Chiral cyclohexane 1,3-diones as inhibitors of mutant SOD1-dependent protein aggregation for the treatment of ALS

Article abstract of DOI:10.1021/ml3000963

Cyclohexane 1,3-diones were identified as a class of molecules exhibiting a protective effect against mutant SOD1 induced toxicity in PC-12 cells, but an optimized analogue had little or no effect on life extension in the G93A SOD1 mouse model for amyotro

Full text of DOI:10.1021/ml3000963

Letter
pubs.acs.org/acsmedchemlett
Chiral Cyclohexane 1,3-Diones as Inhibitors of Mutant SOD1-
Dependent Protein Aggregation for the Treatment of ALS
Yinan Zhang,† Radhia Benmohamed,‡ Wei Zhang,† Jinho Kim,§ Christina K. Edgerly,§ Yaoqiu Zhu,∥
Richard I. Morimoto,⊥ Robert J. Ferrante,§ Donald R. Kirsch,‡ and Richard B. Silverman*,†
†Department of Chemistry, Department of Molecular Biosciences, Chemistry of Life Processes Institute, Center for Molecular
Innovation and Drug Discovery, Northwestern University, Evanston, Illinois 60208-3113, United States
‡Cambria Pharmaceuticals, Cambridge, Massachusetts 02142, United States
§Neurological Surgery, Neurology, and Neurobiology Departments, University of Pittsburgh, Pittsburgh, Pennsylvania 15213, United
States, and the Geriatric Research Educational and Clinical Center (00-GR-H), V.A. Pittsburgh Healthcare System, 7180 Highland
Drive, Pittsburgh, Pennsylvania 15206, United States
∥MetabQuest Research & Consulting, Beijing, 100871, China
⊥Department of Molecular Biosciences, Rice Institute for Biomedical Research, Northwestern University, Evanston, Illinois
60208-3500, United States
S
* Supporting Information
ABSTRACT: Cyclohexane 1,3-diones were identified as a class of molecules
exhibiting a protective effect against mutant SOD1 induced toxicity in PC-12
cells, but an optimized analogue had little or no effect on life extension in the
G93A SOD1 mouse model for amyotrophic lateral sclerosis (ALS). Additional
testing showed that these compounds were inactive in neurons, and further
analogue synthesis was carried out to identify compounds with neuronal activity.
Starting from two racemic derivatives that were active in cortical neurons, two
potent analogues (1b and 2b) were resolved, which were protective against mutant SOD1 induced toxicity in PC-12 cells. Both
compounds were found to be active in cortical neurons and presented good ADME profiles in vitro. On the basis of these results,
an ALS mouse trial with 1b was carried out, which showed slightly greater life extension than the FDA-approved ALS drug
riluzole, thereby validating cyclohexane 1,3-diones as a novel therapeutic class for the treatment of ALS.
KEYWORDS: Cyclohexane 1,3-diones, superoxide dismutase 1 (SOD1), protein aggregation, amyotrophic lateral sclerosis (ALS),
mutant SOD1, PC-12 cells, cortical neurons
myotrophic lateral sclerosis (ALS) is a rare and fatal
neurodegenerative disease characterized by progressive
neurons, indicating that SOD1 is a viable target for SALS.8
Also, because mutant SOD1 leads to oxidative stress, protein
misfolding, and aggregation, all of which are associated with
ALS pathogenesis,9 it is reasonable to include inhibitors of
mutant SOD1-induced protein aggregation as a viable strategy
to identify novel ALS therapeutics.
