- Email: [email protected]
Involvement of the septum in the regulation of paraventricular vasopressin neurons by the subfornical organ in the rat
Neuroscience Letters, 92 (1988) 187-191
187
Elsevier Scientific Publishers Ireland Ltd.
NSL 05582
Involvement of the septum in the regulation of paraventricular vasopressin neurons by the subfornical organ in the rat Junichi T a n a k a , H i d e o Saito a n d K a t s u o Seto Department of Physiology, Kochi Medical School, Kochi (Japan) (Received 22 February 1988; Revised version received 8 June 1988; Accepted 8 June 1988)
Key words." Subfornical organ; Septum; Paraventricular nucleus; Efferent pathway; Vasopressin; Electrophysiology Extracellular recordings were obtained from 58 phasically active neurosecretory neurons in the hypothalamic paraventricular nucleus (PVN) of urethane-anesthetized male rats. Of these PVN neurons, 39 exhibited an increase and 11 displayed a reduction in ongoing activity following electrical stimulation of the subfornical organ (SFO), while the remaining neurons were unresponsive. Microinjection of the local anesthetic lidocaine into the medial septum reversibly abolished the SFO stimulus-eveoked reduction in 7 out of 9 PVN neurons tested, whereas similar injection was without effect on the stimulus-evoked increase in 18 out of 20 PVN neurons tested. These results suggest that the SFO efferents through the medial septum to the PVN exert a predominantly inhibitory influence on the excitability of putative vasopressin (VP)-secreting neurons in the PVN.
It has been postulated that the subfornical organ (SFO) of the rat plays an important role in regulating vasopressin (VP) release from the posterior pituitary in response to circulating angiotensin II. Destruction of the SFO [3, 4], transection of the SFO efferent projections [5] or pretreatment with the angiotensin II antagonist in the SFO [14] prevents the rise in the excitability of VP-secreting cells in the hypothalamic paraventricular (PVN) [3, 14] and supraoptic (SON) [3] nuclei and the elevation in plasma VP concentrations [4, 5] following peripheral administration of angiotensin II. Neuroanatomical studies have demonstrated that the SFO sends efferent fibers to both the PVN and SON as well as to the other brain regions that in turn project to both the sites [6, 7, 11, 12]. These findings suggest that not only the direct SFO efferents to the PVN and SON but also the indirect SFO efferents through the other brain regions may participate in the control of VP release. Since the septum is known to connect with both the SFO [6, 7] and the PVN and SFO [1, 8, 11, 12], we focused on the involvement of this region in the above-mentioned action of the SFO. In the Correspondence: J. Tanaka, Department of Physiology, Kochi Medical School, Nankoku, Kochi 781 51, Japan. 0304-3940/88/$ 03.50 © 1988 Elsevier Scientific Publishers Ireland Ltd.
188
present study, experiments were done to examine the effect of blockage of neuronal activity in the medial septum produced by the microinjection of the local anesthetic lidocaine on responses of putative VP-secreting neurons in the PVN to electrical stimulation of the SFO. Experiments were conducted on male Wistar rats weighing 300-360 g anesthetized with intraperitoneal urethane (1.25 g/kg), the animals were placed in a stereotaxic frame in a prone position. A concentric bipolar stainless-steel electrode (28 gauge, tip-ring separation 0.3 mm) was stereotaxically lowered in the SFO site 7.6-7.9 mm anterior to the interaural line, 0.0-0.3 mm lateral to the midline, 4.1-4.5 mm below the surface of the cortex, according to the atlas of Paxinos and Watson [9]. Two 30gauge stainless-steel cannulas fused side by side so that the tip of one protruded by 0.8-1.0 mm were placed in the region of the medial septum (Fig. 1D). The electrode A
0,4
,
0.8,
0.4
mA
D
LSI:~i
"
i¸
2mln
B
0.5
0.8
0;5
0.8
' I min ' 0.5
mA
J C
0.5
.s
.5 IN
~
0.8
~o.
