Discovery and preliminary structure-activity relationship of arylpiperazines as novel, brain-penetrant antiprion compounds

Article abstract of DOI:10.1021/ml300472n

Creutzfeldt-Jakob disease and kuru in humans, BSE in cattle, and scrapie in sheep are fatal neurodegenerative disorders. Such illnesses are caused by the conversion and accumulation of a misfolded pathogenic isoform (termed PrP ^Sc) of a normall

Full text of DOI:10.1021/ml300472n

Letter
pubs.acs.org/acsmedchemlett
Discovery and Preliminary Structure−Activity Relationship of
Arylpiperazines as Novel, Brain-Penetrant Antiprion Compounds
Zhe Li,†,‡,∥ Joel R. Gever,†,‡ Satish Rao,†,‡ Kartika Widjaja,† Stanley B. Prusiner,*,†,‡
and B. Michael Silber†,‡,§,⊥
†Institute for Neurodegenerative Diseases, ‡Department of Neurology, and §Department of Bioengineering and Therapeutic Sciences,
University of California, San Francisco, California 94143, United States
S
* Supporting Information
ABSTRACT: Creutzfeldt-Jakob disease and kuru in humans, BSE in cattle, and scrapie in sheep
are fatal neurodegenerative disorders. Such illnesses are caused by the conversion and
accumulation of a misfolded pathogenic isoform (termed PrPSc) of a normally benign, host
cellular protein, denoted PrPC. We employed high-throughput screening enzyme-linked
immunosorbent assays to evaluate compounds for their ability to reduce the level of PrPSc in
Rocky Mountain Laboratory prion-infected mouse neuroblastoma cells (ScN2a-cl3).
Arylpiperazines were among the active compounds identified, but the initial hits suffered from
low potency and poor drug-likeness. The best of those hits, such as 1, 7, 13, and 19, displayed
moderate antiprion activity with EC50 values in the micromolar range. Key analogues were
designed and synthesized on the basis of the structure−activity relationship, with analogues 41,
44, 46, and 47 found to have submicromolar potency. Analogues 41 and 44 were able to
penetrate the blood−brain barrier and achieved excellent drug concentrations in the brains of
mice after oral dosing. These compounds represent good starting points for further lead
optimization in our pursuit of potential drug candidates for the treatment of prion diseases.
KEYWORDS: Neurodegenerative diseases, prion disease, Creutzfeldt-Jakob disease, piperazine, arylpiperazine,
structure−activity relationship
rion diseases are transmissible, rapidly progressive, fatal,
of potent, metabolically stable, BBB-penetrant leads from this
Pand untreatable neurodegenerative disorders that afflict series. These compounds are excellent starting points for
both humans and animals.1−3 An early key pathogenic event in
further lead optimization, with the goal of identifying
candidates for further testing in scrapie- and CJD-infected
animal models of prion disease.
diseases caused by the prion protein (PrP) is the conversion of
the host cellular PrP, denoted PrPC, into an aberrant disease-
causing isoform, PrPSc, which accumulates in the central
nervous system (CNS) of the infected host.4−6 The exact
mechanism for the conversion of PrPC to PrPSc remains
unknown.
Currently, no treatments exist for prion diseases. Several
classes of antiprion compounds that target levels of PrPSc or
PrPC, including approved drugs (for other uses) in humans as
well as experimental small molecules, have been reported.7,8
Some FDA-approved drugs, including quinacrine, phenothia-
zines, and statins, showed antiprion activity in cell culture but
demonstrated no in vivo efficacy when tested in prion-infected
mice.9−11
We have sought a strategy to treat prion disease by lowering
the levels of PrPSc in prion-infected cell models. Reduced PrPSc
levels are correlated with decreased prion infectivity in cultured
cells and delayed development of clinical signs. Using HTS, we
identified small molecules that lowered the levels of PrPSc in an
RML prion-infected murine neuroblastoma cell line (ScN2a-
cl3),12 among which were the arylpiperazines (B. M. Silber et
al., manuscript in preparation). Herein, we report SAR analysis
and preliminary lead optimization to improve the potency and
drug-likeness of the arylpiperazines and describe the discovery
From our HTS screening efforts of 53000 compounds, we
identified 881 arylpiperazine analogues, 108 of which reduced
PrPSc levels in ScN2a-cl3 cells by ≥30%. Of these 108 actives,
58 were subjected to a confirmatory assay and 41 of them were
confirmed, i.e., reduced the level of PrPSc by ≥30% without
affecting cell viability (≤30% inhibition). A majority (n = 29) of
these compounds, including the top 22 actives, are derivatives
of N-arylpiperazine having a para methylketone or acetyl group
on the phenyl ring (1, 7, 13, 14, 18, and 19) (Table 1).
