Design and synthesis of aminothiazole modulators of the gamma-secretase enzyme

Article abstract of DOI:10.1016/j.bmcl.2016.07.011

The design and construction of a series of novel aminothiazole-derived γ-secretase modulators is described. The incorporation of heterocyclic replacements of the terminal phenyl D-ring of lead compound 1 was conducted in order to align potency with favorable drug-like properties. γ-Secretase modulator 28 displayed good activity for in vitro inhibition of Aβ42, as well as substantial improvement in ADME and physicochemical properties, including aqueous solubility. Pharmacokinetic evaluation of compound 28 in mice revealed good brain penetration, as well as good clearance, half-life, and volume of distribution which collectively support the continued development of this class of compounds.

Full text of DOI:10.1016/j.bmcl.2016.07.011

Bioorganic & Medicinal Chemistry Letters 26 (2016) 3928–3937
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design and synthesis of aminothiazole modulators of the gamma-
secretase enzyme
Kevin D. Rynearson a, Ronald N. Buckle b, Keith D. Barnes b, R. Jason Herr b, Nicholas J. Mayhew b,
William D. Paquette b, Samuel A. Sakwa b, Phuong D. Nguyen a, Graham Johnson c, Rudolph E. Tanzi d,
Steven L. Wagner a,
⇑
a Department of Neurosciences, University of California, San Diego, 9500 Gilman Drive MC 0624, La Jolla, CA 92093-0624, United States
b Department of Medicinal Chemistry, AMRI, East Campus, 3 University Place, Rensselaer, NY 12144, United States
c NuPharmAdvise, 3 Lakeside Drive, Sanbornton, NH 03269, United States
d Genetics and Aging Research Unit, Department of Neurology, Massachusetts General Hospital, Charlestown, MA 02129, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
The design and construction of a series of novel aminothiazole-derived
described. The incorporation of heterocyclic replacements of the terminal phenyl D-ring of lead com-
pound 1 was conducted in order to align potency with favorable drug-like properties. -Secretase mod-
ulator 28 displayed good activity for in vitro inhibition of Ab42, as well as substantial improvement in
ADME and physicochemical properties, including aqueous solubility. Pharmacokinetic evaluation of com-
pound 28 in mice revealed good brain penetration, as well as good clearance, half-life, and volume of dis-
tribution which collectively support the continued development of this class of compounds.
Ó 2016 Published by Elsevier Ltd.
c-secretase modulators is
Received 4 May 2016
Revised 1 July 2016
Accepted 4 July 2016
Available online 6 July 2016
c
Keywords:
Alzheimer’s disease
c
-Secretase modulators
Aminothiazoles
Alzheimer’s disease (AD) is a devastating neurological disorder
and is currently estimated to affect 5.3 million Americans. In 2015,
AD and other related dementias will cost the United States
$226 billion, and if the current disease trajectory is maintained,
the associated costs could rise to as high as $1.1 trillion by
2050.1 AD is the only top ten cause of death in America that cannot
be prevented, cured or slowed, and thus there is an urgent medical
need for disease modifying therapeutic agents.1
first cleaved by BACE-1 which yields membrane bound carboxy-
terminal fragment-b (CTF-b) known as C99. Further cleavages of
the C99 fragment by
c-secretase releases the APP intracellular
domain (AICD) and produces extracellular Ab peptides which vary
in length from 37 to 43 amino acids.5
Directly implicated in the production of Ab peptides, BACE-1
and
c-secretase represent attractive targets for the development
of AD therapies.3–5 Consequentially, small molecule compounds
AD is an irreversible and progressive neurodegenerative disease
affecting the brain which insidiously destroys memory, thinking
skills and cognition. Originally described in 1906 by German physi-
cian Alois Alzheimer, AD is characterized by the presence of intra-
neuronal neurofibrillary tangles of hyperphosphorylated
microtubule associated protein tau and extraneuronal neuritic pla-
ques primarily composed of amyloid b-42 (Ab42).2 A large body of
histopathological and genetic evidence implicates that the process-
ing and deposition of Ab peptides within the brain is the primary
driver of AD progression ultimately leading to dementia.3 The Ab
peptides are formed as the result of sequential proteolytic cleav-
ages of the amyloid precursor protein (APP) by two aspartyl pro-
which curtail the formation of all Ab species by inhibition of either
BACE-1 or c c-sec-
-secretase have been intensely pursued.6 Potent
retase inhibitors (GSI) such as semagacestat and avagacestat were
developed and demonstrated robust reduction of Ab peptides.7
Unfortunately, these GSIs also displayed severe side effects and
counter efficacy which resulted in accelerated cognitive decline
during clinical evaluation, and thus
c
-secretase inhibition strate-
-secretase
gies were abandoned.8 As an alternative to inhibition,
c
modulation offers an attractive approach with the goal of attenuat-
ing the production of the neurotoxic and aggregation prone Ab42
isoform, the primary component of neuritic plaques, while not
affecting endogenous c
-secretase function.9 Based on the fact that
teases, b-secretase (BACE-1) and
c
-secretase, respectively.4 APP is
the vast majority of the more than 200 FAD-linked genetic muta-
tions appear to cause a 2-fold increase in the ratio of the longer
Ab42 peptide to the shorter Ab40 peptide, and a large body of data
pointing specifically to Ab42 in pathogenesis, a therapeutic ratio-
⇑
Corresponding author.
E-mail address: slwagner@ucsd.edu (S.L. Wagner).
nale that modulates
c-secretase activity to reduce the level of
http://dx.doi.org/10.1016/j.bmcl.2016.07.011
0960-894X/Ó 2016 Published by Elsevier Ltd.

