Ibotenate-induced cell death in the hypothalamic paraventricular nucleus: differential susceptibility of magnocellular and parvicellular neurons

Brain Research, 383 (1986) 367-372 Elsevier

367

BRE 21798

Ibotenate-induced cell death in the hypothalamic paraventricular nucleus: differential susceptibility of magnocellular and parvicellular neurons JAMES P. HERMAN 1'2and STANLEY J. WIEGAND1 1Department of Neurobiology and Anatomy and the 2Centerfor Brain Research, University of Rochester, School of Medicine and Dentistry, Rochester, NY 14642 (U.S.A.) (Accepted 3 June 1986) Key words: Ibotenic acid - - Magnocellular neuron - - Oxytocin - - Paraventricular nucleus - - Parvicellular neuron - - Vasopressin

Injections of ibotenic acid (IBO) (0.5-2.0/~g) were made unilaterally into the hypothalamic paraventricular nucleus (PVN) of Sprague-Dawley rats. A distinct loss of parvicellular neurons was evident at all dosages examined, and cell death appeared uniform throughout the injection region at doses of 1.0/~g or greater. In contrast, magnocellular neurons were present following injections of all dosages, with a slight loss of neurons evident only at 2.0/~g doses. Magnocellular neurons killed at 2.0/~g doses were primarily vasopressinergic. Neuropeptide content per se was not the critical determinant of susceptibility to IBO, as parvicellular oxytocin neurons were destroyed by low to moderate doses of IBO. The results suggest that IBO may be a useful neurotoxin with which to investigate differential connectional and functional properties of parvicellular and magnocellular PVN neurons.

The paraventricular nucleus (PVN) is a complex hypothalamic structure implicated in the regulation of numerous central integrative processes, particularly those related to autonomic and neuroendocrine function. Structurally, the P V N can be divided into a number of discrete subnuclei based on cytoarchitectonic and connectional criteria 1'1°. In recently proposed parcellations the P V N has been divided into as many as 8 identifiable subnuclei, including 3 magnocellular and 5 parvicellular divisions 1°. The magnocellular subnuclei contain primarily vasopressin and oxytocin immunoreactive cells which project to the posterior pituitary gland and function in hormonal regulation of renal water resorption, vascular smooth muscle tone, milk ejection, and parturition 12. The parvicellular divisions project to median eminence (anterior, medial and periventricular subdivisions) n,13 or brainstem and spinal cord (lateral and dorsal subdivisions) l°'n and are believed to affect adenohypophysial and autonomic function, respectively 12. Accurate determination of functional properties of anatomically distinct divisions of this important

nucleus has thus far been confounded by the inability to make discrete lesions of these divisions without disrupting the neuronal organization of other subnuclei. Previous studies have suggested that the large neurosecretory neurons of the supraoptic nucleus and accessory magnocellular nuclei are resistant to excitatory neurotoxins, including kainic acid, ibotenic acid, quisqualic acid, and N-methyl aspartate 5"7,15. A resistance of PVN magnocellular neurons, specifically, has been reported following kainic acid injections 7'15. These data suggest that neurotoxic lesions may allow the dissociation of functional and connectional properties of magnocellular vs parvicellular PVN neurons. However, parametric data concerning the effect of neurotoxins on parvicellular subdivisions of PVN has been wanting. This report represents an examination of the relative toxicity of ibotenic acid on parvicellular and magnocellular PVN neurons, allowing us to determine whether excitatory neurotoxins may be an appropriate tool with which to selectively destroy specific populations of PVN neurons.

Correspondence: J.P, Herman, Department for Neurobiology and Anatomy, Box 603, University of Rochester, School of Medicine and Dentistry, Rochester, NY 14642, U.S.A. 0006-8993/86/$03.50 © 1986 Elsevier Science Publishers B .V. (Biomedical Division)