Compounds bearing other substituents on the N-phenyl ring,
regardless of whether they are electron-rich (6, 12, 17, and 22)
or electronic-deficient (5, 11, 16, and 21) or their positions on
the N-phenyl ring (compare 3 and 4, 9 and 10, 15 and 16, and
20 and 21), were found to be inactive under our assay
conditions (Table 1). This finding argues for the importance of
the 4-acetyl group of the N-aryl ring in determining the potency
of those compounds. We initially postulated that the oxygen of
the acetyl group is involved in a hydrogen bond (HB) with the
Received: December 21, 2012
Accepted: February 17, 2013
Published: March 25, 2013
© 2013 American Chemical Society
397
dx.doi.org/10.1021/ml300472n | ACS Med. Chem. Lett. 2013, 4, 397−401

ACS Medicinal Chemistry Letters
Letter
Table 1. PrPSc Reduction for Selected Arylpiperazine
Analogues
and 2-thiophenylmethylene (13 and 14), are well tolerated and
show similar antiprion activities (Table 1).
In our A-ring SAR exploration, we initially tested the
requirement for the 4-acetyl group. Compounds with various
substituents, including 4-acetyl, chloro, fluoro, and methoxy as
well as their regioisomers [11, 12, and 23−26 (Table 1 of the
Supporting Information)], were purchased and tested. All six
compounds showed ≤30% reduced PrPSc levels and EC50
values of >32 μM; thus, they were considered inactive in our
assays.
On the B-ring side, several examples of different (hetero)-
aryls representing phenyl (1), 3-pyridyl (7), thiophenyl (13),
and furanyl (19) scaffolds were selected to test the influence of
the B-ring on antiprion activity. Compounds 1, 7, 13, and 19
were found to possess only moderate activities, with EC50
values ranging from 0.71 to 4.62 μM (Supporting Information,
Table 1). For our lead compounds, we aimed for >10-fold
potency compared to those identified in the ScN2a-cl3 cell
assay. Thus, we embarked on an SAR effort to improve the
antiprion potency of the arylpiperazine series. To ensure the
decrease in the PrPSc level was not due to cytotoxicity, cell
viability in ScN2a-cl3 cells was also determined to accompany
each EC50 enzyme-linked immunosorbent assay measurement
using the fluorescent probe calcein-AM, as reported pre-
viously.13
We noted from our early SAR that the hydrogen bond
acceptor (HBA) acetyl group is clearly important for antiprion
potency of these analogues. We sought to replace it with
alternative functionality that could preserve the HB interaction
with the putative target. Initially, we surmised that a five-
membered heterocycle, such as a 1,3-oxazole, might be a good
first choice as an acetyl isostere. Both nitrogen and oxygen
atoms of the oxazole are positioned optimally and are capable
of functioning as an HBA. We also synthesized and examined
the effect of fused benzoxazole systems for possible increased
potency (Figure 1b). While the nitrogen or oxygen of the
benzoxazole ring may function as an HBA, the fused phenyl
ring may also provide additional binding affinity.
