6-Methylflavanone, a more efficacious positive allosteric modulator of γ-aminobutyric acid (GABA) action at human recombinant α2β2γ2L than at α1β2γ2L and α1β2 GABAA receptors expressed in Xenopus oocytes

Article abstract of DOI:10.1016/j.ejphar.2005.02.034

6-Methylflavanone acted as a positive allosteric modulator of γ-aminobutyric acid (GABA) responses at human recombinant α₁β₂γ_2L, α ₂β₂γ_2L and α₁β ₂ GABA_A receptors expressed in Xenopus laevis oocytes. It was essentially inactive at ρ₁ GABA_C receptors. The EC₅₀ values for 6-methylflavanone for the positive modulation of the EC_10-20 GABA responses were 22 μM, 10 μM and 6 μM and the maximum potentiations were 120%, 417% and 130% at α₁β ₂γ_2L, α₂β₂γ _2L and α₁β₂ GABA_A receptors respectively. Thus 6-methylflavanone was much more efficacious as a positive modulator at α₂β₂γ_2L than at α₁β₂γ_2L and α₁β₂ GABA_A receptors. This may be significant since diazepam-induced anxiolysis is considered to be mediated via α₂-containing GABA_A receptors, while sedation is thought to be mediated via α₁-containing GABA_A receptors. We have previously reported that 6-methylflavone (1-100 μM) produced positive allosteric modulation at α₁β ₂γ_2L and α₁β₂ GABA_A receptors with no significant difference between the enhancement seen at either receptor subtype. In the present study, 6-methylflavone was tested at α₂β₂γ _2L GABA_A receptors and found to maximally potentiate the EC_10-20 GABA response by 183 ± 39% which is similar to that previously observed for 6-methylflavone at α₁β ₂γ_2L GABA_A receptors. Thus, 6-methylflavone did not show a preference for α₂β ₂γ_2L over α₁β ₂γ_2L GABA_A receptors in terms of efficacy. Compared to 6-methylflavone, 6-methylflavanone is more efficacious as a positive allosteric modulator at α₂β₂γ _2L GABA_A receptors, and less efficacious at α₁β₂γ_2L GABA_A receptors. This may represent a relatively unique type of selectivity for positive modulators of GABA-A receptor subtypes based on efficacy as distinct from potency. As was previously shown for 6-methylflavone at α₁β₂γ_2L GABA_A receptors, the positive modulation of GABA responses at α ₁β₂γ_2L and α₂β ₂γ_2L GABA_A receptors by 6-methylflavanone was insensitive to antagonism by flumazenil, indicating that this action is not mediated via "high-affinity" benzodiazepine sites.

Full text of DOI:10.1016/j.ejphar.2005.02.034

European Journal of Pharmacology 512 (2005) 97–104
www.elsevier.com/locate/ejphar
6-Methylflavanone, a more efficacious positive allosteric modulator of
g-aminobutyric acid (GABA) action at human recombinant a2h2g2L than
at a1h2g2L and a1h2 GABAA receptors expressed in Xenopus oocytes
Belinda J. Halla, Mary Chebibb, Jane R. Hanrahanb, Graham A.R. Johnstona,
T
aAdrien Albert Laboratory of Medicinal Chemistry, Department of Pharmacology, Faculty of Medicine, University of Sydney, Sydney, NSW 2006, Australia
bPharmaceutical Chemistry, Faculty of Pharmacy, University of Sydney, Sydney NSW 2006, Australia
Received 25 November 2004; received in revised form 17 February 2005; accepted 22 February 2005
Available online 7 April 2005
Abstract
6-Methylflavanone acted as a positive allosteric modulator of g-aminobutyric acid (GABA) responses at human recombinant a1h2g2L
a2h2g2L and a1h2 GABAA receptors expressed in Xenopus laevis oocytes. It was essentially inactive at U1 GABAC receptors.
,
The EC50 values for 6-methylflavanone for the positive modulation of the EC10–20 GABA responses were 22 AM, 10 AM and 6 AM and
the maximum potentiations were 120%, 417% and 130% at a1h2g2L, a2h2g2L and a1h2 GABAA receptors respectively.
Thus 6-methylflavanone was much more efficacious as a positive modulator at a2h2g2L than at a1h2g2L and a1h2 GABAA receptors. This
may be significant since diazepam-induced anxiolysis is considered to be mediated via a2-containing GABAA receptors, while sedation is
thought to be mediated via a1-containing GABAA receptors.
We have previously reported that 6-methylflavone (1–100 AM) produced positive allosteric modulation at a1h2g2L and a1h2 GABAA
receptors with no significant difference between the enhancement seen at either receptor subtype. In the present study, 6-methylflavone was
tested at a2h2g2L GABAA receptors and found to maximally potentiate the EC10–20 GABA response by 183F39% which is similar to that
previously observed for 6-methylflavone at a1h2g2L GABAA receptors. Thus, 6-methylflavone did not show a preference for a2h2g2L over
a1h2g2L GABAA receptors in terms of efficacy.
Compared to 6-methylflavone, 6-methylflavanone is more efficacious as a positive allosteric modulator at a2h2g2L GABAA receptors,
and less efficacious at a1h2g2L GABAA receptors. This may represent a relatively unique type of selectivity for positive modulators of
GABA-A receptor subtypes based on efficacy as distinct from potency.
As was previously shown for 6-methylflavone at a1h2g2L GABAA receptors, the positive modulation of GABA responses at a1h2g2L and
a2h2g2L GABAA receptors by 6-methylflavanone was insensitive to antagonism by flumazenil, indicating that this action is not mediated via
bhigh-affinityQ benzodiazepine sites.
D 2005 Elsevier B.V. All rights reserved.
Keywords: GABA (g-aminobutyric acid); 6-methylflavanone; 6-methylflavone; Benzodiazepine; Flavonoid; GABAA receptor
1. Introduction
GABAA receptors are hetero-oligomeric receptors forming a
pentameric structure. At least 16 human GABAA receptor
subunits have been described and are classified under 6
subfamilies of protein subunits: a, h, g, y, q, u (Chebib and
Johnston, 2000).
GABAA receptors belong to the superfamily of ligand-
gated ion channels that include GABAC, nicotinic acetyl-
choline, strychnine-sensitive glycine and 5-HT3 receptors.