0~5 ' m)
~
.5
l.O
~
0.5 ~
0.8 1()
i_o
0;8 lwO
mA Hz
q~ w
Fig. 1. A~E: rate-meter records obtained from three phasic.ally active neurosecretory cells in the hypothalamic paraventricular nucleus (PVN). Solid and open horizontal oblongs abo~e the records indicate time of electrical stimulation of the subfornical organ (SFO) and period of microinjection of lidocalne (Lid) into the medial septum (MS), respectively. The top (0.4-0.8) and next (5-20) numbers above the records refer to stimulus intensity (mA) and frequency (Hz), respectively. Repetitive stimulation of the SFO prodnced either an increase (A,B) or a reduction (C) in the ongoing activity. The injection of Lid into the MS reversibly abolished the SFO stimulus-evoked increase in B and reduction in C, but did not affect the increase in A. D: closed circles on two schematic transverse sections (9.2 and 9.7 mm anterior to the interaural line) depict the location of the cannula tips in the MS where microinjections of Lid prevented the SFO stimulus-evoked responses. 2n, optic nerve; LSI, lateral septal nucleus, intermediate part; LSV, lateral septal nucleus, ventral part; MS, medial septal nucleus; VDBV, nucleus of the ventrical limb of the diagonal band, dorsal part; VDBV, nucleus of the ventrical limb of the diagonal band, ventral part.
189
and cannulas were attached to the skull with jeweler's screws and dental cement. The stimulating electrode and injection cannulas were connected to an isolated stimulation unit that delivered 0.05 ms current pulses of 0.1-1.0 mA and to two microinjectors (N. Takahashi Co., BV-341) using polyethylene tubes, respectively. The animal was then placed in the supine position. The basal hyothalamus was exposed via a transpharyngeal route. An additional bipolar electrode used to activate the terminals of neurohypophyseal PVN neurons was positioned in the posterior pituitary. Extracellular single-unit recordings from the PVN were obtained through glass micropipettes filled with 0.5 M sodium acetate solution containing 2% Pontamine sky blue. Signals were amplified conventionally, bandpass filtered, visualized on a storage oscilloscope and led to a signal processer (Nihon Koden, ATAC-450) programmed for spike train analysis. PVN cells projecting to the posterior pituitary were identified by antidromic activation (i.e. all-or-none constant latency responses at threshold, constant latency following of two stimuli at intervals under 10 ms, and collision cancellation of antidromic responses by spontaneous action potentials at appropriate intervals). The effects of electrical stimulation of the SFO on the excitability of PVN neurons were classified as increase, decrease, or no change in the frequency of action potentials. Such changes were assessed by examining the mean firing rate during SFO stimulation. A 30% change in the mean intraburst frequency was chosen as a minimum arbitrary level of change. Each microinjection of 2% iidocaine (Fujisawa) or isotonic saline into the medial septum was administered in a volume of 0.2/d at a rate of 0.01 /d/s. The lidocaine injection was achieved through one of the cannulas mentioned above. When the effect of the drug was not observed, another one of the cannulas was used for the injection. At the end of the experiment, the recording and stimulating sites were marked by depositing a small amount of dye and iron, respectively. The tip positions of microinjection sites were marked by electrolytic lesion. The animal was then sacrificed with an overdose of urethane and perfused with 10 % formalin containing 3% potassium ferrocyanide and ferricyanide. The marking sites were verified histologically in 50/tm coronal sections cut on a microtome and stained with Neutral red. The stereotaxic coordinates for marking sites were determined according to the atlas of Paxinos and Watson [9] (Fig. 1D). In the rat, magnocellular neurosecretory PVN cells display phasic, continuous or irregular firing [10]. The phasic activity is thought to be a characteristic of VP-secreting cells [I 0]. Thus, the phasically firing PVN units, which had been identified by their antidromic responses (n=58; mean _+ S.D. latency, 13.9 _+ 1.9 ms), were tested for a response to repetitive electrical stimulation (30-200 pulses at 5 20 Hz) of the SFO. Of the units tested, 39 displayed an increase in ongoing excitability and an initiation of discharge in the silent period (Fig. 1A,B), 11 exhibited a suppression in ongoing activity and often yieled a silent period (Fig. 1C), while the remainder (n= 8) were unresponsive. The efficiency of such stimuli was enhanced by increasing the number, intensity, and frequency of stimuli (Fig. IA-C). The results indicate that the SFO efferents have both facilitatory and inhibitory influences on the excitability of putative VP-secreting neurons in the PVN, and are in agreement with the previous observations [2, 13].
