Isosteric replacements for benzothiazoles and optimisation to potent Cathepsin K inhibitors free from hERG channel inhibition

Article abstract of DOI:10.1016/j.bmcl.2012.07.012

The discovery of nitrile compound 4, a potent inhibitor of Cathepsin K (Cat K) with good bioavailability in dog is described. The compound was used to demonstrate target engagement and inhibition of Cat K in an in vivo dog PD model. The margin to hERG ion channel inhibition was deemed too low for a clinical candidate and an optimisation program to find isosteres or substitutions on benzothiazole group led to the discovery of 20, 24 and 27; all three free from hERG inhibition.

Full text of DOI:10.1016/j.bmcl.2012.07.012

Bioorganic & Medicinal Chemistry Letters 22 (2012) 5563–5568
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Isosteric replacements for benzothiazoles and optimisation to potent
Cathepsin K inhibitors free from hERG channel inhibition
Alexander G. Dossetter a, , Jonathan Bowyer a, Calum R. Cook a, James J. Crawford a, Jonathan E. Finlayson a,
⇑
Nicola M. Heron a, Christine Heyes a, Adrian J. Highton a, Julian A. Hudson a, Anja Jestel b, Stephan Krapp b,
Philip A. MacFaul a, Thomas M. McGuire a, Andrew D. Morley a, Jeffrey J. Morris a, Ken M. Page a,
Lyn Rosenbrier Ribeiro a, Helen Sawney a, Stefan Steinbacher b, Caroline Smith a
a AstraZeneca R&D, Mereside, Alderley Park, Macclesfield, Cheshire, SK10 4TG, UK
b Proteros Biostructures, Am Klopferspitz 19, D-82152 Martinsried, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
The discovery of nitrile compound 4, a potent inhibitor of Cathepsin K (Cat K) with good bioavailability in
dog is described. The compound was used to demonstrate target engagement and inhibition of Cat K in an
in vivo dog PD model. The margin to hERG ion channel inhibition was deemed too low for a clinical can-
didate and an optimisation program to find isosteres or substitutions on benzothiazole group led to the
discovery of 20, 24 and 27; all three free from hERG inhibition.
Received 28 May 2012
Revised 29 June 2012
Accepted 4 July 2012
Available online 15 July 2012
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
Cathepsin K
Covalent inhibitor
Molecular complexity
Structure based design
hERG
Pharmacokinetics
Matched molecular pair analysis
MMPA
Synthesis
The lysosomal cysteine protease Cathepsin K (Cat K) is highly
expressed in osteoclasts and plays a key role in bone resorption
by the degradation of type I cartilage.1 Cat K knockout mice exhibit
osteopetrosis (abnormally dense bone) and abnormal joint mor-
phology2 indicating the potential role of this enzyme in key bone
pathologies such as osteoporosis (OP), osteoarthritis (OA), and
metastatic bone disease (MBD).3–5 Osteoarthritis is a group of
degenerative disorders characterised by joint pain, and loss of
function in the absence of chronic autoimmune or autoinflamma-
tory mechanisms.6 They have an increasing prevalence with age
and are thus a growing socioeconomic burden.2 Articular cartilage
breakdown is a prominent feature of joint degeneration and, as a
result, has received significant attention from the pharmaceutical
industry. Several companies have trialed compounds in the OP
and OA disease areas. Novartis has completed Phase II studies in
OP and OA with Balicatib (1).7 The Phase II OP trial reported
positive outcomes on bone mineral density measures. Merck has
completed a Phase II trial in OP with Odanacatib (2), is recruiting
for Phase III in MBD and has abandoned plans for a Phase II study
in OA.8–10 Both compounds are electrophilic nitrile-containing
compounds that bind covalently and reversibly to the Cat K en-
zyme preventing type I collagen degradation.
O
O
N
N
H
N
N
H
N
N
O
S
O
F
N
N
F
O
F F
1 Balicatib
2 Odanacatib
O O
N
O
O
S
N
F
N
N
N
N
N
N
N
H
H
3
4
⇑
Corresponding author. Tel.: +44 (0) 1625 517673; fax: +44 (0) 1625 516667.