Three different scaffolds, arylsulfanylpyrazolones (ASP),
pyrimidine-2,4,6-triones (PYT), and cyclohexane-1,3-diones
(CHD), were identified by a high-throughput cell-based
screen10 based on cell lines developed by Morimoto and co-
workers.11 Extensive modification of the ASP12,13 and PYT14
leads afforded excellent therapeutic candidates, with favorable
potency, pharmacokinetics, toxicity, and life extension in the
ALS mouse model. However, the most potent of the CHD
derivatives did not show any significant extension of life in the
ALS mouse model, despite having comparable potency in the
PC-12 cell assay and favorable pharmacokinetic properties.15
Aggregation of mutant G93A SOD1 is induced in the PC-12
A
motor neuron loss in the central and peripheral neuron
systems, leading to clinical muscle atrophy, paralysis, and final
death from respiratory failure, generally, in 3−5 years.1 It is
estimated that the incidence of ALS is 1−2 cases per 100,000
people, with an increased risk for military personnel.2,3
Although there has been progress in the identification of
potential targets for the disease, and many new therapeutics
have been tested in animals and in clinical trials over the last
two decades,4 no effective treatment is currently available; the
only FDA-approved drug, riluzole, a presumptive antiglutama-
tergic drug, extends survival by only 2−3 months.5
Although ALS is principally a sporadic disease, approximately
10% of all cases are familial (FALS), and over 100 genes are
potentially responsible for FALS.6 Mutations in Cu/Zn
superoxide dismutase (SOD1) are the most common cause
of FALS.7 Although mutations in SOD1 account for only 2% of
ALS patients, it has recently been shown that astrocytes from
both FALS and sporadic ALS (SALS) patients are similarly
toxic to motor neurons and that knockdown of SOD1
significantly attenuates astrocyte-mediated toxicity of motor
Received: April 19, 2012
Accepted: May 22, 2012
Published: May 22, 2012
© 2012 American Chemical Society
584
dx.doi.org/10.1021/ml3000963 | ACS Med. Chem. Lett. 2012, 3, 584−587

ACS Medicinal Chemistry Letters
Letter
a
assay, which produces a concomitant loss in cell viability. Cell
viability is restored through treatment with compounds that
reduce protein aggregation. The proposed explanation was the
lack of in vitro activity in cortical neurons. Two racemic
analogues (1 and 2) were identified with enhanced activity in
cortical neurons that retained their activity in the PC-12 assay.
Here we have synthesized the enantiomers of the active
compounds and show that both enantiomers of each scaffold
penetrate cortical neurons, that the pharmacokinetics of the
eutomers are favorable, and that one of the isomers produces a
slightly greater extension of life in the ALS mouse model than
riluzole, the only FDA-approved drug for ALS.
Scheme 2. Synthesis of 2
a
Reagents and conditions: (a) Triethyl phosphonoacetate, NaH, THF,
0 °C → room temp, overnight; (b) DIBAL, DCM, 0 °C, 2 h, 67% for
the two steps; (c) CH2I2, ZnEt2, DCM, 0 °C → room temp, overnight,
88%; (d) PCC, silica gel, DCM, room temp, 3 h; (e) 1-
(triphenylphosphoranylidene)-2-propanone, THF, room temperature,
overnight, 58% for the two steps; (f) diethyl malonate, EtONa, EtOH,
room temp, overnight; (g) 2 N NaOH, room temp, 4 h; (h) 1 N HCl,
90 °C, 1 h, 47% for the three steps.
As shown in Scheme 1, starting from commercially available
ethyl lactate (3) and 3,5-ditrifluoromethyl phenol (4), the
synthesize the enantiomers of 2, starting from the enantiomers
of 11.
Compound activity was assessed using a previously described
cytotoxicity protection assay.10 The EC50 values of these
analogues are summarized in Figure 1. The potencies of the
a
Scheme 1. Synthesis of 1
a
Reagents and conditions: (a) PPh3, DEAD, THF, room temp,
overnight, 98%; (b) DIBAL, DCM, −78 °C, 1 h, 96%; (c) 1-
(triphenylphosphoranylidene)-2-propanone, THF, room temp, over-
night, 82%; (d) diethyl malonate, EtONa, EtOH, room temp,
overnight; (e) 2 N NaOH, room temp, 4 h; (f) 1 N HCl, 90 °C, 1
h, 53% for the three steps.
Figure 1. Cytotoxicity protection assay for the CHD analogues.
condensed ether (5) was formed using a Mitsunobu reaction.
Treatment with DIBAL at −78 °C provided aldehyde 6 in a
high yield, which was used directly in a Wittig reaction to afford
enone 7 in a 5:1 trans to cis ratio. A one-pot procedure, which
includes a Michael addition, cyclization, hydrolysis, and
decarboxylation, was carried out to give 1 in high yield; starting
from chiral ethyl lactates, the two enantiomers (1a and 1b)
were readily obtained.
