Stimulation of cortical bone formation with thienopyrimidine based inhibitors of Notum Pectinacetylesterase

Article abstract of DOI:10.1016/j.bmcl.2016.02.021

A group of small molecule thienopyrimidine inhibitors of Notum Pectinacetylesterase are described. We explored both 2-((5,6-thieno[2,3-d]pyrimidin-4-yl)thio)acetic acids and 2-((6,7-thieno[3,2-d]pyrimidin-4-yl)thio)acetic acids. In both series, highly potent, orally active Notum Pectinacetylesterase inhibitors were identified.

Full text of DOI:10.1016/j.bmcl.2016.02.021

Bioorganic & Medicinal Chemistry Letters xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Stimulation of cortical bone formation with thienopyrimidine
based inhibitors of Notum Pectinacetylesterase
a
a
a
a
James E. Tarver Jr. a, , Praveen K. Pabba , Joseph Barbosa , Qiang Han , Michael W. Gardyan ,
Robert Brommage b, Andrea Y. Thompson b, James M. Schmidt b, Alan G. E. Wilson b, Wei He a,
Victoria K. Lombardo a, Kenneth G. Carson a
⇑
a Lexicon Pharmaceuticals, 350 Carter Road, Princeton, NJ 08540, United States
b Lexicon Pharmaceuticals, 8800 Technology Forest Place, The Woodlands, TX 77381, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
A group of small molecule thienopyrimidine inhibitors of Notum Pectinacetylesterase are described. We
explored both 2-((5,6-thieno[2,3-d]pyrimidin-4-yl)thio)acetic acids and 2-((6,7-thieno[3,2-d]pyrimidin-
4-yl)thio)acetic acids. In both series, highly potent, orally active Notum Pectinacetylesterase inhibitors
were identified.
Received 13 December 2015
Revised 6 February 2016
Accepted 9 February 2016
Available online xxxx
Ó 2016 Elsevier Ltd. All rights reserved.
Keywords:
Notum Pectinacetylesterase
Osteoporosis
Thienopyrimidine
SAR
Femur cortical bone thickness
Bone is a dynamic tissue continuously remodeled during life
with bone mass depending on the coordinated activities of bone-
forming osteoblasts and bone-resorbing osteoclasts. An imbalance
in bone turnover between these anabolic and catabolic activities
results in postmenopausal, aging-related and glucocorticoid-
induced osteoporosis. Besides good nutrition, including adequate
calcium and vitamin D intakes, current osteoporosis treatments
include estrogens, bisphosphonates, selective estrogen receptor
modulators (raloxifene and bazedoxifene) and the anti-RANKL
antibody denosumab. These anti-resorptive agents all minimize
additional bone loss by inhibiting osteoclastic bone resorption.
Teriparatide, the amino-terminal fragment of parathyroid hor-
mone given by daily subcutaneous injections, is the sole anabolic
osteoporosis drug stimulating osteoblasts to produce new bone.
Potential osteoporosis drugs under late stage clinical development
include the anti-resorptive odanacatib inhibiting cathepsin K, the
teriparatide analog abaloparatide, and anabolic anti-sclerostin
antibodies. Because of the paucity of available anabolic drugs for
osteoporosis treatment, there is an urgent need to develop
orally-active small molecule therapies to treat this disease that
are nontoxic, cost-effective, and easy to administer.1–4
The WNT signaling pathway, involving 19 secreted WNTs, 10
Frizzled membrane receptors, Lrp4/5/6 coreceptors, plus secreted
Dickkopf (DKK) and Secreted Frizzled-Related Protein (SFRP) inhi-
bitors, plays a key role transducing physical activity to new bone
formation5,6 and is a key pathway for drug development.7,8 Lexi-
con’s gene knockout mouse phenotyping campaign9 identified
the gene Notum Pectinacetylesterase as an important contributor
to cortical bone mass and follow-up studies, subsequently con-
firmed by two independent laboratories,10,11 demonstrated that
NOTUM is a WNT-inactivating lipase that removes the palmi-
toleate essential for binding to Frizzled receptors. Inhibiting
NOTUM stimulates bone formation on all endocortical (marrow-
facing) bone surfaces.12–14
Herein we report a group of small molecules, namely thienopy-
rimidines, as potent inhibitors of Notum Pectinacetylesterase. IC50
values were determined by incubation of conditioned media con-
taining mouse or human NOTUM with trisodium 8-octanoy-
loxypyrene-l,3,6-trisulfonate (OPTS),
a water-soluble enzyme
substrate for fluorimetric assays of esterases and lipases. Com-
pound EC50 values were determined using a cell-based TCF/LEF
CellSensorÒ assay as previously described.12–14
We used the OPTS assay as a primary screen and highly potent
compounds in the OPTS assay were selected for additional profiling
in mouse and human cellular assays. It was generally noticed that
high potency in the OPTS assay (IC50 <50 nM) was required to show
⇑
Corresponding author. Tel.: +1 215 898 4193.
E-mail address: james.tarver2@gmail.com (J.E. Tarver).
http://dx.doi.org/10.1016/j.bmcl.2016.02.021
0960-894X/Ó 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Tarver Jr., J. E.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.02.021