K. D. Rynearson et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3928–3937
3929
H
N
Ab42 relative to the shorter Ab peptides (i.e., Ab40, Ab38 and Ab37)
without affecting overall -secretase function may prove to be an
efficacious course for interrupting AD progression.10 Several
classes of -secretase modulators have been discovered which
S
O
R
H2N
c
NH
R
F
F
Br
N
S
a
3
c
N
N
alter the Ab cleavage pattern in favor of shorter Ab peptides,
including compounds derived from non-steroidal anti-inflamma-
tory drugs (NSAIDs),11 aryl imidazoles5,12 and triterpenes.13
Previously our lab discovered a series of aminothiazole-derived
N
N
2
4
Scheme 1. Reagents and conditions: (a) EtOH, reflux, overnight, 16–79%,
R = tetrahydroindazole, 3-tert-butylpyrazole or cyclopentapyrazole containing
scaffold.
c-secretase modulators (AGSMs) through rational hit to lead opti-
mization efforts which demonstrate remarkable potency for lower-
ing Ab42 (>1000-fold more potent than the NSAID-like GSM
tarenflurbil) and exhibit moderate brain penetrance.14 This novel
class of compounds is characterized by a tetracyclic scaffold com-
posed of bridged linear aromatics (Fig. 1). Compounds within this
series have been shown to specifically reduce the levels of Ab42
and Ab40 production while simultaneously increasing the levels
of Ab38 and Ab37, thereby leaving the total amount of Ab produced
was envisioned to mirror the alkyl functionalization of AGSM 1
while rigidifying the D-ring and preserving favorable hydrophobic
interactions of the parent molecule. 2-Substituted tetrahydroinda-
zoles 5–9 displayed modest activity for the reduction of Ab42 in a
SHSY5Y neuroblastoma cell line stably over-expressing human APP
and demonstrated a slight gain in potency as the alkyl chain
increases in size (compounds 5–7 Table 1).16,17 However, the alter-
ation in N-alkyl connectivity of tetrahydroindazole 7 providing iso-
propyl derivative 8, as well as the introduction of a tert-butyl group
in compound 9 led to a slight reduction in activity. Removal of the
2-postion substituent as illustrated by tetrahydroindazole 10 abro-
gates activity, underscoring the importance of the alkyl functional-
ity with respect to Ab modulation.
Based on the pivotal role the 2-postion substituent plays with
respect activity, in addition to the overall tolerance for alkyl sub-
stituents of various size, analogs 11–26 were prepared to thor-
oughly probe the chemical space for favorable interactions. The
constrained and small 2-cyclobutyl tetrahydroindazole 11 dis-
played a 2-fold increase in activity when compared to isopropyl
unchanged. Importantly, AGSMs do not affect
c-secretase-medi-
ated cleavage of other critical substrates, including Notch and E-
cadherin. The selectivity for Ab products likely stems from the
observed binding of AGSMs to Pen-2 and PS-1 NTF of the c-secre-
tase enzymatic complex which shifts rather than inhibits endoge-
nous function. Additionally, in vivo studies showed AGSMs were
potent and effective at decreasing the levels of Ab42 and Ab40 in
the plasma and brain of APP transgenic mice. Chronic administra-
tion of AGSMs to Tg2576 APP transgenic mice resulted in dramatic
reduction of AD-like pathology in the absence of GSI-related effects
such as intestinal goblet cell hyperplasia due a distinct mode of
action.14 Despite the overall efficacy of these compounds, the poor
aqueous solubility (<0.1 lM at neutral pH) of the AGSMs presents a
analog 8, suggesting that this region of c-secretase features a nar-
significant liability especially when attempting to achieve the
supraefficacious exposures required for safety and toxicology stud-
ies during preclinical development.
row hydrophobic pocket. This assertion is further supported by
ligands 12–16 which illustrate that as the size of the substituent
increases from the 2-cyclopropyl to the larger 2-cyclohexyl deriva-
tive (12 vs 13) the observed activity decreases. Moreover, increas-
ing the polarity of the 2-position substituent through the
incorporation of heteroatoms as exemplified by the tetrahydrofu-
ran 14, tetrahydropyran 15 and 1-methylpiperidine 16 signifi-
cantly erodes potency. In general, deviation from linear N-alkyl
substitution is deleterious to suppressing Ab42 levels, and the
introduction of heteroatoms served to exacerbate this trend.
Efforts to increase the flexibility of the heterocyclic substituent,
as well as extend beyond the constrained hydrophobic region
through the insertion of an ethyl or propyl linker failed to restore
the loss of activity as evidenced by analogs 17, 18 and 19. Conse-
quentially, the incorporation of smaller functionalities within the
active 2-postion alkylated analogs was explored. 2-Trifluo-
rorethane 21 demonstrated similar activity to parent analog 6,
whereas 2-fluoroethane 20 displayed a slight increase in activity
for the reduction of Ab42. The insertion of oxygen with the alkyl
chain was poorly tolerated as demonstrated by 2-methoxyethane
22, 2-hydroxyethane 23, and 2-hydroxypropane 24 which are
approximately 2-fold less active than the corresponding aliphatic
analog. However, increasing hydrophobicity of analog 23 through
the addition of two flanking methyl groups resulting in 2-
hydroxy-2-methylpropane compound 25 showed an unexpected
3-fold improvement over primary analog 7 thus reinforcing the
strong preference for hydrophobic character within this region of
Herein, we describe the development of second generation GSM
compounds aimed at improving critical physicochemical proper-
ties while maintaining the potent activity of the parent AGSM.
Ligand design focused on identifying heterocyclic replacements
for the hydrophobic D-ring as a means to improve ADME (absorp-
tion, distribution, metabolism and excretion) parameters, as well
as aqueous solubility. Replacement of the alkyl rich phenyl D-ring
by a pyrazole containing scaffold was anticipated to ameliorate the
property related shortcomings of the AGSMs. Based on the spatial
arrangement of the substituents within AGSM 1, as well as molec-
ular modeling overlays, tetrahydroindazole, 3-tert-butylpyrazole
and cyclopentapyrazole D-ring analogs were selected for prepara-
tion. The construction of the three unique series would be accom-
plished utilizing established Hantzsch chemistry from a common
bromoacetophenone precursor enabling the rapid preparation of
a diverse set of pyrazole containing scaffolds permitting rigorous
evaluation of these novel D-ring substitutions (Scheme 1).15
A series of 3-aminotetrahydroindazole analogs were synthe-
sized in order to explore the local spatial constraints of the c-sec-
retase binding cavity with respect to this novel D-ring substitution.
