The operant conflict test is one of the most commonly used animal models of anxiety. It was developed by Geller and Seifter (1960) and is therefore often referred to as the Geller-Seifter test. It is based on the approach- avoidance test of Masserman and Yum (1946). Ap- proach behavior is induced by stimulating ingestive behavior by food deprivation but is simultaneously re- duced by punishment (an aversive event such as electric shock), thus inducing conflict in the animal. The con- flict paradigm consists of two components in which freely moving rats are trained to press a lever for food re- ward. In one component, lever-pressing is reinforced with food after a variable interval of time (i.e., the “un- punished” period). In the second component, a short signal (a tone or light) precedes the delivery of a mild electric shock in conjunction with food reward (i.e., the “punished” period). After several weeks of training, the animal presses the lever much less frequently when the signal is on because it is assumed that the animal is anx- ious about the impending delivery of the shock. Thus, the anxiety in this model is conditioned. Both benzodi- azepines and barbiturates increase the response during
the punished period (also called anticonflict effect); that is, the animal accepts significantly more electric shocks, which is taken to reflect anxiolytic properties of the drug under study (for reviews, see Iverson 1980; Pollard and Howard 1990). Modification of the Geller-Seifter paradigm by using incremental shock levels beginning at zero has provided a distinct improvement over the original paradigm and is the procedure most often used today. Employment of this paradigm produced drug ef- fects that were qualitatively similar to those reported when the original procedures were used.
A variant of the Geller-Seifter model of anxiety is Vogel’s conflict test (Vogel et al. 1971). In this test, wa- ter-deprived naïve animals are placed in an experimen- tal chamber with a water tube; they are allowed to drink briefly and then are periodically punished with electric shocks for licking water, thus suppressing licking behav- ior. Benzodiazepines and barbiturates reverse this sup- pression. The advantage of this model over the Geller- Seifter test is that it is faster and simpler and does not involve training of animals. However, Vogel’s original conflict test used unconditioned suppression of drink- ing and could not distinguish nonspecific effects such as sedation or ataxia, which confounded the interpretation of the results. Consequently, this model has been fur- ther modified by the inclusion of an unpunished re- sponding component and a tone signaling the onset of the shock, as well as training of animals to stable base- lines (Ford et al. 1979). This model is often referred to as the conditioned suppression of drinking or conditioned lick-suppression test. Comparisons with the Geller- Seifter test found that this test is equally sensitive to the antianxiety properties of benzodiazepines and barbitu- rates. In the remainder of the discussion, the effects of various drug treatments that use the Geller-Seifter test, Vogel’s conflict test, or conditioned suppression of drinking test are described together.
Treatment with clinically effective anxiolytic drugs outside of the benzodiazepine or barbiturate class has produced inconsistent results in the operant conflict tests. Chronic treatment with tricyclic antidepressants (TCAs) and monoamine oxidase inhibitors, but not the selective serotonin reuptake inhibitor (SSRI) fluoxetine, which are effective in the treatment of panic disorder and other anxiety disorders, resulted in a gradual in- crease in punished drinking behavior without affecting water intake (Table 3–1). Remarkably, this effect paral- lels the slow onset of therapeutic effect observed in panic disorder patients receiving antidepressant treat- ment. Clearly, additional work with SSRIs and the
TABLE 3–1. Effects of chronic treatment with antidepressant drugs in various animal models of anxiety
Model Compound Species Effect Reference
Operant conflict Imipramine Rats + Fontana and Commissaris 1988
Desipramine Rats + Fontana et al. 1989
Amitriptyline Rats + Fontana et al. 1989
Phenelzine Rats + Fontana et al. 1989
Pargyline Rats + Commissaris et al. 1995
