Neuroscience Vol. 71, No. 2, pp. 543 554, 1996
Elsevier ScienceLtd Copyright © 1996 IBRO Printed in Great Britain. All rights reserved 0306-4522/'96 $15.00 + 0.00
OXYTOCIN A N D VASOPRESSIN m R N A EXPRESSION IN RAT HYPOTHALAMUS FOLLOWING KAINIC A C I D - I N D U C E D SEIZURES Q. S U N , * t S. P R E T E L , t C. D. A P P L E G A T E ~ and D. T. P I E K U T t tDepartment of Neurobiology and Anatomy, and ;~Department of Neurology, University of Rochester, Rochester, NY 14642, U.S.A. Abstract--In this study, the regulation of hypothalamic oxytocin and vasopressin messenger RNA expression following the induction of seizures was investigated by in situ hybridization. Following kainic acid-induced seizures, a significant increase in oxytocin messenger RNA in the paraventricular nucleus was demonstrated at 1.5 h, one and two weeks; its level decreased at three weeks and was significantly increased again at four weeks; at eight weeks the messenger RNA level still remained higher than that of controls. Vasopressin messenger RNA in the paraventricular nucleus was increased significantly only at 1.5 h following induction of seizures. The oxytocin messenger RNA level in the supraoptic nucleus was also increased early at 1.5 h and later at four weeks following seizures; however, these increases did not last as long as those in the paraventricular nucleus. Vasopressin messenger RNA in the supraoptic nucleus was also increased after the initial seizures; however, its messenger RNA level vacillated up and down throughout the post-seizure times studied. The earliest significant increase of vasopressin messenger RNA was at one week after seizures, and there was a late significant increase of vasopressin messenger RNA at three weeks after seizures. The present study demonstrates that following kainic acid-induced seizures both, the oxytocin and vasopressin messenger RNA expressions, were up-regulated and these up-regulations were long-term events. The increase of oxytocin messenger RNA in the paraventricular nucleus was more persistent than the others. The pattern of messenger RNA up-regulation was different for oxytocin and vasopressin, and different in the paraventricular nucleus and supraoptic nucleus. These different patterns of messenger RNA elevations suggest that the different components of the rat hypothalamus were regulated differentially by kainic acid-induced seizures. Key words: paraventricular nucleus, supraoptic nucleus, peptides, excitatory amino acid, epilepsy, in situ
Organisms often encounter stressors, which will disrupt the body's homeostasis; the body tends to regain and maintain its homeostasis through stressresponses. Stress-response systems include the hypothalamus-pituitary-adrenal (HPA) axis and the autonomic nervous system. These two systems interact with each other at multiple levels during stress-responses. The hypothalamus is a complex integration and co-ordination center for stressresponses and activation of specific neuronal populations will increase the adrenocorticotrophic hormone ( A C T H ) secretion from the pituitary. In turn, the A C T H release will cause the secretion of cortisol (corticosterone for rodent) from the adrenal cortex, *To whom correspondence should be addressed. ACTH, adrenocorticotrophic hormone; ATP, adenosine triphosphate; HPA, hypothalamuspituitary-adrenal; i.p., intra-peritoneal; KA, kainic acid; N-methyl-o-aspartate; PBS, phosphate-buffered saline; PVN, paraventricular nucleus; SON, supraoptic nucleus; SSC, standard sodium citrate; TdT, terminal deoxynucleotidyl transferase.
which will further modulate the functions of its target organs. The release of A C T H is traditionally associated with those hypothalamic neurons that synthesize and release corticotrophin-releasing factor; however, the release of A C T H can also be induced by the secretion of oxytocin and vasopressin. 29'44 Oxytocin receptors have been demonstrated in the anterior lobe of the pituitary 2 and two pathways by which oxytocin and vasopressin may modulate corticotrophin cells in the anterior lobe of the pituitary have been hypothesized. 44 One is via the long portal veins, through which oxytocin and vasopressin, released from parvocellular neurons into the median eminence, are carried to the anterior lobe. The other is via the short portal veins, which transport the oxytocin and vasopressin, released by magnocellular neurons at the posterior lobe, to the anterior lobe. Oxytocin and vasopressin can either stimulate A C T H secretion directly 36 or potentiate corticotrophin-releasing factor induced A C T H release. 18,54 Oxytocin and vasopressin also may contribute in maintaining basal
Q. Sun et al.
secretion of A C T H since the immunoneutralization of oxytocin and vasopressin by their antisera decreased the basal A C T H level. 16 Recently, the synthesis and release of both peptides have been implicated in the physiological response to different stressors, such as osmotic stimulation, 34'41,59 acute or chronic restraint stress, 3'9A7"21'23ether exposure ~2'17and forced swimming. 33 The occurrence of seizures is also regarded as a severe stressor since many changes, which occur during and after the elicitation of seizures, such as intensive recurrent neuronal discharges, neurochemical changes and progressive neuronal damages, will cause severe disruption of the homeostasis. Consequently, previous evidence has indicated that the H P A axis is activated by seizures. A clinical study of idiopathic or post-traumatic epilepsy and alcoholwithdrawal seizures has shown increased plasma A C T H and vasopressin concentration increased after the seizure activity, which was followed by a subsequent cortisol increase. ~ A recent case report of a prolonged temporal lobe epileptic seizure revealed that both vasopressin and oxytocin plasma concentrations were elevated during the generalized phase of seizure. 38 The modification of oxytocin and vasopressin synthesis and release t h a t has been observed during and following stressful conditions clearly demonstrates a wider role for these peptides in the stressresponsiveness than had been assumed previously. We therefore were interested in examining whether the synthesis and, implicitly, the activity of the hypothalamic oxytocin and vasopressin neurons might be altered by the induction of seizures, and if potential alterations of their activity would be transient or long-lasting. To answer these questions, quantitative in situ hybridization histochemistry for oxytocin and vasopressin m R N A was applied to examine the responses of the rat hypothalamus, specifically the paraventricular nucleus (PVN) and supraoptic nucleus (SON), at different post-seizure survival times following kainic acid (KA)-induced seizures. EXPERIMENTAL PROCEDURES
Animal preparation Male Sprague-Dawley rats (Charles River) were treated to induce fully generalized seizures by i.p. injections of KA (Sigma) at a dosage of 17 mg/kg in 1 ml phosphate-buffered saline (PBS), pH 7.2. Rats were injected between 8.30 and 9.30 a.m. under light ether anesthesia and the progress of seizure behavior was closely observed. Only animals that reached stage 5 (fully generalized) seizures 43 were used for the subsequent study. Controls were prepared for each of