Figure 1. Nitrile Cathepsin K inhibitors.
E-mail
addresses:
al.dossetter@astrazeneca.com,
al@dossetter.plus.com
(A.G. Dossetter).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.07.012

5564
A. G. Dossetter et al. / Bioorg. Med. Chem. Lett. 22 (2012) 5563–5568
Table 1
SAR for thiazole variations
O O
S
N
O
O
N
S
N
N
N
S
N
S
N
S
N
O
N
Compound
Cat K pIC50
hERG pIC50
4
8.0
5.1
20
7.6
<4.0
24
7.8
<4.0
27
7.6
<4.0
F
S
N
S
N
S
N
S
N
S
N
S
N
S
F
N
F
F
S
F
F
N
F
Cl
S
F
F
N
F
S
N
N
S
N
O
O
S
S
N
S
N
N
aBinding affinity Cat K, L, S, B versus FRET substrate, mean of greater than n = 4 tests, unless where stated.
bMean of n = 2 tests.
cIn vitro human liver microsomal turnover mean of at least n = 2 tests (
l
L/min/mg).
dIn vitro rat liver hepatic clearance (RHC) turnover mean of at least n = 2 tests ( L/min/106 cells).
l

A. G. Dossetter et al. / Bioorg. Med. Chem. Lett. 22 (2012) 5563–5568
5565
Electrophilic nitrile-containing cysteine protease inhibitors,
such as 1 and 2 (Fig. 1), have been observed to interact with Cat
K by reversible formation of a covalent bond between the nitrile
and an active site cysteine (Cys-25). Screening of nitrile containing
compounds from the corporate collection identified (R,R)-1,2-
cyclohexyl-diamide-acetonitriles as potent and efficient inhibitors
of the enzyme. We developed this lead series with an initial fo-
cussed library of rigid tertiary amides to bring groups into contact
with the S3 groove,11–13 and identified piperazines, such as 3 and 4,
as potential starting points for optimisation. Both 3 and 4 had good
Cat K inhibition, physical properties and stability to in vitro human
liver microsomes (Table 1).13
Studies in rat pharmacokinetics experiments showed 4-fluoro-
phenylpiperazine 3 was the inferior of the two compounds when
no appreciable compounds levels were detected after oral dosing
compared to benzothiazole-piperazine 4 with bio-availability of
20%. It was unclear why 3 had poor PK as the compound had good
rodent in vitro stability and aqueous solubility. The in vitro rat he-
patic stability for 4 was significantly higher and led us to suspect
an absorption issue with 3 (Table 1). An excellent pharmacokinetic
profile was found when 4 was dosed orally and intravenously to
dog: bio-availability of 78%, CL 13.0 mL/min/kg from doses of
1.0 mg/kg. The blood samples from this PK study were analysed
further by measuring the concentration of C-telopeptide (CTX-I),
the degradation product of the action of Cat K on type I colla-
gen.14–16 As blood concentrations of 4 rose after oral dosing to
dog, a rapid decrease in CTX-I levels was observed as Cat K enzyme
activity was abated as expected (Fig. 2). A large effect was observed
in both dogs, demonstrating target engagement and inhibition of
enzyme activity. As blood levels of 4 diminished, so enzyme activ-
ity returned, as shown by rising CTX-I levels over the second 12-
hour period of the experiment. Compound 4 was clearly a good
quality tool compound suitable for further pre-clinical disease
model work, which will be published in due course.
Compound 4 was prioritised for further optimisation due to the
in vivo profile above and good selectivity with respect to other
cathepsins; at worst, 80-fold against Cat S. The absence of a basic
centre in the molecule was also seen as important because basicity
is associated with compound accumulation in lysosomes (lysoso-
motropism), which is considered a potential cause of toxicity.17,18
However, the compound was not suitable for further clinical devel-
opment as inhibition of the hERG ion channel, measured at pIC50
5.1 (9.0 lM), needed improvement to provide a better safety mar-
gin.19–21 With this in mind a set of thiazole piperazines (10–19)
was designed as an initial training set to cover a range of lipophi-
licities and reduce the molecular complexity of the compounds so
we could incorporate structure based design, using the results we
obtained.