190 TABLE I RESPONSES OF PHASICALLY ACTIVE NEUROSECRETORY CELLS IN THE PVN TO ELECTRICAL STIMULATION OF THE SFO AND THE EFFECTIVENESS OF MICROINJECTED LIDOCAINE OR SALINE INTO THE MEDIAL SEPTUM IN BLOCKING THE RESPONSES In the cells that displayed excitatory response, the remaining 14 cells were not tested. Response
Number of cells
Lidocaine
Saline
Excitatory response Inhibitory response No response
39 11 8
2 / 20* 7/ 9 . . .
0/ 5 0!2 .
.
.
*Number of cells that displayed blockage/Number of cells tested.
The contribution of the medial septum to the effects of SFO stimulation on PVN units was assessed by examining whether the evoked response could be blocked by the treatment with lidocaine. Microinjection of lidocaine into the medial septum abolished the stimulus-evoked reduction in 7 out of 9 units tested (Fig. 1C), but did not affect the remainder. Similar injection, on the other hand, could not prevent the stimulus-evoked increase in 18 out o f 20 units tested (Fig. I A), while only 2 units showed the lidocaine blockage (Fig. I B). In all the units that displayed the blockage, the effect of lidocaine was reversible with a duration o f action of less than 15 min (Fig. I B,C). As a control, isotonic saline injected into the medial septum did not cause a change in both the stimulus-evoked reduction and increase in ongoing activity. The results are summarized in Table I. These findings suggest that the SFO efferents to the PVN through the medial septum may exert a predominantly inhibitory influence on the excitability of putative VP-secreting neurons in the PVN. Although the lidocaine blockage of the responses allows for possible effects on the vicinity of the medial septum, such as the lateral part of the septum, the present data suggest that the greater part of the inhibitory pathways from the SFO to putative VP-secreting neurons in the PVN may pass through, or terminate in, the region of the septum. It has been reported that the septum principally acts to suppress the activity of magnocellular neurosecretory cells in the PVN [8]. A recent study has demonstrated that the SFO efferents activate neurons in the medial septum [1]. Thus, it might be expected that medial septum neurons relay activations of SFO neuronal activity to putative VP-secreting neurons in the P V N which result in inhibited activity. Further studies using antagonistic substances of possible SFO efferent transmitters will be necessary to clarify this hypothesis. In addition, because of several investigations have shown that the SFO facilitates the release of VP in response to intravenous angiotensin II [3-5, 14], the functional significance o f the inhibitory SFO efferents via the septum to VP-secreting neurons in the P V N remains unclear. 1 Ferguson, A.V., Bourque, C.W. and Renaud, L.P., Subfornical organ and supraoptic nucleus connections with septal neurons in rats, Am. J. Physiol., 249 (1985) R214--R218. 2 Ferguson, A.V., Day, T.A. and Renaud, L.P., Subfornical organ efferents influence the excitability
191
