Synthesis and evaluation of N-(4-benzylphenyl)piperazines as VGF inducers

Article abstract of DOI:10.1016/j.bmcl.2018.05.047

A series of compounds was discovered that induce the production of VGF mRNA in SH-SY5Y cells and exhibit cytoprotection under tunicamycin induced endoplasmic reticulum (ER) stress. The aminophenol ring and linker chain of the template SUN N8075 (1) was modified to yield compounds with higher efficacy and lower propensity for adverse effects.

Full text of DOI:10.1016/j.bmcl.2018.05.047

Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and evaluation of N-(4-benzylphenyl)piperazines as VGF inducers
⁎
Tetsuya Tobaa,b, , Ryosuke Suzukia,c, Junko Futamura-Takahashid, Yoshito Kawamotob,
Shigeki Tamurab, Mariko Kurodab, Yoshiari Shimmyoe, Michinori Kadokurab, Kazumichi Gotob,
Teruyoshi Inouec, Tsuyoshi Mutob, Hirokazu Annourab
Asubio Pharma Co., Ltd., 6-4-3 Minatojima-Minamimachi, Chuo-ku, Kobe 650-0047, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
VGF inducer
Cytoprotection
Structure-activity relationship
SUN N8075
A series of compounds was discovered that induce the production of VGF mRNA in SH-SY5Y cells and exhibit
cytoprotection under tunicamycin induced endoplasmic reticulum (ER) stress. The aminophenol ring and linker
chain of the template SUN N8075 (1) was modified to yield compounds with higher efficacy and lower pro-
pensity for adverse effects.
SH-SY5Y
VGF (non-acronymic) is a neuropeptide precursor whose proteolysis
fragments are involved in a variety of functions including the regulation
of energy homeostasis, reproductive mechanisms, synaptic plasticity,
and neuronal apoptosis.1 Among its myriad of functions, VGF has at-
tracted attention for its involvement in amyotrophic lateral sclerosis
(ALS), a fatal progressive neurodegenerative disorder affecting motor
neurons which eventually results in respiratory deficiency. VGF levels
are reduced in the plasma and fibroblasts2 of ALS patients, as well as in
cerebrospinal fluid (CSF) which are in parallel with development of ALS
symptoms.3 The Hara group reported that VGF depletion participates in
the onset and/or progression of ALS.4 VGF gene expression can be in-
duced by neurotrophins such as nerve growth factor (NGF),5 brain-
derived neurotrophic factor (BDNF),6 and neurotrophin-3 (NT3),7 but
to the best of our knowledge, SUN N8075 (1)4 and clioquinol (2)8 are
the only small-molecule inducers of VGF reported to date (Fig. 1).
Compound 1, a dual Na+/Ca2+ channel blocker and antioxidant,9 in-
duces the production of VGF peptides, rescues human neuroblastoma
(SH-SY5Y) cells from tunicamycin induced endoplasmic reticulum (ER)
stress, and prolongs the life span of in vivo mouse ALS models with
SOD1 mutation.4 Although 1 is an attractive candidate for the ALS
therapy, it contains the chemically and biologically unstable p-amino-
phenol moiety. Drugs containing aminophenol often exhibit hepato-
toxicity,10–13 and this structure is a known substrate that can cause
mechanism-based inactivation (MBI) of CYP450 enzymes,11 possibly
leading to adverse drug-drug interactions.
Our goal was to discover novel compounds based on SUN N8075
scaffold with a refined profile. The effect of 1 on the survival of SOD1-
G93A transgenic mice was comparable to that of riluzole,4,14,15 the only
FDA-approved ALS treatment until the recent approval of Radicava® in
2017. Thus, the main objective of this study was to transform 1 into
compounds with a reduced risk for adverse effects while retaining/
augmenting its efficacy. In this letter, the syntheses of multiple deri-
vatives of 1 as VGF inducers are disclosed, and their structure-activity
relationship is discussed.