Thiazole piperazines were synthesised in one of two basic
methods, either displacement of leaving group from the 2—posi-
tion of the substituted thaizole or employing the classic Hantzsch
reaction (Scheme 1). Nucleophilic aromatic substitution of a
2-chloro, 2-methylsuphide, 2-methylsulphone on thiazole 5 with
carbamate protected piperazine 6 was high yielding but required
heating at high temperatures. Although diversity was introduced
early on, condensation of an a-haloketone (7) with carbamate pro-
tected piperazine thio-urea 8 was a short and convenient route.
Following either route, simple deprotection with acid yielded the
thiazole-piperazines (10a–28a) ready for coupling. Final com-
pounds were synthesised by ring opening of chiral anhydride
9—(R,R) at ambient temperature with the piperazines (10a–28a)
to yield an intermediate acid. Final compounds (10–28) were then
obtained by amide formation with 1,1-amino-cyclopropylnitrile
and
a suitable coupling reagent O-(7-azabenzotriazol-1-yl)-
N,N,N0,N0-tetramethyluronium hexafluorophosphate (HATU) and
di-iso-propylethylamine (DIPEA) in yields between 7% and 78%.22
This method proved to be an excellent ’one-pot’ convergent meth-
od and amenable to scale-up on multigram amounts; for example,
4.8 g of 4 was synthesised in an overall yield of 62%.
Table 1 details the results of the initial set of thiazole Cat K
inhibitors compared to 4. The simple thiazole 10 showed that the
benzothiazole ring of 4 does not add significant Cat K potency (or
indeed change selectivity), however activity was diminished along
with inhibition at the hERG channel. The drop in lipophilicity, in
line with prediction, resulted in thiazole 10 achieving a high point
in ligand lipophilicity efficiency (LLE) for this series at 6.3. Com-
pounds 11 through 17 steadily increase in molecular size and lipo-
philicity, which as a set made very little difference in Cat K
inhibition. The more lipophilic examples were less stable to
in vitro microsomal and hepatic incubation experiments (e.g.,
Figure 2. Dog PK/PD experiment for 4—the blood samples taken at a given time
point after dosing were split and used to measure [4] and [CTX-1]. s—PK Mean [4] %
ng/mL at 1.0 mg/kg uid to both dogs (406 and 476); }—[CTX-1] as % baseline
1.0 mg/kg uid Dog 406; +—[CTX-1] as % baseline 1.0 mg/kg for Dog 476. Dog
enzyme Cat K inhibition for 4 was measured at 0.018 0.010
bound thus 1ꢀ free [4] was 49.2 27.4 ng/mL.
l
M and dog PPB 83%
R1
S
a
X
N
N BOC
R2
+
O
N
O
O
9 - (R,R)
X = Cl, S-CH3, SO2-CH3
5
6
R1 [Cl,Br]
R1
R1
R2
c
b
S
N
S
O
S
O
N
N BOC
N H
N
+
R2
N
N
N
H
R2
O
N
N
N
N
H2N
7
8
10a - 28a
10 - 28
Scheme 1. Synthesis of cyclohexyl-1,2-diamides 10–27. Reagents and conditions: (a) (i) EtOH, 50 °C, 1 h, concentrate, (ii) 5 equiv 1.0 M HCl in dioxane, rt, 1 h or TFA, CH2Cl2,
overall 15–52% yield; (b) (i) DIPEA, DMF, toluene, xylene, 60–140 °C, 1–5 h, (ii) 5 equiv 1.0 M HCl in dioxane, rt, 1 h or TFA, CH2Cl2, Overall 25–68% yield; (c) (i) 9—(R,R),
CH2Cl2, rt, 1 h, concentrate, (ii) DMF or CH2Cl2, DIPEA (5 equiv), HATU (1.1 equiv), NCC(CH2CH2)NH3,Cl (1.0 equiv), rt, 12 h. 12–78% yield.

5566
A. G. Dossetter et al. / Bioorg. Med. Chem. Lett. 22 (2012) 5563–5568
between a glycine rich loop and Tyr-26 such that only the ‘edge’
A
was in contact with the protein surface (Fig. 3—B). This projected
the 4-fluorophenyl in the same plane as the piperazine so that two
aromatic C–H groups were pointing towards the protein surface.
