Brain Research, 538 (1991) 263-268
263
Elsevier BRES 16229
Differentiation of dentate granule cells in slice cultures of rat hippocampus: a Golgi/electron microscopic study B. Heimrich and M. Frotscher Institute of Anatomy, University of Freiburg, Freiburg (ER.G.) (Accepted 7 August 1990)
Key words: Organotypic culture; Fascia dentata; Granule cell; Neuronal development
The differentiation of granule cells in organotypic cultures of rat hippocampus was studied by means of the Golgi/electron microscope (EM) technique. Like in vivo, the granule cells have a small round or ovoid cell body which gives rise to apical dendrites densely covered with spines. However, the apical dendrites of the cultured granule cells are more horizontally oriented than in the normal fascia dentata where they form a cone-shaped dendritic arbor. Granule cells in slice cultures occasionally have basal dendrites invading the hilar region. Electron microscopic examination revealed many synaptic contacts on identified apical and basal dendrites of the gold-toned granule cells in culture. This suggests that a considerable synaptic reorganization takes place since all extrinsic afferents normally innervating the granule cells are lost. Several granule cells displayed deep infoldings of their nuclei which are known from in vivo studies to be a characteristic feature of non-granule cells in this region, i.e. basket cells. The presence of basal dendrites and nuclear infoldings indicates an increased variability of this cell type which in situ displays a rather stereotyped morphology.
INTRODUCTION The granule cells represent the main cell type in the fascia dentata. In rodents, these cells have a remarkably uniform appearance and several typical features 18. The
affect their characteristic dendritic and synaptic organization. For this purpose we applied the combined Golgi/electron microscope (EM) procedure which allows a correlated light and electron microscopic analysis of single identified n e u r o n s and their processes.
cell bodies are densely packed in a U-shaped cell layer on the fringe of the hippocampal formation. In situ, this cell type can be easily detected by its small round cell body and its characteristic bipolarity. From the apical pole the cone-shaped dendritic arbor emerges whereas the axon, the mossy fiber, originates from the basal pole and projects to hilar n e u r o n s and to the hippocampus proper forming giant synaptic boutons 2'15. A t least in rodents, granule ceils lack basal dendrites. Basal dendrites and an increased morphological variability are, however, characteristic features of granule cells in the primate hippocampus 22. Thus, by their relative simple uniform appearance, granule cells in the rodent fascia dentata provide a model for studies of n e u r o n a l plasticity. Organotypic slice cultures offer the possibility of studying n e u r o n a l differentiation in the absence of all extrinsic afferents. In vivo, a well-defined set of afferent fibers impinges on the granule cell dendrites in a laminated fashion 18. It was the aim of the present study to evaluate to what extent the lack of these specific extrinsic afferents as well as a somewhat altered spatial a r r a n g e m e n t of the granule cells u n d e r culture condition
MATERIALS AND METHODS Two- to 3-day-old rat pups were sacrificed by decapitation. The hippocampi were dissected under aseptic conditions and cut into 400 /zm thick slices. Two hippocampal slices were mounted together on a glass coverslip. The distance between both explants was about 1 cm to avoid slice to slice interactions. They were embedded in a plasma clot and coagulated by the addition of a drop of thrombin. For cultivation we applied the roller-tube technique developed by G~ihwiler13. Cultures were fed 3 times a week and incubated for 20 + 2 days. The cultures were fixed for 15 min in a solution of 1% glutaraldehyde and 1% paraformaldehyde in 0.12 M phosphate buffer (pH 7.3). Golgi-impregnation was performed by employing a modified Golgi section procedure 6, see ref. 8. Five to 10 cultures, with intervening Parafilm, were piled on top of each other, and embedded in 5% agar thereby forming a 'tissue block'. These blocks were immersed in a solution of 1 g osmium tetroxide and 12 g potassium dichromate in 500 ml distilled water for 6 days at 8 °C. Thereafter, the cultures were incubated in 0.75% AgNO 3 solution for 14-21 h at room temperature. Prior to gold-toning4 the cultures were dipped in glycerol and illuminated for 2 h at 21 °C1'5. They were then rehydrated in graded series of glycerol/potassium dichromate solutions. Gold-toning was performed in 0.05% gold tetrachloride in distilled water containing 20% glycerol for 75-120 min at 4 °C. The cultures were rinsed in distilled water, deimpregnated in sodium thiosulphate, postfixed in osmium tetroxide, and fiat-embedded in Araldite (for details, see
Correspondence: B. Heimrich, Institute of Anatomy, University of Freiburg, Albertstr. 17, D-7800 Freiburg, F.R.G. 0006-8993/91/$03.50 (~) 1991 Elsevier Science Publishers B.V. (Biomedical Division)