compd
R1
X
PrPSc reduction (%)
1
4-MeCO-
H
CH
CH
CH
CH
CH
CH
N
94
5
2
3
3-Cl
7
4
4-Cl
6
5
4-F
7
6
4-MeO
4-MeCO-
H
7
7
73
4
8
N
9
3-Cl
N
4
10
11
12
4-Cl
N
0
4-F
N
27
10
4-MeO
N
compd
R2
Y
R3
PrPSc reduction (%)
13
14
15
16
17
18
19
20
21
22
4-MeCO-
4-MeCO-
3-Cl
S
H
91
78
34
12
16
96
78
6
S
4-Me
3-Br
2-Br
H
S
4-Cl
S
4-MeO
4-MeCO-
4-MeCO-
3-Cl
S
O
O
O
O
O
H
2-Me
H
4-Cl
H
13
6
4-MeO
2-Br
To ensure adequate coverage on the B-ring side, three aryl
and heteroaryl rings, including 3-pyridine, 2-thiophene, and 5-
methyl-2-furan, were selected for these initial SAR studies. A
total of three oxazoles (38−40) and three benzoxazoles (41−
43) were synthesized (Table 2 and Schemes 1 and 2). Almost
without exception, we found that arylpiperazine analogues did
not affect cell viability in ScN2a-cl3 cells (Table 2).
putative target protein (Figure 1a). Conversely, the substitution
requirement on the N′ atom of the piperazine is less stringent,
and both arylmethylenes and heteroarylmethylenes, including
benzyl (1), 3-pyridinylmethylene (7), 2-furanylmethylene (19),
With respect to antiprion potency, oxazoles 38−40 were
found to be only weakly active, with EC50 values of ≥10 μM,
suggesting that a simple replacement of the HBA group by an
oxazole ring was not sufficient to maintain potency. The
additional phenyl ring, as seen in benzoxazoles 41−43 (Table
2), was beneficial to antiprion activity: all three analogues
showed EC50 values of <10 μM. The 3-pyridyl analogue 41 had
an EC50 of 0.25 μM, which was ≥13-fold more potent than
both furanyl 42 and thiophenyl 43 (Table 2). While we cannot
specify the exact reason for such a potency increase for 41
versus those of other similar analogues with different B-rings,
the pyridyl nitrogen atom in 41 may be situated to establish a
second HB with the putative molecular target, contributing to
its antiprion activity (Figure 1c).
To evaluate the BBB permeability of 41 and its potential as a
CNS drug lead, we administered it to mice using a previously
reported pharmacokinetic (PK) protocol.14 Rather than
creating a full PK profile, we utilized an abbreviated protocol
Figure 1. (a) Arylpiperazine lead, showing HB interaction between its
4-acetyl group and the putative target. (b) Design of piperazine
analogues, via replacement of acetyl with oxazole or benzoxazole. (c)
Compound 41, showing its HB interactions with the putative target.
398
dx.doi.org/10.1021/ml300472n | ACS Med. Chem. Lett. 2013, 4, 397−401

ACS Medicinal Chemistry Letters
Letter
a
Table 2. Antiprion Potencies (EC50) and Cell Viabilities
(LC50) for Selected Arylpiperazine Analogues
Scheme 2. Synthesis of Compounds 41−48
a
Reaction conditions: (a) (i) 2-aminophenol or 2-amino-4-methyl-
phenol, O2, TEMPO, toluene, reflux overnight, 23−81%; (ii) TFA/
DCM, room temperature, 3 h, 66−77%; (b) (i) tert-butyl
piperazinecarboxylate/t-BuOK, Pd2(dba)3 (5% mmol), BINAP (15%
mmol), toluene, reflux, 12 h, 60%; (ii) TFA/DCM, 25 °C, 3 h, 85%;
(c) RCHO/HCOOH (2 equiv), anhydrous DMF, 100 °C, 3 h, 6−
44%.
synthesized a methylated benzoxazole [48 (Scheme 2)] to
probe the size constraints of the A-ring binding pocket.
Compound 48 was less potent than 41, with an EC50 of 1.65
μM, indicating that substitution of benzoxazole with a methyl
group (48) was not well-tolerated (data not shown).
We also examined the effect of hydrophobicity on binding
affinity by replacing the benzoxazole ring (41) with a more
hydrophobic benzothiazole ring [44 (Scheme 2)]. The sulfur
atom of 44 is not capable of acting as an HBA, a role assumed
by the oxygen atom in the benzoxazole 41; the outcome of this
compound may allow us to determine which atom (nitrogen or
oxygen) in 41 is involved in the HB with the putative target.