Therefore only half the piperazine/aromatic surface atoms made
energetically favourable interactions with the protein. Virtual sub-
stitution of the 4-fluorophenyl of 3 with benzothiazole (4) and
molecular minimisation logically placed this aromatic in the same
plane as the piperazine, to maintain electronic interaction, such that
sulphurcontacts the proteinsurface in preference to nitrogen, which
pointed towards solvent (Fig. 3—B). This result in the benzenoid
moiety placed further down the groove between Asp-61 and Glu-
59 towards a more polar area and open area. Applying knowledge-
based design using matched molecular pair analysis we chose to
make the methylsulfonyl substituent at C6, which has been shown
to improve the physicochemical properties of compounds29 and hu-
man liver microsomal stability30–32, in the hope this group would be
tolerated or even yield a favourable interaction. This produced 20,
with reduced hERG binding, increased in vitro stability, but a three-
fold drop in Cat K inhibition, showing we were incorrect. The orien-
tation of the thiazole (sulphur towards protein) was consistent with
our experimental observations, in that alkyl substituents from the
C4 position of thiazole (12–16) would orientate away from any
favourable Van-de-Waals interactions with the protein, so explain-
ing the fairly consistent level of Cat K inhibition. Therefore combin-
ing the knowledge gained we refined our tactics to focus on (i)
substitution at the C5 position, maintaining LogD7.4 equal or below
2.0, or (ii) finding a non aromatic isostere for benzothiazole that
could pick up interactions with Asp-61 or Glu-59, to increase Cat K
inhibition and eliminate hERG binding.
Asp-61
S3
S2
Tyr-26
Glu-59
B
Asp-61
The first of these approaches, substitution at C5 position, proved
disappointing with the best example represented by 21 (Table 1),
where the C5—isobutyronitrile group gave a compound with
LogD7.4 = 1.9 and hERG inhibition pIC50 <4.0, but with inhibition
of Cat K reduced to pIC50 7.4 (fourfold less). Focusing more on
Tyr-26
´
the second approach, we found that Walczynski et al., as part of
a program towards optimised H3-antagonists, had made all four
aza-benzothiazoles to reduce lipophilicity and find a polar interac-
tion.33 The routes outlined in their paper fitted our synthetic
approach to Cat K inhibitors and all the four aza-benzothiazoles
were made and profiled (Table 2). Using the virtual structure of
benzothiazole 4 we reasoned upfront that 7-azabenzothioazole
22 would be less potent as the nitrogen lone pair was likely to
point towards a non polar surface region. Despite this we still
chose to make 22 for further medicinal chemistry understanding
and as predicted it was less active but pleasingly free from hERG
liability and remarkably stable in in vitro incubations, which is
noteworthy. All other three isosteres were below pIC50 of 4.0 for
hERG, and in a similar range for Cat K inhibition. The best example,
5-azabenzothiazole 24, had similar Cat K inhibition to 4, as well as
good physical properties and in vitro stability making this an
attractive isostere. The 5-aza nitrogen lone pair would, we predict,
point away from protein surface and towards either solvent or
Asp-61, which fits with the inhibition result. By the same rationale,
we expected 4-aza compound 25 to show an increase in inhibition
by better alignment with Asp-61. To our surprise a more significant
drop was observed, which we cannot rationalise at this point. Fi-
nally for three isomers we measured cytochrome P450 inhibition
potential using isolated enzymes (3A4, 2D6, 2C9, 2C19 and 1A2).
4 and 25 are in-active against all five isoforms, whereas 23 and
24 were inhibitors of three out the five so ruling these out as suit-
able compounds (see Supplementary data).
Glu-59
Figure 3. Compound 3 (green) bound to Cathepsin K (PDB 4DMX) and with 4 (pink)
overlaid by atom substitution. Structures shown from two aspects, where
illustrate the open nature of the S3 groove.
