Conformations of 3-carboxylic esters essential for neurotoxicity in veratrum alkaloids are loosely restricted and fluctuate

Article abstract of DOI:10.1016/j.bmc.2007.12.037

The lipid-soluble veratrum alkaloids, veratridine and cevadine, are plant neurotoxins that are agonists of voltage-gated sodium channel. Their conformations in a hydrophobic environment were analyzed by NMR spectroscopy in solution phase chloroform at low temperatures. The conformations around the 3-carboxylic esters which is essential for their neurotoxicity, was completely different from the previously reported X-ray crystallographic structure. The carbonyl oxygen atom (O28) of the carboxylic ester forms a weak intramolecular hydrogen bond with the OH proton at C4 (4-OH) that loosely restricts the conformation of the 3-veratroyl ester in veratridine and the 3-angeloyl ester in cevadine. Methylation at C4 hydroxyl group of veratridine had much reduced its neurotoxic activity relating to voltage-gated sodium channel. The results suggested that the loose conformational restrictions of the carboxylic esters are important for neurotoxicity of the veratrum alkaloids.

Full text of DOI:10.1016/j.bmc.2007.12.037

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 3025–3031
Conformations of 3-carboxylic esters essential for neurotoxicity
in veratrum alkaloids are loosely restricted and fluctuate
Ai Niitsu, Masanori Harada, Tohru Yamagaki* and Kazuo Tachibana
Department of Chemistry, School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113-0033, Japan
Received 2 November 2007; revised 15 December 2007; accepted 19 December 2007
Available online 11 January 2008
Abstract—The lipid-soluble veratrum alkaloids, veratridine and cevadine, are plant neurotoxins that are agonists of voltage-gated
sodium channel. Their conformations in a hydrophobic environment were analyzed by NMR spectroscopy in solution phase chlo-
roform at low temperatures. The conformations around the 3-carboxylic esters which is essential for their neurotoxicity, was com-
pletely different from the previously reported X-ray crystallographic structure. The carbonyl oxygen atom (O28) of the carboxylic
ester forms a weak intramolecular hydrogen bond with the OH proton at C4 (4-OH) that loosely restricts the conformation of the 3-
veratroyl ester in veratridine and the 3-angeloyl ester in cevadine. Methylation at C4 hydroxyl group of veratridine had much
reduced its neurotoxic activity relating to voltage-gated sodium channel. The results suggested that the loose conformational restric-
tions of the carboxylic esters are important for neurotoxicity of the veratrum alkaloids.
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
excellent strategy for the lipid-soluble neurotoxins to
incorporate their own structural flexibility, even though
they retain the preferable, but not fixed, conformations
required for specific binding to the receptor proteins.
These neurotoxins include such contradictory structural
properties and thus elegantly solve this biological
problem.
Interactions between small molecules and target pro-
teins have been studied to elucidate their biological func-
tions, and in pharmaceutical chemistry these types of
studies are applied to drug design and drug screening.
Recently, attention has been focused on the relation-
ships between structural fluctuations and protein func-
tions. Fluctuations can also be important for protein–
ligand binding. Biologically active organic molecules
have partial flexibility in their structures. For example,
the marine natural products ciguatoxins,1 brevetoxins,2
and maitotoxins,3 include seven- to nine-membered
rings, and the middle part of these molecules can fluctu-
ate due to those rings. Such structural flexibility is
thought to be essential for their toxic activity and for
optimal binding to their target proteins.4,5 Particularly
in the case of lipid-soluble natural products binding spe-
cifically to target membrane proteins, these types of fluc-
tuations are essential. Molecular motions generally
become much slower in biomembranes than in solution,
hence specific interactions between ligands and their
receptors are statistically less frequent. Thus, it is an
Veratridine 1 is one of the neurotoxic veratrum alkaloids
isolated from sabadilla seed, whose primary targets are
voltage-gated sodium channel.6–8 It has a steroidal back-
bone 4a,9-epoxycevane substituted by 3,4-dim-
ethoxybenzoate at C-3; in other words, it has a rigid
backbone and a highly flexible and freely rotatable ester
substitution (Fig. 1). Veratridine is known to bind to
and to activate voltage-gated sodium channel, which
makes the ion channel open and thereby depolarizes bio-
membranes.6–8 A diverse set of lipid-soluble neurotoxins
relating to sodium channel has been well studied, includ-
ing tetrodotoxin, saxitoxin, brevetoxin, batrachotoxin,
and veratridine.9,10 Competitive binding studies with
radiolabeled neurotoxin analogues have identified at least
eight distinct neurotoxin binding sites, and those specific
for veratrum alkaloids, as well as batrachotoxin, aconitin,
and grayanotoxin,9–14 were categorized as ‘site-2’.11–14
Keywords: Alkaloid; Conformation; NMR; Hydrogen bonding; Struc-
ture fluctuation; Veratridine; Sodium channel.
