound concen-
tration.15 After 2.5 h the livers of the animals were collected, and
the 14C labeled products were evaluated by thin-layer chromatog-
raphy (TLC), using quantitative phosphoimaging to analyze for the
delta-5 desaturase mediated conversion of 14C DGLA to 14C
Arachidonic acid, and comparing those levels to vehicle control.8
The results from these experiments are summarized in Table 4.
The 30 mg/kg doses of compounds 19, 20, 21, and 26 produced a
range of concentrations in the plasma at 0.5 h, and factoring in
the potencies of the compounds, it can be seen that the compound
concentrations are 151–652 fold over the IC50 values. This in turn
manifests itself in the inhibition of delta-5 desaturase in the liver
in the experiments, showing inhibition of the conversion of DGLA
to Arachidonic acid from 18% to 75%. Compound 26 was also dosed
at 60 mg/kg, and at this dose the compound plasma concentration
at 0.5 h is 1753 fold above the IC50 value, which achieves 97%
control
19
20
21
26
26
84
59
58
8
30
30
30
30
60
12,699
9448
14,072
5215
151
160
242
652
1753
42
18
49
75
97
8
14,025
enzyme inhibition-displaying the dose dependent inhibition of
delta-5 desaturase of compound 26. Analysis of the ratio of plasma
concentration for a range of structurally related analogs with
respect to their respective in vitro IC50 values shows a very good
correlation with the in vivo inhibition of delta-5 desaturase in
the liver, Figure 2, showing the strong relationship between phar-
macokinetics and pharmacodynamics in these experiments.
Scheme 3. Reagents and conditions: (a) Ar-NH2. HCI, 180 °C; (b) Ar-B(OH)2, Cu(OAc)2, MeOH, air, 20 °C; (c) Ar-I, Pd(bda)2, BINAP, Cs2CO3, PhME, MeCN, 140 °C. For lists of Y
and Z Table 3.

S. D. P. Baugh et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3836–3839
3839
120
100
80
60
40
20
0
References and notes
1. Moller, D. E. Nature 2001, 414, 821.
2. Grundy, S. M. Nat. Rev. Drug Disc. 2006, 5, 295.
3. Moller, D. E.; Kaufman, K. D. Annu. Rev. Med. 2005, 56, 45.
4. Hughes, S.; York, D. A. Biochem. J. 1985, 225, 307.
5. Jump, D. B.; Clarke, S. D. Ann. Rev. Nutr. 1999, 19, 63.
6. For a review on desaturases, see: Nakamura, M. T.; Takayuki, Y. N. Ann. Rev. Nut.
2004, 24, 345.
7. Zambrowicz, B. P.; Sands, A. T. Nat. Rev. Drug Disc. 2002, 2, 37.
8. Barbosa, J.; Baugh, S. D. P.; Han, Q.; Lombardo, V. K.; Nguyen, H. V.; Pabba, P. K.;
Terranova, K. M. Patent WO2008089310A2.
9. Shimizu, S.; Akimoto, K.; Shinmen, Y.; Kawashima, H.; Sugano, M.; Yamada, H.
Lipids 1991, 26, 512.
0
500
1000
1500
2000
10. Obukowicz, M. G.; Raz, A.; Pyla, P. D.; Rico, J. G.; Wendling, J. M.; Needleman, P.
Biochem. Pharmacol. 1998, 55, 1045.
Concentraꢀon 0.5 hours/ IC50
11. Suzuki, H.; Fujimoto, T.; Yamamoto, T. Patent US2010190747; Matsunaga, N.;
Suzuki, H.; Asano, K.; Tokuhara, H.; Yamamoto, K.; Okamoto, R. Patent
US20130131050.
12. Obukowicz, M. G.; Welsch, D. J.; Salsgiver, W. J.; Martin-Berger, C. L.; Chinn, K.
S.; Duffin, K. L.; Raz, A.; Needleman, P. J. Exp. Ther. 1998, 287, 157.
13. For a review on intra-molecular hydrogen bonding see: Kuhn, B.; Mohr, P.;
Stahl, M. J. Med. Chem. 2010, 53, 2601.
14. Hodge, C. N.; Pierce, J. Bioorg. Med. Chem. Lett. 1993, 3, 1605.
15. The plasma fraction was immediately separated by centrifugation at 4 °C, and
then stored at À20 °C until sample analysis. To quantify compound
concentrations, plasma samples were extracted with acetonitrile–water–
Figure 2. Correlation of pharmacokinetics and pharmacodynamics in the in vivo
experiments.
In summary, we have developed a series of novel inhibitors of
delta-5 desaturase, and four of these have displayed in vivo efficacy
in reducing the enzymatic activity of delta-5 desaturase in the
liver.
formic acid (80:20:0.1) containing verapamil as internal standard at
a
concentration of M. Extracted samples were analysed by LC/MS–MS.
Compound levels were calculated by normalizing the area counts of
compound to the area counts of the internal standard (relative area counts).
1
l
Acknowledgment
The authors would like to acknowledge Theodore C. Jessop and
Kenneth G. Carson for their assistance during the preparation of
the manuscript.