The fused ring system of the 2-substituted tetrahydroindazole 5
S
C
N
NH
F
B
the D-ring by the c-secretase enzyme. Finally, an N-methyl deriva-
D
N
tive of compound 26 was synthesized in order to ascertain the role
of the proton with respect to ligand affinity. Unfortunately, N-
methylated analog 26 loses all activity suggesting either the pres-
ence of a hydrogen bond between the ligand and enzyme or an
overall reduction in ligand flexibility stemming from the sterical
interactions between the D-ring and the added methyl which
restrict access to conformations required for target interaction.
A
N
1
Aβ42 IC50 = 10 nM
clogP=5.96
Figure 1. General scaffold of optimized lead aminothiazole-derived
c-secretase
modulator (AGSM). The rings are labeled A–D for clarity.

3930
K. D. Rynearson et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3928–3937
Table 1
3-Aminotetrahydroindazole D-ring analogs of AGSM 1
S
R
F
N
N
N
clogPb
Kinetic aqueous
solubilityc
a
Compd
R
Ab42 IC50
NH
NH
NH
NH
NH
NH
NH
N
N
5
208
3.54
ntd
ntd
ntd
ntd
ntd
ntd
<1.6
N
N
6
163
3.88
4.35
4.30
4.37
3.59
4.33
N
N
7
148
N
N
8
195
N
N
9
216
NH
N
10
11
>1000
112
N
N
NH
N
N
12
225
4.73
ntd
NH
NH
NH
ntd
ntd
ntd
N
N
13
14
15
280
289
766
5.13
3.32
3.37
O
N
N
O
N
N
N
NH
NH
N
N
16
17
>1000
475
3.52
4.24
2.8
8.1
N
N
N

K. D. Rynearson et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3928–3937
3931
Table 1 (continued)
a
Compd
R
Ab42 IC50
clogPb
Kinetic aqueous
solubilityc
O
NH
NH
NH
NH
NH
NH
NH
NH
N
N
N
18
>1000
3.18
ntd
N
O
N
N
19
20
21
22
23
24
25
270
110
179
272
362
229
55
3.23
3.70
4.52
3.38
2.80
2.85
3.29
4.13
16
F
N
N
ntd
ntd
2.3
15
CF3
O
N
N
N
N
OH
N
N
OH
N
N
ntd
<1.6
ntd
OH
N
N
N
N
N
26
>1000
a
IC50 represents the concentration in nM of compound required for reducing Ab42 levels by 50%. The IC50 values are the mean of at
least 2 determinations.16,17
b
Calculated partition coefficient of the ratio of the compound’s concentration in octanol to the compound’s concentration in water
using ChemAxon.
c
Kinetic solubility measured at pH 7.4 by UV/Vis absorbance in PBS buffer (
nt = not tested—compound did not meet minimum activity threshold (Ab42 IC50 6 100 nM).
l
M).
d
Optimization of the tetrahydroindazole series led to the
pyrazole core of this series to determine the effect of the fused ring
system on ligand activity, as well as on aqueous solubility.
In order to improve upon the tetrahydroindazole family of com-
pounds, as well as further enhance our understanding of D-ring
discovery of two moderately potent
c-secretase modulators,
2-cyclobutane 11 and 2-hydroxy-2-methylpropane 25 which were
subsequently evaluated in a kinetic aqueous solubility assay.
Unfortunately, neither analog demonstrated improved solubility
(Table 1). However, the incorporation of heteroatoms within the
D-ring does lead to compounds which exhibit ideal solubility as
demonstrated by analogs 17, 19 and 23, albeit at the expense of
modulator activity. Modest improvements in aqueous solubility
were also observed for 1-methylpiperidine 16 and 2-methox-
yethane 22, thus validating our synthetic strategy for addressing
this property related shortcoming. In addition, integrating heteroa-
toms within the 2-position substituent of the tetrahydroindazole
family of compounds was found to result in ligands with poor
microsomal stability (data not shown). Thus, our efforts to main-
tain activity for inhibiting the formation of Ab42 while improving
ligand solubility through the introduction of novel tetrahydroinda-
zole D-ring scaffolds provided ligands with fair activity, but lacking
the desired physicochemical properties of a drug-like compound.
Therefore, additional analogs were explored which retain the
interactions with the c-secretase enzyme, a novel set of 1-substi-
tuted 3-tert-butylpyrazole ligands was designed and synthesized
with the goal of simultaneously increasing potency and improving
aqueous solubility. The placement of substituents was again mod-
eled on preserving favorable hydrophobic interactions present in
the D-ring of AGSM 1. Additionally, the tert-butyl substitution
was envisaged to allow the exploration of new regions within
the target through the use of a bulky substituent directed into a
specific region of space while simultaneously increasing overall
ligand flexibility when compared to its fused ring counterpart.
Alkylated pyrazole analogs 27–31 were prepared based on the
moderate Ab42 reduction observed in the 2-substituted tetrahy-
droindazole series (Table 2). The novel 1,3-disubstitiuted pyrazole
compounds displayed modest to good activity, with N-ethyl
derivative 28 displaying the best Ab42-lowering activity with an
IC50 = 63 nM. The incorporation of longer alkyl chains did not lead

3932
K. D. Rynearson et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3928–3937
Table 2
1-Substituted 3-tert-butylpyrazole D-ring analogs of AGSM 1
S
R
F
N
N
N
a
Compd
R
Ab42 IC50
clogPb
Kinetic aqueous
solubilityc
NH
NH
NH
NH
NH
N
N
27
294
4.41
ntd
N
N
28
29
30
31
32
33
63
86
87
85
394
67
4.75
5.22
5.16
5.24
4.09
4.25
4.7
N
N
<1.6
5.5
N
N
N
N
2.0
NH
NH
N
N
ntd
N
N
<1.6
NH
N
N
34
35
110
121
4.01
4.50
<1.6
ntd
F3C
NH
N
N
F3C
NH
CF3
N
N
36
87
5.38
<1.6
NH
CF3
N
N
37
38
771
4.64
2.72
ntd
ntd
F
3C
NH
N
N
>1000
a
IC50 represents the concentration in nM of compound required for reducing Ab42 levels by 50%. The IC50 values are the mean of at
least 2 determinations.16,17
b
Calculated partition coefficient of the ratio of the compound’s concentration in octanol to the compound’s concentration in water
using ChemAxon.
c
Kinetic solubility measured at pH 7.4 by UV/Vis absorbance in PBS buffer (
nt = not tested—compound did not meet minimum activity threshold (Ab42 IC50 6 100 nM).
l
M).
d

K. D. Rynearson et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3928–3937
3933
Table 3
to improved potency as observed with the tetrahydroindazoles.