of neurohypophysial and tuberoinfundibular paraventricular nucleus neurons in the rat, Neuroendocrinology, 39 (1984) 423~428. 3 Ferguson, A.V. and Renaud, L.P., Systemic angiotensin acts at subfornical organ to facilitate activity of neurohypophysial neurons, Am. J. Physiol., 251 (1986) R712 R717. 4 lovino, M. and Steardo, L., Vasopressin release to central and peripheral angiotensin II in rats with lesions of the subfornical organ, Brain Res., 322 (1984) 365-368. 5 Knepel, W., Nutto, D. and Meyer, D.K., Effect of transection of subfornical organ efferent projections on vasopressin release induced by angiotensin or isoprenaline in the rat, Brain Res., 248 (1982) 180 184. 6 Lind, R.W., Van Hoesen, G.W. and Johnson, A.K., An HRP study of the connections of the subfornical organ of the rat, J. Comp. Neurol., 210 (1982) 265-277. 7 Miselis, R.R., The efferent projections of the subfornical organ of the rat: a circumventricular organ within a neural network subserving water balance, Brain Res., 230 (1981) 1~23, 8 Negoro, H., Visessuwan, S. and Holland, R.C., Inhibition and excitation of units in paraventricular nucleus after stimulation of the septum, amygdala, and neurohypophysis, Brain Res., 57 (1973) 479 483. 9 Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, Academic, New York, 1982. 10 Poulain, D.A. and Wakerley, J.B., Electrophysiology of hypothalamic magnocellular neurones secreting oxytocin and vasopressin, Neuroscience, 7 (1982) 773--808. 11 Sawchenko, P.E. and Swanson, L.W., The organization of forebrain afferents to the paraventricular and supraoptic nuclei of the rat, J. Comp. Neurol., 218 (1983) 121 144. 12 Silverman, A.J., Hoffman, D.L. and Zimmerman, E,A., The descending afferent connections of the paraventricular nucleus of the hypothalamus (PVN), Brain Res. Bull., 6 (1981) 47 61. 13 Tanaka, J., Kaba, H., Saito, H. and Seto, K., Electrophysiological evidence that circulating angiotensin I 1 sensitive neurons in the subfornical organ alter the activity of hypothalamic paraventricular neurohypophyseal neurons in the rat, Brain Res., 342 (1985) 361 365. 14 Tanaka, J., Saito, H., Kaba, H. and Seto, K., Subfornical organ neurons act to enhance the activity of paraventricular vasopressin neurons in response to intravenous angiotensin II, Neurosci. Res., 4 (1987) 424~427.
187
Elsevier Scientific Publishers Ireland Ltd.
NSL 05582
Involvement of the septum in the regulation of paraventricular vasopressin neurons by the subfornical organ in the rat Junichi T a n a k a , H i d e o Saito a n d K a t s u o Seto Department of Physiology, Kochi Medical School, Kochi (Japan) (Received 22 February 1988; Revised version received 8 June 1988; Accepted 8 June 1988)
Key words." Subfornical organ; Septum; Paraventricular nucleus; Efferent pathway; Vasopressin; Electrophysiology Extracellular recordings were obtained from 58 phasically active neurosecretory neurons in the hypothalamic paraventricular nucleus (PVN) of urethane-anesthetized male rats. Of these PVN neurons, 39 exhibited an increase and 11 displayed a reduction in ongoing activity following electrical stimulation of the subfornical organ (SFO), while the remaining neurons were unresponsive. Microinjection of the local anesthetic lidocaine into the medial septum reversibly abolished the SFO stimulus-eveoked reduction in 7 out of 9 PVN neurons tested, whereas similar injection was without effect on the stimulus-evoked increase in 18 out of 20 PVN neurons tested. These results suggest that the SFO efferents through the medial septum to the PVN exert a predominantly inhibitory influence on the excitability of putative vasopressin (VP)-secreting neurons in the PVN.