these experimental rats by i.p. injection of 1 ml PBS, also under light ether anesthesia. The pairs of rats, consisting of one experimental and one control animal, were always treated under the same conditions in the rest of the study. They were kept for different post-seizure survival times: 1.5 h, one, two, three, four, six and eight weeks. Three to six pairs of rats were prepared for each time point. Before and after treatment, animals had free access to water and food. Animals were killed by injecting an overdose of sodium pentobarbital (i.p. 1 ml per animal) and perfused through an intra-aortic cannula. Prior to perfusion with PBS, rats were injected with 0.5 ml 1% sodium nitrate and 0.3 ml heparin (about 1500unit) into the left cardiac ventricle. Then, rats were perfused with cold PBS, followed by 4% paraformaldehyde in PBS. For cryoprotection, 15% sucrose in PBS was perfused following the fixation. The brains were subsequently kept in 30% sucrose in PBS overnight at 4°C. Brains were quickly frozen with liquid nitrogen and stored in -20°C freezer for a few days until cutting. Frozen sections of rat brains were cut on a cryostat at 14#m thickness. Sections through the rostrocaudal extent of the rat PVN and SON were thaw-mounted on poly-L-lysine-coated slides 26 and stored at - 2 0 ° C until hybridization. All experiments were conducted in agreement with the Animal Welfare Act and Public Health Service Policy. In situ hybridization The probes for oxytocin and vasopressin mRNA were 48-mer oligonucleotides (generously provided by Dr S. Young III, NIH). The oxytocin probe was complementary to bases of 247-294 of the oxytocin gene; the vasopressin probe was complementary to bases of the vasopressin gene, coding for the last 16 amino acids of a glycoprotein, which had no homology with the oxytocin gene. 27 Both probes were Y-end labeled with terminal deoxynucleotidyl transferase (TdT, Gibco) and [35S]deoxy-ATP (> 1000Ci/mM, NEN) to a specific activity of 1-2 x 10 -6 Ci/mM. The protocol for in situ hybridization has been described in detail previouslyfl In brief, brain sections were pre-treated with 0.25% acetic anhydride in 0.1 M triethanolamine for 10 min, dehydrated in a series of graded ethanol and delipidated with chloroform. Hybridization solution (1 2 #1 probe per 100/d hybridization buffer per two rat brain sections) was applied to the tissue sections. Brain sections were hybridized overnight in a humidified chamber at 37°C. Sections were then washed with 1 × standard sodium citrate (SSC), 4 × 15min at 55°C and l h at room temperature, followed by 0.1 × SSC for 1 h at room temperature. Slides were dipped briefly in deionized water and air-dried. Hybridized tissue sections were exposed to autoradiographic film (Hyperfilm-fl max tm, Amersham) for three to five days at 4°C with sections from the paired experimental and control rats on the same film. Slide autoradiography was obtained by dipping the hybridized slides in NBT2 emulsion (Kodak) and exposed them for five to seven days at 4°C. Films and slides were developed with D-19 (Kodak) and fixed with Kodak Fixer. Image analysis The presence of potential quantitative differences of oxytocin and vasopressin mRNA in PVN and SON between the experimental and control rats was examined by densitometry analysis of the film images. The amount of mRNA for oxytocin or vasopressin in the tissue sections was reflected in the
Abbreviations used in the figures and tables
CON KA KA~20N
control animal kainic acid treated animal difference between kainic acid treated and control animals
PVN~)X PVN-VAS SON~)X SON VAS
oxytocin in vasopressin oxytocin in vasopressin
paraventricular nucleus in paraventricular nucleus supraoptic nucleus in supraoptic nucleus
Regulation of hypothalamic peptides following seizures grey levels of the film images. For this study, the mean density of film images with labeling of oxytocin or vasopressin mRNA in PVN and SON were analysed with the NIH IMAGE program. In brief, film images of the hypothalamus were captured with a video camera (Cohu, Inc.) connected to the computer through a frame grabber (Perceptics, Inc.). The positive area of oxytocin mRNA or vasopressin mRNA in PVN or SON was selected and the mean density (i.e. unit density per pixel area) of the PVN and SON, as well as the mean density for the negative background, was measured. Background areas were selected from negative tissue sites around each of the measured positive areas. The mean density value for the background area was subtracted from the mean density of the positive area, and the resulting value was used to calculate a mean density value of oxytocin or
vasopressin mRNA in PVN or SON. The relative changes of the amount of oxytocin or vasopressin mRNA in PVN or SON, caused by KA-induced seizures, were shown by the differences of mean densities between the'paired experimental and control rats. Statistics
The results of the measurements were expressed in the differences between paired experimental and control animals. Differences between experiments and controls for the whole study and for each post-seizure time point were analysed by two-tailed t-test. Variances between the time points were analysed with ANOVA for repeated measures. The significant level was P < 0.05. I f the ANOVA analysis was significant, data would be further analysed by the Bonferroni/Dunn test.
Fig. 1. Oxytocin and vasopressin mRNA containing PVN and SON neurons are presented in bright field photomicrographs of control rats. A is an image of oxytocin mRNA expressing neurons from the medial PVN and these neurons are concentrated in the magnocellular division (10 x objective). The insert is a high power image of the neuron indicated in A, which shows the autoradiographic silver grains in neuronal cytoplasm (100x objective). B is an image of vasopressin mRNA expressing neurons in the medial PVN (10 x objective). Its AP level is slightly posterior to that of image A and the vasopressin neurons form a round shaped core structure in the magnocellular division, which is dorsolateral to the main structures for oxytocin neurons. C is an image of oxytocin mRNA expressing neurons in SON, which are concentrated at the dorsal SON (10 x objective). D is an image of VAS mRNA expressing neurons in SON (10 x objective). They occupied a larger area than oxytocin mRNA expressing neurons and are concentrated in the ventral part of the SON.
Q. Sun et al. RESULTS
Following a single injection of K A (17 mg/kg, i.p.), more than 90% of the treated animals showed a progressive development of seizure behavior from low stages to high stages. More than half of them reached stage 5 seizures in 1.5-2 h following the K A injection. One-fourth of the treated animals reached stage 5 seizures with an additional K A injection of 25-50% of the original dosage. Ten percent of the
treated animals, which had less severe seizures and failed to reach stage 5 seizures, were not included in the study. More than 90% of these stage 5 seizure animals survived the whole period of the experiment. Following the first stage 5 seizure, the animals usually experienced several stage 5 or stage 4 seizures and/ or a period of status epilepticus, but calmed down within 4 - 6 h. A l m o s t all the animals regained their normal activities on the next day. Occasionally seizure behavior was observed, but convulsive seizures were
Fig. 2. A series of photomicrographs showing autoradiographic film images of oxytocin mRNA labeled structures in the rat hypothalamus (2 x objective). Images from A to D are arranged from anterior to posterior.
