[1] 11w Dopaminergic denervation reverses behavioral deficits induced by prenatal exposure to phenobarbital

[1] 167w Mice were prenatally exposed to phenobarbital. As adults, these mice (B animals) were deficient in the hippoeampally related eight-arm maze performance, a behavior apparently dependent on the integrity of the septohippocampal cholinergic pathways. Preliminary studies suggest possible parallel alterations in their hippocampal cholinergic innervations. The dopaminergic septal innervations are known to indirectly inhibit the septohippocampal cholinergic innervations. Consequently, the septal dopaminergic innervations of adult B mice were destroyed by 6-hydroxydopamine (6-OHDA). Mice treated with 6-OHDA had an improvement in maze performance which was most marked with increased experience. Concomitant increase in choline acetyltransferase (CHAT) was also demonstrated in these mice (79%, P < 0.001). Similar increase in ChAT could be demonstrated in control mice after 6-OHDA treatment, but the behavioral changes were small and did not reach statistical significance, possibly due to the ceiling effect of the studied behavior. Thus, the dopaminergic innervations in the septum regulate cholinergic activity and its related behaviors along the septohippocampal pathway, and thereby ameliorate behavioral deficits induced by early phenobarbital administration.

[1] 122w The barbiturates are drugs of use and abuse in various situations including pregnancy 6. Animal models for the effect of early phenobarbital administration on the CNS and behavior were established 19 '25, but the mechanism of phenobarbital action is not understood. An inherent methodological problem in these studies is the lack of specificity of the drug which effects numerous brain processes resulting in many behavioral deficits. In the present model on the effect of early phenobarbital administration, a behavior was chosen which is known to be largely related to a certain brain region and specific innervation. Thus, early exposure to phenobarbital induced in mice deficits in eight-arm maze performance 26, a behavior dependent on the integrity of the septohippocampal cholinergic pathway 14.

[2] 132w Preliminary studies suggest possible parallel alterations in the hippocampal muscarinic cholinergic receptors. Furthermore, transplantation of cholinergic but not noradrenergic neurons into the impaired hippocampus of mice, almost completely reversed the early phenobarbital-induced deficits in maze performance 26. Dopaminergic projections from Alo to the septum are known to indirectly inhibit the septohippocampal cholinergic pathways 2,9. Destruction of these dopaminergic pathways enhanced the septohippocampal cholinergic pathways and their related behaviors 9. The present study was designed to gain further insight on the relationships between the hippocampus and its connections, and the early phenobarbitalinduced deficits in eight-arm maze behavior. Mice were exposed to phenobarbital during prenatal development. At adulthood, their Alo dopaminergic projections to the hippocampus were destroyed and the effect of this treatment on hippocampal cholinergic innervation and restoration of normal behavior were investigated.

[1] 33w Treatment with 6-OHDA reduced septal dopamine level by 54% from 16.0 + 3.7 (ng/mg protein + S.E.M.) to 7.4 + 2.7 (n = 6 in each group, P < 0.05 one tailed t-test).

[2] 128w As can be seen in Fig. 1, ChAT levels of animals who were treated with 6-OHDA was greater than in their respective control groups. Thus, the increase in 40. C/DA from C/S level was 57%, and in B/DA, 79% above B/S level (P < 0.001). There were no differences in ChAT level between B and control, and their response to the 6-OHDA injection was to the same extent (no prenatal phenobarbital x 6-OHDA treatment interaction). This made it possible to pool the data across prenatal treatment. Thus, ChAT activity of the pooled sham-injected was 20.0 + 3.3 nM ACh/mg protein/h (+ S.E.M., n = 13) and the ChAT activity of the 6-OHDA injected mice was 33.2 ___ 2.2 (n = 25); an increase of 66% (P < 0.01).

[3] 66w The results of the eight-arm maze test are given in Figs. 234. As can be seen in Fig. 2, after 5 days of testing 86% C/S mice reached criterion in the eight-arm maze, but only 32% of B/S reached criterion (P < 0.01). However, injection of 6-OHDA to B mice (B/DA group) increased their score to 75% (P < 0.01 for the difference from B/S level).

[4] 68w On the first days of testing, B/DA mice had similar scores to B/S which were 5-fold lower than C/S (P < 0.05). Only on the third day they started to show an accelerated rate of improvement and gradually approached C/S level. C/DA group appeared to have higher scores than C/S for this parameter in the eight-arm maze. However, these differences were small and did not reach statistical significance.