264 ref. 12). Following re-embedding in plastic capsules, serial ultrathin sections of identified granule cells (n = 5) were cut on an LKB ultratome 3, mounted on slot grids coated with Formvar film, and stained with uranyl acetate and lead nitrate for electron microscopic analysis. RESULTS Most h i p p o c a m p a l cultures preserved the typical Ushaped structure of the cell layers although the cells were less densely p a c k e d than in ,vivo. Of the fascia dentata, the early forming suprapyramidal (ectal) blade mainly developed. We have, accordingly, focused in this report on granule cells located in this blade. Golgi-impregnation
resulted in the staining of the various hippocampal and dentate cell types. Several granule cells are shown in Fig. 1. They have a small round cell body which gives rise to an apical dendritic arbor. All dendrites are densely covered with spines (Fig. lb). Unlike in vivo, the dendritic arbor appears more variable ranging from the typical cone-shaped to almost horizontal orientation (Fig. la). Also, many granule cells displayed basal dendrites extending into the hilar region (Fig. la, open arrows). In vivo basal dendrites have only been o b s e r v e d in yet undifferentiated granule cells of postnatal stages P 5 P1019. H o w e v e r , the granule cells cultured for 20 days have long, well-differentiated dendrites which are
/
Fig. 1, a: photomontage of Golgi-impregnated and gold-toned granule cells in slice culture. Note almost horizontal course of apical dendrites and the occurrence of basal dendrites (open arrows). The axons (black arrows) originate as normal from the basal pole of the small granule cell body. Curved arrow marks an apical dendrite with spines that is shown at higher magnification in b. a, x 200; b, × 315.
265 densely covered with spines (Fig. l a , b). Hence, with regard to dendritic length and spine density the cultured
granule cells cannot be regarded as undifferentiated neurons that have preserved an immature appearance.
t
~:i¸¸¸~i~ f
\
. -
\
\
\\
Fig. 2. Golgi-impregnated granule cell in vitro, a: cell body and apical dendritic arbor as seen after gold-toning and fiat-embedding in Araldite. Black arrow points to the mossy fiber axon originating from the basal pole of the cell body. Open arrow indicates two somatic spines, x 200. b: electron micrograph of the cell body region of the granule cell shown in a. Gold grains are diffusely dispersed in the cytoplasm which forms a small perinuclear rim. Note nuclear infoldings (black arrows). Open arrow labels a somatic spine, x 11,900.
266
Fig. 3. a: ultrathin section of an apical dendrite of the identified granule cell shown in Fig. 2. Numerous spines are visible along this gold-toned dendrite. An arrow points to a complex spine which is shown at higher magnification in b. x 11,900. b: vesicle-filled bouton establishing asymmetric synaptic contact with a complex spine (black arrow), The synaptic contact zone is divided by a spinule (open arrow). × 42,000. c: asymmetric synaptic contact formed with a dendritic shaft of this identified granule cell. × 42,000. The granule cell axons arise, as normally, from the basal pole (Fig. la) and project towards the hilar region of the culture. Unfortunately, mossy axons and their hilar collaterals could not be traced over long distances. The electron microscopic analysis of Golgi-impregnated gold-toned granule cells revealed characteristic fine structural features of these neurons such as small amounts of perinuclear cytoplasm. However, several granule cell nuclei displayed infoldings (Fig. 2b) which in vivo is characteristic for non-pyramidal cells of the hippocampus, i.e. the basket cells. In previous combined Golgi/EM studies of developing granule cells no such
nuclear indentations have been observed ~9. Granule cells from 5-day-old rats displayed instead the characteristic smooth, round nucleus surrounded by small amounts of perinuclear cytoplasm. Preliminary studies of glutamate decarboxylase-immunoreactive non-pyramidal cells in hippocampal slice cultures have shown that these cells also have characteristic deep nuclear infoldings, Often somatic spines were observed on the cell bodies of the cultured granule cells (Fig. 2a,b). Spines on granule cell somata are present in developing as well as in adult rats 7. Numerous presynaptic boutons formed contacts on somatic and dendritic spines (Fig. 3a,b) as well as on