1-Substituted 3-tert-butylpyrazole D-ring analogs of AGSM 1
Also distinct, the larger N-isopropyl analog 30 and N-tertbutyl ana-
log 31 did not adversely impact the potency of this series. The
improved activity of compounds 28–31 was also accompanied by
a modest improvement in aqueous solubility; pyrazoles 28, 30
and 31 displayed improved solubility, while compound 29 main-
tained the poor solubility associated with many of the ligands
developed thus far.
S
R
F
N
N
N
a
Compd
R
Ab42 IC50
clogPb
Kinetic
The initially prepared 1-substituted 3-tert-butylpyrazole ligands
(28–31) demonstrated consistent potency over a wide range of
alkyl substitutions. The improved properties of the 3-tert-
butylpyrazole set of ligands led us to explore additional analogs
of N-ethyl pyrazole 28. Ligand design focused on potential modifi-
cations to the pyrazole substitution pattern which probe the subtle
interactions between the compound and the target. Introduction of
a 3-isopropyl substituent (32) within the pyrazole resulted in a
6-fold reduction in activity when compared to compound 28
(Table 2). On the other hand, 3-methylcyclopropyl 33 maintained
potency, suggesting the overall hydrophobic bulk within the
3-position of the pyrazole is relevant to binding affinity. Surpris-
ingly, the integration of a smaller 3-trifluoromethyl substituent in
pyrazole 34 was tolerated. In addition, similar to the loss in potency
observed between 3-tert-butylpyrazoles 28 and 31, N-tertbutyl-3-
trifluoromethyl analog 35 only showed a minor loss in activity
associated with the larger amino substituent. The preservation of
activity associated with the 3-trifluoromethyl analogs 34 and 35
is hypothesized to result from the maintained favorable hydropho-
bic interactions of the N-substituent within the enzyme target, as
well as a substantial reduction in the penalty associated with the
lost interactions of the 3-tertbutyl substituent through unantici-
pated electronic effects of the trifluoromethyl replacement. The
incorporation of N-trifluoroethyl was also explored resulting in
derivative 36 which showed comparable activity to 3-trifluo-
romethyl pyrazole 34. However, N-trifluoroethyl-3-trifluoromethyl
pyrazole 37 displays a greater than 10-fold loss in activity lending
support to the notion that the presence of a hydrophobic sub-
stituent within the D-ring is required in order to maintain ligand
affinity. Furthermore, the removal of the 3-position substituent as
illustrated by N-ethylpyrazole analog 38 results in a complete loss
of activity for the suppression of Ab42 formation.
aqueous
solubilityc
NH
NH
NH
NH
NH
NH
NH
NH
N
N
39
40
41
42
43
44
45
>1000
41
3.14
3.49
3.90
3.98
3.94
4.12
2.40
ntd
N
N
<1.6
<1.6
ntd
N
N
66
N
N
175
92
N
N
<1.6
<1.6
<1.6
CF3
N
N
94
OH
N
N
>1000
N
N
46
62
4.71
2.1
The focused set of additional pyrazole analogs allowed the iden-
NH
tification of three additional moderately potent c-secretase modu-
N
N
47
48
49
5.04
4.28
2.6
ntd
lators, 33, 34 and 36 which were subsequently screened in a
kinetic solubility assay. Unfortunately, neither the architectural
changes in analog 33 nor the integration of trifluoro substituents
in analogs 34 and 36 resulted in an increase in solubility despite
the decrease in the clogP observed for compounds 33 and 34. Even
though few ligands within the series showed improved solubility,
overall the prepared tertbutylpyrazole family of compounds pro-
vided tremendous insight into the specific interactions within the
D-ring which govern biological potency.
NH
N
N
114
a
IC50 represents the concentration in nM of compound required for reducing
Ab42 levels by 50%. The IC50 values are the mean of at least 2 determinations.16,17
b
As a complement to the tetrahydroindazoles and the substituted
Calculated partition coefficient of the ratio of the compound’s concentration in
octanol to the compound’s concentration in water using ChemAxon.
tert-butylpyrazoles,
a novel set of compounds containing a
c
Kinetic solubility measured at pH 7.4 by UV/Vis absorbance in PBS buffer (
nt = not tested—compound did not meet minimum activity threshold (Ab42
lM).
cyclopentapyrazole D-ring were also explored. The primary design
objective within this series was to evaluate whether contraction of
the fused ring system would mitigate potential unfavorable clashes
within the binding cleft resulting from the larger tetrahydroinda-
zole scaffold while reducing the overall hydrophobic character of
the substituent thereby improving aqueous solubility. Additionally,
the substituted cyclopentapyrazoles were envisaged to broaden the
structure-activity relationship (SAR) data through the incorpora-
tion of structural features which conferred activity to the previously
prepared analogs within the context of this distinct scaffold.
Cyclopentapyrazole analogs 39–45 were synthesized based on
the most active compounds within the tetrahydroindazole series.