Regulation of hypothalamic peptides following seizures rare. Their eating and drinking behavior appeared normal, and their physical condition healthy. However, animals, which were kept longer than one week of post-seizure survival time, appeared to be more nervous; i.e. they were more sensitive to touch and tended to over-react and jump when handled. Distributions of neurons containing the m R N A for oxytocin in the PVN of control rats are shown in Figs IA and 2. The cytoplasm of the m R N A positive neurons was covered by a dense accumulation of autoradiographic silver grains (Fig. 1A, insert). The majority of these neurons were located in the magnocellular division throughout most of the rostrocaudal extent of the PVN. In the anterior and medial magnocellular division of the PVN, oxytocin m R N A labeled neurons occupied two relatively small areas at either sides of the third ventricle. In the posterior magnocellular division, oxytocin m R N A positive neurons formed a large cluster located slightly anteromedial to the core structure of vasopressin neurons (Fig. 1A, B). Parvocellular neurons of the dorsal cap and ventromedial areas of the PVN were also labeled and included in the analysis. On the autoradiographic film, positively labeled magnocellular and parvocellular neurons formed a dark gray image usually with its center slightly denser than its circumference (Fig. 2). PVN oxytocin gene expression was examined by quantitative densitometry of the oxytocin m R N A level in the PVN of each animal. Density values for each animal were expressed as mean density of the whole PVN. For this study, mean density of PVN oxytocin m R N A was calculated from an average of 15.9 + 6.9 tissue sections per animal. For control animals, the variation between different time points was not significant (ANOVA, P > 0.05). To identify the general trend of changes caused by KA-induced seizures, density differences between experimental and control rats for the whole set of experiments, including all post-seizure time groups, were analysed. The density of PVN oxytocin m R N A for K A rats was 98.84 + 2.00 (unit density per pixel area) and that for controls was 90.7 + 72.69 (unit density per pixel area; Table 1). The results showed that there was a significant increase of oxytocin m R N A following the KA-induced seizures (two-tailed paired t-test, P < 0.001; Table 1). The time course of changes in oxytocin gene expression following KA-induced seizures is summarized in Fig. 3A and Table 3. At all
post-seizure time points studied, the levels of oxytocin m R N A in the PVN were higher in experimental animals than those in controls. Although the net increases of mean densities were different from time point to time point (ranged from 0.72 to 15.48), these variances between the groups were not significant (ANOVA, P > 0.05). Shortly (1.5h) following the first stage 5 seizure, there was a significant increase of PVN oxytocin m R N A expression compared with control m R N A expression (P < 0.05). The elevation of PVN oxytocin m R N A levels remained significantly high at the following time points: one week and two weeks, but returned to control level at the three week point. At four weeks after the initial stage 5 seizures, the PVN oxytocin gene expression again increased significantly over control values and formed a second peak (P <0.01), with an even larger increase in density than the first peak at 1.5 h (Fig. 3A). At later time points (six and eight weeks), the oxytocin gene expression remained elevated but the differences of PVN oxytocin m R N A were not statistically significant any more (P > 0.05). Neurons expressing the m R N A for vasopressin shared the same morphological appearance with those expressing oxytocin m R N A (Fig. 1). Their main distribution in the PVN was concentrated more posteriorly in the magnocellular PVN, and formed a round shaped core structure at the dorsolateral portion of the hypothalamus (Fig. 1B). Vasopressin synthesizing neurons also included some parvocellular neurons of the dorsal cap and ventromedial aspect of the PVN. The quantitative analysis of the vasopressin m R N A expression at the PVN was achieved by using an average of 5.8 _+ 3.3 sections per animal and the results are summarized in Tables 1 and 3, and Fig. 3B. The analysis of the density differences for the whole set of experiments revealed a significantly higher vasopressin m R N A level in K A seizure rats (98.42 _+ 1.85 unit density per pixel area) than that of control rats (92.95 _+ 1.86 unit density per pixel area; P < 0.001). At 1.5 h following the first stage 5 seizure, the vasopressin m R N A level was significantly higher than control level (P < 0.05). However, this elevation of vasopressin m R N A was short lived, since the vasopressin m R N A levels at one, two and three weeks were only slightly higher than control levels (P > 0.05). At the four week point, however, there was a large, but not significant, increase of vasopressin m R N A over control (P > 0.05). At the six and eight week points,
Table 1. Densities of oxytocin and vasopressin mRNA labeling in KA and control (CON) rats, as well as the differences of density between KA rats and their paired controls (KA-CON) in PVN and SON are shown in unit density values (per pixel area) PVN~)X PVN-VAS SON4~)X SON-VAS
98.94 _+2.00 98.42 + 1.85 104.66 ± 1.78 111.29 _+ 1.65
90.77 _+2.69 92.95 + 1.86 98.11 + 2.36 106.18 _+ 1.88
KA-CON 8.14 +_ 1.36"** 5.22 4- 1.34"** 5.72 _+ 1.66"* 5.32 _+ 1.77'*
Values are presented as mean + standard error. *P < 0.05; **P < 0.01; ***P < 0.001 determined with the two-tailed paired t-test.
Q. Sun et al.
221 28 ~
4 ,Ira 2
I w 2w 3w 4w 6w P o s t - s e i z u r e suruival times
20 86 *"
I w 2w 3w 4w 6w P o s t - s e i z u r e survival times
s°N°x 22e 468 I (c
lw 2w 3w 4w 6w P o s t - s e i z u r e survival times
I w 2w 3w 4w 6w P o s t - s e i z u r e survlual times
Fig. 3. Graphs show the density differences of mRNA labeling between KA rats and their paired controls at different post-seizure times in the PVN (A and B) and SON (C and D) for oxytocin (OX, A and C) and vasopressin (VAS, B and D) mRNA expressing neurons. The bars at different time points represent mean + standard error. The dashed line at zero level indicates control values. The statistical results were obtained with the two-tailed paired t-test and the significance levels are indicated with *(P < 0.05), **(P < 0.01) or ***(P < 0.001).