Veratridine 1 and cevadine 2 have similar neurotoxic
activities, but veracevine 3 is not toxic.15 Previous re-
search has shown that the functional groups at C-3,
*
Corresponding author. Fax: +81
yamagaki@chem.s.u-tokyo.ac.jp
3
5841 8059; e-mail:
0968-0896/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.12.037

3026
A. Niitsu et al. / Bioorg. Med. Chem. 16 (2008) 3025–3031
These C-3 acyl groups seem to be highly flexible in the
veratrum skeleton because the carboxylic esters have
three free rotational axes. Considering the selectivity
and toxic activity of veratrum alkaloids for targeted so-
dium channel, the conformations at the C-3 functional
groups on the steroidal skeleton are expected to be re-
stricted. However, the monomer structure indicated no
conformational restriction of the 3-veratroyl group in
the crystal structure previously reported.17
In this study, we investigated the conformational restric-
tion and/or flexibility of the 3-carboxylic esters of vera-
trum alkaloids in chloroform as a representative
hydrophobic environment, varying the temperature
and using 1H NMR to evaluate changes in the molecular
dynamics of the system.
2. Results and discussion
2.1. Hydroxyl group proton signals and refined
assignments
Figure 1. Structures of veratridine 1, cevadine 2, and veracevine 3.
1
All signals in the H and 13C NMR spectra of 1–3 were
re-assigned through DQF-COSY, HSQC, HMBC 2D-
the aromatic and aliphatic carboxylic esters, are essen-
tial for the neurotoxicity of veratrum alkaloids.15,16
Table 1. Revised Assignments of 1H and 13C NMR chemical shifts of veratridine 1, cevadine 2 and veracevine 3
Veratridine
Cevadine
Veracevine
Position
dH
dC
32.6
Position
dH
dC
32.6
Position
dH
dC
31.9
1
2a
3
1.62, 1.62
1.73, 2.16
5.10
1
2a
1.55, 1.55
1.53, 1.65
4.98
1
1.49, 1.61
1.47, 1.58
3.76
26.7
75.7
105.0
46.1
18.9
16.9
44.7
94.5
45.5
42.0
81.7
36.9
80.3
31.1
70.9
72.0
51.3
18.9
75.7
15.4
63.5
18.3
29.0
27.4
61.2
17.0
167.1
122.2
112.4
148.8
153.5
110.3
123.8
56.0
56.0
26.7
75.1
105.1
46.0
18.8
16.9
44.7
94.4
45.5
42.0
81.7
36.9
80.3
31.0
70.9
72.0
51.3
19.0
75.7
15.4
63.4
18.3
29.0
27.4
61.2
17.0
168.9
127.3
139.7
16.0
20.6
2
27.7
73.5
106.4
44.1
18.9
17.0
44.7
94.4
45.7
42.0
81.7
36.9
80.3
31.0
70.9
72.0
51.3
19.1
75.7
15.4
63.4
18.3
29.0
27.4
61.2
17.1
3
3
4
—
2.07
4
—
2.00
4
—
2.20
5
5
5
6a
7
1.73, 1.95
1.73, 2.00
2.66
6a
7
1.73, 1.92
1.73, 2.00
2.66
6
—
7
1.78, 1.89
2.66
—
8
8
8
9
—
—
9
—
—
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23a
24a
25
26
27
28
29
30
31
32
33
34
35
36
10
11
12
13
14
15
16
17
18
19
20
21
22
23a
24a
25
26
27
28
29
30
31
32
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
—
1.81, 2.18
—
2.10
1.80, 2.18
—
2.10
1.79, 2.18
—
2.10
—
—
—
1.75, 1.89
4.10
—
1.76, 1.90
4.10
—
1.75, 1.91
4.12
—
2.61
0.99
2.61
0.98
2.60
0.96
—
1.11
—
1.11
—
1.13
2.45
2.46
2.45
1.40, 1.55
1.55, 1.55
1.88
1.40, 1.55
1.55, 1.55
1.88
1.40, 1.55
1.55, 1.55
1.88
2.39, 2.66
1.06
—
2.39, 2.66
1.06
—
2.40, 2.65
1.07
—
7.48
—
6.11
—
—
1.97
1.88
6.84
7.61
3.88
3.90
a The assignments labeled by asterisks were re-assigned to different signals from those in a previous report.18

A. Niitsu et al. / Bioorg. Med. Chem. 16 (2008) 3025–3031
3027