Compounds 5–8 were synthesized in a similar manner to 19 by
coupling the intermediate epoxide 3 and 1-(4-(4-fluorobenzyl)phenyl)
piperazine (4) in 2-propanol under reflux (Scheme 1). Compound 10
was synthesized by coupling 4 with intermediate 9 prepared from 6-
hydroxyquinoline and 1,3-dibromopropane. Compounds 18–27 with a
fluorinated linker were prepared as shown in Scheme 2. The fluorinated
malonic acid esters 11a and b were reduced to their corresponding
diols 12, which were mono-protected with tert-butyl diphenyl silyl
(TBDPS) group to afford 13. The free hydroxy group of 13 was triflated
and reacted with 4 to yield intermediate 15, which were subsequently
desilylated, triflated, and reacted with various aromatic hydroxy groups
to afford target compounds 18–27. The (S)-enantiomer of 23 was pre-
pared from commercially available (S)-glycidyl trityl ether (29)
(Scheme 3). Compounds 28 and 29 were heated to give a propanetriol
⁎
Corresponding author.
E-mail address: toba.tetsuya.zh@daiichisankyo.co.jp (T. Toba).
These authors contributed equally to this work.
Present address: Shinagawa R&D Center, Daiichi Sankyo Co., Ltd., Tokyo 140-8710, Japan.
Present address: Head Office, Daiichi Sankyo Co., Ltd., Tokyo 103-8426, Japan.
Deceased, September 14, 2017. This paper is dedicated to the memory of Ms. Junko Futamura-Takahashi.
Present address: Shonan Research Center, Takeda Pharmaceutical Co., Ltd., Kanagawa 251-8555, Japan.
a
b
c
d
e
https://doi.org/10.1016/j.bmcl.2018.05.047
Received 29 March 2018; Received in revised form 23 May 2018; Accepted 24 May 2018
0960-894X/©2018ElsevierLtd.Allrightsreserved.
Pleasecitethisarticleas:Toba,T.,Bioorganic&MedicinalChemistryLetters(2018),https://doi.org/10.1016/j.bmcl.2018.05.047

T. Toba et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Fig. 1. Structures of known VGF inducers.
Scheme 3. Reagents and conditions: (a) K2CO3, EtOH, reflux, 45%; (b) Deoxo-
fluor®, CH2Cl2, rt; (c) TFA, CH2Cl2, rt, 80% (2 steps); (d) i: Tf2O, iPr2NEt,
CH2Cl2, −78 to −20 °C, ii: 4, CH2Cl2, rt, 80%.
human liver microsomes and test compounds with and without
NADPH.18 Their difference shows the degree of inactivation by the
metabolized test compound during preincubation, proceeding only in
the presence of NADPH. Since MBI is caused by reactive metabolite
formation at the active site of CYP450, metabolic stability towards
human liver microsomes (HLMs) in vitro was also evaluated.19 In the
cell-based VGF mRNA induction and cytoprotection assays, the efficacy
of test compounds was normalized to that of 1. For the cytoprotection
assay, the mean viability of tunicamycin-treated cells throughout the
study was 45% (standard deviation [SD] 6.6%) of untreated control
cells, and the mean viability with additional treatment of 1 at 3 μM was
82% (SD 11%) of untreated controls. The protection from tunicamycin
treatment by the reference compound 1 was defined 100%, and those
by the test compounds were expressed as the ratio to 1. For the VGF
induction assay, mean increase in VGF mRNA throughout the study by
1 at 3 μM was 5.1-fold (SD 2.8) over untreated control. The results are
listed in Table 1.
Scheme 1. Reagents and conditions: (a) 2-PrOH, 50 °C to reflux, 19% (6), 70%
(7), 79% (8), then 4 N HCl/dioxane for 5, 81%; (b) 4, K2CO3, DMF, rt, then 4 N
HCl/EtOAc, 86%.
derivative 30, which was deoxyfluorinated with Deoxo-Fluor®. Detri-
tylation followed by triflation and coupling with 4 yielded the free base
of (S)-23 in a 97% ee. The enantiomer (R)-23 was prepared along with
(S)-23 by the HPLC resolution of racemic 23 (free base) using CHIRAL-
PAK® IC column. The stereochemistry was determined by comparing
retention times with (S)-23 prepared as above.