B
14), although some stability was regained with fluorine substitu-
tion (15–17). A near iso-lipophilic compound 15, compared to 4,
did not have the same level of hERG inhibition (3.2-fold) suggest-
ing structural elements and not just lipophilicity were controlling
this. This observation was supported by examples 16–18, where
LogD7.4 was higher than 4, but hERG inhibition was lessened. These
observations are consistent with molecular models for this ion
channel, where either charged alkyl amino groups or aromatics
bind in a ring of phenyl anilines and additional lipophilicity in-
creases affinity.23–26 These data indicated we should aim to keep
lipophilicity down to LogD7.4 of 2.0 or less and avoid benzothiaz-
oles (The data indicated a weak correlation between hERG inhibi-
tion and LogD7.4—see Supplementary data).
X-ray structure determination of a co-crystal of 3 with Cat K re-
vealed the cyclohexane ring was occupying the relatively shallow
S2 pocket of the enzyme27,28 which placed the tertiary amide, and
thus the piperazine, in the comparatively open S3 pocket with the
carbonyl oxygen exposed to solvent (Fig. 3—A). The protein molecu-
lar surface in the S3 pocket was relatively flat and lacked concave po-
lar regions that could be exploited. The piperazine group was
Table 3 shows the comparison data for partially saturated fused
bicyclic rings which allow the introduction of an oxygen atom to
help control lipophilicity or indeed to introduce lipophilicity
capacity so that additional alkyl groups can be added to increase

A. G. Dossetter et al. / Bioorg. Med. Chem. Lett. 22 (2012) 5563–5568
5567
Table 2
SAR for aza-benzothiazoles
N
O O
R N
N
N
H
Compd R—group
Cat K
pIC50
Cat L
pIC50
Cat S
pIC50
Cat B
pIC50
LogD7.4 Mol Wt.
hERG inhib. Aqueous solubility
HLMc
min/mg)
(
l
L/
RHCd
(
l
L/min/
a
a
a
a
(Da)
pIC50
pH 7.4 (
l
M)
106 cells)
S
4
8.0
<5.1
6.2
5.8
2.6
—
437.6
438.6
5.1
1000
<2.0
64
N
N
S
22
6.5b
<4.0b
<5.0b
4.6b
<4.0
—
<2.0
—
N
S
N
N
23
24
25
7.5
7.8
7.3
<4.0
<4.0
<4.0
<5.7
<5.6
5.8
5.4
5.5
5.4
1.6
1.7
1.4
438.6
438.6
438.6
<4.0
<4.0
<4.0
>1600
>1500
>2000
11
<2.0
5.2
N
S
4.2
<2.0
N
S
<2.0
N
N
a
b
c
Binding affinity Cat K, L, S, B versus FRET substrate, mean of greater than n = 4 tests, unless where stated.
Mean of n = 2 tests.
In vitro rat liver microsomal turnover mean of at least n = 2 tests (
l
L/min/mg).
d
In vitro rat liver hepatic clearance (RHC) turnover mean of at least n = 2 tests (
l
L/min/106 cells).
Table 3
SAR for partially saturated benzothiazoles
N
O O
R N
N
N
H
Compd R—group
Cat K
pIC50
Cat L
pIC50
Cat S
pIC50
Cat B
pIC50
LogD7.4 Mol Wt.
hERG inhib. Aqueous solubility
HLMc
min/mg)
(
l
L/
RHCd
(
l
L/min/
106 cells)
a
a
a
a
(Da)
pIC50
pH 7.4 (lM)
S
N
4
8.0
<5.1
6.2
5.8
2.6
437.6
427.6
443.6
5.1
1000
<2.0
64
S
26
7.9b
7.6
<4.3b
<4.0
5.5b
5.6
5.7b
5.5
>3.0
1.4
—
—
11
160
N
S
O
O
27
28
<4.0
>3600
<2.4
<3.0
N
S
7.9
<4.0
<5.6
5.6
2.4
471.6
<4.0
1700
33
23
N
a
b
c
Binding affinity Cat K, L, S, B versus FRET substrate, mean of greater than n = 4 tests, unless where stated.
Mean of n = 2 tests.
