- Email: [email protected]
ventromedial hypothalamic nuclei in the rat
EXPERIMENTAL
92,563-570
NEUROLOGY
(1986)
Electrophysiologic Evidence for Connections between the Supraoptic and the Arcuate/Ventromedial Hypothalamic Nuclei in the Rat DAVID Laboratory
SAPHIER
AND SHAUL FELDMAN’
of Neurophysiology. Department of Neurology. Hebrew University-Hadassah McJdical School, Received
September
6. 1985; revision
Hadassah Jerusalem
received
January
University Hospital 91120. Israel
and
6. 1986
Extracellular action potentials were recorded from 48 single units located in the hypothalamic arcuate and ventromedial nuclei. Fifteen percent ofthe cells were identified as projecting to the median eminence and some of these cells may have belonged to the tuberoinfundibular dopaminergic systems.Responses of all cells to stimulation of the ipsilateral supraoptic nucleus were recorded, 17% of ventromedial nucleus neurons were antidromically identified as projecting to the supraoptic nucleus. None of the latter cells was also identified as projecting to the median eminence. Three of six identified tuberoinfundibular and eight unidentified ventromedial nucleus cells were found to be excited by stimulation of the supraoptic nucleus. One arcuate cell identified as projecting to the median eminence was nonresponsive to supraoptic stimulation. Orthodromic inhibitory responses were recorded from 17% of all cells recorded but no inhibitory responses were recorded from cells identified as projecting to the median eminence. We suggestthat these results may provide some neurophysiologic explanations for the observed interrelationships between oxytocin and prolactin secretion, and between vasopressin and growth hormone secretion. 0 1986 Academic press, Inc. INTRODUCTION
The supraoptic nucleus (SON) contains magnocellular cells of the neurohypophyseal neurosecretory system that synthesize oxytocin and vasopressin (12). These hormones are released in response to the suckling stimulus and elevation of plasma osmotic pressure, respectively. The neural control of their Abbreviations: SON-supraoptic nucleus, ARC/VMH-arcuate/ventromedial nuclei, PVNparaventricular nucleus, ME-median eminence. ’ This research was supported by the Lena P. Harvey Endowment Fund for Neurological Research. 563 0014-4886/86 $3.00 Copyright Q 1986 by Academic Press. Inc. All rights of reproduction in any form reserved.
564
SAPHIER AND FELDMAN
activation has been the subject of many studies and has been reviewed extensively (14, 15, 25, 26). Anatomical evidence has demonstrated that substantial interconnections exist between the SON and adjacent nuclei of the mediobasal hypothalamus ( 18). Magnocellular neurons of the SON and paraventricular nucleus (PVN) have also been shown electrophysiologically to send axon collaterals to these other regions of the endocrine hypothalamus (10, 2 1). This suggests interactions between neurons controlling neurohypophyseal secretory activity and those in the tuberal hypothalamus concerned with the regulation of anterior pituitary secretory activity. The suckling stimulus is also known to cause an elevation of plasma prolactin (3 1), believed to be primarily under inhibitory control by dopaminesynthesizing and -secreting tuberoinfundibular neurons in the arcuate and ventromedial nuclei (ARC/VMH) of the mediobasal hypothalamus (3, 8,23, 32). Dopamine is also known to stimulate the release of oxytocin (4, 22) and increase the firing of putative oxytocin-secreting cells of the SON (20). Thus it is possible that any physiologic release of oxytocin mediated by dopamine could arise from dopaminergic tuberoinfundibular neuron axon collaterals or from other dopaminergic, or other systems (1,2,24). As it has been shown that more than one-half of mediobasal hypothalamic neurons are excited by the iontophoretic application of sodium ions (17) it is possible that these neurons may be involved in osmoregulatory processes. Such processes are primarily regulated by vasopressin-secreting cells of the SON and PVN (12, 26), but local neural circuits may also be of importance (10, 11, 18, 2 1, 29). In view of the above considerations, we examined any possible connections between the SON and ARC/VMH. METHODS Male albino rats (body weight, 260 to 320 g) of the Hebrew University strain were anesthetized with a single injection of urethane (ethyl carbamate: 1.2 g/kg, i.p. as a 25% W/V solution). The animals were then secured in a stereotaxic frame, and a concentric bipolar stimulating electrode was lowered into the SON from a dorsal aspect. This electrode consisted of an outer pole of 500-pm diameter insulated to within 300 pm of the tip. The inner pole was 120-pm-diameter wire insulated to within approximately 150 pm of the tip and separated from the cathode by the same distance. The apparatus was subsequently inverted and the median eminence and pituitary stalk exposed according to a modification of the technique of Dreifuss and Ruf (7). A nonconcentric bipolar stimulating electrode, consisting of two 120~pm-diameter Nichrome wires insulated to within 700 pm of the tip and separated by approximately 1 mm, was lowered and rested upon the median eminence (ME).