[5] 187w The number of entries needed to reach all 8 arms was decreased with trials in all groups (Fig. 3). However, B/S needed up to 37% more entries than C/S in order to reach all 8 arms (second testing day), and differences remained for the rest of the testing days (P < 0.001). The score of B/DA was intermediate between C/S and B/S on the first two days of testing, but it approached C/S level on the remaining testing days (P < 0.01 for the differences from B/S levels). Again, C/DA appeared to have needed fewer entries than C/S in order to reach all 8 arms, but these differences were small and did not reach statistical significance. As is shown in Fig. 4, C/S animals reached criterion after an average of 1.8 days while B/S needed 4.7 days (P < 0.001). However, administration of 6-OHDA to phenobarbital-treated mice (B/DA group) improved their score to 3.0 days (P < 0.01 for the difference from B/S level). C/DA appeared to have needed fewer days than C/S to reach criterion, but this difference was small and did not reach statistical significance.

[1] 59w In the present studies, early exposure of mice to phenobarbital caused long-term deficits in the hippocampus-related eight-arm maze behavior. However, disinhibition of the septohippocampal cholinergic pathways via destruction of the dopaminergic innervations in the septum augmented the impaired eight-arm maze performance of these mice to a near control level. Parallel increases could be shown in the activity of CHAT.

[2] 147w Early (pre-or neonatal) exposure to phenobarbital induced in mice long-term reduction in eightarm maze performance 25, in spontaneous alternations 25 and in water maze 29. The results on the eight-arm maze were reconfirmed in the present study. All these behaviors are thought to be largely related to the hippocampusZS; more specifically, to the septohippocampal cholinergic pathways. It has been demonstrated that specific neurons in the hippocampus are activated according to the specific location of the animals within the maze 13"17. Being hippocampus-dependent does not contradict the fact that other regions in the brain also affect this behavior. For instance, lesions of the cortex 1° or even cerebellum 8 alter the maze performance. However the central role of the hippocampus and its afferents and efferents in this behavior is well established. Animals with hippocampal lesions performed poorly in the eight-arm maze, with no evidence of functional recovery 14.

[3] 85w Several neurotransmitters were implicated in the eight-arm maze performance 4"16"21"23. However, the role of the cholinergic innervations appears most significant. The cholinergic antagonist, scopolamine, reduces 3, while the acetylcholine agonist physostigmine improves, the performance in the eight-arm maze ~z. Furthermore, improvement of performance in eight-arm maze could be achieved by brain tissue transplantation from the cholinergic-rich embryonic septal area into animals with hippocampal lesions 11. In our hands, destruction of the cholinergic innervations with the neurotoxin AF64A caused marked impairment in eight-arm maze performance 28.

[4] 147w Our preliminary finding suggest that prenatal exposure to phenobarbital induced in mice longterm change in the number (Bmax) of the muscarinic receptors 28. The activity of acetylcholinesterase was changed after neonatal but not prenatal exposure and these changes were only transient. An extensive study (n --81) suggested that the activity of ChAT is not affected by prenatal phenobarbital admini-stration28; this lack of effect was confirmed in the present study. Indirect evidence for the role of the septohippocampal cholinergic pathways in the deficits in the eight-arm maze came from our transplantation studies. Thus, transplantation of septal cholinergic neurons but not of noradrenergic neurons reversed the early phenobarbital-induced deficits in eight-arm maze behavior 26. Yet, more components of the hippocampal cholinergic innervation as well as other neurotransmitters should be assayed in relation to the phenobarbital-induced deficits in eightarm maze behavior. This issue is the subject of our current investigations.

[5] 279w The possible involvement of the septohippocampal cholinergic pathways in the phenobarbitalinduced behavioral deficits made it pertinent to investigate its regulating pathways. The septohippocampal cholinergic innervation originate from the septum and the nucleus of the diagonal band where the cell bodies reside. Dopaminergic innervations which prevail in these nuclei originate from the A m area 2 and exert inhibitory influence on the septohippocampal cholinergic pathways. Their innervation with the cholinergic cells is mostly mediated via GABAergic neurons 2°. Previous studies showed that destruction of these dopaminergic innervations enhanced the biochemical activity of the septohippocampal pathways. Furthermore, hippocampus related behaviors, including the performance in the eight-arm maze, were also enhanced by the destruction of the dopaminergic innervation in the septum 7. Therefore, it appeared pertinent in the present experiment to attempt to ameliorate the impaired behavior of the phenobarbital-exposed mice by disinhibiting the septohippocampal pathways via the destruction of the septal dopaminergic innervations. The major effect of the Alo dopaminergic pathway on the septohippocampal cholinergic activity and eight-arm maze performance is now well established. The evidence was supported by other manipulations of the septal dopaminergic innervations in addition to the 6-OHDA treatment 2°. Yet, further studies should look into the possible effect of the 6-OHDA treatment on other neurotransmitters, mainly norepinephrine. Possibly, such effects may also have some influence on the behavior under study. Another consideration is that 6-OHDA may induce neurostructural alterations which also play a role in the behavioral changes. The effect of 6-OHDA treatment on the behavior was not related to its possible effect on general motor activity, since the performance in the eight-arm maze test, as conducted in the present study, is independent of speed or activity.