In vitro rat liver microsomal turnover mean of at least n = 2 tests (
l
L/min/mg).
d
In vitro rat liver hepatic clearance (RHC) turnover mean of at least n = 2 tests (
l
L/min/106 cells).
inhibition by either protein contact or hydrophobicity. Firstly the
lipophilic compound 26 did not increase potency, showing that
hydrophobicity alone was not a suitable approach to increasing
inhibition. This was not surprising given the openness and expo-
sure of the S3 groove to solvent. However, the 6,7-dihydro-4H-pyr-
ano[4,3-d][1,3]thiazol-2-yl group of 27 proved to be a good
isostere, with LLE of 6.1 driven by Cat K inhibition of pIC50 = 7.6
for a LogD7.4 of 1.4, but with the benefit of greatly improved
in vitro stability. Encouraged by this we designed 28, using the
combined SAR of 27 with 14, the most active C4-alkyl, yielding
an additive effect on Cat K inhibition and the desired increase in
hERG margin (9000-fold). Sadly human in vitro stability was in-
creased at least 10-fold compared to 4 and 27, although rat hepatic
clearance was improved relative to 4, otherwise this would have
been the desired isostere.
Table 4 summarises rodent in vivo and in vitro pharmacokinet-
ics for 20, 24 and 27, compared to the lead 4. All compounds are
comparable with respect to protein binding in rodent and human

5568
A. G. Dossetter et al. / Bioorg. Med. Chem. Lett. 22 (2012) 5563–5568
Table 4
Serum protein binding and pharmacokinetic data
Compd
Species
Protein bindinga (% free)
In vitro hepatocyte clintb
(l
L/min/106 cells)
Clp (mL/min/kg)
Vdss (L/kg)
iv half life (h)
Bioavailability (%)
4
Ratc
17
16
17
43
42
64
35
53
49
64 1.3
5.6 0.2
<2.0 (n = 4)
7.1 6.5
<2.0 (n = 2)
5.2 (n = 1)
<3.0 (n = 2)
<3.0 (n = 4)
<2.0 (n = 5)
47 5.1
13 4.9
1.5 0.3
1.5 0.5
0.8 0.2
1.4 0.2
20 9.0
78 21
Dogd
Human
20
24
27
Ratc
23e
0.54
1.4
3.3e
Human
Ratc
19e
0.8
0.63
17e
Human
Ratc
32 8.1
1.3 0.3
2.0 0.3
33 11
Human
a
b
c
All results are a mean at least n = 2 experiments.
(ꢁ)—number of repeat experiments were result was below limit of detection.
Alderley Park Han Wistar male rats—dosed to fed animals, po 2.0 mg/kg as a suspension in 5% DMSO/95% HPMC/Tween, iv dosed as solution in 40% DMA/Water at 2.0 mg/
kg.
d
Beagle dog—dosed to fasted animals, po 1.0 mg/kg as a suspension in 5% DMSO/95% HPMC/Tween, iv dosed as solution in 10% DMSO/90% Sorensons at 1.0 mg/kg. All
pharmacokinetic experiments reported were single dosed compounds in at least two experiments involving two animals in each, unless otherwise stated.
e
From a single experiment with two animals.
5. Stoch, S.; Wagner, J. Clin. Pharmacol. Ther. (NY, U. S.) 2008, 83, 172.
6. Felson, D. T. N. Engl. J. Med. 2006, 354, 841.
7. Vasiljeva, O.; Reinheckel, T.; Peters, C.; Turk, D.; Turk, V.; Turk, B. Curr. Pharm.
plasma, and are highly unbound. Scaling in vitro rat hepatic clear-
ance to whole liver predicted the measured in vivo clearance of 4
but under predicts for the rest of the compounds. We suspect that
clearance was higher for these compounds due to renal elimina-
tion. Disappointingly, sulphone 20 has low bio-availability, which
given the low clearance and polarity of the compound must be
due to poor passive absorption (not indicated by CACO2 or MDCK
assays). All four compounds had low human hepatic in vitro clear-
ance suggesting HLM was predicting well for the series. The 5-aza-
benzothaizole 24 was orally bio-available and similar to 4 with
lower clearance and overall represented the best isostere found
from these studies. Finally 4 and 27 were tested in a low
through-put in vitro primary human cell osteoclast ‘pit’ resorption
assay and were found to have similar potency of pIC50 7.2 and 6.9,
respectively. This technically demanding assay measured the abil-
ity of compounds to stop the osteoclast cells dissolving cartilage
and bone, thereby stopping the observed formation of ‘pits’ in
the bone sample (see Supplementary data). Results were only con-
sidered qualitatively as no relationship was found to PD end points
in the dog CTX-I model (data not shown).
