S. M. Lynch et al. / Bioorg. Med. Chem. Lett. 25 (2015) 48–52
51
Table 1
FLIPR potencies and kinetic solubility measurements for azacyclic analogs
Table 4
a
Rat pharmacokinetic data for selected compounds
Compound
FLIPRa
Na 1.7 IC50 (lM)
Solubilityb
pH 1.2
Solubilityb
pH 6.8
Compound
T
max (h)
C
max (ng/ml)
AUC (h ng/ml)
v
19
1.5
2.5
3.0
5.8
2.5
51 ± 11
94 ± 32
334 ± 22
562 ± 87
1898 ± 455
2728 ± 640
133 ± 57
9
9
9
1
1
1
a
b
c
2a
2b
3
>20
0
0
0.03
0.08
0
0
0
0.02
0.08
0
21
22
24
25
0.91 ± 0.02
0.36 ± 0.02
0.34 ± 0.03
0.40 ± 0.04
1.92 ± 0.30
186 ± 30
212 ± 43
21 ± 8
a
Dose = 3 mg/kg, po.
18
0.59
a
Mean ± SEM (standard error of the mean), n P3.
b
lM.
MDCK AB apparent permeability (Papp <1) and high efflux ratio
ER >30) and reduced CNS exposure. While compound 14 clearly
(
does not meet these criteria, we were pleased to see that the bis-
amide 15 did show a trend toward reduced permeability (Table 3).
Amides 9, 12, and 15 were found to suffer from low kinetic sol-
ubility at both acidic and neutral pH (Tables 1 and 3). At this junc-
ture we felt compelled to address this trend which we believed
could be attributed to the pyridyl amide motif. To this end, we
discovered that replacement of the C-ring aryl amide of 15 with
a linear diol provided an analog (21) which displayed superior
solubility while maintaining potency and limited MDCK perme-
ability. Thus we chose to concentrate on the diol-containing scaf-
fold for additional studies designed to probe the SAR at diazepine
N-4.
Table 2
a
FLIPR and electrophysiological potencies for azacyclic analogs
Compound
FLIPR Na
v
1.7 IC50
(l
M)
Electrophysiology
Na 1.7 K M)
hERG
(% inh at 1 lM)
v
i
(
l
1
1
1
2
2
2
2
2
2
2
4
5
9
0
1
2
3
4
5
6
0.61 ± 0.09
0.38 ± 0.06
1.05 ± 0.06
0.64 ± 0.03
0.66 ± 0.04
0.47 ± 0.02
0.58 ± 0.04
1.32 ± 0.13
0.97 ± 0.10
0.67 ± 0.05
0.40 ± 0.20
0.30 ± 0.03
0.30 ± 0.07
0.44 ± 0.12
0.63 ± 0.13
0.23 ± 0.03
nd
0.55 ± 0.09
0.45 ± 0.09
nd
nd
28.7 ± 5.6
8.7 ± 1.6
30 ± 5.9
13.9 ± 2.6
21.4 ± 5.8
nd
95.6 ± 0.9
22.6 ± 2.6
nd
All of the N-4 substituted polar analogs synthesized displayed
a
v
some degree of potency in both the Na 1.7 FLIPR and EP assays,
Mean ± SEM (standard error of the mean), n P3. nd = not determined.
with a good correlation observed between the two assays. This
data highlights the broad tolerance for substitution at this site.
Secondary profiling in the MDCK assay revealed several com-
pounds such as amide 19, sulfonamide 22 and urea 25 which were
of particular interest based on their low permeability, high efflux
potential and reasonable solubility. On the other hand, analogs
bearing a lesser degree of polarity at N-4, such as nitrile 23 and
cyclopropyl derivative 24 proved to be more permeable or have
lower efflux potential.
Table 3
Kinetic solubility and MDCK permeability measurements for selected compounds
Compound Solubilitya pH 1.2 Solubilitya pH 6.8 MDCK (Papp)b Effluxc
AB
3.5
0.7
BA
5.4
11.2
1
1
1
2
2
2
2
2
2
2
4
5
9
0
1
2
3
4
5
6
62
0.02
0.22
0.01
1
5
39
5
9
61
1
1.6
16
92
84
80
79
77
86
66
87
0.15 34.8 232
0.24 31.2 130
0.9
0.13 20.9 160
0.52
2.4
Since blockade of the hERG potassium channel can lead to long
QT syndrome and potentially fatal arrhythmias, we counter
28.3
31
22
screened several of our promising analogs against this target.
Although the cyclopropyl analog 25 was found to be a potent
inhibitor of hERG, most other analogs tested were found to only
weakly interact with this channel (Table 2). The risk for most of
the polar analogs was deemed acceptable at this stage of
exploration.
7.0
11.0
nd
13
5
nd
nd
0.2
0.1
34.3 172
a
b
c
l
M.
À6
1
0
cm/s.
Based on the favorable in vitro pharmacological activity, several
compounds possessing variable permeability profiles were
selected for additional characterization in rat pharmacokinetic
BA/AB. nd = not determined.
2
3
(
PK) assays. The data collected upon oral administration at a dose
To study the effect of heteroatoms within the azepine ring,
several close analogs to 9c were prepared and tested. Both of the
regioisomeric oxazepines were found to be tolerated in our
v
Na 1.7 primary screen (Table 1). Although the 4-unsubstituted dia-
of 3 mg/kg are highlighted in Table 4. The results implied that sul-
fonamide 22 could attain reasonable exposures, similar to cyclo-
propyl analog 24, even though it had a much lower intrinsic
permeability. Conversely, acetyl derivative 21, amide 19 and urea
25 showed reduced exposures. Designing a molecule which is gut
penetrant, but CNS restricted is a formidable challenge. Perhaps
the abundance of hydrogen bond donors in the amide and urea
analogs limits gut permeability whereas the dual polar carbonyl
moieties (hydrogen bond acceptors) of the sulfonamide impart a
more favorable balance between CNS peripheralization and oral
zepine 13 was less active, it was discovered that simple capping
with methyl or acetyl groups enabled a recovery in potency. In
order to confirm activity in a more physiologically relevant setting,
electrophysiology (EP) experiments were conducted on analogs 14
1
8
and 15. Both compounds displayed good functional potency in
this assay (Table 2).
2
4
In order to evaluate the peripheral nature of newly synthesized
compounds, several potent analogs were tested in a MDCK cell per-
exposure.
In conclusion, we have discovered a new series of N-aryl azacy-
clic sodium channel blockers which are predicted to be peripher-
ally restricted. Members of this class showed good potency for
1
9
meability assay. It is known that compounds possessing a high
affinity for efflux transporters, as measured by in vitro assays,
are likely to demonstrate limited brain penetration.2
0,21
In our
v
Na 1.7 in FLIPR-based and electrophysiological assays, selectivity
experience, there is a strong correlation between relatively low
over hERG and reasonable exposure in rat PK studies. Future work