Developmental Brain Research, 4 (1982) 103-108 103 Elsevier Biomedical Press

Short Communications

Kainic acid sensitivity of mammalian Purkinje cells in monolayer cultures

WARREN GIBBS, ELAINE A. NEALE* and GUSTAVE MOONEN**

Laboratory of Developmental Neurobiology, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD 20205 (U.S.A.)

(Accepted February 16th, 1982)

Key words: Purkinje cells - - cerebellum -- monolayer cultures - - kainic acid -- HRP staining

Purkinje cells from fetal mouse cerebellum were studied in a monolayer culture system. These neurons, labeled by intracellular iontophoresis of HRP, showed one or more thick, spine-covered dendrites and an axon whose swellings contacted other neuronal processes and Purkinje cell somata. In short-term experiments, kainic acid was cytotoxic to Purkinje neurons and not to granule cells in monolayer cultures of fetal rat cerebellum.

Successful long-term monolayer cultivation of embryonic rat cerebellum was recently reported using a dissociation procedure referred to as the microexplant technique, which maintains cellular relationships critical for Purkinje cell survival while allowing cellular dispersal so that intracellular mi- croelectrode recording of these neurons is possible s. The essentially non-afferented cultured Purkinje cells were morphologically characterized at both the light and electron microscopic levels after intracel- lular iontophoresis of horseradish peroxidase (HRP) (refs. 8, 10). Electrophysiological analysis demon- strated a Purkinje cell chemosensitivity to glutamate (ref. 4), the putative parallel fiber transmitter. In the present communication, we report results using this culture technique for fetal mouse cerebellum in which intracellular HRP iontophoresis was used to characterize the morphology of the Purkinje cells. In addition, kainio acid neurotoxicity was used to de- monstrate the glutamate chemosensitivity of Purkin- je cells in cultures prepared from rat cerebellum.

The cultivation technique has been reported else- where s. Briefly, cerebella from 17-19-day fetal mice were dissected and dissociated into microexplants by sieving through 215/tin nylon mesh. These microex-

plants were plated on calfskin collagen-coated 35 mm Falcon culture dishes. Cells were grown in Eagle's minimal essential medium (Gibco) supplemented with glucose (up to 600 mg~o), fetal calf serum (10 ~o) and heat-inactivated horse serum (10 ~o). Cul- tures were treated with 5'-fluoro-2'-deoxyuridine (FdU) from culture day 4 to day 6. From the fourth day in culture onward, fetal calf serum was omitted from the medium. Intraeellular H RP iontophoresis and subsequent processing of the cultures were per- formed as described earlier 9. Kainic acid (final con-

centration, 75/zM) was added to 28-day-old cultures and the cytotoxic effect of this treatment was moni- tored with phase contrast microscopy.

Fig. 1 is a phase-contrast photomicrograph of a 31-day-old culture, prepared from mouse cerebel- lum, showing phase-bright neurons on top of flat- tened non-neuronal cells. The macroneurons have rounded or ovoid somata approximately 20 #m in diameter from which emerging processes are clearly visible. Smaller approximately 8 #m diameter phase- bright cells are present in the same field. Intracellu- lar iontophoresis of HRP was achieved in the ma- croneurons, and Fig. 2 shows two such injected cells. One or two large dendritic processes emanate from

* To whom correspondence should be addressed at: Bldg. 36, Rm. 2A21, LDN, NICHD, NIH, Bethesda, MD 20205, U.S.A. ** Research Associate, FNRS, Institute of Medicine, Department of Neurology, University of Liege, Belgium.

104

Fig. 1. Phase-contrast photomicrograph of neurons Jn a 3 l-day-old culture prepared b3 the m~croexplant :echmque from 17-day fetal mouse cerebellum. This field shows 5 Purkinje cells and several mlcroneurons carrowheads). Bar "~5 ,m 450.

the cell soma and generate secondary and tertiary

branches. Numerous spines cover the surface of

these dendrites (Fig. 2. inset al. The shape of the

dendritic arbor and the presence of dendritic spines

characterize these macroneurons as Purkinje cells

(consult ref. 8 for further discussion of the morpho-

logy of these neurons). The thinner, smooth-surfaced axon is easily discriminated and is seen to form a

variable number of swellings (Fig. 2. inset b) in contact with neighboring neurons and neuronal pro-

cesses. Electron microscopic studies have demon-

strated that similar swellings contained pleomorphic

vesicles 10. Also visible within this field are large non- injected Purkinje neurons and a greater number or

microneurons, many of which are presumably gra- nule cells. It was a consistent finding that microneu- ron survival was greater and that Purkinje cell den-

dritic arborizatiorL was somewhat more elaborate in cultures prepared from mice than in those from rats.

