Identification of 2-(2′-fluoro-[1,1′-biphenyl]-2-yl)acetamide as a Sodium Valproate-like broad spectrum anti-epileptic drug candidate

Article abstract of DOI:10.1016/j.bmcl.2018.12.014

By further optimizing compound A [2′-fluoro-N-methyl-[1,1′-biphenyl]-2-sulfonamide], we identified DSP-0565 [2-(2′-fluoro-[1,1′-biphenyl]-2-yl)acetamide, 17a] as a strong, broad-spectrum anti-epileptic drug (AED) candidate. Our efforts mainly focused on finding an alternative polar group for the sulfonamide in order to improve ADME profile of compound A including good metabolic stability and no reactive metabolic production. This led to the identification of biphenyl acetamide as a new scaffold for development of broad-spectrum AED candidates. DSP-0565 showed anti-convulsant activity in various models (scPTZ, MES, 6 Hz and amygdala kindling) with good safety margin, and was therefore selected as a clinical candidate.

Full text of DOI:10.1016/j.bmcl.2018.12.014

Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification of 2-(2′-fluoro-[1,1′-biphenyl]-2-yl)acetamide as a Sodium
Valproate-like broad spectrum anti-epileptic drug candidate
Tomoyuki Tanaka⁎, Nana Yajima, Tomoko Kiyoshi, Yoshiki Miura, Yoshifumi Inoue,
Takuya Nishimaki, Seiji Iwama
Drug Research Division, Sumitomo Dainippon Pharma Co., Ltd, 3-1-98 Kasugade-naka Konohana-ku, Osaka 554-0022, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
By further optimizing compound A [2′-fluoro-N-methyl-[1,1′-biphenyl]-2-sulfonamide], we identified DSP-0565
[2-(2′-fluoro-[1,1′-biphenyl]-2-yl)acetamide, 17a] as a strong, broad-spectrum anti-epileptic drug (AED) can-
didate. Our efforts mainly focused on finding an alternative polar group for the sulfonamide in order to improve
ADME profile of compound A including good metabolic stability and no reactive metabolic production. This led
to the identification of biphenyl acetamide as a new scaffold for development of broad-spectrum AED candidates.
DSP-0565 showed anti-convulsant activity in various models (scPTZ, MES, 6 Hz and amygdala kindling) with
good safety margin, and was therefore selected as a clinical candidate.
1,1′-Biphenyl-2-acetamide
Epilepsy
In vivo screening
MES
PTZ
Introduction
activity and minimal side effects are highly desired.7 Under above
consideration, we hypothesized that VPA-like AEDs which can sup-
Epilepsy is a complex neurological disorder that affects about 50
million people worldwide.1 Epileptic seizures can be classified as partial
or generalized, and are difficult to control in up to 30% of patients,
despite optimal use of available anti-epileptic drugs (AEDs).2,3 Al-
though a number of 2nd generation AEDs, such as Zonisamide, Lamo-
torigine, Topiramate, and Levetiracetam (Fig 1) have been developed,
patients refractory epileptic ratio has barely changed over the years.4 In
addition to their limited efficacy, most AEDs have safety concerns, in-
cluding CNS toxicity, drug-drug interaction, skin rash, teratogenicity,
and hepatotoxicity. So there are still much unmet medical needs in
epilepsy therapy and new class AEDs with enough safety profile were
highly required.
pressed both type of seizures may be required multiple mode of action
and in vivo screening using both MES and scPTZ models is the most
effective way to discover such broad spectrum AEDs and it might be
work for refractory epilepsy.
During the course of identifying a new scaffold of AEDs, we have
previously found that the 2′-fluoro-N-methyl-[1,1′-biphenyl]-2-sulfo-
namide A8 shows potent anti-convulsant effect in both MES and scPTZ
For the AED drug discovery to identify new structural feature, the
maximal electroshock seizure model (MES) and the subcutaneous
pentylenetetrazol model (scPTZ) are well-established and widely used.5
Almost all conventional AEDs, such as Carbamazepine and Phenytoin as
well as recently launched AEDs, such as Topiramate and Lamotrigine
have potent anticonvulsant effect in the MES model, but no or weak
efficacy in the scPTZ model.6 This has clinically been confirmed as most
AEDs show efficacy for partial seizure, but only limited effect for gen-
eralized seizure. Only Sodium Valproate (VPA) shows anti-convulsant
activity in both MES and scPTZ models and is currently prescribed for
both classes of seizure6 with side effects, including dizziness, wobble
and lassitude. Therefore, compounds with broad VPA-like anti-epileptic
Fig. 1. Chemical Structures of compounds A and existing AEDs.