368 Subjects were male and female Sprague-Dawley rats weighing between 280-360 g at the time of surgery. Animals were anesthesized with chloral hydrate (0.6 ml/100 g b. wt. of a 7% solution) and mounted in a Kopf stereotaxic apparatus. The head was positioned with bregma and lambda leveled according to the coordinate system of Paxinos and Watson ~ and a burr hole drilled 1.8 mm posterior to bregma and 0.5-0.7 mm lateral to midline. The dura was carefully incised and a glass micropipette (10-30 ~m tip) containing ibotenic acid (IBO; Natural Products, Vashon, WA) dissolved in 50 mM sodium phosphatebuffered saline adjusted to pH 7.2 was lowered to a depth of 7.5 mm relative to the dural surface. Using a pressure injection system, 2.0, 1.0, 0.75 or 0.50/~g of IBO in a total volume of 100 nl was delivered into the region of the PVN over a period of l0 min. The pipette was left in place for an additional 10 min following completion of the injection to minimize reflux of the solution along the pipette track. All injections were unilateral, and were directed towards the central portion of the PVN at the level of the posterior magnocellular division of Swanson and Kuypers l~. Animals were allowed to survive for 21-28 days following surgery, at which time they were perfused with either 4% paraformaldehyde or Zamboni's fixative. Brains perfused with 4% paraformaldehyde were post-fixed for one week. Brains perfused with Zamboni's fixative were post-fixed overnight at 4 °C. Following post-fixation the brains were blocked in the coronal plane and placed in 30% sucrose-0.1 M sodium phosphate buffer at 4 °C for 12-48 h. Frozen sections were subsequently cut at 25/~m through the hypothalamus. Sections from rats perfused with 4% paraformaldehyde were divided into two series and processed for visualization of normal fibers using the silver stain of Spada, Diaz and de Olmos 2 and for Nissl substance using Cresyl violet. The brains of animals perfused with Zamboni's fixative were divided into four series and stained for Nissl substance using Cresyl violet, or processed for immunocytochemical localization of arginine-vasopressin and oxytocin. Rabbit antisera directed against arginine-vasopressin and oxytocin were obtained from Miles Laboratories and ImmunoNuclear, respectively. Vasopressin antiserum was utilized at a dilution of 1:4000 and anti-oxytocin at 1:2000. Absorption of the dilute antisera with 20 #g of homologous peptide completely

blocked oxytocin- and vasopressin-specific staining. Vasopressin and oxytocin antisera were cross-absorbed with oxytocin and arginine-vasopressin, rcspectively, prior to tissue incubation to obviate potential cross-reactivity. Immunohistochemistry was performed according to standard avidin:biotin:peroxidase (ABC-P) procedure, using biotinylated goat anti-rabbit y-globulin as a secondary antibody (Vector Labs). Representative Nissl-stained sections from an animal receiving 0.5 and 1.0/~g dose of IBO are presented in Fig. 1. It is clear that injection of IBO into the PVN results in extensive destruction of parvicellular PVN elements at the 1.0#g dose (Fig. lc and d). On the non-lesioned side (Fig. ld), parvicellular neurons are evident in the periventricular zone and in the medial and dorsal subdivisions of the nucleus. The lesioned side (Fig. lc) exhibits an almost total loss of neurons in the parvicellular subdivisions of the PVN and the adjacent anterior hypothalamus, along with a light to moderate infiltration of gtia and dilation of the third ventricle. In contrast, the majority of magnocellular neurons in this region of the PVN are clearly resistant to IBO at this dosage, with viable cells present at the center of the injection site. Injections centered on more anterior PVN loci reveal that resistance to IBO is evident in all magnocellular subdivisions at this dose. Injections of 2.0 #g IBO result in a more marked gliosis in the region of PVN and extensive ventricular dilation. Many magnocellular neurons are evident following injections of this dosage, and indeed appear to be the only viable neurons in the region of injection. However, the number of magnocellular neurons is clearly reduced. Injections of low doses of IBO (0.75 or 0.50#g) result in a more limited cell death in the PVN, with an obvious sparing of considerable numbers of parvicellular as well as virtually all magnocellular neurons at the lowest dose (Fig. la, b). The overall appearance of the injection site varies considerably with dose and location. Dimensions of lesions at the doses utilized are summarized in Table I. Generally, injection sites exhibit a roughly spherical zone of neuronal cell death encompassing a circumscribed central region of moderate gliosis surrounded by an irregularly-shaped peripheral zone of light gliosis. Injections of 2.0j~g of IBO into the PVN result in cell death and moderate gliosis in the adja-