It has been postulated that the subfornical organ (SFO) of the rat plays an important role in regulating vasopressin (VP) release from the posterior pituitary in response to circulating angiotensin II. Destruction of the SFO [3, 4], transection of the SFO efferent projections [5] or pretreatment with the angiotensin II antagonist in the SFO [14] prevents the rise in the excitability of VP-secreting cells in the hypothalamic paraventricular (PVN) [3, 14] and supraoptic (SON) [3] nuclei and the elevation in plasma VP concentrations [4, 5] following peripheral administration of angiotensin II. Neuroanatomical studies have demonstrated that the SFO sends efferent fibers to both the PVN and SON as well as to the other brain regions that in turn project to both the sites [6, 7, 11, 12]. These findings suggest that not only the direct SFO efferents to the PVN and SON but also the indirect SFO efferents through the other brain regions may participate in the control of VP release. Since the septum is known to connect with both the SFO [6, 7] and the PVN and SFO [1, 8, 11, 12], we focused on the involvement of this region in the above-mentioned action of the SFO. In the Correspondence: J. Tanaka, Department of Physiology, Kochi Medical School, Nankoku, Kochi 781 51, Japan. 0304-3940/88/$ 03.50 © 1988 Elsevier Scientific Publishers Ireland Ltd.
188
present study, experiments were done to examine the effect of blockage of neuronal activity in the medial septum produced by the microinjection of the local anesthetic lidocaine on responses of putative VP-secreting neurons in the PVN to electrical stimulation of the SFO. Experiments were conducted on male Wistar rats weighing 300-360 g anesthetized with intraperitoneal urethane (1.25 g/kg), the animals were placed in a stereotaxic frame in a prone position. A concentric bipolar stainless-steel electrode (28 gauge, tip-ring separation 0.3 mm) was stereotaxically lowered in the SFO site 7.6-7.9 mm anterior to the interaural line, 0.0-0.3 mm lateral to the midline, 4.1-4.5 mm below the surface of the cortex, according to the atlas of Paxinos and Watson [9]. Two 30gauge stainless-steel cannulas fused side by side so that the tip of one protruded by 0.8-1.0 mm were placed in the region of the medial septum (Fig. 1D). The electrode A
0,4
,
0.8,
0.4
mA
D
LSI:~i
"
i¸
2mln
B
0.5
0.8
0;5
0.8
' I min ' 0.5
mA
J C
0.5
.s
.5 IN
~
0.8
~o.
0~5 ' m)
~
.5
l.O
~
0.5 ~
0.8 1()
i_o
0;8 lwO
mA Hz
q~ w
Fig. 1. A~E: rate-meter records obtained from three phasic.ally active neurosecretory cells in the hypothalamic paraventricular nucleus (PVN). Solid and open horizontal oblongs abo~e the records indicate time of electrical stimulation of the subfornical organ (SFO) and period of microinjection of lidocalne (Lid) into the medial septum (MS), respectively. The top (0.4-0.8) and next (5-20) numbers above the records refer to stimulus intensity (mA) and frequency (Hz), respectively. Repetitive stimulation of the SFO prodnced either an increase (A,B) or a reduction (C) in the ongoing activity. The injection of Lid into the MS reversibly abolished the SFO stimulus-evoked increase in B and reduction in C, but did not affect the increase in A. D: closed circles on two schematic transverse sections (9.2 and 9.7 mm anterior to the interaural line) depict the location of the cannula tips in the MS where microinjections of Lid prevented the SFO stimulus-evoked responses. 2n, optic nerve; LSI, lateral septal nucleus, intermediate part; LSV, lateral septal nucleus, ventral part; MS, medial septal nucleus; VDBV, nucleus of the ventrical limb of the diagonal band, dorsal part; VDBV, nucleus of the ventrical limb of the diagonal band, ventral part.