N-Methyl derivative 39 was unexpectedly inactive toward
d
IC50 6 100 nM).
reducing the generation of Ab42. However, N-ethylcyclopentapyra-
zole 40, N-isopropylcyclopentapyrazole 41 and N-trifluoroethylcy-
clopentapyrazole 44 displayed good potency and were 4-fold,
3-fold and 2-fold superior to the corresponding tetrahydroinda-
zoles 6, 8 and 21, respectively. However, N-tert-butyl compound
42 and N-cyclobutyl compound 43 were unaffected by the contrac-
tion of the fused ring system and displayed commensurate activity
to the analogous tetrahydroindazoles 9 and 11, respectively.
Efforts to incorporate a more polar amino substituent resulted in

3934
K. D. Rynearson et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3928–3937
Table 4
a complete loss of activity as evidenced by N-hydroxyethyl analog
45. The absence of a general tendency toward improved activity
indicates that the smaller D-ring is not solely responsible for
improved potency of the cyclopentapyrazole series. Additionally,
increasing the size of the fused ring system to the corresponding
cycloheptapyrazole, 48, provides an analog with activity superior
to the analogous tetrahydroindazole 6, demonstrating the absence
of a direct connection between size and activity within this partic-
ular region of the D-ring. Thus, the specific interactions of the non-
planar fused D-ring system, as well as the alkyl amino substituent
are crucial to facilitating favorable interactions within this region
In vitro ADMET properties of pyrazole 28
S
NH
F
N
N
N
N
N
28
Ab42 IC50 (nM)
Kin. sol. pH 7.4 (
clogP
63
l
M)a
4.7
4.75
3.64
51
64
95
of the c-secretase enzyme.
logD
In order to correlate the binding mode of the substituted tert-
butylpyrazoles within the cyclopentapyrazole family of com-
pounds, 6,6-dimethylcyclopentapyrazole analogs 46 and 47 were
also synthesized. The dimethyl substituent of compound 46 was
hLM (%)b
rLM (%)b
mLM (%)b
CYP450 IC50
3A4
1A2
(l
M)c
16
anticipated to occupy the hydrophobic region of
c-secretase in a
>100
>100
82
>100
>30
0.97
1.6
similar fashion to tert-butyl group of pyrazole 28, however within
the context of a more rigid scaffold. In addition, the slightly larger
trimethylcyclopentapyrazole 47 was conceived to explore the
breadth of the binding pocket. Surprisingly, both ligands main-
tained similar activity to the parent pyrazole 40, displaying IC50 val-
ues of 62 nM and 49 nM for the inhibition of Ab42, respectively.
The unanticipated activity of both ligands demonstrates that there
is little to no penalty associated with incorporation of methyl sub-
stituents at either the 4- or 6-positions of cyclopentapyrazole signi-
fying the presence of a wide cavity which may yield, upon further
exploration, favorable interactions with less hydrophobic moieties.
The focused set of cyclopentapyrazoles produced six com-
pounds with sufficient activity to warrant the evaluation of kinetic
solubility. Despite the improved activity of the series, these com-
pounds remain stalwartly insoluble (Table 3), indicating the need
for additional chemical modifications which address aqueous solu-
bility. However, the general tendency of adding of more hydropho-
bic bulk to the D-ring in order to produce compounds with the
desired activity runs counter to the overarching goal of increasing
solubility and improving other related physicochemical properties.
Therefore, D-ring optimization may not be an ideal avenue to
address the underlying property related issues of the cyclopen-
tapyrazole scaffold.
2C9
2C19
2D6
hERG IC50
(
l
M)d
e
Papp
Efflux ratiof
% PPBg
99.4
a
Kinetic solubility measured at pH 7.4 by UV/Vis absorbance in PBS buffer (
l
M).
% remaining after 30 min upon incubation with human (hLM), rat (rLM) and
M test compound concentration.
b
mouse (mLM) liver microsomes (1 mg/mL), at 1
l
c
Five recombinant CYP isoforms were tested for inhibition by the test compound
using fluorescence-based assays.
d
Patch-Xpress patch-clamp assay; compounds were tested (n = 3) in a five-point
concentration–response on HEK-293 cells stably expressing the hERG channel.
Apparent permeability (A–B; Â10À6 cm/s) determined in MDR1-MDCK cell
e
monolayers.
f
Calculated efflux ratio determined using MDR1-MDCK cell monolayers (B–A/A–B).
g
% PPB = plasma protein binding.
Table 5
Mouse in vivo pharmacokinetic study of pyrazole 28
iv
po
Mouse pharmacokinetics of 28a
Dose (mg/kg)
Cmax (ng/mL)
Tmax (h)
t1/2 (h)
1
5
1005
1.0
6.0
5393
5720
161
0.083
2.4
Chemical optimization of AGSM 1 through the construction of
three related D-ring series of compounds led to the discovery of
various novel ligands with good activity for suppressing the forma-
AUClast (ngÁh/mL)
335
498
2009
7152
100
0.64
ncb
AUCinf (ngÁh/mL)
tion of Ab42 through modulating the activity of the
c-secretase
Cl (mL/h/kg)
enzyme. Unfortunately, many of the ligands identified suffered
from poor aqueous solubility. However, a subset of the ligands
within the 1-substituted 3-tert-butylpyrazoles series displayed
both good activity, as well as improved aqueous solubility. The
most potent compound discovered within the series, N-ethylpyra-
zole 28, displayed good potency for suppressing Ab42 with an IC50
value of 63 nM, in addition to moderate solubility (Table 4). Conse-
quentially, further in vitro ADMET (Absorption, Distribution, Meta-
bolism, Excretion and Toxicology) analysis of N-ethylpyrazole 28
was undertaken in order to identify potential liabilities within this
family of compounds which would preclude further development.
N-Ethylpyrazole 28 demonstrated ideal CyP inhibition potential
and hERG ion channel inhibition potential (all IC50 values
VSS (mL/kg)
F (%)
Brain/plasma ratio (@ 1 h post dose)
Brain/plasma ratio (@ 4 h post dose)
Brain/plasma ratio (@ 8 h post dose)
>100
0.50
0.53
0.59
ncb
a
CD-1 male mice n = 24.