2
J. E. Tarver Jr. et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
ides or ketoacids 6. Similarly, commercially available or prepared12
Reduction to alcohol or
CH2 not tolerated
thienopyrimidine-4-ols 7 or other heterocyclic pyrimidines were
converted to thioacetic acids 9 or other corresponding acids.
Our SAR work began by exploring point substitutions at each
position. A logical step was to see if improvements could be made
to the 5,6-di-methyl substitution. Most of these compounds
showed significant losses in potency. As stated above, 2-((5,6-
O
S
OH
No substitution tolerated
Replacement with
N or O not tolerated
9
4
5
3
N
dimethylthieno[2,3-d]pyrimidin-4-yl)thio)acetic acid
OPTS IC50 of 2 nM . The demethylated analog, compound 10,
showed a reduction in potency, OPTS IC50 to 2.8 M. Mono-methyl
1 had an
6
S
2
Substitution reduces activity
N
1
l
compounds substituted at the 5- or 6-position, 11 and 12, had
IC50’s of 92 nM and 152 nM respectively, demonstrating a prefer-
ence for bis-substitution on the thiophene ring. Expanding dialkyl
substitution by incorporating a ring in compound 16 led to a
noticeable loss of activity but suggested that larger substitutions
in this region of the molecule could be tolerated. Replacing the
thiophene with a thiazole led to dramatic loss of activity, com-
pound 13 exhibited a micromolar IC50. Phenyl, pyrrole and pyra-
zole, compounds 15, 17, and 18, also proved poor surrogates for
the thiophene, registering significant losses in potency. Transpos-
ing the thiophene ring in compound 14, however, had similar
activity to its constitutional isomers, compounds 11 and 12.
A survey of the carboxyl terminus based on converting acid 1 to
esters, amides or other heterocycles is summarized in Table 2.
These modifications generally led to modest losses in potency. In
the case of the amide derivatives, we found that they exhibited
poor metabolic stability and pharmacokinetic issues. In contrast,
we observed that the parent carboxylic acids generally had good
ADME profiles, and so we chose to focus our efforts on thioacetic
acids.
Figure 1. Preliminary SAR triage.
potency in cells. Once the OPTS assay potencies reached single-
digit nanomolar range, we used mouse and human cellular assay
EC50 values to further differentiate compounds, referred to as
mEC50 and hEC50 respectively. Based on internal high throughput
screening, we identified 2-((5,6-dimethylthieno[2,3-d]pyrimidin-
4-yl)thio)acetic acid 1 as a small molecule lead (Fig. 1). Compound
1 had OPTS IC50 of 2 nM, mouse EC50 of 1020 nM and human EC50
of 570 nM, which offered a good starting point for our SAR work.
From the lead compound 1, variations at different positions were
explored (Fig. 1). The general observation was that replacement
of the carboxylic acid with esters, heterocycles, sulfonamides and
carboxamides was tolerated. No substitutions on the carbon at
the 9-position were tolerated. Replacement of the thienopyrim-
idine ring with other heterocycles generally gave much lower
potency.
The syntheses of our derivatives were carried out following the
general synthetic methods shown in Scheme 1. The appropriate
commercially available or prepared12 thienopyrimidine-4-ols 2
were treated with phosphorus oxychloride to generate the 4-
chlorothienopyrimidines 3. Subsequent treatment with methyl 2-
mercaptoacetate gave esters 4. Saponification of the esters with
aqueous base yielded carboxylic acids 5, which could be further
functionalized to amides, heterocycles, sulfonamides, carboxam-
Further SAR of 2-((5,6-thieno[2,3-d]pyrimidin-4-yl)thio)acetic
acids is summarized in Table 3. The 5-ethyl-6-methyl compound
29 had a hEC50 of 22 lM. But shifting the positions, 5-methyl-6-
ethyl compound 30 made a major impact on potency, with a
hEC50 of 835 nM. Retaining the 5-methyl and changing the 6-posi-
tion to bromo (31) further improved hEC50 potency to 144 nM,
Scheme 1. Reagents and conditions: (a) POCl3; (b) HSCH2CO2Me, Et3N; (c) NaOH, THF/H2O; (d) HATU, R1NHR2.
Please cite this article in press as: Tarver Jr., J. E.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.02.021