Compound 1 showed low metabolic stability and moderate risk of
MBI in the CYP450 assay. Compound 5 with an unsubstituted phenyl
ring exhibited a marked decrease in VGF induction. Together with the
complete loss of action with the truncation of the phenyl ring (not
shown), an aromatic ring and polar functional group were both
The analogs prepared were evaluated in vitro for their VGF mRNA
induction capabilities by quantitative real-time PCR16 in parallel with
cytoprotection under tunicamycin induced ER stress in SH-SY5Y cells.17
The MBI risk was evaluated by the ratio of metabolite formation of
testosterone (a probe substrate for CYP3A4) after preincubation of the
Scheme 2. Reagents and conditions: (a) NaBH4, THF, 50% aq. EtOH, 0 °C to rt, 85% (12a), 44% (12b); (b) NaH, TBDPSCl, THF, 0 °C to rt, 52% (13a), 51% (13b); (c)
Tf2O, iPr2NEt, CH2Cl2, -70 to -20 °C; (d) 4, iPr2NEt, DMF, 80 °C, 53% (15a, 2 steps), 83% (15b, 2 steps); (e) TBAF, THF, rt, 86% (16a), 87% (16b); (f) Tf2O, iPr2NEt,
CH2Cl2, -65 to -30 °C, 92% (17a); (g) R2OH, K2CO3, DMF, rt, 17–94%; (h) R2OH, Cs2CO3, DMF, 0 °C to rt, then 4 N HCl/EtOAc, CHCl3, 53–94%; (i) R2OH, Cs2CO3,
DMF, rt, then TFA, CH2Cl2, rt, then 4 N HCl/EtOAc, CHCl3, 71%.
2

T. Toba et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 1
VGF induction, cytoprotection, MBI risk and metabolic stability of compounds 1, 5–8, 10, and 18–19.
Compound
R1
R2
R3
H
VGF induction (%)a
Cell viability (%)a
MBI indexb
(% remaining)
Metabolic rate
(pmol/min/mg protein)
Mean
SD
Mean
SD
1c
OH
(100)
(100)
57
153
5d
6
OH
OH
H
H
8
26
43
33
11
12
NTe
88
28
27
32
41
41
7
OH
H
30
43
8
59
79
8
OH
H
H
H
H
F
36
52
77
68
21
10
8
64
75
89
67
27
7
70
94
82
93
122
47
10d
18d
19d
F
6
79
F
31
13
34
a
At 3 μM, in reference to 1.
b
c
The remaining activity of CYP3A4 after preincubation with the test compound and NADPH.
MsOH salt.
HCl salt.
NT: Not tested.
d
e
assumed to be essential for proper function. Among the several het-
eroatom-bearing aromatic derivatives, quinoline derivative 8 showed
moderate efficacy in VGF induction and cytoprotection. Removal of the
hydroxy group in the linker retained the activity, showed improved
stability, and decreased the risk for MBI (compound 10). Substitution of
the hydroxy group with a fluorine atom led to further potentiation
(compound 18). Addition of another fluorine, to form gem-difluoro
compound 19, showed good stability comparable to the non-substituted
linker, but was not as effective as 18 in the induction of VGF.
The potential of aromatic ring derivatization was further in-
vestigated with the monofluoride linker (Table 2). The aminophenol
ring same as 1 showed a similar but slightly improved profile (com-
pound 20). Although the trimethylphenyl ring without the amino group
resulted in decreased efficacy, the MBI risk became negligible, in-
dicating that the p-aminophenol ring is primarily responsible for the
MBI risk (compound 21). The dimethylpyridine analog 23 showed
comparable activity towards VGF induction to the quinoline derivative
18, while analog 22 with unsubstituted pyridine showed negligible
induction. Although 22 lost its activity, its metabolic stability was
significantly improved, indicating that the methyl group on the ring is
another metabolic site. Enantiomeric (R)-23 and (S)-23 showed similar
profiles when racemic 23 was optically resolved, but the metabolic
stability of (S)-23 was somewhat higher than that of (R)-23. Benzene-
sulfonamide (25) but not benzamide (24) was effective as a surrogate
for the quinoline ring and exhibited sufficient stability. The mono-
methylated benzenesulfonamide derivative 26 showed a marked en-
hancement in VGF induction/cytoprotection equipotent to 1, with
greatly improved metabolic stability and negligible MBI risk. The
trimethyl derivative 27 was found to be the most potent VGF inducer
exceeding the level of 1.
It should be noted that VGF induction and cytoprotective activity
increased in a somewhat parallel manner. The expression of VGF re-
presents the major mechanism of action of 1 as indicated by siRNA
experiments.4 From the result presented herein it is reasonable to as-
sume that the VGF induction in this series contributes to the neuronal
cell protection under ER stress. On the other hand, even the cytopro-
tection of the most potent compounds did not exceed that of 1. Cyto-
protection by 1 has likely reached the maximum level detectable by the
assay used in this study, presumably with the participation of other
mechanisms.