267 dendritic shafts (Fig. 3c) of the gold-toned identified granule cell dendrites. Characteristic complex spines that protrude a spinule towards the presynaptic bouton, thereby dividing the synaptic contact, are also found in vitro (Fig. 3b). Other types of spines as described for granule cells of adult rats, i.e. stubby spines and mushroom-shaped spines 3, were observed as well. However, in this first qualitative analysis we did not monitor their distribution, and no spine counts were performed. We regard it as an important finding that all these different types of spines were contacted by presynaptic terminals which formed mature synaptic contacts. DISCUSSION The results of the present study have demonstrated that dentate granule cells survive and differentiate in slice cultures lacking all extrinsic afferent input. Our observations thus confirm and extend previous studies 1°' 12,14,23.25 which similarly showed the cell-specific development of pyramidal cells and GABAergic neurons in slice cultures of hippocampus. However, when compared with the differentiation of dentate granule cells in vivo, some striking differences are noted. First, the infrapyramidal (ental) blade of the fascia dentata, which appears late in ontogeny (around P024) and matures postnatally, is poorly developed in the cultures. Certainly, the age of the donors plays an important role in this respect since the infrapyramidal blade was more developed when the cultures were taken from 6-8-day-old animals 25. In vivo, malformation of the infrapyramidal blade has recently been reported after selective destruction of meningeal cells 16. It appears that intact meninges are a prerequisite for the normal development of the fascia dentata. It is interesting in this context that in both experimental paradigms, granule cells in the remaining suprapyramidal blade develop basal dendrites. These granule cells then resemble very much the more variable granule cells in the primate fascia dentata a2. A more developed infrapyramidal blade in slice cultures from older donors may explain the lack of basal dendrites on granule cells in these studies 25. An increase in variability also holds true for the apical dendritic arbor of the cultured granule cells which often have a more horizontal orientation when compared with normal granule cells fixed in situ by transcardial perfusion. Probably the orientation of dendrites is under control of extrinsic afferents since granule cells in dentate transplants which similarly lack their normal afferent input also show numerous horizontally oriented dendrites9. It may also be that granule cells cultured for 3 weeks retain some immature structural features. It has been shown, for instance, that immature granule cells transiently form basal dendrites 19'21. The
appearance of basal dendrites on the cultured granule cells may also be explained by more space between cell bodies. Intercellular space increases by the death of some neurons and flattening of the cultures. We have recently shown that pyramidal cells in CA1 develop a complex horizontal and basal dendritic arbor under these culture conditions ~2. A similarly complex horizontal and basal dendritic arbor is not formed by pyramidal cells in the normal rodent hippocampus but is characteristic for CA1 pyramidal cells in the primate hippocampus where these cells are loosely distributed 11. Like in vivo, the granule cell axons arise from the basal pole of the cell body and run through the hilus to the regio inferior. There they establish characteristic giant synapses on large spines of CA3 pyramidal cells as known from EM studies of HRP-filled CA3 pyramidal cells in slice cultures of hippocampus 1°. We were impressed by the large number of terminals that formed synaptic contacts on the identified gold-toned granule cell dendrites in the cultures. These dendrites are normally contacted by the various extrinsic fiber systems terminating in a laminated fashion in the molecular layer. Obviously, the lack of extrinsic input in the cultures is largely compensated for by collaterals of the cultured neurons themselves. This may explain why postsynaptic dendritic spines are formed, albeit less than in vivo. However, the appearance of characteristic types of spines, like the complex spines, stubby spines and mushroom-shaped spines 3 suggests that their development is not dependent upon specific extrinsic afferents. Some of the terminals establishing synaptic contacts on the cultured cells have been identified by intracellular labeling with horseradish peroxidase 1° and by immunostaining for glutamate decarboxylase, the GABA-synthesizing enzyme x2. As observed after partial deafferentation in vivo 7"17, the granule cell axons themselves take part in the sprouting process since we observed an increased number of Timm-stained supragranular mossy fiber boutons in the cultures 2°. In conclusion then, the development of granule cells in slice cultures results in a larger variability of these cells which concerns the dendritic orientation, the occasional occurrence of basal dendrites, and fine structural characteristics such as infolded nuclei. In spite of this, granule cells are still easily identified in the cultures and may serve as a model to analyze principal steps of neuronal differentiation in vitro.
Acknowledgements. The authors wish to thank E. Thielen for technical assistance and Dr. H. Schwegler for reading the manuscript. This study was supported by the Deutsche Forschungsgemeinschaft: Fr 620/2-3.
