Peter L.Jeffrey,* Vladimir J. Balcar,* Ornella Tolhurst,* Ron P. Weinberger,* and Jenny A. Meany*
*Developmental Neurobiology Group Children's Medical Research Institute Westmead, New South Wales 2145 Australia
^Institute for Biomedical Research University of Sydney Sydney, New South Wales 2600 Australia
^Oncology Research Unit Children's Hospital at Westmead Westmead, New South Wales 2145 Australia
I. Introduction II. Methods and Systems
A. Cerebellar Cell Culture
B. Coculture Systems
D. Morphological Analysis
E. Glutamate Uptake in Purkinje Neurons III. Applications
A. Astrocyte Effects on Purkinje Neurons
B. Granule Cell Effects on Purkinje Neurons
C. Glutamate Transport in Developing Purkinje Neurons References
METHODS IN CELL BIOLOGY, VOL. 71 Copyright 2003, Elsevier Science (USA). All rights reserved. 0091-679X/03 $35.00
An in vitro coculture system is described to study the avian Purkinje neuron and the interactions occurring with astrocytes and granule cells during development in the cerebellum. Astrocytes initially and granule cells later regulate Purkinje neuron morphology. The coculture system presented here provides an excellent system for investigating the morphological, immunocytochemical, and electrophysiological differentiation of Purkinje neurons under controlled conditions and for studying cell-cell interactions and extrinsic factors, e.g., glutamate in normal and neuropathological conditions.
The myriad neuronal interactions, both cellular and synaptic, that occur in a particular central nervous system (CNS) region, both during development and in the mature system, can be better understood by first distinguishing the components of the area. This is the strength of being able to study individual elements of a particular neuronal region in vitro and also to study combinations of these elements.
The cerebellum is a layered and foliated structure, consisting of the molecular layer, the Purkinje neuron monolayer, and the internal granule cell layer (Ramon y Cajal, 1995). Because it is easy to recognize and dissect out, it lends itself to in vitro study. There are two main in vitro systems used: dissociated cell culture, where neurons can be investigated individually in a two-dimensional layer after preparation from a tissue source, and slice culture (Haydar et al., 1999), where the system better reflects the in vivo situation, but limits resolution of visualizing individual neurons and glia. Experiments using dissociated culture followed by slice culture would give a more complete analysis of a particular neuronal cell type. A more recently reported method of in vitro analysis involves the FACS analysis of green fluorescent protein-labeled neurons (Tomomura et al., 2001).
Dissociated Purkinje neuron cultures from both mammalian mouse and rat cerebellum have been prepared and refined by various laboratories (Baptista et al., 1994; Brorson et al., 1991; Furuya et al., 1988; Gillard et al., 1997; Gruol and Franklin, 1987; Hockberger et al., 1994; Linden et al., 1991; Mintz and Bean, 1993; Okubo et al., 2001; Schilling et al., 1991; Tabata et al., 2000).
Preparation of the mammalian dissociated Purkinje neuron culture is an uncomplicated procedure, and the end product is a mixed neuronal culture where Purkinje neurons comprise approximately 5% of the culture (Baptista et al., 1994). Mammalian cultures that require a higher density of Purkinje neurons require more complicated procedures. This is where our unique chick Purkinje neuron in vitro system offers advantages when compared to the mammalian system. The chick cerebellum has a higher in vivo Purkinje neuron density than either mouse or rat (Ito, 1984). This also means that at the time of dissection (embryonic day 8, ED8), high-density cultures of Purkinje neurons can be prepared that do not require multiple purification steps, and there is little contamination by other cerebellar neuronal cell types (Feirabend, 1990; Feirabend et al., 1985). In addition, astrocytes of varying ages, and granule cells, can be purified and added to the initial Purkinje cell cultures to study cell-cell interactions and outcomes (Jeffrey et at., 1996).
The cerebellum offers an ideal model for neurobiological research: neuronal migration (Koster and Fraser, 2001; Yang et at., 2002), pattern formation (Armstrong and Hawkes, 2000; Karram et at., 2000) development (Blanco et at., 2002; Cingolani et at., 2002), mouse mutants (Marti et at., 2002; Fernandez-Gonzalez et at., 2002), electrophysiology (Bushell et at., 2002; Shinmei et at., 2002), glutamate uptake (Brasnjo and Otis, 2001; Reichelt and Knopfel, 2002), synaptic interactions (Hirono et at., 2001; Lossi et at., 2002), and regeneration of axotomized Purkinje neuron axons (Dusart et at., 1997).