SUPRAOPTIC-MEDIOBASAL
HYPOTHALAMIC
CONNECTIONS
565
This electrode allowed for the antidromic identification of tuberoinfundibular neurosecretory cells (8). Extracellular recordings were made from single cells located within the ARC/VMH. All recording sites were ipsilateral with the stimulating electrode in the SON and were made through glass micropipets filled with a solution of 0.5 M sodium acetate in 2 M sodium chloride and containing fast green dye. Single-unit signals were amplified through a Grass Instruments (model P5 11) preamplifier and monitored with a Grass Instruments audio monitor, using conventional and storage oscilloscopes. Stimuli were delivered to the ME as single (0.2-Hz), bipolar square-wave pulses of 1-ms total duration and up to I-mA peak-to-peak current strength. The same parameters were used to stimulate the SON, except that the current strength was 350 to 750 /IA. Responses of the cells were recorded as antidromic if the antidromic action potential occurred after the stimulus at a constant latency, and if the cell was able to follow trains of repeated stimuli at frequencies greater than 150 Hz. A third criterion, that could be tested only for spontaneously active neurons, was the presence of collision between spontaneous action potentials and antidromically evoked action potentials. Orthodromic excitation was distinguished by the appearance of evoked action potentials of variable latency, and orthodromic inhibition by the observation of extended periods of silence following the stimulation that exceeded the unstimulated interspike interval mode. At the termination of each experiment the final site of recording was marked by ejection of the green dye from the tip of the electrode. Previous recording sites could then be estimated according to the calibrated movements of the hydraulic microdrive. Histological verification of the electrode sites was by examination of 45-pm frozen sections cut from the perfused (10% Formalinsaline) and postfixed tissue. In all animals the stimulating electrode track was seen to terminate in the SON. RESULTS Extracellular action potentials were recorded from a total of 48 single units, of which 12 were located within the ARC and the remaining 36 within the VMH. The cells were located at mean depths from the median eminence of 0.40 + 0.04 and 1.15 f 0.07 mm, respectively. The ARC/VMH division was defined at 0.7 mm above the ME. Firing rates were not found to be significantly different, although ARC cells were found to fire somewhat more slowly at 1.4 f 0.5 Hz (X + SE), with VMH cells firing at 2.0 2 0.7 Hz. Six VMH cells (17%) were antidromically identified as projecting to the SON, exhibiting a mean antidromic invasion latency of 10.6 f 4.1 ms; None of those cells could be antidromically invaded following stimulation of the
566
SAPHIER AND FELDMAN
ME. Stimulation of the SON evoked orthodromic, i.e., synaptic, responses from 56% of the VMH neurons. Of the responsive VMH cells, 11 (31%) showed orthodromic excitation following SON stimulation and 3 of these cells were identified as projecting to the ME. Six nonidentified cells were inhibited by the stimulation and the remaining 13 were nonresponsive. Of the nonresponsive VMH population, three were identified as projecting to the ME. Orthodromic responses were recorded from nine (75%) of the ARC cells and none of these cells was identified as projecting to the ME. One identified ARC cell was nonresponsive to SON stimulation and two other unidentified ARC cells were also nonresponsive. A further two cells were inhibited and seven excited, after stimulation of the SON. No significant differences were detected comparing the distribution of ARC and VMH response categories (chi-square test). Neither were any differences recorded between these groups for the onset or offset latencies of excitatory or inhibitory responses (t test). The response latencies are shown in Table 1 and the principal findings are also summarized in Fig. 1. DISCUSSION This study demonstrated several points that may assist in the interpretation of certain neuroendocrine observations. The ARC and VMH have been demonstrated as containing cell bodies that synthesize dopamine (3). This catecholamine is widely believed to be the primary prolactin-inhibiting factor (32) and is also known to stimulate the release of oxytocin from rat hypothalami TABLE 1 Responses of Arcuate (ARC) and Ventromedial Hypothalamic (VMH) Nucleus Neurons to Stimulation of the Ipsilateral Supraoptic Nucleus (SON) in Male Rats’
N(s)
Onset (ms)
Offset (ms)
ARC cells Orthodromic inhibition Orthodromic excitation
2 (17) 7 (58)
0.0 f 18.4 f
0.0 3.5
35.0 f 5.0 30.7 k 7.8
VMH Cells Orthodromic inhibition Orthodromic excitation Antidromic invasion
6 (17) 13 (36) 6 (17)
11.6+_ 11.5 19.2 _+ 3.9 10.6 f 4.1
115.8 f 25.5 52.6 k 17.1 -
’ Values are X + SE. Not included in the table are three ARC neurons (25%) and 13 VMH neurons (36%) that were nonresponsive to the stimulation. Onset of inhibitory responses in the ARC were difficult to determine accurately due to the slow firing rate of the two cells examined.
SUPRAOPTIC-MEDIOBASAL
HYPOTHALAMIC
CONNECTIONS
567
il.2 Hz
FIG. 1. This figure illustrates the experimental set-up for the recording of arcuate/ventromedial hypothalamic nucleus (ARC/VMH) single units and their responses to electrical stimulation of the ipsilateral supraoptic nucleus (SON). Stimulation of the median eminence (ME) for the antidromic identification of tuberoinfundibular neurons was also included in the experimental protocol, although the stimulating electrode for this is not illustrated. Insets a, and an show, respectively, oscilloscope photographs of constant latency and collision tests for the antidromic identification of the ARC/VMH cell shown; b-responding with orthodromic excitation following stimulation of the SON (timebar (a), 5 ms; timebar (b), 10 ms). c-an inhibitory response of an unidentified ARC/VMH cell following stimulation of the SON (timebar, IO ms). d-the antidromic invasion of an ARC/VMH cell following stimulation of the SON (timebar, 5 ms). Each inset represents a minimum of IO superimposed stimulus-triggered sweeps of the oscilloscope. See text for details of stimulation parameters. Abbreviations: OC-optic chiasm, da?-putative dopamine-secreting neuron. ap-anterior pituitary, pp-posterior pituitary, prl-prolactin. oxy-oxytocin, vp-vasopressin.