Regulation of hypothalamic peptides following seizures Table 2. Positively, i.e. oxytocin and vasopressin mRNA, labeled areas in PVN and SON for KA and control rats, as well as the differences in positively labeled areas between KA rats and their paired controls are shown in unit pixels PVN~)X PVN-VAS SON43X SON-VAS
571.19 + 36.71 392.86 _+30.25 472.31 ___29.75 488.02 ___22.58
483.74 + 31.34 339.53 _ 23.68 409.85 + 25.14 437.87 __+21.87
KA-CON 81.51 ___19.07"** 48.73 __+19.94" 51.60 __ 19.63" 51.27 ___24.76*
Values are presented as mean __.standard error. *P < 0.05; **P < 0.01; ***P < 0.001 by two-tailed paired t-test. the level of vasopressin m R N A decreased, but never to control level. The longitudinal shaped SON is located in the ventral portion of the hypothalamus, laterally adjacent to the optic tract, and its rostrocaudal extension is longer than that of the PVN. In the SON, oxytocin neurons are concentrated in its dorsal portion (Fig. 1C). Quantitative analysis of the density of the oxytocin m R N A in the SON for the whole set of experiments is shown in Table 1. The densities of SON oxytocin m R N A labeling for K A rats and controls were calculated from an average of 17.2 + 7.5 sections per animal and their values were 104.6 _ 1.78 (unit density per pixel area) and 98.11 + 2.36 (unit density per pixel area), respectively. This difference in density between K A and control rats was highly significant (P <0.01). The changes of SON oxytocin gene expression at different post-seizure times are shown in Fig. 3C and Table 3. At 1.5 h following seizure, the density of SON oxytocin m R N A for K A rats increased significantly compared with that of control rats (P < 0.01). However, this increase in oxytocin m R N A levels did not last long and became not significant at the following time points: one, two a n d three weeks (P > 0.05). At the four week time point again, the SON oxytocin m R N A level was significantly elevated (P < 0.01) and formed a second peak. At the next time point, six weeks, the difference of SON oxytocin m R N A levels between K A and control rats was similar to that for four weeks, but it was not significant (P > 0.05). At eight weeks, the SON oxytocin m R N A level for K A rats dropped below control level. Variances of the SON oxytocin m R N A levels caused by K A seizures between the different post-seizure time points were analysed with A N O V A for repeated measures and appeared significant
(ANOVA, P < 0.05); specifically, the analysis with the Bonferroni/Dunn test showed that the differences of SON oxytocin m R N A values between 1.5 h and eight weeks were significant (P < 0.001). Neurons of the SON containing the vasopressin m R N A were distributed more evenly than oxytocin m R N A expressing neurons (Fig. 1D), but they extended rostrocaudally approximately the same distance as oxytocin synthesizing neurons in the SON. These neurons were concentrated in the ventral portion of SON. Quantitative analysis of the SON vasopressin m R N A values was achieved by using an average of 15.0 ___6.9 sections per animal and the results are shown in Tables 1 and 3, and Fig. 3D. A general comparison of the density differences of vasopressin m R N A between the K A rats (111.29 ___1.64 unit density per pixel area) and control rats (106.18 + 1.88 unit density per pixel area) for the whole set of experiments demonstrated a highly significant increase of the density in SON vasopressin m R N A labeling in experimental rats (P <0.01, Table I). However, the pattern of the SON vasopressin m R N A expression changes was different from that for vasopressin or oxytocin m R N A in the PVN or oxytocin m R N A in SON (Fig. 3). The vasopressin m R N A levels in the SON were not maintained at a certain level, but vacillated up and down. There was no significant early increase in vasopressin m R N A in the SON, i.e. 1.5 h following the initial seizures; instead the m R N A value increased to a significant level only one week following the seizure. At the next time point (two weeks), the SON vasopressin m R N A returned to control level; at three weeks, its level increased significantly (P < 0.05), but decreased again at four weeks (P > 0.05). Although not significant, the SON vasopressin m R N A level was obviously
Table 3. Density differences of oxytocin and vasopressin mRNA labeling between KA rats and their paired controls at different post-seizure times in PVN and SON are shown in.unit density (per pixel area) 1.5 h 1 week 2 weeks 3 weeks 4 weeks 6 weeks 8 weeks
9.73 + 3.49 (6)* 9.37 __+3.72 (6)* 6.82 + 2.57 (6)* 0.72 _+3.35 (6) 15.48 _+2.17 (5)** 9.70 ___3.99 (4) 6.30 + 4.05 (4)
9.10 + 2.11 (5)* 4.76 _+ 3.58 (4) 4.60 __+4.26 (5) 1.87 __+3.57 (5) 9.20 _+2.90 (4) 2.81 __+6.61 (3) 3.07 _+0.82 (3)
SON~OX 13.95 __+3.14 5.05 +_4.40 3.19 + 3.00 4.39 ___4.24 9.46_ 3.32 9.18 +4.39 -7.39 _ 6.46
(6)** (6) (6) (6) (5)** (4) (4)
SON-VAS 7.03 __+3.14 (6) 10.42 __+3.46 (6)* -0.83 _+4.66 (6) 6.40 + 2.26 (5)* 3.54 ___7.70 (5) 13.34 ___5.12 (4) - 1.21 + 4.33 (5)
Values are presented as mean + standard error. Numbers in parentheses indicate numbers of the experimental and control animals. *P < 0.05; **P < 0.01; ***P < 0.001 by two-tailed paired t-test. NSC 7 1 2 - - I
Q. Sun et al.
Table 4. Differences in oxytocin and vasopressin mRNA labeled areas between KA rats and their paired controls at different post-seizure times in PVN and SON are shown in unit pixels PVN-OX
h5 h 1 week 2 weeks 3 weeks 4 weeks 6 weeks 8 weeks
44.87 ___49.12 (6) -8.85 _ 30.15 (6) 107.74 -t- 57.44 (6) 103.69 _ 24.46 (6)** 89.74 _+26.40 (5)* 91.36 + 49.75 (4) 179.24 + 100.54 (4)
46.43 + 51.65 (5) 9.88 _ 6.03 (4) 65.54 + 53.04 (5) 42.67 +__53.94 (5) 17.34 +_79.83 (4) 13.9 _+ 17.47 (3) 163.09 + 60.38 (3)
- 12.38 + 30.73 (6) 21.81 __+26.13 (6) 95.11 _ 84.73 (6) 91.99 + 53.39 (6) 44.29 _ 20.03 (5) 95.63 + 9.52 (4)** 31.51 + 88.98 (4)
46.77 + 47.87 (6) 70.64 __+51.61 (6) -43.32 __+90.02 (6) 113.29 + 77.15 (5) 110.29 + 71.24 (5) 92.29 + 78.90 (4) -6.96 + 25.70 (4)
Values are presented as mean + standard error. Numbers in parentheses indicate numbers of the experimental and control animals. *P < 0.05; **P < 0.01; ***P < 0.001 by two-tailed paired t-test.
increased over control levels at the six week point and returned to control level at the eight week point (Fig. 3). A quantitative analysis of the oxytocin and vasopressin m R N A labeled areas in PVN and SON of the rat hypothalamus is shown in Tables 2 and 4. In general, these positive areas were significantly larger in K A rats than those for control rats (Tables 2, P < 0.05). At individual post-seizure times, however, only a few of them reached the significant level (three and four week points for PVN oxytocin, and the six week point for SON oxytocin). Microscopic analysis of the autoradiographic slides did not reveal any obvious changes of the oxytocin and vasopressin m R N A positive neuronal population following KA-induced seizures.