Des. 2007, 13, 387.
8. Gauthier, J. Y.; Chauret, N.; Cromlish, W.; Desmarais, S.; Duong, L. T.;
Falgueyret, J.; Kimmel, D. B.; Lamontagne, S.; Leger, S.; LeRiche, T.; Li, C. S.;
Masse, F.; McKay, D. J.; Nicoll-Griffith, D. A.; Oballa, R. M.; Palmer, J. T.; Percival,
M. D.; Riendeau, D.; Robichaud, J.; Rodan, G. A.; Rodan, S. B.; Seto, C.; Therien,
M.; Truong, V.; Venuti, M. C.; Wesolowski, G.; Young, R. N.; Zamboni, R.; Black,
W. C. Bioorg. Med. Chem. Lett. 2008, 18, 923.
9. Eisman, J. A.; Bone, H. G.; Hosking, D. J.; McClung, M. R.; Reid, I. R.; Rizzoli, R.;
Resch, H.; Verbruggen, N.; Hustad, C. M.; Da, S.; Carolyn; Petrovic, R.; Santora,
A. C.; Ince, B. A.; Lombardi, A. J. Bone Miner. Res. 2011, 26, 242.
10. Stoch, S. A.; Zajic, S.; Stone, J.; Miller, D. L.; Van, D. K.; Gutierrez, M. J.; De, D. M.;
Liu, L.; Liu, Q.; Scott, B. B.; Panebianco, D.; Jin, B.; Duong, L. T.; Gottesdiener, K.;
Wagner, J. A. Clin. Pharmacol. Ther. (N. Y., NY, U. S.) 2009, 86, 175.
11. Dossetter, A. G.; Howard Beeley, H.; Bowyer, J.; Cook, C. R.; Crawford, J. J.;
Finlayson, J. E.; Heron, N. M.; Christine Heyes, C.; Highton, A. J.; Hudson, J. A.;
Jestelb, A.; Kenny, P. W.; Krappb, S.; Martin, S.; MacFaul, P. A.; McGuire, T. M.;
Morentin-Gutierrez, P.; Morley, A. D.; Morris, J. J.; Page, K. M.; Rosenbrier-
Ribeiro, L.; Sawney, H.; Steinbacherb, S.; Smith, C.; Vickers, M. J. Med. Chem.
2012, 55, in press. http://dx.doi.org/10.1021/jm3007257.
12. Dossetter, A. G.; Heron, N. M. PCT Int. Appl. 2008, 98.
13. Dossetter, A. G.; Heron, N. M. PCT Int. Appl. 2008, 66.
14. Garnero, P.; Ferreras, M.; Karsdal, M.; Nicamhlaoibh, R.; Risteli, J.; Borel, O.;
Qvist, P.; Delmas, P. D.; Foged, N. T.; Delaisse, J. M. J. Bone Miner. Res. 2003, 18,
859.
In conclusion our studies of benzothiazole 4 found this potent
inhibitor of Cat K had good dog PK and as such was used to study
in vivo target engagement and pharmacodynamic effects success-
fully. We strove to improve on this start point by increasing the
margin to hERG binding by the combined tactics of reducing lipo-
philicity and structure based design. Whilst clear improvements
were made, we failed to maintain or increase Cat K inhibition, or
indeed selectivity for other cathepsins, which remained fairly flat
across the compounds in this study. However substituted benzo-
thiazole 20, and isosteres 21, 24 and 27 were all within 4 fold of
Cat K inhibition, below the assay limit for hERG binding and main-
tained similar rodent PK properties when compared to 4.
15. Herrmann, M.; Seibel, M. Clin. Chim. Acta 2008, 393, 57.
16. Okuno, S.; Inaba, M.; Kitatani, K.; Ishimura, E.; Yamakawa, T.; Nishizawa, Y.
Osteoporosis Int. 2005, 16, 501.