Not collected.
b
As shown in Table 5, in vivo dosing of N-ethylpyrazole 28 in
male CD-1 mice revealed that kinetically compound 28 has a rel-
atively high clearance (CL) and a short plasma half-life (t ),
½
despite the high metabolic stability observed in the in vitro
mouse microsomal stability assay.18 Following IV administration
at 1 mg/kg in mice, the clearance of compound 28 was
2009 mL/hr/kg (or 33.5 mL/min/kg) which is about 40% of mouse
>10 lM). In addition, compound 28 showed fair microsomal stabil-
ity. However, poor membrane permeability as measured by MDR1-
MDCK assay and high plasma protein binding were also observed
for N-ethylpyrazole 28. Despite the limited in vitro membrane per-
meability, compound 28 was selected for in vivo pharmacokinetic
(PK) studies in order to gauge critical PK parameters, determine
whether this ligand can reach the intended enzyme target within
the brain and establish a baseline for future compound develop-
ment with respect to SAR/SPR.
hepatic blood flow (ꢀ80 mL/min/kg), while its half-life (t ) was
½
2.4 h. Although compound 28 has a relatively large volume of dis-
tribution (Vss = 7.15 L/kg), which is about 11-fold higher than
body water (ꢀ0.65 L/kg), it has a short half-life due mainly to
its high rate of clearance. The bioavailability (F) of N-ethylpyra-
zole 28 is unexpectedly high and is estimated to be 230%.

K. D. Rynearson et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3928–3937
3935
O
O
O
F
F
Br
a
b
F
F
N
N
N
N
50
49
2
Scheme 2. Reagents and conditions: (a) 4-Methyl-1H-imidazole, K2CO3, DMSO, 55 °C, overnight, 58%; (b) Br2, CHCl3, 30% HBr in acetic acid, rt, 3 h, 100%.
H
N
H
N
CN
O
N
N
N
H
N
a
b
H2N
c
N
N
N
H2N
Ph
S
S
O
n
n
n
n
54
51
52
53
Scheme 3. For substrates n = 1 or 2. Reagents and conditions: (a) Ethylhydrazine oxalate, EtOH, reflux, overnight, 45%; (b) benzoyl isothiocyanate, 60 °C, 3.5 h, 99%; (c) K2CO3,
MeOH, THF, rt, overnight, 90%.
H
N
N
N
H2N
54
S
O
S
NH
F
F
Br
N
N
N
n
a
N
N
N
N
n
2
n =
n =
1
2
40,
6,
Scheme 4. Reagents and conditions: (a) EtOH, reflux, overnight, n = 1, 78%, n = 2, 24%.
H
N
H
N
N
N
O
N
H
N
a
b
H2N
c
N
N
N
H2N
NC
Ph
S
S
O
55
56
57
58
Scheme 5. Reagents and conditions: (a) Ethylhydrazine oxalate, EtOH, reflux, overnight, 61%; (b) benzoyl isothiocyanate, acetone, reflux, 3.5 h, 88%; (c) K2CO3, MeOH, THF, rt,
overnight, 92%.
H
N
N
N
H2N
58
S
O
NH
S
F
F
Br
N
N
N
a
N
N
N
N
28
2
Scheme 6. Reagents and conditions: (a) EtOH, reflux, overnight, 23%.
Theoretically, the bioavailability should not exceed 100%, and one
would expect based on the clearance of compound 28 being 40%
of mouse hepatic blood flow that significant hepatic first-pass
metabolism would result in bioavailability much less than 100%,
even when the compound is completely absorbed from the
intestinal lumen. The high bioavailability is hypothesized to be
the result of saturation of hepatic drug metabolism during oral
absorption. Following the oral dose of 5 mg/kg, the concentration
of N-ethylpyrazole 28 is very high in the portal vein which satu-
rates the drug metabolizing enzymes in the liver and results in
decreased compound clearance leading to a marked increase in
the oral AUC and an over estimation of F. N-Ethylpyrazole 28 also
shows good brain penetrance for both routes of administration, in
contrast to the in vitro permeability assay results. Kinetically, the
mouse PK parameters (CL, t and Vss) of compound 28 are reason-
½
ably good and support further development efforts surrounding
this class of compounds.
In order to facilitate the preparation of 2-aminothiazoles con-
taining novel D-ring substitutions, a convergent synthetic route
was selected which requires the synthesis of common precursor

3936
K. D. Rynearson et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3928–3937
2. Straightforward aromatic substitution of 3,4-difluoroacetophe-
none using 4-methylimidazole provided access to 2-imidazolylace-
tophenone intermediate 50 in fair yield (Scheme 2).19 Subsequent
acid-catalyzed bromination allowed isolation of the desired
bromoacetophenone compound 2 in quantitative yield.20 The con-
struction of the complementary cyclopentapyrazole fragment
toward the synthesis of 2-aminothiazole 40 commenced with the
condensation of 2-cyanocyclopentanone with ethylhydrazine in
order to afford 3-aminocyclopentapyrazole 52 (Scheme 3).21 Sub-
sequent reaction of intermediate 52 with benzoyl isothiocyanate
furnished benzoyl thiourea 53 in good yield. Isolation of the
desired cyclopentapyrazole coupling partner 54 was accomplished
following removal of the benzoyl group using K2CO3. Assembly of
the desired final product was then accomplished by Hantzsch con-
densation of thiourea 54 with bromoethanone 2 to afford aminoth-
iazole 40 in good yield (Scheme 4). This general approach provided
access to tetrahydroindazoles 5–26, as well as cyclopentapyrazoles
39–48 by selecting the appropriate starting material and utilizing
an assortment of commercially available substituted hydrazines
in the synthetic sequence.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2016.07.
011.