The effect o f kainic acid (75 #M) on cultured rat cerebellar neurons is shown in Fig. 3. Whereas Pur- kinje cells (Fig. 3A, B and C) degenerate within 6 h of treatment, small phase-bright neurons (Fig. 3 D. E

and F) appear unaffected after at tong as 24 h in kai-

nic acid. No microneurons could bc seen in the area

around the cell illustrated in Fig 3A-C. The Pur-

kinje cell degeneration appeared grossly as a cyto-

plasmic vacuotation similar to that observed in vivo

(ref. 2). Purkinje cell dendrites showed the first signs

of deterioration followed by the soma and finally,

the axon. After 24 h of treatment, there were no morphologically identifiable Purkinje cells remain-

ing in the cultures. Control cultures prepared from rat tissue contained relatively (ew microneurons.

and these had been identified by ultrastructural

features as granule cells 8,t°. Occasional scattered microneurons, possibly stellate or basket cells, were

lost during kainic acid treatment, although micro- neurons in small aggregates (granule cells) appeared spared. After prolonged exposure (6-17 days) to kainic acid, numbers of granule cells in intact cere- bella 1 and in explant cultures t2 were markedly re- duced, although these cells were relatively unaffec- ted by short-term treatment. It is not known wheth- er a longer treatment interval would have resulted in

105

~;~i ~ iii:ii~:~i.

Fig. 2. Bright-field photomicrograph of a cerebellum microexplant, 28 days in culture, showing many Purkinje cells and a greater number of small (approximately 6 pm diameter) neurons, presumably granule cells. Two of the Purkinje cells have b~en injected with HRP. Neuron A displays a rounded soma, one primary dendritic trunk, and an axon that emerges from the soma at the pole opposite the dendrite. Neuron B has an ovoid soma and two main dendrites. The dendrites of both neurons ramify into secondary and tertiary branches which are covered with spines (see inset a). Axonal branches are slender and smooth-surfaced, and show occa- sional varicosities along their length. Inset b shows several such varicosities contacting an uninjected Purkinje cell Arrows indicate origin of Purkinje cell axons. Mouse cerebellum. Inset b, phase-contrast optics. Bar = 25/zm. x 320. Insets, x 800.

106

Fig. 3. Kainic acid neurotoxicity demonstrated by phase-contrast microscopy of neurons before (A, D] and a~tcl - 6 (B, E) and 24 (C, F) h of treatment. A, B, and C: an isolated bipolar Purkinje cell; there were no microneurons in this area. A fret 6 h in kainic acid. the cellular morphology of the Purkinje neuron is grossly altered, and after 24 h, only debris remains. D. E, ~tnd F: microneurons in the same culture. In contrast to the Purkinje cells, these neurons appear unaffected by up to 24 h of kainic acid ~realment. Final con- centration of kainic acid, 75 uM. Bar = 25 !~m. x 450.

107

further degeneration of microneurons in the cultures

studied here. In a previous study on rat embryo Purkinje cells

in culture, we showed chemosensitivity of Purkinje cells to glutamate, even in the nearly complete ab- sence of granule cells. Furthermore, glutamate 'hot spots' present on Purkinje cell dendrites were tenta- tively correlated with the distribution of spines 4 which, although largely devoid of synaptic contact, exhibited membrane specializations similar to 'dif- ferentiated' postsynaptic densities 10. In the present study, a different strategy demonstrates the presence of glutamate receptors on non-afferented Purkinje cells. Kainic acid, a structural analog of glutamate, has excitotoxic properties. Although the effects of kainic acid were first thought to be related to overex- citation of glutamate receptors11 the exact mecha- nism of kainate toxicity is not completely under- stoodS, 13. Microinjection of kainic acid into the cerebellum in vivo caused a degeneration of neurons that receive synaptic input from parallel fibers 2,16,17, although these results have been disputed 3. In post- natal cerebellar dissociated cell cultures, which were devoid of Purkinje cells, short-term treatment with kainic acid was cytotoxic to GABAergic microneu- rons, although granule cells seemed morphologically intact 6. In cerebellar explants, kainic acid treatment resulted in a major reduction or complete loss of large cortical neurons and a relative sparing of gra- nule cells 12. The degeneration of Purkinje cells oc- curred even when the cultures were treated with kai- nic acid prior to the development of parallel fiber- Purkinje cells synapses 12, and in cultures in which granule cells had been virtually eliminated 14. There was no evidence of Purkinje cell loss in cultures treated with D- or L-glutamic acid 15. Furthermore, intracerebellar nucleus neurons which do not receive parallel fiber input, and did not demonstrate an elec-

trophysiologic response to applied glutamate, also degenerated as a result of kainic acid treatmenP 2. These results were taken as evidence that the neuro- toxic effects of kainic acid may not be mediated ex- clusively by an action on glutamate receptors.

In the short-term (24 h) experiments presented in this paper, the cellular resolution of the preparation allowed us to monitor directly, using phase-contrast microscopy, the cytotoxic effect of kainic acid on Purkinje cells. Even those Purkinje cells in areas of the culture which appeared completely devoid of granule cells showed severe morphologic damage. This observation is compatible with our previous pharmacologic data showing that Purkinje cells pos- sess glutamate chemosensitivity independently of glutamatergic input 4. While kainic acid neurotoxici- ty ultimately may involve more than a direct action on the glutamate receptor, the action of kainic acid on Purkinje neurons in these relatively 'mature' cul- tures appears rapid and selective.

