Melatonin was tested in an ongoing attempt to find the endogenous antagonists of quinolinic acid, an endogenous convulsant. Among a great number of metabolites that have been tried before, only a few were found (cerulein and quinaldic acid in mice and kynurenic acid in rats). In SHR (bred from Swiss) male mice, intracerebroventricular (i.c.v.) pretreatment with melatonin (1.25-10.0 microg) attenuated (in the descending order of potency) the convulsant effect of i.c.v. administered kainate, quinolinate, glutamate, N-methyl-D-aspartate, and pentylenetetrazole. Melatonin was ineffective against i.p. administered pentylenetetrazole. Systemically (intraperitoneal, i.p.) administered melatonin (12.5-100.0 mg/kg) attenuated the convulsant effect of quinolinate, while the action of other convulsants used remained unaltered. It is suggested that melatonin could be tried against grand mal seizures in epileptic patients.

Quinolinic acid, a metabolite of tryptophan on the kynurenine pathway, shortened the duration of social contacts (sniffings) in C57BL/6 mice which had been previously isolated for 24 h. This effect was observed at the following time intervals after i.c.v. administration: 2-6, 22-26 and 32-36 min. Locomotion was significantly less inhibited and only during the first interval. L-Kynurenine sulphate was less active. It shortened the duration of contacts only during the 32-36 min interval after i.c.v. administration. Grooming was significantly reduced by quinolinic acid at 7-11, 12-16 and 17-21 min after administration. These effects of quinolinic acid in the social interaction test are similar to those of standard anxiogens and suggest that quinolinic acid belongs to the putative endogenous anxiogens (and not only to the endogenous convulsants). The same assumption about L-kynurenine based on data in other models of anxiety has been made previously.

Daily testing of SHR and C57BI/6 mice during 3 days shows an increase of the number of mice with a high level of anxiety, which was determined according to the ratio of the number of entries into the light arms to the total sum of entries into the light and dark arms of an elevated T-maze. Such results were obtained after 3 testings every 3 days in C57BI/6 but not in SHR mice. This procedure is proposed for the selection of anxious individuals as objects for an anxiety model. Key Words: anxiety; mice; selection; elevated T-mazeIn experiments with putative endogenous anxiogens [3] and anxiolytics [4] in an elevated T-maze it was previously noted that in the control under conditions of repeated tests the number of mice with a high level of anxiety was increased according to the test indexes. We decided to verify this observation with the aim of determining whether repeated tests can be used for the selection of anxious mice for experiments with anxiolytics and anxiogens. MATERIALS AND METHODSOutbred (SHR) and inbred (C57BI/6) male mice aged 6 weeks weighing 19-20 g (Rappolovo nursery) were kept in groups of 8 animals in 20x15• metal boxes prior to the first test in the maze. According to the results of the first test the animals of both groups were divided into three subgroups as follows: high-anxiety (HA), low-anxiety (LA), and moderateanxiety (MA). The criterion for such a classification was the ratio (RAT) of the number of entries into the V. M. Bekhterev Psychoneurological Institute, St. Petersburg (Presented by S. N. Golikov, Member of the Russian Academy of Medical Sciences) light arms to the sum of entries into the light and dark arms. This index is a standard one for estimating levels of anxiety in mice and rats [1,2]. The lower the index, the higher the level of anxiety, and vice versa. Another index, the amount of time spent in the light arms, which is routinely used in tests involving a maze, including our experiments [3,4], was not used in the present study due to the marked variations of data in the control. Initially the two groups of mice differed in the RAT value: outbred -0.43 (the mean value for 60 mice) and inbred -0.34 (the mean value for 56 mice), that is, the inbred mice were on the whole more anxious. Ratios of different magnitudes were accordingly chosen for the separation of individuals within a group. Outbred HA rats had a the RAT lower than or equal to 0.26, LA a value greater than or equal to 0.40, and MA a value between 0.26 and 0.40. For the inbred animals RAT was 0.20, 0.33, and between 0.20 and 0.33, respectively. Then the subgroups were compared to one another within one group but not between groups. Each animal kept its individual tag throughout the tests, enabling it to be identified and its "transition" to be followed from one subgroup to another during the second and third tests. For the one or two days between tests all 28