incubated in vitro (4, 22) as well as increasing unit discharge of putative oxytocin-secreting cells of the SON (20). In view of the fact that substantial proportions of ARC/VMH cells are probably dopaminergic (2, 3, 8, 28), it is likely that a number of the cells recorded in the present study synthesized and secreted this catecholamine. However, as increased prolactin secretion is believed to be a function of decreased tuberoinfundibular dopaminergic activity (8,28) it would not be expected for those ARC/VMH neurons identified
568
SAPHIER
AND
FELDMAN
as projecting to the ME to also be antidromically identified as projecting to (oxytocinergic cells of) the SON. Indeed, such cells were not found, the only ARC/VMH cells identified as projecting to the SON being those not belonging to the tuberoinfundibular system. Some of these cells may have belonged to the A14 system of incertohypothalamic dopaminergic neurons (2), or some other system affecting oxytocin secretion, such as one synthesizing the endogenous opioid peptides (1,9,24). Furthermore, the majority of ARC/VMH cells were orthodromically excited after SON stimulation, perhaps suggesting a facilitatory action of oxytocin on those cells. Such an action would thus increase their secretory activity and decrease pituitary prolactin secretion. In this situation, an effective dissociation of oxytocin and prolactin release could occur, as has been reported (27). Some of the ARC/VMH neurons identified as projecting to the SON may also have been osmosensitive ( 17), providing an afferent input to vasopressinsecreting cells of the SON. Such cells are known to respond principally to changes in osmotic pressure (26) and some lines of evidence suggest that afferent inputs may be of importance in the regulation of their secretory activity (25). Because the ARC/VHM region is known to be involved in the neural regulation of growth hormone secretion (19), a proportion of the identified tuberoinfundibular neurons may have been involved in such regulation. These may be the VMH cells excited by stimulation of the SON. This is in view of the fact that vasopressin has been shown to release growth hormone in the rat ( 16). Such release, stimulated by vasopressin, is inhibited by pentobarbitone anesthesia (13, 16). Pentobarbital and urethane anesthesia both inhibit in vitro pituitary hormone release in response to luteinzing hormone-releasing hormone (5), and urethane does not inhibit the spontaneous activity of neurons in rat diencephalic islands (6). It therefore seems possible that our results represent a neurophysiologic substrate underlying the demonstrated interrelationships between growth hormone and vasopressin secretion ( 13, 16) assuming a similar action of the anesthesia on pituitary somatotrophs. The nature of the relationship between vasopressin and growth hormone secretion remains incompletely answered but may be related to the possible effects of central vasopressin upon sleep-wake rhythms (30) and the association of such rhythms with growth hormone secretion ( 19). The results of the present study suggest that direct interactions may exist. A further possibility for the function of some of the neurons recorded may be to act as interneurons mediating communication, demonstrated electrophysiologically between the SON and PVN (29). Anatomical studies have shown that the ARC/VMH receives afferent projections from the SON ( 18) and the PVN has been demonstrated as having connections with the ARC/ VMH (11). Thus, complex intrahypothalamic circuitry may serve to provide
SUPRAOPTIC-MEDIOBASAL
HYPOTHALAMIC
CONNECTIONS
569
interactions in the multitude of mechanisms that are apparent in the regulation of anterior and posterior pituitary hormone secretion. REFERENCES 1. BEAUMONT, A., AND J. HUGHES. 1979. Biology of opioid peptides. Annzc. Rev. Pharmacol. To.xicol.
19: 245-267.
2. BJ~RKLUND, A., 0. LINDVALL. AND R. NOBIN. 1975. Evidence of an incertohypothalamic dopamine neurones system in the rat. Brain Rex 89: 29-42. 3. BJ~RKLUND, A., R. Y. MOORE, A. NOBIN, AND U. STENEVI. 1973. The organization of tubero-hypophyseal and reticula-infundibular catecholamine neuron systems in the rat brain. Brain Res. 51: 171-181. 4. BRIDGES, T. E., E. W. HILLHOUSE. AND M. T. JONES. 1976. The effect of dopamine on neurohypophysial hormone release in vivo and from the rat neural lobe and hypothalamus in vitro. J. PhyGol. (London) 260: 647-666. 5. CARTER, D. A.. AND R. G. DYER. 1979. Inhibition by pentobarbitone and urethane of the in vitro response of the adenohypophysis to luteinising hormone-releasing hormone in male rats. Br. J. Pharmacol. 67: 277-28 1. 6. CROSS.B. A.. AND R. G. DYER. 197 1. Unit activity in rat diencephalic islands-the effect of anaesthetics. J. Ph)siol. (London) 212: 467-48 I. 7. DREIFUSS,J. J.. AND K. RUF. 1971. A transpharyngeal approach to the rat hypothalamus. Pages 2 13-228 in G. A. KERBUT, Ed., Experiments in Physiology and Biochemisfry. Vol. 5. Academic Press, New York./London. 