[6] 170w Both transplantation 26 and 6-OHDA treatment (present study) improved the eight-arm maze behavior of animals exposed in utero to phenobarbital. Yet a basic difference exists between the two treatments. Transplantation repaired the phenobarbital-induced damage where it occurred and thus reversed the behavioral deficits. In fact, transplantation to normal animals never improve their behavior 26. On the other hand 6-OHDA improve eight-arm maze performance of all animals, normal or impaired. Thus, the difference between transplantation anti 6-OH-DA treatments is the difference between specific and non-specific treatments. In the present study, the augmentation of performance by 6-OHDA among the control animals was small and did not reach statistical significance. The likely explanation is that control HS/Ibg mice normally have a near maximum score in the eight-arm maze where further improvement is no longer possible (ceiling effect). Either 'raising the ceiling' by using a more difficult test (e.g. 16-arm maze) or using a mouse strain with a lower score in the maze would enable the detection of a 6-OHDA effect in normal animals.

[7] 73w Another basic difference between the outcome of the two treatments is that transplantation improved the behavior of the treated animals almost from the beginning of the testing period so that the scores of the transplanted barbiturate animals always paralleled control level 26. On the other hand, 6-OHDA improved the behavior substantially only after 3 testing days. It appears, therefore, that 6-OHDA treatment enhances the ability of the impaired animal to improve with experience.

[8] 72w Previous studies demonstrated the disinhibition by 6-OHDA or other treatments of the septohippocampal cholinergic pathways on both behavioral and neurochemical levels 7. Acetylcholine turnover rate 2° and high affinity choline uptake 7 served as indicators for the disinhibition of the neurochemical processes. The present study demonstrates 6-OHDA treatment-induced rise in ChAT activity as an additional evidence for the disinhibition of septohippocampai cholinergic activity following destruction of the A m septai dopaminergic pathways.

[9] 84w Although further biochemical studies are still required, the present findings, taken together with the biochemical findings and the results of the transplantation studies, all seem to suggest that phenobarbital given during early development exerts its effect on eight-arm maze performance mainly by altering the septohippocampal cholinergic pathways. Consequently, manipulating this system enables the correction of much of the early phenobarbitalinduced behavioral deficits. Thus, the findings offer a simple model of possible utility for the understanding of certain behavioral birth defects and their related mechanisms.

[1] 32w Pregnant mice were exposed to phenobarbital on gestation days 9-18. The offspring (B-mice) and their respective controls were injected at age 50 days with 6-hydroxydopamine (6-OHDA) or were shaminjected with vehicle solution.

[2] 6w Thus, there were 4 experimental groups:

[3] 47w Control animals (i.e. mice who were not exposed to phenobarbital prenatally) and were Sham-injected later with vehicle solution; B/S-B-animals (i.e. mice who received barbiturate prenatally) and were Sham-injected later with vehicle solution; C/DA-Control animals who were injected later with 6-OHDA; B/DA-B-animals who were injected later with 6-OHDA.

[4] 90w Septal dopamine levels were determined in a sample group in order to assess the extent of 6-OHDA-induced destruction of the dopaminergic innervations. At age 54 days all mice began a schedule of water deprivation and a week later they were tested for their performance in the eight-arm maze. Three days after testing, their brains were removed for a choline acetyltransferase (CHAT) assay. Female and male offspring were employed in approximately equal numbers. Sample sizes are given in the figures. Data were analyzed using Analysis of Variance, t-test and X2-test 22'24.

[5] 35w [3H]Acetyl coenzyme A (33/~Ci/ml) was obtained from Amersham (U.K.). Acetyl coenzyme A (Co-A)-lithium salt, choline bromide, Brilliant blue G, 6-OHDA, and ascorbic acid were all obtained from Sigma. All salts and buffers were Analar grade.

[6] 41w Heterogeneous stock (HS/Ibg) mice 12 were used. Four females and one male were housed in each mating cage and maintained under standard labora-tory conditions of 24 °C and 12-12 h light-dark cycle. Their offspring were the subjects of the present experiment.