NMR techniques, and a few assignments were revised as
shown in Table 1. The mixture of veratrum alkaloids
was measured directly in the previous report,18 but puri-
fied samples were used in this study. The samples and
solutions were prepared carefully as described in Section
4.
The OH proton signals showed temperature depen-
dency; their chemical shifts migrated to the upfield, as
seen in measurements at higher temperatures (Figs. 2
and 3). No migration of chemical shifts and no signal
broadening were observed for the C–H proton signals
under the same conditions. None of the proton signals
of 1–3 depended on the concentrations (data not
shown), indicating the sample molecules are monomeric.
Thus, all the hydrogen bonds in these molecules 1–3 are
intramolecular, rather than intermolecular.
Hydroxyl group (OH) proton signals were observed in
the H NMR spectra of 1–3 in deuterated chloroform,
1
in addition to the C–H proton signals (Fig. 2). Gener-
ally, OH and NH proton signals in deuterated chloro-
form can exchange easily and rapidly to contaminating
deuterium in the solution, and therefore their signals dis-
appear. When the OH and NH groups form hydrogen
bonds, their signals can be detected. Thus, the observed
OH signals indicate the formation of hydrogen bonds.
2.2. b-Hydroxyl group (4-OH) neighboring the C-3
carboxylic ester
The four OH proton signals assigned to 12-OH, 14-OH,
16-OH, and 17-OH were observed commonly among 1–
Figure 2. OH signal region spectra of veratridine 1 (a), cevadine 2 (b), and veracevine 3 (c). The spectrum of 3 was acquired at 35 °C.
Figure 3. OH Signal migrations in the 1H NMR spectra of veratridine (a), cevadine (b), and veracevine (c) at low temperatures. The chemical shifts
are plotted for different temperatures at 35, 15, 0, À15, and À30 °C. The red circles are the C-4 hydroxy group proton signal (4-OH), the blue squares
are the C-12 hydroxy group proton signal (12-OH), the light green diamonds are the C-14 OH proton signal (14-OH), the black triangles are the C-16
OH proton signal (16-OH), and the magenta reversed-triangles are the C-17 OH proton signal (17-OH).

3028
A. Niitsu et al. / Bioorg. Med. Chem. 16 (2008) 3025–3031
3 (Figs. 2 and 3). The 4-OH proton signals of 1 and 2
were detected, in contrast, the 4-OH proton signal of 3
was not. The shape of the 4-OH proton signal changes
most significantly. The 4-OH signal of 1 begun to broad-
en at 55 °C, that of 2 broadened at 35 °C as the measure-
ment temperatures were elevated (Fig. 2). The chemical
shift of 4-OH proton migrates further upfield than the
other four OH proton signals at higher temperatures
(Fig. 3). The OH proton signal migrations on the 12-
OH, 14-OH, 16-OH, and 17-OH are gentle and of sim-
ilar magnitude. These results suggest that the hydrogen
bond of 4-OH is weaker and the proton exchange rate
is faster than the others, and that the hydrogen bonding
at 4-OH becomes strong and the exchange rate gets
slower as the temperature decreases. Since the other four
OH group signal shapes do not change, their hydrogen
bonds must be quite stable. As described later, the
hydrogen bond network in steroidal backbone included
these hydroxyl groups at C12, C14, C17, and C16.