References and notes
1. Alzheimer’s Association Alzheimer’s Demen. 2015, 11, 332.
2. (a) Glenner, G. G.; Wong, C. W. Biochem. Biophys. Res. Commun. 1984, 120, 885;
(b) Masters, C. L.; Simms, G.; Weinman, N. A.; Multhaup, G.; McDonald, B. L.;
Beyreuther, K. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 4245; (c) Iwatsubo, T.; Saido,
T. C.; Mann, D. M.; Lee, V. M.; Trojanowski, J. Q. Am. J. Pathol. 1996, 149, 1823;
(d) Dickson, D. W. J. Neuropathol. Exp. Neurol. 1997, 56, 321; (e) Liao, L.; Cheng,
D.; Wang, J.; Duong, D. M.; Losik, T. G.; Gearing, M.; Rees, H. D.; Lah, J. J.; Levey,
A. I.; Peng, J. J. Biol. Chem. 2004, 279, 37061.
3. (a) Hardy, J.; Selkoe, D. J. Science 2002, 297, 353; (b) Tanzi, R. E.; Bertram, L. Cell
2005, 120, 545.
4. (a) Thinakaran, G.; Koo, E. H. J. Biol. Chem. 2008, 283, 29615; (b) De Strooper, B.;
Vassar, R.; Golde, T. Nat. Rev. Neurol. 2010, 6, 99; (c) Karrran, E.; Mercken, M.;
De Strooper, B. Nat. Rev. Drug Disc. 2011, 10, 698.
5. (a) Oehlrich, D.; Berthelot, D. J.; Gijsen, H. J. M. J. Med. Chem. 2011, 54, 669; (b)
Evin, G.; Barakat, A.; Masters, C. L. Int. J. Biochem. Cell Biol. 2010, 42, 1923.
6. (a) Durham, T. B.; Shepherd, T. A. Curr. Opin. Drug Disc. Devel. 2006, 9, 776; (b)
Stachel, S. J. Drug Dev. Res. 2009, 70, 101; (c) Vassar, R.; Kandalepas, P. C.
Alzheimer’s Res. Ther. 2011, 3, 20.
7. Bateman, R. J.; Siemers, E. R.; Mawuenyega, K. G.; Wen, G.; Browning, K. R.;
Sigurdson, W. C.; Yarasheski, K. E.; Friedrich, S. W.; DeMattos, R. B.; May, P. C.;
Paul, S. M.; Holtzman, D. M. Ann. Neurol. 2009, 66, 48.
A similar synthetic sequence was utilized to isolate the 3-tert-
butylpyrazole series of c-secretase modulators. The condensation
of ethylhydrazine with 4,4-dimethyl-3-oxopentanenitrile fur-
nished substituted aminopyrazole 56 (Scheme 5).21 The subse-
quent reaction of intermediate 56 with benzoyl isothiocyanate
followed by base mediated removal of the benzoyl protecting
group provided the desired coupling partner, pyrazolylthiourea
58, in good overall yield. Finally, reaction between intermediate
58 and common precursor 2 gave the desired aminothiazole 28
(Scheme 6). The use of a convergent synthetic strategy provided
access to a large assortment of novel D-ring ligands enabling rapid
evaluation of structure–activity relationships within the three
related series.
8. (a) Tomita, T. Exp. Rev. Neurother. 2009, 9, 661; (b) Imbimbo, B. P.; Panza, F.;
Frisardi, V.; Solfrizzi, V.; D’Onofrio, G.; Logroscino, G.; Seripa, D.; Pilotto, A.
Expert Opin. Investig. Drugs 2011, 20, 325.
9. (a) Van Broeck, B.; Chen, J. M.; Trenton, G.; Desmidt, M.; Hopf, C.; Ramsden, N.;
Karran, E.; Mercken, M.; Rowley, A. Br. J. Pharmacol. 2011, 163, 375; (b) Wagner,
S. L.; Tanzi, R. E.; Mobley, W. C.; Galasko, D. Arch. Neurol. 2012, 69, 1255.
10. (a) Iwatsubo, T.; Odaka, A.; Suzuki, N.; Mizusawa, H.; Nukina, N.; Ihara, Y.
Neuron 1994, 13, 45; (b) Weggen, S.; Erikson, J. L.; Das, P.; Sagi, S. A.; Wang, R.;
Pietrzik, C. U.; Findlay, K. A.; Smith, T. E.; Murphy, M. P.; Butler, T.; Kang, D. E.;
Marquez-Sterling, N.; Golde, T. E.; Koo, E. H. Nature 2001, 414, 212; (c) Kumar-
Singh, S.; Theuns, J.; Van Broek, B.; Pirici, D.; Vennekens, K.; Corsmit, E.; Cruts,
M.; Dermaut, B.; Wang, R.; Van Broeckhoven, C. Hum. Mutat. 2006, 27, 686; (d)
Rynearson, K. D.; Tanzi, R. E.; Wagner, S. L. In Translation Neuroscience:
Fundamental Approaches for Neurological Disorders; Tuszynski, M. H., Ed.;
Springer Science: New York, 2016; pp 359–368.
In summary, using poorly soluble AGSM 1 as the starting point,
a novel series of c-secretase modulator D-ring analogs possessing a
common pyrazole moiety were synthesized leading to the discov-
ery of compounds from three unique families which demonstrate
moderate activity for suppressing the formation of Ab42. Unfortu-
nately, the heterocyclic D-ring derivatives did not provide the
anticipated broad increase in compound solubility. Furthermore,
the SAR indicated an overall trend in which increasing the
lipophilicity of the D-ring leads to improved activity and conse-
quentially reduced aqueous solubility, as well as less than stellar
drug-like properties. Despite these challenges, ligands within the
1-substituted 3-tert-butylpyrazole series displayed reasonably
good activity and substantially improved properties including
aqueous solubility. Rigorous ADME evaluation of the most potent
molecule within the family, N-ethylpyrazole 28, revealed an over-
all good profile with the exception of poor membrane permeability
and high plasma protein binding. However, the reasonably good
in vivo PK behavior exhibited by pyrazole 28 indicates the need
for further development of this scaffold aimed at improving
potency, as well as crucial ADME parameters including membrane
permeability and aqueous solubility. Based on literature analysis,
modulator design has now shifted toward novel B-ring analogs in
order to address continuing property related issues and these
efforts will be reported in due course.22
11. (a) Bulic, B.; Ness, J.; Hahn, S.; Rennhack, A.; Jumpertz, T.; Weggen, S. Curr.