Replacing the oxygen atom in 41 with a sulfur atom (44) had
little impact on antiprion activity: 44 was slightly less potent
than 41, with an EC50 of 0.38 μM (Table 2). Presumably, 44
still maintains a HB with the putative target via the nitrogen
atom of the benzothiazole ring.
a
Three tests for EC50 and LC50 measurements.
a
Scheme 1. Synthesis of Oxazole Analogues 38−40
On the B-ring side, we believe the nitrogen atom of 41 is
involved as an HBA with the putative target. Addition of a
substituent next to the nitrogen may perturb this interaction
and thus influence the potency of these analogues. Additionally,
because the nitrogen in the pyridine ring is often a target for
cytochrome P450 (CYP450) enzymes, the introduction of an
adjacent group, in addition to modulating the HB potential and
its potency, may also improve the metabolic stability of these
pyridine analogues. Therefore, we designed and synthesized
several analogues with substituents (45−47) next to the
nitrogen atom of the pyridine ring (Scheme 2).
4-Methyl analogue 45 was found to be inactive, with an EC50
value of >10 μM (Table 2); the B-ring binding pocket should
be wide enough to accommodate a small methyl group based
on an early SAR derived from phenyl, thiophene, and furan
analogues. As a result, the loss of the antiprion potency of 45
may be due to interference of HB formation between the
adjacent nitrogen and its target. In contrast, both 4-F and 4-
MeO analogues (46 and 47, respectively) were active, with
EC50 values of 0.38 and 1.11 μM, respectively (Table 2). 4-
Fluoropyridine analogue 46 was 1.5-fold less potent than its
parent pyridine analogue 41. This observation might be
explained by a reduction in the strength of the HB between
the nitrogen and the putative target; the neighboring fluoro
atom reduces the electron density on the nitrogen atom and
thus weakens the strength of the HB. 4-MeO analogue 47 was
a
Reaction conditions: (a) (i) tert-butyl piperazinecarboxylate/t-BuOK,
Pd2(dba)3 (5% mmol), BINAP (15% mmol), toluene, reflux, 12 h,
70%; (ii) TFA, DCM, room temperature, 3 h, 93%; (b) RCHO/
HCOOH (2 equiv), anhydrous DMF, 100 °C, 3 h, 21−37%.
to determine key parameters, including Cmax and AUC, which
are important for initial compound selection and prioritization.
For initial PK studies, all compounds were administered in a
single dose (10 mg/kg) by oral gavage in a formulation
containing 20% propylene glycol, 5% ethanol, 5% labrosol, and
70% PEG400. Brain and plasma concentrations were measured
at 0.5, 2, 4, and 6 h after administration. The Cmax and the AUC
from 0 to 6 h (AUC0−6h) for both brain and plasma were
calculated. Compound 41 exhibited an excellent brain
concentration [Cmax = 3.63 μM (Table 3)]. Its very good
brain exposure (AUC0−6h = 14.2 μM h) was approximately 4-
fold higher than its plasma exposure (AUC0−6h = 4.18 μM h), a
desired characteristic in a CNS drug. Compound 41 was better
absorbed considering its poor microsomal stability [t1/2 = 9.2
min (Table 3)].
Encouraged by the in vitro antiprion potency and in vivo PK
profile, we decided to expand the SAR around lead 41, hoping
to identify more potent analogues. On the A-ring side, we
399
dx.doi.org/10.1021/ml300472n | ACS Med. Chem. Lett. 2013, 4, 397−401

ACS Medicinal Chemistry Letters
Letter
Table 3. In vivo PK Parameters and in vitro Microsomal Stability for Arylpiperazines (single oral 10 mg/kg dose)
brain exposure
AUC (μM h)
plasma exposure
AUC (μM h)
microsomal stability, t1/2 (min)
compd
Cmax (μM)
Cmax (μM)
Mo
Hu
41
44
46
47
3.6 0.9
3.0 0.5
0.4 0.1
1.6 0.8
14.2 0.6
14.4 0.2
1.3 0.2
3.0 1.0
0.9 0.1
0.6 0.0
0.4 0.2
0.7 0.4
4.2 0.6
2.7 0.4
1.2 0.1
1.1 0.5
9.2
17.3
7.3
50.2
34.3
43.6
25.2
>60
Funding
4.5-fold less potent than 41. The larger methoxy group may
hinder and weaken the HB forming ability or strength of the
nitrogen with the target.