⁎ Corresponding author.
E-mail address: tomoyuki-1-tanaka@ds-pharma.co.jp (T. Tanaka).
https://doi.org/10.1016/j.bmcl.2018.12.014
Received 12 November 2018; Accepted 6 December 2018
0960-894X/©2018PublishedbyElsevierLtd.
Pleasecitethisarticleas:Tanaka,T.,Bioorganic&MedicinalChemistryLetters,https://doi.org/10.1016/j.bmcl.2018.12.014

T. Tanaka et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Scheme 1. Synthetic routes to biphenyl compounds.
models. However the ADME evaluation revealed that compound A was
converted into unfavorable metabolite i.e., its reactive metabolites were
trapped as dansyl glutathione (dGSH) adducts in human liver micro-
somes and NADPH. This finding suggested that compound A has the
risk of idiosyncratic toxicity such as drug-induced liver injury (DILI),
Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN)
etc., so there was room to improve the ADME profile of compound A.
Therefore, our goal was to identify an alternative polar group for N-
methyl sulfonamide with acceptable ADME profile including no re-
active metabolite production. Herein we report the ortho-aryl-phenyl
acetamide derivatives as promising broad spectrum AED candidates
with improved ADME profile compared to the former ortho-aryl-phenyl
sulfonamide derivatives. The structure-activity relationships associated
with the conversion of the two phenyl rings A and B are also discussed
in this paper.
All compounds were synthesized according to Scheme 1. The ben-
zamide derivative 4 was synthesized from the commercially available
methyl 2-bromobenzoate 1 via introduction of 2-fluorophenyl group by
Suzuki coupling, followed by hydrolysis of the ester 2 and amidation of
the resulting carboxylic acid 3. The racemic hydantoin 7 was obtained
from the ester 2 via 3 steps including conversion of the ester group into
an aldehyde group followed by Urech hydantoin synthesis. The γ-
Lactam derivative 9 was synthesized by Stille coupling of the com-
mercially available 8. Conversion of ring B (13a-g) was achieved in a
2

T. Tanaka et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 1
Table 2
Anticonvulsant activity of biaryl compounds having various polar groups.
Anticonvulsant activity of acetamide compounds having various B rings.
Compound
Structure
Dose (mg/kg)
MESa)
scPTZa)
Compound
Structure
Dose (mg/kg)
100
MESa)
3/3
scPTZa)
1/3
A
100
50
3/3
3/3
3/3
3/3
3/3
2/3
13a
25
13b
13c
13d
13e
13f
100
100
100
0/3
0/3
0/3
1/3
4
100
50
3/3
3/3
3/3
1/3
0/3
0/3
0/3
25
7
100
100
100
3/3
3/3
0/3
3/3
3/3
1/3
2/3
2/3
0/3
9
50
25
100
3/3
3/3
0/3
2/3
2/3
2/3
16a
17a
50
25
100
3/3
2/3
0/3
3/3
0/3
-
100
50
3/3
3/3
2/3
2/3
2/3
1/3
50
25
25
13g
100
3/3
3/3
2/3
2/3
2/3
1/3
a) Efficacy was assessed based on the number of animals with no convulsion /
the total number of test animals.
50
25
manner similar to that for the synthesis of 4, i.e. 3 steps including Su-
zuki-Miyaura coupling reaction of 2-fluoro phenyl boronic acid and the
commercially available corresponding ethyl (hetero)aryl acetate 10,
hydrolysis of the resultant biaryl ester 11, and amidation of the re-
sulting biaryl carboxylic acid 12. Conversion of ring A (17a-g) was also
carried out in a similar way to that for conversion of ring B, starting
from Suzuki-Miyaura coupling reaction of the commercially available
ethyl 2-(2-bromophenyl)acetate 14 and the corresponding substituted
aryl boronic acid to afford 15, followed by hydrolysis and amidation to
give the carboxylic acid 16 and amide 17, respectively.
a) Efficacy was assessed based on the number of animals with no convulsion /
the total number of test animals.