Figure 4. Structure of veratridine in solution phase calculated at the
HF/3-21G(d) level of theory.
2.3. Hydrogen bond networks
group at C-3 (O3) is not an acceptor of the hydrogen
atom in 4-OH. In addition, these vicinal hydroxyl
groups (O3 and 4-OH) are not in the same plane,
although they are both b-orientations, so it is difficult
to form a hydrogen bond between them. Consequently,
in veratridine and cevadine, the carbonyl oxygen atom
O28 forms an intramolecular hydrogen bond with the
4-OH hydrogen atom. The structure of veratridine 1
was calculated with the Gaussian HF/3-21G(d) level of
theory based on all intramolecular hydrogen bonds indi-
cated by the NMR experiments as shown in Figure 4.
The hydrogen bonding in the crystal structure of 1 was
reported.17 The hydroxyl groups at C-12, C-14, C-17,
and C-20 form intramolecular hydrogen bonds both in
liquid phase and crystal structure. The steroidal back-
bone conformation was stabilized by extensive intramo-
lecular hydrogen bonds. The suggested hydrogen bond
networks from the NMR data and that of the crystal
structure are shown in Table 2. The hydrogen bond net-
works of 12-OH, 14-OH, 17-OH, and 20-OH were sim-
ilar among the three veratrum alkaloids 1–3. The
difference between the liquid phase structure and the
crystal structure is the hydrogen bond including 4-OH.
The 4-OH in the liquid phase forms an intramolecular
hydrogen bond, in contrast, that in the crystal structure
forms an intermolecular one.
2.4. Hydrogen bond around 3-carboxylic esters and their
toxic activity
Considering the necessity of 3-carboxylic ester for verat-
ridine neurotoxicity, it is significant that the acceptor of
the hydrogen bond of the 4-OH is the carbonyl oxygen
atom O28 in the 3-veratroyl ester, rather than the ester
oxygen atom (O3). If the hydrogen bonding acceptor
was the O3 ester oxygen, there would be no factors
restricting the rotation at the 3-veratroyl ester. The
hydrogen bond formed between the 4-OH and the car-
bonyl oxygen atom O28 enables to restrict the conforma-
tion of the 3-veratroyl group to the steroidal backbone as
shown in Figure 4. Moreover, this hydrogen bond is not
strong, and the 3-veratroyl group conformation is
loosely restricted rather than rigid, as described above.
These structural features and properties are suggested
to be important for neurotoxicity, that is, the structurally
fixed steroidal backbone and the fluctuating 3-veratroyl
group having a specific but loosely restricted conforma-
tion. This is an excellent strategy that allows the lipid-sol-
uble neurotoxins to effectively target and specifically
bind their target membrane proteins. Such strategies
for binding membrane proteins are thought to be one
of the common matters among the neurotoxins.
The 4-OH signal was observed in the spectra of 1 and 2,
but that was not in the spectrum of 3 (Figs. 2 and 3) as
described above. The steroidal backbone is common
among these three compounds (Fig. 1), but only 3 does
not have the carboxylic ester. Therefore, the acceptor of
the 4-OH proton is suggested to be the carbonyl oxygen.