This work was funded by National Institutes of Health Grants
AG021601, AG031220, AG002132, and AG010770 and by gifts
from the Sherman Fairchild Foundation, Lincy Foundation,
Rainwater Charitable Foundation, and Schott Foundation for
Public Education.
Next, we tested the metabolic stability of compounds 44, 46,
and 47 in mouse and human microsomes (Table 3). Compared
to benzoxazole 41, benzothiazole 44 had approximately 2-fold
improvement (t1/2 = 17.3 and 9.2 min for 44 and 41,
respectively) in mouse microsomes but was less stable in
human microsomes (t1/2 = 34.3 and 50.2 min for 44 and 41,
respectively). 4-Fluoro analogue 46 was less stable in both
mouse and human microsomes than 41. The fluorine atom
could affect the oxidation of the pyridine ring, but it also
increases the overall hydrophilicity (ChemDraw calculated logP
values of 3.86 and 4.09 for 41 and 46, respectively) of the
molecule and may open the possibility of oxidation in other
parts of the molecule. Methoxy analogue 47 had an
approximately 3-fold increase in stability in both human and
mouse microsomes.
We then profiled compounds 44, 46, and 47 via in vivo PK
studies as described above, focusing on their BBB permeability
and brain exposure (Table 3). Both B-ring analogues 46 and 47
had poor exposure in brain and plasma compared to their
parent pyridine analogue 41. Benzothiazole analogue 44,
however, displayed a favorable PK profile compared to that
of 41, with a 5-fold brain:plasma exposure ratio.
In summary, the initial arylpiperazine HTS hits with low
micromolar activity (2.94 μM for 7) were optimized to yield
several leads with a >10-fold antiprion potency improvement.
The acetyl group in the original HTS hits was replaced with the
more druglike benzoxazole ring. These efforts culminated with
the identification of 41 and 44, both demonstrating robust
antiprion potency, favorable PK properties, and good CNS
penetration. Further evaluation of these compounds in RML-
and CJD-infected animal models is underway.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Mr. Phillip Benner for preparing animal dosing and
collecting samples in PK studies, Dr. Kurt Giles and the staff of
the Hunter’s Point animal facility for animal studies, and Drs.
John Nuss and Robert Wilhelm for reviewing the manuscript.
ABBREVIATIONS
■
BBB, blood−brain barrier; BINAP, 2, 2′-bis-
(diphenylphosphino)-1,1′-binaphthyl; CJD, Creutzfeldt-Jakob
disease; DCM, dichloromethane; EC50, half-maximal effective
concentration; HBA, hydrogen bond acceptor; HBD, hydrogen
bond donor; HTS, high-throughput screening; Pd2(dba)3,
tris(dibenzylideneacetone)dipalladium(0); PrPC, normal cellu-
lar isoform of the prion protein; PrPSc, pathogenic isoform of
the prion protein; RML, Rocky Mountain Laboratory; SAR,
structure−activity relationship; TEMPO, (2,2,6,6-tetramethyl-
piperidin-1-yl)oxyl or (2,2,6,6-tetramethylpiperidin-1-yl)-
oxidanyl; TFA, trifluoroacetic acid.
REFERENCES
■
(1) Collinge, J. Molecular neurology of prion disease. J. Neurol.,
Neurosurg. Psychiatry 2005, 76, 906−919.
(2) Aguzzi, A.; Sigurdson, C.; Heikenwaelder, M. Molecular
mechanisms of prion pathogenesis. Annu. Rev. Pathol. 2008, 3, 11−40.
(3) Prusiner, S. B. A unifying role for prions in neurodegenerative
diseases. Science 2012, 336, 1511−1513.
