In previous research, the drug flumazenil has been categorized both as
a pure benzodiazepine antagonist and as a benzodiazepine partial agon
ist. The following studies used an elevated plus maze to test whether
flumazenil would exert any antianxiety action in rats. While chlordiaz
epoxide (3.0 mg/kg), ethanol (0.75 g/kg), and the atypical benzodiazep
ine zolpidem (1.0 mg/kg) all significantly increased time spent on the
open arms and percent open arm entries, flumazenil (1-10 mg/kg) alone
did not produce any anxiolytic effects on the maze. Withdrawal from c
hronic ethanol treatment led to a decrease in open arm time and percen
t open arm entries. Flumazenil (3.0 mg/kg) blocked these changes, sugg
esting that the effects of flumazenil are at least partially dependent
upon the levels of stress or anxiety in the subjects. An anxiolytic a
ction of flumazenil was not seen following the central administration
of the neuropeptide corticotropin-releasing factor (CRF), which reduce
d open arm time on the elevated plus maze. These results support the h
ypothesis that the mechanism of action for flumazenil effects on the a
nxiety observed during ethanol withdrawal involves antagonism of an en
dogenous benzodiazepine inverse agonist, rather than activity as a par
tial agonist or blockade of CRF-mediated effects.