8. DYER, R. G., AND D. J. SAPHIER. 1981. Electrical activity of antidromically identified tuberoinfundibular neurones during stimulated release of luteinizing hormone and prolactin in pro-oestrous rats. J. Endocrinol. 89: 35-44. 9. HALDAR, J.. D. L. HOFFMAN. AND E. A. ZIMMERMAN. 1979. Action of beta-endorphin on oxytocin release. Fed. Proc. 38: 270. IO. HATTON, G. I., P. COBBETT. AND A. K. SALM. 1985. Extranuclear axon collaterals of paraventricular neurons in the rat hypothalamus: intracellular staining. immunocflochemistry and electrophysiology. Brain Rex Bull. 14: 123- 132. 1 I. HARRIS, M. C.. AND M. SANGHERA. 1974. Projection of medial basal hypothalamic neurones to the preoptic anterior hypothalamic area and the paraventricular nucleus in the rat, Brain Rex 81: 401-41 I. 12. HELLER. H., AND M. GINSBURG. 1966. Secretion. metabolism and role of the posterior pituitary hormones. Pages 330-373 in G. W. HARRIS. AND B. T. DONOVAN, Eds. The Pit2&zrJJ Gland. Vol. 3. Butterworths, London. 13. KNOBIL, E.. V. MEYER, AND A. V. SCHALLY. 1968. Hypothalamic extracts and the secretion of growth hormone in the rhesus monkey. Pages 226-237 in A. PECILE, AND E. MOLLER Eds., Growlh Hormone. Excerpta Medica, Amsterdam. 14. LINCOLN, D. W., AND J. B. WAKERLEY. 1974. Electrophysiological evidence for the activation of supraoptic neurones during the release of oxytocin. J. Physiol. (London) 242: 533-554. 15. LINCOLN, D. W., AND J. B. WAKERLEY. 1975. Factors governing the periodic activation of supraoptic and paraventricular neurosecretory cells during suckling in the rat. J. Physiol. (London) 250: 443-46 1. 16. MALACARA. J. M.. R. VALVERDE, AND S. REICHLIN. 1972. Elevation of plasma radioimmunoassayable growth hormone in the rat induced by porcine hypothalamic extract. Endocrinologv 91: 1189-I 198. 17. MANDELBROD. I., S. FELDMAN, AND R. WERMAN. 1983. Mediobasal hypothalamic neurons are excited by the iontophoretic application of sodium. Brain Res. 273: 35-44.
570
SAPHIER AND FELDMAN
MARANI, E., W. J. RIETVELD, ANDP. G. M. LUITEN. 1984. Interconnections ofthe mediobasal hypothalamus nuclei in the adult rat. Neurosci. Left. (Suppl.) 18: S62. 19. MARTIN, J. B. 1976. Brain regulation of growth hormone secretion. Pages 129-168 in L. MARTINI AND W. F. GANONG, Eds., Frontiers in Neuroendocrinology, Vol. 4. Raven Press, New York. 20. MASON, W. T. 1983. Excitation by dopamine of putative oxytocinergic neurons in the rat supraoptic nucleus in vitro: evidence for two classesof continuously firing neurons. Brain Res. 267: 113-121. 2 1. MASON, W. T., Y. W. Ho, AND G. I. HATTON. 1984. Axon collaterals of supraoptic neurons: anatomical and electrophysiological evidence for their existence in the lateral hypothalamus. 18.
Neuroscience 22. 23.
24. 25. 26.
11: 169-182.
Moos, F., ANDP. RICHARD. 1982. Excitatory effectof dopamine on oxytocin and vasopressin reflex releases in the rat. Brain Res. 241: 247-260. NEILL. J. D. 1980. Neuroendocrine regulation of prolactin secretion. Pages 129-155 in L. MARTINI AND W. F. GANONG, Eds., Frontiers in Neuroendocrinology, Vol. 6. Raven Press, New York. PALKOVITS, M. 1984. Distribution of neuropeptides in the central nervous system: a review of biochemical mapping studies. Prog. Neurobiol. 23: 15 1- 189. POULAIN, D. A., ANDJ. B. WAKERLEY. 1982. Electrophysiology ofhypothalamic magnocegular neurones secreting oxytocin and vasopressin. Neuroscience 7: 773-808. POULAIN, D. A., J. B. WAKERLEY, AND R. E. J. DYBALL. 1977. Electrophysiological differentiation of oxytocin and vasopressin secreting neurons. Proc. Roy. Sot. Land., Ser. B. 196: 367-384.
SAMSON,W. K.. J. K. MCDONALD, AND M. S. LUMPKIN. 1985. Naloxone-induced dissociation of oxytocin and prolactin release. Neuroendocrinology 40: 68-7 1. 28. SAPHIER,D. J., AND R. G. DYER. 198 1, Effectsof neonatal exposure to monosodium glutamate on the electrical activity of neurones in the mediobasal hypothalamus, and on the plasma concentrations of thyroid-stimulating hormone and prolactin, following stimulation of the rostra1 hypothalamus in adult female rats. J. Endocrinol. 89: 379-387. 29. SAPHIER, D., AND S. FELDMAN. 1985. Electrophysiologic evidence for neural connections between the paraventricular nucleus and neurons of the supraoptic nucleus in the rat. 27.
Exp. Neural.
89: 289-294.
30. URBAN, I. J. A. I98 1. Intraseptal administration of vasopressin and oxytocin affects hippocampal electroencephalogram in rats. Exp. Neurol. 74: 13 l- 147. 3 1. VOOGT, J. L., AND L. A. CAM. 1974. Plasma prolactin levels and hypothalamic catecholamine synthesis during suckling. Neuroendocrinology 16: 108-l 18. 32. WEINER, R. I., M. K. SELMANOFF, C. Y. CHEUNG, M. J. CRONIN, AND P. C. GOLDSMITH. 1978. Brain catecholamines and control of LH and prolactin secretion in the rat. Pages 101-l 15 in J. D. VINCENT AND C. KORDON, Eds., Cell Biology of Hypothalamic Neurosecretion. CNRS, Paris.