The results are summarized in Table 2. Conversion of the phenyl ring
into any pyridine ring (13a, 13b and 13c) resulted in loss of anti-
convulsant activity in scPTZ models. On the other hand, introduction of
a fluorine atom at any position of the phenyl ring B resulted in accep-
table anti-convulsant activity in both MES and scPTZ models (com-
pound 13d, 13e, and 13g), except for the para position of A ring (i.e.
compound 13f). Among the obtained compounds, 13g showed a strong
anti-convulsant activity equivalent to that of compound 17a. However
dGSH adduct was obsearved in reactive metabolite production assay of
13g. Based on the results, introduction of a fluorine atom at any posi-
tion of the B ring brought no beneficial effect on anti-convulsant ac-
tivity. We therefore determined that a simple phenyl ring without any
substituent is the best for ring B.
First, we focused on finding an alternative polar group of the N-
methyl-sulfonamide in compound A. The results are summarized in
Table 1. Conversion of the sulfonamide into carboxamide (4) resulted in
maintenance of the potent anti-MES activity, but loss of anti-PTZ ac-
tivity. The Hydantoin 7 and lactam 9 also gave similar results. However
Compound 17a, a homologation analogue of compound 4, showed
strong anti-convulsant effects in both seizure models. Compound 17a
even showed anti-convulsant activity against scPTZ model at 25 mg/kg.
On the other hand, the carboxylic acid 16a showed no activity in both
MES and scPTZ models, indicating that the acetamide is crucial for anti-
convulsant effect. As for liver metabolic stability of 17a, more than 95%
of parent compound were recovered following incubation with rats or
human hepatic microsome and NADPH for 30 min at 37 °C. As for re-
active metabolite production, no dGSH adduct of 17a was observed.
These results suggested that the acetamide is appropriate as substitute
for N-methylsulfonamide in A and therefore we chose compound 17a
for further optimization.
Finally, we focused on optimization of the A ring moiety. The results
are summarized in Table 3. Based on our previous study, introduction
of a substituent at the ortho position is preferable for strong anti-con-
vulsant activity. Therefore, we assessed various electron withdrawing /
donating groups at the ortho position of A ring. With regard to electron
withdrawing groups, introduction of a chlorine atom (17b) maintained
the anti-convulsant activity in both MES and scPTZ models, however its
effect at 25 mg/kg p.o. in those models was weaker than that of 17a.
Introduction of other electron withdrawing groups, such as tri-
fluoromethyl (17c) or cyano (17d) obviously decreased the anti-
Next, we turned our attention to optimization of the B ring moiety.
3

T. Tanaka et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 3
Table 5
Anticonvulsant activity of acetamide compounds having various A rings.
Comparison of the effects of DSP-0565 and VPA in the MES, PTZ and 6 Hz
models, and in the Rotarod test.
Compound
Compound
Structure
Dose (mg/kg)
MESa)
scPTZa)
MESa) scPTZa) 6 Hza) Rotarodb) TI (MES)c) TI (6 Hz)d)
17b
100
50
3/3
3/3
1/3
3/3
3/3
0/3
17a (DSP-0565) 27
27
56
129
396
4.8
1.2
2.3
1.0
25
VPA 309
336
343
a) ED50 values (mg/kg, oral administration) were determined in our laboratory,
b) TD50 values (mg/kg, oral administration) were determined in our laboratory,
c) TI (MES) means Therapeutic index defined as TD50 / ED50 (MES), d) TI (6 Hz)
means Therapeutic index defined as TD50 / ED50 (6 Hz).
17c
17d
17e
17f
100
50
3/3
0/3
1/3
0/3
convulsant activity in scPTZ model even at 100 mg/kg p.o. As for
electron donating groups, introduction of a methoxy group (17e) or a
methyl group (17f) also decreased the anti-convulsant activity in both
model at 100 mg/kg. These findings suggested that both A and B rings
should be slightly twisted and that a fluorine atom provides a suitable
angle for strong anti-convulsant activity. Next, we evaluated the ac-
tivity of bis-fluorine or fluoropyridyl substituted compounds (17g-j).