The neighboring ester oxygen atom O3 could possibly
accept a hydrogen bond with 4-OH by forming five-
membered ring. If the 4-OH hydroxyl group formed a
hydrogen bond to the neighboring O3 ester oxygen, they
could bind each other in compound 3 without the 3-acyl
substitution at the hydroxyl group. However, the 4-OH
signal disappeared in the spectrum of 3 (Fig. 2), indicat-
ing that the oxygen atom of the neighboring hydroxyl
Table 2. Hydrogen bonding network
NMR
Crystal structureb
aO16–H16Á Á ÁO20
O17–H17Á Á ÁO12
O12–H12Á Á ÁO14
O14–H14Á Á ÁO49
O20–H20Á Á ÁO16
O17–H17Á Á ÁO12
O12–H12Á Á ÁO14
O14–H14Á Á ÁO49
2.5. 4-Methoxy veratridine and its neurotoxicity
a The difference between liquid phase and crystal structure.
b Ref. 17.
A previous structure-activity relationship study showed
4,16-diacetyl veratridine and 3a-epiveratridine had no

A. Niitsu et al. / Bioorg. Med. Chem. 16 (2008) 3025–3031
3029
Scheme 1. Preparation of 4-methoxy veratridine. 4-OH of veratridine was methylated using cation exchange resin in methanol.
toxicity against the housefly.16 The acetylation at the C-
4 hydroxyl group in the former compound and the a-ori-
entation of the 3-carboxylic ester in the latter compound
prevent the formation of an intramolecular hydrogen
bond between 4-OH and the carbonyl oxygen atom (4-
OH::::O28). These results suggest that the 4-OH is
important for veratridine neurotoxic activity and that,
moreover, the conformational restriction of the 3-car-
boxylic esters in the veratrum alkaloids with an intramo-
lecular hydrogen bond to 4-OH is important for their
neurotoxicity. Here, we investigated the toxic activity
of 4-methoxy veratridine to prove the importance of
intramolecular hydrogen bonding for veratridine neuro-
toxicity, because that hydrogen bond cannot form in 4-
methoxy veratridine. 4-methoxy veratridine was pre-
Figure 5. Selective activity for voltage-gated sodium channel in the
neuroblastoma N18-RE-105 hybrid cell. The value of absorbance at
570 nm caused by formazan, reduced compound from 3-(4,5-Dimeth-
ylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, is correlated with cell
viability. The cytotoxicity indicated the selective activation activity for
sodium channel in this assay system. The gray bar labeled with (—) on
the left side is the negative control in which only 1% (v/v, final
concentration) ethanol solution was added to the cell. The dotted line
drawn over 0.15 (abs 570 nm) is the positive control in which the
1.7 mM adriamysin (final concentration) was added. Curve approxi-
mation was performed using a Graph Pad PrismÒ sigmoidal dose–
response program.
pared from veratridine using a cation exchange resin
and methanol (Scheme 1).19
To investigate toxic activity of 4-methoxy veratridine
relating to voltage-gated sodium channel, a newly de-
vised toxicity assay system was used.20 A neuroblastoma
cell expressing voltage-gated sodium channel (N18-RE-
105 cell) was used in this assay system.21 In general,
the intracellular accumulation of sodium ion leads to
cell apoptosis.22 In the neuroblastoma cells, a channel
opener such as veratridine can induce apoptosis in the
presence of an inhibitor of Na+/K+ ATPase such as oua-
bain, because the intracellular Na+ concentration is usu-
ally stabilized by the plasma membrane Na+/K+ ATPase
(see the overview of the assay system in Supporting
information 1).
mations of 3-veratroyl ester in veratridine and 3-angeloyl
ester in cevadine. It should be difficult that these hydro-
gen bonding restrictions and conformations were pre-
dicted computationally based on their crystal
structures. The neurotoxicity relating to sodium channel
was effectively reduced in 4-methoxy veratridine that
does not have the hydrogen bond around 3-veratroyl es-
ter. Therefore, the intramolecular hydrogen bond
around 3-carboxylic ester is suggested to be essential
for the neurotoxicity of veratrum alkaloids.
Cell viability against 4-methoxy veratridine was much
higher than that against veratridine at each concen-
tration (Fig. 5), suggesting that the activity of 4-
methoxy veratridine was effectively reduced by the
methylation. Therefore, 4-OH is essential for veratri-
dine neurotoxicity relating to voltage-gated sodium
channel.