Classically, benzodiazepine agonists act as positive
modulators at a subpopulation of GABAA receptors by
increasing the frequency of chloride channel openings
(Chebib and Johnston, 2000; Rogers et al., 1994). The
T Corresponding author. Tel.: +61 2 9351 6117; fax: +61 2 9351 2891.
E-mail address: grahamj@mail.usyd.edu.au (G.A.R. Johnston).
0014-2999/$ - see front matter D 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.ejphar.2005.02.034

98
B.J. Hall et al. / European Journal of Pharmacology 512 (2005) 97–104
benzodiazepine binding sites for this action, which can be
blocked by flumazenil, are considered to lie between the a
and g subunits (Klausberger et al., 2001). Flumazenil-
insensitive positive modulation of GABAA receptors has
been described in receptors lacking a g subunit by
benzodiazepines at micromolar concentrations. For exam-
ple, Malherbe et al. (1990) found that recombinant a1h1
GABAA receptors from rat brain were sensitive to potentia-
tion by benzodiazepine receptor ligands, with both diaze-
pam and flumazenil acting as positive modulators. Walters
et al. (2000) found that classical benzodiazepines produce
biphasic potentiation at rat recombinant a1h2g2 GABAA
receptors via two distinct mechanisms. This biphasic
potentiation is believed to be mediated via two sites referred
to as high-affinity and low-affinity benzodiazepine sites.
Both sites are present at receptors composed of a1h2g2
GABAA subunits and low-affinity potentiation can be
selectively observed at receptor combinations lacking a g
subunit such as a1h2 GABAA receptors. Furthermore, low-
affinity potentiation at both receptor combinations is
insensitive to flumazenil.
O
O
2
3
H3C
Fig. 1. 6-Methylflavanone.
Some examples of flavanones that have been investigated
for an ability to displace [3H]diazepam are eriodictyol
(3V,4V,5,7-tetrahydroxyflavanone), hesperetin (3V,5,7-trihy-
droxy-4V-methoxyflavanone), 7-hydroxyflavanone, kaemp-
ferol (3,4V,5,7-tetrahydroxyflavanone), 4V-methoxyflavanone
and naringenin (4V,5,7-trihydroxyflavanone). All were found
to be inactive or only weakly active as determined by having
an IC50 value greater than 25 AM (Ai et al., 1997).
Flavanone itself had no effect on [3H]flunitrazepam
binding up to a concentration of 40 AM while flavone
inhibited [3H]flunitrazepam binding with a Ki of 1 AM
(Marder et al., 1996). Also, the flavone apigenin inhibited
benzodiazepine binding with a Ki of 4 AM (Viola et al.,
1995) while naringenin, the flavanone analogue of
apigenin, had no effect up to a concentration of 25 AM
(Ai et al., 1997). Thus, it would appear that flavones are
more active than flavanones at binding to classical
benzodiazepine sites. According to structure–activity
analysis, in order to bind to the classical benzodiazepine
site at GABAA receptors, flavonoids should be planar in
structure (Dekermendjian et al., 1999). Thus, the less
planar structure of flavanones would be expected to
reduce their affinity for classical benzodiazepine binding
sites. However, it has also been found that a planar
conformation is not an absolute requirement for flavonoid
activity at the benzodiazepine site of GABAA receptors
(Hong and Hopfinger, 2003). Hence, there remain
discrepancies regarding the structure–activity requirements
for the flavonoid pharmacophore at the benzodiazepine
site.
Flavonoids are a group of structurally related phenyl-
benzopyrones subdivided into the following main classes:
flavones, flavanones, flavonols, isoflavones, flavans and
flavan-3-ols. They are found in all plants and therefore form
part of our diet. Our estimated average daily intake of
flavonoids of 1–2 g is derived from fruits, vegetables,
chocolate and beverages such as tea, coffee and red wine as
well as herbal preparations.
Flavonoids are known to exert both peripheral and central
nervous system effects. Much of the research into flavonoids
has focussed on their peripheral actions, however more
recently, they have been the subject of many studies using rat
or bovine brain membrane binding assays which suggest that
they possess a selective affinity for the benzodiazepine
binding site at GABAA receptors (Dekermendjian et al.,
1999; Hong and Hopfinger, 2003; Huang et al., 2001; Hui et
al., 2000; Kahnberg et al., 2002; Marder et al., 1997, 1998;
Medina et al., 1998; Nielsen et al., 1988). Many behavioural
studies have also been carried out, often in conjunction with
binding studies, which indicate that flavonoids exert
anxiolytic effects in rodents without many of the unwanted
side effects of benzodiazepines (Griebel et al., 1999; Marder
et al., 1995, 1996; Salgueiro et al., 1997; Viola et al., 1995,
1997, 2000a; Wolfman et al., 1994, 1996, 1998). More
recently, this has been extended to include functional studies
of flavonoid activity at recombinant GABAA receptors (Ai et
al., 1997; Hall et al., 2004; Hanrahan et al., 2003; Huen et al.,
2003; Kavvadias et al., 2004; Viola et al., 2000b).
Previously, we described the flumazenil-insensitive
positive modulatory action of 6-methylflavone at human
recombinant a1h2g2L and a1h2 GABAA receptors (Hall et
al., 2004). To date, no studies have investigated the action
of 6-methylflavanone (Fig. 1) at GABAA receptors. Based
on previous findings, 6-methylflavanone would be pre-
dicted to be less active than 6-methylflavone on benzo-
diazepine binding. However, functional studies in our
laboratory have already shown that 6-methylflavone is a
positive modulator of GABA-induced currents through a
mechanism other than an action at classical benzodiazepine
sites. Hence, testing 6-methylflavanone in a functional
study at GABAA receptors is also likely to yield different
results to those expected to be obtained in a benzodiaze-
pine binding study.
Flavanones have not been studied as GABA receptor
modulators as extensively as the flavones. Flavanones differ
from flavones by the absence of the 2–3 double bond (Fig.
1), resulting in a less planar structure and a chiral centre at
the 2 position of the pyrone ring, giving rise to two or four
possible isomers depending on the substituents at the 3
position.