Both the 2,6-difluoro derivative (17g) and 2,5-difluoro derivative
(17 h) at 50 mg/kg p.o. showed anti-convulsant activity comparable to
that of 17a. However, at 25 mg/kg p.o. both compounds exhibited no
activity in the scPTZ model, while compound 17a showed a slight ef-
ficacy in this model. On the other hand, the 2,4-difluoro derivative 17i
did not show enough anti-convulsant activity in the scPTZ model even
at the dose of 100 mg/kg p.o. Conversion of the 2-fluorophenyl ring
into a fluoropyridyl ring (17j) decreased the anti-convulsant effect in
the MES model even at 100 mg/kg p.o. From the above pharmacolo-
gical data for biphenyl acetamide-type compounds, we selected com-
pound 17a (DSP-0565)9 for further evaluation using other convulsion
models to confirm its potential as new AED candidate.
100
1/3
0/3
100
50
2/3
0/3
1/3
-
100
1/3
1/3
17g
100
50
3/3
3/3
3/3
3/3
2/3
0/3
25
Initially, we tested the rat amygdala kinding model for predicting
10
drug’s efficacy against partial epilepsy.
As summarized in Table 4,
17h
17i
100
50
3/3
3/3
0/3
3/3
2/3
0/3
DSP-0565 dramatically improved alteration of after-discharge
threshold (ADT) and after-discharge duration (ADD). This result sug-
gests that DSP-0565 could be show anti-epileptic activity in patients
with partial seizure.
25
100
50
3/3
1/3
1/3
Next, we evaluated the beneficial effect of DSP-0565 in the 6 Hz
model, which has recently been recognized as a drug-resistant epilepsy
model.11 The results are summarized in Table 5. Both DSP-0565 and
VPA showed anti-convulsant activity in this model. We also determined
the ED50 values of DSP-0565 in MES, scPTZ and 6 Hz models found
their values were in a similar range (27, 27 and 56 mg/kg, respec-
tively). A comparison of both compounds’ neurotoxicity using the ro-
tarod test revealed that the therapeutic Index (TI), i.e. TD50/ED50 value
of DSP-0565 is 2–4 times superior to that of VPA. Thus, the effective
dosages of DSP-0565 in those model were lower than that of rotarod
test. On the other hand, VPA is same range as rotarod TD50. These re-
sults suggest that DSP-0565 could be more useful drug than VPA.
Mode of action (MoA) of DSP-0565 was also investigated. Firstly we
25
17j
100
0/3
2/3
a) Efficacy was assessed based on the number of animals with no convulsion /
the total number of test animals.
Table 4
Anticonvulsant effect of 17a (DSP-0565) and VPA in the rat amygdala kindling model.
Compound
dose : mg/kg, p.o.
ADTa)
%change
ADDb) (sec)
Before
%change
SSc)
(μA)
(score)
(No. of rats)
Before
After
After
Before
After
control
- (6)
68
68
0
109
101.7
72.3
−5.5
–22.9
–22.9
11.1
5
5
17a (DSP-0565)
50 (3)
100 (3)
300 (3)
600 (3)
63.3
45.3
39.3
42.3
114.0
80.0
36.7
80.0
75.9
80.1
−5.4
81.3
89.2
89.7
104.7
99.7
5.0
5
4.3
4.3
4.7
3.3
70.8
VPA
115.7
79.0
5.0
5.0
–22.5
a) Alteration of after-discharge threshold (ADT): Point when abnormal EEG was observed.
b) after-discharge duration (ADD), c) seizure stage (SS): classified as follows; (1) mouth and facial movements; (2) head nodding; (3) unilateral forelimb clonus; (4)
rearing and bilateral forelimb clonus; and (5) rearing and falling.
4

T. Tanaka et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
conducted GABA-A receptors binding assays using radio ligands such as
flunitrazepam, Muscimol, Ro-15-1788, TBOB and TBPS according to
reported methods 12–20, because scPTZ is known as a block of GABA-A
receptor. The unbound brain concentration of DSP-0565 was calculated
as 24.6 μmol/L at 30 mg/kg oral administration, which is approxi-
mately the ED50 values of this compound in MES or scPTZ model.