J. E. Tarver Jr. et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
3
Table 1
SAR on thieno[2,3-d]pyrimidin-4-yl)thio)acetic acid ring for Notum Pectinacetylesterase
Table 3
SAR
of
2-((5,6-thieno[2,3-d]pyrimidin-4-yl)thio)acetic
acids
for
Notum
Pectinacetylesterase
O
OH
S
R
Compounds
OPTS IC50
0.002
(lM)
Compounds
OPTS IC50
0.193
(lM)
S
N
N
Compounds
R5
R6
mEC50 (nM)
hEC50 (nM)
S
N
N
1
14
1
H
Et
H
Me
Et
Br
Me
Cl
Me
Et
i-Pr
c-Pr
Pr
1020
65,394
2637
485
375
2561
409
1442
828
1006
342
570
21,928
835
144
208
885
107
393
477
29
30
31
32
33
34
35
36
37
38
Me
Me
Br
Me
Cl
Cl
Cl
Cl
Cl
N
N
2.8
1.2
S
N
N
10
15
N
N
250
88
0.092
0.620
S
S
N
N
16
11
N
these analogs were more potent than the dialkyl congeners. 5-
Methyl-6-chloro compound 33 decreased the potency to 885 nM,
while 5-chloro-6-methyl compound 34 improved hEC50 potency
to 107 nM. We decided next to fix the 5-chloro while varying the
6-alkyl group. This effort led to a 6-ethyl and 6-isopropyl com-
pounds 35 and 36 that increased potency to 393 nM and 477 nM
respectively. However, cyclopropane 37 slightly improved potency
to 250 nM and n-propyl compound 38 continued the trend with
hEC50 of 88 nM.
N
0.152
15
15
15
N
H
S
N
12
N
17
N
N
N
N
S
N
N
N
18
H
13
Although many early modifications of the heterocyclic core
structure were not fruitful (as in Table 1), one of the most tolerated
core changes was the transposition of the thiophene ring, such as
compound 14. Our next approach was to explore the change in
while 5-bromo-6-methyl isomer 32 had a hEC50 of 208 nM. Com-
pound 32 suggested that electronics were important for increasing
activity compared with the hydrophobic nature of bromine, as both
Table 2
SAR of the carboxyl group for thieno[2,3-d]pyrimidin-4-yl)thio)acetic acid
O
R
S
N
N
S
Compounds
R
OPTS IC50 (nM)
2
hEC50 (nM)
570
Compounds
R
OPTS IC50 (nM)
hEC50 (nM)
1
–OH
O
O
N
19
–NHCN
2
400
25
10
280
NH
20
21
–NHOH
–OMe
2
2
556
445
O
HN
26
27
5
2
249
168
O
O
S
N
22
–NHMe
27
789
N
23
24
–NH2
2
857
O
HN
–NHSO2Me
11
1462
28
6
211
O
Please cite this article in press as: Tarver Jr., J. E.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.02.021