B.J. Hall et al. / European Journal of Pharmacology 512 (2005) 97–104
99
The present study set out to characterise the effects of 6-
methylflavanone at human recombinant a1h2g2L, a2h2g2L,
a1h2 and U1 GABA receptors expressed in Xenopus laevis
oocytes.
(eluent: 3% ethyl acetate in hexane). Recrystallisation from
ethanol afforded the title compound as orange crystals (1.27
g; 10%). yH (300 MHz; CDCl3) 2.36 (3 H, s, ArCH3), 6.95 (1
H, d, J 8.1, C(6)–H), 7.33 (1 H, dd, J 8.4, 2.1, C(5)–H), 7.44–
7.46 (2 H, m, C(2V)–H), 7.65–7.70 (5 H, m, HC=CH, C(3)–
H, C(3V)–H, C(4V)–H), 7.92 (1 H, d, J 15.6, HC =CH), 12.65
(1 H, s, OH).
2. Materials and methods
2.1. Drugs
2.2.4. 6-Methylflavanone
2V-Hydroxy-5V-methylchalcone (1.27 g, 5.3 mmol) was
mixed with glacial acetic acid (~50 ml) and phosphoric acid
(0.3 ml, 5.3 mmol) then heated under reflux overnight. The
acetic acid was removed under reduced pressure and the
Diazepam and flumazenil were gifts from Hoffman-La
Roche (Nutley, NJ, USA). GABA and dimethyl sulfoxide
(DMSO) were obtained from Sigma (St Louis, MO, USA).
ꢀ
product extracted into dichloromethane (2 50 ml) to give a
2.2. Synthesis of 6-methylflavanone
yellow oil (1.14 g). The crude product was recrystallised
twice from ethanol to afford the title compound as white
crystals (0.14 g; 10.7%), m.p. 107–1098C. yH (300 MHz;
CDCl3) 2.34 (3 H, s, ArCH3), 2.88 (1 H, dd, J 16.8, 3, C(3)–
HV), 3.08 (1 H, dd, J 16.8, 13.2, C(3)–H), 5.46 (1 H, dd, J
13.5, 3, C(2)–H), 6.97 (1 H, d, J 8.4, C(8)–H), 7.33 (1 H,
dd, J 8.4, 2.1, C(7)–H), 7.39–7.50 (5 H, m, C(2V)–H, C(3V)–
H, C(4V)–H), 7.73 (1 H, d, J 2.1, C(5)–H).
2.2.1. p-Cresyl benzoate
Acetic anhydride (9.2 g; 90 mmol) was added slowly
with stirring to a solution of p-cresol (8.00 g; 74 mmol) in
pyridine (100 ml). The reaction mixture was heated to
reflux, stirred overnight, then cooled to room temperature
and the solvent removed under vacuum. The crude product
was extracted into dichloromethane and the solvent
removed to give a yellow oil (10.03 g; 91%). yH (300
MHz; CDCl3) 2.28 (3 H, s, ArCH3), 2.34 (3 H, s, COCH3),
6.96 (2 H, dd, J 6.6, 2.4, ArH), 7.17 (2 H, dd, J 6.6, 2.4,
ArH).
2.3. Expression of recombinant GABA receptors in X. laevis
oocytes
The procedures involving the use of X. laevis were
carried out according to those described by Huang et al.
(2003) and were approved by the Animal Ethics Commit-
tee of the University of Sydney. Female X. laevis were
anaesthetised by immersion in 0.17% 3-aminobenzoic acid
ethyl ester with 0.02% NaCl for 10–15 min, and a lobe of
the ovaries was surgically removed. The lobe was rinsed
with oocyte releasing buffer 2 (OR2; 82.5 mM NaCl, 2
mM KCl, 1 mM MgCl2.6H2O, 5 mM HEPES, pH 7.5)
and treated with Collagenase A (2 mg/ml in OR2,
Bohringer Manheim, Germany) for 2 h to separate oocytes
from connective tissue and follicular cells. Released
oocytes were rinsed in ND96 dwashT solution (96 mM
NaCl, 2 mM KCl, 1 mM MgCl2.6H2O, 1.8 mM CaCl2, 5
mM HEPES, pH 7.5). Stage V–VI oocytes were collected
and stored on an oscillator at 16 8C in ND96 dstorageT
solution (ND96 dwashT solution supplemented with 2.5
mM pyruvate, 0.5 mM theophylline and 50 Ag/ml
gentamycin).
2.2.2. 2-Acetyl-4-methylphenol
Aluminium chloride (8.90 g, 67 mmol) was added slowly
(over approximately 1 h) to p-cresyl benzoate (10.00 g, 67
mmol) at 130 8C. The temperature was raised to 160 8C and
kept there for 45 min. The reaction mixture was cooled to
room temperature and 1 M hydrochloric acid (20 ml) was
added. The crude product was extracted into dichloro-
methane and the solvent removed to give dark brown
crystals (8.51 g; 85%). yH (300 MHz; CDCl3) 2.32 (3 H, s,
ArCH3), 2.63 (3 H, s, COCH3), 6.92 (1 H, d, J 8.1, C(6)–H),
7.33 (1 H, dd, J 8.4, 2.1, C(5)–H), 7.52 (1 H, d, J 1.2, C(3)–
H), 12.10 (1 H, s, OH).