Neuropharmacol. 2011, 9, 598; (b) Jumpertz, T.; Rennhack, A.; Ness, J.; Baches,
S.; Pietrzik, C. U.; Bulic, B.; Weggen, S. PLoS One 2012, 7, e30484.
12. (a) Wan, Z.; Hall, A.; Jin, Y.; Xiang, J. N.; Yang, E.; Eatherton, A.; Smith, B.; Yang,
G.; Yu, H.; Wang, J.; Ye, L.; Lau, L. F.; Yang, T.; Mitchell, W.; Cai, W.; Zhang, X.;
Sang, Y.; Wang, Y.; Tong, Z.; Cheng, Z.; Hussain, I.; Elliot, J. D.; Matsuoka, Y.
Bioorg. Med. Chem. Lett. 2011, 21, 4016; (b) Velter, A. I.; Bischoff, F. P.; Berthelot,
D.; De Cleyn, M.; Oehlrich, D.; Jaroskova, L.; Macdonald, G.; Minne, G.; Pieters,
S.; Rombouts, F.; Van Brandt, S.; Van Roosbroeck, Y.; Surkyn, M.; Trabanco, A.
A.; Tresadern, G.; Wu, T.; Borghys, H.; Mercken, M.; Masungi, C.; Gijsen, H.
Bioorg. Med. Chem. Lett. 2014, 24, 5805; (c) Pettersson, M.; Johnson, D. S.;
Humphery, J. M.; am Ende, C. W.; Evrard, E.; Efremov, I.; Kauffman, G. W.;
Stepan, A. F.; Stiff, C. M.; Xie, L.; Bales, K. R.; Hajos-Korcsok, E.; Murrey, H. E.;
Pustilnik, L. R.; Steyn, S. J.; Wood, K. M.; Verhoest, P. R. Bioorg. Med. Chem. Lett.
2015, 25, 908.
13. (a) Findeis, M. A.; Schroeder, F.; McKee, T. D.; Yager, D.; Fraering, P. C.; Creaser,
S. P.; Austin, W. F.; Clardy, J.; Wang, R.; Selkoe, D.; Eckman, C. B. ACS Chem.
Neurosci. 2012, 3, 941; (b) Fuller, N. O.; Hubbs, J. L.; Austin, W. F.; Creaser, S. P.;
McKee, T. D.; Loureiro, R. M. B.; Tate, B.; Xia, W.; Ives, J. L.; Findeis, M. A.; Bronk,
B. S. ACS Med. Chem. Lett. 2012, 3, 908.
14. Kounnas, M. Z.; Danks, A. M.; Cheng, S.; Tyree, C.; Mao, L.; Comer, D.; Yu, C.;
Pleynet, D.; Nyugen, P.; Digregorio, P. J.; Page, N.; Ackerman, E.; Velicelebi, G.;
Stauderman, K. A.; Sisodia, S. S.; Comer, W. T.; Tanzi, R. E.; Wagner, S. L. Neuron
2010, 67, 769.
15. (a) Hantzsch, A. Ber. Dtsch. Chem. Ges. 1881, 14, 1637; (b) Hantzsch, A. Ann.
Chem. 1892, 1.
16. Wagner, S. L.; Zhang, C.; Cheng, S.; Nguyen, P.; Zhang, X.; Rynearson, K. D.;
Wang, R.; Sisodia, S. S.; Mobley, W. C.; Tanzi, R. E. Biochemistry 2014, 53, 702.
17. Ab42 levels were quantified in human SHSY5Y neuroblastoma cells stably
over-expressing human APP751 following treatment for 24 h with either
vehicle or compound. The quantity of Ab42 was then measured by ELISA, see
Supplemental information for additional experimental details.
18. Male CD-1 mice were administered a single dose of vehicle, 1 mg/kg GSM 28 by
IV or 5 mg/kg GSM 28 by oral gavage. Brain and plasma samples were collected
at 1 h, 4 h, 8 h and 24 h post-dose. See Supplemental information for additional
experimental details.
Funding sources: This work was supported by grants from the
Cure Alzheimer’s Fund; and the National Institutes of Health Grant
U01-NS074501.
Acknowledgments
19. Wagner, S. L.; Cheng, S.; Mobley, W. C.; Tanzi, R. E. U.S. Patent US20130165416
A1, 2013.
20. (a) Möhlau, R. Ber. Dtsch. Chem. Ges. 1882, 15, 2464; (b) Hanifin, J. W.; Johnson,
B. D.; Menschik, J.; Ridge, D. N.; Sloboda, A. E. J. Pharm. Sci. 1979, 68, 535; (c)
The authors would like to thank Jiunn H. Lin of Biopharm Con-
sulting Partners for his contribution to the in vivo pharmacokinetic
studies.

K. D. Rynearson et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3928–3937
3937
Ridge, D. N.; Hanifin, J. W.; Harten, L. A.; Johnson, B. D.; Menschik, J.; Nicolau,
G.; Sloboda, A. E.; Watts, D. E. J. Med. Chem. 1979, 22, 1385.
21. (a) Taylor, E. C.; Barton, J. W. J. Am. Chem. Soc. 1959, 81, 2448; (b) Adachi, I.;
Yamamori, T.; Hiramatsu, Y.; Sakai, K.; Sato, H.; Kawakami, M.; Uno, O.; Udea,
M. Chem. Pharm. Bull. 1986, 35, 3235.
22. (a) Portelius, E.; Van Broeck, B.; Andreasson, U.; Gustavsson, M. K.; Mercken,
M.; Zetterberg, H.; Borghys, H.; Blennow, K. J. Alzheimer’s Dis. 2010, 21, 1005;
(b) Nakano-Ito, K.; Fujikawa, Y.; Hihara, T.; Shinjo, H.; Kotani, S.; Suganuma, A.;
Aoki, T.; Tsukidate, K. Toxicol. Sci. 2013, 137, 249.