369

Fig. 1. Representative Nissl-stained sections from animals receiving unilateral injection of 0.5 (a, b) and 1.0 (c, d)/~g IBO into the region of PVN, with injected sides on the left (a, c) and uninjected sides on the right (b, d). All photomicrographs are taken at the level of the posterior magnocellular division of PVN. a: injection of 0.5/~g IBO results in considerable but incomplete destruction of the parvicellular neurons within the PVN. Magnocellular neurons are clearly unaffected by IBO at this dosage, b: contralateral PVN, same animals as a. The divisions of PVN at this level are delineated in b. Dense aggregations of parvicellular neurons are present in the medial and dorsal parvicellular subdivisions, and numerous small neurons are present in the periventricular zone. c: injection of 1.0/~g IBO results in extensive cell death in all parvocellular divisions at this level of the PVN. Note the presence of increased numbers of glial cells throughout the nucleus, and a distinct proliferation of glia in the region of the micropipette tip (arrow). In contrast, the denselypacked magnocellular neurons of the posterior magnocellular division remain viable following injection of IBO. d: contralateral PVN from the same animal as c. PM, posterior magnocellular PVN; MP, medial parvicellular PVN; DP, dorsal parvicellular PVN; LPv, lateral parvicellular PVN; PVN, ventral portion; PZ, periventricular zone; III, third ventricle. TABLE I

c e n t a n t e r i o r h y p o t h a l a m u s a n d v e n t r a l t h a l a m u s , in-

Dimensions of ibotenic acid lesions

cluding large p o r t i o n s of t h e ipsilateral a n t e r i o r hy-

Mean medial-lateral and dorsal-ventral extent of ibotenic acid lesions centered on PVN, measured at the center of injection. Numbers in parentheses indicate number of animals at each dosage. Dose

Mediallateral (mm)

Dorsalventral (mm)

Patternof cell death

0.5/~g (2) 0.75/~g (2) 1.00 k~g (4)

0.57 0.45 0.73

0.63 0.95 1.04

2.00/~g (2)

0.95

1.43

irregular irregular uniform, except PVN magnocellular neurons uniform, except PVN magnocellular neurons

p o t h a l a m i c n u c l e u s , z o n a i n c e r t a , and t h e nucleus r e u n i e n s of t h a l a m u s . Cell d e a t h is e v i d e n t t h r o u g h out t h e r o s t r o c a u d a l e x t e n t o f the P V N r e g i o n at this dosage. I n j e c t i o n s of 1.0/~g I B O yield m o r e l o c a l i z e d cell d e a t h , largely c o n f i n e d to the P V N and its i m m e diate s u r r o u n d . I n t e r e s t i n g l y , the t h a l a m i c n u c l e u s r e u n i e n s a p p e a r s to be particularly susceptible to the n e u r o t o x i c actions of I B O . A d m i n i s t r a t i o n of 1.0/.tg I B O o f t e n results in c o n s i d e r a b l e cell d e a t h in nucleus r e u n i e n s e v e n w h e n it is s i t u a t e d at the p e r i p h ery of an i n j e c t i o n site c e n t e r e d in the P V N . In c o m p a r i s o n , regions of the a n t e r i o r h y p o t h a l a m u s at an

37O equal distance from the center of the injection show little cell loss. Injections of low doses of IBO {0.75 or 0.50/,g) result in distinct but non-uniform cell death among parvicellular neurons within the zone of the injection. Material stained with silver for visualization of normal fibers reveals that, as has been reported previously 3, IBO does not cause significant damage to fibers of passage. There is no obvious disruption of fibers in the fornix, the corticohypothalamic tract, or periventricular hypothalamic region despite injection of the largest dose of IBO (2.0/~g). Immunocytochemical localization of vasopressin and oxytocin reveals that neurons staining for both neurohypophysial peptides are evident in the magnocellular subdivisions of PVN following IBO lesions at

all doses utilized (Fig. 2). Similar numbers ot oxvn>cin and vasopressin neurons are present on rejected and uninjected sides following a dose of IB() sufficient to kill virtually all of the parvicellular neurons in the PVN (1.0 ug) (Fig. 2a-d). Quite striking is the observation that similar numbers of magnocellular oxytocin-containing neurons are present following all injection doses. However, while magnocellular vasopressin neurons are also present in all animals, their numbers are clearly reduced in animals receiving 2.0 ¢~g IBO (Fig. 3). The dorsal and lateral parvicellular subnuclei of the PVN normally contain a considerable number of parvicellular oxytocin-containmg neurons and a more modest quantity of vasopressinpositive elements. However, these regions are depleted of fusiform, oxytocin- and vasopressin-immu-

Fig. 2. Representative immunocytochemically-stained sections through the PVN of an animal receiving a 1.0/*g injection of IBO, demonstrating that magnocellular PVN neurons are resistant to IBO. Nissl sections from this animal are depicted in Fig. l a, b. a and b: posterior magnoceUular division of the PVN, stained for oxytocin. Numerous magnocellular oxytocin neurons are present on both the injected (a) and uninjected (b) sides, c and d: posterior magnocellular subdivision of PVN, stained for vasopressin. Level is slightly caudal to that of a and b. Similar numbers of magnocellular vasopressin neurons are present on both the injected (c) and uninjected (d) sides. This animal showed extensive cell loss in parvicellular PVN regions on alternate series stained for Nissl substance (Fig. la, b). III, third ventricle.