[7] 136w Female parent mice were checked daily at 08.00 h for insemination (as evidenced by a vaginal plug). Females with plugs (considered as gestation day 1 = GD-1) were separated from the males and housed with other pregnant females. On GD-9 the females were placed in individual cages. Treated females then received milled mouse food containing 3 g/kg phentobarbital in acid form (their only food source) and water, both available ad libitum. This method of administration resulted in blood and brain phenobarbital levels of about 100/tg/m125. Control females received milled food and water. Drug administration continued until GD-18, when the phenobarbital and control diets were replaced by standard mouse food. The treated and control offspring were weaned at 25 days of age. Males and females were separated at age 30 days and housed in groups of four.

[8] 256w The test procedure in the radial eight-arm maze is described in detail elsewhere26, Mice were placed on a regimen of water deprivation that consisted of the administration of water for 30 min once a day. After a week, at age 50 days, the mice were introduced individually into the maze for 10 min of habituation without reinforcement of water. The first 16 entries were recorded. In the following 5 days of the test, the mice received reinforcement with water drops of 50 /A. Unlike the habituation day (Day 1), the animals were left in the maze until they had either entered all the 8 arms or until they had made 16 entries, whichever occurred first. Animals were considered to have reached criterion if there were 8 correct entries out of the first 8 entries for two consecutive days. Thus, it became possible to determine (1), the number of trials needed to enter all arms through 16 trials and (2), the number of days it took to reach criterion. Animals that reached criterion on day 1 and on day 2 (8 correct responses out of 8 trials in the pretest-Day 1 plus day 2) received the score 0. Animals that reached criterion after day 1 received the score of 1, etc. Animals that achieved 8 correct responses out of 8 entries only on the final day, day 6, received the score 5. Animals that never reached criterion and did not even succeed in making 8 correct responses in 8 entries on Day 6 received the score 6.

[9] 121w The assay was conducted to confirm the denervation made by 6-OHDA. The method was described before 18'27. Briefly, septal tissues were sonicated in approximately 20 vols. of ice-cold 0.1 M perchloric acid containing 2 mM Na2EDTA and 10 /zl/ml of NaHSO3 (0.1 M). After centrifugation (15,600 g for 5 min), 180 pl of supernatant was taken for analysis. Dopamine was isolated by the alumina extraction technique. Extracts were subjected to reverse-phase (C18 ODS column, DuPont) highperformance liquid chromatography and were eluted with 0.1 M monochloroacetate buffer (pH 3.05) containing 1 mM Na2EDTA and 70 mg/1 of sodium octyl-sulfate as an ion-pairing agent. The catechois were quantitated using an LC-4A glassy carbon electrode (Bioanalytical Systems) with an applied potential of 0.7 V.

[1] 75w At age 50 days control and B offspring were divided into 6-OHDA-treated and sham-injected groups. They were anesthetized with 80 mg/kg pentobarbital, placed in a Kopf stereotaxic apparatus and injected bilaterally into the septum with 6-OHDA (1 #g 6-OHDA dissolved in 1 pt of 0.2% ascorbic acid solution to each side). The coordinates were 0.7 mm anterior to the bregma, Lo = _+ 0.3 and DV = 3.3. Sham-injected groups received vehicle (ascorbic acid) solution.

[2] 59w Brains were rapidly dissected on ice and the hippocampi were removed and kept as pairs from each brain in -80 °C freezer until use. Before the biochemical assay, the tissues were homogenized in 20 vols. of homogenization buffer (10 mM EDTA disodium, 20 mM phosphate buffer-sodium, 0.5 Triton X-100, pH 7.4) for 10 s using an IKA-WERK Ultra Turax.

[3] 189w The assay performed is according to the method of Fonnum 5. Reaction mixtures in a 5-ml scintillation vial were composed of 10/A of homogenate and 20 /d reaction buffer [CHAT buffer (24 mM choline bromide, 60 mM EDTA-disodium, 150 mM disodium buffer, 0.3 B.S.A.) 50, Eserine (physostigmine, 2.5 mg in 12 ml D.D.W.) 30, and [3H]acetyl-CoA (2 mM, 33/~Ci/ml) 20]. Triplicates were taken of both reaction and protein. Reaction was performed for 20 min in a 37 °C shaking water bath. The reaction was terminated by adding 1 ml of 10 mM sodium phosphate buffer. A biphasic scintillation 'cocktail' consisting of 167 mg POPOP, 416 mg PPO and 833 ml Toluene mixed with 166 ml acetonitrile and 833 mg 'calignost' (sodium tetraphenyl boron) was used to separate the [3H]acetylcholine produced from the [3H]acetyl-CoA substrate. 33 ml of cocktail were added to the reaction vial and the samples were then counted in a Packard Tri Carb Liquid Scintillation Spectrometer of 60% efficiency for tritium. Total count was measured using LUMAX as the scintillation fluid. Results were expressed in nM ACh/mg protein/h. Protein was determined by the procedure of Bradford 1.