268
REFERENCES 1 Blackstad, T.W., Tract tracing by electron microscopy of Golgi preparations. In L. Heimer and M.J. RoBards (Eds.), Neuroanatomical Tract-Tracing Methods, Plenum, Press, New York, 198l, pp. 407-440. 2 Blackstad, T.W. and Kjaerheim, A., Special axodendritic synapses in the hippocampal cortex: electron and light microscopic studies on the layer of mossy fibers, J. Comp. Neurol., 117 (1961) 113-159. 3 Desmond, N.L. and Levy, W.B., Granule cell dendritic spine density in the rat hippocampus varies with spine shape and location, Neurosci. Lett., 54 (1985) 219-224. 4 Fair6n, A., Peters, A. and Saldanha, J., A new procedure for examining Golgi impregnated neurons by light and electron microscopy, J. Neurocytol., 6 (1977) 311-337. 5 Fair6n, A., DeFelipe, J. and Martinez-Ruiz, R., The Golgi EM procedure: a tool to study neocortical interneurons. In Eleventh International Congress of Anatomy: Glial and Neuronal Cell Biology, Liss, New York, 1981, pp. 291-301. 6 Freund, T. and Somogyi, P., The section Golgi impregnation procedure. I. Description of the method and its combination with histochemistry after intracellular iontophoresis or retrograde transport of horseradish peroxidase, Neuroscience, 9 (1983) 463-474. 7 Frotscher, M. and Zimmer, J., Lesion-induced mossy fibers to the molecular layer of the rat fascia dentata: identification of postsynaptic granule cells by the Golgi-EM technique, J. Comp. Neurol., 215 (1983) 299-311. 8 Frotscher, M. and Leranth, C., The cholinergic innervation of the rat fascia dentata: identification of target structures on granule cells by combining choline acetyltransferase immunocytochemistry and Golgi impregnation, J. Comp. Neurol., 243 (1986) 58-70. 9 Frotscher, M. and Zimmer, J., Intracerebral transplants of the rat fascia dentata: a Golgi/electron microscope study of dentate granule cells, J. Comp. Neurol., 246 (1986) 181-190. 10 Frotscher, M. and G~ihwiler, B.H., Synaptic organization of intracellularly stained CA3 pyramidal neurons in slice cultures of rat hippocampus, Neuroscience, 24 (1988) 541-551. 11 Frotscher, M., Kraft, J. and Zorn, U., Fine structure of identified neurons in the primate hippocampus: a combined Golgi/EM study in the baboon, J. Comp. Neurol., 275 (1988) 254-270. 12 Frotscher, M., Heimrich, B. and Schwegler, H., Plasticity of identified neurons in slice cultures of hippocampus: a combined Golgi/electron microscopic and immunocytochemical study. In J.
Storm-Mathisen, J. Zimmer and O.P. Ottersen (Eds.), Progress in Brain Research, VoL 83, Elsevier, Amsterdam, 1990, pp. 323-337. 13 G~hwiler, B.H., Organotypic monolayer cultures of nervous tissue, J. Neurosci. Methods, 4 (1981) 329-342. 14 G~hwiler, B.H., Development of the hippocampus in vitro: cell types, synapses and receptors, Neuroscience, 11 (1984) 751-760. 15 Hamlyn, L.H., The fine structure of the mossy fiber endings in the hippocampus of the rabbit, J. Anat., 96 (1962) 112-120. 16 Hartrnann, D., Frotscher, M., Sievers, J. and Pehlemann, F.W., Altered development of the dentate gyrus after neonatal destruction of meningeal cells: analysis of fiber connections and cell morphology. In N. Eisner and W. Singer (Eds.), Dynamics and Plasticity in Neuronal Systems, Thieme, New York, 1989, 393 pp. 17 Laurberg, S. and Zimmer, J., Lesion-induced rerouting of hippocampal mossy fibers in developing but not in adult rats, J. Comp. Neurol., 190 (1980) 627-650. 18 L/Jbbers, K. and Frotscher, M.. Fine structure and synaptic connections of identified neurons in the rat fascia dentata, Anat. Embryol., 177 (1987) 1-14. 19 L~ibbers, K. and Frotscher, M., Differentiation of granule cells in relation to GABAergic neurons in the rat fascia dentata: combined Goigi/EM and immunocytochemical studies, Anat. Embryol., 178 (1988) 119-127. 20 Schwegler, H., Heimrich, B., Keller, E, Renner, P. and Crusio, W.E., Strain-specific development of the mossy fiber system in organotypic cultures of the mouse hippocampus, Neurosci. Lett., 87 (1988) 7-10. 21 Seress, L. and Pokorny, J., Structure of the granular layer of the rat dentate gyms. A light-microscopic and Golgi study, J. Anat., 133 (1981) 181-195. 22 Seress. L. and Frotscher, M., Morphological variability is a characteristic feature of granule cells in the primate fascia dentata: a combined Golgi/electron microscope study, J. Comp. Neurol., 293 (1990) 253-267. 23 Streit, P., Thompson, S.M. and G~ihwiler, B.H., Anatomical and physiological properties of GABAergic neurotransmission in organotypic slice cultures of rat hippocampus. Eur. J. Neurosci., 1 (1989) 603-615. 24 Wenzel, J., Stender, G. and Duwe, G., Zur Entwicklung der Neuronenstruktur der Fascia dentata bei der Ratte. Neurohistologisch-morphometrische, ultrastrukturelle und experimentelle Untersuchungen, J. Hirnforsch., 22 (198t) 629-683. 25 Zimmer, J. and G~ihwiler, B.H., Cellular and connective organization of slice cultures of the rat hippocampus and fascia dentata, J. Comp. Neurol., 228 (1984) 432-446.