The route shown in Scheme 2 was used to synthesize 2. 3,5-
Ditrifluoromethyl benzaldehyde (8) was treated with triethyl
phosphonoacetate and then reduced to obtain mostly trans-allyl
alcohol 9 in 67% yield. An enantioselective Simmons−Smith
cyclopropanation was performed with bifunctional boron ester
1016 in a high yield and excellent enantioselectivity. The alcohol
intermediate (11) was converted to an enone (12) by PCC
oxidation and a Wittig reaction. A one-pot procedure of a
Michael addition, cyclization, hydrolysis, and decarboxylation
was carried out to give trans-2. This method was used to
ether linker CHD (1) were superior to those of the cyclopropyl
linker compounds (2). The enantiomers of the ether linker
analogues (1a and 1b) showed a greater potency difference
than their cyclopropyl counterparts (2a and 2b); S-enantiomer
1b was 4−5-fold more potent than R-enantiomer 1a, but 2b
was only 1.5−2-fold more potent than 2a. Ether 1b was the
most potent among all of the CHD analogues tested.15
It was previously found that 13 had little or no effect on life
extension in the ALS mouse model and was not active with
cortical neurons. As shown in Figure 2, all of the compounds in
Figure 1 had cortical neuron activity except 13. Furthermore,
compound 1b exhibited more than 90% neuronal activity at 3
μM, while, as a control, the best ASP compound (see Figure 2)
required a concentration of 10 μM to reach maximum recovery.
The aqueous solubilities of 1b and 2b were evaluated by
dilution from a stock solution in DMSO to a final
585
dx.doi.org/10.1021/ml3000963 | ACS Med. Chem. Lett. 2012, 3, 584−587

ACS Medicinal Chemistry Letters
Letter
Maximum blood levels (245 μM) of 1b by ip administration
(500 mg/kg) occurred at 12 h, the blood half-life; brain
penetration was 8.3 μM with a Tmax of 12 h. As the commonly
used ALS animal model, transgenic mice expressing human
G93A mutant SOD1 develop a series of similar symptoms to
those observed in both familial and sporadic ALS patients.19
Control and transgenic mice of the same age ( 3 days) and
from the same “f” generation were selected from multiple litters
to form experimental cohorts. The tolerable dose range for 1b
was determined in wild-type mice by increasing the dose b.i.d.,
and the maximum tolerated dose was 1280 mg/kg. On the basis
of the ADME and MTD studies, the dose levels of 10, 20, and
30 mg/kg were administered daily, starting from 6 weeks of age
to the end of life of the G93A mice. Administration of 1b
resulted in a 13% extension in survival at 20 mg/kg compared
to the case of untreated G93A mice (Figure 3). This result is
slightly better than that observed for the only FDA approved
drug riluzole, which showed a lifespan extension of 10−11% at
22 mg/kg in the same animal model.20
Figure 2. Qualitative primary cortical neurons protection assay.
Reference: compound 13 in ref 13.
concentration of 1% DMSO in PBS. The solubility limit was
the highest concentration with no precipitation. Both
compounds were found to have high aqueous solubility17
(≥100 μM).
The in vitro plasma stability half-life for 1b was >60 min, and
that for 2b was 71 min. The human and mouse microsomal
stabilities of these compounds were tested at 1 μM at 37 °C for
1 h in the presence and absence of NADPH (Table 1). Both of
a
Table 1. In Vitro Microsomal Stability of 1b and 2b
NADPH-dependent
NADPH-absent
b
c
b
c
CLint
T1/2
CLint
T1/2
compd (mL min−1 kg−1
)
(min) (mL min−1 kg−1
)
(min)
human
mouse
1b
2b
1b
2b
31
36
45
82
74
64
52
28
10
12
45
63
>180
>180
52
37
a
b
Data were obtained from Apredica. Microsomal intrinsic clearance.
c
Half-life.
Figure 3. Kaplan−Meier plot of 1b-treated SOD1 G93A ALS mice:
untreated group, 125.7 4.3 days; group 1 (10 mg/kg), 139.2 8.3
days; group 2 (20 mg/kg), 142.0 9.1 days (p < 0.03); group 3 (30
mg/kg), 127.9 4.7 days (p < 0.63).
the compounds showed moderate clearance with human liver
microsomes (31−36 mL/(min kg)) and moderate to near high
clearance with mouse liver microsomes (45−82 mL/(min
kg));18 both had half-lives in human microsomes greater than 1
h. Metabolite identification studies indicated the only metabolic
product was insertion of an oxygen atom somewhere other than
on the bis(trifluormethyl)phenyl ring (see Supporting
Information).