To investigate the structural common features of the
agonists of sodium channel in their pharmacological
activity, the structure superposition study was per-
formed in the crystal structure of batrachotoxine and
veratridine,17 however, the superposed veratridine struc-
ture was limited in the steroidal backbone because, as
one of the possibilities, the restriction of 3-carbonyl ester
was not observed in the crystal structure of veratridine
monomer. In these studies, the conformation study
including the weak and labile restrictions or fluctuations
becomes more important to understand the mechanisms
how those pharmacologically important agonists and
antagonists active to their receptors.
3. Conclusions
The neurotoxicity of veratrum alkaloids is well known to
depend on the chemical structures of 3-carboxylic esters
from the structure-activity relationship research. Never-
theless, their conformations have not been discussed be-
cause there was no useful structure information. In this
study, hydrogen bond analyses of veratrum alkaloids in
1
liquid phase using H NMR spectroscopy indicate the
intramolecular hydrogen bond around the 3-carboxylic
ester, which cannot be observed in the X-ray crystal
structure. The intramolecular hydrogen bond between
4-OH and 3-carbonyl oxygen atom restricts the confor-
Interestingly, the intramolecular hydrogen bond of 3-
carboxylic ester is weaker than those of steroidal back-

3030
A. Niitsu et al. / Bioorg. Med. Chem. 16 (2008) 3025–3031
Table 3. Assignments of 1H and 13C NMR chemical shifts of 4-
methoxy veratridine
by using reversed-phase HPLC (JASCO HPLC System,
JASCO Corp., Hachioji, Japan). The effluents were de-
1
tected by UV absorbance at 245 nm. The H and 13C
Position
dH
dC
32.6
NMR signal assignments are summarized in Table 1.
The detected mass is [M+H]+ 592.4 in MALDI-TOF-
MS (AXIMA-CFRs MALDI TOF mass spectrometer,
Shimadzu, Co., Kyoto, Japan).
1
2
1.62, 1.62
1.73, 2.16
5.58
26.7
67.3
107.1
46.8
19.2
16.9
44.7
95.6
45.5
42.0
81.7
36.9
80.3
30.9
70.9
72.0
51.3
19.2
75.7
15.4
63.4
18.3
29.0
27.4
61.2
17.0
165.3
122.6
112.2
148.8
152.5
110.3
123.6
56.0
56.0
50.5
3
4
—
2.29
5
6
1.73, 1.95
1.73, 2.00
2.66
Veracevine was prepared from veratridine by alkaline
hydrolysis.23 Veratridine (3.8 mg) in 100 lL of methanol
was treated with 4 lL of 5 N aqueous NaOH and stirred
at 0 °C for 24 h. The solution was diluted with 200 lL of
ice water and extracted 10 times with 20 lL portions of
chloroform. The extract was washed with 40 lL of
water, evaporated, and applied to normal phase silica
gel column chromatography, and then it was applied
to reversed-phase HPLC. The effluent was monitored
on an ISI RI-981 refractive index (LabSystem Instru-
ments Ltd., Ome, Japan). The purified veracevine yield
7
8
9
—
—
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
1.81, 2.18
—
2.10
—
1.75,1.89
4.14
—
1
was 2.1 mg (80%). The H and 13C NMR signal assign-
2.61
0.99
ments are summarized in Table 1. The detected mass
was [M+H]+ 510.2 in MALDI-TOFMS.
—
1.11
2.46
4.2. NMR and signal assignments
1.40, 1.55
1.55, 1.55
1.88
All NMR data were acquired on a JEOL ECA500
instrument (1H, 500 MHz; JEOL Ltd., Tokyo, Japan)
with Nanolac Probe and JEOL ECX400 instrument
(13C, 100 MHz; JEOL Ltd., Tokyo, Japan) with TH5
Probe in deuterium chloroform (Sigma, Japan). The
chemical shifts were calibrated from the solvent signal
CHCl3 (dH 7.24, dC 77.0).