263
Elsevier BRES 16229
Differentiation of dentate granule cells in slice cultures of rat hippocampus: a Golgi/electron microscopic study B. Heimrich and M. Frotscher Institute of Anatomy, University of Freiburg, Freiburg (ER.G.) (Accepted 7 August 1990)
Key words: Organotypic culture; Fascia dentata; Granule cell; Neuronal development
The differentiation of granule cells in organotypic cultures of rat hippocampus was studied by means of the Golgi/electron microscope (EM) technique. Like in vivo, the granule cells have a small round or ovoid cell body which gives rise to apical dendrites densely covered with spines. However, the apical dendrites of the cultured granule cells are more horizontally oriented than in the normal fascia dentata where they form a cone-shaped dendritic arbor. Granule cells in slice cultures occasionally have basal dendrites invading the hilar region. Electron microscopic examination revealed many synaptic contacts on identified apical and basal dendrites of the gold-toned granule cells in culture. This suggests that a considerable synaptic reorganization takes place since all extrinsic afferents normally innervating the granule cells are lost. Several granule cells displayed deep infoldings of their nuclei which are known from in vivo studies to be a characteristic feature of non-granule cells in this region, i.e. basket cells. The presence of basal dendrites and nuclear infoldings indicates an increased variability of this cell type which in situ displays a rather stereotyped morphology.
INTRODUCTION The granule cells represent the main cell type in the fascia dentata. In rodents, these cells have a remarkably uniform appearance and several typical features 18. The
affect their characteristic dendritic and synaptic organization. For this purpose we applied the combined Golgi/electron microscope (EM) procedure which allows a correlated light and electron microscopic analysis of single identified n e u r o n s and their processes.
cell bodies are densely packed in a U-shaped cell layer on the fringe of the hippocampal formation. In situ, this cell type can be easily detected by its small round cell body and its characteristic bipolarity. From the apical pole the cone-shaped dendritic arbor emerges whereas the axon, the mossy fiber, originates from the basal pole and projects to hilar n e u r o n s and to the hippocampus proper forming giant synaptic boutons 2'15. A t least in rodents, granule ceils lack basal dendrites. Basal dendrites and an increased morphological variability are, however, characteristic features of granule cells in the primate hippocampus 22. Thus, by their relative simple uniform appearance, granule cells in the rodent fascia dentata provide a model for studies of n e u r o n a l plasticity. Organotypic slice cultures offer the possibility of studying n e u r o n a l differentiation in the absence of all extrinsic afferents. In vivo, a well-defined set of afferent fibers impinges on the granule cell dendrites in a laminated fashion 18. It was the aim of the present study to evaluate to what extent the lack of these specific extrinsic afferents as well as a somewhat altered spatial a r r a n g e m e n t of the granule cells u n d e r culture condition
MATERIALS AND METHODS Two- to 3-day-old rat pups were sacrificed by decapitation. The hippocampi were dissected under aseptic conditions and cut into 400 /zm thick slices. Two hippocampal slices were mounted together on a glass coverslip. The distance between both explants was about 1 cm to avoid slice to slice interactions. They were embedded in a plasma clot and coagulated by the addition of a drop of thrombin. For cultivation we applied the roller-tube technique developed by G~ihwiler13. Cultures were fed 3 times a week and incubated for 20 + 2 days. The cultures were fixed for 15 min in a solution of 1% glutaraldehyde and 1% paraformaldehyde in 0.12 M phosphate buffer (pH 7.3). Golgi-impregnation was performed by employing a modified Golgi section procedure 6, see ref. 8. Five to 10 cultures, with intervening Parafilm, were piled on top of each other, and embedded in 5% agar thereby forming a 'tissue block'. These blocks were immersed in a solution of 1 g osmium tetroxide and 12 g potassium dichromate in 500 ml distilled water for 6 days at 8 °C. Thereafter, the cultures were incubated in 0.75% AgNO 3 solution for 14-21 h at room temperature. Prior to gold-toning4 the cultures were dipped in glycerol and illuminated for 2 h at 21 °C1'5. They were then rehydrated in graded series of glycerol/potassium dichromate solutions. Gold-toning was performed in 0.05% gold tetrachloride in distilled water containing 20% glycerol for 75-120 min at 4 °C. The cultures were rinsed in distilled water, deimpregnated in sodium thiosulphate, postfixed in osmium tetroxide, and fiat-embedded in Araldite (for details, see
Correspondence: B. Heimrich, Institute of Anatomy, University of Freiburg, Albertstr. 17, D-7800 Freiburg, F.R.G. 0006-8993/91/$03.50 (~) 1991 Elsevier Science Publishers B.V. (Biomedical Division)