92,563-570
NEUROLOGY
(1986)
Electrophysiologic Evidence for Connections between the Supraoptic and the Arcuate/Ventromedial Hypothalamic Nuclei in the Rat DAVID Laboratory
SAPHIER
AND SHAUL FELDMAN’
of Neurophysiology. Department of Neurology. Hebrew University-Hadassah McJdical School, Received
September
6. 1985; revision
Hadassah Jerusalem
received
January
University Hospital 91120. Israel
and
6. 1986
Extracellular action potentials were recorded from 48 single units located in the hypothalamic arcuate and ventromedial nuclei. Fifteen percent ofthe cells were identified as projecting to the median eminence and some of these cells may have belonged to the tuberoinfundibular dopaminergic systems.Responses of all cells to stimulation of the ipsilateral supraoptic nucleus were recorded, 17% of ventromedial nucleus neurons were antidromically identified as projecting to the supraoptic nucleus. None of the latter cells was also identified as projecting to the median eminence. Three of six identified tuberoinfundibular and eight unidentified ventromedial nucleus cells were found to be excited by stimulation of the supraoptic nucleus. One arcuate cell identified as projecting to the median eminence was nonresponsive to supraoptic stimulation. Orthodromic inhibitory responses were recorded from 17% of all cells recorded but no inhibitory responses were recorded from cells identified as projecting to the median eminence. We suggestthat these results may provide some neurophysiologic explanations for the observed interrelationships between oxytocin and prolactin secretion, and between vasopressin and growth hormone secretion. 0 1986 Academic press, Inc. INTRODUCTION
The supraoptic nucleus (SON) contains magnocellular cells of the neurohypophyseal neurosecretory system that synthesize oxytocin and vasopressin (12). These hormones are released in response to the suckling stimulus and elevation of plasma osmotic pressure, respectively. The neural control of their Abbreviations: SON-supraoptic nucleus, ARC/VMH-arcuate/ventromedial nuclei, PVNparaventricular nucleus, ME-median eminence. ’ This research was supported by the Lena P. Harvey Endowment Fund for Neurological Research. 563 0014-4886/86 $3.00 Copyright Q 1986 by Academic Press. Inc. All rights of reproduction in any form reserved.
564
SAPHIER AND FELDMAN
activation has been the subject of many studies and has been reviewed extensively (14, 15, 25, 26). Anatomical evidence has demonstrated that substantial interconnections exist between the SON and adjacent nuclei of the mediobasal hypothalamus ( 18). Magnocellular neurons of the SON and paraventricular nucleus (PVN) have also been shown electrophysiologically to send axon collaterals to these other regions of the endocrine hypothalamus (10, 2 1). This suggests interactions between neurons controlling neurohypophyseal secretory activity and those in the tuberal hypothalamus concerned with the regulation of anterior pituitary secretory activity. The suckling stimulus is also known to cause an elevation of plasma prolactin (3 1), believed to be primarily under inhibitory control by dopaminesynthesizing and -secreting tuberoinfundibular neurons in the arcuate and ventromedial nuclei (ARC/VMH) of the mediobasal hypothalamus (3, 8,23, 32). Dopamine is also known to stimulate the release of oxytocin (4, 22) and increase the firing of putative oxytocin-secreting cells of the SON (20). Thus it is possible that any physiologic release of oxytocin mediated by dopamine could arise from dopaminergic tuberoinfundibular neuron axon collaterals or from other dopaminergic, or other systems (1,2,24). As it has been shown that more than one-half of mediobasal hypothalamic neurons are excited by the iontophoretic application of sodium ions (17) it is possible that these neurons may be involved in osmoregulatory processes. Such processes are primarily regulated by vasopressin-secreting cells of the SON and PVN (12, 26), but local neural circuits may also be of importance (10, 11, 18, 2 1, 29). In view of the above considerations, we examined any possible connections between the SON and ARC/VMH. METHODS Male albino rats (body weight, 260 to 320 g) of the Hebrew University strain were anesthetized with a single injection of urethane (ethyl carbamate: 1.2 g/kg, i.p. as a 25% W/V solution). The animals were then secured in a stereotaxic frame, and a concentric bipolar stimulating electrode was lowered into the SON from a dorsal aspect. This electrode consisted of an outer pole of 500-pm diameter insulated to within 300 pm of the tip. The inner pole was 120-pm-diameter wire insulated to within approximately 150 pm of the tip and separated from the cathode by the same distance. The apparatus was subsequently inverted and the median eminence and pituitary stalk exposed according to a modification of the technique of Dreifuss and Ruf (7). A nonconcentric bipolar stimulating electrode, consisting of two 120~pm-diameter Nichrome wires insulated to within 700 pm of the tip and separated by approximately 1 mm, was lowered and rested upon the median eminence (ME).