Previously, we had prepared a compound (13) that was a
potent inhibitor of protein aggregation in PC-12 cells
expressing mutant G93A SOD1 with very good pharmacoki-
netic properties but which was inactive in vivo in the ALS
mouse model. It was found to be inactive in cortical neurons,
which led to the design of two racemic compounds (1 and 2)
that were active in cortical neurons. Chiral syntheses of the
enantiomers of 1 and 2 were carried out, and both enantiomers
of each compound were found to be active in both the PC-12
and cortical neuron assays. The eutomers of each racemic
compound (1b and 2b) had good pharmacokinetic properties,
and the more potent of these (1b) was shown to extend the life
of the ALS mouse by 13%, which is slightly better than that
previously reported for riluzole, the only FDA-approved drug
for ALS, in the same mouse model. These studies demonstrate
the importance of investigating the cortical neuron activity of
compounds prior to the expensive and time-consuming task of
an ALS mouse trial. They also validate the cyclohexane 1,3-
dione class of compounds as a potential therapeutic scaffold for
the treatment of ALS.
Compounds 1b and 2b were further evaluated for their
ability to penetrate Caco-2 cell monolayers, which is correlated
with intestinal permeability in vivo. As shown in Table 2, both
Table 2. In Vitro Caco-2 Permeability of 1b and 2ba
b
b
Papp (A→B) (10−6 Papp (B→A) (10−6 efflux ratio (B→A)/
compd
cm/S)
cm/S)
(A→B)
1b
2b
24.1
21.7
1.5
0.1
0.1
1.1
a
b
Data were obtained from Apredica. Apparent permeability.
had high permeability from the A side to the B side. Moreover,
the low efflux ratio (Papp(B→A)/Papp(A→B)) indicates these
compounds are unlikely to be substrates of efflux transport
proteins, which is especially important for CNS drugs.
Compound 1b was selected for in vivo testing on the basis of
its potency and in vitro predicted pharmacology profile.
586
dx.doi.org/10.1021/ml3000963 | ACS Med. Chem. Lett. 2012, 3, 584−587

ACS Medicinal Chemistry Letters
Letter
aggregation. Potential application for the treatment of amyotrophic
lateral sclerosis. Bioorg. Med. Chem. 2011, 19, 613−622.
ASSOCIATED CONTENT
* Supporting Information
Detailed description of experimental procedures, biological
evaluations, and characterization of 1 and 2. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
(13) Chen, T.; Benmohamed, R.; Kim, J.; Smith, K.; Amante, D.;
Morimoto, R. I.; Ferrante, R. J.; Kirsch, D.; Silverman, R. B. ADME-
guided design and synthesis of aryloxanylpyrazonlone derivatives to
block mutant superoxide dismutase 1 (SOD1) cytotoxicity and
protective aggregation: Potential application for the treatment of
amyotrophic lateral sclerosis. J. Med. Chem. 2012, 55, 515−527.
(14) Xia, G.; Benmohamed, R.; Kim, J.; Arvanites, A. C.; Morimoto,
R. I.; Ferrante, R. J.; Kirsch, D.; Silverman, R. B. Pyrimidine-2,4,6-
trione derivatives and their inhibition of mutant SOD1-dependent
protein aggregation. Toward a treatment for amyotrophic lateral
sclerosis. J. Med. Chem. 2011, 54, 2409−2421.
AUTHOR INFORMATION
Corresponding Author
*Tel: 847-491-5663. Fax: 847-491-7713. E-mail: Agman@
chem.northwestern.edu.
■
Funding
(15) Zhang, W.; Benmohamed, R.; Arvanites, A. C.; Morimoto, R. I.;
Ferrante, R. J.; Kirsch, D. R.; Silverman, R. B. Cyclohexane 1,3-diones
and their inhibition of mutant SOD1-dependent protein aggregation
and toxicity in PC12 cells. Bioorg. Med. Chem. 2012, 20, 1029−1045.
(16) Charette, A. B.; Juteau, H. Design of amphoteric bifunctional
ligands: Application to the enantioselective Simmons-Smith cyclo-
propanation of allylic alcohols. J. Am. Chem. Soc. 1994, 116, 2651−
2652.
We thank the National Institutes of Health (Grant
1R43NS057849), the ALS Association (TREAT program),
and the Department of Defense (AL093052), for their
generous support of this research.
Notes
The authors declare no competing financial interest.
(17) Kerns, E. H.; Di, L. Drug-like Properties: Concepts, Structure,
Design, and Methods; Academic Press: 2008; p 65.
(18) Nassar, A. F. Drug metabolism Handbook: Concepts and
ABBREVIATIONS
■
ADME, absorption, distribution, metabolism, excretion; ALS,
amyotrophic lateral sclerosis; CHD, cyclohexane 1,3-dione;
CNS, central nervous system; FALS, familial ALS; PBS,
phosphate buffered saline; PK, pharmacokinetics; SALS,
sporadic ALS; SOD1, Cu/Zn superoxide dismutase
applications; Wiley: 2009; p 150.