4
J. E. Tarver Jr. et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
Table 4
somewhat alter the electronics of the system. The SAR is summa-
rized in Table 4. With R6 = Me, R7 = Cl, compound 39 had a hEC50
of close to 2 lM. By exchanging the R6 and R7 substituents, the
SAR
of
2-((6,7-thieno[3,2-d]pyrimidin-4-yl)thio)acetic
acids
for
Notum
Pectinacetylesterase
potency of 40 improved to 435 nM. Fixing R7 as the methyl and
changing R6 to bromo gave hEC50 of 417 nM, similar to chloro com-
pound 40, and changing R6 to methyl 42 reduced the potency to
881 nM. When we selected R6 = Me and introduced cyclopropyl
at R7 in compound 43, we obtained a major boost in potency to
115 nM. Keeping R7 as cyclopropyl, and modifying R6 substitution
produced compound 44 (R6 = Cl), hEC50 of 21 nM, and compound
45 (R6 = CF3), hEC50 of 12 nM.
As shown in Table 5, when dosed orally to mice at 10 mg/kg,
compounds 34 and 44 were both able to achieve high Cmax and
AUC while having low clearance and volume of distribution. Com-
pound 34 was advanced into in vivo mouse models of bone growth
and was found to increase cortical bone thickness.12
The in vivo effect of 2-((6-chloro-7-cyclopropylthieno[3,2-d]
pyrimidin-4-yl)thio)acetic acid 44 was determined by treating F1
male hybrid (129xC57) mice for 25 days with the compound, start-
ing at 8.7 weeks of age. The compound was administered by daily
oral gavage (vehicle = 0.1% Tween 80 in water). Four groups of mice
(N = 13) were used: control, 3 mg/kg compound, 10 mg/kg com-
pound, and 30 mg/kg compound. Midshaft femur cortical thickness
Compounds
R6
R7
mEC50 (nM)
hEC50 (nM)
39
40
41
42
43
44
45
Me
Cl
Br
Me
Me
Cl
Cl
5508
1444
1360
4468
408
1965
435
417
881
115
21
Me
Me
Me
c-Pr
c-Pr
c-Pr
55
43
CF3
12
Table 5
Mouse pharmacokinetic data for compounds 34 and 44
Compounds
Cmax
10 mg/kg po
(
l
M) at
AUC (
at 10 mg/kg po
l
M * h)
Clearance
(mL/min/kg)
Vd (L/
kg)
%F
was measured by microCT (Scanco
lCT40). As shown in Figure 2,
an increase in cortical bone thickness was observed at all doses
compared to control: 7% (p = 0.003) at 3 mg/kg; 10% (p <0.001) at
10 mg/kg; and 13% (p <0.001) at 30 mg/kg.
34
44
157
129
515
1533
1.22
0.49
0.16
0.13
95
65
In summary, we have developed novel inhibitors of Notum
Pectinacetylesterase showing in vivo efficacy in significantly
increasing midshaft femur cortical bone thickness in mice and rats.
Mids haft Femur C ortical Thicknes s (μm)
ANOVA P < 0.001
References and notes
300
280
260
240
220
1. Allen, J. G.; Fotsch, C.; Babij, P. J. Med. Chem. 2010, 53, 4332.
2. Guo, H.; Shao, H.; Yang, Z.; Xue, S.; Li, X.; Liu, Z.; He, X.; Jiang, J.; Zhang, Y.; Si, S.;
Li, Z. J. Med. Chem. 2010, 53, 1819.
3. Makras, P.; Delaroudis, S.; Anastasilakis, A. D. Metabolism 2015, 64, 1199.
4. Cairoli, E.; Zhukouskaya, V. V.; Eller-Vainicher, C.; Chiodini, I. J. Endocrinol.
Invest. 2015, 38, 303.
5. Baron, R.; Kneissel, M. Nat. Med. 2013, 19, 179.
6. Maupin, K. A.; Droscha, C. J.; Williams, B. O. Bone Res. 2013, 1, 27.
7. Kahn, M. Nat. Rev. Drug Disc. 2014, 13, 513.
Δ = 13%
P < 0.001
Δ = 10%
P < 0.001
Δ = 7%
P = 0.003
8. Brommage, R. J. Cell. Biochem. 2015, 116, 2139.
9. Brommage, R.; Liu, J.; Hansen, G. M.; Kirkpatrick, L. L.; Potter, D. G.; Sands, A. T.;
Zambrowicz, B.; Powell, D. R.; Vogel, P. Bone Res. 2014, 2, 14034.
10. Kakugawa, S.; Langton, P. F.; Zebisch, M.; Howell, S. A.; Chang, T. H.; Liu, Y.;
Feizi, T.; Bineva, G.; O’Reilly, N.; Snijders, A. P.; Jones, E. Y.; Vincent, J. P. Nature
2015, 519, 187.
11. Zhang, X.; Cheong, S. M.; Amado, N. G.; Reis, A. H.; MacDonald, B. T.; Zebisch,
M.; Jones, E. Y.; Abreu, J. G.; He, X. Dev. Cell 2015, 32, 719.
12. Barbosa, J.; Carson, K. G.; Gardyan, M. W.; He, W.; Lombardo, V.; Pabba, P.;
Tarver, J. Jr., U.S. Patent 65, 200, 2012.
Vehicle
3 mg/kg
10 mg/kg 30 mg/kg
Figure 2. Change in cortical bone thickness after treatment for 25 days.
13. Brommage, R.; Feng, X.; Hong, S.; Landes, G.; Liu, J.; Potter, D. G.; Powell, D. R.
U.S. Patent 302,346, 2013.
14. Han, Q.; Pabba, P. K.; Barbosa, J.; Mabon, R.; Healy, J. P.; Gardyan, M. W.;
Terranova, K. M.; Brommage, R.; Thompson, A. Y.; Schmidt, J. M.; Wilson, A. G.
E.; Xu, X.; Tarver, J. E., Jr.; Carson, K. G. Bioorg. Med. Chem. Lett. 2016, 26, 1184.
thiophene orientation to give 2-((6,7-thieno[3,2-d]pyrimidin-4-yl)
thio)acetic acids. In this manner, we hoped to preserve the geom-
etry of the core and the molecular arrangement of substituents but
Please cite this article in press as: Tarver Jr., J. E.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.02.021