SUPRAOPTIC-MEDIOBASAL
HYPOTHALAMIC
CONNECTIONS
565
This electrode allowed for the antidromic identification of tuberoinfundibular neurosecretory cells (8). Extracellular recordings were made from single cells located within the ARC/VMH. All recording sites were ipsilateral with the stimulating electrode in the SON and were made through glass micropipets filled with a solution of 0.5 M sodium acetate in 2 M sodium chloride and containing fast green dye. Single-unit signals were amplified through a Grass Instruments (model P5 11) preamplifier and monitored with a Grass Instruments audio monitor, using conventional and storage oscilloscopes. Stimuli were delivered to the ME as single (0.2-Hz), bipolar square-wave pulses of 1-ms total duration and up to I-mA peak-to-peak current strength. The same parameters were used to stimulate the SON, except that the current strength was 350 to 750 /IA. Responses of the cells were recorded as antidromic if the antidromic action potential occurred after the stimulus at a constant latency, and if the cell was able to follow trains of repeated stimuli at frequencies greater than 150 Hz. A third criterion, that could be tested only for spontaneously active neurons, was the presence of collision between spontaneous action potentials and antidromically evoked action potentials. Orthodromic excitation was distinguished by the appearance of evoked action potentials of variable latency, and orthodromic inhibition by the observation of extended periods of silence following the stimulation that exceeded the unstimulated interspike interval mode. At the termination of each experiment the final site of recording was marked by ejection of the green dye from the tip of the electrode. Previous recording sites could then be estimated according to the calibrated movements of the hydraulic microdrive. Histological verification of the electrode sites was by examination of 45-pm frozen sections cut from the perfused (10% Formalinsaline) and postfixed tissue. In all animals the stimulating electrode track was seen to terminate in the SON. RESULTS Extracellular action potentials were recorded from a total of 48 single units, of which 12 were located within the ARC and the remaining 36 within the VMH. The cells were located at mean depths from the median eminence of 0.40 + 0.04 and 1.15 f 0.07 mm, respectively. The ARC/VMH division was defined at 0.7 mm above the ME. Firing rates were not found to be significantly different, although ARC cells were found to fire somewhat more slowly at 1.4 f 0.5 Hz (X + SE), with VMH cells firing at 2.0 2 0.7 Hz. Six VMH cells (17%) were antidromically identified as projecting to the SON, exhibiting a mean antidromic invasion latency of 10.6 f 4.1 ms; None of those cells could be antidromically invaded following stimulation of the
566
SAPHIER AND FELDMAN
ME. Stimulation of the SON evoked orthodromic, i.e., synaptic, responses from 56% of the VMH neurons. Of the responsive VMH cells, 11 (31%) showed orthodromic excitation following SON stimulation and 3 of these cells were identified as projecting to the ME. Six nonidentified cells were inhibited by the stimulation and the remaining 13 were nonresponsive. Of the nonresponsive VMH population, three were identified as projecting to the ME. Orthodromic responses were recorded from nine (75%) of the ARC cells and none of these cells was identified as projecting to the ME. One identified ARC cell was nonresponsive to SON stimulation and two other unidentified ARC cells were also nonresponsive. A further two cells were inhibited and seven excited, after stimulation of the SON. No significant differences were detected comparing the distribution of ARC and VMH response categories (chi-square test). Neither were any differences recorded between these groups for the onset or offset latencies of excitatory or inhibitory responses (t test). The response latencies are shown in Table 1 and the principal findings are also summarized in Fig. 1. DISCUSSION This study demonstrated several points that may assist in the interpretation of certain neuroendocrine observations. The ARC and VMH have been demonstrated as containing cell bodies that synthesize dopamine (3). This catecholamine is widely believed to be the primary prolactin-inhibiting factor (32) and is also known to stimulate the release of oxytocin from rat hypothalami TABLE 1 Responses of Arcuate (ARC) and Ventromedial Hypothalamic (VMH) Nucleus Neurons to Stimulation of the Ipsilateral Supraoptic Nucleus (SON) in Male Rats’
N(s)
Onset (ms)
Offset (ms)
ARC cells Orthodromic inhibition Orthodromic excitation
2 (17) 7 (58)
0.0 f 18.4 f
0.0 3.5
35.0 f 5.0 30.7 k 7.8
VMH Cells Orthodromic inhibition Orthodromic excitation Antidromic invasion
6 (17) 13 (36) 6 (17)
11.6+_ 11.5 19.2 _+ 3.9 10.6 f 4.1
115.8 f 25.5 52.6 k 17.1 -
’ Values are X + SE. Not included in the table are three ARC neurons (25%) and 13 VMH neurons (36%) that were nonresponsive to the stimulation. Onset of inhibitory responses in the ARC were difficult to determine accurately due to the slow firing rate of the two cells examined.
SUPRAOPTIC-MEDIOBASAL
HYPOTHALAMIC
CONNECTIONS
567
il.2 Hz
FIG. 1. This figure illustrates the experimental set-up for the recording of arcuate/ventromedial hypothalamic nucleus (ARC/VMH) single units and their responses to electrical stimulation of the ipsilateral supraoptic nucleus (SON). Stimulation of the median eminence (ME) for the antidromic identification of tuberoinfundibular neurons was also included in the experimental protocol, although the stimulating electrode for this is not illustrated. Insets a, and an show, respectively, oscilloscope photographs of constant latency and collision tests for the antidromic identification of the ARC/VMH cell shown; b-responding with orthodromic excitation following stimulation of the SON (timebar (a), 5 ms; timebar (b), 10 ms). c-an inhibitory response of an unidentified ARC/VMH cell following stimulation of the SON (timebar, IO ms). d-the antidromic invasion of an ARC/VMH cell following stimulation of the SON (timebar, 5 ms). Each inset represents a minimum of IO superimposed stimulus-triggered sweeps of the oscilloscope. See text for details of stimulation parameters. Abbreviations: OC-optic chiasm, da?-putative dopamine-secreting neuron. ap-anterior pituitary, pp-posterior pituitary, prl-prolactin. oxy-oxytocin, vp-vasopressin.