(4) De Gioia, L.; Selvaggini, C.; Ghibaudi, E.; Diomede, L.; Bugiani,
O.; Forloni, G.; Tagliavini, F.; Salmona, M. Conformational poly-
morphism of the amyloidogenic and neurotoxic peptide homologous
to residues 106−126 of the prion protein. J. Biol. Chem. 1994, 269,
7859−7862.
(5) Hegde, R. S.; Tremblay, P.; Groth, D.; Prusiner, S. B.; Lingappa,
V. R. Transmissible and genetic prion diseases share a common
pathway of neurodegeneration. Nature 1999, 402, 822−826.
(6) Legname, G.; DeArmond, S. J.; Cohen, F. E.; Prusiner, S. B.
Pathogenesis of prion diseases. In Protein Misfolding, Aggregation, and
Conformational Diseases; Uversky, V. N., Fink, A. L., Eds.; Springer:
New York, 2007; pp 125−146.
(7) Trevitt, C. R.; Collinge, J. A systematic review of prion
therapeutics in experimental models. Brain 2006, 129, 2241−2265.
(8) Weissmann, C.; Aguzzi, A. Approaches to therapy of prion
diseases. Annu. Rev. Med. 2005, 56, 321−344.
(9) Collins, S. J.; Lewis, V.; Brazier, M.; Hill, A. F.; Fletcher, A.;
Masters, C. L. Quinacrine does not prolong survival in a murine
Creutzfeldt-Jakob disease model. Ann. Neurol. 2002, 52, 503−506.
(10) Barret, A.; Tagliavini, F.; Forloni, G.; Bate, C.; Salmona, M.;
Colombo, L.; De Luigi, A.; Limido, L.; Suardi, S.; Rossi, G.; Auvre, F.;
ASSOCIATED CONTENT
■
S
* Supporting Information
Table 1, experimental procedures, and synthetic procedures for
all intermediates. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Phone: (415) 476-4482. Fax: (415) 476-8386. E-mail:
stanley@ind.ucsf.edu.
Present Addresses
∥Z.L.: Global Blood Therapeutics, Inc., South San Francisco,
CA 94080.
⊥B.M.S.: ELMEDTECH, LLC, San Francisco, CA 94123.
Author Contributions
All authors have given approval to the final version of the
manuscript.
400
dx.doi.org/10.1021/ml300472n | ACS Med. Chem. Lett. 2013, 4, 397−401

ACS Medicinal Chemistry Letters
Letter
Adjou, K. T.; Sales, N.; Williams, A.; Lasmezas, C.; Deslys, J. P.
Evaluation of quinacrine treatment for prion diseases. J. Virol. 2003,
77, 8462−8469.
(11) Nakajima, M.; Yamada, T.; Kusuhara, T.; Furukawa, H.;
Takahashi, M.; Yamauchi, A.; Kataoka, Y. Results of quinacrine
administration to patients with Creutzfeldt-Jakob disease. Dementia
Geriatr. Cognit. Disord. 2004, 17, 158−163.
(12) Ghaemmaghami, S.; May, B. C. H.; Renslo, A. R.; Prusiner, S. B.
Discovery of 2-aminothiazoles as potent antiprion compounds. J. Virol.
2010, 84, 3408−3412.
(13) May, B. C. H.; Zorn, J. A.; Witkop, J.; Sherrill, J.; Wallace, A. C.;
Legname, G.; Prusiner, S. B.; Cohen, F. E. Structure-activity
relationship study of prion inhibition by 2-aminopyridine-3,5-
dicarbonitrile-based compounds: Parallel synthesis, bioactivity and in
vitro pharmacokinetics. J. Med. Chem. 2007, 50, 65−73.
(14) Gallardo-Godoy, A.; Gever, J.; Fife, K. L.; Silber, B. M.; Prusiner,
S. B.; Renslo, A. R. 2-Aminothiazoles as therapeutic leads for prion
diseases. J. Med. Chem. 2011, 54, 1010−1021.
401
dx.doi.org/10.1021/ml300472n | ACS Med. Chem. Lett. 2013, 4, 397−401