2.2.3. 2V-Hydroxy-5V-methylchalcone
Chilled aqueous sodium hydroxide (27 ml of 10% w/v)
was added to a cooled stirred solution of benzaldehyde (5.86
g; 55 mmol) and 2-acetyl-4-methylphenol (8.29 g; 55 mmol)
in ethanol (~30 ml). Ayellow precipitate formed immediately
and became a dark orange oil after a few hours. The reaction
mixture was stirred for 48 h then poured into ice/water (100
ml) to become a tacky yellow solid. The crude product was
extracted into dichloromethane and the organic layer was
Human a1, a2, h2 and g2L DNA in pcDM8 were
provided by Dr Paul Whiting (Merck, Sharpe and Dohme
Research Labs, Harlow, UK). Human U1 DNA in
pcDNA1.1 (Invitrogen, San Diego, CA, USA) was provided
by Dr George Uhl (National Institute for Drug Abuse,
Baltimore, MD, USA). Plasmids containing a1, a2, h2 or
g2L DNA and U1 DNA were linearised using the restriction
enzymes NOT1 and Xba1, respectively. RNA was synthes-
ised from linearised DNA using the dmMessage mMachineT
kit from Ambion (Austin, TX, USA). A dNanojectT
(Drummond Scientific Co., Broomali, PA, USA) was used
ꢀ
washed with brine (2 100 ml) and dried over anhydrous
magnesium sulphate to give a dark orange oil (12.77 g). Thin
layer chromatography (TLC) analysis showed that the
reaction only went to 25% completion. Crude 2V-hydroxy-
5V-methylchalcone along with starting material was separated
from other impurities by vacuum column chromatography

100
B.J. Hall et al. / European Journal of Pharmacology 512 (2005) 97–104
to inject either a1h2g2L (20 ng/50 nl), a2h2g2L (20 ng/50
nl), a1h2 (20 ng/50 nl) or U1 (10 ng/50 nl) RNA dissolved in
nuclease-free water into the cytoplasm of the oocytes, which
were then stored at 16 8C in ND96 storage solution. For
a1h2g2L and a2h2g2L GABAA receptors the RNA for each
subunit was injected in the ratio 1:1:2 respectively to favour
incorporation of the g subunit. For a1h2 GABAA receptors
the RNA for each subunit was injected in a 1:1 ratio. Sham-
injected oocytes were prepared by injection with nuclease-
free water.
nonlinear regression fit was performed using a sigmoidal
dose–response (variable slope), the equation of which is—
nH
I ¼ Imax=ð1 þ ½EC50=ðAފ Þ
where [A] is the agonist concentration, I is the current and
Imax is the maximum current. EC50 is the concentration of
agonist that produces a response that is 50% of the
maximum current and nH is the Hill coefficient. EC50
values are expressed as mean with 95% confidence
intervals. Hill coefficients (nH) are expressed as meanF
S.E.M. Significant differences were determined using two-
way analysis of variances (ANOVAs) producing F and P
values which have been quoted where appropriate.
2.4. Recording from oocytes
Two to seven days after injection of RNA, receptor
activity was measured by two-electrode voltage clamp
recording. Glass micropipettes for the electrodes were made
using a micropipette puller (Narishige Scientific Instrument
Lab, Tokyo, Japan) and filled with 3 M potassium chloride.
Oocytes were transferred into a cell bath, impaled by two
micropipettes, and voltage clamped using a Geneclamp 500
amplifier (Axon Instruments Inc., Foster City, CA, USA).
The membrane potential was clamped at ꢁ60 mV and the
electrodes had resistance values between 0.5 and 2 Mg. In
the cell bath, oocytes were continuously superfused with
ND96 solution and drugs were applied in the perfusate.
Current traces were recorded using a Mac Lab 2e recorder
(ADInstruments, Sydney, NSW, Australia) and Chart
v.3.5.2.
3. Results
3.1. Direct activity at GABA receptors
6-Methylflavanone (100 AM) did not have any effect at
sham-injected oocytes (n=3). Also, 6-methylflavanone (100
AM) had no activity at a1h2g2L, a2h2g2L, a1h2 and U1
GABA receptors when administered alone.
3.2. 6-Methylflavanone as a positive allosteric modulator
with a higher efficacy at a2b2c2L GABAA receptors
compared to a1b2c2L and a1b2 GABAA receptors
Oocytes expressing a1h2g2L or a2h2g2L GABAA recep-
tors were screened with 10 AM Zn2+ in the presence of two
concentrations of GABA (10 AM and 100 AM) to ensure
that the g subunit was incorporated. GABAA receptors
without a g subunit are sensitive to inhibition by zinc,
whereas those expressing a g subunit are not (Hosie et al.,
2003). To test for positive modulation, low GABA doses at
each subtype were chosen for the control doses since
benzodiazepines produce greatest enhancement of the
GABA response at lower GABA doses. 5 AM GABA, 3
AM GABA and 1 AM GABA were used as control GABA
doses at a1h2g2L, a2h2g2L and a1h2 GABAA receptors
respectively as these doses corresponded to approximately
an EC10 response at each subtype. Stock solutions in DMSO
of 100 mM (6-methylflavone, diazepam and flumazenil)
concentrations were prepared. GABA stock solutions (100
mM) were prepared in milli-Q water. For experiments where
any drugs were dissolved in DMSO, all drug solutions were
standardised to contain 0.8% DMSO, which had no
significant effect on the oocytes.
6-Methylflavanone enhanced the response to 3 AM GABA
at a2h2g2L GABAA receptors (n =4–5; deviation from zero:
significant, F=55.45, P b0.0001). The EC50 for this effect
was 9.9 AM (95% CI: 7.01 to 14.02) with a Hill coefficient of
1.36F0.28. 6-Methylflavanone maximally enhanced the
response to 3 AM GABA by 417F30% at a concentration
of 60 AM (n=3; Fig. 2A) and shifted the GABA dose–
response curve at a2h2g2L GABAA receptors to the left (n =3;
two-way ANOVA: F =105.41, Pb0.0001), decreasing the
mean GABA EC50 from 18.2 AM (95% CI: 14.46 to 22.77) to
2.60 AM (95% CI: 1.935 to 3.481; Fig. 3A).
6-Methylflavanone also enhanced the response to a low
dose of GABA at a1h2g2L (n =4–5; deviation from zero:
significant, F=60.56, P b0.0001; Fig. 2B) and a1h2 (n=3–
6; deviation from zero: significant, F =26.45, Pb0.0001;
Fig. 2C) GABAA receptors, however the level of enhance-
ment was lower than that observed at a2h2g2L GABAA
receptors. At a1h2 GABAA receptors, the EC50 value for 6-
methylflavanone-induced enhancement was 5.72 AM (95%
CI: 3.09 to 10.6). At a1h2g2L GABAA receptors, an
accurate EC50 value could not be determined as the dose–
response curve was incomplete due to the solubility limit for
6-methylflavanone. At a concentration of 100 AM, 6-
methylflavanone enhanced the response to 5 AM GABA
at a1h2g2L GABAA receptors and 1 AM GABA at a1h2
GABAA receptors by 120F11% (n=5) and 116F36%
(n =6) respectively. 6-Methylflavanone shifted the GABA
2.5. Data analysis
Standardised responses were plotted against drug con-
centration on a semi-logarithmic scale using GraphPad
Prism version 2.0. Logarithmically transformed data were
tested for significance using a linear regression fit. Provided
the slope of the curve significantly deviated from zero, a

B.J. Hall et al. / European Journal of Pharmacology 512 (2005) 97–104
101
A
3.3. 6-Methylflavone as a positive allosteric modulator at
a2b2c2L GABAA receptors
500
400
300
200
100
0
γ
α2β2 2L GABAA receptors
6-Methylflavone also enhanced the response to 3 AM
GABA at a2h2g2L GABAA receptors (n =3–5, deviation
from zero: significant, F=35.35, P b0.0001), however this
was to a lesser extent than that by 6-methylflavanone.