Various diseases are also associated with Purkinje neuron pathology. Ataxia telangiectasia (AT) is an autosomal recessive disorder where progressive ataxia is accompanied by Purkinje neuron degeneration (Paula-Barbosa et at., 1983). In-activation of the ATM gene causes a large decrease in numbers of Purkinje neurons and granule neurons. Without the ATM protein, oxidative stress is increased markedly in Purkinje neurons, elevating superoxide levels (Quick and Dugan, 2001). Spinocerebellar ataxia types 1 and 6 (SCA1 and SCA6) are part of a family of disorders induced by polyglutamine repeats (Gomez et at., 1997; Cum-mings et at., 2001). SCA1 has been studied using a transgenic mouse model (Burright et at., 1995). SCA6 is an autosomal disorder that encodes the P/Q type Ca2+ channel in the CNS, particularly in Purkinje neurons and granule neurons in the cerebellum (Piedras-Renteria et at., 2001). Synaptic alterations in the cerebellum have also been associated with Alzheimer's disease (Baloyannis et at., 2000). Besides the well-known deleterious effects of Creuzfeldt-Jakob disease, Hsp72 accumulates in Purkinje neurons, possibly conferring a protective mechanism on these neurons (Kovacs et at., 2001). All these pathologies present obvious studies with an in vitro Purkinje neuron system.
We have used our chick in vitro Purkinje neuron system to study the distribution of glutamate transporters, and also glutamate uptake in developing Purkinje neurons, and found that GLT-1, the glutamate transporter normally associated with astrocytes, has a neuronal function at early developmental stages in the chick cerebellum. We also showed that inhibition of glutamate transport reduced neur-ite outgrowth greatly in immature chick Purkinje neurons, inferring a relationship between the transport of L-glutamate and the morphological development of Purkinje neurons (Meaney et at., 1998).
The link between neuronal sensitivity to excess glutamate and the function of glutamate transporters has implications for neuronal and glial homeostasis at the individual cell level, and also at the level of the whole tissue. Several studies have demonstrated that L-glutamate transport (GluT) is of crucial importance both for the maintenance of low extracellular levels of neurotransmitter L-glutamate and for the normal functioning of neurotransmitter metabolism (for reviews, see Rae et at., 2000; Danbolt, 2001; Balcar, 2002). Inhibition of GluT has been shown to be neurotoxic both in culture and in vivo (Danbolt, 2001; Balcar, 2002), and circumstantial evidence shows that deficient or altered GluT is associated with neurological disease. Thus Scott et al. (1995) reported subtle changes in the characteristics of GluT in the brains of patients dying from Alzheimer disease, and Rothstein et al. (1995) noticed severe deficits in one of the transporters mediating GluT (GLT) in spinal cords of patients who had died of amyotrophic lateral sclerosis. More recent findings show that glutamate uptake is reduced in astrocytes derived from human AD patients, but that estrogen treatment can increase glutamate uptake (Liang et al., 2002). Experimental autoimmune encephalomyelitis (EAE) results in activation of the AMPA glutamate receptor (Ohgoh et al., 2002). The connection between the AMPA receptor and the expression of glutamate transporters in the spinal cord of the Lewis EAE model has shown that EAAC1 protein and mRNA is upregulated dramatically in the EAE model, whereas other transporters, GLT-1 and GLAST, were both downregulated in the spinal cord. In a rat portacaval anastomosis model, analysis of expression of glutamate transporters in the cerebellum showed that transient alterations in the transporters GLAST and GLT-1 caused an accumulation of excess glutamate in the extracellular space (Suarez et al., 2000). Maric et al. (2002) also identified a cytoplasmic LIM protein, Ajuba, that interacts with the aminoterminus of the GLT protein, proposing it as the scaffolding protein that allows interactions among it, GLT, and the cytoskeleton. Other transporter-associated proteins have been identified (for a review, see Balcar et al., 2001), but at present there is no evidence that such proteins have any affect on the activity of GluT.
