A review of 25 years of the social interaction test
Section snippets
Background
The social interaction test was developed 25 years ago (File and Hyde, 1978) as the first animal test of anxiety that endeavoured to use ethologically relevant sources of anxiety, and to use a natural form of behaviour as the dependent measure. It, therefore, avoided the use of food or water deprivation and electric shock, and did not require extensive training of the animal. The dependent variable is the time spent by pairs of male rats in social interaction (e.g. sniffing, following or
Behavioural correlates
In the original validation of this test (File and Hyde, 1978), it was shown that the reduction in anxiety that resulted from a brightly lit or unfamiliar arena was not mediated by olfactory cues (anosmic rats showed the same pattern of decreases in social interaction across the four test conditions). It was also shown that reduced social interaction was not the result of the rats spending more time in other activities, such as walking round or exploring the unfamiliar test arena. In fact,
Mice
de Angelis and File (1979) showed that Swiss albino mice showed the same decrease in social interaction as rats when the light level was manipulated, but that the changes in response to the familiarity of the test arena were less reliable. Lister and Hilakivi (1988) found similar results, using male NIH Swiss strain of mice and, furthermore, drugs that have clear anxiolytic and anxiogenic profiles in the rat test did not have specific effects in the mouse test Lister and Hilakivi, 1988,
Corticotropin-releasing factor (CRF)
Because the social interaction test was not developed solely to detect the effects of the benzodiazepines, it is not surprising that it has proved to be sensitive to the effects of a wide range of drugs and neuropeptides. Intracerebroventricular injection of corticotropin-releasing hormone has an anxiogenic effect (Dunn and File, 1987), and this could be reversed by chronic, but not acute, treatment with fluoxetine (To et al., 1999). Combined infusions of CRF and arginine vasopressin into
Benzodiazepine receptor antagonists and inverse agonists
Although agonists at the benzodiazepine receptor have anxiolytic effects (see Section 7), the social interaction test was the first to show that some compounds acting at this receptor had opposite effects. The benzodiazepine receptor antagonist, flumazenil, had an anxiogenic action (File et al., 1982) that could be reversed by subchronic administration of chlordiazepoxide (File and Pellow, 1984). Another benzodiazepine receptor antagonist, ZK 93426, also had an anxiogenic action (File et al.,
Benzodiazepine withdrawal
Whilst the benzodiazepines have clear anxiolytic effects after subchronic treatment (e.g. 5 days), after longer treatment (e.g. 21 days), tolerance develops to this effect (Vellucci and File, 1979), and after even longer treatment (28 days) with a high dose, an anxiogenic effect can be seen (Fernandes et al., 1999). When rats are withdrawn from chronic treatment with benzodiazepines, an anxiogenic response is seen and this can be reversed by administration (in the drug-free state) of the
Benzodiazepines, barbiturates and ethanol
The benzodiazepines are still the most widely used and effective anxiolytic drugs, and many of these have been found to be effective in the social interaction test (see Table 1 for summary).
The social interaction test reflects well the difference between acute and chronic benzodiazepine treatment. With acute administration, the benzodiazepines have their usual sedative effect in all of the test conditions File et al., 1976, File, 1979b, File, 1982. In contrast, after 5 days of pretreatment,
Amygdala
The amygdala has long been considered to be an important structure mediating changes in anxiety and following injection of the specific neurotoxin 5,7-dihydroxytryptamine, resulting in 80% depletion of 5-HT, rats spent significantly less time in social interaction and had reduced locomotor activity (File et al., 1981). The lesion, therefore, seemed to be having a nonspecific effect of decreasing spontaneous behaviours. After direct injection into the amygdala, Higgins et al. (1991) observed
Conclusions
The social interaction test of anxiety has proved extremely useful in detecting both anxiolytic and anxiogenic effects of environmental factors and systemically administered drugs and peptides. It has also proved invaluable in unravelling the neural basis of anxiety. Most of the research over the past 25 years has been on the state of anxiety that is induced by the test itself. This is thought to mimic the state of anxiety most similar to that experienced in generalised anxiety disorder.
References (244)
- et al.
5-HT1A receptors in the median raphe nucleus and dorsal hippocampus may mediate anxiolytic and anxiogenic behaviours respectively
Eur. J. Pharmacol.
(1994) - et al.
Caffeine-induced anxiogenesis: the role of adenosine, benzodiazepine and noradrenergic receptors
Pharmacol. Biochem. Behav.
(1989) - et al.
Antagonistic effects of caffeine and yohimbine in animal tests of anxiety
Eur. J. Pharmacol.
(1989) - et al.
The differential activities of R(+)- and S(−)-zacopride as 5-HT3 receptor antagonists
Pharmacol. Biochem. Behav.
(1990) - et al.
Ketotifen and its analogues reduce aversive responding in the rodent
Pharmacol. Biochem. Behav.
(1990) - et al.
The interaction of R(+)- and S(−)-zacopride with PCPA to modify rodent aversive behaviour
Eur. J. Pharmacol.
(1992) - et al.
The functional significance of biochemical alterations in streptozotocin-induced diabetes
Physiol. Behav.
(1991) Anxiogenic activity of centrally administered scorpion (Mesobuthus tamulus) venom in rats
Toxicon
(1995)- et al.
Anxiolytic–antidepressant activity of Withania somnifera glycowithanolides: an experimental study
Phytomedicine
(2000) - et al.
Effects of chlorisondamine and restraint on cortical [3H]ketanserin binding, 5-HT2A receptor-mediated head shakes, and behaviours in models of anxiety
Neuropharmacology
(1994)