Solubility constraints prevented the determination of an
accurate EC50 for this effect of 6-methylflavone at a2h2g2L
GABAA receptors; however, 6-methylflavone (100 AM)
maximally potentiated the control response to GABA by
183F39% (n =5; Fig. 2A). 6-Methylflavone (60 AM)
0.01
0.1
1
10
100
Flavonoid Concentration (µM)
B
500
400
300
200
100
0
A
GABA + 60µM 6-Methylflavanone
γ
100
α1β2 2L GABAA receptors
GABA
50
α2β2γ2L GABAA receptors
0.01
0.1
1
10
100
Flavonoid Concentration (µM)
0
0.01
0.1
1
10
100
1000
C
GABA Concentration (µM)
500
400
300
200
100
0
B
α1β2 GABAA receptors
GABA + 60µM 6-Methylflavanone
100
GABA
50
0
α1β2γ2L GABAA receptors
0.01
0.1
1
10
100
Flavonoid Concentration (µM)
0.1
1
10
100
1000
Fig. 2. Enhancement of the response to a low dose of GABA by 6-
methylflavanone (n; solid line) and 6-methylflavone (5; dotted line) at (A)
a2h2g2L, (B) a1h2g2L and (C) a1h2 GABAA receptors. Control GABA
doses used at each receptor subtype were 3 AM, 5 AM and 1 AM
respectively. Data for 6-methylflavone in B, C are from Hall et al. (2004).
Data are expressed as meanFS.E.M.
GABA Concentration (µM)
C
GABA + 60µM 6-Methylflavone
100
50
0
GABA
dose–response curve at a1h2g2L (n =4; two-way ANOVA:
F =19.04, P b0.0001) GABAA receptors to the left, decreas-
ing the mean GABA EC50 from 21.47 AM (95% CI: 17.18
to 26.84) to 13.39 AM (95% CI: 8.87 to 20.22; Fig. 3B).
6-Methylflavanone (0.01–100 AM) did not enhance the
response to 0.3 AM GABA at U1 GABAC receptors (data not
shown). Linear regression analysis produced a line with a
slope that did not significantly deviate from zero (n=3;
F =1.668, P=0.2129). 6-Methylflavanone also had no
significant effect on the GABA dose–response curve at U1
GABAC receptors (n =5; two-way ANOVA: F =0.13,
P=0.7214; data not shown).
α2β2γ2L GABAA receptors
0.1
1
10
100
1000
GABA Concentration (µM)
Fig. 3. Enhancement of the GABA dose–response curves at (A) a2h2g2L
,
(B) a1h2g2L GABAA receptors by 60 AM 6-methylflavanone and (C)
a2h2g2L GABAA receptors by 60 AM 6-methylflavone. Data are expressed
as meanFS.E.M.

102
B.J. Hall et al. / European Journal of Pharmacology 512 (2005) 97–104
shifted the GABA dose–response curve a2h2g2L (n=4, two-
way ANOVA: F =5.57, P=0.0219) GABAA receptors to the
left, decreasing the mean GABA EC50 from 15.7 AM (95%
CI: 13.13 to 18.74) to 9.46 AM (95% CI: 7.26 to 12.3;
Fig 3C).
dependent on the presence of a g subunit, which is required
for the formation of classical benzodiazepine sites.
Although flumazenil (10 AM) slightly enhanced the
response to 3 AM GABA in the presence of 3 AM 6-
methylflavanone at a2h2g2L GABAA receptors (Fig. 4A), it
is unlikely that this was responsible for the apparent
flumazenil insensitivity of 6-methylflavanone-induced pos-
itive modulation, since under the same conditions, diaze-
pam-induced positive modulation was completely abolished
by the presence of flumazenil (0.1–10 AM; Fig. 4B).
3.4. Flumazenil insensitive action of 6-methylflavanone
Application of increasing concentrations of flumazenil
(0.01–10 AM) in the presence of GABA (3 AM) and 6-
methylflavanone (3 AM) did not block the 6-methylflava-
none-induced enhancement of the GABA response at
a2h2g2L GABAA receptors (Fig. 4A). Similar results were
observed at a1h2g2L GABAA receptors using 5 AM GABA
in the presence of 10 AM 6-methylflavanone. These 6-
methylflavanone concentrations were chosen to test for
flumazenil sensitivity as they represent approximately the
lowest dose producing observable enhancement at each
receptor subtype. In contrast, the enhancement of the GABA
(3 AM) response produced by diazepam (1 AM) was
completely abolished by 0.1 AM flumazenil at a2h2g2L
GABAA receptors (Fig. 4B).
6-Methylflavanone enhanced the response to GABA at
a2h2g2L, a1h2g2L and a1h2 GABAA receptors (Fig. 2A–C),
but not at U1 GABAC receptors. A complete dose–response
curve for 6-methylflavanone action at a1h2g2L GABAA
receptors was not obtained, as the limit of solubility for 6-
methylflavanone in 0.8% DMSO is 100 AM. This makes a
comparison between 6-methylflavanone action at a1h2g2L
and a1h2 GABAA receptors difficult, however two-way
ANOVA showed that there was no significant difference
between the dose–response curves for 6-methylflavanone
enhancement at each receptor subtype ( F =3.01, P=0.088).
It seems that 6-methylflavanone was more potent at a1h2
GABAA receptors, since a maximal response was reached
and an EC50 value obtained. The EC50 value of 5.75 AM
also appears to be lower than any possible estimate for an
EC50 value at a1h2g2L GABAA receptors, since, at these
receptors, 6-methylflavanone at a concentration of 5 AM
produced ~40% enhancement of the GABA response which
was only about a third of the maximal observed enhance-
ment of 120F11% at 100 AM. At a2h2g2L GABAA
receptors, 6-methylflavanone enhanced the control response
to GABA with an EC50 of 9.9 AM which is comparable to,
but slightly higher than, the EC50 at a1h2 GABAA receptors.