incubated in vitro (4, 22) as well as increasing unit discharge of putative oxytocin-secreting cells of the SON (20). In view of the fact that substantial proportions of ARC/VMH cells are probably dopaminergic (2, 3, 8, 28), it is likely that a number of the cells recorded in the present study synthesized and secreted this catecholamine. However, as increased prolactin secretion is believed to be a function of decreased tuberoinfundibular dopaminergic activity (8,28) it would not be expected for those ARC/VMH neurons identified
568
SAPHIER
AND
FELDMAN
as projecting to the ME to also be antidromically identified as projecting to (oxytocinergic cells of) the SON. Indeed, such cells were not found, the only ARC/VMH cells identified as projecting to the SON being those not belonging to the tuberoinfundibular system. Some of these cells may have belonged to the A14 system of incertohypothalamic dopaminergic neurons (2), or some other system affecting oxytocin secretion, such as one synthesizing the endogenous opioid peptides (1,9,24). Furthermore, the majority of ARC/VMH cells were orthodromically excited after SON stimulation, perhaps suggesting a facilitatory action of oxytocin on those cells. Such an action would thus increase their secretory activity and decrease pituitary prolactin secretion. In this situation, an effective dissociation of oxytocin and prolactin release could occur, as has been reported (27). Some of the ARC/VMH neurons identified as projecting to the SON may also have been osmosensitive ( 17), providing an afferent input to vasopressinsecreting cells of the SON. Such cells are known to respond principally to changes in osmotic pressure (26) and some lines of evidence suggest that afferent inputs may be of importance in the regulation of their secretory activity (25). Because the ARC/VHM region is known to be involved in the neural regulation of growth hormone secretion (19), a proportion of the identified tuberoinfundibular neurons may have been involved in such regulation. These may be the VMH cells excited by stimulation of the SON. This is in view of the fact that vasopressin has been shown to release growth hormone in the rat ( 16). Such release, stimulated by vasopressin, is inhibited by pentobarbitone anesthesia (13, 16). Pentobarbital and urethane anesthesia both inhibit in vitro pituitary hormone release in response to luteinzing hormone-releasing hormone (5), and urethane does not inhibit the spontaneous activity of neurons in rat diencephalic islands (6). It therefore seems possible that our results represent a neurophysiologic substrate underlying the demonstrated interrelationships between growth hormone and vasopressin secretion ( 13, 16) assuming a similar action of the anesthesia on pituitary somatotrophs. The nature of the relationship between vasopressin and growth hormone secretion remains incompletely answered but may be related to the possible effects of central vasopressin upon sleep-wake rhythms (30) and the association of such rhythms with growth hormone secretion ( 19). The results of the present study suggest that direct interactions may exist. A further possibility for the function of some of the neurons recorded may be to act as interneurons mediating communication, demonstrated electrophysiologically between the SON and PVN (29). Anatomical studies have shown that the ARC/VMH receives afferent projections from the SON ( 18) and the PVN has been demonstrated as having connections with the ARC/ VMH (11). Thus, complex intrahypothalamic circuitry may serve to provide
SUPRAOPTIC-MEDIOBASAL
HYPOTHALAMIC
CONNECTIONS
569
interactions in the multitude of mechanisms that are apparent in the regulation of anterior and posterior pituitary hormone secretion. REFERENCES 1. BEAUMONT, A., AND J. HUGHES. 1979. Biology of opioid peptides. Annzc. Rev. Pharmacol. To.xicol.
19: 245-267.
2. BJ~RKLUND, A., 0. LINDVALL. AND R. NOBIN. 1975. Evidence of an incertohypothalamic dopamine neurones system in the rat. Brain Rex 89: 29-42. 3. BJ~RKLUND, A., R. Y. MOORE, A. NOBIN, AND U. STENEVI. 1973. The organization of tubero-hypophyseal and reticula-infundibular catecholamine neuron systems in the rat brain. Brain Res. 51: 171-181. 4. BRIDGES, T. E., E. W. HILLHOUSE. AND M. T. JONES. 1976. The effect of dopamine on neurohypophysial hormone release in vivo and from the rat neural lobe and hypothalamus in vitro. J. PhyGol. (London) 260: 647-666. 5. CARTER, D. A.. AND R. G. DYER. 1979. Inhibition by pentobarbitone and urethane of the in vitro response of the adenohypophysis to luteinising hormone-releasing hormone in male rats. Br. J. Pharmacol. 67: 277-28 1. 6. CROSS.B. A.. AND R. G. DYER. 197 1. Unit activity in rat diencephalic islands-the effect of anaesthetics. J. Ph)siol. (London) 212: 467-48 I. 7. DREIFUSS,J. J.. AND K. RUF. 1971. A transpharyngeal approach to the rat hypothalamus. Pages 2 13-228 in G. A. KERBUT, Ed., Experiments in Physiology and Biochemisfry. Vol. 5. Academic Press, New York./London. 