II. Methods and Systems
A. Cerebellar Cell Culture
Trypsin (type XII-S) DNase I
Dulbeccos Modified Eagles Medium (DMEM)
Fetal calf serum (FCS)
Chick embryo extract (CEE)
Cytosine ^-D-arabinoside (AraC)
Phosphate-buffered saline (PBS), calcium and magnesium free
Millicell (30mm, 0.4-^m pore size) culture plate inserts
Cerebellar Purkinje neuron cultures were prepared from chick embryos (ED8). This time was chosen as the inner cortical cell layer, the source of Purkinje neurons appears at this time and the layer is subdivided into Purkinje neuron clusters between days 8 and 11 (Feirabend, 1990).
1. Dissect out ED8 cerebellum in aseptic conditions using sterile instruments in a biohazard hood.
2. Incubate in 0.25% (w/v) trypsin in PBS for 30 min at 37° C.
3. Triturate in 0.15% DNase I (including 0.25% MgSO4).
4. Coat 25-mm glass coverslips with 1 mg/ml poly-L-lysine and 100 yg/ml lami-nin dissolved in sterile distilled water and dried in a tissue culture hood.
5. Plate the coverslips at a density of 2000 cells/mm2 with cells in DMEM containing 5% FCS, 2.5% CEE, and 25 mm KCl (neuronal medium) in 7.5% CO2 at 37°C.
6. After 2 days in culture, add the mitotic inhibitor, AraC, at 5 x 10~6M for 3 days to inhibit astrocyte proliferation on the coverslips.
Fifteen ED8 embryos yield enough cells for 18 coverslips at a density of 2000 cells/mm2.
1. Dissect cerebellum from ED16 chickens under aseptic conditions.
2. Cut into small pieces.
3. Produce a single cell suspension by dissociation with 0.25% trypsin in PBS and triturate with 0.15% DNase I containing 0.25% (w/v) MgSO4 in PBS.
4. Culture for 4 days in six-well, 35-mm Corning plates in DMEM plus 15% FCS in 7.5% CO2 at 37° C to produce an astrocyte monolayer.
5. Coat Millicells (30 mm, 0.4-ym pore size) culture plate inserts with 1/10 Matrigel and allow to air dry.
6. Prepare a single cell suspension from astrocyte monolayers by subculturing in 0.25% trypsin and 0.15% DNase I containing 0.25% (w/v) MgSO4.
7. Coat astrocytes onto Millicells at a density of 2800 cells/mm2 and culture in neuronal medium overnight.
Three embryos will yield enough cells for 18 Millicells. Astrocytes were characterized using antibody to glial fibrillary acidic protein (GFAP). Astrocytes can be prepared from any age using this procedure.
4. Granule Cells
Purified granule cells were isolated utilizing the procedure of Hatten (1985)
developed for mice.
1. Dissect cerebellum from chicks of ED13 or ED15, chop into small pieces, and incubate in 0.25% trypsin, and triturate in 0.15% DNase I, including 0.25% (w/v) MgSO4 for 20min at 37° C.
2. Prepare 30% Percoll (1.5ml Percoll plus 3.5ml DMEM) and 60% Percoll (3 ml Percoll plus 2 ml DMEM) gradient in a centrifuge tube.
3. Layer the single cell suspension onto the Percoll gradient and centrifuge at 3000 rpm (2000 x gmax) for 30min in a Beckman GS-6 centrifuge.
4. Collect cells from the 30/60% Percoll interface and wash twice with PBS at a 5/1 volume ratio to cell pellet by centrifugation at 2200 rpm for 5 min (1000 x gmax).
5. Plate granule cells onto tissue culture plastic dishes coated with 100 yg/ml poly-D-lysine dissolved in sterile distilled water for 20 min. Collect nonad-herent granule cells by centrifugation.
6. Store in neuronal medium at high density (>2 x 106 cells/ml) until the addition to coverslips.
Purified avian granule cells stain positive for the amino acid neurotransmitter glutamate and negative for parvalbumin and calbindin D-28. The cell size (6 ym) and immunohistochemical profile define these cells as granule cells (Gruol and Crimi, 1988).
B. Coculture Systems
A Purkinje neuron coverslip was placed into each well of a six-well 35-mm culture dish, followed by an astrocyte-coated Millicell covered with neuronal medium and cultured for various times.
Granule cells at a ratio of 5/1 to Purkinje neurons were added onto Purkinje neuron coverslips, which had been incubated for either 5 or 7 days in vitro culture (DIV). Isochronic cultures were prepared by adding granule cells from either ED13 or ED15 cerebellum so that granule cells added were the same age as Purkinje neurons on the coverslip, e.g., ED8 plus 5 or 7 DIV.
Was this article helpful?