189
and cannulas were attached to the skull with jeweler's screws and dental cement. The stimulating electrode and injection cannulas were connected to an isolated stimulation unit that delivered 0.05 ms current pulses of 0.1-1.0 mA and to two microinjectors (N. Takahashi Co., BV-341) using polyethylene tubes, respectively. The animal was then placed in the supine position. The basal hyothalamus was exposed via a transpharyngeal route. An additional bipolar electrode used to activate the terminals of neurohypophyseal PVN neurons was positioned in the posterior pituitary. Extracellular single-unit recordings from the PVN were obtained through glass micropipettes filled with 0.5 M sodium acetate solution containing 2% Pontamine sky blue. Signals were amplified conventionally, bandpass filtered, visualized on a storage oscilloscope and led to a signal processer (Nihon Koden, ATAC-450) programmed for spike train analysis. PVN cells projecting to the posterior pituitary were identified by antidromic activation (i.e. all-or-none constant latency responses at threshold, constant latency following of two stimuli at intervals under 10 ms, and collision cancellation of antidromic responses by spontaneous action potentials at appropriate intervals). The effects of electrical stimulation of the SFO on the excitability of PVN neurons were classified as increase, decrease, or no change in the frequency of action potentials. Such changes were assessed by examining the mean firing rate during SFO stimulation. A 30% change in the mean intraburst frequency was chosen as a minimum arbitrary level of change. Each microinjection of 2% iidocaine (Fujisawa) or isotonic saline into the medial septum was administered in a volume of 0.2/d at a rate of 0.01 /d/s. The lidocaine injection was achieved through one of the cannulas mentioned above. When the effect of the drug was not observed, another one of the cannulas was used for the injection. At the end of the experiment, the recording and stimulating sites were marked by depositing a small amount of dye and iron, respectively. The tip positions of microinjection sites were marked by electrolytic lesion. The animal was then sacrificed with an overdose of urethane and perfused with 10 % formalin containing 3% potassium ferrocyanide and ferricyanide. The marking sites were verified histologically in 50/tm coronal sections cut on a microtome and stained with Neutral red. The stereotaxic coordinates for marking sites were determined according to the atlas of Paxinos and Watson [9] (Fig. 1D). In the rat, magnocellular neurosecretory PVN cells display phasic, continuous or irregular firing [10]. The phasic activity is thought to be a characteristic of VP-secreting cells [I 0]. Thus, the phasically firing PVN units, which had been identified by their antidromic responses (n=58; mean _+ S.D. latency, 13.9 _+ 1.9 ms), were tested for a response to repetitive electrical stimulation (30-200 pulses at 5 20 Hz) of the SFO. Of the units tested, 39 displayed an increase in ongoing excitability and an initiation of discharge in the silent period (Fig. 1A,B), 11 exhibited a suppression in ongoing activity and often yieled a silent period (Fig. 1C), while the remainder (n= 8) were unresponsive. The efficiency of such stimuli was enhanced by increasing the number, intensity, and frequency of stimuli (Fig. IA-C). The results indicate that the SFO efferents have both facilitatory and inhibitory influences on the excitability of putative VP-secreting neurons in the PVN, and are in agreement with the previous observations [2, 13].
190 TABLE I RESPONSES OF PHASICALLY ACTIVE NEUROSECRETORY CELLS IN THE PVN TO ELECTRICAL STIMULATION OF THE SFO AND THE EFFECTIVENESS OF MICROINJECTED LIDOCAINE OR SALINE INTO THE MEDIAL SEPTUM IN BLOCKING THE RESPONSES In the cells that displayed excitatory response, the remaining 14 cells were not tested. Response
Number of cells
Lidocaine
Saline
Excitatory response Inhibitory response No response
39 11 8
2 / 20* 7/ 9 . . .
0/ 5 0!2 .
.
.
*Number of cells that displayed blockage/Number of cells tested.