(19) Gurney, M. E.; Pu, H.; Chiu, A. Y.; Dal Canto, M. C.; Polchow,
C. Y.; Alexander, D. D.; Caliendo, J.; Hentati, A.; Kown, Y. W.; Deng,
H. X.; Chen, W.; Zhai, P.; Sufit, R. L.; Siddique, T. Motor neuron
degeneration in mice that express a human Cu, Zn superoxide
dismutase mutation. Science 1994, 264, 1772−1775.
(20) Gurney, M. E.; Cutting, F. B.; Zhai, P.; Doble, A.; Taylor, C. P.;
Andrus, P. K.; Hall, E. D. Benefit of vitamin E, riluzole, and gabapentin
in transgenic model of familial amyotrophic lateral sclerosis. Ann.
Neurol. 1996, 39, 147−157.
REFERENCES
■
(1) Rowland, L. P.; Shneider, N. A. Amyotrophic lateral sclerosis. N.
Engl. J. Med. 2001, 344, 1688−1700.
(2) Cronin, S.; Hardiman, O.; Traynor, B. J. Ethnic variation in the
incidence of ALS. Neurology 2007, 68, 1002−1007.
(3) Weisskopf, M. G.; O’Reilly, E. J.; McCullough, M. L.; Calle, E. E.;
Thun, M. J.; Cudkowicz, M.; Ascherio, A. Prospective study of military
service and mortality from ALS. Neurology 2005, 64, 32−37.
(4) Zinman, L.; Cudkowicz, M. Emerging targets and treatments in
amyotrophic lateral sclerosis. Lancet Neuro. 2011, 10, 481−490.
(5) Bensimon, G.; Lacomblez, L.; Meininger, V. A controlled trial of
riluzole in amyotrophic lateral sclerosis. ALS/Riluzole Study Group. N.
Engl. J. Med. 1994, 330, 585−591.
(6) See http://alsod.iop.kcl.ac.uk/, the ALS online genetics database.
(7) Brown, R. H., Jr.; Robberecht, W. Amyotrophic lateral sclerosis:
pathogenesis. Semin. Neurol. 2001, 21, 131−139.
(8) Haidet-Phillips, A. M.; Hester, M. E.; Miranda, C. J.; Meyer, K.;
Braun, A.; Frakes, L.; Song, S. W.; Likhite, S.; Murtha, M. J.; Foust, K.
D.; Rao, M.; Eagle, A.; Kammesheidt, A.; Christensen, A.; Mendell, J.
R.; Burghes, A. H. M.; Kaspar, B. K. Astrocytes from familial and
sporadic ALS patients are toxic to motor neurons. Nat. Biotechnol.
2012, 29, 824−830.
(9) Bosco, D. A.; Morfini, G.; Karabacak, N. M.; Song, Y.; Gros-
Louis, F.; Pasinelli, P.; Goolsby, H.; Fontaine, B. A.; Lemay, N.;
McKenna-Yasek, D.; Frosch, M. P.; Agar, J. N.; Julien, J. -P.; Brady, S.
T.; Brown, R. H., Jr. Wild-type and mutant SOD1 share an aberrant
conformation and a common pathogenic pathway in ALS. Nat.
Neurosci. 2010, 13, 1396−1403.
(10) Benmohamed, R.; Arvanites, A. C.; Silverman, R. B.; Morimoto,
R. I.; Ferrante, R. J.; Kirsch, D. R. Identification of compounds
protective against G93A SOD1 toxicity for the treatment of
amyotrophic lateral sclerosis. Amyotrophic Lateral Scler. 2011, 12,
87−96.
(11) Matsumoto, G.; Stojanovic, A.; Holmber, C. I.; Kim, S.;
Morimoto, R. I. Structure properties and neuronal toxicity of
amyotrophic lateral sclerosis-associated Cu/Zn superoxide dismutase
1 aggregates. J. Cell Biol. 2005, 171, 75−85.
(12) Chen, T.; Benmohamed, R.; Arvanites, A. C.; Ranaivo, H. R.;
Morimoto, R. I.; Ferrante, R. J.; Watterson, D. M.; Kirsch, D. R.;
Silverman, R. B. Arylsulfanyl pyrazolones block mutant SOD1-G93A
587
dx.doi.org/10.1021/ml3000963 | ACS Med. Chem. Lett. 2012, 3, 584−587