The major finding of the densitometry analysis demonstrates an extensive increase in the hypothalamic oxytocin and vasopressin m R N A expressions following KA-induced seizures. This increase in m R N A levels is present in every component of the magnocellular PVN, as well as in the small or medium sized neurons of the dorsal cap and ventromedial aspect of the PVN. The data also demonstrate a differential regulation of oxytocin and vasopressin in the PVN and SON. Increases of oxytocin m R N A in PVN and SON were demonstrated as early as 1.5 h following the occurrence of fully generalized seizures and were continuously observed up to six weeks in SON and eight weeks in PVN following seizures. Statistically, two significant peak increases of oxytocin m R N A occurred in PVN and SON; the first peak at 1.5 h and the second at four weeks 'following generalized seizures. Similar increases in vasopressin m R N A in PVN were also observed, but fewer were statistically significant. The most obvious and significant increase of vasopressin m R N A occurred at 1.5 h following the K A seizures. The vasopressin m R N A changes in the SON after K A seizures showed a pattern different from that for vasopressin or oxytocin in the PVN or oxytocin in SON. The m R N A level was not maintained at a certain level, but vacillated up and down (ranging from -1.21 to 13.34) with significant
increases at the one and three week time points. The reason for this instability is not known. The area occupied by oxytocin and vasopressin synthesizing neurons did not change significantly throughout the time periods examined. Thus, in contrast to observations made following adrenalectomy, 39,46,61 no change of phenotypic expression in parvocellular PVN neurons was observed following KA-induced seizures. Although previous studies have shown that different types of stimulations and stressors can alter the synthesis of oxytocin and vasopressin, 3'm21'4~ the information about the synthesis of these peptides following seizures is very limited. Besides an increase in vasopressin m R N A in SON following kindled seizures, 19only decreases in oxytocin and vasopressin m R N A have been reported in rat SON following metrazole or KA-induced seizures. 8 These results are in contrast to the result observed in the present study, but are likely due to differences in the design of the two studies, i.e. the K A dose used by Carter and Murphy g was lower (8 mg/kg) than ours (17 mg/kg), and it is not clear from their method description if the animals experienced stage 5 seizures, as those in our study did. Furthermore, their survival time is different from the one used in our study, as it was calculated from the injection time, while that of our animals was calculated from the time the animals had experienced two consecutive stage 5 seizures. Several mechanisms that could influence directly or indirectly the activity of PVN and SON neurons are likely to contribute to the up-regulation of oxytocin and vasopressin gene expression. A direct effect of K A on the hypothalamic neurons is possible and was suggested by Costa et al., m who demonstrated that K A induced arginine vasopressin secretion from hypothalamic explants. However, only a relative low density of K A receptoi" has been identified in the PVN and SON, 5 so that it seems more likely that the effects of K A on PVN and SON neurons are indirect. This is also supported by other studies, which have demonstrated increases of oxytocin and vasopressin release and/or synthesis following induction of seizures using other experimental models, such as kindling, post-trauma or alcohol withdrawal or following temporal lobe seizures and chronic electroconvulsive seizures, u9,20.n.38.60
Regulation of hypothalamic peptides following seizures It is well known that the hypothalamus receives neuronal input from multiple regions and from multiple neurotransmitter systems of the central nervous system; activities of several neurotransmitter systems (e.g. glutaminergic, catecholaminergic, serotonergic and peptidergic systems) were increased shortly following seizures. 48 Therefore, the perturbation of one or more of these transmitter systems and the subsequent changes in neuronal functions could be responsible for the changes in the oxytocin and vasopressin m R N A expressions following KA-induced seizures. Several studies indicate that the glutamate transmitter system might represent a likely candidate to mediate seizure-related increases in oxytocin and vasopressin mRNA. Glutaminergic innervation of the hypothalamus and the presence of glutamate receptors in the hypothalamus have been reported: '32'58 Glutamate release was increased in several brain regions following KA- or kindling-induced seizures, 4s'53 indicating the activation of the glutamate system following seizure induced status epilepticus. Catecholamine and serotonin transmitter systems are also involved in seizure responses and may contribute to the oxytocin and vasopressin mRNA increase following KA-induced seizures. They both stimulated the HPA axis at the level of the hypothalamus, ~° and catecholaminergic innervation of oxytocin neurons in rat PVN has been revealed by electron microscopyY Following KA-induced seizures, the turnover of these two biogenic amines has been increased in wide regions of brain, 4s providing further examples of the perturbations of neurotransmitter systems following KA-induced seizures. KA-induced seizures also modulate the activities of numerous peptidergic systems. 4s Among them, alterations in the function of somatostatin neurons could contribute to the observed increases of oxytocin and vasopressin synthesis. A Fos and somatostatin m R N A double-label study has demonstrated that somatostatin synthesizing neurons in the hippocampal formation were activated by KA,induced seizures, 4° the release of somatostatin was increased during the KA-induced seizures, 5° and the synthesis of somatostatin was increased for long periods of time following KA- induced seizures. 37 An electron microscopic study has also shown that somatostatin-immunoreactive axons terminate on oxytocin neurons in the rat PVN, 24 providing an anatomical basis for a somatostatin and oxytocin interaction. Thus, in addition to a direct effect of KA on hypothalamic neurons, the activation and perturbations of other neurotransmitter systems, such as glutamate, catecholamine, serotonin and somatostatin following KA-induced seizures could well contribute indirectly to the alterations in the oxytocin and vasopressin mRNA expressions observed in the present study. For the oxytocin m R N A in the PVN and SON, there was a second' peak increase at four weeks following KA-induced seizures. These late oxytocin