The contribution of the medial septum to the effects of SFO stimulation on PVN units was assessed by examining whether the evoked response could be blocked by the treatment with lidocaine. Microinjection of lidocaine into the medial septum abolished the stimulus-evoked reduction in 7 out of 9 units tested (Fig. 1C), but did not affect the remainder. Similar injection, on the other hand, could not prevent the stimulus-evoked increase in 18 out o f 20 units tested (Fig. I A), while only 2 units showed the lidocaine blockage (Fig. I B). In all the units that displayed the blockage, the effect of lidocaine was reversible with a duration o f action of less than 15 min (Fig. I B,C). As a control, isotonic saline injected into the medial septum did not cause a change in both the stimulus-evoked reduction and increase in ongoing activity. The results are summarized in Table I. These findings suggest that the SFO efferents to the PVN through the medial septum may exert a predominantly inhibitory influence on the excitability of putative VP-secreting neurons in the PVN. Although the lidocaine blockage of the responses allows for possible effects on the vicinity of the medial septum, such as the lateral part of the septum, the present data suggest that the greater part of the inhibitory pathways from the SFO to putative VP-secreting neurons in the PVN may pass through, or terminate in, the region of the septum. It has been reported that the septum principally acts to suppress the activity of magnocellular neurosecretory cells in the PVN [8]. A recent study has demonstrated that the SFO efferents activate neurons in the medial septum [1]. Thus, it might be expected that medial septum neurons relay activations of SFO neuronal activity to putative VP-secreting neurons in the P V N which result in inhibited activity. Further studies using antagonistic substances of possible SFO efferent transmitters will be necessary to clarify this hypothesis. In addition, because of several investigations have shown that the SFO facilitates the release of VP in response to intravenous angiotensin II [3-5, 14], the functional significance o f the inhibitory SFO efferents via the septum to VP-secreting neurons in the P V N remains unclear. 1 Ferguson, A.V., Bourque, C.W. and Renaud, L.P., Subfornical organ and supraoptic nucleus connections with septal neurons in rats, Am. J. Physiol., 249 (1985) R214--R218. 2 Ferguson, A.V., Day, T.A. and Renaud, L.P., Subfornical organ efferents influence the excitability
191
of neurohypophysial and tuberoinfundibular paraventricular nucleus neurons in the rat, Neuroendocrinology, 39 (1984) 423~428. 3 Ferguson, A.V. and Renaud, L.P., Systemic angiotensin acts at subfornical organ to facilitate activity of neurohypophysial neurons, Am. J. Physiol., 251 (1986) R712 R717. 4 lovino, M. and Steardo, L., Vasopressin release to central and peripheral angiotensin II in rats with lesions of the subfornical organ, Brain Res., 322 (1984) 365-368. 5 Knepel, W., Nutto, D. and Meyer, D.K., Effect of transection of subfornical organ efferent projections on vasopressin release induced by angiotensin or isoprenaline in the rat, Brain Res., 248 (1982) 180 184. 6 Lind, R.W., Van Hoesen, G.W. and Johnson, A.K., An HRP study of the connections of the subfornical organ of the rat, J. Comp. Neurol., 210 (1982) 265-277. 7 Miselis, R.R., The efferent projections of the subfornical organ of the rat: a circumventricular organ within a neural network subserving water balance, Brain Res., 230 (1981) 1~23, 8 Negoro, H., Visessuwan, S. and Holland, R.C., Inhibition and excitation of units in paraventricular nucleus after stimulation of the septum, amygdala, and neurohypophysis, Brain Res., 57 (1973) 479 483. 9 Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, Academic, New York, 1982. 10 Poulain, D.A. and Wakerley, J.B., Electrophysiology of hypothalamic magnocellular neurones secreting oxytocin and vasopressin, Neuroscience, 7 (1982) 773--808. 11 Sawchenko, P.E. and Swanson, L.W., The organization of forebrain afferents to the paraventricular and supraoptic nuclei of the rat, J. Comp. Neurol., 218 (1983) 121 144. 12 Silverman, A.J., Hoffman, D.L. and Zimmerman, E,A., The descending afferent connections of the paraventricular nucleus of the hypothalamus (PVN), Brain Res. Bull., 6 (1981) 47 61. 13 Tanaka, J., Kaba, H., Saito, H. and Seto, K., Electrophysiological evidence that circulating angiotensin I 1 sensitive neurons in the subfornical organ alter the activity of hypothalamic paraventricular neurohypophyseal neurons in the rat, Brain Res., 342 (1985) 361 365. 14 Tanaka, J., Saito, H., Kaba, H. and Seto, K., Subfornical organ neurons act to enhance the activity of paraventricular vasopressin neurons in response to intravenous angiotensin II, Neurosci. Res., 4 (1987) 424~427.