Fluoxetine Rats 0 Beaufour et al. 1999
Open field Imipramine Rats – Dwyer and Roy 1993
Fluoxetine Mice + Dulawa et al. 2004
Citalopram Rats + Kugelberg et al. 2002
Elevated plus-maze Phenelzine Rats 0 File 1995; Johnston and File 1988
Imipramine Rats 0 File and Johnston 1987
Imipramine Mice 0 Cole and Rodgers 1995
Maprotiline Mice + /0 Rodgers et al. 1997b
Mianserin Rats + Rocha et al. 1994
Cianopramine Rats + Griebel et al. 1994
Fluvoxamine Mice 0 Rodgers et al. 1997b
Paroxetine Rats + Cadogan et al. 1992
Sertraline Rats – Kurt et al. 2000
Fluoxetine Rats + Kurt et al. 2000
Fluoxetine Rats 0 Silva and Brandão 2000
Defensive burying Desipramine Rats 0 Beardslee et al. 1990
Desipramine Rats + Fernandez-Guasti et al. 1999
Imipramine Rats + Craft et al. 1988
Social interaction Clomipramine Rats 0 File 1985
Imipramine Rats 0 Pellow and File 1987
Phenelzine Rats – Johnston and File 1988
Maprotiline Mice + /0 Cutler et al. 1997b
Paroxetine Rats + Lightowler et al. 1994
Fluvoxamine Mice 0 Cutler et al. 1997b
Isolation-induced ultrasonic vocalizations Clomipramine Rats 0 Winslow and Insel 1990
Fluoxetine Guinea pigs + Kramer et al. 1998
Substance P recep-
tor antagonists Guinea pigs + Kramer et al. 1998
Maternal deprivation Paroxetine Rats + Ladd et al. 2000
Fluoxetine Rats + Ruedi-Bettschen et al. 2004
Desipramine Rats + MacQueen et al. 2003
selective serotonin-norepinephrine reuptake inhibitors, effective in certain anxiety disorders, is warranted. An anxiolytic profile of 5-HT1A receptor agonists has been observed in this test in chronic treatment in rats but not in mice (Table 3–2). Neither acute nor chronic treat- ment with the β-adrenergic antagonist propranolol, used often to treat fear of public speaking, affected pun- ished responding in the conflict test.
Both compounds that block the γ-aminobutyric acid (GABA) receptor–coupled chloride channel and benzo- diazepine inverse agonists were anxiogenic in the conflict tests, as shown by a decrease in punished responding, which is further evidence of the validity of the paradigm. Central administration of corticotropin-releasing factor (CRF), the peptide that coordinates the endocrine, au-
tonomic, immune, and behavioral responses to stress, suppresses both punished and unpunished responding. Both chlordiazepoxide and ethanol reverse the suppres- sive effects of CRF, as does neuropeptide Y (NPY ) (Britton et al. 2000). CRF1 receptor antagonists produce opposite effects to CRF, which is evidence of their anx- iolytic properties (see, e.g., Arborelius et al. 1999).
The conflict test has several disadvantages. It is time- consuming to perform, because the animals have to un- dergo long training periods and need to be deprived of food or water. Drugs that directly affect hunger or thirst confound the interpretation of the results, and drug treatments also may affect motivation for these rewards. Indeed, animals treated with benzodiazepines increase both their intake of rewards and their rate of responding TABLE 3–2. Effects of chronic treatment with partial 5-HT1A receptor agonists in various animal
models of anxiety
Model Compound Species Effect Reference
Operant conflict Buspirone Rats + Amano et al. 1993
Buspirone Rats + Schefke et al. 1989
Buspirone Rats + Yamashita et al. 1995
Buspirone Mice 0 Martin et al. 1993
Gepirone Rats + Yamashita et al. 1995
Tandospirone Rats + Shimuzu et al. 1987
Elevated plus-maze Buspirone Rats + Söderpalm et al. 1993
Buspirone Rats – Moser 1989
Buspirone Rats – File 1995
Buspirone Mice + Cole and Rodgers 1994
Buspirone Mice 0 Rodgers et al. 1997a
Gepirone Rats + Motta et al. 1992a
Gepirone Rats + Maisonnette et al. 1993a
Gepirone Rats + Silva and Brandão 2000
Ipsapirone Rats 0 Wright et al. 1992
Defensive burying Buspirone Rats + Treit et al. 1981
Perospirone Rats + Sakamoto et al. 1998
Social interaction Buspirone Mice + Cutler 1991
Buspirone Mice + Gao and Cutler 1992
Buspirone Mice 0 Cutler et al. 1997a
Isolation-induced ultrasonic vocalizations Buspirone Rats + Iijima and Chaki 2005
Buspirone Guinea pigs + Kramer et al. 1998
Shock-induced ultrasonic vocalizations Ipsapirone Rats + Baudrie et al. 1993
Note. + =anxiolytic effect; 0=no effect; –=anxiogenic effect; 5-HT1A=serotonin type 1A. aIsolated animal.
for rewards (for an extensive review on criticisms of this model, see Treit 1985).