m R N A increases also lasted several weeks. While a similar pattern of m R N A increases occurred for vasopressin in the PVN, they were, ~t these late time points, not statistically significant. The mechanisms for the later increases in mRNA synthesis are unknown; it cannot be explained by direct KA effects on the hypothalamus since KA in the central nerve system was rapidly decreased after its administration. 35 Instead, mechanisms that resulted in tissue damage of the hippocampus or other limbic structures following KA-induced seizures might be expected to contribute to these late increases of mRNA. One well-known feature of the KA seizure model is the occurrence of progressive neuronal damage of limbic structures. 4 Most vulnerable are the CA3 pyramidal cells and interneurons of the hilus of dentate gyrus of the hippocampus. A common pattern of tissue damage can be induced by KA, regardless of the pathways of its administration. 4,51 One possibility for the damaged hippocampus to contribute to the increased oxytocin and/or vasopressin m R N A expressions could be via loss of its inhibitory influence on the HPA axis. It is known that the hippocampus exerts an inhibitory influence on the hypothalamic structures, especially the P V N . 15'28'31 This inhibitory effect appears to originate in the dorsal hippocampus, since it was eliminated by removal of the dorsal hippocampus or by hypothalamic deafferentation. 14 This pathway appears to utilize GABA neurons in the bed nucleus of stria terminalis as relay neurons) 1 However, excitatory pathways apparently also exist from the hippocampus to the hypothalamus, since direct electrical stimulation of ventral hippocampus resulted in excitatory responses of neurons in the SON. 42 The progressive tissue damage that occurs in the hippocampus following KA-induced seizures is likely to affect the target neurons of the hippocampal efferents, including neurons in the PVN and thus could represent a mechanism, through which in particular the late up-regulation of oxytocin and vasopressin synthesis could occur. Examples of long-lasting alterations in peptide synthesis, which may, at least in part, be attributed to the progressive tissue damage observed following KA-induced seizures, include increases in vasopressin synthesis in the SON (four months), ~9 increases in serotonin turnover (nine days), 49 and increases in somatostatin synthesis (30-60 days) in several areas of the brain) 7 These alterations, which might represent secondary effects of KA- or kindling-induced seizures, could also contribute to the up-regulated mRNA expressions of oxytocin and vasopressin observed in our study. Oxytocin and vasopressin neurons share a number of functional and morphological features; for example, they both regulate aspects of osmolality, they both are synthesized predominantly in magnocellular and only some parvocellular neurons, both types of neurons project to the pituitary, brainstem, spinal cord, etc. In the past, they have often been
Q. Sun et al.
thought of as highly homologous groups of neurons. Yet a differential regulation of oxytocin and vasopressin magnocellular neurons was clearly observed in the present study and has previously been observed by others. For example, oxytocin has been shown to be released following both restrain stress and ether exposure, while vasopressin release was induced only by ether exposure. 17 Serotonin stimulated oxytocin secretion only from rat PVN, but not SON. 55 Somatostatin only innervated the oxytocin neurons, but not vasopressin neurons in rat PVN. 24 While in the present study the synthesis for oxytocin at both PVN and SON, and for vasopressin at the PVN were increased after KA seizures, only the oxytocin mRNA increase in the PVN remained significantly high at one and two week post-seizure survival times. Furthermore, up-regulated vasopressin mRNA expression in the SON occurred in a time course pattern different from that for the others following KA-induced seizures. Since peptides have to be synthesized de novo, the synthesis of peptides has to match their release over time, if the neurons are not to be depleted or overloaded with peptide material. One can thus state that in case of peptide synthesis, increased synthesis implies increased release of the peptides. The upregulation of oxytocin and vasopressin mRNA expressions is therefore likely to alter the activity of those neurons, which receive projections from the hypothalamus. In addition to pituitary cells, which receive projections from the magnocellular neurons of the PVN and SON, neurons in, for example, the amygdala, hippocampus, brainstem and spinal cord, which have the oxytocin and vasopressin receptors 52 and receive projections from parvocellular neurons of the PVN, 45,47'62 are likely to be subjected to the increased release of these peptides. Since the application of oxytocin and vasopressin frequently increased neuronal activity, ~3 enhanced neuronal activation in
several sites can be expected. This in turn is likely to alter the neurochemical balance, and may or may not be translated into observable alterations of behavior. In our study, experimental animals with longer survival times exhibited nervous and overreactive behavior to such a degree that they could easily be distinguished from controls. A presumed increase of neuronal activation in the amygdala, brainstem and spinal cord, as it might occur following increased release of these peptides, could contribute to these behavioral effects. In addition, increased activation of pituitary cells, subsequent increase in ACTH release and activation of the "general stress response" could also contribute to the overreactive behavior. Indeed, the animals appeared to be overreactive to normal stimulations, such as touch, slight movements of the cage, etc., as if the threshold for elicitation of stressresponses had been lowered. It is of interest that these animals did not show alterations in drinking behavior and in general, appeared well groomed and nourished. Thus the osmolality regulation was either not effected, or some compensatory responses were sufficiently effective to assure the retention of normal behavior. CONCLUSIONS In summary, acute and long lasting increases in oxytocin and vasopressin mRNA expressions were observed in neurons of the PVN and SON following the induction of seizures by KA. While increased mRNA expressions of both peptides were observed, the data of the present study clearly indicated a differential regulation of hypothalamic oxytocin and vasopressin mRNA expressing neurons. Acknowledgements--We like to thank Dr S. Young III
(NIH) for providing the synthetic oligonucleotide probes, and Dr J. Olschowka for allowing the usage of his Cryostat. This study was supported by NIH Grants NS 18626 (D.P.) and NS 30925 (S.P.).