In conclusion, we prepared a series of compounds that induce VGF
mRNA production in SH-SY5Y cells. The compounds showed near-
parallel protection of the SH-SY5Y cells under tunicamycin induced ER
stress. The p-aminophenol ring in 1 was shown to be responsible for the
MBI risk, but the aromatic ring and polar functional group(s) were
determined to be necessary for the efficacy of the drug. The methyl
group on the phenyl ring is one of the main metabolic site, but has a
positive effect on cell protection. Removal of the hydroxy group on the
linker or substitution with a fluorine atom led to further potentiation
and ameliorated the MBI risk. Through the modification of the ami-
nophenol ring and linker chain in 1, compounds such as 2720 were
obtained with higher efficacy and a reduced propensity for hepato-
toxicity and/or drug-drug interaction. Further structural optimization
of these VGF inducers including the diphenylmethyl moiety is un-
derway, and will be reported in due course.
3

T. Toba et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 2
VGF induction, cytoprotection, MBI risk and metabolic stability of compounds 20–27.
Compound
R1
VGF induction (%)a
Cell viability (%)a
MBI indexb
Metabolic rate
Mean
SD
Mean
SD
(% remaining)
(pmol/min/mg protein)
20c
126
9
94
9
7
75
106
21c
75
13
60
107
25
22c
23c
13
79
12
10
13
82
5
8
101
67
< 10
41
(R)-23c
(S)-23c
24
78
68
20
14
25
5
93
96
21
27
13
5
83
80
76
71
37
34
25
26
71
28
88
10
99
12
17
114
16
105
14
108
27
172
13
106
15
92
32
a
At 3 μM, in reference to 1.
The remaining activity of CYP3A4 after preincubation with the test compound and NADPH.
HCl salt.
b
c
Acknowledgements
7. Eagleson KL, Fairfull LD, Salton SRJ, Levitt P. J Neurosci. 2001;21:9315–9324.
8. Katsuyama M, Ibi M, Matsumoto M, Iwata K, Ohshima Y, Yabe-Nishimura C. J
Pharmacol Sci. 2014;124:427–432.
9. Annoura H, Nakanishi K, Toba T, et al. J Med Chem. 2000;43:3372–3376.
10. Vermeulen NPE, Bessems JGM, Van de Straat R. Drug Metab Rev. 1992;24:367–407.
11. Teng WC, Oh JW, New LS, et al. Mol Pharmacol. 2010;78:693–703.
12. Kalgutkar AS, Dalvie D, Obach RS, Smith DA. Reactive drug metabolites. John Wiley &
Sons; 2012.
The authors would like to thank Ms. S. Kanki-Mekata and Mr. T.
Tabata for performing metabolic stability assays. Drs. Y. Minamitake, T.
Nishihara, Y. Kita, H. Ogasawara, Y. Fukuda, A. Ogata, and M.
Hasegawa are acknowledged for valuable discussions and support
throughout this study.
13. Ongarora DSB, Gut J, Rosenthal PJ, Masimirembwa CM, Chibale K. Bioorg Med Chem
Lett. 2012;22:5046–5050.
14. Waibel S, Reuter A, Malessa S, Blaugrund E, Ludolph AC. J Neurol.
2004;251:1080–1084.
A. Supplementary data
15. Del Signore SJ, Amante DJ, Kim J, et al. Amyotroph Lateral Scler. 2009;10:85–94.
16.. In vitro VGF mRNA induction assay: To examine the effect of the SUN N8075 analog
on VGF mRNA expression, SH-SY5Y cells were seeded in 12-well plates at a density of
2.0 × 105 cells per well. After the cells were incubated for 72 h, they were exposed to
the various compounds at a concentration of 3 μM in DMEM/F12 with 1% FBS for
24 h. RNA was isolated using RNeasy mini kit (QIAGEN), cDNA was produced using
QuantiTect Reverse Transcription Kit (QIAGEN), and quantitative real-time PCR was
performed using the TaqMan Gene Expression Assay (VGF: Hs00705044-s1), TaqMan
Gene Expression Master Mix (Thermo Fisher Scientific), and Mx3000P QPCR system
(Agilent Technologies). Gene expression was normalized relative to the GAPDH in-
ternal control gene and assessed using the comparative CT method.