Therefore, we selected to concentration at 30 μM for those assays.
Contrary to our expectation, DSP-0565 did not bind to any GABA-A
binding sites where radio ligands described above. This finding sug-
gested that in vivo activity of DSP-0565 in scPTZ may be caused as the
result of its GABAergic function via novel target.
to Keiko Bando for performing elemental analysis. We also would like
to thank Yoshiaki Isobe for his support and encouragement.
References
1. FDA fact sheet: http://www.who.int/mediacentre/factsheets/fs999/en/.
2. Kwan P, et al. Exper Rev Neurother. 2006;6:397.
3. Perucca E, et al. Lancet Neurol. 2007;6:793.
4. Zhibin C, et al. JAMA Neurol. 2018;75:279.
5. Schmidt D, et al. Epilepsy Res. 2002;50:71.
6. Loscher W. Seizures. 2011;20:359.
7. NIH once recommended these “in vivo screening” approach, see ; Stables JP, et al.
“The NIH Anticonvulsant Drug Development(ADD) Program: preclinical antic-
onvulsant screening project” http://www.ninds.nih.gov/research/asp/addadd_re-
view.pdf.
In conclusion, by using the golden standard AED rodent models MES
and scPTZ, we identified 2-(2′-fluoro-[1,1′-biphenyl]−2-yl)acetamide
17a (DSP-0565) as a new drug candidate with broad anti-convulsant
activity. Optimization of the N-methylsulfonamide moiety in the lead
compound A revealed that an acetamide structure is a suitable alter-
native polar group. Optimization of the two phenyl ring moieties re-
vealed that introduction of other substituents or replacement of the
phenyl ring with other aryl group achieved no improvement in the anti-
convulsant activity. Our results also show that DSP-0565 clearly pos-
sesses anti-convulsant activity in the amygdala kindling and 6 Hz
models with acceptable neurotoxicity in the rotarod test. Based on the
results of MoA discovering studies, we found that DSP-0565 possesses
unique GABAergic function which is different from ones VPA or
Benzodiazepines have. It is therefore concluded that DSP-0565 can be
useful in the treatment of refractory epilepsy, which are currently dif-
ficult to control by available AEDs.
8. Tanaka T, et al. Bioorg Med Chem Lett. 2017;27:4118.
9. Spectral data for compound 17a: 1H NMR (CDCl3, 400 MHz)δ 7.33-7.43 (4H, m), 7.
19-7.28 (3H, m), 7.11-7.17 (1H, m), 5.25 (2H, br), 3.48 (2H, s); 13C NMR (CDCl3,
400 MHz) δ173.1, 159.4 (d, J = 971.6 Hz), 136.0, 133.5, 131.6 (d, J = 11.6 Hz),
131.0, 130.3, 129.7 (d, J = 30.8 Hz), 128.7, 127.9 (d, J = 65.2 Hz), 127.5, 124.3 (d,
J = 11.2 Hz), 115.8 (d, J = 88.0 Hz), 40.8; HRMS (ESI+) m/z calcd for C14H13FNO
(M+H)+: 230.0976, found: 230.0973; Anal. Calcd for C14H12FNO; C, 73.35; H, 5.
28; N, 6.11; F, 8.29. Found: C, 73.44; H, 5.28; N, 6.18; F, 8.29.
10. Sato M, et al. Electroenceph Clin Neurophysiol. 1990;76:459.
11. Barton M, et al. Epilepsy Res. 2001;47:217.
12. Damm HW, et al. Res Comm Chem Pathol Pharmacol. 1978;22:597.
13. Speth RC, et al. Life Sci. 1979;24:351.
14. Enna SJ, et al. Mol Pharmacol. 1977;13:442.
15. Martinin C, et al. J Neurochem. 1983;41:1183.
16. McLeod M, et al. Brain Res Mol Brain Res. 2002;104(2):203.
17. Levin AH, et al. Mol Pharmacol. 1989;35:189.
18. Schwartz RD, et al. J Pharmacol Exp Ther. 1988;244:963.
19. Maksay G. J Eur Pharmacol. 1993;246(3):255.
20. Savini L, et al. Bioorg Med Chem. 1998;6(4):389.
Acknowledgements
We are grateful to Sayo Takayama for recording HRMS spectra and
5