2.39, 2.66
1.06
—
—
7.50
—
—
6.87
1.88
1
The H NMR signals of 1–3 were assigned, and in par-
1
ticular the OH groups were assigned with HMBC. H
3.91
3.92
NMR signals depend on pH values for alkaloids, and
we carefully prepared samples to maintain their repro-
ducibility throughout our protocol, to avoid causing
spectral changes by protonation of the amine in acid
purification conditions; the sample dissolved in chloro-
form was washed in a 1-N sodium hydrate aqueous solu-
tion, evaporated, and freeze dried.
3.37
bone, that is, the conformations of 3-carboxylic ester are
loosely restricted, not fixed in the molecules. ‘his flexibil-
ity is thought to be essential to recognize the receptors in
the ligand–receptor interactions. Recently, the structure
fluctuations of proteins as the receptors were studied
well, but those of the organic small molecules as the li-
gands were not so much. These concepts in the molecu-
lar fluctuations are excellent strategy to recognize the
receptor and/or ligand to each other specifically and rap-
idly. These ideas supported by our study will be very
useful and helpful for drug design, and understanding
the small molecule interactions to their receptors in
biomembrane.
Two milligrams of purified cevadine and veracevine, and
10 mg of veratridine, were dissolved in 550 lL of
CDCl3, respectively. 1H NMR spectra were recorded
from À30 to 55 °C.
4.3. 4-Methoxy veratridine19
Veratridine (4.7 mg) in 2 mL of MeOH was stirred with
53.2 mg of cation-exchange resin Dowex 50Wx4 (Sig-
ma–Aldrich, Japan) and heated to reflux for 11 h and
subsequently at 55 °C for 16 h. The solution was filtered,
washed with 1 N HCl/MeOH, 1:1, evaporated, and puri-
fied using reversed-phase HPLC (effluent monitored at
UV 260 nm) to yield 654 lg (14%) of 4-methoxy veratri-
dine. The 1H and 13C NMR signal assignments are sum-
marized in Table 3. The 4-methoxy group was
determined by an HMBC signal between C4 and H37.
A high-resolution Mass (JEOL JMS-SX102 FAB mass
spectrometer, JEOL Ltd., Japan) was found at
[M+H]+ 688.3667 for C37H54NO11 (À3.0 mmu).
4. Experimental
4.1. Samples
Veratridine (>90% purity at TLC) was purchased from
Sigma–Aldrich, Japan, and was used without any fur-
ther purification.
Cevadine was isolated from veratrine hydrochloride, a
mixture of veratrum alkaloids (Sigma–Aldrich, Japan)

A. Niitsu et al. / Bioorg. Med. Chem. 16 (2008) 3025–3031
3031
4.4. Cell culture
Supplementary data
Neuronal mouse-rat hybrid cells N18-RE-105
(IFO50221, subdivided from JCRB Cell Bank, Japan)
were cultured at 37 °C with 5% CO2 in Dulbecco’s mod-
ified Eagle’s medium (Sigma–Aldrich, Japan) containing
1% HAT media supplement (the final reagents’ concen-
trations are 100 lM hypoxanthine, 0.4 lM aminopterin,
16 lM thymidine; Sigma–Aldrich, Japan), 5% penicillin/
streptomycin (P/S, Sigma–Aldrich, Japan) and 5% fetal
bovine serum (Sigma–Aldrich, Japan). Passage of cells
was performed by pipetting. Experiments were carried
out 2–3 days after seeding.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmc.2007.12.037.
References and notes
1. Murata, M.; Legrand, A.-M.; Ishibashi, Y.; Yasumoto, T.
J. Am. Chem. Soc. 1989, 111, 8927.
2. Lin, Y.-Y.; Risk, M.; Ray, S. M.; Engen, D. V.; Clardy, J.;
Golik, J.; James, J. C.; Nakanishi, K. J. Am. Chem. Soc.
1981, 103, 6773–6775.
3. Murata, M.; Iwashita, T.; Yokoyama, A.; Sasaki, M.;
Yasumoto, T. J. Am. Chem. Soc. 1992, 114, 1975.
4. Gawley, R. E.; Rein, K. S.; Kinoshita, M.; Baden, D. G.
Toxicon 1992, 30, 780–785.