More significant however, was the difference in 6-
methylflavanone efficacy at a2h2g2L GABAA receptors
compared with the a1-containing GABAA receptor subtypes
tested. 6-Methylflavanone had a similar efficacy at both
a1h2g2L and a1h2 GABAA receptors, producing a maximal
enhancement at each receptor subtype of 120F11% and
130F24% respectively (t-test showed no significant differ-
ence: P=0.727). In contrast, at a2h2g2L GABAA receptors,
6-methylflavanone maximally enhanced the response to
GABA by 417F30% (t-tests showed significant differences
between a2h2g2L and each a1-containing subtype:
P b0.0001 for a1h2g2L and P=0.0002 for a1h2). Fig. 2A,
B and C are all displayed with a y-axis spanning from 0 to
500% to illustrate this difference more clearly. Two-way
ANOVA showed that there was a significant difference
between the dose–response curves for 6-methylflavanone
enhancement at a1h2g2L and a2h2g2L GABAA receptors
( F=160.25, P b0.0001).
4. Discussion
These results show that 6-methylflavanone, like 6-
methylflavone, is an allosteric positive modulator at human
recombinant a1h2g2L, a2h2g2L and a1h2 GABAA expressed
in X. laevis oocytes, as it produced no response when
administered alone but enhanced the response to a low dose
of GABA. This enhancement by 6-methylflavanone does
not appear to be mediated via classical benzodiazepine sites,
since it was insensitive to antagonism by flumazenil (Fig.
4A). Furthermore, the action of 6-methylflavanone was not
6-Methylflavanone (60 AM) shifted the GABA dose–
response curve to the left at both a1h2g2L and a2h2g2L
GABAA receptors (Fig. 3A–B), but not at U1 GABAC
receptors (data not shown). The greatest shift was seen at
a2h2g2L GABAA receptors, where 6-methylflavanone
Fig. 4. Representative current traces from individual oocytes showing that
potentiation of the GABA (3 AM) response at human recombinant a2h2g2L
GABAA receptors by (A) 6-methylflavanone (3 AM) is not inhibited by
flumazenil (0.01–10 AM) whereas under the same conditions, (B)
potentiation by 1 AM diazepam is inhibited by flumazenil.

B.J. Hall et al. / European Journal of Pharmacology 512 (2005) 97–104
103
decreased the mean GABA EC50 to 2.60 AM (Fig. 3A),
which is only 14% of the original mean GABA EC50 value
of 18.2 AM. At a1h2g2L GABAA receptors, 6-methylflava-
none decreased the mean GABA EC50 to 62% of the
original value (Fig. 3B).
3A), while 6-methylflavone only decreased the mean
GABA EC50 to 60% of its original value (Fig. 3C).
In conclusion, 6-methylflavanone is a flumazenil-insen-
sitive, positive modulator at GABAA receptors. This action
appears to be selective for GABAA receptors, since 6-
methylflavanone was inactive at U1 GABAC receptors. At
GABAA receptors, 6-methylflavanone, unlike 6-methylfla-
vone (Hall et al., 2004), displayed some subtype selectivity,
as it was much more efficacious at GABAA receptors of the
a2h2g2L subtype compared with the a1h2g2L and a1h2
subtypes. The greater efficacy of 6-methylflavanone at
a2h2g2L GABAA receptors compared to the a1-containing
GABAA receptors may be important behaviourally, as it is
known that the actions of benzodiazepines depend on the
GABAA receptor subtype, with the type of a subunit present
being particularly important.
Previously, we investigated the effect of 6-methylfla-
vone at recombinant GABAA receptors expressed in
Xenopus oocytes and found that it is a flumazenil-
insensitive, positive modulator at recombinant GABAA
receptors of a1h2g2L and a1h2 subtypes (Hall et al., 2004).
Comparing our previous results with those of the current
study, it appears that 6-methylflavone had a slightly higher
efficacy at a1h2g2L GABAA receptors, since the maximal
observed enhancement at 100 AM 6-methylflavone
(183F20%; Hall et al., 2004) was higher than that
produced by 6-methylflavanone in the current study
(120F11%). In contrast to 6-methylflavone, 6-methylfla-
vanone is a racemic mixture of two enantiomers, thus the
concentration of the active component may be only half
the stated concentration. At a1h2 GABAA receptors, 6-
methylflavone and 6-methylflavanone appear to have
similar potencies, with EC50 values of 8.7 AM (Hall et
al., 2004) and 5.7 AM respectively. 6-Methylflavanone also
had a similar efficacy to 6-methylflavone at a1h2 GABAA
receptors, the former maximally enhancing the response to
1 AM GABA in the current study by 130F24% compared
with the maximal enhancement of 116F27% produced by
6-methylflavone in the previous study (Hall et al., 2004).
The effect of each compound on the GABA dose–response
curve at a1h2g2L GABAA receptors was similar, as 6-
methylflavanone and 6-methylflavone decreased the mean
GABA EC50 values to 62% and 67% (Hall et al., 2004) of
their original values, respectively.
Rudolph et al. (1999) showed that a1-containing
receptors are responsible for mediating diazepam-induced
sedation, anterograde amnesia and impairment of locomotor
activity, while a2-containing GABAA receptors have been
shown to be responsible for mediating diazepam-induced
anxiolysis (Lo¨w et al., 2000). Therefore, it is possible that a
compound like 6-methylflavanone that has a greater efficacy
at a2h2g2L receptors may be useful for treating anxiety with
a reduction in some of the other effects of classical
benzodiazepines, particularly sedation. Clearly 6-methylfla-
vanone as a racemic mixture is not as potent with an EC50 of
9.9 AM at a2h2g2L GABAA receptors, but resolution of the
active isomer may afford a compound with a likely EC50 of
~5 AM. 6-Methylflavanone provides a lead for the discovery
of more potent analogues as positive GABA modulators
with an even greater selectivity for a2-containing GABAA
receptors.