371

Fig. 3. Representative immunocytochemically-stained sections through the PVN of an animal receiving 2.0/ag IBO. a: posterior magnocellular division of PVN, stained for oxytocin. No difference in number of oxytocin-positive magnocellular neurons is evident between lesion (left) and non-lesion (right) sides, b: posterior magnocellular division of PVN, stained for vasopressin. Note the decrease in number of vasopressin neurons on the lesion (left) side. noreactive parvicellular n e u r o n s following injections of as little as 0.75 ~g I B O (Fig. 4), indicating that resistance to the neurotoxic action of IBO is specific to magnocellular n e u r o n s and not to cells containing these specific neuropeptides. The results of this experiment demonstrate that magnocellular n e u r o n s of the PVN are relatively resistant to the neurotoxic effects of IBO. This resistance is p r o n o u n c e d at doses of neurotoxin sufficient to cause nearly complete loss of parvicellular neurons, and becomes less marked at high doses, where some degree of cell death is observed in magnocellular populations. The magnocellular n e u r o n s eliminated by high doses of I B O are primarily vasopressi-

nergic. The reason for the specific and exceptional resistance of oxytocinergic magnocellular n e u r o n s to IBO is not known. N e u r o t r a n s m i t t e r expression does not seem to be the critical d e t e r m i n a n t of resistance, since parvicellular oxytocin n e u r o n s are eliminated by moderate doses of IBO (0.75-1.0~tg). In contrast, cells of the ventral thalamic area (nucleus reuniens) appear to be peculiarly sensitive to IBO. This region undergoes extensive cell loss even when located at the extreme periphery of injections of an intermediate dose. Previous evidence indicates that the magnocellular n e u r o n s of the PVN are insensitive to kainic acid 7'15 and N-methyl aspartate 7. Similarly, magnocellular

Fig. 4. Representative sections through the posterior PVN, demonstrating the susceptibility of lateral parvicellular neurons to IBO toxicity (1.0 pg dose), a: Nissl-stained section illustrating extensive cell death in the lateral parvicellular PVN and ventral thalamus. Note the conspicuous absence of fusiform, mediolaterally-oriented cell bodies on the injected side (left). b: adjacent section to a, stained for oxytocin. Fusiform, mediolaterally-oriented oxytocin-containing parvicellular neurons are evident on the uninjected side (right) but are absent on the side of injection. Several oxytocin-positive neurons can be seen on the injected side, which are scattered magnocellular neurons present within the medial parvicellular subdivision of the PVN at this level. Unlike magnocellular oxytocin neurons, parvicellular oxytocin neurons are killed by IBO. Divisions of PVN at this level are delineated with arrows on a. LP, lateral parvicellular PVN; MP, medial parvicellular PVN; PZ, periventricular zone; III, third ventricle.

372 neurons of the supraoptic nucleus and accessory

tBO to selectively eliminate parvicellular PVN neu-

magnocellular nuclei are resistant to kainic acid, N-

rons while sparing magnocellular elements renders

methyl-aspartate, quisqualate, and IBO 5'9. The data

this lesion technique a valuable method with which to

reported here indicate that magnocellular PVN neu-

investigate the physiological and connectional prop-

rons display a selective resistance to IBO as well. Previous reports have indicated that PVN neurons 4

erties of magnocellular vs parvicellular PVN neurons. In this context it should be noted that in the

or neurons of medial hypothalamic regions including the PVN 9 are generally resistant to neurotoxins. De-

present experiments an incomplete, but otherwise

tailed analyses of the effects of kainic acid and N-me-

uncharacterized, pattern of cell loss was observed in the parvicellular subnuclei of the PVN following the

thyl aspartate injections into the region of PVN have

administration of low doses (0.75 or 0.5/~g) of IBO.

suggested that parvicellular PVN neurons are destroyed following such treatment 714. Our data indi-

Therefore, it may prove feasible to eliminate specific

cate that injections of IBO preferentially kills parvicellular PVN n e u r o n s in a dose-dependent m a n n e r .