264 ref. 12). Following re-embedding in plastic capsules, serial ultrathin sections of identified granule cells (n = 5) were cut on an LKB ultratome 3, mounted on slot grids coated with Formvar film, and stained with uranyl acetate and lead nitrate for electron microscopic analysis. RESULTS Most h i p p o c a m p a l cultures preserved the typical Ushaped structure of the cell layers although the cells were less densely p a c k e d than in ,vivo. Of the fascia dentata, the early forming suprapyramidal (ectal) blade mainly developed. We have, accordingly, focused in this report on granule cells located in this blade. Golgi-impregnation
resulted in the staining of the various hippocampal and dentate cell types. Several granule cells are shown in Fig. 1. They have a small round cell body which gives rise to an apical dendritic arbor. All dendrites are densely covered with spines (Fig. lb). Unlike in vivo, the dendritic arbor appears more variable ranging from the typical cone-shaped to almost horizontal orientation (Fig. la). Also, many granule cells displayed basal dendrites extending into the hilar region (Fig. la, open arrows). In vivo basal dendrites have only been o b s e r v e d in yet undifferentiated granule cells of postnatal stages P 5 P1019. H o w e v e r , the granule cells cultured for 20 days have long, well-differentiated dendrites which are
/
Fig. 1, a: photomontage of Golgi-impregnated and gold-toned granule cells in slice culture. Note almost horizontal course of apical dendrites and the occurrence of basal dendrites (open arrows). The axons (black arrows) originate as normal from the basal pole of the small granule cell body. Curved arrow marks an apical dendrite with spines that is shown at higher magnification in b. a, x 200; b, × 315.
265 densely covered with spines (Fig. l a , b). Hence, with regard to dendritic length and spine density the cultured
granule cells cannot be regarded as undifferentiated neurons that have preserved an immature appearance.
t
~:i¸¸¸~i~ f
\
. -
\
\
\\
Fig. 2. Golgi-impregnated granule cell in vitro, a: cell body and apical dendritic arbor as seen after gold-toning and fiat-embedding in Araldite. Black arrow points to the mossy fiber axon originating from the basal pole of the cell body. Open arrow indicates two somatic spines, x 200. b: electron micrograph of the cell body region of the granule cell shown in a. Gold grains are diffusely dispersed in the cytoplasm which forms a small perinuclear rim. Note nuclear infoldings (black arrows). Open arrow labels a somatic spine, x 11,900.
266
Fig. 3. a: ultrathin section of an apical dendrite of the identified granule cell shown in Fig. 2. Numerous spines are visible along this gold-toned dendrite. An arrow points to a complex spine which is shown at higher magnification in b. x 11,900. b: vesicle-filled bouton establishing asymmetric synaptic contact with a complex spine (black arrow), The synaptic contact zone is divided by a spinule (open arrow). × 42,000. c: asymmetric synaptic contact formed with a dendritic shaft of this identified granule cell. × 42,000. The granule cell axons arise, as normally, from the basal pole (Fig. la) and project towards the hilar region of the culture. Unfortunately, mossy axons and their hilar collaterals could not be traced over long distances. The electron microscopic analysis of Golgi-impregnated gold-toned granule cells revealed characteristic fine structural features of these neurons such as small amounts of perinuclear cytoplasm. However, several granule cell nuclei displayed infoldings (Fig. 2b) which in vivo is characteristic for non-pyramidal cells of the hippocampus, i.e. the basket cells. In previous combined Golgi/EM studies of developing granule cells no such
nuclear indentations have been observed ~9. Granule cells from 5-day-old rats displayed instead the characteristic smooth, round nucleus surrounded by small amounts of perinuclear cytoplasm. Preliminary studies of glutamate decarboxylase-immunoreactive non-pyramidal cells in hippocampal slice cultures have shown that these cells also have characteristic deep nuclear infoldings, Often somatic spines were observed on the cell bodies of the cultured granule cells (Fig. 2a,b). Spines on granule cell somata are present in developing as well as in adult rats 7. Numerous presynaptic boutons formed contacts on somatic and dendritic spines (Fig. 3a,b) as well as on