In the current study, 6-methylflavone was tested at
a2h2g2L GABAA receptors for comparison with 6-methyl-
flavanone since this subtype was not investigated in our
previous study on 6-methylflavone (Hall et al., 2004). Fig.
2A shows that 6-methylflavanone had both a greater
efficacy and potency than 6-methylflavone. While no
EC50 value was obtained for the enhancing effect of 6-
methylflavone at these receptors, the EC50 value of 9.9 AM
for 6-methylflavanone was well below any EC50 value that
could be estimated for 6-methylflavone, hence 6-methyl-
flavanone appears to be more potent. For instance, at a
concentration of 10 AM, 6-methylflavone produced an
average enhancement which was less than a quarter of the
average enhancement observed at the maximal test concen-
tration of 100 AM. The efficacy of 6-methylflavanone at
a2h2g2L GABAA receptors was much higher than that of 6-
methylflavone since the former produced a maximal
enhancement of 417F30% compared with 183F39% for
6-methylflavone. Furthermore, 60 AM 6-methylflavanone
shifted the GABA dose–response curve at a2h2g2L GABAA
receptors to the left to a greater extent than 60 AM 6-
methylflavone. 6-Methylflavanone decreased the mean
GABA EC50 to a mere 14% of its original value (Fig.
Acknowledgements
We are grateful to Dr Paul Whiting (Merck, Sharpe and
Dohme Research Laboratories, Harlow, UK) for the gift of
human a1, a2, h2 and g2L DNA and Dr George Uhl
(National Institute for Drug Abuse, Baltimore, MD, USA)
for the gift of human U1 DNA. We are also grateful to Dr
Hue Tran, Kong Li, Dr Erica Campbell and Suzanne Habjan
for performing the surgery to provide the oocytes and the
National Health and Medical Research Council of Australia
for financial support.
References
Ai, J., Dekermendjian, K., Wang, X., Nielsen, M., Witt, M., 1997. 6-
Methylflavone, a benzodiazepine receptor ligand with antagonistic
properties on rat brain and human recombinant GABAA receptors in
vitro. Drug Dev. Res. 41, 99–106.
Chebib, M., Johnston, G.A.R., 2000. GABA-activated ligand gated ion
channels: medicinal chemistry and molecular biology. J. Med. Chem.
43, 1–21.

104
B.J. Hall et al. / European Journal of Pharmacology 512 (2005) 97–104
Dekermendjian, K., Kahnberg, P., Witt, M.R., Sterner, O., Nielsen, M.,
Liljefors, T., 1999. Structure–activity relationships and molecular
modeling analysis of flavonoids binding to the benzodiazepine site of
the rat brain GABAA receptor complex. J. Med. Chem. 42, 4343–4350.
Griebel, G., Perrault, G., Tan, S., Schoemaker, H., Sanger, D.J., 1999.
Pharmacological studies on synthetic flavonoids: comparison with
diazepam. Neuropharmacology 38, 965–977.
ligand for the central benzodiazepine receptors is a member of a family
of active flavonoids. Biochem. Biophys. Res. Commun. 223, 384–389.
Marder, M., Zinczuk, J., Colombo, M.I., Wasowki, C., Viola, H., Wolfman,
C., Medina, J.H., Ruveda, E.A., Paladini, A., 1997. Synthesis of
halogenated/nitrated flavone derivatives and evaluation of their affinity
for the central benzodiazepine receptor. Bioorg. Med. Chem. Lett. 7,
2003–2008.
Hall, B.J., Chebib, M., Hanrahan, J.R., Johnston, G.A.R., 2004. Flumaze-
nil-independent positive modulation of g-aminobutyric acid action by
6-methylflavone at human recombinant a1h2g2L and a1h2 GABAA
receptors. Eur. J. Pharmacol. 491, 1–8.
Marder, M., Viola, H., Bacigaluppo, J.A., Colombo, M.I., Wasowki, C.,
Wolfman, C., Medina, J.H., Ruveda, E., Paladini, A., 1998. Detection of
benzodiazepine receptor ligands in small libraries of flavone derivatives
synthesized by solution phase combinatorial chemistry. Biochem.
Biophys. Res. Commun. 249, 481–485.
Hanrahan, J.R., Chebib, M., Davucheron, N.L.M., Hall, B.J., Johnston,
G.A.R., 2003. Semisynthetic preparation of amentoflavone: a
negative modulator at GABAA receptors. Bioorg. Med. Chem. Lett.
13, 2281–2284.
Medina, J.H., Viola, H., Wolfman, C., Marder, M., Wasowski, C., Calvo,
D., Paladini, A.C., 1998. Neuroactive flavonoids: new ligands for the
benzodiazepine receptors. Phytomedicine 5, 235–243.
Hong, X., Hopfinger, A.J., 2003. 3D-pharmacophores of flavonoid binding
at the benzodiazepine GABAA receptor site using 4D-QSAR analysis.
J. Chem. Inf. Comput. Sci. 43, 324–336.
Nielsen, M., Frokjaer, S., Braestrup, C., 1988. High affinity of the
naturally-occurring biflavonoid, amentoflavon, to brain benzodiazepine
receptors in vitro. Biochem. Pharmacol. 37, 3285–3287.
Hosie, A.M., Dunne, E.L., Harvey, R.J., Smart, T.G., 2003. Zinc-mediated
inhibition of GABAA receptors: discrete binding sites underlie subtype
specificity. Nat. Neurosci. 6, 362–369.
Rogers, C.J., Twyman, R.E., Macdonald, R.L., 1994. Benzodiazepine and
h-carboline regulation of single GABAA receptor channels of mouse
spinal neurons in culture. J. Physiol. 475, 69–82.
Huang, X.Q., Liu, T., Gu, J.D., Luo, X.M., Ji, R.Y., Cao, Y., Xue, H.,
Wong, J.T.F., Wong, B.L., Pei, G., Jiang, H.L., Chen, K.X., 2001. 3D-
QSAR model of flavonoids binding at benzodiazepine site in GABAA
receptors. J. Med. Chem. 44, 1883–1891.
Rudolph, U., Crestani, F., Benke, D., Brunig, I., Benson, J.A., Fritschy, J.,
Martin, J.R., Bluethmann, H., Mohler, H., 1999. Benzodiazepine
actions mediated by specific g-aminobutyric acidA receptor subtypes.