I. Aminoff M. J., Simon R. P. and Wiedemann E. (1984) The hormonal responses to generalized tonic-clonic seizures. Brain 107, 569-578. 2. Antoni F. A. (1986) Oxytocin receptors in the adrenohypophysis: evidence from radioligand binding studies. Endocrinology 119, 2393-2395. 3. Bartanusz V., Aubry J., Jezova D., BaffiJ. and Kiss J. Z. (1993) Up-regulation ofvasopressin mRNA in paraventricular hypophysiotrophic neurons after acute immobilization stress. Neuroendocrinology 58, 625-629. 4. Ben-Ari Y. (1985) Limbic seizure and brain damage produced by kainic acid: mechanisms and relevance to human temporal lobe epilepsy. Neuroscience 14, 375-403. 5. Brann D. W. (1995) Glutamate: a major excitatory transmitter in neuroendocrine regulation. Neuroendocrinology 61, 213-225. 6. Bruhn T. O., Sutton S. W., Plotsky P. M. and Vale W. W. (1986) Central administration of corticotropin-releasing factor modulates oxytocin secretion in the rat. Endocrinology 119, 1558-1563. 7. Buma P. and Nieuwenhuys R. (1987) Ultrastructural demonstration of oxytocin and vasopressin release sites in the neural lobe and median eminence of the rat by tannic acid immunogold methods. Neurosci. Lett. 74, 151-157. 8. Carter D. A. and Murphy D. (1993) Acute down-regulation of oxytocin and vasopressin mRNA levels following metrazole-induced seizures in the rat. Neurosci. Lett. 160, 135-138. 9. Carter D. A. and Lightman S. L. (1986) Diurnal pattern of stress-evoked neurohypophyseal hormone secretion: sexual dimorphism in rats. Neurosci. Lett. 71, 252-255. 10. Costa A., Yasin S. A., Hucks D., Forsling M. L. and Besser G. M. (1992) Differential effects of neuroexcitatory amino acids on corticotropin-releasing hormone-41 and vasopressin release from rat hypothalamic explants. Endocrinology 131, 2595-2602.
Regulation of hypothalamic peptides following seizures
11. Cullinan W. E., Herman J. P. and Watson S. J. (1993) Ventral subicular interaction with the hypothalamic paraventricular nucleus: evidence for a relay in the bed nucleus of the stria terminalis. J. comp. Neurol. 332, 1-20. 12. Ding J. M., Carver W. C., Terracio L. and Buggy J. (1994) Proto-oncogene c-fos and the regulation of vasopressin gene expression during dehydration. Molec. Brain Res. 21, 247-255. 13. Dreifuss J. J. and Raggenbass M. (1993) Oxytocin-responsive cells in the mammalian nervous system. Regul. Pept. 45, 109-114. 14. Feldman S. and Weidenfeld J. (1995) Neural mechanisms involved in the corticosteroid feedback effects on the hypothalamo-pituitary-adrenocortical axis. Prog. Neurobiol. 45, 129-141. 15. Feldman S. and Weidenfeld J. (1993) The dorsal hippocampus modifies the negative feedback effect of glucocorticoids on the adrenocortical and median eminence CRF-41 responses to photic stimulation. Brain Res. 614, 227-232. 16. Franci C. R., Anselmo-Franci J. A., Kozlowski G. P. and McCann S. M. (1993) Actions of endogenous vasopressin and oxytocin on anterior pituitary hormone secretion. Neuroendocrinology 57, 693-699• 17. Gibbs D. M. (1984) Dissociation of oxytocin, vasopressin and corticotropin secretion during different types of stress. Life Sci. 35, 478-491. 18. Gillies G• E., Linton E. A. and Lowry P. J. (1982) Corticotropin releasing activity of the new CRF is potentiated several times by vasopressin. Nature 299, 355-357. 19. Greenwood R. S., Meeker R. B., Abdou A. and Hayward J. N. (1994) Kindled seizures induce a long-term increase .in vasopressin mRNA. Molec. Brain Res. 24, 20-26. 20. Greenwood R. S., Meeker R. B. and Hayward J. N. (1991) Amygdala kindling elevates plasma vasopressin. Brain Res. 538, 9-14. 21. Herman J. P. and Sherman T. G. (1993) Acute stress upregulates vasopressin gene expression in parvocellular neurons of the hypothalamic paraventricular nucleus. Ann. N.Y. Acad. Sci. 689, 546-549. 22. Herman J. P., Schafer K. H., Sladek C. D., Day R., Young E. A., Akil H. and Watson S. J. (1989) Chronic electroconvulsive shock treatment elicits up-regulation of CRF and AVP mRNA in select populations of neuroendocrine neurons. Brain Res. 501, 235-246. 23. Higuchi T., Honda K. and Negoro H. (1986) Influence of oestrogen and noradrenergic afferent neurons on the response of LH and oxytocin to immobilization stress. J. Endocr. 110, 245-250. 24. Hisano S., Chikamori-Aoyama M., Aizawa T. and Fukui Y. (1993) Somatostatin-like immunoreactive axon terminals on oxytocin-like immunoreactive neurons in the paraventricular nucleus of the rat hypothalamus. Neurosci. Lett. 156, 21-23. 25. Horie S., Shioda S. and Nakai Y. (1993) Catecholaminergic innervation of oxytocin neurons in the paraventricular nucleus of the rat hypothalamus as revealed by double-labeling immunoelectron microscopy. Acta Anat. 147, 184-192• 26. Huang W. M., Gibson S. J., Facer P., Gu J. and Polak J. M. (1983) Improved section adhesion for immunocytochemistry using high molecular weight polymers of L-lysine as a slide coating. Histochemistry 77, 275-279. 27. Ivell R. and Richter D. (1984) Structure and comparison of the oxytocin and vasopressin genes from rat. Proc. nam. Acad. Sci. U.S.A. 81, 2006-2010. 28. Jacobson L. and Sapolsky R. (1991) The role of the hippocampus in feedback regulation of the hypothalamic-pituitary adrenocortical axis. Endocr. Rev. 12, 118-134. 29. Jenkins J. S. and Nussey S. S. (1991) The role of oxytocin: present concepts. Clin. Endocr. 34, 515-525. 30. Jones M. T., Gillham B., Campbell E. A., A1-Taher A. R. H., Chuang T. T. and Di Sciullo A. (1987) Pharmacology of neural pathways affecting CRH secretion. Ann. N.Y. Acad. Sci. 512, 162-175• 31. Knigge K. M. and Hays M. (1963) Evidence of inhibitive role of hippocampus in neural regulation of ACTH release. Proc, Soc. exp. Biol. Med. 114, 67-69• 32. Kus L., Beitz A. J., Kerr J. E. and Handa R. J. (1993) NMDA R1 receptor mRNA expression in the hypothalamus of intact, castrate and DHT-treated male rats. Soc. Neurosci. Abstr. 19, 919. 33. Lang R. E., Heil J. W., Ganten D., Hermann K., Unger T. and Rascher W. (1983) Oxytocin unlike vasopressin is a stress hormone in the rat. Neuroendocrinology 37, 314-316. 