In summary, Purkinje cells from fetal mouse cere- bellum were studied in a monolayer culture system. Morphologic features of these Purkinje neurons were analyzed after intracellular HRP iontophore- sis. In short-term experiments, kainic acid was cyto- toxic to Purkinje neurons and not to granule cells in monolayer cultures of fetal rat cerebellum. It has been reported that granule cells from agranular neurological mutants of mice survive in vitro al- though they degenerate in vivo 7. Study of the in vitro behavior of Purkinje cells from such mutants, using the microexplant technique, could provide further insight into the pathogenesis of those muta- tions which affect cerebellar histogenesis.

We are grateful to Linda Bowers and Raymond Rusten for excellent assistance in the photographic darkroom, and to Dr. Phillip G. Nelson for his generous support.

1 Foster, A. C. and Roberts, P. J., Morphological and bio- chemical changes in the cerebellum induced by kainic acid in vivo, J. Neurochem., 34 (1980) 1191-1200.

2 Herndon, R. M. and Coyle, J. T., Selective destruction of neurons by a transmitter agonist, Science, 198 (1977) 71-72.

3 Lovell, K. L. and Jones, M. Z., Kainic acid neurotoxicity in the mouse cerebellum, Brain Res., 186 (1980) 245-249.

4 Macdonald, R. L., Moonen, G., Neale, E. A. and Nelson, P. G., Cerebellar macroneurons in microexplant cell cul-

ture: postsynaptic amino acid pharmacology, Develop. Brain Res., 5 (1982) in press.

5 McGeer, P. L. and McGeer, E. G., Use of the neurotoxic agents kainic acid and tetanus toxin in the extrapyramidal system. In B. Ceccarelli and F. Clementi (Eds.), Neuro- toxins: Tools in Neurobiology, Advances in Cytopharmaco- logy, Voi. 3, Raven Press, New York, 1979, pp. 437-446.

6 Messer, A. and Maskin, P., Short-term effects of kainie acid on rat cerebellar cells in monolayer cultures, Neuro- sci. Lett., 19 (1980) 173-177.

i0~

7 Messer, A. and Smith, D. M., In vitro behavior of granule cells from Staggerer and Weaver mutants of mice, Brain Res., 130 0977) 13-23.

8 Moonen, G., Neale, E. A., Macdonald, R. L., Gibbs, W. and Nelson, P. G., Cerebellar macroneurons in micro- explant cell culture: methodology, basic electrophysiolo- gy, and morphology after horseradish peroxidase injec- tion, Develop. Brain Res., 5 (1982) in press.

9 Neale, E. A., Macdonald, R. L. and Nelson, P. G., Intra- cellular horseradish peroxidase injection for correlation of light and electron microscopic anatomy with synaptic physiology of cultured mouse spinal cord neurons, Brain Res., 152 (1978) 265-282.

10 Neale, E. A., Moonen, G., Macdonald, R, L. and Nelson, P. G., Cerebellar macroneurons in microexplant cell culture: ultrastruCtural morphology, Neuroscience, (1982) in press.

II Olney, J. W., Neurotoxicity of excitatory amino acids. In E. G. McGeer, J. W. Olney and P. L. McGeer (Eds.), Kainic Acid as a Tool in Neurobiology, Raven Press, New York, 1978, pp. 95-121.

12 Sell, F. J., Blank, N. K. and Leiman, A. L., Toxic effects

of kalnic acid on mouse cerebellum m tissue culture, Brai~z Res., 161 (1979) 253-265.

13 Sell, F. J., Blank, N. K., Woodward, W. R. and Leiman, A. L., Kainate effects in cerebe!tar cultures. In G, Di- Chiara and G. L. Gessa (Eds,~, Ghltamate as a Neuro.. transmitter, Advances in Biochemical P.vychz~pharmacohJ£,~ Vol. 27, Raven Press, New York, !981, pp. 347 354~

14 Sell, F. J. and Woodward, W. R., Kainicacid neurotoxici- ty in granutoprival cerebellar cuin~res, Brain Rei~:; 197 (1980) 285 289.

15 Sell, F. J., Woodward, W, R., Blank, N. K. and Leimam A. L., Evidence against chronic depolarization as a me-. chanism of kainic acid toxicity in mouse cerebel!ar cul- tures, Brain Res., 159 (1978) 43 ! 435.

16 Snider, S. R. and Snider, R S.. Kainic acid: enduring alterations in cerebellar morpholog,/and in cerebral cate- cholamine and GABA concentrt~ti0ns after cerebellar injection in lhe rat, Neun)sci. LeH.. t2 (1979) 3392342.

17 Tan Tran, V. and Snyder, S. M., Amino acid neuromms- mitter candidates in rat cerebellum: selective effeOs of kainic acid lesions, Brain Res., 167 (1979J 345 -353.