8. DYER, R. G., AND D. J. SAPHIER. 1981. Electrical activity of antidromically identified tuberoinfundibular neurones during stimulated release of luteinizing hormone and prolactin in pro-oestrous rats. J. Endocrinol. 89: 35-44. 9. HALDAR, J.. D. L. HOFFMAN. AND E. A. ZIMMERMAN. 1979. Action of beta-endorphin on oxytocin release. Fed. Proc. 38: 270. IO. HATTON, G. I., P. COBBETT. AND A. K. SALM. 1985. Extranuclear axon collaterals of paraventricular neurons in the rat hypothalamus: intracellular staining. immunocflochemistry and electrophysiology. Brain Rex Bull. 14: 123- 132. 1 I. HARRIS, M. C.. AND M. SANGHERA. 1974. Projection of medial basal hypothalamic neurones to the preoptic anterior hypothalamic area and the paraventricular nucleus in the rat, Brain Rex 81: 401-41 I. 12. HELLER. H., AND M. GINSBURG. 1966. Secretion. metabolism and role of the posterior pituitary hormones. Pages 330-373 in G. W. HARRIS. AND B. T. DONOVAN, Eds. The Pit2&zrJJ Gland. Vol. 3. Butterworths, London. 13. KNOBIL, E.. V. MEYER, AND A. V. SCHALLY. 1968. Hypothalamic extracts and the secretion of growth hormone in the rhesus monkey. Pages 226-237 in A. PECILE, AND E. MOLLER Eds., Growlh Hormone. Excerpta Medica, Amsterdam. 14. LINCOLN, D. W., AND J. B. WAKERLEY. 1974. Electrophysiological evidence for the activation of supraoptic neurones during the release of oxytocin. J. Physiol. (London) 242: 533-554. 15. LINCOLN, D. W., AND J. B. WAKERLEY. 1975. Factors governing the periodic activation of supraoptic and paraventricular neurosecretory cells during suckling in the rat. J. Physiol. (London) 250: 443-46 1. 16. MALACARA. J. M.. R. VALVERDE, AND S. REICHLIN. 1972. Elevation of plasma radioimmunoassayable growth hormone in the rat induced by porcine hypothalamic extract. Endocrinologv 91: 1189-I 198. 17. MANDELBROD. I., S. FELDMAN, AND R. WERMAN. 1983. Mediobasal hypothalamic neurons are excited by the iontophoretic application of sodium. Brain Res. 273: 35-44.
570
SAPHIER AND FELDMAN
MARANI, E., W. J. RIETVELD, ANDP. G. M. LUITEN. 1984. Interconnections ofthe mediobasal hypothalamus nuclei in the adult rat. Neurosci. Left. (Suppl.) 18: S62. 19. MARTIN, J. B. 1976. Brain regulation of growth hormone secretion. Pages 129-168 in L. MARTINI AND W. F. GANONG, Eds., Frontiers in Neuroendocrinology, Vol. 4. Raven Press, New York. 20. MASON, W. T. 1983. Excitation by dopamine of putative oxytocinergic neurons in the rat supraoptic nucleus in vitro: evidence for two classesof continuously firing neurons. Brain Res. 267: 113-121. 2 1. MASON, W. T., Y. W. Ho, AND G. I. HATTON. 1984. Axon collaterals of supraoptic neurons: anatomical and electrophysiological evidence for their existence in the lateral hypothalamus. 18.
Neuroscience 22. 23.
24. 25. 26.
11: 169-182.
Moos, F., ANDP. RICHARD. 1982. Excitatory effectof dopamine on oxytocin and vasopressin reflex releases in the rat. Brain Res. 241: 247-260. NEILL. J. D. 1980. Neuroendocrine regulation of prolactin secretion. Pages 129-155 in L. MARTINI AND W. F. GANONG, Eds., Frontiers in Neuroendocrinology, Vol. 6. Raven Press, New York. PALKOVITS, M. 1984. Distribution of neuropeptides in the central nervous system: a review of biochemical mapping studies. Prog. Neurobiol. 23: 15 1- 189. POULAIN, D. A., ANDJ. B. WAKERLEY. 1982. Electrophysiology ofhypothalamic magnocegular neurones secreting oxytocin and vasopressin. Neuroscience 7: 773-808. POULAIN, D. A., J. B. WAKERLEY, AND R. E. J. DYBALL. 1977. Electrophysiological differentiation of oxytocin and vasopressin secreting neurons. Proc. Roy. Sot. Land., Ser. B. 196: 367-384.
SAMSON,W. K.. J. K. MCDONALD, AND M. S. LUMPKIN. 1985. Naloxone-induced dissociation of oxytocin and prolactin release. Neuroendocrinology 40: 68-7 1. 28. SAPHIER,D. J., AND R. G. DYER. 198 1, Effectsof neonatal exposure to monosodium glutamate on the electrical activity of neurones in the mediobasal hypothalamus, and on the plasma concentrations of thyroid-stimulating hormone and prolactin, following stimulation of the rostra1 hypothalamus in adult female rats. J. Endocrinol. 89: 379-387. 29. SAPHIER, D., AND S. FELDMAN. 1985. Electrophysiologic evidence for neural connections between the paraventricular nucleus and neurons of the supraoptic nucleus in the rat. 27.
Exp. Neural.
89: 289-294.
30. URBAN, I. J. A. I98 1. Intraseptal administration of vasopressin and oxytocin affects hippocampal electroencephalogram in rats. Exp. Neurol. 74: 13 l- 147. 3 1. VOOGT, J. L., AND L. A. CAM. 1974. Plasma prolactin levels and hypothalamic catecholamine synthesis during suckling. Neuroendocrinology 16: 108-l 18. 32. WEINER, R. I., M. K. SELMANOFF, C. Y. CHEUNG, M. J. CRONIN, AND P. C. GOLDSMITH. 1978. Brain catecholamines and control of LH and prolactin secretion in the rat. Pages 101-l 15 in J. D. VINCENT AND C. KORDON, Eds., Cell Biology of Hypothalamic Neurosecretion. CNRS, Paris.