Nature 401, 796–800.
Huang, S.H., Duke, R.K., Chebib, M., Sasaki, K., Wada, K., Johnston,
G.A.R., 2003. Bilobalide, a sesquiterpene trilactone from Gingko
biloba, is an antagonist at recombinant a1h2g2L GABAA receptors.
Eur. J. Pharmacol. 464, 1–8.
Salgueiro, J.B., Ardenghi, P., Dias, M., Ferreira, M.B.C., Izquierdo, I.,
Medina, J.H., 1997. Anxiolytic and synthetic flavonoid ligands of the
central benzodiazepine receptor have no effect on memory tasks in rats.
Pharmacol. Biochem. Behav. 58, 887–891.
Huen, M.S.Y., Hui, K., Leung, J.W.C., Sigel, E., Baur, R., Wong, T., Xue,
H., 2003. Naturally occurring 2V-hydroxyl-substituted flavonoids as
high-affinity benzodiazepine site ligands. Biochem. Pharmacol. 66,
2397–2407.
Viola, H., Wasowki, C., Levi de Stein, M., Wolfman, C., Silveira, R., Dajas,
F., Medina, J.H., Paladini, A., 1995. Apigenin, a component of
Matricaria recutita flowers, is a central benzodiazepine receptors-
ligand with anxiolytic effects. Planta Med. 61, 213–216.
Hui, K.M., Wang, X.H., Xue, H., 2000. Interaction of flavones from the
roots of Scutellaria baicalensis with the benzodiazepine site. Planta
Med. 66, 91–93.
Viola, H., Marder, M., Wolfman, C., Wasowki, C., Medina, J.H., Paladini,
A., 1997. 6-Bromo-3V-nitroflavone, a new high affinity benzodiazepine
receptor agonist recognizes two populations of cerebral cortical binding
sites. Bioorg. Med. Chem. Lett. 7, 373–378.
Kahnberg, P., Lager, E., Rosenberg, C., Schougaard, J., Camet, L., Sterner,
O., Nielsen, E.O., Nielsen, M., Liljefors, T., 2002. Refinement and
evaluation of a pharmacophore model for flavone derivatives binding to
the benzodiazepine site of the GABAA receptor. J. Med. Chem. 45,
4188–4201.
Viola, H., Marder, M., Wasowski, C., Giorgi, O., Paladini, A.C., Medina,
J.H., 2000a. 6,3V-Dibromoflavone and 6-nitro-3V-bromoflavone: new
additions to the 6,3V-disubstituted flavone family of high-affinity
ligands of the brain benzodiazepine binding site with agonistic
properties. Biochem. Biophys. Res. Commun. 273, 694–698.
Viola, H., Wolfman, C., Marder, M., Goutman, J.D., Bianchin, M.,
Wasowski, C., Calvo, D.J., Izquierdo, I., Paladini, A.C., Medina,
J.H., 2000b. 6-Chloro-3V-nitroflavone is a potent ligand for the
benzodiazepine binding site of the GABAA receptor devoid of intrinsic
activity. Pharmacol. Biochem. Behav. 65, 313–320.
Kavvadias, D., Sand, P., Youdim, K.A., Qaiser, M.Z., Rice-Evans, C., Baur,
R., Sigel, E., Rausch, W., Riederer, P., Schreier, P., 2004. The flavone
hispidulin, a benzodiazepine receptor ligand with positive allosteric
properties, traverses the blood–brain barrier and exhibits anticonvulsive
effects. Br. J. Pharmacol. 142, 811–820.
Klausberger, T., Sarto, I., Ehya, N., Fuchs, K., Furtmuller, R., Mayer,
B., Huck, S., Sieghart, W., 2001. Alternate use of distinct
intersubunit contacts controls GABAA receptor assembly and stoichi-
ometry. J. Neurosci. 21, 9124–9133.
Walters, R.J., Hadley, S.H., Morris, D.W., Amin, J., 2000. Benzodiazepines
act on GABAA receptors via two distinct and separable mechanisms.
Nat. Neurosci. 3, 1274–1281.
Lfw, K., Crestani, F., Keist, R., Benke, D., Brunig, I., Benson, J.A.,
Fritschy, J., Rulicke, T., Bluethmann, H., Mohler, H., Rudolph, U.,
2000. Molecular and neuronal substrate for the selective attenuation of
anxiety. Science 290, 131–134.
Wolfman, C., Viola, H., Paladini, A., Dajas, F., Medina, J.H., 1994.
Possible anxiolytic effects of chrysin, a central benzodiazepine receptor
ligand isolated from Passiflora Coerulea. Pharmacol. Biochem. Behav.
47, 1–4.
Malherbe, P., Draguhn, A., Multhaup, G., Beyreuther, K., Mohler, H., 1990.
GABAA-receptor expressed from rat brain a- and h-subunit cDNAs
displays potentiation by benzodiazepine receptor ligands. Mol. Brain
Res. 8, 199–208.
Wolfman, C., Viola, H., Marder, M., Wasowski, C., Ardenghi, P., Izquierdo,
I., Paladini, A.C., Medina, J.H., 1996. Anxioselective properties of 6,3V-
dinitroflavone, a high-affinity benzodiazepine receptor ligand. Eur. J.
Pharmacol. 318, 23–30.
Marder, M., Viola, H., Wasowski, C., Wolfman, C., Waterman, P.G.,
Medina, J.H., Paladini, A., 1995. 6,3V-Dinitroflavone, a novel high
affinity ligand for the benzodiazepine receptor with potent anxiolytic
properties. Bioorg. Med. Chem. Lett. 5, 2717–2720.
Wolfman, C., Viola, H., Marder, M., Ardenghi, P., Wasowki, C., Schroder,
N., Izquierdo, I., Ruveda, E., Paladini, A., Medina, J.H., 1998.
Pharmacological characterisation of 6-bromo-3V-nitroflavone, a syn-
thetic flavonoid with high affinity for the benzodiazepine receptors.
Pharmacol. Biochem. Behav. 61, 239–246.
Marder, M., Viola, H., Wasowki, C., Wolfman, C., Waterman, P.G., Cassels,
B.K., Medina, J.H., Paladini, A., 1996. 6-Bromoflavone, a high affinity