17.. In vitro cytoprotection assay: To assess the protective effect of the prepared com-
pounds against ER stress-induced cell death, SH-SY5Y cells (ECACC) were suspended
in Dulbecco's Modified Eagle Medium (DMEM, Thermo Fisher Scientific) containing
15% fetal bovine serum (FBS, EQUITECH-BIO), and 50 U/mL of penicillin-
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.bmcl.2018.05.047.
References
1. Ferri GL, Noli B, Brancia C, D’Amato F, Cocco C. J Chem Neuroanat. 2011;42:249–261
and references cited therein.
2. Brancia C, Noli B, Boido M, et al. PLoS One. 2016;11:e0164689.
3. Zhao Z, Lange DJ, Ho L, et al. Int J Med Sci. 2008;5:92–99.
4. Shimazawa M, Tanaka H, Ito Y, et al. PLoS One. 2010;5:e15307.
5. Levi A, Eldridge JD, Paterson BM. Science. 1985;229:393–395.
6. Alder J, Thakker-Varia S, Bangasser DA, et al. J Neurosci. 2003;23:10800–10808.
4

T. Toba et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
streptomycin (Thermo Fisher Scientific). The cells were then cultured using a col-
NADPH to that measured without NADPH.
lagen type-I coated 96-well plate (Greiner) at a density of 1 × 104 cells/well in 5%
CO2 incubated at 37 °C. After 24 h, the culture media were changed to DMEM con-
taining 1% FBS and 50 U/mL of penicillin-streptomycin, then the prepared com-
pounds were added to the cells at a final concentration of 1–10 μM. After incubation
with the compounds for 1 h, the cells were treated with tunicamycin at a final
concentration of 2 μg/mL. To assess cell viability, a 10% WST-8 solution was added
to the cells and incubated for 4 h in a CO2 incubator 48 h after the tunicamycin
treatment. The O.D. absorbance was recorded at 450 nm with background absor-
bance at 650 nm using a SpectraMax (Molecular Devices) instrument. The inhibition
of ER-stress induced cell death for each compound was calculated from the mean of
at least three experiments.
19.. Metabolic stability assay: For the incubation of the mixtures containing 1 μM test
compound, 0.2 mL of HLMs (0.5 mg of microsomal protein/mL) and phosphate buffer
(pH 7.4) were prepared in a 96-well plate. Reactions were initiated by the addition of
NADPH (final concentration of 1 mM) and kept in a shaking water bath at 37 °C.
Reactions were terminated by adding 50 μL of the incubation mixture to 100 μL of
acetonitrile containing 100 ng/mL of an antipurine internal standard. The sampling
point for t = 0 min was taken immediately after NADPH addition, and additional
samples were taken at 3, 10, and 30 min. The samples were subsequently centrifuged
for 5 min at 3000 rpm to pellet the precipitated microsomal protein, and the su-
pernatant was subjected to LC-MS/MS analysis without further treatment. The me-
tabolic rate was calculated according to the following equation: Metabolic rate
[pmol/min/mg protein] = {1000 [nM]–(conc. of residual compound [nM])}/reac-
tion time [min]/0.5 [mg protein].
18.. MBI assay: A mixture of human liver microsomes (0.5 mg/mL) and 10 µM test
compound was pre-incubated in phosphate buffer (pH 7.4) for 30 min at 37 °C with
and without NADPH (1 mM). The mixture was then added to the assay buffer con-
taining testosterone (a probe substrate of CYP3A4) and NADPH, and incubated for an
additional 10 min at 37 °C. The reaction was quenched with acetonitrile, and the
testosterone metabolite concentration was measured by liquid chromatography-
tandem mass spectrometry (LC-MS/MS) to evaluate the remaining CYP3A4 activity.
MBI was depicted as the ratio of metabolized testosterone after preincubation with
20.. Analytical data for 27: 1H NMR (400MHz, CDCl3) δ = 7.15–7.09 (m, 2H), 7.08–7.03
(m, 2H), 6.98–6.92 (m, 2H), 6.88–6.83 (m, 2H), 6.59 (s, 1H), 5.14–4.93 (m, 1H), 4.76
(s, 2H), 4.29–4.15 (m, 2H), 3.87 (s, 2H), 3.21–3.15 (m, 4H), 2.92–2.76 (m, 2H),
2.76–2.70 (m, 4H), 2.67 (s, 3H), 2.62 (s, 3H), 2.19 (s, 3H); HRMS (FAB) m/z cal-
culated for C29H36F2N3O3S+ 544.2440, observed 544.2438. See also Supplementary
Data for the synthetic procedure.
5