neurochemical and/or connectional subclasses of parvicellular neurons using IBO or related excitotoxins, further enhancing the degree of precision with which

Utilization of IBO to destroy parvicellular PVN neurons was the method of choice in these experiments,

discrete cell populations within the PVN may be studied.

since IBO does not produce seizures, lesions distant to the injection site or hemorrhagic necrosis, which may occur following injection of kainic acid 3'6A4. Appropriate doses of quisqualate or N-methyl-aspartate may prove reasonable substitutes for kainic acid in this paradigm, although the utility of N-methyl-aspartate may be limited by its p r o n o u n c e d epileptogenic properties la. The ability of moderate doses of

1 Armstrong, W.E., Warash, S., Hatton, G.1. and McNeill, T.H., Subnuclei in the rat hypothalamic paraventricular nucleus: a cytoarchitectural, horseradish peroxidase and immunocytochemical analysis, Neuroscience, 5 (1980) 1931-1952. 2 Blackstad, T.W., Heimer, L. and Mugnaini, E., Experimental neuroanatomy. General approaches and laboratory procedures. In L. Heimer and M.J. Robards (Eds.), Neuroanatomical Tract-Tracing Methods, Plenum, New York, 1981, pp. 1-53. 3 Coyle, J.T. and Schwartz, R., The use of excitatory amino acids as selective neurotoxins. In A. Bj6rklund and T. Hokfelt (Eds.), Handbook of Chemical Neuroanatomy, Vol. 1, Elsevier, Amsterdam, 1983, pp. 508-527. 4 Hastings, M.H., Roberts, A.C. and Herbert, J., Neurotoxic lesions of the anterior hypothalamus disrupt the photoperiodic but not the circadian system of the Syrian hamster, Neuroendocrinology, 40 (1985) 316-324. 5 Hastings, M.H., Winn, P. and Dunnett, S.B., Neurotoxic amino acid lesions of the lateral hypothalamus: a parametric comparison of the effects of ibotenate, N-methyl-D,Laspartate and quisqualate in the rat, Brain Research, 360 (1985) 248-256. 6 Kohler, C. and Schwarcz, R., Comparison of ibotenate and kainate neurotoxicity in the rat brain: a histological study, Neuroscience, 8 (1983) 819-835. 7 Olney, J.W. and de Gubareff, T., Extreme sensitivity of olfactory cortical neurons to kainic acid toxicity. In E.G. McGeer, J.W. Olney and P.L. McGeer (Eds.), Kainic Acid as a Tool in Neurobiology, Raven, New York, 1978, pp.

We would like to thank Kim A. Geseli for typing the manuscript, Barbara T u r n e y and Dorothy Herrera for expert technical assistance, and Dr. D o n M. Gash for helpful comments concerning this m a n u script. This work is supported by MH08883 to J. P. H. and NS19900 to S.J.W.

201-217. 8 Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, Academic Press, New York, 1982. 9 Peterson, G.M. and Moore, R.Y., Selective effects of kainic acid on diencephalic neurons, Brain Research, 202 (1980) 165-182. 10 Swanson, L.W. and Kuypers, H.G.J.M., The paraventricular nucleus of the hypothalamus: cytoarchitectonic subdivisions and the organization of projections to the pituitary, dorsal vagal complex and spinal cord as demonstrated by retrograde fluorescence double-labeling methods, J. Comp. Neurol., 194 (1980) 555-570. 11 Swanson, L.W., Sawchenko, P.E., Wiegand, S.J. and Price, J.L., Separate neurons in the paraventricular nucleus project to the median eminence and to the medulla or spinal cord, Brain Research, 198 (1980) 190-195. 12 Swanson, L.W. and Sawchenko, P.E., Hypothalamic integration: organization of the paraventricutar and supraoptic nuclei, Annu. Rev. Neurosci., 6 (1983) 269-326. 13 Wiegand, S.J. and Price, J.L., The cells of origin of afferent fibers to the median eminence in the rat, J. Comp. Neurol., 192 (1980) 1-19. 14 Zaczek, R. and Coyle, J.T., Excitatory amino acid analogues: neurotoxicity and seizures, Neuropharmacologyl 21 (1982) 15-26. t5 Zhang, T.-X. and Ciriello, J., Kainic acid lesions of paraventricular nucleus neurons reverse the elevated arterial pressure after aortic baroreceptor denervation in the rat, Brain Research, 358 (1985) 334-338.