34. Lightman S. L. and Young W. S. III (1989) Lactation inhibits stress-mediated secretion of corticosterone and oxytocin and hypothalamic accumulation of corticotropin-releasing factor and enkephalin messenger ribonucleic acids. Endocrinology 124, 2358-2364. 35. Lothman E. W., Bertram E. H. III and Stringer J. L. (1991) Functional anatomy of hippocampal seizures. Prog. Neurobiol. 37, 1-82. 36. Makara G. B. (1992) The relative importance of hypothalamic neurons containing corticotropin-releasing factor or vasopressin in the regulation of adrenocorticotrophic hormone secretion. Ciba Found. Symp. 168, 43-53. 37. Marksteiner J. and Sperk G. (1988) Concomitant increase of somatostatin, neuropeptide Y and glutamate decarboxylase in the frontal cortex of rats with decreased seizure threshold. Neuroscience 26, 379-385. 38. Meierkord H., Shorvon S. and Lightman S. L. (1994) Plasma concentrations of prolactin, noradrenaline, vasopressin and oxytocin during and after a prolonged epileptic seizure. Acta Neurol. Scand. 90, 73-77. 39. Piekut D. T, and Joseph S. A• (1986) Co-existence of CRF and vasopressin immunoreactivity in parvocellular paraventricular neurons of the rat hypothalamus. Peptides 7, 891-898. 40. Pretel S., Applegate C. D. and Piekut D. T. (1995) Seizure induced activation of enkephalin and somatostatin neurons. Peptides 16, 951-957. 41. Pretel S. and Piekut D. T. (1989) Mediation of changes in paraventricular vasopressin and oxytocin mRNA content to the medullary vagal complex and spinal cord of the rat. J. chem. Neuroanat. 2, 327-334. 42. Prutskova N. P. and Petrov Yu. A. (1990) Electrophysiological investigation of the hippocampal projections to the neurosecretory cells of the supraoptic nucleus of the rat hypothalamus. Neurosci. Behav. Physiol. 20, 194-200. 43. Racine R. J. (1972) Modification of seizure activity by electrical stimulation--II. Motor seizure. Electroenceph. clin. Neurophysi'ol. 32, 281-284. 44. Raff H. (1993) Interactions between neurohypophysial hormones and the ACTH-adrenocortical axis. Ann. N.Y. Acad. Sci. 689, 411-425. 45. Sawchenko P. E. and Swanson L. M. (1982) Immunohistochemical identification of neurons in the paraventricular nucleus of the hypothalamus that project to the medulla or to the spinal cord in the rat. J. comp. Neurol. 205, 260-272•
Q. Sun et al.
46. Sofroniew M. V. (1985) Vasopressin, oxytocin and their related neurophysins. In Handbook of Chemical Neuroanatomy, Iiol. 4, GABA and Neuropeptides in the CNS, Part I (eds Bj6rklund A. and HSkfelt T.), pp. 93-165. Elsevier, Amsterdam. 47. Sofroniew M. V. and Schrell U. (1981) Evidence for a direct projection from oxytocin and vasopressin neurons in the hypothalamic paraventricular nucleus to the medulla oblongata: immunohistochemical visualization of both the horseradish peroxidase transported and the peptide produced by the same neurons. Neurosci. Lett. 22, 211-217. 48. Sperk G. (1994) Kainic acid seizures in the rat. Prog. Neurobiol. 42, 1-32. 49. Sperk G., Lassmann H., Baran H., Kish S. J., Seitelberger F. and Hornykiewicz O. (1983) Kainic acid induced seizures: neurochemical and histopathological changes. Neuroscience 10, 1301-1315. 50. Sperk G., Weiser R., Widmann R. and Singer E. A. (1986) Kainic acid induced seizures: changes in somatostatin, substance P and neurotensin. Neuroscience 17, 1117-I 126. 51. Tanaka T., Tanaka S., Fujita T., Takano K., Fukuda H., Sako K. and Yonemasu Y. (1992) Experimental complex partial seizures induced by a microinjection of kainic acid into limbic structures. Prog. Neurobiol. 38, 317-334. 52. Tribollet E., Barberis C., Jard S., Dubois-Dauphin M. and Dreifuss J. J. (1988) Localization and pharmacological characterization of high affinity binding sites for vasopressin and oxytocin in the rat brain by light microscopic autoradiography. Brain Res. 442, 105-118. 53. Ueda Y. and Tsuru N. (1994) Bilateral seizure-related changes of extracellular glutamate concentration in hippocampi during development of amygdaloid kindling. Epilepsy Res. 18, 85-88. 54. Vale W., Vaughan J., Smith M., Yamamoto G., Rivier J. and Rivier C. (1983) Effects of synthetic ovine corticotropinreleasing factor, glucocorticoids, catecholamines, neurohypophysial peptides and other substances on cultured corticotropic cells. Endocrinology 113, 1121-1131. 55. Van de Kar L. D., Rittenhouse P. A., Li Q., Levy A. D. and Brownfield M. S. (1995) Hypothalamic paraventricular, but not supraoptic neurons, mediate the serotonergic stimulation of oxytocin secretion. Brain Res. Bull. 36, 45-50. 56. Van de Kar L. D., Richardson-Morton K. D. and Rittenhouse P. A. (1991) Stress: neuroendocrine and pharmacological mechanisms. Methods Achier. expl Path. 14, 133-173. 57. Van den Pol A. N., Hermans-Borgmeyer I., Hofer M., Ghost P. and Heinemann S. (1994) Ionotropic glutamatereceptor gene expression in the hypothalamus: localization of AMPA, kainate and NMDA receptor RNA with in situ hybridization. J. comp. Neurol. 343, 428-444. 58. Van den Pol A. N. (1991) Glutamate and aspartate immunoreactivity in hypothalamic presynaptic axons. J. Neurosci. I1, 2087-2101. 59. Van Tol H. H., Voorhuis D. T. and Burbach J. P. (1987) Oxytocin gene expression in discrete hypothalamic magnoeellular cell groups is stimulated by prolonged salt loading. Endocrinology 120, 71-76. 60. Weinger M. B., Partridge L. B., Hauger R., Mirow A. and Brown M. (1991) Prevention of the cardiovascular and neuroendocrine response to electroconvulsive therapy--II. Effects of pretreatment regimens on catecholamines, ACTH, vasopressin, and cortisol. Anesth. Analg. 73, 563-569. 61. Wolfson B., Manning R. W., Davis L. G., Arontzon R. and Baldino F. Jr (1985) Co-localization of corticotropin releasing factor and vasopressin mRNA in neurons after adrenalectomy. Nature 315, 59-61~ 62. Zimmerman E. A., Nilaver G., Hou-Yu A. and Silverman A. J. (1984) Vasopressinergic and oxytocinergic pathways in the central nervous system. Fedn Proc. 43, 91-96. (Accepted 10 October 1995)