267 dendritic shafts (Fig. 3c) of the gold-toned identified granule cell dendrites. Characteristic complex spines that protrude a spinule towards the presynaptic bouton, thereby dividing the synaptic contact, are also found in vitro (Fig. 3b). Other types of spines as described for granule cells of adult rats, i.e. stubby spines and mushroom-shaped spines 3, were observed as well. However, in this first qualitative analysis we did not monitor their distribution, and no spine counts were performed. We regard it as an important finding that all these different types of spines were contacted by presynaptic terminals which formed mature synaptic contacts. DISCUSSION The results of the present study have demonstrated that dentate granule cells survive and differentiate in slice cultures lacking all extrinsic afferent input. Our observations thus confirm and extend previous studies 1°' 12,14,23.25 which similarly showed the cell-specific development of pyramidal cells and GABAergic neurons in slice cultures of hippocampus. However, when compared with the differentiation of dentate granule cells in vivo, some striking differences are noted. First, the infrapyramidal (ental) blade of the fascia dentata, which appears late in ontogeny (around P024) and matures postnatally, is poorly developed in the cultures. Certainly, the age of the donors plays an important role in this respect since the infrapyramidal blade was more developed when the cultures were taken from 6-8-day-old animals 25. In vivo, malformation of the infrapyramidal blade has recently been reported after selective destruction of meningeal cells 16. It appears that intact meninges are a prerequisite for the normal development of the fascia dentata. It is interesting in this context that in both experimental paradigms, granule cells in the remaining suprapyramidal blade develop basal dendrites. These granule cells then resemble very much the more variable granule cells in the primate fascia dentata a2. A more developed infrapyramidal blade in slice cultures from older donors may explain the lack of basal dendrites on granule cells in these studies 25. An increase in variability also holds true for the apical dendritic arbor of the cultured granule cells which often have a more horizontal orientation when compared with normal granule cells fixed in situ by transcardial perfusion. Probably the orientation of dendrites is under control of extrinsic afferents since granule cells in dentate transplants which similarly lack their normal afferent input also show numerous horizontally oriented dendrites9. It may also be that granule cells cultured for 3 weeks retain some immature structural features. It has been shown, for instance, that immature granule cells transiently form basal dendrites 19'21. The
appearance of basal dendrites on the cultured granule cells may also be explained by more space between cell bodies. Intercellular space increases by the death of some neurons and flattening of the cultures. We have recently shown that pyramidal cells in CA1 develop a complex horizontal and basal dendritic arbor under these culture conditions ~2. A similarly complex horizontal and basal dendritic arbor is not formed by pyramidal cells in the normal rodent hippocampus but is characteristic for CA1 pyramidal cells in the primate hippocampus where these cells are loosely distributed 11. Like in vivo, the granule cell axons arise from the basal pole of the cell body and run through the hilus to the regio inferior. There they establish characteristic giant synapses on large spines of CA3 pyramidal cells as known from EM studies of HRP-filled CA3 pyramidal cells in slice cultures of hippocampus 1°. We were impressed by the large number of terminals that formed synaptic contacts on the identified gold-toned granule cell dendrites in the cultures. These dendrites are normally contacted by the various extrinsic fiber systems terminating in a laminated fashion in the molecular layer. Obviously, the lack of extrinsic input in the cultures is largely compensated for by collaterals of the cultured neurons themselves. This may explain why postsynaptic dendritic spines are formed, albeit less than in vivo. However, the appearance of characteristic types of spines, like the complex spines, stubby spines and mushroom-shaped spines 3 suggests that their development is not dependent upon specific extrinsic afferents. Some of the terminals establishing synaptic contacts on the cultured cells have been identified by intracellular labeling with horseradish peroxidase 1° and by immunostaining for glutamate decarboxylase, the GABA-synthesizing enzyme x2. As observed after partial deafferentation in vivo 7"17, the granule cell axons themselves take part in the sprouting process since we observed an increased number of Timm-stained supragranular mossy fiber boutons in the cultures 2°. In conclusion then, the development of granule cells in slice cultures results in a larger variability of these cells which concerns the dendritic orientation, the occasional occurrence of basal dendrites, and fine structural characteristics such as infolded nuclei. In spite of this, granule cells are still easily identified in the cultures and may serve as a model to analyze principal steps of neuronal differentiation in vitro.
Acknowledgements. The authors wish to thank E. Thielen for technical assistance and Dr. H. Schwegler for reading the manuscript. This study was supported by the Deutsche Forschungsgemeinschaft: Fr 620/2-3.
