Rationale: Some anxiety disorders may be treated in a different way than no
rmal anxiety. Objective: This study was aimed at investigating the action o
f the benzodiazepine receptor antagonist flumazenil, compared to that of th
e benzodiazepine receptor full agonist chlordiazepoxide, in an animal model
of generalised anxiety disorder (the BALB/c mouse). Methods: Flumazenil (0
.0001, 0.001, 0.01, 0.1 and 1 mg/kg) or chlordiazepoxide (5 mg/kg) were adm
inistered to BALB/c or C57BL/6 mice subjected to the light/dark test, the e
levated plus maze or a passive avoidance step-through paradigm. Results: Ch
lordiazepoxide and flumazenil (at all doses tested in the elevated plus maz
e and at the doses of 0.001 and 0.01 mg/kg in the light/dark test) induced
a strong anxiolytic effect in BALB/c mice. Flumazenil did not induce anxiol
ysis in C57BL/6 mice, whatever the behavioral test or the dose used. Howeve
r, chlordiazepoxide elicited anxiolysis in this strain in both procedures.
In the passive avoidance test. chlordiazepoxide was amnesic in both strains
but flumazenil had no effect. Conclusion: Flumazenil induces partial agoni
st-like effects in BALB/c and not in C57BL/6 mice, suggesting a possible be
nzodiazepine receptor set point shift toward the agonistic direction in som
e pathological anxiety states such as generalised anxiety disorder.