5. Sasaki, M.; Noguchi, T.; Tachibana, K. J. Org. Chem.
2002, 67, 3301–3310.
6. Rando, T. A. J. Gen. Physiol. 1989, 93, 43–65.
7. Leibowitz, M. D.; Sutro, J. B.; Hille, B. J. Gen. Physiol.
1986, 87, 25–46.
8. Ulbricht, W.; Flacke, W. J. Gen. Physiol. 1965, 48, 1035–
1046.
9. Wang, S-Y.; Wang, G. K. Cell. Signal. 2003, 15, 151–159.
10. Anger, T.; Madge, D. J.; Mulla, M.; Riddall, D. J. Med.
Chem. 2001, 44, 115–137.
11. Trainer, V. L.; Brown, G. B.; Catterall, W. A. J. Biol.
Chem. 1996, 271, 11261–11267.
12. Catterall, W. A. Annu. Rev. Pharmacol. Toxicol. 1980, 20,
15–43.
13. Brown, G. B. Int. Rev. Neurobiol. 1988, 29, 76–116.
14. Wang, S.-Y.; Wang, G. K. Proc. Natl. Acad. Sci. U.S.A.
1998, 95, 2653–2658.
4.5. Voltage-gated sodium channel specific activity
assay16
One hundred microliters of cell suspensions (100,000
cells/mL) in the medium was seeded in each well of a
96-well cell culture plate and incubated for 24 h. Cell
culture medium (100 lL) containing 1 mM ouabain (fi-
nal 500 lM, Sigma–Aldrich, Japan) and various concen-
tration of veratridine or 4-methoxy veratridine (final
0.781, 1.56, 3.13, 6.25, 12.5, 25.0, 50.0, and 100 lM
each) in ethanol (solvent max 1%), or an aqueous solu-
tion of adriamysin (final 1.7 mM, Sigma–Aldrich, Ja-
pan) as a positive control or ethanol (final 1%) as a
negative control were added to each well and incubated
for 48 h. 50 lL of aqueous 3-(4,5-Dimethylthiazol-2-yl)-
2,5-diphenyltetrazolium bromide (5 mg/mL, Sigma–Al-
drich, Japan) was added and incubated at 37 °C for
4 h. The culture supernatant was then decanted and
100 lL of dimethylsulfoxide was added as a solvent.
The absorbance was measured at 570 nm subsequently
using a Thermo Varioskan Flash micro plate reader
(Thermo Fisher Scientific K.K., Japan).
15. Ikawa, M.; Dicke, R. J.; Allen, T. C.; Link, K. P. J. Biol.
Chem. 1945, 159, 517–524.
16. Ujvary, I.; Eya, B. K.; Grendell, R. L.; Toia, R. F.;
Casida, J. E. J. Agric. Food Chem. 1991, 39, 1875–1881.
17. Codding, P. W. J. Am. Chem. Soc. 1983, 105, 3172–3176.
18. Krishnamurthy, V. V.; Casida, J. E. Magn. Reson. Chem.
1988, 26, 980–988.
Acknowledgments
19. Cheethan, N. W. H.; Sirimanne, P. Carbohydr. Res. 1981,
96, 126–128.
We thank our laboratory colleague Hiroaki Suzuki and
Dr. Kazuhiko Fukui (National Institute of Advanced
Industrial Science and Technology) for technical help
with calculations in theoretical chemistry. We also thank
our laboratory colleague Takahisa Genji for his helpful
discussions. This study was financially supported by the
Suntory Institute for Bioorganic Research (SUNBOR),
Osaka, Japan.
20. Weiser, T. J. Neurosci. Methods 2004, 137, 79–85.
21. Carter, O. D.; Conejero, C.; Tornatore, C.; Poltorak, M.;
Freed, J. W. Cell Tissue Res. 1998, 291, 191–205.
22. Urenjak, J.; Obrenovitch, T. P. Amino Acids 1998, 14,
151–158.
23. Pelleter, S. W.; Jacobs, W. A. J. Am. Chem. Soc. 1953, 75,
3248–3252.