268
REFERENCES 1 Blackstad, T.W., Tract tracing by electron microscopy of Golgi preparations. In L. Heimer and M.J. RoBards (Eds.), Neuroanatomical Tract-Tracing Methods, Plenum, Press, New York, 198l, pp. 407-440. 2 Blackstad, T.W. and Kjaerheim, A., Special axodendritic synapses in the hippocampal cortex: electron and light microscopic studies on the layer of mossy fibers, J. Comp. Neurol., 117 (1961) 113-159. 3 Desmond, N.L. and Levy, W.B., Granule cell dendritic spine density in the rat hippocampus varies with spine shape and location, Neurosci. Lett., 54 (1985) 219-224. 4 Fair6n, A., Peters, A. and Saldanha, J., A new procedure for examining Golgi impregnated neurons by light and electron microscopy, J. Neurocytol., 6 (1977) 311-337. 5 Fair6n, A., DeFelipe, J. and Martinez-Ruiz, R., The Golgi EM procedure: a tool to study neocortical interneurons. In Eleventh International Congress of Anatomy: Glial and Neuronal Cell Biology, Liss, New York, 1981, pp. 291-301. 6 Freund, T. and Somogyi, P., The section Golgi impregnation procedure. I. Description of the method and its combination with histochemistry after intracellular iontophoresis or retrograde transport of horseradish peroxidase, Neuroscience, 9 (1983) 463-474. 7 Frotscher, M. and Zimmer, J., Lesion-induced mossy fibers to the molecular layer of the rat fascia dentata: identification of postsynaptic granule cells by the Golgi-EM technique, J. Comp. Neurol., 215 (1983) 299-311. 8 Frotscher, M. and Leranth, C., The cholinergic innervation of the rat fascia dentata: identification of target structures on granule cells by combining choline acetyltransferase immunocytochemistry and Golgi impregnation, J. Comp. Neurol., 243 (1986) 58-70. 9 Frotscher, M. and Zimmer, J., Intracerebral transplants of the rat fascia dentata: a Golgi/electron microscope study of dentate granule cells, J. Comp. Neurol., 246 (1986) 181-190. 10 Frotscher, M. and G~ihwiler, B.H., Synaptic organization of intracellularly stained CA3 pyramidal neurons in slice cultures of rat hippocampus, Neuroscience, 24 (1988) 541-551. 11 Frotscher, M., Kraft, J. and Zorn, U., Fine structure of identified neurons in the primate hippocampus: a combined Golgi/EM study in the baboon, J. Comp. Neurol., 275 (1988) 254-270. 12 Frotscher, M., Heimrich, B. and Schwegler, H., Plasticity of identified neurons in slice cultures of hippocampus: a combined Golgi/electron microscopic and immunocytochemical study. In J.
Storm-Mathisen, J. Zimmer and O.P. Ottersen (Eds.), Progress in Brain Research, VoL 83, Elsevier, Amsterdam, 1990, pp. 323-337. 13 G~hwiler, B.H., Organotypic monolayer cultures of nervous tissue, J. Neurosci. Methods, 4 (1981) 329-342. 14 G~hwiler, B.H., Development of the hippocampus in vitro: cell types, synapses and receptors, Neuroscience, 11 (1984) 751-760. 15 Hamlyn, L.H., The fine structure of the mossy fiber endings in the hippocampus of the rabbit, J. Anat., 96 (1962) 112-120. 16 Hartrnann, D., Frotscher, M., Sievers, J. and Pehlemann, F.W., Altered development of the dentate gyrus after neonatal destruction of meningeal cells: analysis of fiber connections and cell morphology. In N. Eisner and W. Singer (Eds.), Dynamics and Plasticity in Neuronal Systems, Thieme, New York, 1989, 393 pp. 17 Laurberg, S. and Zimmer, J., Lesion-induced rerouting of hippocampal mossy fibers in developing but not in adult rats, J. Comp. Neurol., 190 (1980) 627-650. 18 L/Jbbers, K. and Frotscher, M.. Fine structure and synaptic connections of identified neurons in the rat fascia dentata, Anat. Embryol., 177 (1987) 1-14. 19 L~ibbers, K. and Frotscher, M., Differentiation of granule cells in relation to GABAergic neurons in the rat fascia dentata: combined Goigi/EM and immunocytochemical studies, Anat. Embryol., 178 (1988) 119-127. 20 Schwegler, H., Heimrich, B., Keller, E, Renner, P. and Crusio, W.E., Strain-specific development of the mossy fiber system in organotypic cultures of the mouse hippocampus, Neurosci. Lett., 87 (1988) 7-10. 21 Seress, L. and Pokorny, J., Structure of the granular layer of the rat dentate gyms. A light-microscopic and Golgi study, J. Anat., 133 (1981) 181-195. 22 Seress. L. and Frotscher, M., Morphological variability is a characteristic feature of granule cells in the primate fascia dentata: a combined Golgi/electron microscope study, J. Comp. Neurol., 293 (1990) 253-267. 23 Streit, P., Thompson, S.M. and G~ihwiler, B.H., Anatomical and physiological properties of GABAergic neurotransmission in organotypic slice cultures of rat hippocampus. Eur. J. Neurosci., 1 (1989) 603-615. 24 Wenzel, J., Stender, G. and Duwe, G., Zur Entwicklung der Neuronenstruktur der Fascia dentata bei der Ratte. Neurohistologisch-morphometrische, ultrastrukturelle und experimentelle Untersuchungen, J. Hirnforsch., 22 (198t) 629-683. 25 Zimmer, J. and G~ihwiler, B.H., Cellular and connective organization of slice cultures of the rat hippocampus and fascia dentata, J. Comp. Neurol., 228 (1984) 432-446.