17. Black, W. C.; Percival, M. D. ChemBioChem 2006, 7, 1525.
18. Falgueyret, J.; Desmarais, S.; Oballa, R.; Black, W. C.; Cromlish, W.; Khougaz, K.;
Lamontagne, S.; Masse, F.; Riendeau, D.; Toulmond, S.; Percival, M. D. J. Med.
Chem. 2005, 48, 7535.
19. De, P.; Fabrizio; Poluzzi, E.; Cavalli, A.; Recanatini, M.; Montanaro, N. Drug Saf.
2002, 25, 263.
20. De, B. M. L.; Pettersson, M.; Meyboom, R. H. B.; Hoes, A. W.; Leufkens, H. G. M.
Eur. Heart J. 2005, 26, 590.
21. Jamieson, C.; Moir, E. M.; Rankovic, Z.; Wishart, G. J. Med. Chem. 2006, 49, 5029.
22. Crawford, J. J.; Dossetter, A. G.; Finlayson, J. E.; Heron, N. M. PCT Int. Appl. 2008,
295.
23. Waring, M. J.; Johnstone, C. Bioorg. Med. Chem. Lett. 2007, 17, 1759.
24. Aronov, A. M. Drug Discovery Today 2005, 10, 149.
25. Cavalli, A.; Poluzzi, E.; De, P.; Fabrizio; Recanatini, M. J. Med. Chem. 2002, 45,
3844.
Supplementary data
26. Cianchetta, G.; Li, Y.; Kang, J.; Rampe, D.; Fravolini, A.; Cruciani, G.; Vaz, R. J.
Bioorg. Med. Chem. Lett. 2005, 15, 3637.
27. Crane, S. N.; Black, W. C.; Palmer, J. T.; Davis, D. E.; Setti, E.; Robichaud, J.;
Paquet, J.; Oballa, R. M.; Bayly, C. I.; McKay, D. J.; Somoza, J. R.; Chauret, N.;
Seto, C.; Scheigetz, J.; Wesolowski, G.; Masse, F.; Desmarais, S.; Ouellet, M. J.
Med. Chem. 2006, 49, 1066.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2012.
07.012.
28. Robichaud, J.; Bayly, C. I.; Black, W. C.; Desmarais, S.; Leger, S.; Masse, F.;
McKay, D. J.; Oballa, R. M.; Paquet, J.; Percival, M. D.; Truchon, J.; Wesolowski,
G.; Crane, S. N. Bioorg. Med. Chem. Lett. 2007, 17, 3146.
29. Leach, A. G.; Jones, H. D.; Cosgrove, D. A.; Kenny, P. W.; Ruston, L.; MacFaul, P.;
Wood, J. M.; Colclough, N.; Law, B. J. Med. Chem. 2006, 49, 6672.
30. Dossetter, A. G. Bioorg. Med. Chem. 2010, 18, 4405.
References and notes
1. Teitelbaum, S. L. Science 2000, 289, 1504.
2. Lawrence, R. C.; Felson, D. T.; Helmick, C. G.; Arnold, L. M.; Choi, H.; Deyo, R. A.;
Gabriel, S.; Hirsch, R.; Hochberg, M. C.; Hunder, G. G.; Jordan, J. M.; Katz, J. N.;
Kremers, H. M.; Wolfe, F.; National Arthritis Data Workgroup Arthritis Rheum.
2008, 58, 26.
3. Gupta, S.; Singh, R. K.; Dastidar, S.; Ray, A. Expert Opin. Ther. Targets 2008, 12, 291.
4. Yasuda, Y.; Kaleta, J.; Broemme, D. Adv. Drug Delivery Rev. 2005, 57, 973.
31. Lewis, M. L.; Cucurull-Sanchez, L. J. Comput. Aided Mol. Des. 2009, 23, 97.
32. Hajduk, P. J.; Sauer, D. R. J. Med. Chem. 2008, 51, 553.
´
33. Walczynski, K.; Zuiderveld, O. P.; Timmerman, H. Eur. J. Med. Chem. 2005, 40,
15.